Material Information

Inspiration come to the headwaters : proceedings of the fifteenth annual conference of the Association of State Floodplain Managers, June 10-14, 1991, Denver, Colorado
Series Title:
Special publication ;
Portion of title:
Proceedings of the fifteenth annual conference of the Association of State Floodplain Managers, June 10-14, 1991, Denver, Colorado
Association of State Floodplain Managers -- Conference, 1991
University of Colorado, Boulder -- Natural Hazards Research and Applications Information Center
Place of Publication:
Boulder, Colo
Madison, Wis
Natural Hazards Research and Applications Information Center
Association of State Floodplain Managers
Publication Date:
Physical Description:
1 online resource (xxiii, 370 p.) : ill. ;


Subjects / Keywords:
Floodplain management -- Congresses -- United States ( lcsh )
Flood control -- Congresses -- United States ( lcsh )
government publication (state, provincial, terriorial, dependent) ( marcgt )
bibliography ( marcgt )
conference publication ( marcgt )
non-fiction ( marcgt )


Includes bibliographic references.
General Note:
Description based on print version record.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
002025190 ( ALEPH )
432312730 ( OCLC )
F57-00091 ( USFLDC DOI )
f57.91 ( USFLDC Handle )

USFLDC Membership

Added automatically
Natural Hazards Center Collection

Postcard Information



This item has the following downloads:

Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200337Ia 4500
controlfield tag 001 002025190
005 20090928143531.0
006 m d
007 cr bn|||||||||
008 090821s1991 coua sb s100 0 eng d
datafield ind1 8 ind2 024
subfield code a F57-00091
2 110
Association of State Floodplain Managers.
b Conference
n (15th :
d 1991 :
Denver, Colo.)
1 0 245
h [electronic resource] :
come to the headwaters : proceedings of the fifteenth annual conference of the Association of State Floodplain Managers, June 10-14, 1991, Denver, Colorado.
3 246
Proceedings of the fifteenth annual conference of the Association of State Floodplain Managers, June 10-14, 1991, Denver, Colorado
Boulder, Colo. :
Natural Hazards Research and Applications Information Center ;
Madison, Wis. :
Association of State Floodplain Managers,
1 online resource (xxiii, 370 p.) :
Special publication ;
v 24
Description based on print version record.
Includes bibliographic references.
Floodplain management
z United States
Flood control
United States
University of Colorado, Boulder.
Natural Hazards Research and Applications Information Center.
i Print version:
Association of State Floodplain Managers. Conference (15th : 1991 : Denver, Colo.).
t Inspiration.
Boulder, Colo. : Natural Hazards Research and Applications Information Center ; Madison, Wis. : Association of State Floodplain Managers, [1991?]
w (OCoLC)25530686
Special publication (University of Colorado, Boulder. Natural Hazards Research and Applications Information Center) ;
Natural Hazards Center Collection
4 856


Inspiration: Come to the Headwaters Proceedings of the Fifteenth Annual Conference of the Association of State Floodplain Managers June 10-14, 1991 Denver, Colorado


The opinions contained in this volume are those of the authors and do not necessarily represent those of the funding or sponsoring organizations or those of the Association of State Floodplain Managers. Cover photo courtesy of Jeff Andrew, Colorado Tourism Board. This volume is available from: The Natural Hazards Research and Applications Information Center Institute of Behavioral Science Campus Box 482 University of Colorado Boulder, CO 80309-0482 and The Association of State Floodplain Managers P.O. Box 2051 Madison, WI 53701


Preface "Inspiration: Come to the Headwaters" was the theme for the Fifteenth Annual Conference of the Association of State Floodplain Managers. Coming to Denver truly was inspirational for those of us fortunate enough to make the trip. Added to the inspirational value of the technical sessions and interactions with our peers during the conference was the sheer beauty of Colorado. This year's conference was hosted by the Colorado Water Conservation Board, the Urban Drainage and Flood Control District, the University of Colorado at Colorado Springs, and the Colorado Association of Stormwater and Floodplain Managers. Our hosts ensured that the nearly 400 conference attendees not only benefitted from excellent technical sessions but enjoyed themselves in the process. Broad-ranging issues including flood hazard mitigation, hydrologic and hydraulic modeling, local flood warning, stormwater, arid regions, and multi-objective management provided something for everyone. Each year our conferences touch on additional issues that relate to floodplain management. As we continue to widen our sphere of influence, floodplain management becomes a more challenging field. We are continually learning as the issues increase, yet we need to stay focused on our mission of reducing flood losses and human suffering and to protect the natural and beneficial values of floodplains. We hope that these proceedings will help you to learn and retain information that will contribute to the accomplishment of our mission. iii Jerry Louthain, Chair Association of State Floodplain Managers


This page IS intentionally blank


Acknowledgements It takes the combined efforts of many people to make a successful conference. The conference team would like to extend our sincere appreciation to those individuals and organizations throughout the country who gave unselfishly of their time and talent to make this annual event so inspiring. First, we thank the Colorado Water Conservation Board (CWCB), the Urban Drainage and Flood Control District (UDFCD) the University of Colorado at Colorado Springs (UCCS) and the Colorado Association of Stormwater and Floodplain Managers (CASFM) for co-hosting the conference. They set the stage. Second, we thank all the exhibitors, moderators, panelists, authors, and speakers who brought the technical content to the conference. It is one thing to recognize an idea is worthy of sharing with others; it is another to take the time to carefully write it down and present it in front of one's colleagues. They are the scriptwriters and actors. Third, we thank the crew. Students from UCCS; staff members from the CWCB and the UDFCD; volunteers from the ASFPM, the CASFM, and consulting engineers; and employees from our conference contractors and the hotel played a big part behind the scenes. Corporate sponsors this year included Eveready Flood Control, Michael Baker, Jr., Inc., Dewberry & Davis, French and Associates, Les Bond and Associates, CH2M Hill, and Boyle Engineering Corporation. The inclusion of several informal events helps set the ASFPM conferences apart from the others. These proceedings are made possible by the generous support of several subscribers, including the U.S. Army Corps of Engineers, the Tennessee Valley Authority, the U.S. Environmental Protection Agency, the Federal Fmergency Management Agency, the USDA/Soil Conservation Service, NOAA/National Weather Service, the U.S. Geological Survey; and the states of Arizona, California, and Hawaii. Special thanks to the Natural Hazards Research and Applications Information Center for editing and publishing this document, which we feel reflects the range and depth of the program and will serve as an excellent reference for anyone in the wide variety of disciplines around the world concerned with floodplain management. Finally, we thank the audience. Of the 389 attendees this year, there was not one who did not also participate in some way. For all of you who came to the headwaters, we hope these proceedings and the memories of a quality performance, with Colorado's backdrop of majestic mountains, will be a source of inspiration for the future. William P. Stanton, Conference Director Eve Gruntfest, Program Chair Dan Accurti, Exhibits Chair Bill DeGroot, Field Trip Coordinator v


This page IS intentionally blank


Table of Contents List of Speakers and Panelists ......................... xv Part One: Plenary Addresses Mission Impossible(?): Working Together in Floodplains and River Corridors OJristopher N. Brown . . . . . . . . . . . . . . . . . 3 EPA and the National Environmental Policy Act: Facilitators of Creative Floodplain and Natural Resources Management in the 90s Shannon E. Cunniff and Richard E. Sanderson .................. 9 Stone Soup and Hazard Mitigation Edward A. Thomas . . . . . . . . . . . . . . . . .. 14 Perspectives: A Status Report on the Nation's Floodplain Management Activity Frank Thomas . . . . . . . . . . . . . . . . . . .. 19 A Review of the Status Report on the Nation's Floodplain Management Activity, April and September 1989 Gilbert F. White .................................... 21 Part Two: Flood Hazard Mitigation Applying Floodplain Regulations to the Real World Cynthia A. Baumann and G. Nicholas Textor .................. 27 Flood Hazard Regulations in King County, Washington's Sensitive Areas Ordinance Thomas C. Bean .................................... 30 Update 1991: Emergency Management Institute Natural Hazards Mitigation Training Courses Dan Bondrolf . . . . . . . . . . . . . . . . . . . 34 vii


An Assessment Methodology and Solutions to Barriers and Constraints to Floodplain Management-The Tennessee Example Contents George E. Bowen . . . . . . . . . . . . . . . 38 Mullet and Bishop Creeks Drainage Basin Improvements Robert E. Johnson and Daniel Shabledeen . . . . . . . . 42 Multi-Level Government Cooperation on a Small Scale Floodway Acquisition Project: Stanley Township, Cass County, North Dakota Jeff Klein . . . . . . . . . . . . . . . . . . . . 45 South Platte River Central Platte Valley Improvements Brian S. Kolstad, John M. Pflaum, and Richard E. McLaughlin ...... 48 Basalt River Stabilization Plan Tom Newland ...................................... 53 Speaking Plain English H. James Owen . . . . . . . . . . . . . . . . . . 57 Flood Hazard Mitigation In Englewood, Colorado: The Little Dry Creek Success Story John M. Pflaum . . . . . . . . . . . . . . . . . . 62 Implementing a National Flood Mitigation Grant Program under Pending Legislation Michael Robinson, John Gambel, and A. Todd Davison ............ 67 What Happened to the Eccleston Brook Watershed? Tales of a Floodplain Management Study Elizabeth A. Rogers .................................. 69 If There's Water in My Living Room, I Must Be in a Floodplain Ronald L. Rossmiller ................................. 72 Developing a State Floodplain Management Plan: Content and Process James M. Wright ............................ 76 viii


Contents Part Three: The National Flood Insurance Program Arapahoe County, Colorado Flood Risk Directory Project Methods and Applications Douglas Gore . . . . . . . . .. 83 How to Get Flood Insurance Study Data and What to Do When It's Vanished Maggie Mathis and Lisa Bourget . . . . . . . . . . . . . 87 Flood Risk Directory Production Myles E. Powers and Vince DiCamillo ...................... 91 Acquiring Data from the FIA Archives-A Step-by-Step Approach Richard A. Wild .................................... 95 Part Four: Hydrology and Hydraulics More On "Backwater" Surface Profile George R. Alger and Henry S. Santeford A Comparison of a Gaged Urban Watershed and Computer Modeling Using HYMO 103 aifford E. Anderson and Richard 1. Heggen ................. 107 Two Dimensional Modeling in an NFIP Community: Experiences and Lessons Learned Lawrence P. Basich and Karl L. Krcma . . . . . . . . . .. 113 Modeling Floods in Urban Areas G. Braschi, M. Gallati, and L. Natale. . . . . . . . . . .. 117 Combined Geomorphic, Hydraulic, and Sediment Transport Analyses: Application to a Sedimentation Problem Karin 1. Fischer, Michael D. Harvey, and Lyle W. Zevenbergen ..... 123 Geomorphic Response of Lower Feather River to 19th Century Hydraulic Mining Operations Karin 1. Fischer and Michael D. Harvey . . . . . . . . . .. 128 ix


Contents Two Dimensional Modeling for Flood Hazard Delineation George V. Sabol and Kenneth A. Stevens . . . . . . . . . .. 133 Extrapolation of Regional Flood Frequency Relations Based on Flow Variability Kenneth L. Wahl . . . . . . . . . . . . . . . . .. 138 Geomorphic and Sediment Transport Analyses for the Napa River Flood Control Project C. Gary Wolff and Mark R. Peterson ...................... 143 Two-Dimensional Hydrodynamic Model of the Colusa Flood Overflow Weir on the Sacramento River Lyle W. Zevenbergen, Mark R. Peterson, and Michael D. Harvey .... 148 Part Five: Local Flood Warning Can We Have Too Much Warning Time? A Study of Rockhampton, Australia John Handmer .................................... 155 Flood Warning Services Market Survey Study for Maricopa County, Arizona Laurie T. Miller, Fred K. Duren, Jr., and J.M. RU1rUmn Using Appropriate Flood Warning Technology for Communities at Risk 160 Mark E. Nelson . . . . . . . . . . . . . . . . . .. 164 Managing Flood Warning Systems: The United Kingdom Experience Dennis J. Parker and Sylvia Tunstall ...................... 168 Local Flood Warning Systems in New Jersey Robert D. Schopp and Rebecca J. Burns .................... 172 Community Alert Network: Dam Safety and Flood Warning Steven E. Smead ................................ .. 176 Summertime Precipitation in the Higher Elevations of Colorado Larry Tunnell ..................................... 179 x


Contents Part Six: Stormwaler Management Innovative Stormwater Management Techniques in Northeastern lllinois-A Case Study Gerald J. Kauffman and Anwer Ahmed .................... 185 Navigating Political Waters: Successful Implementation of Urban Stormwater Management Programs Sharon L. Oakes ................................... 191 Stormwater Quality Monitoring As a Planning Tool-A Case Study Cynthia L. Paulson and Dottie Nazarenus ................... 194 Successful Municipal Stormwater Management: Key Elements Andrew J. Reese ................................... 202 NPDES Stormwater Permitting-Is There Something Missing? William P. Ruzzo . . . . . . . . . . . . . . . . .. 206 A Countywide Stormwater and Floodplain Ordinance Jonathon P. Steffen and Joseph E. Stuber ................... 211 From Liability to Resource: The City of Aurora, Colorado's Changing Approach to Drainageway Design William E. Wenk ................................... 215 Incorporating Stormwater Quality Enhancement Features Into Urban Flood Control Projects James T. Wulliman . . . . . . . . . . . . . . . . .. 221 Part Seven: Arid Region Issues Fluvial Geomorphology Principles Applied to a Stream Regulating Program Joseph V. Borgione and Chad Gourley ..................... 227 Correlation Between Flow, Slope, Roughness Coefficient, and Regime Width for Upper Incised Channels on Alluvial Apron of Summerlin Area, Las Vegas, Nevada Donald W. Davis .................................. 232 xi


Contents Estimating Sediment Delivery And Yield On Alluvial Fans Michael D. Harvey and Robert C. MacArthur ................. 238 Considering Storage Effects On Alluvial Fan Flooding Stephen G. King and Edward R. Mifflin . . . . . . . . . .. 243 Optimization Modeling for Flood Control on Distributary Flow Areas in the Southwest James R. Morris ................................... 247 Continuing Research on Two-Dimensional Modeling of Flood Hazards on Alluvial Fans J.S. O'Brien and W.T. Fullerton ......................... 251 FEMA Method for Predicting Flood Hydraulic Boundaries on Alluvial Fans Requires Verification J.S. O'Brien and W.T. Fullerton ......................... 256 Floodproofing Development on Alluvial Fans Vassilios A. Tsihrin!Vs, Blake N. Murillo, Michael E. Mulvihill, and William J. Trott. . . . . . . . . .. 261 Part Eight: Innovative Software Applications Visualization Techniques for Flooding Models G. Braschi, S. Braschi, and L. Natale. . . . . . . . . . .. 267 Floodplain Management With MacProject II Albert H. Halff and Henry M. Halff . . . . . . . . . . .. 273 Practical Applications of the HEC Flood Damage Analysis Package Troy Lynn Lovell, Walter E. Skipwith, and Michael A. Moya ....... 277 The Alternative Futures Assessment Process: Building a Consensus for Coastal Management in Texas Andrew Mangan, Thomas Bonnicksen, and Jim LeGrotte Computer Aided Evaluation of Floodplain Development Risk 283 Mark R. Peterson, C. Gary Wolff, and Jay D. Schug ............ 288 xii


Contents Part N"me: Multi-Objective PIanning Reduced Flood Losses in Oregon and Oregon's Statewide Planning Program George M. Currin .................................. 295 Comprehensive Floodplain Management: The Denver Area Experience Bill DeGroot . . . . . . . . . . . . . . . . . . .. 299 Management of Natural and Beneficial Floodplain Values: The Commonwealth of Virginia's Strategy Douglas J. Plasenda ................................ 303 The Missouri River Corridor Project, or the Beginnings of a River Renaissance John H. Sowl ..................................... 306 Santa Rosa Creek Restoration Project Linda Stonier ..................................... 312 South Platte River Improvements: Floodplain Management Achieves Multiple Objectives Steven R. Williams, Ben Urbonas, and R. Jay Nelson ............ 316 Part Ten: Geographic Information Systems and Flood Hazard Mapping Guidelines and Specifications for Erosion Studies Mark Crowell and Michael K. Buckley ..................... 321 Multi-Objective River Corridor Planning Using Geographic Information System Methods: A Case Study of the Cache La Poudre River Corridor Duane A. Holmes. . . . . . . . . . . . . . . . . .. 324 NFIP Map Revision Process Made Simple Mary Anne Lyle, Vince DiCamillo, and Mary Jean Pajak. . . . .. 328 xiii


Flood Mapping Studies in Northern Italy Luigi Natale, Fabrizio Savi, Contents Paolo Bonaldi, and Pier Giorgio Mandola ................... 332 Channel Migration On The Tolt And Raging Rivers, King County, Washington Susan J. Perkins ................................... 338 Automated Digital Line Graph Data Capture Procedures John M. Taylor .................................... 342 Part Eleven: New Techniques in Ice Jam Control Israel River Ice Control Structure Kathleen D. Axelson . . . . . . . . . . . . . . . .. 349 Ice Jam Flood Frequency Analysis Techniques Jon E. Zufelt and James L. Wuebben ...................... 353 Part Twelve: Lessons from Recent Floods Lessons from the 1990 Taiwan Typhoon Season Spenser W. Havlick ................................. 359 Snohomish County, Washington, Thanksgiving Day Flood, 1990 Sky Miller ....................................... 364 The Shadyside, Ohio, Flash Floods: June 14, 1990 William L. Read ................................... 368 xiv


List of Speakers and Panelists Christopher R. Adams Lawrence P. Basich Daniel L. Bondroff Colorado State University FEMAlNatural and FEMAlEmergency Cooperative Institute for Technological Hazards Management Institute Research in the Division 16825 South Seton Avenue Atmosphere 130 228th Street, SW Emmitsburg, MD 21727 7003 East Warren Drive Bothell, WA 98021 (301) 447-1278 Denver, CO 80224 (206) 487-4703 (303) 756-4320 Joseph V. Borgione Cynthia A. Baumann Utah Division of Water Anwer R. Ahmed Envirodyne Engineers, Rights, #220 Donohue & Associates, Inc. 1636 West North Temple Inc., Suite 200 E 168 North Clinton Street Salt Lake City, UT 84116 1501 Woodfield Road Chicago, IL 60606 (801) 538-7377 Schaumberg, IL 60173 (312) 648-1700 (708) 605-8800 Lisa Bourget Thomas C. Bean Dewberry and Davis George R. Alger King County Surface METS Division, 3rd Floor Civil Engineering Water Management 8401 Arlington Boulevard Department Division Fairfax, V A 22031 Michigan Tech University 730 Dexter Horton (703) 849-0476 Houghton, MI 49922 Building (906) 487-2568 710 Second Street George E. Bowen Seattle, WA 98104 School of Planning Clifford E. Anderson (206) 296-6519 #11 Hensen Hall Albuquerque Metro University of Tennessee Arroyo Martin Becker Knoxville, TN 37909 Flood Control Authority 425 Chandler Building (615) 588-6897 2600 Prospect, NE 127 Peachtree Street, N.E. Albuquerque, NM 87107 Atlanta, GA 30303 Terry A. Bowen (505) 884-2215 (404) 523-3900 Boyle Engineering Corporation, Suite 200 Kathleen D. Axelson Leslie A. Bond 165 South Union U.S. Army Corps of Leslie A. Bond Associates Boulevard Engineers Post Office Box 397 Lakewood, CO 80228 72 Lyme Road Arivaca, AZ 85601 (303) 987-3443 Hanover, NH 03255 (602) 398-9286 (603) 646-4187 xv


Speakers and Panelists Margaret F. Bowker Dan Bunting George M. Currin Nimbus Engineers EI Paso County Building FEMAlRegion X 3710 Grant Drive, #D Department Federal Regional Center Reno, NV 89509 101 West Costilla 130 228th Street, S.W. (702) 689-8630 Colorado Springs, Bothell, WA 98021 CO 80903 (206) 487-4679 A. Jean Brown (719) 578-6230 California Department of Tim P. D'Acci Water Resources Curt M. Chandler Washington State DepartFPM Branch City of Henderson ment of Ecology 1416 9th Street, Room Public Works Department Shorelands, MS PV 11 723 240 Water Street Olympia, WA 98504 Sacramento, CA 95814 Henderson, NV 89015 (206) 459-0796 (916) 445-6249 (702) 565-2329 Steve Daroff Christopher N. Brown Dave G. Clark Transamerica Flood Rivers and Trails King County Surface Hazard Cert. Conservation Water Management Post Office Box 607 National Park Service Division Elmwood Park, NJ 07407 Post Office Box 37127 710 Second Avenue (800) 247-3384 Washington, DC 20013 Seattle, WA 98104 (202) 343-3775 (206) 296-6519 Donald W. Davis Boyle Engineering William M. Brown, III Carl L. Cook, Jr. Corporation U.S. Geological Survey FEMAlRegion X 1785 East Sahara, #300 BGRAINLIC, Box 25046 130 228th Street, SW Las Vegas, NV 89104 MS-966 Bothell, WA 98021 (702) 731-5511 Denver Federal Center (206) 487-4687 Denver, CO 80225 Todd Davison (303) 236-0616 Robert R. Cox FEMAlFederal Insurance Louisiana Department of Administration Michael K. Buckley Transportation and Office of Loss Reduction FEMAlFIAlRisk Development 500 C Street, S.W. Assessment Floodplain Management Washington, DC 20472 500 C Street, S.W. Section, Room 430 (202) 646-3448 Washington, DC 20472 Post Office Box 94245 (202) 646-2756 Baton Rouge, LA 70804 William G. DeGroot (504) 379-1408 Urban Drainage and Flood Control District, #156B 2480 West 26th Avenue Denver, CO 80211 (303) 455-6277 xvi


Speakers and Panelists Vincent DiCamillo Mark W. Glidden John Handmer Greenhome and O'Mara, Boyle Engineering Centre for Resources and Inc., #700 Corporation Environmental Studies 7500 Greenway Center 165 South Union Australia National Drive Boulevard University Greenbelt, MD 20770 Lakewood, CO 80228 GPO Box 4 (301) 220-1873 (303) 987-3443 Canberra ACT 2601 Australia Hugh Duffy Douglas A. Gore 062494277 National Park Service FEMNRegion VIII River, Trail and Natural and Technological Michael D. Harvey Conservation Hazards Division Water Engineering and 12795 West Alameda Denver Federal Center Technology, Inc. Parkway Building 710, Box 25267 419 Canyon Lakewood, CO 80228 Denver, CO 80225 Fort Collins, CO 80521 (303) 969-2850 (303) 235-4830 (303) 482-8201 Thomas L. Fischer David J. Greenwood Spenser W. Havlick Stanley Township Board Michael Baker, Jr., Inc. College of Environmental RR 1, Box 85 1420 King Street Design Fargo, ND 58104 Alexandria, V A 22314 Campus Box 314 (701) 235-8396 (703) 838-0400 University of Colorado Boulder, CO 80309-0314 Ronald D. Flanagan Eve Gruntfest (303) 492-6936 R.D. Flanagan and Department of Geography Associates, Suite 70 University of Colorado Robert C. Henchbarger 201 West 5th Street 1420 Austin Bluffs Michael Baker, Ir., Inc. Tulsa, OK 74103 Parkway 1420 King Street (918) 587-7166 Colorado Springs, Alexandria, VA 22314 CO 80933 (703) 838-0400 William Fullerton (719) 593-3513 FLO Engineering, Inc. Debra J. Hendrickson Post Office Box 1659 Albert H. Halff King County Surface Breckenridge, CO 80424 Albert H. Halff Water Management (303) 453-6394 Associates, Inc. Division 8616 Northwest Plaza River Planning Program John Gambel Drive 730 Dexter-Horton FEMNFINOffice of Loss Dallas, TX 75225 Building Reduction (214) 739-0094 Seattle, WA 98104 500 C Street, S.W. (206) 296-6519 Washington, DC 20472 (202) 646-2724 xvii


Speakers and Panelists Duane A. Holmes Robert E. Johnson Larry A. Larson National Park Service Greiner, Inc. Wisconsin Department of River and Trail Post Office Box 31646 Natural Resources Conservation Tampa, FL 33631 Bureau of Water 12795 West Alameda (813) 286-1711 Regulation, W216 Pkwy Post Office Box 7921 Lakewood, CO 80228 Karen C. Kabbes Madison, WI 53703 (303) 969-2850 Illinois Department of (608) 266-1926 Transportation George R. Hosek Division of Water Shirley Laska Michigan Department of Resources Environmental Institute Natural Resources 201 West Center Court Department of Sociology Land and Water Schaumburg, IL 60196 University of New Orleans Management Division (708) 705-4341 New Orleans, LA 70148 Post Office Box 30028 (504) 286-6472 Lansing, MI 48909 Gerald J. Kauffman (517) 335-3182 Donohue and Associates, Jim LeGrotte Inc., Suite 200E FEMAlRegion VI Rebecca Quinn Hughes 1501 Woodfield Road 800 North Loop 288 State of Maryland Schaumberg, IL 60173 Denton, TX 76201-3698 Flood Management (708) 605-8800 (817) 898-9162 Division D-3 Tawes State Office Bldg Jeff Klein John Liou Annapolis, MD 21401 North Dakota State Water FEMAlRegion VIII (301) 974-3825 Commission Denver Federal Center, 900 East Boulevard Ave Building 710 Brian R. Hyde Bismarck, ND 58505 Denver, CO 80225 Colorado Water (701) 224-2752 (303) 235-4830 Conservation Board 1313 Sherman Street, Brian S. Kolstad David W. Lloyd #721 McLaughlin Water Urban Drainage and Flood Denver, CO 80203 Engineers, Ltd. Control, #156B (303) 866-3441 2420 Alcott Street 2480 West 26th Avenue Denver, CO 80211 Denver, CO 80211 Alan A. Johnson (303) 458-5550 (303) 455-6277 FEMAlFIAlORAlRSD 500 C Street, S.W. Jon Kusler David J. Love Washington, DC 20472 Association of State Love and Associates, Inc. (202) 646-3403 Wetland Managers 2995 Centergreen Court Post Office Box 2463 South, Suite C Berne, NY 12023 Boulder, CO 80301 (518) 872-1804 (303) 440-3439 xviii


Speakers and Panelists Nancy B. Love Love and Associates, Inc. 2995 Centergreen Court South, Suite C Boulder, CO 80301 (303) 440-3439 Troy L. Lovell Albert H. Halff Associates, Inc. 4000 Fossil Creek Blvd Fort Worth, TX 76137 (817) 847-1422 Andrew I. Mangan Texas General Land Office 1700 North Congress Austin, TX 78701 (512) 463-5193 Margaret Mathis Dewberry and Davis METS Division, 3rd Floor 8401 Arlington Boulevard Fairfax, VA 22031 (703) 849-0330 Mark N. Mauriello New Jersey Division of Coastal Resources 501 East State Street Trenton, NJ 08625 (609) 292-8262 Edwin L. May NOAA/OARfERL/FSL U.S. Department of Commerce 325 Broadway Boulder, CO 80303 (303) 938-2088 Patricia McDermott Dewberry and Davis METS Division, 3rd floor 8401 Arlington Boulevard Fairfax, VA 22031 (703) 849-0245 Edward J. McKay NOAA/National Geodetic Survey, Room 313 11400 Rockville Pike Rockville, MD 20852 (301) 443-8567 Ross McKay FEMAlFederal Insurance Administration 500 C Street, S.W. Washington, DC 20472 (202) 646-2717 Richard E. McLaughlin McLaughlin Water Engineers, Ltd. 2420 Alcott Street Denver, CO 80211 (303) 458-5550 Jeanne M. Melanson Environmental Protection Agency Wetland Division (A 1 04-F) 401 M Street, S.W. Washington, DC 20460 (202) 863-1799 Laurie T. Miller James M. Montgomery Cons. Engineer 6245 North 24th Parkway, Suite 208 Phoenix, AZ 85106 (602) 954-6781 xix Matthew B. Miller FEMAlOffice of Risk Assessment 500 C Street, S.W. Washington, DC 20015 (202) 646-3461 Sky D. Miller Snohomish County Surface Water Management Third floor Wall Street Building 2930 Wetmore Street Everett, WA 98201 (206) 388-3464 Terri Miller Arizona Department of Water Resources flood Management Section 15 South 15th Avenue Phoenix, AZ 85007 (602) 542-1541 Michele C. Monde Michael Baker, Jr., Inc. 1420 King Street Alexandria, VA 22314 (703) 838-0400 James R. Morris Arizona Department of Water Resources flood Management Section 15 South 15th Avenue Phoenix, AZ 85007 (602) 542-1541


Speakers and Panelists Virginia K. Motoyama Sharon L. Oakes Jerome Q. Peterson FEMAlRegion VIII Disaster Recovery U.S. Army Corps of Natural and Technological Resources Engineers Hazards Division 1021 N.E. 4th Street 20 Massachusetts Avenue, Building 710 Gainesville, FL 32601 N.W. Denver Federal Center (904) 375-6124 Washington, DC 20314 Denver, CO 80225 (202) 272-0169 (303) 235-4830 H. James Owen Flood Loss Reduction Mark Peterson Mary Fran Myers Associates Water Engineering and Natural Hazards Center 4145 Maybell Way Technology, Inc. Campus Box 482 Palo Alto, CA 94306 Suite 225 University of Colorado (415) 493-7198 419 Canyon Avenue Boulder, CO 80309-0482 Fort Collins, CO 80521 (303) 492-2150 Mary Jean Pajak (303) 482-8201 Greenhorne and O'Mara, Luigi Natale Inc. John M. Pflaum Dip. Ingegneria Idraulica 7500 Greenway Center McLaughlin Water Facolta di Ingegneria Drive, #700 Engineers, Ltd. via Abbiategrasso, 213 Greenbelt, MD 20770 2420 Alcott Street Pavia, Italy, (301) 220-1871 Denver, CO 80211 (303) 458-5550 Mark E. Nelson Dennis Parker U.S. Army Corps Flood Hazard Research Clancy Philipsborn of Engineers Centre The Mitigation Assistance Omaha District Middlesex Polytechnic Corporation 215 North 17th Street Queensway Enfield Post Office Box 382 Omaha, NE 68137 EN34SF Boulder, CO 80306 (402) 221-3109 United Kingdom (303) 494-4242 Tom A. Newland Cynthia Paulson William H. Phillips Pitkin County Government Brown and Caldwell House of Representatives 530 East Main, 3rd Floor 7535 East Hampden Banking Committee Aspen, CO 81611 Avenue, #403 Finance and Urban Affairs (303) 520-5200 Denver, CO 80231 139 Ford Building (303) 750-3983 Washington, DC 20515 Jimmy S. O'Brien (202) 225-1271 FLO Engineering, Inc. Philip A. Pearthree Post Office Box 1659 Arizona Geological Survey Breckenridge, CO 80424 845 North Park Avenue (303) 453-6394 Tucson, AZ 85719 (602) 882-4795 xx


Speakers and Panelists Douglas J. Plasencia Ronald L. Rossmiller Roy D. Sedwick Virginia Division of Soil HDR Engineering, Inc. Texas Floodplain and Water Conservation Suite 200, Building C Management Association Bureau of Flood 11225 SE Sixth Street Post Office Box 162632 Protection Bellevue, WA98004 Austin, TX 78716 203 Governor Street, Suite (206) 869-2282 (512) 264-1556 206 Richmond, VA 23219 Victor Rothacker W. Louis Sidell, Jr. (804) 371-6136 Arizona Floodplain Office of Comprehensive Management Associates Planning Andrew J. Reese Floodplain Management State House Station 130 ERCE Section Augusta, ME 04333 3325 Perimeter Hill Drive 201 North Stone, 4th floor (207) 289-6800 Nashville, TN 37211 Tucson, AZ 85701 (615) 331-3520 (602) 740-6350 Steven E. Smead Community Alert Network Perry E. Rhodes William P. Ruzzo 301 Nott Street Dewberry and Davis Brown and Caldwell Schenectady, NY 12305 METS Division, 3rd Floor 7535 East Hampden (518) 382-8007 8401 Arlington Boulevard Avenue, #403 Fairfax, V A 22031 Denver, CO 80231 Anthony F. Smith (703) 849-0390 (303) 750-3983 Grand River Conservation Authority Jack D. Riessen George V. Sabol 400 Clyde Road, Box 729 Iowa Department of George V. Sabol Cambridge, Ontario Natural Resources ConSUlting Engineers NIR 5W6 Canada Wallace State Office 1351 East 141 Avenue (519) 621-2761 Building Brighton, CO 80601 Des Moines, IA 50319 (303) 457-4015 Stanley L. Smith (515) 281-5029 Maricopa County Flood Richard E. Sanderson Control District Michael F. Robinson U.S. Environmental 3335 West Durango FEMAIFederal Insurance Protection Agency Phoenix, AZ 85009 Administration A-1OOEA (602) 262-1501 500 C Street, S.W. 401 M Street, S.W. Washington, DC 20472 Washington, DC 20460 David A. Smutzer (202) 646-2716 (202) 475-8200 Pima County Flood Control District Elizabeth A. Rogers C.M. (Bud) Schauerte Planning and Development USDA/Soil Conservation Federal Insurance 201 North Stone, Service Administration Suite 400 139 Wolf Den Road 500 C Street, S.W. Tucson, AZ 85701 Brooklyn, CT 06234 Washington, DC 20472 (602) 740-6350 (203) 774-0224 (202) 646-2781 xxi


Speakers and Panelists Robbin B. Sotir Joseph E. Stuber Joseph J. Tram Robbin B. Sotir and CH2M HILL Maricopa County Associates 310 West Wisconsin Flood Control District 627 Cherokee Street, Avenue 3335 West Durango Street N.E., Suite 11 Milwaukee, WI 53203 Phoenix, AZ 85009 Marietta, GA 30060 (414) 272-2426 (602) 262-1501 (404) 424-0719 John M. Taylor Vassilios A. Tsihrintzis John H. Sowl Michael Baker, Jr., Inc. Psomas and Associates National Park Service 1420 King Street 3420 Ocean Park Midwest Regional Office Alexandria, V A 22314 Boulevard 1709 Jackson Street (703) 838-0400 Santa Monica, CA 90405 Omaha, NE 68102 (213) 450-1217 (402) 221-3485 Nicholas Textor Envirodyne Engineers, L. Scott Tucker William P. Stanton Inc. Urban Drainage and Flood Colorado Water 168 North Clinton Street Control District, #156B Conservation Board Chicago, IL 60614 2480 West 26th Avenue 721 Centennial Building (312) 648-1700 Denver, CO 80211 1313 Sherman Street (303) 455-6277 Denver, CO 80203 Ed Thomas (303) 866-3441 FEMNRegion I Lawrence Tunnell J. W. McCormack Post National Weather Service Kevin G. Stewart Office and Courthouse 10230 Smith Road Urban Drainage and Flood Building Denver, CO 80239 Control District Boston, MA 02109-4595 (303) 361-0666 Floodplain Management (617) 223-9500 Program Donald B. Von Wolffradt 2480 West 26th Avenue, Frank H. Thomas USDNSoil Conservation #156-B FEMNFederal Insurance Service Denver, CO 80211 Administration 14709 Dunbar Lane (303) 455-6277 500 C Street, SW Woodbridge, VA 22193 Washington, DC 20472 (202) 382-8769 Linda Stonier (202) 646-2717 National Park Service Kenneth L. Wahl NPS Western Region William L. Trakimas U.S. Geological Survey 600 Harrison Street, Suite ISO Commercial Risk Water Resources Division 600 Services, Suite 175 Denver Federal Center San Francisco, CA 94107 7321 Shadeland Station Lakewood, CO 80225 (415) 744-3975 Indianapolis, IN 46256 (303) 236-936 (317) 845-1750 xxii


Speakers and Panelists Alan R. Wald Washington State Department of Ecology Floodplain Management Mail Stop PV-11 Olympia, WA 98504 (206) 438-7419 French Wetmore French and Associates, Ltd. 153 Nanti Park Forest, IL 60466 (708) 747-5273 Gilbert F. White Institute of Behavioral Science Campus Box 482 University of Colorado Boulder, CO 80309 (303) 492-6311 Richard A. Wild Michael Baker, Jr., Inc. 1420 King Street Alexandria, VA 22314 (703) 838-0400 Steven R. Williams Hydro-Triad, Ltd., #100 1310 Wadsworth Boulevard Lakewood, CO 80215 (303) 238-6022 Wallace A. Wilson Michigan Department of Natural Resources Land and Water Management Division Post Office Box 30028 Lansing, MI 48909 (517) 335-3194 Robert L. Wold The Mitigation Assistance Corporation Post Office Box 382 Boulder, CO 80306 (303) 494-4242 Ann K. Woods CH2M Hill 6060 South Willow Drive Post Office Box 22508 Denver, CO 80111 (303) 771-0900 James M. Wright Tennessee Valley Authority Evans Building, Room 1A 524 Union Avenue Knoxville, TN 37902 (615) 632-4792 Kenneth R. Wright Wright Water Engineers 2490 West 26th Street, Suitc 100A Denver, CO 80211 (303) 480-1700 James T. WuIIiman CH2M Hill 6060 South Willow Drive Greenwood Village, CO 80222 (303) 771-0900 Anna Zacher U.S. Army Corps of Engineers Los Angles District c/o P.O. Box 8096 Calabasas, CA 91372 (213) 894-2028 xxiii Lyle W. Zevenbergen Love and Associates 2995 Centergreen Court South, Suite C Boulder, CO 80301 (303) 440-3439 Jon E. Zufelt U.S. Army Corps of Engineers 72 Lyme Road Hanover, NH 03755 (603) 646-4275


Part One Plenary Addresses


This page IS intentionally blank


MISSION IMPOSSmLE(?): WORKING TOGETHER IN FLOODPLAINS AND RIVER CORRIDORS Christopher N. Brown National Park Service The agencies, constituencies, and consciences represented here have numerous mandates and agendas. Given our diverse missions, is a common, multiobjective approach a "Mission Impossible"? In offering my opinion on this question, I emphasize several points about multiobjective river corridor management (which I will refer to as MORC). First, my own working definition views MORC as those individual projects or management regimens that attempt to meet a number of objectives, some of which would seem at first glance to be mutually exclusive. Possible objectives include flood loss reduction, navigation, water quality, wildlife habitat, recreation, soil conservation, historical and cultural resource preservation, and biological diversity; when put together, espedally in unusual combinations, they constitute MORC. And MORC should be distinguished from "multipurpose." Traditional multipurpose projects strive for flood control, hydropower, and recreation; or perhaps water quality improvement, wetlands restoration, and habitat enhancement. MORC connotes a more expansive rendering of a project (see the references, below, for a National Park Service publication containing several prime case studies). I offer two suggestions for making MORC the basic planning approach in our floodplains and river corridors: 1. Develop a sense of reverence for rivers andjloodplains. It is fundamental to the success of MORC that we develop our sense of reverence for the values in rivers and floodplains. My talk is entitled "Mission Impossible;" I ask you to think of the word "mission" in its sense as a calling, a purposeful undertaking. I do not think MORC can succeed without a deep and abiding appreciation of the natural and beneficial values of rivers and floodplains. It is not just the wild and scenic rivers that occupy a special niche in peoples' lives. Rivers like Penobscot, Mississippi, Atchafalaya, Cache La Poudre, Niobrara, and Snoqualmie are places of legend, adventure, enterprise, and refreshment for Americans. Rivers are in a sense sacred for us, just as the Ganges and Nile are for others. 2. Find-or develop"implementation geniuses" in your program. Jon Kusler points out that MORC requires a shift not only with regard to goals but to factors considered and the implementation techniques. Many of the incentives and much of the "know-how" for multiobjective river corridor management are already in place in the United States. A revolution in river management goals is not needed, nor is extensive new legislation or huge new budgets. What is needed is to much more broadly apply what is known about hydrology, geo-


4 Mission Impossible (?) morphology, and ecology in creative "consensus" ways to manage waters and adjacent floodplains and wetlands as integrated landscape features. We have the technical knowledge, the skills, and the approach; but how do we put them into practice? Hans Bleiker teaches a workshop on getting your agency's plans implemented. Hans says: The public agency 1) that has successfully analyzed the problem that it is supposed to solve, and 2) that, therefore has developed good solutions to that problem, but 3) that cannot implement those solutions-fails to fulfill its societal mission just as much as the agency that doesn't understand the problem that it's supposed to solve. Bleiker maintains that, despite the sophisticated planning our agencies employ, we have a major weak link: implementation. Hans sees two skills that are both essential to an agency's success: 1) technical works (plans, solutions, project proposals) that, when implemented, accomplish an agency's mission; and 2) "con sent-building" (outreach skills needed to insure implementation). Bleiker calls the second group of skills our" Achilles heel." And he challenges: "Focus the attention of your organization's best brains on citizen participation and 'consent-building'. Think of that! How many of us are willing to put our organizations' best brains on public involvement? And yet, we hear from each other time and again about the extreme frustration and paralysis of not being able to get a well-conceived plan implemented or to get "locals" to buy into a project. For a few thousand dollars, Hans says he will teach you to be an "implemen tation genius" who can cultivate "informed consent" from the public, so that "opponents . go along with a course of action that they-actually-are still opposed to. For free, I will tell you what I think his lesson boils down to: have the best brains in your organization develop an inclusive and broad-based public involvement program, and commit your project team to early, meaningful involve ment of all interested parties. Implementation Using the -Riverwork" Process. The program I work with-the National Park Service's Rivers, Trails, and Conservation Assistance Program, has devoted a good deal of energy to developing a successful process for effective public involvement. This "Riverwork" process (described in our Riverwork Book) is not the only way to involve the public, but it is one example of an open process that develops the powerful force of public "ownership" of a project. The Riverwork process rests on five tenets: 1) Client-driven-We help at the request of state and local governments and communities. 2) Cooperative-National Park Service (NPS) and local partners share responsi bilities, as formalized in a memorandum of understanding.


Brown 5 3) Cost-shared-While there is no set formula, all projects involve significant local cost-sharing. Typically NPS puts in well under half of the funds and time devoted to the project. 4) Catalytic-The NPS plays the role of catalyst and facilitator, helping its project cooperators define what they want and come up with an effective process for attaining it. 5) Concrete results-NPS measures its effectiveness by counting tangible successes, like acres of land protected, miles of new trail built, legislation passed, and coalitions formed. Our landscape architects, planners, and other resource professionals work on about 100 greenway, trail, river corridor, and open space projects each year; public participation and implementation are trademarks of our approach. Some examples of NPS projects with a MORC slant include the Missouri River Corridor Project (Nebraska); Santa Rosa Creek (California); Portneuf Project (Idaho); Lackawanna River (pennsylvania); Spickett River (Massachusetts); Trinity River (Texas); Four-Mile Creek Greenway (Iowa); and Grand Junction, Colorado. One example, the Bear River Greenway in Evanston, Wyoming, is presented in Appendix 1. Please contact my office for more information on these projects or our program. 1991 Knoxville Workshop A recent workshop in Knoxville, Tennessee, helped move MORC a step forward (see Appendix 2 for other milestones). For the first time ever, nine federal agencies (the Environmental Protection Agency, the U.S. Army Corps of Engineers, the U.S. Geological Survey, the Soil Conservation Service, the Federal Emergency Management Agency, the National Oceanic and Atmospheric Administration, the Fish and Wildlife Service, the National Park Service, and our host, the Tennessee Valley Authority) held a joint session on the natural and beneficial values of flood plains. We successfully avoided the paralysis of falling back on our individual agency mandates and constraints, and instead thought creatively about forming new partnerships. Participants gained an understanding of the need for skills in the area of facilitating and developing consensus. I believe that we will embrace MORC as a concept central to accomplishing our agencies' missions, for, as Nobel-winning economist John Maynard Keynes once said, "We will do the rational thing, but only after exploring all the other alternatives. I hope that one result of this conference will be a rethinking of the accepted. Some thoughts: The fact that we serve people and protect, preserve, and enhance the values of floodplains unites us professionally. Let us see ourselves not as flood-loss reduction specialists, wetlands scientists, etc., but as floodplain managers, with the vision and ethical responsibility befitting a profession.


6 Mission Impossible (!) Why not rename the organization ("Friends of Floodplains") to better reflect the personal, as well as the technical, commitment we have to floodplains and river corridors? How about working with conservation organizations (they have many different perspectives and political connections) as partners in national activities and local projects? Let each agency and organization here renew its sense of mission, and assign its "best brains" to public involvement. Kusler, Jon 1991 References Multiobjective River Corridor Management: An Introduction to Issues. Association of Wetland Managers. National Park Service 1989 The Economic Impacts o/Protecting Rivers, Trails, and Greenway Corridors: A Resource Book. 1991 1988 Riley, Ann 1990 Innovative Stream Conservation: A Case Book on Multi-Objective River Corridor Management (draft). Riverwork Book. Some Basics on Urban Stream Restoration (technical paper). National Park Service. Tennessee Valley Authority 1990 Conserving Your Valuable Floodplain Resources (brochure). Wright, James M. 1991 Opportunities and Efforts/or Multiobjective Management o/River Corridors and Coastal Zones in Virginia.


Brown Appendix 1 Bear River Greenway 7 NPS Rivers and Trails staff have helped a range of participants join forces and establish a four-mile greenway along the Bear River in Evanston, Wyoming. The recently improved Bear River State Park is just outside town, but the river within Evanston was badly degraded by years of neglect. National Park Service assistance with a two-year cooperative planning effort (1989-1990) led the city and a new nonprofit organiza tion to produce a comprehensive plan for the river corridor within the city limits. Federal, state, and local agencies have been eager to contribute to the project. As of April 1991, the following public contributions have been made: U.S. Geological Survey (hydrologic studies)$150,OOO U.S. Army Corps of Engineers (bridge and access design) ao,ooo USDA-Soil Conservation Service (diversion designs) (n/a) Land & Water Conservation Fund (a federal program) 25,000 Wyoming State Game and Fish Department 73,500 Wyoming Recreation Comm. (Bear River State Park) 1,372,000 Wyoming State Government (various agencies have committed unspecified time and resources to the project) Bridger Valley Conservation District Uinta County Youth Services and State Hospital Adolescent Treatment Unit (2,000 hours @ $5.00/hr. est.) Uinta County (in-kind labor on stream bank stabilization) City of Evanston (land purchases) City of Evanston (in-kind labor-stream stabilization) National Park Service (staff time, expenses) (n/a) 8,000 10,000 13,000 250,000 9,000 $42,000 Total Documented Funds Committed to the Project: $1,982,500 The NPS was involved with the nonprofit group in arranging for a $100,000 donation by the Chevron Corporation, which will be used to construct bridges and other greenway improvements. Chevron also challenged other corporations in the region to participate, and its corporate leadership has been mirrored by smaller contributions by many individuals and small businesses. This effort has resulted in the purchase of approximately 60 acres of land along the river in Evanston-green open spaces that are now available for recreation and wildlife habitat. 1979 1987 1988 Appendix 2 Recent Milestones in Multiobjective River Corridor and Floodplain Planning and Management Revisions to Water Resources Council's report on the Unified National Program on Roodplain Management incorporated concern for natural and beneficial values of floodplains President's Commission on Americans Outdoors calls for greenways along stream valleys, floodplains, corridors Paper by J. Glenn Eugster on Multiobjective planning


8 1989 1989 1989 1989 1990 1990 1990 1990 1990-1991 1991 1991 1991 1991 1992 Mission Impossible (?) Congressman Joseph McDade and the NPS Rivers & Trails program sponsor six workshops on MORC; 600 people attend Urban Stream Corridor and Stormwater Management conferenceColorado Springs (March) Multiobjective Management of River Corridors and River Restoration (hosted by TVA, EPA, Association of Wetlands Managers) International Symposium on Wetlands and River Corridor Manage ment-Charleston, South Carolina (July) Congressman Joseph McDade introduces HR 4250-State & Local Multi-Objective River Corridor Assistance Act of 1990 Omnibus Water bill strengthens Corps' environmental role MORC projects in Tulsa, Scottsdale, Prairie du Chien, and Soldiers Grove continue to succeed and inspire others "Country in the City" Conference, Portland, Oregon (April) focuses on urban wetland protection & restoration Papers by Jim Wright, Ann Riley, and Jon Kusler -TVA brochure shows benefits of undeveloped floodplains MORC Casebook from NPS Rivers and Trails program Economics of Greenways handbook from NPS R& T program Southeastern Regional Workshop on Natural and Beneficial Values/ Floodplain Management-Knoxville (March), cooperatively sponsored by nine federal agencies H.R. 1236, Flood Insurance, Mitigation, and Erosion Management Act of 1991 passes House-promotesfederal cooperation to protect natural and beneficial functions (planned) Floodplain Management Resource Center (Boulder, Colo.) compiles multiobjective floodplain publications (planned) Multiobjective Floodplain Managementworkshop-Pittsburgh, Pennsylvania (planned) Multiobjective theme for 16th Annual ASFPM ConferenceMichigan


EPA AND THE NATIONAL ENVIRONMENTAL POLICY ACT: FACll..ITATORS OF CREATIVE FLOODPLAIN AND NATURAL RESOURCES MANAGEMENT IN THE 90s Shannon E. Cunniff and Richard E. Sanderson Environmental Protection Agency Introduction Demand for environmentally sound flood risk reduction at the regional and watershed levels can be expected to increase. The National Environmental Policy Act (NEPA), which pertains to major federal action having the potential to affect the human environment, can promote comprehensive and innovative floodplain management that is harmonious with or advances environmental risk reduction goals if properly and creatively applied. The flexible nature of NEPA lends itself to use as a vehicle for carrying out floodplain risk reduction and environmental risk reduction with full public participation. The Environmental Protection Agency's (EPA) NEPA oversight responsibility stems from both NEPA and Section 309 of the Clean Air Act, which requires that EPA review and comment in writing on the environmental impact of any matter relating to the duties and responsibilities of the administrator. EPA will be using its NEPA review responsibility to encourage creative planning that enhances environ mental quality and/or reduces environmental risks. EPA believes that reduction of environmental risk comports well with reduction of flood hazard risks. NEPA documents that go beyond site-specific, single-purpose proposals will serve to integrate the variety of issues (e.g., floodplain encroachment, wetland losses, non-point source pollution, recreation, etc.) directly or peripherally related to watershed management. Traditionally, EPA's NEPA and other environmental review responsibilities have played an important role in ensuring that environmental considerations are part of the policy framework for other federal agencies whose activities affect environmental quality. In the future, EPA will use its NEP A oversight responsibility, as well as other environmental authorities, to try to foster cooperation among federal, state, and local governmental entities to arrange integration of these cross-media, mUltiobjective issues. NEPA Requirements and Values Implementation of NEPA is one of the most successful elements of the overall commitment the federal government has made to environmental protection over the last two decades. Unique among the many federal environmental statutes, NEP A is procedural rather than regulatory in nature. NEPA simply requires that the decision maker and the public be aware of the effect of a proposed federal action. In the NEPA process, the direct, indirect, and cumulative impacts of a proposed federal action and its alternatives are evaluated and compared, and mitigative measures identified in a publicly reviewed and commented-upon document.


10 EPA and NEPA NEP A is the only federal environmental law that requires analysis of cross-media impacts and allows agencies to discuss the impact of their action on the totality of the human and natural environment. The potential of NEPA is often overlooked. NEPA can be used to achieve a number of objectives; it can provide a process for consideration of new policy direction and research in management decisions; provide a forum for long-term, region-wide and/or watershed planning; assess cumulative effects and analyze management areas; and carry out analysis and disclosure of site-specific environmental impacts of individual projects. NEPA and comparable state statutes can and should be utilized to facilitate development of creative solutions to floodplain management that reduce environmen tal risk and further national, regional, and local environmental protection goals. Agencies and the public should view NEPA as an opportunity to initiate early planning and creatively integrate multiobjective planning within the context of environmental impact assessment. The basic components of the NEPA procedurealternatives analysis; assessment of direct, indirect, and cumulative impacts; mitiga tion; and public involvement-are appropriate tools that can be utilized to realize integration of floodplain management within the broader context of water and natural resources management; integration of nonstructural, structural, and environmentally enhancing solutions to floodplain management; multiobjt!ctivt! floodplain management; assessment and establishment of a floodplain's natural and beneficial uses and values; and watershed approach to floodplain management. The values of NEPA are many and local jurisdictions should seek to take advantage of the federal impact analysis to define local goals and/or further their implementation. EPA's Science Advisory Board (EPA, 1990) recognized that the environment is an interrelated whole, and that society's environmental protection efforts should be integrated as well. Specifically, the Science Advisory Board report recommended that government agencies assess the range of environmental problems of concern and then target protective efforts at the problems that seem to be the most serious. As a result, EPA will be encouraging federal agencies to undertake cross media analyses and will encourage the use of NEP A to determine and evaluate multiobjective goals where such can contribute to enhancing the natural environment or reducing environmental risk. NEPA is an appropriate vehicle to promote early involvement in planning and risk reduction-both in terms of reducing environmental risk but also natural hazard risks and the risk of depleting the national flood insurance fund. Through NEPA's alternative analysis process, EPA will be emphasizing full exploration of the


Cunniff and Sanderson 11 nonstructural flood protection options. EPA believes that these options can address environmental goals as well as flood protection goals. For example, the preservation of flood storage areas, greenways, wetlands, and coastal barriers are mutually beneficial measures that reduce flood risk and losses as well as enhance water quality and biodiversity. Conversely, it has been EPA's experience that irrevocable damage frequently results from structural measures, correlated floodplain development, and associated cumulative impacts. EPA believes that societal and environmental benefits in terms of both pollution prevention and ecological enhancement can be demon strated as greater when nonstructural options and their cumulative impacts (both beneficial and adverse) are fully explored. The Council on Environmental Quality's (CEQ) regulations on the implemen tation of NEPA require that all reasonably foreseeable future actions be included in impact analyses so that decision making is based on the total probable future condition of an area. This requirement links a site-specific proposal to the broader floodplain management issues of the local jurisdiction. It is the NEPA document's cumulative impacts analysis that locals should utilize to reveal and assess long-term trends and develop policies to address the desirability of such trends, their compatibility with other local goals, and develop means of coordinating local objectives. A complete cumulative impacts analysis is necessary to provide a meaningful assessment of the full societal costs of nonstructural and structural flood protection options. Indirect and cumulative impacts must be addressed in the NEPA document's impact assessment and mitigation sections, regardless of the federal agency's responsibility for the impact or its commitment to perform the mitigation. Logically, involvement oflocals in the environmental impact statement (EIS) process is crucial for development of meaningful evaluations and to determine which societal benefits are desirable and which societal costs are acceptable. Furthermore, for the NEP A process to work to its fullest potential, local jurisdictions should use the NEPA document to publicly indicate their capability and willingness to implement measures to offset environmentally or socially unsatisfactory impacts. CEQ's NEP A implementation regulations provide for tiered analysis and decision making. The tiered process promotes watershed planning and consolidation of flood control into overall floodplain management issues and facilitates early coordination and planning. It is the cornerstone of the floodplain management process that integrates nonstructural and structural options with other local goals such as water quality improvement, recreation, and aesthetics. In federal flood control projects, the first tier usually addresses the overall feasibility of a proposal. In such cases, local jurisdictions can still utilize that tier to identify and implement policies, programs, andlor actions required to achieve local goals. The U.S. Army Corps of Engineers' Water Resources Development Act of 1986 enhanced cost-sharing responsibilities for local project sponsors and effectively increased the level of local participation in planning and mitigating flood control projects. Local sponsors are not utilizing their involvement in federal flood control planning and the associated NEPA process to its fullest potential. In some cases the effect, unfortunately, has been to limit the scope of alternatives by establishing minimal locally acceptable flood protection goals. Localjurisdictions should consider


12 EPA and NEPA these and other merits of expanding the NEPA analysis to address related local decisions and thereby facilitate integrated floodplain management. At a minimum, NEPA documents should be used to assess and establish a floodplain's natural and beneficial uses and values; local governments can use this information to assist subsequent floodplain management programs. The first tier can address broader policy or programmatic issues. A tiered NEP A process can ensure that policy or program decisions utilize a systematic, comprehen sive, and public assessment of the total human and natural environment, including cumulative economic, social, and environmental impacts. Nonstructural options of flood risk reduction should be fully explored at this initial tier and integrated into all subsequent planning. The tiered process should be explored when a series of similar or related actions are expected to occur within 10 or 20 years or when several activities are occurring within a single watershed or floodplain area and need to be coordinated. A programmatic NEPA analysis can be conducted to address activities requiring Clean Water Act Section 404 permits for placement of dredged or fill material, as well as other activities within the floodplain that may effect its beneficial uses but that do not require a federal permit. Despite the planning and mitigation advantages to such an approach, other practical reasons exist to promote use of the tiered process. Such NEPA analyses can facilitate addressing site-specific actions through subsequent abbreviated impact analyses and permitting procedures, saving government agencies and developers time and money. Both large-scale and expanded purpose site-specific NEP A processes can be used to enhance public participation in a comprehensive planning process which can encourage "buying into" or committing to an overall program. The tiered approach can also be utilized to accomplish coordinated implementa tion of federal and state programs in a region. For example, a NEPA analysis can bring together the U.S. Army Corps of Engineers, the Federal Emergency Management Agency, the U.S. Fish and Wildlife Service, the National Oceanic and Atmospheric Administration's coastal zone management group, and appropriate state or local officials in a cooperative effort to address common goals. The NEPA process also offers a chance to modify existing federal projects in ways that will offer substantial environmental benefits. Locals should encourage the federal agencies to periodically review their operations to determine if means exist to reduce environmental impacts or to determine if opportunities exist to coordinate operations to enhance environmental quality and other local goals. State and local floodplain management and environmental agencies may need encouragement to utilize NEPA to its fullest potential. For a variety of reasons, regulatory agencies may shy away from programmatic NEP A analyses that have no immediate federal action. This is partly because of the regulatory burden that the agencies are under; however, comprehensive planning can reduce the time and effort needed for subsequent NEPA analyses. Another unfortunate reason for the lack of willingness to undertake or participate in programmatic NEP A analyses is the lack of recognition given by internal tracking mechanisms because there is no federal action to credit at this phase. Agencies may resist broadening the scope of analysis as they may desire that the NEPA document only address the specific federal action.


Cunniff and Sanderson 13 These attitudes, coupled with the lack of value assigned to programmatic efforts, limit comprehensive planning and regional issue resolution. EPA will be working to remove these barriers. Active solicitation of federal agency support for 1) a tiered EIS program, or 2) the broadening of a single purpose NEP A document to include goals and actions beyond those which are the sole responsibility of the federal government must be pursued by local and state entities as well as EPA. These actions are necessary to integrate flood control structure construction into floodplain management and floodplain management into the broader context of water and natural resources management. Conclusion EPA believes that NEPA represents a valuable tool that federal, state, and local entities can utilize more effectively to realize significant benefits to both local floodplain management and environmental protection. EPA will be doing more to foster cooperation among the agencies to ensure better integration of floodplain management within the broader context of water and natural resources management. However, EPA cannot undertake this effort alone. Strong local recognition of the value of such integration is necessary to encourage multiobjective floodplain manage ment. References Council on Environmental Quality 1978 "Regulations for Implementing the Procedural Provisions of the National Environmcntal Policy Act." 43 Federal Register 55978 (November 29, 1987). Council on Environmental Quality 1970 National Environmental Policy Act. Public Law 91-190, 42 U.S.C. 4321-4347, January 1, 1970, as amended by Public Law 94-52, July 3, 1975, and Public Law 94-83, August 9, 1975. Environmental Protection Agency 1970 Clean Air Act. Public Law 91-604,42 U.S.C 7609. 1990 Reducing Risk: Setting Priorities and Strategies/or Environmental Protection. Washington, D.C.: Environmental Protection Agency


STONE SOUP AND HAZARD MITIGATION Edward A. Thomas Federal Emergency Management Agency Region I Postdisaster Hazard Mitigation Is Not Just Report Writing Anymore Postdisaster hazard mitigation was once limited to sporadic efforts to insure compliance with the substantial damage rebuilding requirements of the National Flood Insurance Program. Then, in 1980, an interagency agreement was signed that committed the federal government to facilitating the preparation of intergovern mental, interagency reports which analyzed the causes of disasters and made recommendations about how such disasters could be prevented in the future. These reports were a great step forward; however, all too often there were insufficient funds or staff time available to implement these recommendations. Thanks to recent changes to the disaster assistance legislation,1 postdisaster hazard mitigation is not just report writing anymore. It is "stone soup." Stone Soup and Hazard Mitigation Gilbert White in his presentation at the beginning of this session beautifully and clearly showed that there is no unified program for floodplain management in the United States. Certainly this is correct. Yet this in no way means that hazard mitigation, or, more precisely, postdisaster hazard mitigation, cannot be made to work. Let's think of it in the context of an old fable. "Stone Soup" is a traditional tale from many lands. As it is told in parts of France, the story goes like this: Two hungry strangers come to a village. Everyone in the village is starving because they had a natural disaster that year-the crop failed. The strangers announce that they will make soup from stones to feed everyone in the village. First, the strangers ask for a pot, then water and firewood to cook the stones-no problem. Then, individually, they go around asking for enhancements for the soupa little salt, a few spices, a few greens, some meat, etc. Most people have some contribution to make the soup a little better. Finally, to the amazement of everyone, a huge pot of soup is prepared that feeds the entire village. The moral of the story is simple. Individually these people, confronted by a disaster, were unable to solve a problem. They needed a catalyst, some direction and some help. Postdisaster hazard mitigation is a lot like stone soup. No one party can solve the problem. In each city/region/disaster area, we need to develop a package of solutions. Some of the principal programs that can be used as part of the mix in this package are: Section 406 of the Stafford Act. Under Section 406 of the Stafford Act, FEMA may fund, on a 75-25 % cost sharing basis, cost-effective hazard mitigation efforts at a site eligible for public assistance. 2


Thomas 15 Section 404 of the Stafford Act. Under Section 404 of the Stafford Act, FEMA may make grants on a 50-50 matching basis for cost-effective hazard mitigation measures. 3 The 50% match for these grants may be in hard cash or soft, in-kind services, or a combination of these approaches. 4 Depending on the size of the disaster,5 a significant amount of money may be available to help mitigate the consequences of disasters in the afflicted area. This program has gotten off to a slow start, but in places like New Hampshire, where the state and FEMA have worked well together to explain the program to local governments, it is oversubscribed. Use ofFEMA Staff. FEMA has staff personnel who often can be made available to provide technical assistance to state and local personnel who are performing postdisaster hazard mitigation duties. In addition, it is increasingly common for FEMA and sometimes state staff to be present at Disaster Application Centers to provide advice on hazard mitigation to the victims. Small Business Administration. In those disasters involving the Small Business Administration, Disaster Loan Program-eligible individuals and businesses can receive loan increases beyond the repair of their damages to cover compliance with existing codes and ordinances, 6 as well as up to a 20 % increase in the loan amount to cover hazard mitigation beyond that required by state and local codes and ordinances. Individual and Family Grants. In those disasters involving the use of the Individual and Family Grant Program, the state sometimes requests, and the federal coordinat ing officer sometimes approves, the use of grant funds for individuals to carry out hazard mitigation efforts. Additional Assistance. Other programs and organizations that assist disaster victims, such as the Red Cross, the various Voluntary Agencies Active in Disasters, and FEMA's Cora Brown Fund may be available to meet victims' needs, including mitigation, if the other programs available in a disaster do not meet these needs. Examples of Stone Soup There are many examples of hazard mitigation stone soup being made around the country. Two of the most recent come to us from Tennessee and Maine. Storms and Floods in Tennessee in January 1991.1 The governor and the governor's authorized representative for the floods that took place in Tennessee in January of this year expressed great interest in solving the repetitive flood damage that has afflicted several parts of the disaster area. Based on this interest, FEMA worked with the state and the Interagency Hazard Mitigation Team to develop a hazard mitigation program to meet the needs of these areas: 1. In the Sandtown area of Rhea County, a special program was developed to relocate the residents from an extremely hazardous area which was subject to erosion, stream meander, and high velocity flood waters. With the assistance of the Red Cross, the state used Individual and Family Grants to


16 Stone Soup and Hazard Mitigation purchase homes for the victims outside the floodplain. As a condition for the assistance, the residents signed options for the sale of their land to the county at market value. A Section 404 grant application for the purchase of this land is now being developed by the county with assistance from the Tennessee Emergency Management Agency. Eventually, this hazardous location will be a county park. 2. In Obion County it was determined that the best solution to the repetitive flooding in the city of Obion was a ring levee that was then under design by the U.S. Army Corps of Engineers (USACE). At the request of the federal coordinating officer, the USACE accelerated its design, and we hope that a Section 404 application will soon be forthcoming for this project. 3. In the Bogota section of Dyer County there had been frequent, low-velocity flooding of a broad flat floodplain tributary of the Mississippi. Relocation, as was done in Rhea County, was not necessary because this floodplain was not particularly hazardous, nor was it practicable because in this area there were few sites safe from flooding. In this area, it was determined that the state and FEMA would supply Individual and Family Grant Funds to elevate the houses three to four feet to the level of the l00-year flood. This project is currently under way. Ice Jam Flooding in Maine, April 1991. In April 1991 there was tremendous flooding from ice jams and snow melt in Aroostock County in northern Maine. The Village of Dicky in the small town of Allagash was particularly hard hit. 8 FEMA, the state, and the American Red Cross developed a unique approach to mitigating future flood damage. 1. Technical Assistance-Hydrology and Hydraulics. At the request of the federal coordinating officer for the disaster, FEMA flood insurance staff calculated the level of the l00-year flood in Allagash, thus providing a basis for proper mitigation. 2. Technical Assistance-Design. At the request of the state and the Red Cross, FEMA, in cooperation with the State Emergency Management Agency and state flood insurance coordinator, developed model mitigation strategies for all the types of structures damaged in this flood. These strategies ranged from elevating furnaces a few feet in basements to relocating structures from the most hazardous sections of floodplain. 3. Program Coordination. Staff from the Red Cross are working with the victims individually to put together packages of assistance to meet their disaster-induced needs, including mitigation. The programs involved include: funds from the Maine Office of Energy; loans and grants from the Farmers Home Administration; loans from the Small Business Administration; rebuilding assistance from the Voluntary Agencies Active in Disaster, especially the Mennonites; and financial assistance from the American Red Cross. 9


Thomas 17 Conclusion Lack of a unified national program to address floods and other natural hazards does not mean that solutions cannot be developed for many hazardous situationsespecially in a postdisaster context. We can work together to develop win-win, stone soup strategies which will both prevent many of our citizens from becoming disaster victims again from future floods and provide substantial credit points under the flood insurance program's Community Rating System. All you need to start is stones. 10 Notes 1. The Stafford Act, Public Law 93-288, as amended. 2. When the president determines that a disaster is of a magnitude such that supplemental federal assistance is needed, public assistance may be provided to fund the rebuilding of government-related items such as roads, bridges, sewage treatment plants, and parks. 3. Among the items that may be suitable for funding under Section 404 are flood prone land acquisition, natural hazard warning systems, hazard education programs, stormwater management, and similar, common sense, cost-effective measures. 4. Many communities are able to submit applications for Section 404 projects with little or no cash match by using their own staff and equipment to carry out projects. 5. Available funding for postdisaster mitigation is calculated as follows: multiply the funding for all permanent restorative work and the Public Assistance Program by 0.2. One half of this will be FEMA's share. The other half is matched as described above. 6. This would include local requirements such as elevation to met the standard of the National Flood Insurance Program, or even a local requirement to move from a hazardous location such as a floodway. 7. For a complete description of the disaster see Interagency Hazard Mitigation Team Report, Storm December 1990, available from FEMA, Disaster Assistance Programs Division, Suite 729, 1371 Peachtree Street, Atlanta, GA 30309. 8. For a more complete description of this flood see Hazard Mitigation Survey Team Report, Maine, June, 1991, FEMA-DR-0901-ME, available from FEMA, LW. McCormack Post Office and Courthouse Building, Room 436, Boston, MA 02109. 9. N.B.: FEMA's Individual and Family Grant Program was not included in this disaster, therefore many other programs were used to meet the needs of victims. 10. Further information on the programs mentioned in this paper can be found in: FEMA DR&R-l (3/81) Item #8.0056 Handbook for Applicants FEMA DR&R-2 (7/81) Item #8.0057 Eligibility Handbook


18 Stone Soup and Hazard Mitigation DAP-12 (6/86) Item #8.0569 Making Hazard Mitigation Work-A Practical Handbook for State Officials DAP-16 (3/86) Item #8.0383 When You Return to a Storm-Damaged Home DR&R-18 (1/87) Item #8.0514 Individual and Family Grant Program Handbook Pursuant to Public Law 93-288 DAP-19 (3/87) Item #8.0600 A Guide to Federal Aid in Disaster DAP-21 (6/89) item #8.0721 Digest Federal Disaster Assistance Programs These publications may be requested by letter addressed to FEMA, Post Office Box 702 74, Washington, DC 20024. All requests should include the reference identification number and full title of the publication. Please include your full name, address, and zip code.


PERSPECTIVES: A STATUS REPORT ON THE NATION'S FLOODPLAIN MANAGEMENT ACTIVITY Frank Thomas Federal Insurance Administration Federal Emergency Management Agency The Federal Floodplain Management Task Force is completing a multiyear effort to assess the status of floodplain management in the United States. Due to be published in 1992, this assessment discusses the status of activity, identifies trends, and recommends actions for improving floodplain management. My remarks briefly describe the assessment and then highlight some major findings and opportunities for improving floodplain management. Gilbert White will follow with his perspective on actions needed to improve floodplain management. The Status Report on the Nation's Floodplain Management Activity seeks to identify opportunities for improving the effectiveness of floodplain management. Its approach is twofold: first, compile data describing the current milieux of economic, political, and social forces affecting floodplain decisions and what has been done with loss reduction strategies and tools; and second, evaluate the meaning of this compilation using the observations and recommendations of representative stakeholders and a national panel of experts. The assessment will be published as a comprehensive reference volume accompanied by a summary volume containing major findings and recommendations. Trends Six national, often related trends bear upon decisions affecting floodplains and the effectiveness of tools. In the last 30 years, state and local It:adership expanded while the federal role became focused on research, standard setting, information dissemination, and technical assistance. Concurrently, management philosophy and strategy has shifted significantly; primary reliance on flood control structures has been replaced by a balanced structural-nonstructural approach wherein land use management regulations and building codes are a primary loss reduction tool. Regional basin-wide approaches have lost favor to community-focused management. Legislatively, there has been a growing emphasis on program linkages as exemplified by the denial offederal flood insurance benefits in areas of environmentally sensitive coastal barriers. The number of skilled state and local floodplain managers has rapidly expanded as evidenced by the emergence of the Association of State Floodplain Managers. Finally, the legal force of liability for negligence has become a strong motivational force. Accomplishments Three broad areas of accomplishment stand out, given the continued rapid urbanization along the nation's coasts. First, in spite of a lack of high quality flood loss measurement data, evidence indicates that the per capita average annual loss of life is down from earlier in the century, average annual property damage has held


20 A Status Report on the Nation's Activity steady compared to gross national product, and a new awareness of the value of floodplain environmental functions has begun to reduce loss of the functions. Second, a more balanced management approach has emerged with nonstructural and structural strategies being jointly built into "best mix" solutions tailored to local loss reduction problems. Third, the shift away from federal leadership has pushed decision making to the local level particularly through nonfederal cost sharing eligibility requirements for flood control projects and through local land use regulation eligibility require ments for federal disaster assistance and flood insurance. Problems Three groups of problems stand out. First, flood risk must be more effectively addressed through the location of new infrastructure, mitigation programs for older, high loss structures, and the integration of loss reduction strategies for human life and property, with similar strategies for floodplain environmental functions. Second, floodplain data, information dissemination, and education are inadequate and receive low priority, particularly with respect to local officials and smaller communities. Finally, the accepted floodplain management strategies and tools suffer from several deficiencies, including bias against low income communities and individuals, competing mandates of some governmental programs, the absence of a widely accepted floodplain management definition and measurable goals, and reluctance to enforce floodplain management regulations. Conclusion Based on the assessment, one must applaud the significant progress of the past 30 years. However, nobody wins a cigar, only a pat on the back with encouragement to continue to press forward and realize the opportunities laid out before us.


A REVIEW OF THE STATUS REPORT ON THE NATION'S FLOODPLAIN MANAGEMENT ACTIVITY, APRIL AND SEPTEMBER 1989 Gilbert F. White University of Colorado The prospective release of the Status Report on the Nation's Floodplain Management Activity calls fresh attention to the findings in that document, the evolution of thinking leading to it, and the range of public action that properly should follow in the train of the conclusions of the federal interagency task force, as reported by Frank Thomas. Situation in Brief Although public discussion of the multiple problems of floodplain management first took shape in connection with the controversies over the flood control acts of 1936 and 1938, the formal need for an integrated program of floodplain management did not come into public focus until 1966, when a Bureau of the Budget Task Force brought forth House Document 465 with its recommendations for a variety of actions by public agencies. In 1976 those activities were reviewed and, in most instances, reinforced by the revised report on A Unified National Program, and then in 1986 a further revision was undertaken. The document under review brings the basic information up to date and also poses opportunity for further constructive action. Two Goals It is clear that there have been two major goals in managing t100dplains and that the first held preponderant attention until very recently: it was to reduce vulnerability to flood danger and damage by using a whole arsenal of possible private and public measures to do so. At the same time the public interest was in preserving and enhancing the natural values of the nation's floodplains, and this too was expressed in a complementary set of measures by both public and private agencies. The whole floodplain management effort recently has been reinforced by the interest and activities of agencies concerned with this second goa\. Assessment A national review committee (listed at the end of this paper) reviewed the materials in the assessment report as well as comments from interested nonfederal and state agencies, and called attention to 10 characteristics of the report that deserved more thorough discussion. 1. The data available on floodplain use and flood damages are inadequate to permit a genuinely accurate and comprehensive review of the national situation.


22 Review of the Status Report 2. The current activities are neither unified nor national in scope. 3. Federal agencies have been diligent for the most part in carrying out sectors of the unified program, but they differ greatly in the extent to which they foster local and state efforts. 4. There has been no central direction or coordination for the activities of the more than half a dozen federal agencies. 5. There have been advances in shifting responsibility in some sectors to state and local agencies, such as in the National Flood Insurance Program participating communities, the limitations on subsidies for new construction, the development of new forecaSt and mitigation technologies, and the identification of wetlands. 6. The field policies and practices have been inconsistent in assimilating the wetland concept. 7. There has been a patchwork of federal legislation to support or extend the aspirations of the unified program. 8. Data collection continues to be uncoordinated and incomplete without central direction from anyone federal agency. 9. States and communities in many instances lack the resources to carry out activities that would be indicated under the unified program and are not covered by federal practice. 10. There is a great difference among states and communities in the degree to which they have been able to respond to federal initiatives and incentives. Factors Affecting Further Activity Some of the direct factors that will influence what happens next are: the amount and quality of information that affects public perception of the severity and characteristics of flood problems, the difficulty of reconciling programs in particular drainage areas to different modes of social analysis, and the barriers that exist between management at federal, state, and local levels. On the indirect side, the national situation in general and more specifically in some regions suggests that: we are dealing with an aging public infrastructure, new forms and patterns of land use are emerging on floodplains, interest in recreational use of waterside lands has increased, public concern with overall environmental protection has grown rapidly, the scale and sophistication of urban planning in the local communities has increased, and there will continue to be major problems of boundary effects between different forms of regulation. Recommended Action The review group considered these and a number of other factors and outlined six directions along which it felt action should be pursued with the support of the federal agencies and of all state and local groups concerned.


White 23 1. There is pressing need to integrate considerations of loss vulnerability and natural values into broader state and community development and resource management programs so that floodplain management is only one feature of those development activities. This would include fostering state floodplain managementplans,requiringcomprehensivemanagementplanstoparticipate in the National Flood Insurance Program (NFIP), issuing a new executive order that would seek to consolidate federal activities in a consistent fashion with state and local activities, and improving and demonstrating coordinated federal activity in particular areas. 2. The data base should be improved by remapping urban areas subject to rapid land use change, providing better support for special conditions such as alluvial plains, lake levels, ice jams, moving channels, subsidence, storm drainage, and mud flows; gathering flood loss data on a consistent basis; and analyzing the full costs and benefits of selected management measures. 3. All such activities should be modified to give greater weight to differences resulting from local landscape conditions, and this could be supported by possible use of performance standards and in the extension of the Com munity Rating System under the NFIP. 4. The present conflicts and divergences among federal programs should be remedied so far as practicable by organizing an Office of Management and Budget task force that would address those issues and seek to find adminis trative solutions under current legislation. S. A major need is for reducing the vulnerability of existing buildings by adding to the approaches currently in use, by reporting what, in fact, is happening, and by critically appraising the current preparedness and retrofitting activities. 6. The current weakness of professional skill and public education relating to the whole effort of floodplain management integrated with other develop ment activities could be advanced by new training programs under federal auspices and by activities at all levels in expanding public education programs. The national review committee hopes that these issues will receive critical public attention as soon as the full assessment report is in the public domain. This will be a partial responsibility of the Association of State Floodplain Managers, but it needs to be shared much more widely among other state, local, and nongovernment organizations. The National Review Committee Raymond J. Burby Gerald E. Galloway James E. Goddard James G. Gosselink H. James Owen Rutherford H. Platt William E. Riebsame, Vice-chair John R. Sheaffer French Wetmore Gilbert F. White, Chair Stanley M. Williams Mary Fran Myers, Staff


This page IS intentionally blank


Part Two Flood Hazard Mitigation


This page IS intentionally blank


APPLYING FLOODPLAIN REGULATIONS TO THE REAL WORLD Cynthia A. Baumann and G. Nicholas Textor Envirodyne Engineers In 1988, the Illinois Department of Transportation, Division of Water Resources, proposed rules regulating construction within all regulatory floodways in the northeastern portion of the state. These rules were signed into law on November 29, 1988. This law generated an avalanche of permit applications that temporarily overloaded the division. Consequently, four consulting firms were hired to aid in the review of these applications. Working as a consultant for the State Division of Water Resources provided a unique insight into how and why floodplain regulations are written and, more importantly, interpreted and applied. Federal, State, and Local Regulators The first, and probably the most common, misunderstanding of the rules was with the definition of a floodway. The state rules, in this case, only govern construc tion within a floodway, yet numerous applicants submit applications for construction that take place within the flood fringe of a stream where a regulatory floodway exists. The following definitions are taken from the state's rules: "Regulatory Floodway"-The channel and that portion of the floodplain adjacent to a stream or watercourse which is needed to store and convey the anticipated future loo-year frequency flood discharge with no more than a 0.1 foot increase in flood stage due to loss in conveyance or storage, and no more than a 10% increase in velocities. "Floodplain"-That land adjacent to a body of water with ground surface elevations at or below the loo-year frequency flood elevation. "Flood Fringe"-That portion of the floodplain outside the regulatory floodway. It should be noted here that this is only one state's definition of these terms; however, the differences are indeed universal. In Illinois, the local governments are responsible for regulating the flood fringe areas. These regulations generally only relate to the storage within the area and therefore are less restrictive than the state's floodway rules. Another misunderstood portion of the law deals with changes to the regulatory floodway. No state has the authority to revise a regulatory floodway. Only the Federal Emergency Management Agency (FEMA) can change any hydrologic or hydraulic property relating to a regulatory floodway or floodplain. The state of Illinois now works with FEMA in order to make the application process as simple as possible. When a permit application, where a change to the floodway will occur, is submitted to the state, it will be forwarded to FEMA for


28 Applying Floodplain Regulations to the Real World review. Both the state and FEMA must agree with the application before a state permit or FEMA Letter of Map Revision will be issued. Submitting a permit application to the correct regulator decreases the amount of work the applicant and engineer must do. Technical Details The first detail that needs to be considered when preparing a permit application for any activity within the floodplain is that of appropriate use. The State of Illinois has delineated the appropriate uses that vary from flood control structures to roads and sidewalk. This quickly became a bone of contention with applicants who had submitted applications for activities that were not appropriate and were immediately denied a permit-even though all their hydraulic calculations were correct and the parameters were met. A word of advice to an applicant: before you spend a lot of time and money performing difficult hydraulic calculations, examine the local law to determine if there is a list of appropriate uses, and if one exists, check to be sure the activity is allowed. Once an appropriate use has been defined the following technical aspects should be considered. The hydrologic or hydraulic model used as a baseline will be checked to be sure it is the regulatory model. In Illinois this is the model FEMA used for the National Flood Insurance Maps. This model must be used as a baseline even if it appears to have errors and inconsistencies in it. The errors, if any exist, will first be corrected to obtain an existing condition model. Next, any valley sections or other existing information will be added to the existing condition model to acquire revised existing condition model. Finally, any proposed changes must be added to the proposed condition model. These models will be compared to one another to assure the requirements, or definition of a floodway, are met. A second aspect that will be examined is the method of obtaining conveyance. Transition sections must be used when entering or exiting a constriction in flow. The state of Illinois requires a transition of 1:4 when water is flowing from a narrow section to a wider section, and a transition of 1: 1 when water is flowing from a wide section to a narrow section. This requirement is often misapplied or ignored by the applicant. Another misconception found when reviewing conveyance calculations is assuming there will be conveyance through an on-line storage pond. Storage and conveyance are two separate and specific properties and thus must be treated as such. A point of contention between the state of Illinois and many of the applicants is the state's requirements for the use of Manning's "n" values. The state requires the "n" value be varied between the existing conditions model and the revised conditions model to the point where the water surface elevations at each section within the two models remain the same. This requirement at times results in some grossly inaccurate "n" values and thus a model that, in fact, does not reflect the actual conditions of the stream. This is one point where the regulators should look at their interpretation and perhaps alter it to be more realistic.


Baumann and Textor 29 A final technical point to be considered here is what happens when the modeling has been done and the base flood elevation in the proposed condition model is lower than it was in the revised existing conditions model. The state of lllinois has interpreted this to mean the hydrologic conditions have been changed since the regulatory hydrologic model was done and will require the applicant to revise the hydrology. This is also an element within the review where the applicant and the regulator may have a disagreement. The applicant must spend a considerable amount of time and money revising the hydrology, when possibly all that is being done is installing a culvert, which obviously will not change the hydrology. However, it is not uncommon, especially in a rapidly developing area such as this area in Illinois, that the hydrology has been changed. Often though, the state is requiring a party not responsible for this change to pay for updating the modeling. This is another point that should be re-evaluated by the regulators. Conclusion Only recently have floodplain regulations become specific and rigorous. State and local governments are learning how to formulate laws while applicants for permits are learning how to meet and adhere to them. As the process continues to evolve, in order for the regulators to receive what they are asking for and for applicants to get the desired permit with the minimum amount of work, we must all learn to work together to achieve the desired result of flood protection.


FLOOD HAZARD REGULATIONS IN KING COUNTY, WASHINGTON'S SENSITIVE AREAS ORDINANCE Thomas C. Bean King County Surface Water Management Division Introduction King County has six major rivers and hundreds of tributary streams that total 2,235 miles in length. The floodplains that adjoin these streams represent a significant share of King County's land base. Rapid economic and population growth in the region have increased development pressures on these floodplains. At the same time, four presidentially declared flood disasters in the last five years have made clear the hazards associated with floodplain development. Despite flood hazard regulations that met Federal Emergency Management Agency (FEMA) minimum standards, King County has seen repetitive floods cause loss of human life as well as millions of dollars in public and private property damages. On August 29, 1990, the King County Council adopted a new Sensitive Areas Ordinance (SAO). In addition to increasing protection of wetlands, stream corridors, steep slopes, and other sensitive areas, the new SAO substantially strengthened King County's flood hazard regulations. The SAO requires a Sensitive Area Special Study of every development that contains or is adjacent to any sensitive area. The study must define the extent of the sensitive area as well as any impacts of the proposed development. Floodplains are considered sensitive areas-their definition is required, even for streams not mapped by FEMA. The most significant SAO regulations that exceed FEMA minimum standards are: zero-rise floodway, compensatory storage, migratory river regulations, flood protection elevations, future basin-condition flows, and subdivision restrictions. This paper describes these regulations and the reasons each is necessary. Zero-Rise Floodway Regulation. Defines and regulates dual floodway corridors: a standard one-foot floodway and a larger zero-rise floodway. The one-foot flood way is defined to meet FEMA minimum standards. No flow obstruction is allowed within this corridor. Washington state prohibits construction or substantial improvement of residential structures within the one-foot floodway. The zero-rise floodway includes the one-foot floodway within a broader conveyance corridor. The entire floodplain is included in the zero-rise floodway unless HEC-2 modeling demonstrates no change (0.00 ft.) in either water surface or energy grade due to cumulati ve encroachment outside the corridor. Flow obstructions are not allowed within the zero-rise floodway unless easements are granted by all affected landowners.


Bean 31 Rationale. Development that causes a backwater surcharge (Le., a "rise") will additionally inundate upstream and adjacent properties. This regulation prevents such upstream impacts by requiring that floodplain conveyance functions be identified and preserved. The zero-rise standard avoids setting an arbitrary level to which increased flood damages are considered acceptable. It is analogous to King County's on-site detention standards, which set no arbitrary acceptable peak outflow increase for the design storm event. The dual floodway definition meets Washington state and FEMA requirements while allowing lesser restrictions within the broader zero-rise floodway corridor. SAO allows reconstruction of existing floodplain residences where not prohibited by Washington state. The easement option preserves the ability to locate bridge piers and abutments where not prohibited by FEMA. Compensatory Storage Regulation. Requires replacement of any natural flood storage displaced by project grading. Replacement storage must be at same elevations (equivalent areas within each one-foot elevation contour) and must drain effectively to the channel. Rationale. Displacement of natural floodplain storage can increase downstream peak flow rates. This regulation prevents such downstream impacts by identifying and preserving floodplain storage functions. Even with the zero-rise regulation, significant storage can be eliminated from ineffective conveyance areas. The importance of floodplain storage in ineffective flow areas has been documented for King County's Soos Creek (technique adapted from Carlton, 1989). Hydrologic modeling was used to simulate the downstream effects of floodplain storage losses. The study results (tabulated below) do not include effects of any increased impervious surface that might accompany the modeled storage displacement. They merely represent the effect of storage loss. These figures clearly document that peak flows can be sensitive to even the relatively small storage losses associated with ineffective flow areas not protected by zero-rise regulations. Table 1 documents that downstream peak flows are much more sensitive to encroachment of ineffective flow areas than to encroachment that causes a backwater surcharge. This is largely because, by causing a surcharge, the encroachment acts like a retention pond outlet. It forces additional storage to occur upstream, thus mitigating the storage volume displaced by the encroaching project. Migratory River Regulation Regulation. New structures must be safe from stream bank erosion, including that associated with lateral channel migration. Rationale. Many reaches of King County rivers laterally migrate across large alluvial fans. This migration hazard can exceed all other regulatory flood hazard boundaries,


32 Calibrated Floodplain Conditions Encroached to Zero-Rise Floodway Encroached to One-Foot Floodway King County, Washington's Sensitive Areas Ordinance Table 1 Downstream Peak Flow Sensitivity Storage Volume Eliminated (ac ft) o 7 170 Simulated Peak Flow at Lake Youngs Way (cfs) 10-Yr l00-Yr 81 124 111 160 145 223 including the mapped floodplain. The SAO provides a regulatory framework for managing land use in these hazard areas. Flood Protection Elevation Regulation. Finished floor must be at least one foot above the base flood elevation. Rationale. Safety margin for modeling uncertainties, flows exceeding regulatory frequency, and construction errors. Reduced flood insurance premiums for landowners. Future Basin-Condition Flows Regulation. In basins for which King County has adopted a basin plan, including estimates of future flow frequency, these estimates must be used in defining floodplain, floodway, and all other flood hazard areas. Rationale. King County's Basin Planning Program uses hydrologic models to anticipate the effects of basin development. Once the model is calibrated to existing basin conditions, the input is changed to reflect complete basin buildout under current zoning and other land use controls. Output from this modeling procedure represents our best estimate of future flood hazards. These estimates were already being used for design of King County capital facilities. They are equally valuable for design of other floodplain development proposals.


Bean 33 Subdivision Restrictions Regulation. Subdivisions must demonstrate that, given the above regulations, each proposed lot has a suitable building pad. SAO explicitly requires a 5,OOO-square-foot building pad outside the zero-rise floodway on every proposed lot. Warnings must be shown on plat face, along with flood protection elevation data. Rationale. Prohibits creation of lots that cannot be reasonably used without SAO exemptions. (Note: exemptions explicitly allow for reasonable use of all lots created. before this SAO became law). Conclusion The SAO requires development proposals to determine and respect floodplain functions and hazards. These regulations have been adopted and now apply to development proposals in unincorporated King County. Of the 31 cities that also regulate floodplain development along King County rivers and streams, none has all of these flood hazard regulations. References Carlton, David K., Bruce Barker, Ralph Nelson, and Jeanne Stypula 1989 "Effect of Lost Floodplain Storage on Flood Peaks." In Partner ships: Effective Flood Hazard Management, Proceedings of the Thirteenth Annual Conference of the Association of State Floodplain Managers. Boulder, Colorado: University of Colorado, Institute of Behavioral Science. Environmental Protection Agency 1984 Hydrologic Simulation Program Fortran. Federal Emergency Management Agency 1985 Flood Insurance Study Guidelines and Spedjications for Study Contractors. 1989 Flood Insurance Study for King County, Washington. Hydrology Subcommittee, Water Resources Council 1984 Guidelinesfor Determining Flood Flow Frequency, Bulletin 817B. King County Surface Water Management Division 1989a Flood Hazard Regulations: Discussion of Major Regulatory Changes and Issues. 1989b 1991 Comprehensive Flood Control Management Plan Phase 1 Report. Draft Flood Hazard Reduction Plan.


UPDATE 1991: EMERGENCY MANAGEMENT INSTITUTE NATURAL HAZARDS MITIGATION TRAINING COURSES Dan Bondroff Federal Emergency Management Agency Emergency Management Institute Introduction Breaking the cycle of destruction-reconstruction-destruction again has become the rallying call of natural hazards mitigation advocates everywhere. The Emergency Management Institute (EM!) training courses in mitigation should be viewed as an important link in providing the training necessary for state and local governments to break this cycle. I will cover three areas in this chapter: 1) an overview of EMI, 2) current training courses in natural hazards mitigation, and 3) general information on the registration process. Overview of EMI EMI is located at the Federal Emergency Management Agency's (FEMA) National Emergency Training Center in Emmitsburg, Maryland. It is co-located there with the National Fire Academy. The 107-acre campus is 75 miles north of Washington, D.C. Teaching facilities at EMI include large classroom and lecture areas, an operations area for emergency response and recovery simulations, small group breakout rooms, a student Learning Resource Center, and a new micro computer laboratory. Six residence halls can accommodate 382 students in single occupancy rooms. Current Training Courses The focus of this section will be on mitigation courses of most interest to those attending the ASFPM conference. However, I will touch upon other mitigation training courses offered through EM!. In addition, there are professional develop ment courses offered through the EMI field training program that also might be of interest. The professional development courses are offered at the state level. Natural Hazards Mitigation and Recovery Course. This 411z-day course is designed to teach participants how to develop and implement an effective hazard mitigation program and provides knowledge about the disaster recovery planning process. The first 1 liz days is spent on mitigation, with presentations on conflicts and compatibility in mitigating different hazards, concepts and strategies, planning, and FEMA programs such as the NFIP. The next 1 liz days covers the disaster recovery process, including presentations on the federal programs, longand short-term activities, volunteer agency assistance, and the presidential disaster declaration process. The course also features presentations by various local communities and a field trip. This course is currently offered in residence only and is intended for city, county, and state officials, including upper and middle management staff from departments of


Bondroff 35 planning and zoning, economic development, code enforcement, public works, and emergency management. Course dates for Fiscal Year 1992 are March 16-20, 1992; May 18-22, 1992; and July 20-24, 1992. Community Rating System (CRS) Train-the-Trainer. The CRS was created by the Federal Insurance Administration (PIA) as an integral part of the National Flood Insurance Program (NFIP). The 41h-day course includes sections on how a community applies, the various activities that are creditable, and how to provide technical assistance on the CRS to community officials. There will also be a module on methods and techniques of adult learning. Attendance is open to NFIP state coordinators, local and regional officials, those performing floodplain management services for local governments, and others interested in learning about the CRS in order to provide technical assistance to communities seeking to apply for CRS credits. The course dates at EMI are November 4-8,1991; February 3-7, 1992; May 4-8, 1992; and September 21-25, 1992. Community Floodplain Management Course. The purpose of this 4 1h-day course is to provide an organized training opportunity for local officials responsible for implementing and administering the NFIP. The course focuses on flood loss reduction, including floodplain management and the NFIP. The first pilot offering of this course was held at EMI from June 3-7, 1991. The primary audience is local floodplain management administrators, building inspectors, code enforcementlzoning officers, planners, attorneys, engineers, public works officials, city/county managers, members of planninglzoning/variance/conservation boards, and others responsible for the administration of local floodplain management ordinances. Course dates for Fiscal Year 1992 at EMI are October 21-25, 1991, and March 2-6, 1992. Regional, State, and Local Hazard Mitigation Planning Courses. Offered through the EMI field program, these courses are designed for state or local government officials responsible for writing or implementing Section 409 hazard mitigation plans or predisaster mitigation plans. The courses contain optional modules that can be tailored to meet the specific needs of the state or local audience. Interagency Hazard Mitigation Team Training. This EMI field course provides members of the Interagency Hazard Mitigation Teams with knowledge and skills to perform as effective members of a team. In addition, the course covers concepts and operations. Earthquake courses. Natural hazards mitigation courses offered through EMI in our earthquake series include: Nonstructural Earthquake Hazard Mitigation for Hospitals and Other Health Care Facilities, Earthquake Hazard Mitigation for Utility Lifeline Systems, National Earthquake Hazard Reduction Program (NEHRP): Seismic Building Provisions Course, and Earthquake and Fire Hazards in High-Rise Buildings.


36 Natural Hazards Mitigation Training Courses Professional development courses. As I mentioned earlier, EMI offers professional development courses. The courses listed below are offered in the field by state offices of emergency management in order to make them available to as many professionals as possible. These courses include: Introduction to Emergency Management, Leadership and Influence, Decision Making and Problem Solving, Effective Communications, Creative Financing, and Developing Volunteer Resources. The Registration Process Generally, anyone involved in emergency management can participate in resident (EMI campus) and nonresident (state or local level) training activities of EM!. Applicants must, however, meet the prerequisites specified for the course. Selection may also be based on the following considerations: the impact the applicant will have on emergency preparedness in their work place, the utilization potential of acquired skills, and the distribution that the applicant represents within the total emergency management community. Application for admission to EMI is made by using the General Admission Application Form (FEMA Form 75-5). Forms can be obtained from state or local emergency management offices. An application is also on the back page of each EMI catalog. In the standard enrollment process for EMI resident courses, applications should be coordinated with the local Emergency Program Manager and must be reviewed and approved by: the state emergency management director, the FEMA regional director, and the Admissions Office, National Emergency Training Center. Nonresident training is a nationwide program of instruction offered through FEMA regional offices. The vast majority of this training is conducted in partnership with state emergency management offices. For example, the professional develop ment courses mentioned above would fall into this category. Financial and administrative assistance is provided through FEMA regional offices. For further information on nonresident training (field courses), contact your state emergency management training office or FEMA regional training office. The address and phone number for these offices are located in the back of the EMI catalog.


Bondroff 37 References Federal Emergency Management Agency 1990 1990/91 Catalog of Activities. Washington, D.C.: Emergency Management Institute.


AN ASSESSMENT METHODOLOGY AND SOLUTIONS TO BARRIERS AND CONSTRAINTS TO FLOODPLAIN MANAGEMENTTHE TENNESSEE EXAMPLE George E. Bowen The University of Tennessee Floodplain management constraints and barriers are diverse and complex. A barrier is defined as an obstacle, problem, or reason for inaction in the management of floodplains. Specifically, a barrier is something that restricts or impedes the rational solutions to a problem. In order to understand the constraints and barriers in Tennessee, a Delphi process was undertaken with experts in the state. The experts defined the barriers, and then they rated and prioritized the barriers. They then offered alternative solutions to major barriers. Once the solutions were defined, they then rated the solutions with regard to effectiveness and feasibility. The effectiveness and feasibility scores were added to determine a Policy Value Index (PYI). The categories of barriers and constraints the experts assessed were: 1) legal, 2) political/ governmental, 3) cultural, and 4) economic/financial. The most important barriers and constraints were as follows: Legal 1. Failure of state and local agencies to enforce implementation controls 2. No effective legal sanctions to force local agencies to cooperate, adopt, and implement floodplain programs 3. No legal authority Political/Governmental 1. Lack of commitment from local elected officials 2. Low priority on local political agenda 3. Public participation programs do not generate the support required to make legal changes Cultural 1. No constituency and resentment of enforcement and requirements 2. Strong resistance to government regulation and control 3. Resistance to land use controls 4. Politicians do not implement protection if they think it is unpopular Economic 1. Lack of adequate local matching funds


Bowen 39 2. Economic conflict between goals for flood protection and increased growth and development 3. State inclined not to allocate money for floodplain protection Proposed Solutions Legal barriers. To the problem of failure to enforce implementation controls, five solutions had a PVI of more than 12. The number one solution (15.0) was to develop effective penalties for noncompliance. It was also considered the most effective solution. The second most important solution was to undertake legal actions against violators, with a PVI of 13.5. Tied for third most important, with a 12.75 PVI score, were structuring enforcement measures and obtaining local planning commission approval. To the problem of no effective legal sanctions, nine solutions were proposed. The most effective was to cut government funds that support floodplain development, the most feasible was to make floodplain problems a priority for information and education programs. The four best solutions based on the PVI are: 1) tougher penalties for noncompliance, 2) cut funds that support development of floodplain, 3) elevate floodplain protection as a priority, and 4) education programs. To the problem of no legal authority, eight solutions were proposed. The highest feasibility score was to add a comprehensive floodplain protection law to the Tennessee Water Quality Act. The highest efficiency score was to enact laws that prohibit development in the floodplain. The PVI shows the first priority would be to develop a state plan to be implemented by the governor's office (13.25). Political/governmental barriers. To the problem oflack of commitment, 11 solutions were proposed. Federal penalties were considered most effective and education and workshops were considered the most feasible. The PVI indicates that the top six solutions are: 1) federal penalties, 2) multiobjective management of floodplains, 3) education programs, 4) develop linkages, 5) build local interest and support, and 6) obtain state commitment and develop training workshops. To the problem of low government priority, seven solutions were proposed. Information and education programs were rated as the most feasible solutions, while cutting government funds that support floodplain development was rated the most effective. They were also the highest PVI solutions. To the problem oflack of support to get local laws and ordinances adopted, there were eight proposed solutions. This indicates that something can be done about the lack of support. The top three solutions with a PVI of 13.5 are: 1) use "windows of opportunity, 2) raise the priority of information, and 3) develop education programs and make lots of presentations before groups. Cultural barriers. One problem is that it is hard to perceive environmental hazards caused by non-point sources. For this problem, five solutions were proposed. Two solutions received a PVI of more than 12. The first, develop programs with economic incentives, was also rated the most effective solution. The second solution was to use more forceful building codes. The problem of resistance to government


40 Solutions to Barriers and Constraints to Floodplain Management regulations and control elicited eight solutions. The top solution and one considered most effective was to develop linkages. The second most effective and most feasible solution was to assign priority to information and education programs. Local resistance to land use controls and regulations is another problem that needs to be addressed. The respondents enumerated nine potential solutions. Developing linkages was the highest PYI solution with a 14.5 score. It was also considered to be the most effective solution. The second highest PYI score was 14.0, which pointed out floodplain protection needed to be a higher priority and information and education programs should be used to raise the priority level. Politicians will not implement floodplain protection if they perceive it to be unpopular. Eight solutions were proposed to overcome this barrier. Developing linkages was the top solution and the highest ranked in terms of effectiveness. Second was making floodplain protection a priority for information and education. This is also considered to be the most feasible solution. Economic barriers. Lack of local matching funds is a major barrier to floodplain protection. For this problem the experts developed seven potential solutions. The solution that had the highest PYI, effectiveness score, and feasibility score was elevate floodplain protection management as a priority for information and education programs by a coordinated effort of state, federal, and conservation groups. There is a perceived conflict between protection and increased growth. Nine solutions to this problem were proposed. There were six solutions with a PYI of more than 12. The most important solution, with a PYI of over 15, was to maintain adherence to minimum federal regulations. The second most important solution, with a PYI of 14, was to make information and education programs a priority. This was also rated the most feasible solution. To the problem of the state not allocating money for floodplain protection, eight solutions were proposed. Five of the eight solutions had a PYI of over 12. In first place (13.25) was to raise priority for information and education programs. This was also rated the most feasible solution. In second place was to develop funding packages with a PYI of 12.75. Conclusions and Recommendations The best way to deal with legal barriers is to have effective penalties for noncompliance. Laws should also prohibit development in the floodplains and have rigorous enforcement procedures and administrative structure. It would also be good in Tennessee to have the governor directly involved by having the governor's office implement a state plan that requires local planning commissions to adopt local elements. There is also a need for education and information programs to be developed and implemented. Solving political/government barriers will also require education and information programs especially aimed at developing political input. This also suggests that it is important to take advantage of windows of opportunity by highly publicizing flood problems. It will also be important to develop linkages with other programs and


Bowen 41 agencies, like recreation and open space, and take a mUltiobjective view of floodplains. Penalties are also important, and not funding development in floodplains is essential. A constituency also has to be developed and an active promotion campaign undertaken. Cultural barriers, by their nature, may be underlying reasons why other problems exist. Economic incentives are one way to overcome some of these barriers, but more important in the long run will be changing public attitudes. To do this you must involve people and provide information and education programs. You must also develop linkages with other agencies, programs, and communities. If people's attitudes about regulations and controls were different, then protection and conservation would be easier. Overcoming economic barriers also requires public and political support, which will require information and education programs. Incentives have to be linked to compliance. Federal regulations have to be the minimum requirement and growth that protects environmental factors has to be encouraged. It will also require developing some funding incentive packages that will help change the economic equation. Support and money are essential features of floodplain protection and manage ment. Without public and political support, money is not easy to come by. So it would be prudent to focus on information and education programs and developing citizen and elected offices support. Once support has been mustered, then a good legislative/legal base needs to be addressed. This should be based on plans that were developed through an interactive process. Once support and legislation is in place, the money to establish administration and enforcement mechanisms can be sought. With the right mix of people, information, and legislation, the management of floodplains can be accomplished. Floodplains can then contribute to an enhanced quality of life in a community, rather than be a detracting factor.


MULLET AND BISHOP CREEKS DRAINAGE BASIN IMPROVEMENTS Robert E. Johnson and Daniel Shabledeen Oreiner, Inc. Introduction The three-square-mile Mullet Creek and 1.4-square-mile Bishop Creek drainage basins are located in the northeastern portions of Pinellas County in west-central Florida. Within the past five to 10 years, the basins have become increasingly developed. As a result of the increased development and land use changes, several areas within the basins experience street and house flooding and erosion problems during frequent storm events. These flooding problems can be attributed to limited channel capacity, hydraulically inadequate structures, sedimentation, debris, vegetative obstructions, development in flood-prone areas, and urbanized land uses in areas adjacent to the creek. Erosion problem areas are also prevalent within each basin and are due to development within the floodplain, erodible soils, steep channel side slopes, and excessive flow velocities within the channels. Past efforts to control channel erosion include the use of sandbag and rubble riprap, wooden retaining walls, concrete ditch pavement, and sheet pile retaining walls. The Mullet and Bishop Creeks drainage basin improvements were developed to reduce threatening conditions for residents along the creeks by reducing flooding and erosion problems. Drainage Basin Improvement Plan Development The design criteria for the proposed Mullet and Bishop creeks drainage basin improvement plan included both containing the 25-year storm within the confines of the creek channel or designated floodplain and preventing the flooding of existing buildings during the lOO-year storm event. The proposed improvement plan included data collection, computer analysis, and alternatives evaluation and design. The data collection tasks included acquiring existing reports, historical documentation of past flooding, soils maps, land use maps, and aerial topography; conducting an ecological overview; reviewing current stormwater regulations; performing an existing structure inventory; and developing hydrologic parameters. A major focus of the project involved updating the existing Pinellas County Master Drainage Plans and hydrologic and hydraulic computer models. The U.S. Soil Conservation Service hydrologic computer model, TR-20, was used to model the existing condition hydrology for Mullet and Bishop creeks. Peak discharge values generated from the TR-20 model were input into the U.S. Army Corps of Engineers HEC-2 computer model to simulate the hydraulics of the drainage basins and to determine the flood stages for the design storm event(s). Headlosses through culverts in the HEC-2 model were evaluated using the HYDRP (Hydraulic Rating Program) computer program and entered into the HEC-2 model. Many drainage improvement strategies were considered in the study. Strategy options included development


lohnson and Shabledeen 43 restrictions, wetland creation, interbasin transfer, special drainage criteria, channel maintenance, stormwater injection wells, storm water irrigation, structure improve ments, channel improvements, detention systems, retention systems, new channel construction, pumping, erosion control (structural and nonstructural), and secondary system improvements. From these improvement strategies, three alternatives were developed for each creek system. The alternatives consisted of both structural and nonstructural components. Proposed structural alternatives included 1) the improvement of conveyance at culverts and the main channel, and 2) the provision of upstream storage areas for peak discharge attenuation. Regional stormwater detention areas were considered for each alternative. However, due to the degree of existing development and the topography within the basins, only two locations for these large detention areas were available. These improvements would reduce peak discharges while improving the drainage from areas where water collects and stores. Proposed nonstructural alternatives included development restrictions in flood-prone areas, channel maintenance, and requiring special drainage criteria to be used for new development within the basins. The detailed hydrologiclhydraulic models were then altered to reflect the proposed improvements. Ultimate land use conditions (year 2010) were estimated from the Pinellas County Future Land Use Plan Maps (FLUP). Ultimate condition discharge rates and flood elevations throughout the basins were then predicted for the alternatives. Environmental impacts due to the proposed alternatives were evaluated to assess the feasibility of each improvement measure. The environmental impacts of the project were determined to be minimal. However, regulatory agencies and public perception of the project will still require public education, best management practices (BMPs) wetland enhancement, and wetland creation to improve the likelihood of project implementation. Estimated construction costs were also prepared for each alternative. The construction cost estimate included the costs of the proposed structures, channel improvements, erosion control, right-of-way acquisition, mobilization, utility relocation, maintenance of traffic environmental protection, surveying, testing, and construction inspection. The impacts of each alternative on flood elevations, environmental impacts, construction and right-of-way costs, and utility conflicts were determined. To simplify the evaluation of the alternatives, matrices were developed to provide a comparison of each alternative. A public meeting was held to present the proposed alternatives to the public. A comment box and court reporter were available so citizens could officially record their questions and comments and provide input on the project. Based on the results of the alternatives analysis, recommendations were made regarding the preferred alternatives for Mullet and Bishop creeks. The preferred alternatives included structure replacements at selected locations, channel improve ments, erosion control measures, sediment sumps, implementation of a watershed wide channel maintenance program, development restrictions within the loo-year floodplain, and, to meet regulatory permitting requirements, implementation of


44 Mullet and Bishop Creeks Improvements wetland mitigation measures. Due to the high land acquisition costs and the limited degree of improvement in flood stages, the detention ponds were not included in the preferred alternatives. The project results were summarized in an Alternatives Evaluation Report and Schematic Plans and presented to the local governmental officials who selected the alternatives that would be implemented for design. Design and permitting of the selected drainage basin improvements is under way and is expected to be completed in early 1992. Construction of the proposed improvements will be completed as funds become available.


MULTI-LEVEL GOVERNMENT COOPERATION ON A SMALL SCALE FLOODWAY ACQUISITION PROJECT: STANLEY TOWNSHIP, CASS COUNTY, NORTH DAKOTA Jeff Klein North Dakota State Water Commission Introduction Stanley Township is a rural township in eastern North Dakota bordering the Red River of the North and the state of Minnesota, and lying immediately south of Fargo, the state's largest city and fastest growing area. The township population is approximately 2,000 people, with land use consisting of residential subdivisions and farmland. The township lies within an ancient glacial lakebed, has very level topography, and experiences pressure to build homes along native riverine woodlands. The Forest River Subdivision experienced four flood events during the 1970s. In April 1989, spring snowmelt and rainfall caused the Red River to flood, damaging numerous communities and subdivisions including the Forest River Subdivision of Stanley Township. As a result of this flood event, two presidential flood disasters were declared on each side of the river in eastern North Dakota and western Minnesota. Due to the 1989 flood damages, the township pursued a FEMA Section 1362 acquisition project, which removed two substantially damaged residential structures from the floodway. The project, though modest in scope, was accomplished with the help of numerous political entities. The project entailed the commitment of local officials and the cooperation of various governmental entities in making the acquisition project a reality. Project Features Stanley Township sponsored the project with very limited financial resources to devote to the effort. The township has an annual budget of approximately $40,000, three-quarters of which is spent on road system maintenance and snow removal. With cost estimates as high as $37,800 for demolition and site restoration for the two properties qualifying for acquisition, the township enlisted assistance from state and local entities. Contributions of the entities varied, all playing a vital role in project completion. Participants in the project were FEMA, the staff of U.S. Senator Kent Conrad, Stanley Township, Cass County, the Southeast Cass Water Resource District, the Horace Rural Fire District, the City of Fargo, and the North Dakota State Water Commission. Stanley Township, in sponsoring the project, was not dismayed by the time commitment necessary of its officials nor with costs directly or indirectly related to its undertaking. The township was able to enlist a variety of human and financial resources to compliment FEMA's acquisition assistance. The small scale of the project actually proved beneficial.


46 Multi-Level Government Cooperation FEMA's participation provided the financial cornerstone of the property acquisition effort. FEMA paid $345,000 in flood insurance claims and property acquisition costs to acquire two residences, two accessory buildings, and seven lots and place them in ownership by Stanley Township. The staff of the FEMA regional office packaged the Section 1362 project application, which was approved within 60 days after submission to the national office. The involvement of Senator Conrad's office began following the 1989 flood with response to a damaged property owner's request for assistance. The property owner, after learning of the Section 1362 program at a FEMA Disaster Field Office, contacted the senator for help in selling her property and moving from the floodway. This inquiry proved fruitful-the property owner later became one of the two project recipients. One member of the senator's staff was instrumental in working with the township, arranging meetings, and working behind the scenes in moving the Section 1362 application forward. The senate staffer remained involved with the project from inception to completion. The Cass County Commission offered the assistance of its road department personnel and equipment and the county planning office. The road department was unable to provide help to the project. The county planner assisted directly with advice, application development, and in the site re-use plan. The planner continues to work on postproject development activities involving the acquired land. The county lent its continued support for the project by reserving nine adjacent tax delinquent floodway lots from purchase without township knowledge. The three-member Southeast Cass Water Resource District lent its support to the township by approving $8,600 toward restoring one housing site. As the project evolved, the money was not needed. A rural sewage project under construction in the township and sponsored through the Southeast Cass Water Resource District provided the basis for the environmental impact statement needed by the Section 1362 process. The involvement of the water resource district was also key in obtaining the participation of the State Water Commission. The Horace Rural Fire District provided services by burning debris after the moving and salvage operations were completed on each property site. The burning, done within the basement foundation walls, reduced the volume of debris on site. Volunteer firefighters provided time and equipment to oversee the burning as a training activity. The cost to the township was $216 for the services of the rural fire district. The city of Fargo offered technical expertise through its Engineering Department. Jurisdictionally unable to assist the township, the city had a keen interest in the project because of its location near the city. The Fargo Park District remains very interested in the acquired property with its potential to be developed as a natural area park. The property could eventually become part of the Fargo park system. The North Dakota State Water Commission signed an agreement with the town ship to provide personnel and equipment to reclaim the property sites and restore them to natural condition. A five-person construction crew, with engineering oversight, was provided to help complete this final phase of the project. The crew moved on location after the salvage, moving, and burning operations were complete.


Klein 47 Foundation walls were collapsed and the two sites were leveled and landscaped. The work performed was valued at $9,400 and constituted the Water Commission's contribution to the project. Summary Federal, state, county, city, and township resources removed two substantially and repetitively damaged residential structures from the floodway. Stanley Township was able to leverage its location in a Statistical Metropolitan Area to blend outside assistance sources with those available within the township. The accomplishment of the project is not unique, but contributions by a number of entities made the project possible, which otherwise was beyond the capability of the township. One local official volunteered his time to coordinate the project. The commitment of motivated local officials provided the catalyst in completing this acquisition effort. Small scale projects do contribute to reducing future flood losses, and cooperative teamwork made the project a success. Ironically, both former floodway property owners who sold their properties in the project have purchased replacement residences in identified floodplain areas nearby.


SOUTH PLATTE RIVER CENTRAL PLATTE VALLEY IMPROVEMENTS Brian S. Kolstad, John M. Pflaum, and Richard E. McLaughlin McLaughlin Water Engineers Introduction The South Platte River is the major drainageway through the metropolitan Denver area. This paper describes the planning and design of flood control and recreation improvements along the river through the Central Platte Valley, adjacent to downtown Denver. A two-mile reach of the South Platte River forms the western boundary of the area, extending from the confluence with Cherry Creek upstream to 8th Avenue. Interstate 25 (1-25) is a major north-south thoroughfare and is adjacent to the South Platte River. The highway is on the east side from 8th Avenue to 17th Avenue, where it crosses to the west side of the river. The downstream area is known as Confluence Park, which commemorates the historic founding of the city of Denver. Confluence Park, completed in 1975, was the pioneer project of the Platte River Greenway improvements, and includes an amphitheater, trail, pedestrian bridges, plaza, and a white water boat chute. Subsequent greenway projects extended the concrete bike path, added parks, and removed obstructions to allow recreational boating along the river. Background The discovery of gold in the nearby mountains provided impetus for early development in Denver in the mid-1800s. Denver grew into a key railroad hub, and as the rail yard grew in the Central Platte Valley, the river channel was realigned to the west. As urbanization continued, the capacity of the river was gradually reduced by building encroachment, roadway and utility crossings, and diversion dams. In the Central Platte Valley, the channel begins to overflow at events slightly greater than the 10-year flood. Following a disastrous flood in 1965, the U.S. Army Corps of Engineers (US ACE) constructed Chatfield Dam in the southwest metropolitan area. In July 1977, a second flood-control dam was completed on Bear Creek, a major tributary. These structures provide control of the upstream basin areas; however, the l00-year peak discharge for the river at Denver is 19,400 cfs, due mainly to urbanization below the dams. This flow results in a large, expansive floodplain in the Central Platte Valley. Elitch Gardens is an amusement park that was constructed in Denver in 1891. They have outgrown their current location in West Denver and are planning to relocate. Denver wants to maintain the Elitch Gardens heritage and has encouraged their relocation to the Central Platte Valley.


Kolstad, Pflaum, and Mclaughlin 49 Preliminary Design The preliminary design began in the summer of 1988. The Urban Drainage and Flood Control District and the city and county of Denver (hereafter referred to as Denver) retained McLaughlin Water Engineers (MWE) to investigate alternatives and to develop a preliminary design of the selected alternative. Because of the various interest in the land that is inundated by the floodplain and affected by possible improvements, the district and Denver decided to create an advisory group that would meet periodically to review information, look at alternatives, and generally give input to the study. The group consisted of landowners, development interests, and several departments from Denver. The study reach was defined with the confluence with Cherry Creek as the downstream limit and just south of 8th Avenue as the upstream limit. This reach can be isolated because the floodplain is confined in the channel just upstream and downstream of this reach. The purpose of this study was to investigate and evaluate different alternatives that would substantially reduce or eliminate the floodplain outside of the channel, which currently does not have sufficient capacity to contain the loo-year flow rate of 19,400 cfs. Constrictions include utilities, bridges, buildings, railroad tracks, diversion dams, and other encroachments. The flow currently leaves the channel between 8th and 14th streets and inundates the land and buildings east of the river, and some areas on the west side due to the terrain. East of the river, the overflow continues to the north until it is blocked by the Speer Boulevard embankment and directed back to the channel. The floodplain covers a large portion of the Central Platte Valley. After initial analyses were completed, it was found that the project could be further divided into two reaches. The Lower Reach from Cherry Creek to the crossing of 1-25, and the Upper Reach from 1-25 to 8th Avenue. The different alternatives in one reach did not affect the alternatives in the other reach. Each reach is dominated by a diversion dam. The Lower Reach dam is located just upstream of the confluence with Cherry Creek and diverts water into the Farmers and Gardeners Ditch. The dam, diversion, and ditch system is owned and operated by the Denver Water Department under a senior water right. The Upper Reach dam is located just downstream of 13th Avenue and diverts water into the Public Service Company's (PSCo) Zuni Power Plant for cooling. The initial alternatives were to either widen or lower the channel to gain sufficient flood capacity. The alternatives included removing the dams and replacing them with other means to divert the water, or eliminating the need for diversion. The cost estimates for the lower reach were not decisive for either alternative. An alternative for rebuilding the dam and widening the river up to 1-25 gained the most following. In the Upper Reach, removing the dam and replacing it with a pump system was the most favorable and economical alternative. PSCo is currently exploring a closed system for cooling that would require substantially less river water.


50 Central Platte Valley Improvements The alternatives were complicated and expensive due to the encroachment next to the channel. In order to avoid expensive flood walls, both substantial widening and/or lowering of the channel was required. The Confluence Park area is difficult to analyze using common hydraulic models, but a reasonable estimate was made for planning purposes. A physical model was recommended to be completed during the final design phase of the project. In the Lower Reach, lowering the dam at Confluence Park could be accomplished if the efficiency of the Farmers and Gardeners Ditch system could be improved. Calculations showed that if 300 feet of the conduit was replaced, the dam could be lowered one and one-half feet. The more the dam was lowered, the less fill and floodwalls would be required upstream. In the Upper Reach, the railroad, buildings, utilities, and bridges presented the biggest obstacles. By eliminating the dam near 13th Avenue, the channel could be lowered. This would require some major utility work, but was more economical than removing buildings from either the east or west bank. Some alternatives also affected nearby street grades. A special intake system was proposed upstream of the PSCo plant at 13th Avenue where water could be pumped into a pipe or channel and be used by the PSCo plant for cooling. This system would have the least impact on the PSCo intake pumps and cooling system currently being used. The alternatives for the two reaches were presented in a report and submitted to Denver, the district, and the members of the advisory group. The report compared the alternatives, presented the costs for the different concepts, and explained the advantages and disadvantages of each concept. The selected alternative for the Lower Reach includes rebuilding the Farmers and Gardeners Dam at Confluence Park at a lower elevation, replacing part of the Farmers and Gardeners conduit, lowering and widening the boat chute, lowering the channel from the dam to Speer Boulevard, lowering and widening the channel between Speer and 1-25, lowering the water and gas lines at 7th Street, and improving the storm sewer outlets. Other work includes relocating four of the large PSCo transmission towers on the east bank, and adding a maintenancelrecreation road to the east bank with new landscaping and wetlands planting. In order to increase the useable people space at Confluence Park, a plaza has been added to the plans on the east bank. The selected alternative for the Upper Reach includes removing the existing dam at 13th Avenue and replacing it with a pump station, lowering the channel, lowering a 48-inch water line at 11th Avenue, lowering a 115 KV electrical line downstream of 13th Avenue, and widening the channel to the west between 14th Avenue and 8th A venue. The widening includes removing the railroad track from the west bank south of the spur line near 14th Avenue. The Alternative Evaluation and Preliminary Design Report was completed in December 1989. The report presents the alternatives evaluated, their comparison, including costs, and a detailed description of the selected alternative. As the report was being completed, Denver was in the process of presenting a bond issue to the voters. One item contained in the bond issue was infrastructure work (including flood control improvements) in the Central Platte Valley to facilitate the relocation of Elitch Gardens to the Central Platte Valley.


Kolstad, Pflaum, and McLaughlin 51 Final Design The bond issue was approved and negotiations with Elitch Gardens and the existing landowners continued in a positive manner. After receiving proposals and conducting interviews, Denver and the district selected McLaughlin Water Engineers for the design of the Lower Reach, from the Confluence with Cherry Creek to 1-25. The Preliminary Design Report recommended a field test of the Farmers and Gardeners conduit to determine actual flow conditions versus theoretical calculations. This test showed limited capacity, possibly due to a high roughness coefficient. The line was then surveyed with a TV camera to determine its condition. The results showed the line contained a large amount of debris, root intrusion, and other constraints. It was decided to replace an additional 800 feet of conduit. Because of the complex hydraulics at the dam, diversion, and boat chute, and the cost impact upstream, the decision was made to conduct a physical model study to confirm the 100-year water surface profile previously calculated. Colorado State University Hydraulics Laboratory was retained to build the model and make the flow simulation. The model was constructed with some flexibility to see what impacts the different aspects of the project would have on the water surface elevations. In addition, the boat chute was modeled so as to allow the designer to observe the complex flow and to make changes in the chute design to create interesting white water effects that would not be obtainable by typical analytical methods. Due to the size of the area to be modeled, the scale was chosen to be one inch in the model to 20 feet in the prototype. The area modeled is approximately 1,000 feet along the river and 350 feet wide, resulting in a model size of about 60 feet long by 18 feet wide, including piping and measuring equipment. The model is supplied with up to 10.8 cfs, which represents the l00-year flood event of 19,400 cfs for the actual structure. The model testing assisted with evaluation of effects on the l00-year water surface when sediment is present just upstream of the dam. Since this model was not designed as a sediment transport model, the results are approximate. The model testing was completed in December 1990. The results of the testing were presented in a report published in 1991. The results of the model were then incorporated into the final design so the optimum l00-year surface elevation could be obtained and expensive floodwalls avoided. Other details were incorporated into the design based upon information observed during the model testing. The design project was completed in February 1991 and submitted to the district and Denver for review. The final construction documents will be ready for bidding in July 1991. The city is continuing to work with adjacent property owners for rights-of-way and coordination during construction. It is anticipated that work will begin on the project in the latter part of 1991 and continue into 1993. A 404 permit is currently being processed by USACE, and a CLOMR package is being prepared by MWE for submittal to FEMA. The Elitch Gardens Amusement Park will move to a site adjacent to the South Platte River between Speer Boulevard and 1-25. They plan to begin work in the fall of 1991. Project coordination is needed as the material excavated from the river


52 Central Platte Valley Improvements bottom and east bank will be placed as fill on the adjacent Elitch Gardens property. Along with the substantial improvements at Confluence Park, the river corridor will include extension of the greenway trail and complete landscaping. These river improvements, combined with the impressive redevelopment plans for adjacent properties, make the future of the Central Platte Valley very exciting for Denver.


BASALT RIVER STABILIZATION PLAN Tom Newland Pitkin County, Colorado in cooperation with Sid Fox, Eagle County, Colorado, Mark Chain, Carbondale, Colorado, and Alan Czenkusch, Colorado Division of Wildlife Introduction The development and implementation of flood control measures in rural, mountainous areas has always involved a struggle between the protection of private property and the desire to preserve the natural riparian ecosystem associated with the watercourse. A prime example of a desirable outcome to this struggle between people and the natural environment is found within the Basalt River Stabilization Plan. In the past, the portion of the Roaring Fork River between Wingo Junction and Basalt has been very sensitive to high water flooding, most recently in the springs of 1983 and 1984. The landowners of ranches and other agricultural operations along the river have, in the past, conducted permitted and unpermitted work within the floodplain in an attempt to protect their private property. This work has been conducted in an unorganized, piece-meal fashion, which tended to adversely impact lands above and below stream from the project sites. Background The Basalt stretch of the Roaring Fork River can be classified as having a transitional stream pattern with multiple channels found in short reaches. Islands are formed between the channels, which are usually stable and mayor may not be inundated during significant flooding events. The various projects conducted by ranchers in this area consist primarily of rudimentary bank stabilization and channelization activities that help to protect the problem area, but tend to promote the creation of new channels and severely erode river banks downstream. The meandering channel pattern offers the easiest path for water to flow and natural streams seldom follow a straight course for any substantial distance. When working on the river, property owners failed to consider the whole river system. By attempting to place improvements that straightened the river, these property owners felt they were protecting their properties against future flood damage. However, this work tended to cause further, more extreme damage elsewhere on the river when the watercourse attempted to adjust to the new conditions. In one area, where a property owner placed rudimentary riprap along an outside bank, the increase in water velocity resulted in the formation of two new channels downstream where none existed before. These new channels destroyed a prime agricultural hay field on his neighbor's lands. In another instance, a property owner filled into the floodplain and armored the fill area. This work caused the main channel of the river to relocate itself onto previously undisturbed terrain, severely impacting fragile riparian vegetation.


54 Basalt River Stabilization Plan Since 1984, the local governmental entities in the area (pitkin County, Eagle County, and the Town of Basalt) have attempted to formulate a comprehensive river stabilization program that would address the entire reach of the river and still be sensitive to every property owners needs with regard to flood protection. This "pro active" approach was embraced by both local and state agencies. In 1985, the Division of Wildlife, which has interests in this section of the Roaring Fork River because of its "gold medal" fishery status, joined forces with the local governments to pursue development of a river stabilization plan for the Basalt area. A specialist in the field, Dr. Reichmuth of Geomax, Inc., was brought in to develop a conceptual plan that addressed both the property owners' concerns and the local governments' desires to formulate a comprehensive approach to the problem. The goals of the plan are to reduce, to the greatest extent possible, bank erosion and new channel production brought on by both natural and human-made processes within the watercourse. The plan was to be sensitive to environmental quality and attempt to avoid the use of improved structures such as levees, dikes, or other hardened bank improvements. The plan developed by Dr. Reichmuth employs the use of what are known as "drop structures." These structures can be envisioned as small dams across the river that are made by placing a row of boulders perpendicular to the current. Figure 2 is an illustration of a typical drop structure. This row of boulders creates a drop in the water surface and causes a back-water ponding effect that substantially reduces bank erosion upstream from the structure. In addition, the velocity of the current is dissipated when the water falls over the drop, reducing scouring along the banks downstream. The drop structures are constructed in an inverted "V" pattern with the center of the V pointing upstream. This configuration directs the current into the center of the stream and away from the banks. This V configuration also gives the structure the ability to "steer" the thalweg of the river and direct current flows. The drop structure is, in effect, a check dam that regulates the velocity and direction of TYPICAL DROP STRUCTURE CUT OItTAIL Figure 1. Typical drop structure.


Newland 55 the river current at determined or needed intervals. The Division of Wildlife has supplemented the drop structures with a revegetation overlay aimed at fortifying eroding river banks with cabled tree trunks and aggressive riparian revegetation techniques. Issues The Reichmuth plan was taken by the local entities and "engineered" or other wise fine-tuned into a comprehensive plan for the entire reach of the river in the Basalt area. The final plan, developed by Robillard and Associates of Silverthorne, consists of 13 drop structures placed in the river along a 3.5 mile stretch. The plan, when fully implemented, will fix the river in its existing configuration and will fortify eroding banks with natural revegetation and cabling techniques. The Division of Wildlife is especially interested in the plan as it provides natural ponds and eddies within the river course that are important to the health and promulgation of native trout species. The local entities are satisfied because the plan offers a pro-active, comprehensive approach to the problem. When a landowner now applies for a permit to do river work, the local government has an approved plan with engineered drawings for that landowner to follow. The Reichmuth plan solves most, if not all, of the bank erosion problems associated with the predominately agricultural lands affected. The plan reduces bank erosion by restricting river velocity and directing flows. The plan offers a comprehensive approach to solving the problems of bank erosion and new channel formation that have been particularly troublesome to the ranches surrounding the river. However, it must be stated that the plan does nothing to stop or curtail the broad flooding of lands that occurs naturally within the floodplain. The Reichmuth plan is not a flood control plan. Its emphasis is on bank stabilization with the goal of retaining the river within the existing banks. Riprap and other hardening techniques are not desired. Rather, the plan relies on aggressive revegetation of eroded banks coupled with "soft" bank re-enforcement in the form of cabled tree trunks. The plan utilizes materials on site or readily available near the site. Because of this, placement of the improvements is relatively inexpensive. An average drop structure can be placed for between $3,000 and $7,000. Total engineering and easement costs for the project have not exceeded $20,000. Financial assistance for the entire project was pursued by the local entities through federal "404" funding but could not be approved by the U.S. Army Corps of Engineers. Two of the structures have been placed to date. It is anticipated that all 13 structures will be placed eventually by private landowners.


56 Basalt River Stabilization Plan References Binns, N. Allen 1986 Stabilizing Eroding Stream Banks in Wyoming. Cheyenne, Wyoming: Wyoming Game and Fish Department. Colorado Riparian Association 1990 Keeping the Green Line Green, Proceedings of the Second Annual Convention, November 8-9, 1990, Glenwood Springs, Colorado. Colorado State Soil Conservation Board 1989 Streambank Erosion and Fluvial System Management Symposium, July 31-August 2, 1989, Snowmass Village, Colorado. Hunter, Christopher J. 1991 Better Trout Habitat, A Guide to Stream Restoration and Manage ment. Washington, D.C.: Island Press.


SPEAKING PLAIN ENGLISH H. James Owen Flood Loss Reduction Associates Floodplain managers have demonstrated the capability to assess flood-related problems, plan and design appropriate measures to eliminate or reduce potential losses, and operate such programs after their implementation. The greatest obstacle to improved floodplain management has been the difficulty in persuading the city council, board of supervisors, or the public to implement the needed measure. It is at that three-way interface between the floodplain manager, the political and financial decision makers, and the public that projects and programs most often fail. This problem will become increasingly serious in the future. Trends in cost sharing and the increasing use of measures requiring local operation will make progress even more dependent on floodplain managers' ability to persuade local decision makers and the public to fund and participate in flood loss reduction efforts. Research has been conducted aimed at identifying why floodplain managers have been relatively unsuccessful in the past in persuading financial and policy decision makers to implement recommended measures and what might be done to improve the situation. The research has focused on the success of floodplain managers in communicating information on flood risk. The first steps in the research included a literature review, followed by a collection and evaluation of materials presently used by states and federal agencies to inform people about flood risk. The findings of that part of the research were reported at the Association of State Floodplain Managers annual conference in Nashville in 1989 and can be found in the proceedings of that conference. More recently, the research has focused on a case study in the Sacramento, California, area where a major flood control program is being planned by a consortium made up of the U.S. Army Corps of Engineers (USACE), California Department of Water Resources, and several local governments. The objective of the case study is to identify local decision makers' and opinion makers' level of understanding of flood-related matters and to track any changes in that level of understanding over time. The subjects in the case study are a mix of local and state officials, representatives of environmental and other organizations, and media representatives. Together, they make up an elite set of surrogates for the full complement of opinion makers and decision makers in the Sacramento area. The initial step in the case study was to become familiar with the flood-related situation in the Sacramento area, including of collection and review of materials that had been prepared by USACE, including a reconnaissance report. The next step was to formulate a questionnaire and use it to test the case study participants. As finally developed, the first part of the questionnaire asked participants to indicate whether or not they understood each of 53 terms taken from reports issued as part of the local flood control studies or commonly used in flood-related reports. Overall, the case study participants claimed to know only 57% of the terms. This level of understanding of flood-related terminology was considered surprisingly low


58 Speaking Plain English in view of the participants' education, experience, position, and opportunities to encounter the terminology. The remainder of the questionnaire consisted of multiple choice questions dealing with the definition of key floodplain management terms, basic concepts relating to flooding and floodplain management, and the nature and scope of the local flood problem. All of the multiple choice questions had two parts. The first part asked the participant to select the right answer. The second part asked the participant to indicate their degree of certainty that the answer chosen was the correct one. Table 1 indicates, based on the data collected to date, the approximate number of city council members and county supervisors understanding various terms. Table 2 shows typical scores on the questionnaire for various groups based on limited testing. With respect to the multiple choice questions concerning the meaning of key terms related to floodplain management, the participants averaged 58 % correct answers, with an average 57% certainty that their answers were correct. This level of certainty suggests that many of the answers were guesses and that the participants actually know the meaning of fewer of the key terms than indicated by the 58 % correct answers. Again, this was considered to be a surprisingly low level of knowledge for the participants since the participants were not asked to develop a definition of each term. They were only asked to choose from among several possible answers. On the questions pertaining to general information on floods and flooding, participants scored 60% correct answers, with a degree of certainty of 49%, with one person claiming a degree of certainty of only 3 % for the set of questions. Several of the questions concerned the arrangements for planning and its status. Only about half of the participants knew who was responsible for the local participation in the USACE study or who was putting up the million or so dollars for the nonfederal part of the cost for the USACE feasibility study. Only slightly more than half knew what levels of protection were being considered in the study, even though that was one of the key issues being debated by representatives of the various organizations. The final part of the questionnaire concerned knowledge of the flood risk in the Sacramento area. High scores were expected on these questions because of the amount of pUblicity of flood problems over the past two years, the key role of the participants in dealing with the flood issue and the general consensus of the participants that the flood problem was one of the highest priority issues facing the Sacramento area. In spite of this, only a few of the participants knew the season when floods were most likely, the order of magnitude of development at risk, annual flood losses, or the current level of flood protection. The major and most striking fact about the near flood in 1986 was that the water level came within inches of overtopping levees in the area. That fact had been repeated over and over in videotapes, newspapers, local magazines, slide shows, and so on. Yet two-thirds of the participants could not seem to relate the flood height to


Owen 59 Table 1 Understanding of Flood-Related Terminology by City Council Members and County Supervisors Terms Understood by More than 80% of Participants Feasibility Study Terms Understood by 70-80% of Par ticipants Multi-purpose Acre-foot Watershed Benefit/cost ratio Spillway Terms Understood by 60-70% of Par ticipants Peak flood stage Local benefits Riparian Multi-objective Hydrology Hydrologic model Terms Understood by 50-60% of Par ticipants Riprap Flood insurance rate map Backwater Design flow Channel degradation Fishery releases Terms Understood by 40-50% of Par ticipants Off stream storage In-stream flow requirement Slurry wall Recurrence interval Hydrograph Geotechnical Terms Understood by 3040% of Par ticipants River stages Reconnaissance study Levee crown width Embankment clumping Weir Setback levees First costs Cofferdam Terms Understood by 20-30% of Participants Sunk costs Residual flood losses Minimum pool Flood proofing Distributaries Structural measure Riverine Forebay Terms Understood by 10-20% of Par ticipants Wave run-up Power afterbay Flood pool Nonstructural measure Wing levee Stage-frequency curves Objective outflow Afterbay Terms Understood by Less than 10% of Participants Wind setup Reservoir re-regulation Reservoir re-operation NED Plan


60 Speaking Plain English Table 2 Typical Questionnaire Scores on Flood Terminology and Flood-Related Information Recognition Meaning of General 8.00dof terms, key terms, related info percent percent percent nnderstood correct correct Corps of Eng. Professional Staff1 87 81 68 St. Highway Dept. Engineers 2 68 73 67 Members of CAFCA 67 66 58 City Public Works Staff 53 53 49 County Public Works Staff 40 55 47 West Point Cadets 3 39 61 38 Members, League of Women Voters 34 55 46 City Council Members 38 53 41 Members, Kiwanis Clubs 43 50 41 Firefighters 26 48 38 Red Cross Volunteers 14 52 38 High School Teachers 24 49 30 Police 18 48 31 1. Primarily hydrologists and engineers. 2. Primarily engineers who ordinarily deal with flood problems or perform coordination with state flood control agency. 3. Senior cadets upon completion of three hours of lecture on flood control. the level of protection provided by the levees. For the one-third who got the right answer on this point, their degree of certainty indicated that they were largely guessing. The data and information collected so far in the study have led to a three part hypothesis, namely: Floodplain managers use extensive amounts of jargon in their reports and presentations. Floodplain managers overestimate, by a considerable margin, the level of understanding of words and concepts relating to floodplain management on the part of local officials and the public. Proposals for floodplain management projects and programs often fail because the reports and presentations urging action are laced with jargon and otherwise based on wrong assumptions about level of understanding, to the point that a majority of local officials and the public cannot fully compre hend either the extent of flood risk or the proposed action.


Owen 61 It was recognized that, if true, this hypothesis does two important things. First, it explains to some extent why floodplain managers across the nation have such a generally poor record of persuading local officials and the public to implement floodplain management measures. Second, it points the way to improving implemen tation of floodplain management measures in the United States by smoothing out that interface between the planner and the decision maker where projects and programs are often won or lost. The preliminary work has suggested strongly that the hypothesis is true. However, proving it beyond doubt will require more extensive testing of a variety of local officials and representatives of the public. That broader testing has begun, but has not been completed. As the data base becomes large enough to be statistically significant, it will provide some useful guidance for our efforts to improve our communications with local officials and the public. The analysis of responses to the multiple choice questions will help identify which flood-related information is generally known and which is not, so that our educational efforts can be focused where they are needed most. The data on understanding of flood-related terminology will help ensure that those educational efforts use language that is understood by a large share of the intended audience.


FLOOD HAZARD MITIGATION IN ENGLEWOOD, COLORADO: THE LITTLE DRY CREEK SUCCESS STORY John M. Pflaum McLaughlin Water Engineers Introduction The City of Englewood has historically been vulnerable to major flooding along Little Dry Creek, which bisects the community and flows into the South Platte River. Significant flood events occurred in 1913, 1917, 1935, 1965, and 1973. In the mid-1960s, approximately 3,500 feet of the Little Dry Creek channel was enclosed in a reinforced concrete box culvert as part of the construction of a large shopping mall in downtown Englewood. While successful as a retail center, the mall subsequently affected the city in two ways: 1) the existing, older retail businesses along Broadway (the city's main thoroughfare) experienced declines in sales revenues with competition from the new mall, and 2) the box culvert, particularly the entrance structure, performed poorly during large storms, raising suspicions of inadequate capacity. The threat of even worse flooding in downtown Englewood became a concern as urbanization grew in the 25-square-mile upstream basin. In 1974, Englewood participated with the Urban Drainage and Flood Control District and other communities in the preparation of a Major Drainageway Plan for Little Dry Creek. The study developed several flood control alternatives that combined concrete channelization through Englewood with flood storage at a number of possible dam sites in the upstream basin. Additional studies in 1976 and 1981 modified these alternatives to include new rainfall data and a newly completed flood control dam. These studies also introduced a grass-lined channel concept in response to the community's desire to enhance the creek corridor. A significant finding was that the city would suffer an estimated $14 million in damages from a 100-year flood along Little Dry Creek unless some flood control improvements could be con structed. Concurrently, the city council, working with the Urban Renewal Authority, the school district, and major businesses in the community, developed a plan to revitalize the downtown. Their plan included removing the area from the l00-year floodplain, demolishing some existing businesses and residences, and constructing new commercial development. Funding for such a project would come from four sources: 1) tax-increment bonds, sold by the Urban Renewal Authority, 2) private monies from redevelopers, 3) use tax revenues from the city, and 4) funds from the Urban Drainage and Flood Control District. Planning and Preliminary Design In 1981, McLaughlin Water Engineers (MWE) was hired to be the city's prime consultant on the Little Dry Creek improvement project. MWE was responsible for the hydrologic, hydraulic, and civil engineering aspects of the project and


Pflawn 63 management of subconsultants responsible for structural, mechanical, electrical, and geotechnical engineering, urban design, and landscape architecture. Following confirmation of basin hydrology, MWE conducted detailed analyses to determine the capacity of the existing box culvert. Hydraulic calculations indicated that flow conditions within the conduit were extremely unstable, fluctuating between subcritical and supercritical regimes. At bends in the conduit, supercritical shock waves striking the ceiling created pressure flow conditions upstream, reducing the theoretical capacity of the conduit. To confirm the analyses and proposed designs for an extension of the conduit and improved inlet structure, a model of the conduit was constructed at the Engineering Research Center at Colorado State University, Fort Collins, Colorado. Tests conducted with the model verified the hydraulic calculations and enabled refinement of the designs for conduit improvements to optimize its capacity. Once the culvert capacity had been established at 3,650 cfs, as compared to the lOO-year flow of 4,400 cfs, the next problem was what to do with the excess runoff. A storage facility that would reduce stream peak flow to the capacity of the culvert was needed. An upstream storage site would also reduce the cost of bridge and channel improvements along the 0.7-mile reach upstream of the culvert entrance. The city began negotiations with the Englewood School District to use the land between Englewood High School and the creek, near the upstream city limit. This land was being used as a baseball field and for maintenance storage. If the city used the area as an intermittent detention facility, these other uses would have to be replaced. The district drove a hard bargain, since this land was essentially all that was available unless the city purchased existing private property. The school district agreed to an improved facility that would include two soccer fields and a baseball field, complete with press box and bleacher facilities, plus guarantees of maintenance by the city. Preliminary design for the channel reach included a grass-lined channel with a meandering low-flow channel and combination maintenance road/recreation trail, capacity improvements at roadway crossings, architectural walls in constricted areas, and extensive landscaping. The intent was to create a mUltipurpose greenway corridor through the city. The preliminary design was unanimously approved at a joint session of the city council and the Englewood Urban Renewal Authority. Final Design and Construction Major flood control improvements along Little Dry Creek were designed and constructed in the following phases: 1. A mUltipurpose, offstream storage facility, located at the upstream end of the project, provides 95 acre-feet of flood storage during the l00-year storm and provides athletic fields during dry periods. This storage facility was completed in October 1984 at a cost of $2 million. 2. The second phase included flood control improvements along 3,100 feet of channel between South Clarkson Street and South Broadway. The $2.9


64 Little Dry Creek Success Story million channel and bridge crossing improvements were completed in December 1987. 3. The final phase of improvements in the center of downtown Englewood was completed in November 1988 at a cost of$3.7 million. This project included capacity improvements to the existing box culvert conduit and construction of an onstream lake with water features and urban plazas. Offstream storage facility. Since the proposed elevation of the bottom of the storage facility was at or below groundwater levels, an extensive subdrain system was required to maintain dry conditions for the athletic fields during nonflood periods. The system was designed to convey subsurface flows to the northwest corner of the site, where an underground pump station pumps the flow to the creek. A free-draining sand layer was chosen for the athletic fields to provide good drainage and stability. A layer of sand two feet thick was chosen to provide a good hydraulic connection to the subsurface drainage system. Fortunately, excavation for the storage facility provided enough clean sandy material to form the lower 12-inch layer. Imported clean sand was placed in the upper l2-inch layer. For the athletic field surface, a special sand-grown sod was specified and installed. A combination drop/constriction structure in the creek channel provides the hydraulic control by which the upstream water surface is raised and flow is directed over a nO-foot-long grassed side channel spillway into the storage area. The spillway is designed to operate when channel flows reach the 20-year flood level at 2,900 cfs. Approximately 940 cfs of the total loo-year peak flow of 4,400 cfs would spill into the storage facility. A concrete maintenance path, which also serves as a recreational trail, was designed along the crest of the overflow spillway as a means of providing a fixed control of the spillway crest width and elevation. Storage capacity at loo-year flood level is 95 acre feet. Water contained within the storage area drains to the northwest corner, where it is discharged by gravity via a 36-inch reinforced concrete pipeline. A flap gate at the outlet end of the pipe prevents creek flows from backing into the storage area. Channel improvements between Clarkson Street and Broadway. Prior to the project, Little Dry Creek was a constricted waterway with steep banks and limited capacity. Channel widening options were restricted by building encroachment along the channel corridor. To increase channel capacity, cantilever retaining walls were designed for most applications. However, where nearby buildings or structures prevented standard construction methods, retaining walls comprising drilled caissons with structural concrete facings were designed. All channel walls are constructed with buff-colored concrete with a special textured surface pattern and a sandblasted finish. The improved channel is a two-stage design consisting of a low-flow channel that meanders within a grass-lined major channel, which is designed to convey the 100year discharge of 3,650 cfs. The low-flow channel combines a concrete invert with large boulders placed along each side. The boulders were placed with top surfaces level and as close together as possible. Voids between the boulders were filled with


Pflaum 65 smaller rock. Granular bedding material placed behind the boulders enables the channel sides to act as a reverse filter that drains the soils in the grassed overbanks following large storms, providing a stable surface for mowing and maintenance equipment. A lO-foot-wide concrete maintenance road and recreational path traverses the entire length of the channel. A unique feature of the project is the installation of landscape planting in riprap slopes. Sections of 24-inch-diameter polyethylene pipe were utilized for shrub planting wells, and concrete manhole sections were used for tree wells. The planting wells were placed with the rip rap and then backfilled with a combination of rock and backfill material, enabling the wells to integrate with the surrounding riprap. The spreading shrubs and tree plantings serve to soften the structural appearance of the riprap slopes. Plant species were also selected to work with the hydraulic function of the channel. For example, willows were planted in frequently inundated low areas below bridges and drop structures to help dissipate residual energy. Capacity improvements were accomplished at roadway crossings without replacing existing bridges or culverts, resulting in significant savings. At Broadway, the capacity of the existing bridge was increased by channel improvements and the addition of three lO-by-lO-foot reinforced concrete box culverts. The use of precast concrete box culverts enabled timely completion of this portion of the project with less disruption to traffic. At South Sherman Street, an existing concrete box culvert was improved by adding two additional cast-in-place concrete boxes. The existing debris-strewn channel beneath the U.S. 285 bridge was made more efficient by installation of a concrete-lined channel and concrete slope paving. Downtown improvements. The focal point of the downtown project is an onstream lake created by a self-modulating, 8.5-foot high by 42-foot wide fabridam located at the inlet of the box culvert. The water-inflated fabridam adjusts itself for normal fluctuations in creek flow to maintain a stable lake water surface during nonflood periods. When a major flow occurs, the fabridam will automatically deflate, enabling the concrete-lined lake to function as an efficient flood channel. Controls for the fabridam are housed in a nearby subsurface pump station. These controls include an emergency siphon system, whereby the water inside the dam would be automatically evacuated in the event of a power failure during a major flood. Thus, emergency flood capacity is provided without the presence of an operator. The subsurface pump station also houses pumps that serve various water features. An intake located at the conduit inlet structure supplies lake water to the pump station via a 24-inch ductile iron pipeline. Lake water is pumped to a fountain and rock cascade at the upstream end of the lake to supplement creek flows and provide aeration and recirculation of the impounded water. A second pump supplies mUltiple fountains in the main urban plaza area and a "water curtain" feature at the downtown conduit entrance. A third pump serves a curved and terraced "water wall" where water returns to the lake by a series of small fountains, spouts and troughs. All reinforced concrete walls in the downtown project are faced with sandstone veneer. Formal plazas located over the conduit entrance and at the terraced wall


66 Little Dry Creek Success Story feature exposed aggregate concrete pavement sections framed by interlocking paver bands. Extensive landscaping is also a significant part of the project. Conclusion The Little Dry Creek flood control project has received numerous local and national awards as a project that exemplifies the successful integration of recreation and flood control uses. Benefits to the city are threefold: 1) completion of a major redevelopment and revitalization of the downtown area, 2) flood control for Little Dry Creek, and 3) significant park and recreation improvements.


IMPLEMENTING A NATIONAL FLOOD MITIGATION GRANT PROGRAM UNDER PENDING LEGISLATION Michael Robinson, John Gambel, and A. Todd Davison Federal Insurance Administration Office of Loss Reduction The National Flood Insurance, Mitigation, and Erosion Management Act of 1991 (H.R. 1236) would revise the National Flood Insurance Program (NFIP) by establishing a National Flood Mitigation Fund and a mitigation grant program for states, communities, and individuals. Language in the legislation encourages implementation of the grant program by state agencies. The Federal Insurance Administration (FIA) administers the NFIP and will have overall responsibility for developing requirements and allotting funds for the grant program. The FIA Office of Loss Reduction (OLR) therefore conducted a workshop at the 15th Annual Association of State Floodplain Managers (ASFPM) Conference in Denver, Colorado, to solicit input from state and local floodplain managers on their roles in administering and implementing such a grant program. A summary of the major concerns, ideas, and questions that were received during the workshop is listed below: H.R. 1236 not only requires that mitigation techniques be "technically feasible" and "cost effective," but also that they be "in the best interest of the NFIP." Thus, the issuance of grants to protect structures not insured under the NFIP was a concern and topic of debate. Specifically, what percentage or number, if any, of uninsured buildings should be allowed, considering that mitigation fund monies are derived from NFIP policy holders? Several attendees recommended that critical hazard areas (e.g., floodways, alluvial fans, etc.) receive special distinction and consideration in terms of grant issuance and types of permissible mitigation techniques. The formula for distribution of funds from the mitigation fund was discussed. Should there be an equal allocation geographically (i.e., equivalent amounts for each FEMA region), a concentration of funds in areas experiencing the greatest flood damages, or a combination of both? There was a general concern that states with strong floodplain management programs and a Willingness to administer the grant program did not necessarily have severe flood problems and vice versa. The criteria for appropriating money must be cognizant of this point. The grant program should closely coordinate with FEMA Disaster Assistance Programs (DA) and be consistent with existing federal and state hazard mitigation requirements. There is a need to keep the grant application process streamlined and the application requirements straightforward. Some felt that other federal grant programs are overly cumbersome and unnecessarily slow, thereby discourag-


68 A National Flood Mitigation Grant Program ing program participation. However, program efficiency must be balanced by a necessary level of quality control and provisions that ensure credibility and fiscal prudence. Numerous participants wanted the grant program requirements to be consistent with other existing program requirements, such as the Community Rating System (CRS), so that there would be no duplication of effort by grant applicants. General discussion centered around the technical feasibility of floodproofing measures. It was agreed that certain floodproofing measures have been tested in the lab and field and proven effective; however, the problems of human intervention and maintenance still render floodproofing techniques circum spect under the proposed grant program. Within 18 months after enactment ofH.R. 1236, FIA is required to develop qualifying criteria for states wishing to administer their own grant programs. State representatives in the audience were generally reluctant to suggest specific state qualifying criteria at this time. The issue of environmental assessments in the application phase of the grant program was raised. Project applications probably will not warrant detailed environmental impact statements. However, it was agreed that some type of overview of potential environmental impacts should be included for certain types of projects and should be consistent with state and community flood mitigation plans.


WHAT HAPPENED TO THE ECCLESTON BROOK WATERSHED? TALES OF A FLOODPLAIN MANAGEMENT STUDY Elizabeth A. Rogers USDA Soil Conservation Service Introduction Many towns throughout Connecticut have flooding problems and/or extreme development pressures. They are concerned about the increasing cost of flood damage, the pressure to develop in floodplain areas, and the impacts of additional development on flooding. The state of Connecticut does not have statewide regulations that control stormwater management activities, so towns are in need of technical guidance. In addition to the stormwater management issues, towns have: 1) existing flooding problems, 2) floodplain areas where state regulatory programs or other management programs are appropriate, and 3) other interests for the use of the floodplain, such as water quality and quantity, wetland and watercourse habitats, wildlife, fisheries, safety issues, maintenance liability, and health issues. Hydrologic and hydraulic information, which the Soil Conservation Service provided, was needed by the town for stormwater management issues, the flooding problems, and the floodplain issues. The town's commissions also had an interest in the other uses of the floodplain, so they needed information on natural values. The Floodplain Management Study evaluated the floodwater management problems in the Eccleston Brook Watershed in the town of Croton, Connecticut. The town of Croton, the Connecticut Department of Environmental Protection, and the Soil Conservation Service conducted the management study. Program Delivery It was important to explain the concept and goals of the study in an interesting and easily understood manner to local commission members, decision makers, and interested citizens. To accomplish this goal, a presentation was developed in which a "mystery story" was presented. This story solved a mystery and through visuals explained the scene of the crime, the facts, the players, the detectives, the investigations, and the solution. The story ended with a visual that stated it is important to use a team approach to solve a mystery; in this case it was the teamwork of local government and state and federal agencies. In addition to the presentation, which was short and simple, detailed information was available in handout form to individuals who wanted to learn more about the study. References 1988 Plan of Work, Floodwater Management, Floodplain Management Study. November.


70 What Happened to the Eccleston Brook Watershed? Figure 1. The scene of the crime-Groton, Connecticut. Figure 2. The victims-natural resources and the town of Groton.


Rogers 71 Figure 3. The villain-rapid land use change. Doctor Local .. Doctor State .. .... .. Doctor Federal Figure 2. The detectives.


IF THERE'S WATER IN MY LIVING ROOM, I MUST BE IN A FLOODPLAIN Ronald L. Rossmiller HDR Engineering Introduction Many people are flooded each year in locations that are not recognized to be floodplains and/or are not officially designated as floodplains. The definition of a floodplain may well depend on who is writing the definition and what locations are being discussed. How we perceive and define floodplains has important social, technical, and economic implications for homeowners, consulting engineers, various governmental floodplain managers, the federal flood insurance program, and existing flood insurance studies, maps, and zones. This paper examines these implications and uses a flood study done by the author for the city of Beaumont, Texas, as a case study to illustrate them. Floodplain Definitions Several definitions of a floodplain and a floodway have been written, both as a natural geologic feature and from a regulatory perspective. Riverine flooding is the most common: flood waters overtop streambanks and spread out over adjacent lands to some extent. Coastal flooding also occurs: coastal lands are inundated due to tidal fluctuations, storms, hurricanes, and shoreline retreat. The definition of a t100dway has changed over the years. The Flood Control Act of 1928 authorized the U.S. Army Corps of Engineers (US ACE) to construct a floodway, a normally dry portion of the floodplain used to convey a portion of extreme river discharges. In 1962, an ASCE Task Force defined a floodway primarily as a bypass, flood-relief, or diversion channel, and secondarily as the channel of a river or stream and those portions of the floodplains adjoining the channel that are required to carry and discharge the flood water. In 1975, the Federal Insurance Administration developed a definition of a floodway as the watercourse channel and adjacent land areas that must be reserved to carry the base flood (lOO-year flood) without cumulatively increasing the base flood elevation more than a designated height. Some lay people define a floodplain as that area surrounding a stream that has been or could be inundated by flood waters. Some homeowners define a floodplain much more simply-if there's water in my living room, I must be in a floodplain. Size of Watershed and Stream Location Based on the last definition above, the size of watershed upstream of the flooded residence would appear not to be a factor. Whether the watershed contains several square miles or only a relatively few acres is of no consequence. Also, the location of the residence in relation to the stream or river would appear not to be a factor.


Rossmiller 73 IDdeed, the location of a stream or river anywhere in the vicinity would appear not to be a factor. From the individual homeowner's viewpoint, the only important factor is that there is water in the living room. If one must put this into the context of a floodway and floodplain, then the floodway could comprise the street and front yards and the remainder of the floodplain could comprise the houses and backyards. Case Study-Beaumont, Texas The city of Beaumont, Texas, is home to about 120,000 people in an area of about 80 square miles. It is located about 30 miles north of the Gulf of Mexico in Jefferson County, which is immediately west of the state of Louisiana. Land elevations range from 15 to 42. The terrain is flat and the bayous are subject to tidal action and hurricanes. An old saying states that if it rains 12 inches in Beaumont, the water gets a foot deep. Beaumont was established prior to the Civil War and many old drainageways have been filled in, relocated, and straightened as development occurred. The flooding situation in Beaumont as it exists today is shown in Figure 1. The original town was located adjacent to the Neches River and has since spread to the north, west, and south. Drainage paths within the city consist of open channels, streets, storm sewers, and open ditches adjacent to the streets. Because of the flat slopes, tidal effects, and present channel capacities, storm runoff backs up through the system, resulting in street and home flooding. In some areas of the city, the first floor elevations of the homes are at or below street level. The solid areas shown in Figure 1 depict the 25-year floodplain. The hatched areas denote locations where frequent flooding occurs. The solid circles indicate National Flood Insurance Program flood damage claim locations. The figure indicates that very few of the frequently flooded areas coincide with the 25-year floodplain. The figure also indicates that a small percentage of the flood damage locations coincide with either the frequently flooded areas or the 25-year floodplain. Implications The contents of the above paragraph have far-reaching implications for the property owners, for the flood insurance premiums being paid at the present time, and for the underlying precepts of the FEMA flood insurance program. For local property owners. The situation for the local property owners is either a blessing or a curse, depending on one's point of view. It is a curse because many are being flooded throughout the city on a fairly frequent basis. Some of them have experienced flood waters inside the home three times in the last 10 years. However, since most of these homes are located outside any federally designated floodplain, their owners' flood insurance premiums are low.


74 LEGEND If There's Water in My Living Room BEAUMONT, TEXAS Drainage Problem Areas Figure 1 0''1. 4000 20, 20.2.2,.. 25-Year Floodplain w/"//';;/)/1.o/&ft/////l/\ Area of Frequent Flooding NFIP Flood Damage Claim Location Figure 1. Beaumont, Texas, drainage problem areas.


Rossmiller 75 For flood insurance premiums. The vast majority of all flood insurance claims for damage are located outside of any federally designated floodplains. This provides a windfall to the local homeowners and a cost to the federal government. Other than being flooded on a more frequent basis than they would probably like, the local property owners are getting a distinct break on their flood insurance premiums. They are inside a lOO-year floodplain but pay premiums based on the assumption that they are outside (above) the 5OO-year floodplain. The federal government has placed the city of Beaumont on a list for investiga tion because of the numbers of claims filed during the past 10 years and the numbers of repeated claims on the same property in the last 10 years. One reason is that FEMA has adopted studies that indicate the extent of the loo-year floodplain based on the existing system of bayous in the western part of the city. No official floodplains exist in eastern Beaumont because there are no streams in that part of the city. Any that did exist in the past have been filled in and replaced by streets, homes, and businesses. Currently, the streets are the streams in the eastern part of Beaumont. For the FEMA flood insurance program. Three premises of the original flood insurance program were to set standards, promote wise utilization of our floodplains, and to remove existing flood-prone construction from the nation's floodplains. Beaumont seems to have slipped through the cracks. All of western Beaumont was developed after the passage of the flood insurance program in 1968. The designation of a high lower limit of drainage area to be considered in FEMA studies, the delegation of these smaller areas to local control, and the insistence that a floodway and floodplain must contain a natural channel has left many flood-prone properties outside of a designated floodplain. Summary The Beaumont example indicates the shortcomings of the present federal program to prevent flooding of portions of our cities during rainfall events. As long as all properties within a community that is part of the National Flood Insurance Program continue to be eligible for flood insurance, than it may be time to change the rules as to what constitutes a floodplain and the lower limit on the size of drainage area considered in a flood insurance study. Now may be the time to call for a restudy of many of the flood hazard area delineation (FHAD) maps and flood insurance rate maps (FIRM) currently in effect.

PAGE 100

DEVELOPING A STATE FLOODPLAIN MANAGEMENT PLAN: CONTENT AND PROCESS James M. Wright Tennessee Valley Authority Introduction Development of a state floodplain management plan can be both a challenging task and rewarding accomplishment. Such a plan can determine the needs of the state and set forth strategies, measures, and priorities to meet those needs by basically defining the problem situation/assessment, i.e., what is currently being done in flood damage mitigation and maintenance of floodplain resources and what needs to be done, particularly enhancement of the state partnership role. The plan, based on the above analytical foundation, should have a short-term vision (about three years into the future) that identifies immediate and high-priority problems, proposed solutions, and priorities for action. A longer-term vision should extend to at least the end of the decade. The plan can be used to justify use of available resources and requests for additional resources to the executive and legislative branches and provide a basis for the allocation of those resources to provide the greatest impact in floodplain management statewide. Reasons for Preparing a Plan Several reasons or justifications may cause preparation of a plan. The most compelling is to fulfill a legislative or executive directive, which may set forth the purpose, content, and use of the plan and therefore provide some guidelines for its development. Agency justification may be based on the need to develop a plan to better manage the state's floodplains, which would provide for an action agenda to identify needs and establish priorities. Another end result (and reason) of the planning process is to provide for a single source of information-addressing the state's flood-prone areas (including a floodplain inventory), legislation, strategies, and programs-resulting in a complete reference for the staff, other state and federal agencies, and other users. Purpose, Intent, and Use of the Plan In addition to meeting any special legislative or executive requirements, the plan should be a focal point for floodplain management activities within the state by identifying a state strategy for floodplain management to allow for better use of federal, state, local, and private resources. By identifying the state's floodplain management needs, measures and priorities for meeting those needs can be established. Its purpose, intent, and use can be multidimensional. It can first serve as a guide for managing the state's floodplains. The plan can also serve as a data base by providing an analytical basis for planning and action, both short term and long term. Through the provision of a variety of data and information, the plan can serve as a source of information, guidance, and assistance for a variety of users. The

PAGE 101

Wright 77 plan should also be a vision for the future, spelling out what the state wants to accomplish, by what dates, and by what means. Finally, the plan can serve as a baseline for measuring change in future uses of the floodplain, including impact analyses. Content of the Plan If the plan is mandated by a legislative or executive action, it should address any specified items contained in the directive. Unless otherwise restricted, the plan's content, purpose, intent, and use should be far greater than merely meeting any minimum requirements. It should be the best plan the state is able to develop within its resources. Parts of the plan. As a minimum, the plan should have the following six parts. An introduction outlining its purpose, intent, and use. An inventory of the state'sflood plains, based on what is known, to provide an analytical background and basis for subsequent development of other parts of the plan. An examination of the strategies and tools for floodplain management, including those for flood loss reduction and for maintaining natural floodplain resources. An examination of the existing state and federal legislative framework. A determination of the existing roles and responsibilities for floodplain management that provide a framework to carry out various programs and activities. A strategy for managing the state's riverine and coastal floodplains can be developed from information contained in the previous parts of the plan. Information that should be included. In the inventory: Land areas subject to flooding-riverine and coastal Average annual flood damages Extent of floodplain development and use Future potential/projections Account of all significant historic flood events, including damages and fatalities Compilation of floodplain information and flood insurance studies, watershed studies/analyses, project studies, and other floodplain data Information and inventory of natural resources and resource values within flood-prone areas Designated sensitive and critical areas, wetlands, scenic rivers, coastal barrier islands, rare and endangered species, and historical and archaeological sites Information compiled by counties and communities (contacts and assistance, studies, projects, National Flood Insurance Program, other management measures, damages, fatalities, extent of floodplain development and growth, insurance claims, insurance in effect, enforcement of higher standards, greenbelts, and other resource protection)

PAGE 102

78 Developing a State Floodplain Management Plan In strategies and tools: Three strategies for flood loss reduction: 1) modify susceptibility to flooding and flood-related damages, 2) modify flooding, and 3) modify the impacts of flooding on individuals and the community (each have a variety of tools that should be described) Listing of measures in place, being constructed, installed, or planned Two strategies for maintaining natural floodplain resources: 1) restoration, 2) preservation (again describe tools) Listing of measures in place, being put in effect, or planned. In the legislative framework: Listing, summary, or overview of what are considered to be the most meaningful state and federal legislation, regulations, and executive actions for programs that involve or impact the state's floodplains Copies of legislation, regulations (at least title pages and contents), and executive orders in an appendix Existing roles and responsibilities for floodplain management in the state should provide a description of governmental programs and those of the private sector that have an impact or affect use of its floodplains including: State agencies (typically a score or more) Regional planning agencies and districts Local programs Citizen initiative and the private sector Federal agencies (typically FEMA, USACE, SCS, NOAA, USGS, NPS, EPA, FWL, and FHWA). Based on an analysis of the present situation and enumerated needs derived from that examination and analysis, an overall strategy for floodplain mimagement should achieve several goals, some of which may be addressed and accomplished in a short timeframe, while others may take longer periods of time to effectively implement. Among the goals that should be part of the strategy are: Developing an enhanced state role and responsibility for floodplain manage ment as part of a true federal/state/local partnership for dealing with the state's flood problems and in managing its floodplains Ranking by the state of problem or critical areas (both programmatically and geographically) needing floodplain management assistance, including maximizing the allocation of all resources for resolution of problems and needs Recommending and assisting in carrying out programs/projects for specific areas/watersheds, entailing a much greater state role and involvement

PAGE 103

Wright 79 Developing (as appropriate and as needed) standards, guidelines, and regulations designed to meet the particular needs of the state Encouraging and assisting in the multiobjective planning and management of floodplains, stream corridors, coastal zones, and other critical and sensitive areas for both flood loss reduction and environmental protection An appendix provides a place to add a lot of information to complement the plan, to provide a single source of information, and to validate the plan. Among the information that might be included are legislative acts; agency rules and regulations; inventories of floodplain development, use, and resources; statistics for each flood prone locality; lists of floodplain information studies; lists of activities and projects; maps showing various information; and a bibliography of relevant state and federal publications. Finally, the plan text should include a lot of tables, figures, maps, and sketches to aid in the descriptions, to make the text more readable, and to provide for overall embellishment. The Plan Development Process Involvement of those state and federal agencies having important programs affecting the state's floodplains, special districts, and representative localities is critical in developing the plan. They can provide information and data for inclusion in the plan and to review plan drafts, particularly descriptions of their programs and activities for accuracy and completeness. They can also serve as an advisory committee in formulating, developing, reviewing, and selling the plan. The committee should have several meetings during the process. Information for inclusion in the plan should be formally requested by letters to appropriate federal and state agencies and to others. Among the information requested should be a description of programs that involve use of occupancy of floodplain lands or that may be affected by such use or occupancy by others, citations to legislative acts and authorities, floodplain information and resources, adopted rules and regulations, and a list of studies for possible inclusion in the bibliography. Those developing a state floodplain management plan should not underestimate the resources required. To develop a comprehensive plan that presents the current situation, needs, and benefits of greater state involvement, including an implementing strategy to meet those needs and provide those benefits, will likely require at least an equivalent person-year of effort, depending on skill, experience, and expertise. Up to 18 person-months may be more realistic. It will prove to be time and effort well spent. Distribution and Use of the Plan In keeping with the objective of using the plan as a tool and a strategy for better use of federal, state, local, and private resources, it should have wide distribution and utility. Transmittal from the agency head or equivalent level of state government

PAGE 104

80 Developing a State Floodplain Management Plan should be a goal. The plan will also need to be publicized, discussed, and "sold" by state staff in a series of area meetings throughout the state. The plan will need revisiting and revising on a periodic basis (probably in the range of every three to five years) in response to changing needs and situations. Provisions should be made for this situation, including the identification and documentation of needed changes as experience is gained in the use of the plan. A record of those who have copies of the plan should be maintained to receive future changes. Lessons Worth Sharing One of the most difficult aspects is gathering, compiling, analyzing, and determining what data and other information should be included to form the analytical basis for developing the state floodplain management strategy. There is likely a wealth of data and information available that can be overwhelming but which is probably important for inclusion in a plan that should also serve as a valuable resource document. One of the "real world" problems is allocating the resources needed to develop such a comprehensive document in a timely manner. There is a tendency and temptation to put the plan development process aside to deal with day-to-day problems and needs. Finally, do not underestimate the resources required to develop a plan that meets the state's situation and needs and therefore has a realistic chance of implementation. As stated at the beginning, development of a state floodplain management plan can be both a challenging task and, in the end, a rewarding accomplishment. References Virginia Department of Conservation and Recreation 1990 The Floodplain Management Plan for the Commonwealth oj Virginia. Richmond: Division of Soil and Water Conservation.

PAGE 105

Part Three The National Flood Insurance Program

PAGE 106

This page IS intentionally blank

PAGE 107

ARAPAHOE COUNTY, COLORADO, FLOOD RISK DIRECTORY PROJECT METHODS AND APPLICATIONS Douglas Gore Federal Emergency Management Agency Introduction The Arapahoe County, Colorado, Flood Risk Directory was developed under the direction of FEMA, Region VITI by the U.S. Army Corps of Engineers Omaha District in cooperation with local governments in Arapahoe County, Colorado. The project was funded as a pilot project through the National Flood Insurance Program (NFIP) Community Assistance Program-Federal Agency Support Services Element (CAP-FASSE). This paper reviews the project's background, the methods employed, and potential applications for the flood risk directory. Purpose The Arapahoe County Flood Risk Directory Project was initiated to 1) establish a field method for developing a flood risk directory, 2) identify potential applications for the flood risk directory, and 3) to complement FJA's flood risk assessment activities. Ideally, the results could serve as a base for comparing the accuracy and applications of field and automated methods, particularly for flood hazard areas characterized by narrow riverine floodplains. Background Flood risk directories list the addresses of properties located within identified flood hazard areas and provide information about flood-prone buildings. Depending upon the data collected, these directories can service a variety of applications in support of the National Flood Insurance Program (NFIP) and other flood risk management initiatives. A primary application of flood risk directories is to assist lenders in making determinations about the location of properties relative to Special Flood Hazard Areas (SFHAs) identified by NFIP Flood Insurance Rate Maps (FIRMs). Lenders are required to make these determinations pursuant to the Flood Disaster Protection Act of 1973 (public Law 93-234). The Act mandates the purchase of flood insurance as a condition for obtaining most forms of mortgage financing for flood prone buildings. Flood risk directories can also provide valuable information for local officials and others responsible for floodplain management. The Federal Insurance Administration (FIA) of FEMA is currently pursuing an automated risk assessment project to develop flood risk directories for the NFIP. In brief, this effort involves digitizing FIRM panels and overlaying this data on U.S. Bureau of the Census digital street address range maps. The final product is a community flood risk directory listing address ranges (rather than individual properties) located within SFHAs, street name, community name, NFIP community

PAGE 108

84 Flood FUsk [ijrectory Project number, flood insurance zone, Base Flood Elevations (where available) and an "in" or "out" determination relative to the SFHA. Application software is also available for displaying and accessing this data in a computer map format. Site Developed areas of Arapahoe County, Colorado are located in the southern portions of the Denver Metropolitan Area. The county includes eight communities with identified flood hazard areas. These include Arapahoe County (unincorporated areas) and the cities of Aurora, Cherry Hills Village, Columbine Valley, Englewood, Greenwood Village, Littleton, and Sheridan. The estimated combined popUlations of these communities is 234,000. Arapahoe County, Colorado, was selected for this effort because of 1) the current status of the county's Flood Insurance Study and maps, 2) the manageable size of it's SFHAs for the project, 3) the varying sizes and characteristics of the communities to be surveyed, 4) the general commitment of local communities to floodplain management, 5) Arapahoe County's proximity to various Denver institutions likely to participate in field application testing initiatives, and 6) the expertise of the Omaha District of the U.S. Army Corps of Engineers to develop a Flood FUsk Directory for Arapahoe County. Methods The most important aspect of the project was coordination with local officials. A coordination meeting was held with representatives of the eight jurisdictions involved with the project. Further, an effort was made to contact each community before actual field work began in the respective jurisdictions. Local officials were provided the opportunity to submit comments on the data collected before development of the final product. The Arapahoe County Flood FUsk Directory Project utilizes both a conventional field approach for data collection and computer technology for data entry. Directory information was developed from an automobile "wind shield" reconnaissance survey conducted during July through September 1990. The survey was accomplished by two individuals: one who drove and interpreted the floodplain (using FIRMs) and another who entered the data into a lap-top computer. The involvement of a hydraulic engineer and computer specialist in these roles, although not required, was a positive factor in data collection and entry. The data collected included the property address, community name, zip code, flooding source, FIRM panel number and date, FIRM zone, Base Flood Elevation (estimated using FIRM), estimated elevation difference of the grade adjacent to the structure and the lowest floor, floodway or floodway fringe location, an "in" or "out" floodplain determination, structure type, basement according to the NFIP definition, and comments. The dBASE ill PLUS data base management program was used for data entry and prepared before the field work began. A program feature displaying the previous

PAGE 109

Gore 85 data screen made it possible to create a new record simply by amending the fields where data changed, rather than entering data for all fields. This feature is important since most of the information for adjacent properties such as community name, zip code, FIRM panel number, FIRM date, and FIRM zone, etc., are not likely to change. The project was successful in securing information located on 16 FIRM panels. These included seven FIRM panels for portions of Arapahoe County, Colorado, addressing six jurisdictions in a limited county-wide FIRM format; seven FIRM panels for the city of Aurora; and two FIRM panels for the city of Littleton. In total, 2,693 records were placed into the data base during 18 days of field work. This resulted in an average production rate of 149 records per day. Issues of importance include the need to secure an adequate power source to run the lap-top computer and limiting field work to daylight hours and eight-hour days to assure traffic safety. A few advantages of using this method for flood risk directory development include: 1) individual addresses of properties; 2) a field verification of each property surveyed; 3) simultaneous data collection and entry; 4) ease of updating risk directory information; 5) data applicable to insurance agents, lenders, and floodplain managers; 6) the limited need for technical expertise and equipment. 7) capability to search and sort various data fields. Possible disadvantages of the method include: 1) the nature of field surveying as a labor intensive activity; 2) limited information on properties outside of identified flood hazard areas; 3) The absence of a map display capability. Applications The following are examples of possible applications for the Arapahoe County Flood Risk Directory. 1) Assisting insurance agents,lenders, and local government officials in making property determinations. 2) Enhancing flood insurance marketing efforts. 3) Conducting mortgage portfolios reviews for flood insurance coverage. 4) Verifying lender compliance. 5) Evaluating local floodplain management programs. 6) Identifying possible violations of floodplain management standards.

PAGE 110

86 Flood Risk Directory Project 7) Identifying high risk flood-prone properties and potential high repetitive-loss properties before flood losses occur. 8) Updating NFIP Community Information System statistics. 9) Using the directory as a mailing list to contact the occupants of flood hazard areas. 10) Providing a source of information for property data bases. Conclusion The Arapahoe County Flood Risk Directory Project resulted in the development of a feasible field method for developing a flood risk directory addressing a variety of applications of interest to financial institutions and floodplain managers. Further field testing and comparisons with automated approaches will determine the effectiveness of this approach in meeting the needs of various user groups.

PAGE 111

HOW TO GET FLOOD INSURANCE STUDY DATA AND WHAT TO DO WHEN IT'S VANISHED Maggie Mathis and Lisa Bourget Dewberry & Davis Obtaining FIS Data Flood Insurance Study (FIS) data are currently available for a fee from several different sources, including the Federal Emergency Management Agency's (FEMA) Technical Evaluation Contractors (fECs) and the Engineering Study Data Package Facility (ESDP). For many communities participating in the National Flood Insurance Program (NFIP), hydrologic and/or hydraulic models and topographic data are available. Most FIS data are available upon request, with the exception of preliminary data or data that does not reflect "as-built" conditions. All requests for FIS data must be made in writing to the appropriate FEMA regional office, FEMA headquarters, or the ESDP facility. Data requestors should be as specific as possible about the required data in their correspondence with FEMA or the ESDP facility. This correspondence should also include a statement of agreement to pay for research and retrieval time to obtain the requested data and for duplication and mailing costs. Effective with the new fiscal year (October I, 1991) each TEe will be respon sible for processing requests for FIS data for their respective regions, including the requests currently processed by the ESDP facility. It should be noted that the TEC reselection process has resulted in changes in the TEC areas of responsibility for the new fiscal year and beyond. Ultimately, Dewberry & Davis will be responsible for processing all requests for FIS data in Regions 1 through 5, and Baker Engineering will be responsible for filling requests for data in Regions 6 through 10. However, until the transfer of data between TECs is complete, requests should be directed to the TEC currently responsible for the region in which the data are requested. As is currently the practice, all requests must be made in writing; however, a request may be sent directly to the appropriate TEC for processing. Once the request is received, the TEC will determine if the requested FIS data are available and contact the requestor by telephone to inform himlher of the cost to fill the request. The new processing activity scheduled to take effect in October 1991 will be supported by a centralized fee collection system, which will require that payment be made to the National Flood Insurance Program prior to release of the requeste,] FIS data. A standardized fee schedule for charging requestors for FIS data is currently in place at the ESDP facility and will be implemented by the other two TECs effective with the new fiscal year. FEMA's intentions to increase the fees charged for FIS data will be published in the Federal Register prior to implementation. The average request for FIS data will be processed by the TEC in approximately 10 working days and will cost in the range of $80 to $150. Occasionally, FIS data are unavailable from FEMA. One last resort remains: the published FIS report identifies the agency or firm that developed the data. This information is listed in Section 1.3, "Authority and Acknowledgments."

PAGE 112

88 Obtaining Flood Insurance Study Data Dealing with Missing FIS Data If the FIS data are not available from either FEMA or the original source, then the published FIS and Flood Insurance Rate Map (FIRM) provide useful information that can be used to re-create missing hydrologic and/or hydraulic models. FEMA typically requires this re-creation for requests to revise an FIS and FIRM in order to assure a logical transition between revised and unrevised data. Also, the published data are presumed in most cases to be the best available. This is because studies performed under contract to FEMA are developed under specific standards. Data are reviewed by one of the TECs to NFIP specifications, and the final products developed from the data, the FIS and FIRM, are presented to the community and its residents for further scrutiny. Furthermore, time and money can be saved by gleaning valuable information from the published FIS and FIRM, rather than developing all data from scratch. FJS report. The FIS report (Section 3.1) specifies the hydrologic methodology used. This methodology is typically a hydrologic model, a gauge analysis, or a regression equation. For a revision request, the methodology should not be changed unless an improved methodology is proposed; documentation must also be submitted justifying why the change is an improvement. Usually, peak discharges are provided for various frequencies at specific locations along each stream for riverine flooding sources. Drainage areas are also provided for these locations. An example is given in Table 1 below: Table 1 Summary of Discharges Flooding Source Drainage Area and Location (sq. miles) LITTLE CANEY CREEK At Simpson Road 7.62 At Crighton Road 5.53 Peak Discharges (cfs) 100Year 50-Year l00-Year 435 368 1,372 1,204 1,994 1,759 500-Year 3,944 3,497 These discharges may be used directly or may be adjusted via a ratio to drainage area. The FIS report (Section 3.2) also specifies the hydraulic methodology used. As with hydrologic methodologies, the hydraulic methodology should be retained unless documentation can be submitted justifying why a change would be an improvement. Changing from one hydraulic backwater model to another is not usually encouraged; FEMA should be consulted if such a change seems necessary. Section 3.2 also provides information regarding starting water surfaces and Manning's "n" values for riverine flooding sources. Typically, ranges of Manning's

PAGE 113

Mathis and Bourget 89 "n" values are given for both channel and overbank areas. Any changes to Manning's "n" values must be carefully documented. Quite detailed information is available in PIS report Section 4.2 if a stream has a floodway. The floodway data table provides a listing by lettered cross section of channel distance (measured along the streamline or profile baseline shown on the FIRM), floodway width, floodway velocity, and both unencroached and encroached l00-year water-surface elevations to the nearest 0.1 foot. This information is provided in the standard format given below: Table 2 Standard Floodway Data Table FLOODING SOURCE FLOODWAY BASE FLOOD WATER SURFACE ELEVATION 1 SECTION MEAN WIOTtI (I WITHOUT r I, WITH I CROSS SECTION DISTANCE AREA VELOCITY REGULATORY FLOODWAY FLOODWAY INCREASE (FEET) (SQUARE (FEET PER FEET SECOND) Little Caney Creek A 5,356 1,063 6,162 0.32 118.4 B 5,456 1048/314' 5,200 0.38 118.4 'Feet above confluence with Arlls River 2\.lldthf\lldth within corporate limits ]Elevatlons computed based on backwater effects from the Arlia River (FEET NGVO) 118.4' 119.4' 1.0 118.4 119.3 0.9 When the stream in question forms the corporate limit and the widths provided include width within the corporate limits, the floodway at that cross section can be positioned exactly (to the nearest foot). Section 4.1 specifies the scale and contour interval of the topographic map that was used to plot the flood boundaries shown on the FIRM. References are provided to also give the source, title, and date of the map. Plotted flood profiles included in the FIS report give stream invert information and water-surface elevations for various frequency floods, usually the 10-,50-, 100-, and 500-year floods, for riverine flooding sources. Similar information is available from the FIS report for coastal analyses. Section 3.1 provides the methodology used to determine stillwater elevations. Typical methodologies include tidal gauge analyses and link-node models. The Summary of Stillwater Elevations table gives computed stillwater elevations for specific flooding sources and locations for various recurrence intervals. Section 3.2 provides the methodology used to determine any wave heights or wave runup and also may specify fetch lengths. In addition, this section may reference the topographic maps used in developing the coastal analyses and the source for wind data. Finally, the Transect Descriptions table, in conjunction with the Transect Data table and the Transect Location Map, provide the approximate locations at which wave envelope

PAGE 114

90 Obtaining Flood Insurance Study Data elevations were computed, plus the resulting maximum loo-year wave elevations and range of base flood elevations. FIRM. In addition to the information provided in the FIS report, the FIRM provides two important pieces of information for riverine flooding sources: the locations and alignments of selected cross sections, and the configuration of the floodway. Preparing FIS revisions. The data available from the published FIS report should be supplemented as needed and input into the same hydrologic and/or hydraulic model used to create the original FIS. This model should then be calibrated to the FIS flood profiles to obtain loo-year water-surface elevations within an O.I-foot tolerance, as required by the Conditions and Criteria for Map Revisions. If the FIS flood profile cannot be replicated, any differences should be noted, the procedures used in attempting to replicate the profile should be documented, and the reason why the profile cannot be replicated should be explained. The floodway shown on the map, for most communities participating in the NFIP, has been adopted via ordinances and, thus, has gained regulatory status. This regulatory floodway is intended to remain static. Therefore, floodway widths must be specified at the stations of the regulatory floodway. For lettered cross sections, the width from the floodway data table should be distributed across the floodplain so that the resulting floodway matches the mapped floodway (i.e., the floodway may be shifted more to one side of the stream than the other). For areas between lettered cross sections, floodway widths on each side of the stream should be measured from the map and input into the hydraulic model. If resulting surcharges exceed the allowable maximum, the floodway may need to be widened in order to reduce excessive surcharges. Any proposed change to floodway widths must be coordinated with the community that is enforcing the adopted regulatory floodway prior to submitting the proposed change to FEMA. Note that while topographic mapping may guide the configuration of the floodway, it does not entirely dictate it and will not typically override the regulatory status of an adopted floodway. Obtaining and using FIS data will eliminate consistency problems between revision requests and FIS data. In addition, following expected procedures and utilizing available FIS data will smooth FEMA's processing ofFIS revision requests and thus expedite their conclusion.

PAGE 115

FLOOD RISK DIRECTORY PRODUCTION Myles E. Powers and Vince DiCamillo Greenhorne & O'Mara Introduction Several potential products can be realized through the digital capture of flood risk maps for use with Geographic Information Systems (GISs). The most promising prototype product developed thus far is the Flood Risk Directory (FRID). The FRID is a tabular data base product that identifies flood risks by street address ranges within a countywide geographic area. The design of the FRID is based on the information needs of its primary users, who include insurance agents, bankers, real estate agents, and federal, state, and local government officials. These individuals often need to determine the nature of the flood hazards at a particular street address, and the FRID provides a tool that can assist in this effort. The FRID is produced digitally through a map overlay process. This process combines two layers of geographic data (and related tabular records) to produce a table of address ranges, with each address range associated to an assigned flood risk category. The first layer is the digital Flood Insurance Rate Map (FIRM) that defines areas ofland that are at risk, and the second layer is usually the U.S. Bureau of the Census TIGER street map data. If more current and more reliable street data are available, they may be substituted, depending on copyright and license restrictions. The result is a digital file containing a list of address ranges and their associated flood hazard classifications. The methodology used for the production of the FRID is based on research conducted by the Federal Insurance Administration (FIA) and reported in a document titled Automated Flood Risk Assessment for the National Flood Insurance Program: Options for Implementation (FEMA, 1989). The option selected uses the TIGER street map data as the source of geographically referenced street address information because of their low cost. However, the results of FRID verification tests and field use have demonstrated that the FRID is very reliable. For areas where street address range data are available, the FRID and control data (manually derived flood risk determinations) agree more than 90% of the time. Many state and local government agencies are developing GISs for managing property and planimetric data, and as these data become available, it will be possible to produce a more accurate FRID. Perspective The advancement of GIS technology in the National Flood Insurance Program (NFIP) presents an opportunity to solve some of the new and difficult challenges to the risk assessment responsibilities of FlA. Since 1984, the FIA has engaged in a series of successively more complex studies reviewing the feasibility of applying GIS software to the risk assessment aspects of the NFIP. These studies have shown that FIRM panels can be successfully converted to a digital format. The digital FIRM,

PAGE 116

92 Flood Risk Directory Production or DFIRM, has demonstrated several advantages. One advantage is the ability to develop prototype products such as the FRIO. The FRIO applies geocoding techniques to associate individual street addresses to interpolated locations using address-range data and a road-centerline network. The resulting georeferenced graphic of point features is then overlaid digitally with polygons to determine the flood risk category for each point location. Then the individual records are compressed to produce a table of address ranges with each address range associated to an assigned flood risk category. Product formats include an ASCII fixed-length record file, a dBase ill + file, and a bound hard copy printout. Additional development of the FRIO product is being explored to provide an on line query and map display service. A prototype PC product, AUTOFRID, has shown potential for delivery of FRIO information on an individual county basis. However, the development of a centralized on-line query system should provide lower overall cost and increased accessibility to both graphic and tabular data. Techniques Techniques developed in support of the FIA's efforts are relatively general and can be applied to other geocoding applications. All work was done using the Environmental Systems Research Institute (ESRI) GIS software known as ARCIINFO. The basic overlay function required for FRID production is referred to as IDENTITY. However, the ARC/INFO IDENTITY command was not designed to update or maintain the address-range data contained in a street centerline file such as TIGER. Three methods of working around this problem have been used in the development of prototype FRIDs. The first method maintained address range data in an attribute table, and following an identity operation, an INFO program was run to update the address ranges of arcs that had been split. Although this method was reasonably reliable, several disadvantages existed. For example, the order in which the IDENTITY command divided an individual road centerline was not totally predictable, a problem that resulted in some incorrect address interpolations. Another disadvantage was that alias street names, which are found in the Type 5 TIGER file records, could not be used because there can be only one attribute record associated with each linear feature. Finally, the entire process was designed to run on a countywide area at one time; consequently, the process was extremely long and disk storage requirements were high. As a subcontractor to the firm of Dewberry & Davis, ESRI developed a second method that updates address range data stored in a related address table for arcs that were divided during the identity operation. By dividing the countywide area into units approximately the size of U.S. Geological Survey 7.5-minute quadrangles, this method eliminated the need for lengthy computer runs. The use of a related address table also allowed alias street names to be maintained because multiple records can exist for each linear feature. A third method for improving the reliability of the FRID was recently developed by Greenhorne & O'Mara. In this method (iIIustrated in Figure 1), which uses an

PAGE 117

Powers and DiCamillo INFO 4th generation language (4GL) program, the original coverage address range table, including alias addresses, is expanded into a file of individual site addresses. This file and the original ad dress coverage are provided as input to the ARC/INFO ADDRESSMATCH program to produce point coverage. Individual site locations (points) are placed along the length of the road centerline feature by linear interpolation of the ad dress range along each side. The point coverage is then overlaid with a polygon coverage of flood risk information, and the resulting point attribute table is then compressed into address range data to ultimately produce the FRID. This method is reliable and straightforward, and the programming required is not difficult. Also, keeping in mind that flood risk often changes from one side of a street to the other in low-lying 93 Figure 1. Address coverage overlay. areas because roads are usually placed on an elevated grade, the capability of ADDRESSMATCH to offset the points from the road centerline produces more accurate interpretations of the flood risk at individual address sites. Conclusion Overlay analysis of address coverages with the existing ARC/INFO tools can be used to produce useful products. With respect to the FIA's efforts, the creation of a point coverage using the ADDRESSMATCH command has proved to be the most reliable method of FRID production to date. Some of the difficulties encountered with address coverages are primarily a function of the data model used to represent address data. Research into an alternative to the use of address ranges is warranted by the volume of exceptions encountered in address numbering. One solution may be to create a related file that stores the individual street numbers. Greenhorne & O'Mara will continue to explore this and other related geocoding applications. References Federal Emergency Management Agency, Federal Insurance Administration 1986 Flood Hazard Mapping: Geographic Infonnation System Pilot Project. Washington, D.C.

PAGE 118

94 1989 1990a 1990b Flood Risk Directory Production Automated Flood Risk Assessment for the National Flood Insur ance Program: Options for Implementation. Washington, D.C. Historical Data and Technical Procedures for the Orleans Parish, Louisiana Flood Risk Directory. Washington, D.C. FEMA Digital Line Graph and Flood Risk Directory Production Procedures. Washington, D.C.

PAGE 119

ACQUIRING DATA FROM THE FIA ARCHIVESA STEP-BY -STEP APPROACH Richard A. Wild Michael Baker, Jr., Inc. Introduction When they think of a federal agency and hear the word "archives," many people's minds form daunting images of a cold marble building with huge columns to support it, and a mountain of paperwork and red tape to be filed. However, this is not the case with the Federal Insurance Administration (FIA) archives; they are in active library/storage facilities maintained by the specially trained staffs at FIA's technical evaluation and flood map distribution contractors, and the acquisition procedure is straightforward and simple. The purpose of this paper is to provide some specific information on the data that are available from the FIA archives and to document the procedure that should be followed by individuals or organizations that wish to acquire data from these archives. This has been done by answering the following questions: What data are available and where are they stored? How can requesters decide what data they need? How do requesters acquire the data? Will requesters be charged for the data? Data Available From FIA Archives The FIA archives comprises data that can be divided into three major categories: 1) printed reports and maps, 2) "hard" copies of active and historical technical and administrative support data, and 3) microfiche copies of essential historical technical and administrative support data. The types of data that fall into each category are listed in Table 1. A brief summary of where these data are located is provided in the paragraphs that follow. Two groups-the flood map distribution contractor (FMDC) and the Technical Evaluation Contractors (fECs)-are responsible for maintaining the FIA archives. It is through the combined efforts of these dedicated people that the data generated during the processing of Flood Insurance Studies (FISs), Flood Insurance Restudies (RFISs) Map Revisions, and Map Amendments are maintained. The same staff that is responsible for maintaining the FIA archives also assists in the day-to-day handling of printed reports and maps, and in the processing of FISs, RFISs, Map Revisions, and Map Amendments. Because of their hands-on experience with the data, these people are able to keep track of a wide variety of information and consistently give requesters what they need. The printed copies of the effective FIS reports, Flood Insurance Rate Maps (FIRMs), and Flood Boundary and Floodway Maps (FBFMs) are maintained by the

PAGE 120

Printed Materials Effective FIS Reports Effective FHBM Panels Effective FIRM Panels Effective FBFM Panels Table 1. Data Available From Archives Hard-Copy Data Correspondence/Telephone Records for FISs/RFISs Computer Printouts, Cards, & Diskettes Field Survey Notes Cross Section Data and Plots Hydrologic and Hydraulic (H&H) Computations Hydrology Reports FIS/RFIS Contractor Draft Reports and Work Maps Community-Provided Corporate Limit and Street Maps Topographic Maps Preliminary FIS Reports, FIRMs, & FBFMs Revised Preliminary FIS Reports, FIRMs, & FBFMs Floodplain Information Reports Special Flood Hazard Information Reports Correspondence/Telephone Records for Revisions, Amendments, and Appeals Technical Data for Revisions, Amendments, and Appeals Miscellaneous FIA Publications Artwork/Reproduction Materials for FIS Reports/Graphics/Profiles/FHBMs/FIRMs/FBFMs Printed Copies of Previously Effective FHBMs/FIS Reports/FIRMs/FBFMs Microfiche FHBM-Related Correspondence FHBM-Related Telephone Records Backup Data for FHBMs FIS/RFIS-Related Correspondence FIS/RFIS-Related Telephone Records FIS/RFIS-Related Computer Printouts FIS/RFIS-Related Field Survey Notes Cross Section Data and Plots FIS/RFIS-Related H&H Computations FIS/RFIS-Related Hydrology Reports FIS/RFIS Contractor Draft Reports \0 0'1 .g ::;. a8 g-t:r '"

PAGE 121

Wild 97 FMDC, Computer Data Systems, Inc. (CDSI), under the on-site direction of FIA staff. At the flood map distribution facility, maps and reports for more than 20,000 communities are stored. The original hard copies and microfiche copies of the technical and administrative support data are maintained in climate-controlled facilities by the three TECs-Michael Baker, Jr., Inc. (MBJ); Dewberry & Davis (O&D); and Greenhorne & O'Mara (G&O). The states and territories that are assigned to each TEC are listed below. TEC Assigned States and Territories MBJ AK, AZ, CA, CO, Ill, ID, MT, NV, NO, OR, PT, SD, UT, WA,WY D&D AR, CT, DE, DC, LA, ME, MED, MA, NH, NJ, NM, NY, OK, P A, PR, RI, TX, VT, V A, VI, WV G&O AL, FL, GA, IL, IN, IA, KS, KY, MI, MN, MS, MO, NE, NM, NC, OH, SC, TN, WI How To Determine What Data Are Needed To determine what data are needed, requesters must first determine what type of activity they are involved in. The type of activity that could be performed varies widely; however, for the purposes of this paper, discussion has been limited to 1) contractors preparing FISs or RFISs; and 2) local government agencies, private firms, or individual homeowners preparing support data for Map Revisions, Map Amendments, or Appeals. The amount of data needed by FIS/RFIS contractors will depend on the type of analysis that has been performed previously for the areas to be studied and the study approach(es) agreed to with the regional office staff. Therefore, it is recommended that FIS/RFIS contractors call the appropriate TEC to discuss available data before submitting their formal requests. The amount and type of supporting data needed for Map Revisions, Map Amendments, and Appeals varies significantly; therefore, data requesters should decide what process they are to follow before submitting their formal requests for data. The best guidance for this decision is provided in the January 1990 version of the FIA document entitled Appeals, Revisions, and Amendments to Flood Insurance Maps, A Guide jor Corrununity Officials (commonly referred to as "The Blue Book"). However, additional instructive information can be obtained from the following other documents available from FIA: 1) National Flood Insurance Program and Related Regulations; 2) Questions and Answers on the National Flood Insurance Program; 3) Guide to Flood Insurance Rate Maps; 4) Conditions and Criteriajor Map Revisions; 5) Conditions and Criteriajor Floodway Revisions; 6)

PAGE 122

98 Acquiring Data from the FIA Archives Conditions and Criteria for Letters of Map Amendment (LOMAs); and 7) Conditions and Criteria for Issuing Letters of Map Revision (LOMRs) Based on Fill. How To Acquire Data Requests for copies of printed FIS reports, FIRMs, and FBFMs are processed by the FMDC, CDSI, under the on-site direction of Denver Bowman and Sam Morris of FIA. Requests for the printed copies can be made by calling (800) 638-6620. To ensure a timely response, requesters should be able to provide the following information: 1) the full name of the community, 2) the six-digit community identification (CID) number that appears on the FIS report cover and in the title block of every FIRM and FBFM panel, and 3) the FIRM or FBFM panel number(s). Therefore, it is in the best interests of all involved for requesters to review one of the copies of the FIS report and maps that were provided to community officials. However, if such reviews are not possible and only the community name or CID number are known, the FMDC staff will search their computer files and inform the requesters of what is available. If more than one map panel has been prepared for the community and the requester cannot identify the particular FIRM or FBFM panel(s) needed, the FMDC staff will send the requester a copy of the FIRM or FBFM index. The requester can use the indexes, which provide schematic depictions of the panel configurations, to determine the panel(s) needed. To acquire data that are kept in the FIA archives maintained by the TECs, requesters are to submit written requests to the appropriate TEC or to FIA at: Federal Insurance Administration, Office of Risk Assessment, Risk Studies Division, 500 C Street, SW., Room 418, Washington, DC 20472. To ensure a timely and complete response, requesters should include the following in their request letters: 1) the full name of the community (including county and state), 2) a listing of the flooding source(s) for which data are needed, 3) a listing of specific data needed (see Table 1), and 4) descriptions of the area(s) of interest (i.e., stream segment[s] or portion[s] of community). It is important for requesters to remember that additional, specific details about their request will help the TEC staff provide faster service. Therefore, such extra effort as providing copies of FIRM or FBFM panels with the area(s) of interest highlighted on them will also help to ensure a complete response. Fee-Charge System Fees are charged for printed reports and maps on a per report and per panel basis. Current fees can be obtained from the FMDC staff. For the technical and administrative support data maintained by the TECs, FIA has a separate fee system for requests. Under this system, FIA is partially reimbursed for the funds that are expended by the TECs in filling data requests-in effect, establishing a cost-sharing system with the various requesters. Requesters are only billed for the time spent in performing manual or computer searches for data and for reproduction of those data,

PAGE 123

Wild 99 and they are informed of the required fees before data are mailed. Requesters are not billed for the time spent for coordination of the requests. Conclusion As documented in this paper, the procedure for acquiring data from the FIA archives is simple and straightforward. By following this procedure, requesters will assist the TEe and FMDC staffs in assuring a timely and complete response to their requests.

PAGE 124

This page IS intentionally blank

PAGE 125

Part Four Hydrology and Hydraulics

PAGE 126

This page IS intentionally blank

PAGE 127

MORE ON "BACKWATER" SURFACE PROFILE George R. Alger and Henry S. Santeford Michigan Technological University Introduction At the 1990 Annual ASFPM Conference, the authors presented a paper entitled, "Backwater Profile Computations (Is our Approach in Error?)." That paper summarized a six-year study on the hydraulic effects of an ice cover and also presented preliminary findings from a laboratory study aimed at better defining the behavior of backwater profiles in open channels. Subsequent to that presentation, additional work has been completed utilizing a small laboratory flume (six inches wide by approximately 35 feet long). A small reservoir at the upstream end of the flume served as the inlet to the channel. A second reservoir placed at the lower end of the flume was used to induce the backwater. An adjustable sluice gate located at the outlet from the lower reservoir was used to regulate the depth at the lower end of the channel. Water depths were measured at three-inch intervals along the entire length of the flume, including the end reservoirs. However, the discussion of the profiles as presented below is limited to the channel section, which was of constant width and bottom slope. The experimental procedure allowed for variations in discharge, slope, channel roughness, and degree of backwater (defined as the downstream channel depth divided by the normal depth). The backwater or M1 water surface profile is typically defined as the profile that occurs on a mild slope when the flow is decelerating and the flow depth is greater than both normal and critical depths. This is the type of backwater condition generally found in rivers. Under such conditions, the slope of both the water surface and the energy gradeline will be less than the channel slope. At the limiting case when the channel slope is critical, the profile consists of a point defined by the intersection of uniform flow (i.e., critical slope) with a horizontal surface. Theory would suggest that flattening the channel slope will cause the profile, or transition from uniform to horizontal, to lengthen. Typically, backwater or M1 profiles are thought of as being very long (measured in miles). The experimental data suggests the backwater profile as generally defined can actually take two completely different forms, depending upon conditions imposed by the downstream obstruction inducing the backwater. The first of these occurs when the downstream obstruction is small and uniform flow has a chance to develop in the upper portion of the channel reach. The second case occurs when the downstream obstruction is large. There, the backwater effect extends the full length of the channel reach, producing a drowned or inundated condition along the entire reach. Small Obstructions The observed water surface profile for the entire channel, when there is a small degree of backwater, is similar to what would be predicted from a textbook type analysis. For the upstream portion of channel, which had nearly uniform flow, the

PAGE 128

104 More on "Backwater" water surface profile was a straight line parallel to the channel bottom. The water surface profile for the lower portion, where the backwater existed, was (within our limits of measurement) a horizontal surface. Between the two straight lines was a curved transition concave upward. However, instead of being "very long," it actually was very short, often encompassing only one or two measurement positions. Changing the degree of backwater produced a change in the elevation of the downstream nearly horizontal portion and the positioning, but not the shape of the short transition section. Changes in channel slope produced changes in the upper portion, where uniform flow was observed, but had no effect on the downstream portion, where the depth was controlled by the position of the sluice gate. The scale effects were such that no noticeable change could be detected in the short transition section. As expected, changing the channel roughness produced corresponding changes in normal depth for the upper portion of the channel where uniform flow was observed. However, boundary roughness had no effect on the lower nearly horizontal portion of the profile, which was controlled by the position of the sluice gate. Here again, any changes in the short transition section between the two straight line segments was too small to measure. When increased roughness was introduced only in the downstream backwater portion, which exhibited the nearly horizontal water surface, no change was noted in any portion of the profile. Large Obstruction When a large obstruction was placed at the downstream end, such that a drowned condition was produced throughout the entire channel, a condition similar to the classical "two reservoir problem" common to many textbook discussions of open channel hydraulics was encountered. However, the measured water surface profile had no resemblance to those contained in textbooks (Chow, 1959; Henderson, 1966). Instead of the gradual concave upward curve predicted by textbook analysis, the observed profiles were straight lines extending from the lower to the upper end of the channel. Changes in the position of the downstream sluice gate would propagate changes in the downstream depth and the relative elevation of the water surface in the headbox reservoir, but had little if any effect on the water surface slope. Changes in discharge would produce changes in downstream depth, slope of the water surface, and the elevation of the water surface in the headbox reservoir. However, no correlation could be found between water surface slope and mean velocity as defined by discharge/wetted area. Changes in boundary roughness had no effect whatsoever on the observed profiles.

PAGE 129

Alger and Santeford 105 Discussion The results summarized above have more than just an analytical interest. When the actual measured profiles were compared to those predicted by the standard step method, the actual water surface elevations were always less than predicted values. The magnitude of the difference or error between the measured depth and that predicted by the standard step method depended upon the initial starting conditions and the relative position along the profile. The maximum error will be located at the intersection between a horizontal line at an elevation of the downstream pool and normal depth along the channel. For the laboratory data, this difference typically ranged from 20% to 50%, when the actual value of "n" as determined for uniform flow was used in the standard step method computations. As mentioned previously, increasing the boundary roughness in the downstream pool section had no effect on the observed water surface profile. However, when the corresponding "n" value was used in the step method computations, the computed profile increased in elevation, causing the difference or "error" between predicted and measured values to become even greater. Although the results of this study suggest such a procedure is wrong, this is exactly what is done in practice. For example, consider the case where a bridge crossing produces a backwater during flood conditions. In delineating the floodplain, the engineer typically assumes that since the flow is out of the channel, the value of boundary resistance, "n," used in the computation is increased. For the case where the entire reach was inundated, similar differences were observed between measured and computed water surface elevation. The laboratory data suggest that the water surface profile was independent of boundary roughness and not correlated to mean velocity as defined by discharge divided by wetted area. Yet, the computation scheme used for delineating the floodplain boundaries utilizes both parameters. Although the laboratory data collected to date clearly suggest that our current computation scheme is in error, it has not been possible to determine the correct formulation. The primary difficulty has been one of scale. Since the channel slope must be small in order to maintain subcritical condition, either the depths must be small or the channel must be long. The authors clearly recognize the potential for misinterpretation resulting from the small scale laboratory studies. Efforts are currently under way to move the experiments to a much larger facility. Hopefully, we will be able to report on these expanded tests at the next meeting. References Alger, George R., and Henry S. Sante ford 1990 "Backwater Profile Computations: Is Our Approach in Error?" In Challenges Ahead: Proceedings of the Fourteenth Annual Conference of the Association of State Floodplain Managers, Boulder, Colorado: Institute of Behavioral Science, University of Colorado.

PAGE 130

106 Chow, V.T. 1959 Henderson, F.M. More on "Backwater" Open Channel Hydraulics. New York, McGraw-Hill. 1966 Open Channel Flow. New York, Macmillan Co.

PAGE 131

A COMPARISON OF A GAGED URBAN WATERSHED AND COMPUTER MODELING USING HYMO Clifford E. Anderson Albuquerque Metropolitan Arroyo Flood Control Authority Richard J. Heggen University of New Mexico Introduction Prior to 1981, the U.S. Army Corps of Engineers HEC-l Program incorporating initial abstraction and uniform infiltration was commonly used for major drainage studies in Albuquerque, New Mexico. Since 1981, most major drainage studies have utilized a modified version of the USDA Agricultural Research Service (ARS) HYMO computer program (Williams and Hann, 1973). The Federal Emergency Management Agency floodplain studies in Albuquerque used HYMO. The HYMO program was modified to accommodate analysis of urban storm sewers and streets, and to provide the data storage required to efficiently analyze a major drainage system. The HYMO Program has been used to compute runoff from the Hahn Arroyo, a 4.3-square-mile urban watershed where U.S. Geological Survey (USGS) rainfall and streamflow gage data is available. Observed and computed peak flow rates and volumes have been compared. Background In the late 1970s, the USGS established 10 streamflow gages and 11 recording rain gages for undeveloped and urbanized areas. One additional streamflow gage and six recording rain gages were added in the early 1980s. Four urbanized watersheds are included in the gage network. This cooperative program between the USGS and local government agencies is producing useful rainfall-runoff data from major storms. In 1982, rainfall losses were studied using a lO-square-foot rainfall simulator (Sabol et al., 1982). Tests were conducted at 10 sites representing both natural and developing conditions. The study concluded that rainfall-runoff data did "not follow a constant CN line," the "hydrologic soil group and SCS aids for the selection of CN does not in general indicate the most appropriate CN for tested plots," and the SCS rainfall-runoff equation "application, especially in Albuquerque, is based on assumptions which may be too generalized and may be in error. In 1987, 102 split ring infiltration tests were carried out at 32 sites (Heggen, 1987). Data showed that infiltration depends upon land surface treatment and that infiltration rate over the first 30 minutes can be treated as a constant. Hydrologic soil group was not a strong indicator of rate. The infiltration rate for lawns varied extensively, depending upon the timing of irrigation.

PAGE 132

108 Computer Modeling Using HYMO Procedures Available with HYMO The HYMO program was originally developed to utilize the SCS CN method of computing rainfall losses. A specialized unit hydrograph formulation was developed by ARS to allow adjustment of the unit hydrograph shape. The hydrograph shape is controlled by the ratio of the recession constant (K) and the time to peak (tp). Hydrograph computation otherwise followed SCS unit hydrograph procedures. At the Hahn Arroyo, two alternative options were considered, as follows: 1. S.C.S. Curve Number (CN) Method. This is the procedure described in Volume 4, "Hydrology" of the National Engineering Handbook (U.S. Soil Conservation Service, 1969). A single CN is used for each basin; the SCS standard dimensionless unit hydrograph is applied. In order to utilize this procedure, HYMO was modified to include an option to use the SCS dimensionless unit hydrograph in place of the HYMO unit hydrograph formulation. 2. Initial AbstractionlUniform Infiltration (lA/In!) Method. The HYMO program has been modified to use an initial abstraction loss and uniform infiltration rate as an option to the CN procedures (Anderson, 1990). Based on local testing and analysis, the initial abstraction and uniform infiltration parameters were developed for four land treatment classes, as shown in Table 1. Land Treatment Impervious Compacted earth Irrigated lawns Natural Table 1 Initial Abstraction and Infiltration Rates Initial Abstraction On) 0.10 0.35 0.50 0.65 *From 0 to 3 hours; 0.00 after 6.0 hours Infiltration On/hr) 0.04* 0.83 1.25 1.67 A split hydrograph procedure is used to compute pervious and impervious hydrographs separately. The time to peak of the hydrographs is established for the entire basin. The basin hydrograph is obtained by adding the pervious and impervious portions. The unit hydrograph shape is determined by the Kltp formulation available within HYMO, where: Kltp = a + (b x P 60> for watersheds 0 to 40 acres, and Kltp = c + (d x e(-P60 for watersheds greater than 200 acres. P 60 is the 60 minute rainfall and the coefficients a, b, c, d, and e are empirical functions of initial abstraction, infiltration rate, and P 60 Kltp coefficients and the

PAGE 133

Anderson and Heggen 109 resulting unit hydrographs are different for the impervious and pervious portions of the hydrograph computation. Application of HYMO at a Gaged Watershed The USGS rainfall-runoff data has been evaluated for the Hahn Arroyo watershed in order to determine if observed runoff compares with runoff computed by HYMO. The watershed properties are summarized in Table 2. USGS gaging of the Hahn Arroyo allowed analysis at the main channel and the north and south tributary channels. Between 1979 and 1982, gaging consisted of three streamflow gages (main arroyo and two tributaries) and four rain gages. A rain gage was added in late 1982. In 1984, one of the rain gages and the two tributary streamflow gages were removed, and two rain gages were added. Area (Acres) % of area in single family residential Percent impervious Hydrologic soil group Composite CN Number of sub-basins Table 2 Watershed Properties Hahnkroyo North Trib. 2763 482 70 68 31.5 33.0 B B 82.9 82.8 33 5 South Trib. 1722 67 32.0 B 83.0 20 The rainfall and runoff data was input into HYMO to obtain data for the significant storms in the watershed. For this analysis, storms of significance were determined to be those with a rainfall over 0.4 inches at one rain gage and a peak runoff greater than 250 cfs at the main channel. Storm data with any malfunction of rain gages was not included. A storm that may have produced the peak channel flow (July 9, 1988) could not be included because of a streamflow gage malfunction. A summary of observed storm data is shown in Table 3, and the results of the storm computations are summarized in Table 4. Figures A and B show a comparison of observed and computed peak flow rates and runoff volumes. Evaluation Tables 3 and 4 and Figures A and B reveal that the SCS CN method has poor agreement with observed data. The low runoff from these small storms is consistent with previous studies that indicate the CN rainfall-runoff equation will overestimate infiltration during the early portion of storms. The use of this method for determin-

PAGE 134

110 Number of storms Storms dates Mean rainfall On) Maximum rainfall(in) Maximum peak flow (cfs) Mean peak flow (efs) Mean runoff volume On) For the SCS eN Method: Mean peak flow (efs) Mean runoff volume (in) Standard deviation from observed peak flow (efs) 100-year, 24-hour storm For the !Nlnf method: Mean peak flow (efs) Mean runoff volume (in) Standard deviation from observed peak flow (cfs) 100-year, 24-hour storm Computer Modeling Using HYMO Table 3 Observed Storm Data Hahnkroyo North Trib. 16 6 8/SO -9/88 8/SO -8/82 0.70 1.05 2.87 2.58 1080 439 540 244 0.15 0.26 Table 4 Computed Storm Data Hahnhroyo North Trib. 242 181 0.10 0.28 403 92 3827 1152 564 299 0.20 0.39 316 131 3870 1165 South Trib. 6 10/81 -8/82 0.66 0.99 478 301 0.14 South Trib. 130 0.04 219 1542 394 0.17 225 1516 ing runoff from realistic rainfall distributions does not appear to be appropriate. This is especially true for the Albuquerque area, where total rainfall quantities are low. The WInf Method produces a better fit to the observed peak flow rates and volumes. The correlation of observed and actual rainfall for thunderstorm conditions at the Hahn Arroyo is difficult to establish. The scatter observed in Figures A and B is attributed primarily to this factor. Calibration to a single storm event is not likely to provide an actual representation of basin conditions.

PAGE 135

Anderson and Heggen F"lgureA Peak Flow Comparison Figure B Volume Comparison ... I 8" 111 10000 Qp Observed (cis) o IE "C a E 8 c9 o o 0 o o o 0 o <0 10 Volume Observed (at) Conclusion The WInf method appears to produce superior results over SCS CN procedures when used with a realistic rainfall. Comparisons with additional watersheds and storms may allow refinement of parameters for input into HYMO. Additional data from storms near the loo-year event will be particularly critical. It may take many years before such data is available.

PAGE 136

112 Computer Modeling Using HYMO References Anderson, Clifford E. 1990 AHYMO, Problem-Oriented Computer Language for Hydrologic Modeling, AMAFCA Version, Users Manual (Preliminary). Albu querque, New Mexico: Albuquerque Metropolitan Arroyo Flood Control Authority. Heggen, Richard I. 1987 Split Ring Infiltration Basic Data Collection and Interpretation. Report No. PDS 1101210. Albuquerque, New Mexico: Bureau of Engineering Research, University of New Mexico. Sabol, G.V., T.I. Ward, and A.D. Seiger 1982 Phase II, Rainfall Infiltration of Selected Soils in the Albuquerque Drainage Area. Las Cruces, New Mexico: Civil Engineering Department, New Mexico State University. U.S. Soil Conservation Service 1969 "Hydrology." In National Engineering Handbook, Section 4. Washington D.C.: U.S. Department of Agriculture. Williams, l.R. and R.W. Hann 1973 HYMO: Problem-Oriented Computer Language for Hydrologic Modeling, Users Manual. ARS-S-9. Riesel, Texas: Agricultural Research Service, USDA.

PAGE 137

TWO-DIMENSIONAL MODELING IN AN NFIP COMMUNITY: EXPERIENCES AND LESSONS LEARNED Lawrence P. Basich Federal Emergency Management Agency Karl L. Krcma Ogden Beeman and Associates Introduction With over 17,000 communities in the National Flood Insurance Program (NFIP), it is apparent that many of those communities do not have ideal HEC2-type streams-one dimensional, predominantly prismatic streams with a mild stream slope. In actuality, few large watercourses can be modeled using HEC2 and maintain accuracy of plus or minus 0.5 feet across the floodplain. In most cases, however, the intent of the NFIP-to map the nation's floodplains in a cost-effective manner, provide an acceptable degree of technical accuracy, and protect the federal investment from catastrophic flood losses-is met using the U.S. Army Corps of Engineers mainstay, backwater program. There are cases when it is not appropriate to use the standard, but to explore other mechanisms. Background: The Problem In 1984, the FEMA regional office began an investigation into what appeared to be a significant violation of NFIP and local regulations in unincorporated Lane County, Oregon, just east of Springfield, along the McKenzie River. The apparent violation was, in essence, various earth moving activities within a mile-long established island that was totally within the regulatory floodway. The long-term plans were not perfectly clear, but it was apparent that if the immediate issue of the earth moving activities was not addressed, the long-term plans would continue, regardless of what they were. There were several problems at the outset that made a complicated problem even more complicated: since the community was not in the regular phase of NFIP, a preliminary floodway was the enforcement tool. There was insufficient hard, measurable documentation on the exact violation activities; preliminary investigations into the original modeling showed crude modeling techniques; and stereo plotted mapping sections were not within tolerable limits (actually, up to seven feet off). In the interim, the county issued a stop work order to the property owner. While a decision regarding compliance was being made, the property owner hired a consultant who showed that the developments were in compliance. Several meetings were held among the county, the property owner's consultant, and the regional office to discuss this issue. The only thing that was mutually agreed upon was that the original FIS model did not accurately depict the effective flow patterns of this portion of the McKenzie River.

PAGE 138

114 Two Dimensional Modeling A review of the five-mile reach just downstream of this island revealed the following inaccuracies, or areas the HEC2 model could not adequately handle: there were no less than 13 separate split flows areas; there was a six foot discrepancy in the far right overbank in the FIS BFEs and the recorded high water mark for a near loo-year flood event; a privately funded, detailed hydraulic analysis showed the increased flood hazards as a result of development in one of the splits. This study had to be reconciled; the McKenzie River Bridge, located a mile downstream of the island in question, constricts flow from a 4,OOO-foot floodplain to a 4OO-foot opening at a skewed angle, with road overtopping. The key to an accurate study was the correct depiction of the flow characteristics of the area around the bridge. The Solution With the stop-work order issued, NFIP probation pending, and the afore mentioned problems to consider, a determination to go one step above standard operating procedure was made. First, FEMA decided to restudy the area in question. Since this area was going to be a restudy, we had five-foot contour interval maps of the river available from the original study, and two-foot contour interval data from the county. The regional office learned of the availability of the two-dimensional model that was developed by the USGS for the FHW A, FESWMS-2DH, in a PC version. The developers of the model, Dave Froehlich and Jonathan Lee of USGS, were contacted regarding its applicability. The background manuals and details of the program were reviewed by the regional office, and without too much further delay, it was decided that the factors weighing this case were too preponderant to use anything but a model with the capability, utility, and accuracy ofFESWMS. The FIS restudy contract was signed in September 1989. The Model The computer model is a finite element surface water modeling system for two dimensional flow in the horizontal plane. FESWMS uses the Galerkin finite element method to solve the vertically integrated equations of motion and continuity, to obtain depth averages, velocities, and flow depths. FESWMS-2DH is a modular set of computer programs developed to simulate surface water flow. The programs that comprise the modeling system have been designed specifically to analyze flow at bridge crossings where complicated hydraulic conditions exist, although the programs can be used to model many other types of surface water flows. Three separate, but interrelated, programs form the core of the modeling system: 1) The Data Input Module (DINMOD), 2) The Depth-Averaged Flow Module (FLOMOD), and 3) the Analysis of Output Module (ANOMOD).

PAGE 139

Basich and Krcma 115 Topographic data used to create the finite element grid for the McKenzie River project were compiled from a number of sources. For this analysis, a total of 1,760 elements and approximately 6,000 node points were defined in the grid. Calibration consisted of trial and error development of Manning coefficients, boundary conditions, and element configurations to approximately match measured high water marks. Results and Status of the Study The purpose of the study was to show the correct solution to the two dimensional flow problem around the island in question and to demonstrate to our satisfaction the results of the development. In December of 1990, those results were quantified and a meeting was held between the regional office and the county staff. Results clearly indicated that the moving of materials on the island caused the lOO-year water surface elevations to be increased in violation of the NFIP regulations and the local government ordinance. The county was asked to present the findings to the property owner and to seek compensation. In March of 1991, the county did so by letter and notified our office of said attempt. Since that time it has rested in the regional office for action. The regional response has not been formulated at the time of this writing, but it is expected to contain language that will inform the county of our intention to go on with the processing of the study as the conditions exist. The only serious delay encountered with FESWMS-2DH was solving the convergence problem. This problem stemmed from the program's iterative process of turning elements on and off when it saw an element being wet. This was solved by selecting which elements could carry flow from preliminary runs. The major drawback to using the model for NFIP purposes is that it cannot compute a floodway automatically like HEC-2 can. The regional engineer and the consultant decided that the only practicable solution to this problem was to select the initial floodway limits set forth in the Flood Insurance Study and to turn off the ineffective elements in the floodway fringe. Unfortunately, this effort resulted in floodway surcharges that exceeded seven feet. Further refinement using trial and error proved fruitless and this tactic was abandoned. It is expected that the contractor will calibrate a HEC-2 model using the elevations of the 2-D model, realign the cross sections accordingly, and run method four floodways for the reach. At the time of this writing, we are awaiting the final processing of a contract amendment to continue with the above mentioned course of action. As could be expected, there were unexpected costs in this new undertaking, but they were not totally unrealistic. For estimating purposes, the hydraulic phase of this contract was several times that of a standard backwater analysis because of the intense element generation process, data point scrutiny, and iterations to solve the convergence problems. Without at least two-foot contour information available for element generation, this project could have been cost-prohibitive. However, as technology increases, we anticipate that this form of modeling will supplant the existing techniques.

PAGE 140

116 Two Dimensional Modeling References Froehlich, David C. 1989 FESWMS-2DH. Finite Element Surface-Water System. McLean, Virginia: U.S. Department of Transportation, Federal Highway Administration.

PAGE 141

MODELING FLOODS IN URBAN AREAS G. Braschi, M. Gallati, and L. Natale University of Pavia, Italy Introduction Many towns are subject to the risk of flooding; for many of them it is not easy to foresee a priori how the inundation produced by local river bank overtopping or levee breaking will propagate. In such situations, the simplified or traditional hydraulic methods, as summarized for example in Lee and Essex (1983), for mapping flooded areas are not suitable. According to its topography, an occasionally flooded urban road network can be seen, from the hydraulic point of view, as a set of storage locations connected by reaches of channel carrying flow. In fact the water flowing in the town is driven both by the surface slopes and by the urban structures, such as roads, buildings, and squares, which control the advancing water fronts. Usually the network is not a topologically simply connected one and the flow in each reach can change its direction according to the difference between upstream and downstream water levels. Supercritical or subcritical conditions, or both, can take place in a single channel, as the typical road slopes can range from negligible to very high values. The flow situation is further complicated by the dynamic effects, motion of discontinuities, instabilities, local bottom discontinuities, local sinks, etc. It is too difficult or even impractical to try a detailed simulation of such small scale effects; moreover, a small scale time simulation of the phenomenon is not strictly necessary. Following this idea, a numerical code has been prepared (Braschi, Gallati, and Natale, 1989); its main features are described in the following section. As an application, the 1966 flooding of Florence is presented and discussed. Mathematical Model of the Road Network A detailed analysis of the problem is given in Braschi, Gallati, and Natale (1988), where a critical review of the bibliography about hydraulically similar problems is presented, pointing out analogies and differences. The simplified model of the flood propagation in urban areas should produce a large scale description of the flow patterns. The scale of the description is that of the typical road length that is practically well defined in intensively developed cities. Physical phenomena of smaller scale, say the width of the road, are neglected, since a one-dimensional flow description is adopted. Inertial effects are neglected in the dynamic equations. In the model the storage capacity is concentrated in discrete nodal points, placed at gardens, squares, or the crossing of two or more roads. In principle, the node area can change with the water level: starting from the "pure" road area it is increased to take into account the storage in the surrounding area where water can fill

PAGE 142

118 Modeling Floods in Urban Areas courtyards, cellars, etc. In the present simulation, nevertheless, an equivalent invariable area is assigned to each node according to the following rule: the portion of soil surface centered in the node is reduced by the coefficient "urban porosity," that is, the ratio of the unconstructed area to the total plan area, as shown in Figure 1. At each node i the continuity equation states: Ai dz/dt = SQe SQo + Qext (1) where Ai is the node storage area; z is the node water level; SQe, SQo are the total discharges entering and leaving through the channels to the ith node; and Qext is any external inflow given to the node. node --tchannel reach Figure 1. Urban porosity. The connection between two nodes is typically schematized by a rectangular channel modeling the flow carried by a road. The channel flow is simulated by neglecting the inertial effects and assigning the channel storage to the upstream and downstream nodes. The discharge is a function either of the upstream and down stream node levels, if the flow is subcritical or it is controlled by the upstream node level, only if the flow is supercritical: in general the discharge Qik carried by the channel connecting the nodes i, k can be written as: Qik = Q(zi,zk) or Qik= Q(zi) (2) According to the upstream head and to the channel slope, subcritical or supercritical flow conditions can be attained. Depending on whether backwater does affect the upstream section or not, either uniform flow or critical discharge in the channel are respectively assumed. The latter situation is verified in rather long sloping channels. When, on the contrary, backwater effects are not negligible, the flow dynamics equation formally integrated over the channel length gives the appropriate form of Equation 2. When Equation 2 is introduced into Equation 1 for every node, a system of N nonlinear differential equations is obtained for the node water levels zi:

PAGE 143

Braschi, Gallati, and Natale 119 Ai dzi/dt -Sk Qik(zi,zk) Qext = 0 (3) The algebraic sum Sk is extended to all the channels converging to the i th node with i=1,2, .. N. The system is completed by the appropriate initial and boundary conditions. The initial state of the system is defined by the N values of the nodal water depths at the starting time. The value of the external inflow Qext is assigned as a function of time at every node. Alternatively, the water level can be specified at the border nodes (typically at the inflow nodes) while flow rating curves for either a critical or a uniform flow condition are suitable at the outflow border nodes. The system has parabolic behavior and must be solved by a numerical technique appropriate to its strong nonlinearity. The solution of this system can be obtained by several techniques: since an iterative procedure is in any case required by the nonlinearity of the problem, an iterative solution of the system with local updating of the coefficients seems more attractive than a direct solution of the linearized system. The strategy adopted by the present model can be classified as nonlinear Gauss Seidel iteration technique. The Flooding of the City of Florence in 1966 The available information on the mechanics of the 1966 flooding of Florence with the maxima of recorded water depths are reported in Principe and Sica (1967). The reconstruction is based on eyewitness reports, photographs, maxima observed on buildings, etc.: no data, except the rough indication of the flooded areas at several times, are available to describe the evolution of the phenomenon or the local time history at any point. The Arno River first overtopped its right bank at about 11:00 p.m. on the 3rd of November in the upstream neighborhood of Rovezzano. The water was generally contained by railway embankment, locally underpassed only through the road tunnels. The front flowed nearly parallel to the river, then it propagated towards downtown. Due to the concatenated backwater effect of several bridges, the right bank of the river close to the area of the city center (the lowest of the surroundings) was also overtopped; the new flood front met the front advancing from upstream at about 9:00a.m. near San Croce Church. The new water front then advanced toward the inner part of the town, submerging the whole center and the downstream part of the town in the first hours of the afternoon. The inundation reached its maximum expansion at about 4:00 p.m., then the water gradually began to flow back downstream toward the river. The Computer Simulation The quality of the computer simulation was verified, comparing the numerical results to the flood marks (principe and Sica, 1967) recorded in 1966 just after

PAGE 144

120 Modeling Floods in Urban Areas recession of the flood wave on the monuments and along the most important roads at the center of the town. 54.---------------------------------------------------------, 52 .s 50 48 46 44 __ __ _L __ __ __ __ __ __ o 400 800 1200 1600 2000 2400 2800 3200 3600 lenghl (m) Figure 2. Comparison of computed and measured water elevations. In 1989 a survey was carried out in order to verify the ground elevation of the streets of the Florence city center. First of all, in order to evaluate how the accuracy of the topographic survey influences the results, two different computer models were set up: the first one referred to the topography reported on the maps issued by the county of Florence, the second one used data measured in a new survey. The numerical results showed the wrong topographic data of the original maps affected the accuracy of the simulation in a substantial way. Using the measured ground elevations, the road network was modeled by two different hydraulic networks: the more detailed one considered 291 reaches and 165 nodes, the other one used a more rough description (223 reaches and 125 nodes). The results obtained by the two different hydraulic schemes were compared: since the results were practically equivalent, in the following numerical experiments the less dense hydraulic network was used. Many different simulations of the inundation were performed, changing the values of the different model parameters (coefficient of roughness and coefficient of porosity). The sensitivity analysis also considered different boundary situations because the inflow hydrographs were not measured: in fact, for this inundation problem only the initial condition (dry land) is known. As for the boundary conditions, no direct measurements are available; only the estimate of maximum volume stored in the city was reported: it amounted to two million cubic meters. After that, 34 computer simulations were carried on. Experiencing different combinations of parameter values, the model simulated the real phenomenon satisfactorily. The maximum values of the computed and measured water elevations are compared in Figure 2; in this figure is shown a section that crosses the entire center city going parallel to the Arno River.

PAGE 145

Braschi. Gallati. and Natale 121 Conclusions The results obtained in this study show that the mathematical model can lead to useful results also in very complicated situations like the inundation of a big city. Moreover, the use of the model needs only a personal computer. The numerical experiments carried on during this study allow the authors to make the following general observations: 1. The topographical description must be sufficiently accurate, even though the hydraulic computations are insensitive to errors of the magnitude of one foot. 2. The computation is rather influenced by the value of the roughness coeffi cient; nevertheless, its determination can be obtained on the basis of the urban characteristics (Lee and Essex, 1983). For the presented example, five values of Manning'S n between .125 and .05 have been assigned to five areas with estimated homogeneous hydraulic resistance. The values of the Manning coefficient to be used for the flow simulation in the roads are rather higher than those characteristic of the channel flow; this may be due also to the fact that an ancient city has been studied. 3. The value of the urban porosity can be easily estimated from the maps, and its importance is quite immaterial. 4. The knowledge of the inflow discharge hydrographs is essential; on the contrary, outflow boundary condition, at least in the examined case, produces a localized control of the water levels. Finally, it is important to recognize that the flooding over urban areas can produce very high water slopes both in the main direction of the flow (Figure 2) and in the transverse one so that simplified, one-dimensional methods can be inadequate for the simulation. Becchi, I., et al. 1988 References Mappa tecnica del comune di Firenze per interventi di Protezione Civile, GNDCI-Prefettura di Firenze. Braschi G., M. Gallati, and L. Natale 1988 La simulazione delle inondazioni in ambiente urbano. Int. Rep. No. 151. Department of Hydraulic and Environmental Engi neering, University of Pavia, Italy. 1989 "Simulation ofa Road Network Flooding." In Proceedings of the Twentieth Annual Pittsburgh Conference on Modeling and Silmulation.

PAGE 146

122 Modeling Floods in Urban Areas Lee, L.T., and T.L. Essex 1983 "Urban Headwater Flooding Damage Potential." Journal oj Hydraulic Engineering, ASeE, 109 (4) April. Principe I., and P. Sica 1967 "L'inondazione di Firenze del 4 Novembre 1966." L 'Universo 2.

PAGE 147

COMBINED GEOMORPmC, HYDRAULIC, AND SEDIMENT TRANSPORT ANALYSES: APPLICATION TO A SEDIMENTATION PROBLEM Karin J. Fischer and Michael D. Harvey Water Engineering and Technology Lyle W. Zevenbergen Love and Associates Introduction Two Yazoo Basin streams, Abiaca Creek and its principal tributary, Coila Creek, are located in Carrol County, Mississippi. They have a combined drainage area of 95 square miles. Abiaca Creek is a tributary of the Yazoo River in a region known locally as the Yazoo Delta. Current problems in the watershed include downstream flooding and sedimentation in a channelized, leveed reach and in a National Wildlife Refuge wetland on the Yazoo Delta. Within the last 150 years the watershed has been subjected to perturbations that include: land use changes (1830-1910), channelization oflower reaches (1920), implementation of flow control (1962-1977), and sand and gravel mining on upland areas adjacent to both Abiaca and Coila Creeks (1960s-present). The objective of this study was to evaluate the geomorphic responses of the channels to these perturbations so that present channel stability could be evaluated and integrated into sediment management strategies. Incised Channel Evolution Channelization of numerous meandering streams of the upper Yazoo Basin in northern Mississippi has resulted in severe channel erosion (Schumm et al., 1984). Observed morphological adjustments of these incised channels reflect adjustments of cross section area and slope that are required to maintain the continuity of the imposed discharge and sediment load. Incised channels naturally evolve from a state of disequilibrium to a new state of dynamic equilibrium (Schumm et al., 1984). Initially, the incised channel is characterized by nickpoints in the channel bed; little to no sediment storage; and a deep, narrow cross section. As the channel banks exceed heights that are gravitationally stable, bank failure occurs and the channel rapidly widens. The degradation migrates upstream, increasing sediment loads and driving deposition in the incised channel downstream. Continued deposition results in development of a channel in dynamic equilibrium. Berms form and become stabilized by perennial vegetation and these protect the bank from subsequent erosion. Channel sinuosity increases, and a new floodplain develops that is at a lower elevation than the bounding terrace that represents the pre-incision floodplain. Geomorphic History of Abiaca Creek Watershed Each phase of land use change, channelization, flow control, and mining within the Abiaca Creek watershed has produced a channel response, the evidence for which

PAGE 148

124 Combined Transport Analyses is geomorphically identifiable. Individual geomorphic surfaces generated in response to the perturbations border the channel throughout the study reach. These features consist of two distinct terraces (T 1 and T 2) and a modem berm. Each surface represents a distinct cycle of degradation/aggradation in response to a historic perturbation. The highest and oldest terrace, TO' can be related to a pre-European settlement lowering of base level due to changes in the Yazoo River (Figures 1 and 2). Intense agricultural development beginning in 1830 resulted in severe erosion of the Loess Hills in northern Mississippi. Rapid sedimentation rates in the alluvial valley resulted, and up to 16 feet of reworked sands and silts from the highlands were deposited on the valley floor on top of meander belt sediments (Figure 2). These sediments have been referred to as postsettlement alluvium (PSA). During this phase of deposition, a small, low-stage channel was maintained by the dense riparian vegetation, but during flood flows the valley bottom acted as the channel. Valley infilling by hillslope-derived sediments between 1830 and 1920 resulted in a rapid loss of channel capacity and subsequent flooding problems along Abiaca Creek. The lower-most reaches of the creek were channelized in 1920 in order to alleviate the flooding problem. Lowered base level and high channel slopes resulted in basin-wide channel erosion into the valley floor sediments between 1920 and 1940. This incision generated a terrace, referred to as T l' This terrace lies from 5 to 12 feet below TO (Figures 1 and 2). The incision induced by channelization (1920-1940) rejuvenated the upper drainage basin, and rapid channel aggradation occurred as sediment was supplied by upstream channel erosion as well as continued erosion in the Loess Hills. Rapid sedimentation increased the rate of channel widening in the lower reaches. Channel capacity was low and flooding prevalent. This aggradational phase extended from approximately 1940 to 1960 (Figure 2). In 1962, a watershed work plan was developed for Abiaca Creek by the Soil Conservation Service. Sixteen floodwater retarding structures were emplaced in the watershed between 1966 and 1977. The objectives of the work plan were to reduce flooding and to reduce the sediment delivered to the lower reach by 63 %. The rapid reduction of sediment input into the watershed produced yet another phase of channel incision. As a result of clear-water releases from the reservoir, sediment deposited from 1940-60 eroded out of the main channel, leaving a terrace remnant (T 2) perched atop a resistant bank toe (Figures 1 and 2). Differences in the extent of flow control (expressed as percent of watershed area controlled) on Coila (67%) and Abiaca (49%) creeks resulted in varying amounts of channel incision, as Abiaca Creek incised to a deeper level than Coila Creek. The most recent phase of channel evolution in Abiaca Creek watershed is characterized by aggradation in the mid-basin and downstream reaches, and by a cessation of channel erosion in the upstream flow-controlled reaches. The aggra dation has been driven in part by extensive sand and gravel mining in the mid-basin region, where mined sediment is stockpiled directly adjacent to the river and entrained during high flows. Longitudinal berms have developed at an elevation

PAGE 149

Fischer, Harvey, and Zevenbergen ABIACA CREEK CHANNEL INCISION PRE-18.30 BASE LEVEL LOWERING o 125 Figure 1. Schematic cross section showing phases of channel Incision on Ablaca Creek. (Dates reflect historic thalweg elevations. T = Terrace, B = Berm.) z o ---.l W o w m w > w n:: -----1 To 1 I c;>---z is J I-I 111 ABIACA CREEK TERRACE GENERATION I I 1750 1800 1850 1900 1950 2000 YEAR Figure 2. Chronologie account of watershed perturbations and channel thalweg response, Ablaca Creek watershed. Terraces T OT2 generated by channel Incision and berm B by aggradation.

PAGE 150

126 Combined Transport Analyses below T 2 on Abiaca Creek and active deposition occurs on top of T 2 on Coila Creek. HEC-2 water surface profiles were used to determine whether the geomorphic surfaces were correlatable as hydrogeomorphic units (Harvey and Watson, 1988). The correlation showed that the recurrence interval of berm full channel capacity on Abiaca Creek is equivalent to that ofT 2 channel capacity on Coila Creek. Therefore, the berms on Abiaca Creek are hydraulically equivalent to the aggrading T 2 surface on Coila Creek. The berm profile fit the water-surface profile at a discharge of approximately 1,450 cfs, or 25% of a two-year event. Aggradation in the lower reaches has created a flooding threat to agricultural lands of the Yazoo Delta and a sedimentation threat to a National Wildlife Refuge wetland at the mouth of the creek. Management AHernatives A numerical sediment routing model (HEC-2SR) that was calibrated with geomorphic measurements and observations, silt range re-surveys and a specific gage analysis indicated that the sand and gravel mines on Abiaca and Coila Creeks were delivering 34 and 5.8 acre-feetlyr of sediment to the channel, respectively. The determination of the relative geomorphic stability within individual study subreaches was critical to the model calibration. Three alternative solutions to the sedimentation problems were simulated with the numerical routing model. These alternatives were: 1) reducing sediment loads from the gravel mines, 2) setback levees designed to convey flood flows and trap sediment, and 3) a hill line dam designed to attenuate flood flows and trap sediment. Reducing sediment input at the gravel mines would prolong the existence of the National Wildlife Refuge wetland, although this alternative would not solve the flooding problems. The geomorphic history of the watershed suggests that a major reduction in sediment load from the mid-basin area would drive channel incision downstream, which could result in bank instability and channel widening. The hill line dam would have a greater impact on flooding and sedimentation than the setback levees. The setback levees were analyzed for 50-year flood protection, while the hill line dam easily could provide loo-year flood protection for the delta land. The hill line dam traps both bedload and washload while the setback levees trap bedload, but would have little impact on washload sediment yields. Sediment routing indicated that due to the location of the hill line dam, trapping of sediment behind the structure would not drive system-wide channel destabilization, because the dam would provide base level control. Some degradation in the leveed reach might occur. The geomorphic features of Abiaca Creek watershed record the complex action! response relationship characteristic of the drainage basin. Each solution to an existing problem that has been implemented has been followed by geomorphic responses requiring the implementation of further measures. The sensitivity of the system to structural control evidently has been underestimated in the past. The general problem now is one of a sediment supply that exceeds transport capacities in the lower reaches of the watershed. The system-wide response of the Abiaca Creek watershed to historic measures aimed at solving local problems indicates that combined geomorphic and hydraulic engineering studies that address system-wide response

PAGE 151

Fischer, Harvey, and Zevenbergen 127 should be integrated into any watershed management plan so that future channel destabilization can be avoided. References Harvey, M.D., and C.C. Watson 1988 "Channel Response to Grade-Control Structures on Muddy Creek, Mississippi." Regulated Rivers: Research and Management 2: 79-92. Schumm, S.A., M.D. Harvey, and C.C. Watson 1984 Indsed Channels: Morphology, Dynamics and Control. Littleton, Colorado: Water Resources Publications.

PAGE 152

GEOMORPInC RESPONSE OF LOWER FEATHER RIVER TO 19TH CENTURY HYDRAULIC MINING OPERATIONS Karin I. Fischer and Michael D. Harvey Water Engineering and Technology Introduction The Feather River and its major tributaries, the Bear and Yuba Rivers, drain the northern Sierra Nevada and discharge into the Sacramento River near Verona, California. The reach of the Feather River evaluated in this study lies within the Sacramento Valley, extending from the Yuba River confluence at Marysville (RM 28) to the Sacramento River at Verona (RM 0), a distance of 28 river miles (Figure 1). The primary objectives of the study were to determine the effects of 19th century hydraulic mining on in-channel and floodplain sedimentation and to evaluate their effects on present channel stability in the study reach. COLUSA KNIGHTS LANOING MAP Figure 1. General location map of study area

PAGE 153

Fischer and Harvey 129 Hydraulic Mining History During the Tertiary period (two million to 65 million years ago), the rivers of what is now the eastern margin of the Sacramento Valley flowed in a north-south direction, perpendicular to the west-flowing rivers of the Sierran foothills today. Gold derived from the Sierra Nevada range was deposited in these ancient stream channels. Modem rivers subsequently flowed across and eroded into those Tertiary deposits, redepositing gold-bearing sediment to the west in younger channels. A gold rush in 1849 involved placer mining in these modem stream channels. Once these deposits were exhausted, prospectors explored further and discovered immense deposits of gold in the Tertiary-age gravels of the Sierran foothills. The gold-bearing gravels were mined hydraulically since they were commonly hundreds of feet above modem rivers and could not be panned. As the miners began hydraulically mining, immense quantities of sediment were introduced to the drainage ways that are tributaries to the Feather, Yuba, and Bear Rivers. By 1909, more than 600 million cubic yards of hydraulically mined sediment had been washed into the Yuba River. Approximately 255 million cubic yards and 100 million cubic yards of sediment entered the Bear and Feather rivers, respectively. Hall (1880) estimated that 32 million cubic yards of debris lodged in the Feather River study reach in 1880 resulted in an average fill thickness of 20 feet. Meade (1982) showed an increase in the bed elevation of the Yuba River at the Feather River confluence near Marysville of 15 feet between 1850 and 1905 (Figure 2). An increase of nine feet was observed on the Sacramento River at Sacramento. Channel infilling from mining debris resulted in a dramatic loss in channel capacity on the Feather, Bear, and Yuba rivers. Extensive flooding and overbank deposition onto urban areas and agricultural lands surrounding these rivers resulted. The construction of levees in the early 1900s prevented extensive flooding, but only after several flooding disasters in the late 1800s. In 1868, the channel beds of the Feather and Yuba rivers at Marysville were higher than the city streets. A large portion of the city was buried by debris tailings in 1875 due to breaching of city levees. The repeated loss of agricultural land and property generated an uproar among townspeople and farmers of the lower Feather River. As a result of their efforts, the disposal of tailings into drainages by hydraulic mining operators was prohibited with the Sawyer Decision of 1884. The hydraulic mining industry began to decline. By the late 1800s, several plans were developed to reduce sediment infilling in the rivers of the Sacramento. Since the vast majority of sediment derived from hydraulic mining activity entered the Yuba River, initial efforts focused on that system. The strategy implemented in 1901 was aimed at storing sediment within the Yuba River system to prevent passage downstream into the Feather and Sacramento Rivers. The project included barrier dams, training walls, a settling basin and blasting of a low-flow channel. With the evolution of dredging technology, dredge mining of Yuba and Feather river channel deposits became very active in the early 20th century. Dredge mining occurred upstream of the Feather River study reach where immense volumes of dredge spoils were emplaced along the banks of the Up-

PAGE 154

130 Geomorphic Response to 19th Century Mining -E Yuba River at Marysville !c 16 o I-< 14 lIJ Co) a:: 12 (/) a:: 1850 lIJ > 4 Sacramento a:: L&.. 0 z 2 0 I-lIJ 0 lIJ 1850 1900 1950 River at Sacramento 1900 1950 Figure 2. Rise and fall of the annual low-water level of two rivers In California between 1850 and 1950, due mainly to the deposition and subsequent erosion of hydrauliC mining debris In their channels (from Meade, 1982). per Feather and Yuba Rivers. These coarse-grained dredge spoils were utilized as training walls to confine flows within single channels and limit downstream sediment delivery. Geomorphology of the Feather River The geomorphic characteristics of the Feather River document a complex and varied depositional history. Prior to 1850, the river was sinuous and characterized by deep pools and shallow riffles, similar to the present-day Sacramento River (Mendell, 1875). Deposition of hydraulic mining debris has effectively buried the pre-1850 meanderbelt. The diverse deposits characteristic of depositional subenviron ments associated with a meandering channel were replaced with relatively uniform sediments upon aggradation and development of a straight, wide, and shallow channel.

PAGE 155

Fischer and Harvey 131 The caliber of sediment delivered to the Feather River as a result of hydraulic mining varied greatly with time. The extensive mining of the relatively fine-grained upper "red gravel" caused an initial pulse of fine sediment into the channels. Transport downstream further differentiated the grain sizes, so that the sediment that first reached the study reach of the Feather River consisted of silts and clays. In a classic study, G.K. Gilbert (1917) termed the extensive thin bedded deposits formed by these fine-grained silts and clays "slickens." Gilbert (1917) recognized the downstream differentiation in grain size, explaining that "the coarser stuff tarried by the way, building up alluvial deposits on the lower hillslopes, in the flatter creek valleys, and the river canyons. When rains and floods came, the sands and gravels were moved forward toward the lowlands. The coarser-grained sediment prograded over the fine-grained slickens during floods. By 1910, the "sediment wave" had passed the study reach of the Feather River due to the decline of the hydraulic mining industry as well as engineering efforts to store sediment upstream. The sharp reduction in sediment load caused the river to incise into the hydraulic mining debris. The stratigraphy into which the channel incised consists of quartz dominated, mining-derived sands underlain by fine-grained, thinly bedded silts and clays of the slickens. The slickens form a relatively uniform and cohesive bank toe and add stability to the banks of the Feather River. The modern planform of the Feather River is not substantially different from that of the 1920s. Channel incision into the cohesive slickens, as well as construction of flood control reservoirs in the watershed, have helped to maintain the relatively straight channel planform. Planform Stability of the Feather River On the study reach of the Feather River, the existing channel planform is relatively stable due to the uniformity and cohesiveness of the slickens. The planform stability may be reduced, however, if the slickens are eroded through entirely and heterogeneous prehydraulic mining sediments are exposed in the bed and lower bank stratigraphy. Comparative thalweg profiles suggest that the Feather River has incised through the majority of hydraulic mining-derived sediment; if heterogeneous meanderbelt sediments underlying these cohesive deposits are exposed, variable lateral erosion and failure of the bank-forming slickens and coarser hydraulic mining sediments will ensue. The channel planform of the lower Feather River may destabilize as a result, thereby threatening adjacent flood-control levees. The dredge spoils and hydraulic mining debris stored in the Bear, Yuba, and Upper Feather rivers constitute the major sediment source for the Feather River study reach. The installation of Oroville Dam on the Feather River in 1968 greatly reduced the quantity of sediment derived from the upper Feather River watershed. The emplacement of dams on the Yuba (16) and Bear Rivers (4) has reduced sediment delivery to the study reach. The quantity of accessible sediment presently stored in the Yuba River system downstream of the lowest dam is large, consisting of several square miles of dredge spoils and hydraulic mining-derived sediment, but delivery of the sediment is dependent on the occurrence of flood flows. A lesser

PAGE 156

132 Geomorphic Response to 19th Century Mining amount of sediment is stored in the Bear River system, and these deposits are being progressively eroded in a downstream direction and delivered to the Feather River. Continued delivery of the tributary sediment is required to prevent further degradation of the Feather River in the study reach, which in turn will control the lateral stability of the river. Summary The geomorphic behavior of the Feather River has been strongly affected by hydraulic mining in the Upper Feather, Bear, and Yuba River watersheds. Effective river management strategies require a knowledge of the geomorphic history of the river so that potential future problems can be anticipated. The bed and bank stability of the Feather River is currently controlled by sediment load and bank stratigraphy, which are the direct result of 19th century hydraulic mining. If the river incises through the resistant slickens, premining strata will be exposed in the bank toe. Since the pre-1850 Feather River was similar in geomorphic character to the Sacramento River, the erodibility of the underlying sediments is likely variable and locally very high, which could lead to rapid lateral migration of the channel and potential levee threat. To provide effective flood control management on the Feather River, the channel incision should be monitored by permanent ranges and the competency of bank stratigraphy regularly evaluated. Gilbert, G.K. 1917 Hall, W.H. 1880 Mead, R.H. 1982 Mendell, G.H. 1875 References Hydraulic mining debris in the Sierra Nevada. U.S. Geological Survey Professional Paper 105. Report of the State Engineer to Legislature of California, 123rd Session, Part 3, Sacramento. "Sources, sinks and storage of river sediment in the Atlantic drainage of the United States." Journal of Geology 90, No.3: 235. Examination of Sacramento River below Tehama and of Feather River below Marysville, California: Annual Report Upon the Improvements of Rivers and Harbors in California-1875, Appen dix EE5, U.S. Army Corps of Engineers. Washington, D.C.: U.S. Government Printing Office.

PAGE 157

TWO DIMENSIONAL MODELING FOR FLOOD HAZARD DELINEATION George V. Sabol and Kenneth A. Stevens George V. Sabol Consulting Engineers Introduction This report describes the selection of a two-dimensional model for the analysis of flood hazard as a result of dam break for two dams; one on an inactive alluvial fan in Arizona, and another along the front range in Colorado. Both analyses were performed for the purpose of providing input to emergency preparedness plans. The Diffusion Hydrodynamic Model (OHM) was selected for use in Arizona because the dam embankment is on an alluvial fan with no well-defined channels for flood discharges, and it was selected for use in Colorado because of the diverse topography that exists downgradient of the reservoir that would cause large flood flows to break out of the natural drainage channel, resulting in multiple flow paths and lateral flooding. The use of the two-dimensional analysis over traditional one-dimensional techniques is believed to result in significant improvement in defining the flood inundation zones and in defining personal hazard zones. The use of the model does not require expertise beyond that for use of traditional one-dimensional unsteady flow models, nor does the use of the two-dimensional model require a resource commitment greater than what would normally be required for a one-dimensional analysis. The DHM is numerically more stable than the National Weather Service Dam Break model, resulting in improved modeling ease. However, the DHM does require familiarity in its use and has limitations that are not fully documented. Description of Project Areas Arizona. A dam break analysis was performed for the Spook Hill Flood Retarding Structure (Spook Hill FRS), an earthen embankment dam built by the Soil Conservation Service near Mesa, Arizona. The embankment is about 4th miles long and is located on an inactive alluvial fan along the base of an extended mountain range. There are no defined natural drainage watercourses downgradient of the embankment. The flood area below the dam is varied and contains agricultural, residential, industrial, and undeveloped land. The total area for the flood inundation study is about 31 square miles. Colorado. A dam break analysis was performed for the Superior Water Supply Terminal Reservoir near Superior, Colorado. The area contains rolling hills, with the dam constructed in an ephemeral watercourse. The project area, which is about 120 acres, is to be dedicated to residential development with large tracts set aside for parks and recreation areas. There was considerable grading of the natural topography in the process of developing the project area and the natural watercourse was altered by constructing ponds and recreational areas. Several major roads will be constructed through the project area. These roads may serve as conveyance channels for flood

PAGE 158

134 Two Dimensional Modeling flows when they are aligned in a downgradient direction, or the roads may serve as points of flow diversion where the alignments are oblique to the flow direction or where the road is constructed on elevated earthen fill across the watercourse. Description of Flood Conditions Arizona. Two conditions were analyzed for the Spook Hill FRS: 1) emergency spillway operation resulting in the release of about 7,000 acre-feet of water with a peak discharge of about 20,000 cfs, and 2) piping breach failures at mUltiple locations with peak discharges up to about 12,500 cfs and the release of more than 700 acre-feet of water. Colorado. Inflow from storm runoff is restricted from entering the reservoir and therefore embankment overtopping is not considered to be a reasonable assumption for this reservoir. Potential dam break was assumed to only be a consequence of piping. The peak discharge from the piping dam break was estimated as 7,210 cfs. Because the length of the dam is quite short, only one piping breach location was assumed. Selection of DHM Flood inundation analysis is often performed using one-dimensional flow routing models, such as the HEC-2, Water Surface Profiles Program by the U.S. Army Corps of Engineers (1990), and the National Weather Service DAMBRK Program (Fread, 1984). Both programs are applicable to one-dimensional flow-the HEC-2 program for steady, nonuniform flow, and the DAMBRK program for unsteady flow. The flow due to dam break would be unsteady, and, in both cases, the possibility for lateral flows exists. A two-dimensional, unsteady flow routing model was determined to be necessary for these analyses. The DHM (Hromadka and Yen, 1987) is the two-dimensional program that was selected for this study. It is a public domain program that was released by the U.S. Geological Survey (USGS) and has been used in other dam break flood inundation studies (Hromadka et al., 1985; and Hromadka and Lai, 1985). The DHM is capable of representing unsteady flow, backwater affects, ponding, channel overflow due to constrictions, overbank flow, mUltiple flow channels, and flow separations. Results from the DHM for one-dimensional flow have been compared to results obtained from the USGS unsteady, one-dimension flow model (K-634 model) with favorable results (Land, 1980a; 1980b). Description of DHM Theory. The DHM is an unsteady flow routing program that is based on the diffusion equation of motion where noninertial (gravity, friction, and pressure) forces are assumed to dominate the mechanics of flow. The DHM solves the diffusion equation through a finite difference numeric solution. That solution technique requires that the land surface and/or channel network be modeled as a grid of square

PAGE 159

Sabol and Stevens 135 cells of equal size. Flow is calculated to satisfy the principles of the conservation of mass and conservation of energy. Flow velocity (in all four directions) is calculated by Manning's equation. The flow into or out of each grid cell is solved according to the hydraulic properties that are assigned to each grid cell (length, width, and resistance coefficient) and by the hydraulic gradient that exists between the water surface elevation in each grid cell and the water surface elevation that exists at each of the four adjacent grid cells. The numeric solution is executed at incremental time steps. Output from the program is supplied at specified time increments (different from computation time steps). limitations. Very few limitations exist with the use of the DHM. A maximum of 250 grid cells are allowed, and this either limits the size of area being modeled or the size of grid cells that are used. It has been shown (Hromadka and DeVries, 1985) that iflarge velocity flow regimes are developed (greater than 25 feet/second), inaccuracies in results may develop. In typical dam break studies, such velocities are not encountered. Velocities in excess of 25 feet/second may occur at the breach opening, but such velocities will typically dissipate quickly as the flow enters the downgradient channel or the overbank inundation area. Input For each grid cell, a typical land surface elevation, Manning's roughness coefficient (n), and an initial water depth (if existing) must be specified. Since the DHM cannot take into account flow reduction factors such as buildings, bridges, retaining walls, etc., the Manning'S "n" value is varied in each grid cell to account for these effects. Global input to the model, that is, applied to the entire model area, requires the input of minimum and maximum computation time step values, the simulation time, the desired computer output time increment, surface detention depth (water loss), and maximum allowable change in water depth in each grid cell during a computation time step. Surface detention values are used to simulate infiltration and other water losses for the model area. Outflow from the model area is accomplished by defining border grid cells as either no flow boundaries, or as critical flow sections. The DHM can also model channels within the floodplain. Output. Output from the program consists of flow depth and velocities in all four directions for each of the grid cells at each output time step that is selected. This output can be used to define 1) the time history advance of the leading edge of the flood inundation wave, 2) flow depths at any grid cell at any time, and 3) flow velocities at any grid cell at any time. Such output can be used to define critical flood hazard zones where flow depths, flow velocities, or products of flow depth and velocity are greater than specified values. This greatly enhances the ability to define the hydraulics of floodplain and flood inundation zones and to identify areas of partiCUlar hazard.

PAGE 160

136 Two Dimensional Modeling Results Arizona. Input was defined and the model was executed for the emergency spillway operation and for three piping breach locations. The results are presented as a series of four maps: 1) model grid layout, 2) flow arrival times and personal hazard zone delineation, 3) maximum depth contours, and 4) maximum velocity contours. The personal hazard zone is defined as the flood zone where either the maximum flow depth exceeds two feet or where the maximum product of flow depth and velocity exceeds seven. The personal hazard zones indicate areas of special concern or where evacuation should commence. From those maps, a generalized map of flood inundation was prepared that indicates flood hazard and flow arrival times for any breach location along the 4th-mile-long embankment. Colorado. Flood inundation maps indicating flow arrival time were prepared. In addition, because of the unique breakout characteristics of this area, breakout hydrograph information was provided at all locations where flood waters deviated from the natural drainage watercourse. Fread, D.L. 1984 References DAMBRK, The NWS Dam-Break Flood Forecasting Model. Hydrologic Research Laboratory, National Weather Service, Silver Spring, Maryland. 56 pp. Hromadka, T.V., II, Berenbrock, C.E., Freckleton, LR., and Guymon, G.L. 1985 A Two-Dimensional Diffusion Dam-Break Model: Advances in Water Resources. Vol. 8, pg. 7-14. Hromadka, T.V., II, and DeVries, 1.1. 1985 A Two Dimensional Dam-Break Model of the Orange County Reservoir. International Symposium on Urban Hydrology, Hydraulic Infrastructures and Water Quality Control, University of Kentucky, Lexington. pp. 185-193. Hromadka, T.V., II, and Lai, C. 1985 Solving the Two-Dimensional Diffusion Flow Model. Proceedings of the ASCE Hydraulics Division Specialty Conference, Orlando, Florida. Hromadka, T.V., II, and Yen, C. 1987 A Diffusion Hydrodynamic Model. USGS Water-Resources Investigations Report 87-237.

PAGE 161

Sabol and Stevens 137 Land, L.F. 1980a 1980b "Mathematical Simulations of the Toccoa Falls, Georgia, Dam Break Flood." Water Resources Bulletin 16, 6: 1041-1048. Evaluation o/Selected Dam-Break Flood-Wave Models by Using Field Data. USGS Water Resources Investigations 8044. U.S. Army Corps of Engineers 1991 HEC-2 Water Surface Profiles Program (Version 4.6.0). Hydro logic Engineering Center, Davis, California.

PAGE 162

EXTRAPOLATION OF REGIONAL FLOOD-FREQUENCY RELATIONS BASED ON FLOW VARIABILITY Kenneth L. Wahl U.S. Geological Survey Water Resources Division Introduction Regional flood-frequency studies commonly provide equations only for recurrence intervals of 100 years or less, but some applications require estimates of floods with recurrence intervals of more than 100 years. Larger recurrence interval floods, for example the 500-year flood, can be estimated from the 100-year flood if the ratio between the 500-year and 100-year floods is known. However, that ratio varies regionally. This report suggests a method to extrapolate regional flood frequency relations using an approach that accounts for regional variation in flood ratios. This method was used, together with individual station flood-frequency curves and available regional equations for estimating 100-year and 500-year floods, to demonstrate the variation in ratios between the 500-year and 100-year floods over a six-state area. Analysis The Log Pearson type ill (LPIll) distribution is widely used to define the individual station frequency data that serve as the basis for regional relations. For an individual gaging station, the Log Pearson type ill distribution can be used in the general equation: Log(Qt) = M + (Kt)S (1) where Qt = discharge (Q) for the t-year recurrence interval; M = average of the logarithms of annual peak discharges; Kt = Pearson's K for the t-year recurrence interval; and S = standard deviation of the logarithms of annual peak discharges. Pearson's K is a function of both the recurrence interval and the skewness of the logarithms of annual peak discharges. Therefore, K and S reflect the symmetry and magnitude of flow variability. Writing the above equation for the 500-year and 100-year floods yields: Log (Q500) = M + (l(500)S, (2) and Log (Q1OO) = M + (K100)S (3) Substituting for M in either Equation 2 or 3 and rearranging terms yields: Q500/Q1OO = lOS(I(500-K100) (4) The results of Equation 4 for common values of skew and S are summarized in Figure 1. Because individual flood-frequency relations are defined using the LPill distribution, frequency curves defined from the regional relations can be assumed to

PAGE 163

Wahl 3.0 2.6 g 2.2 ..-.z o o Ii) 01 8 1.4 1.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 STANDARD DEVIATION OF LOGARITHMS Figure 1. Relation between ratio of 500-year and 100-year floods and standard deviation of logarithms of annual peak discharges. 139 0.7 also follow the LPm distribution. The skewness of logarithms of annual peak discharges has been mapped for the nation by the U.S. Water Resources Coun cil(l981). Therefore, defining average values for S over a region permits estimating regional values of the ratio between the 5OO-year and loo-year floods. Equation 4 was used to calculate the ratio between the 5OO-year and lOO-year floods in three western states (North Dakota, South Dakota, and Utah) where neither regional equations for the 5OO-year flood nor individual station flood-frequency curves showing the 5OO-year flood were available. In three other western states (Colorado, Montana, and Wyoming) regional equations for the 5OO-year flood or individual flood-frequency curves showing the 5OO-year flood were available. Flood frequency reports for Colorado (Kircher et al., 1985; McCain and Jarrett, 1976) and for Wyoming (Druse et al., 1988; Lowham, 1988) include equations for estimating the Soo-year flood. The ratios shown in Table 1 for Colorado were derived from the estimating equations; those for Wyoming and Montana were calculated as regional averages from the individual station

PAGE 164

140 Regional Flood-Frequency Relations Table 1 Summary of Ratios Between the SOo-Year and 10o-Year Floods Flood frequency region and referat ICe Colorado (Kircher et aI., 1985) Mountain Region Rio Grande Region Southwest Region Northwest Region Plains Region (McCain and Jarrett, 1976; Uvingston and Minges, 1987) Montana (Omang et aI., 1986) West Region Southwest Region Upper Yellowstone Region Northwest Region Northwest Foothills Region Northeast Plains Region East Central Plains Region Southeast Plains Region North Dakota (Crosby, 1975) North and east of Missouri River South and west of Missouri River South Dakota (Becker, 1974,1980) Eastern region (James River and all areas to the east) Western region (area west of James River basin) Utah (Thomas and Undskov, 1983) Northern Mountain Region-high elevation Northern Mountain Region-low elevation Uinta Basin Region-Uinta Mountain Range Uinta Basin Region-south of Uinta Range High Plateaus Region Low Plateaus Region Great Basin Region-high elevation Wyoming (Druse et aI., 1988; Lowham, 1988) Mountainous Region Plains Region High Desert Region Ratio (0500/Q100) 1.2 1.4 1.4 1.2 2.0 1.3 1.3 1.4 2.0 2.2 1.7 1.8 1.8 1.5* 1.8* 1.5 1.8 1.3 1.3 1.3 1.7 1.5 1.7 1.6 1.3 2.0 1.7 Rood frequency equations are limited to the 5O-year flood in North Dakota. The 100-year flood can be approximated statewide by multiplying the 5O-year equation by 1.25. The ratios shown in the above table are the ratios of the 500-year flood to the approximate 100-year flood.

PAGE 165

Wahl 141 frequency curves. In addition, the data for Wyoming were used to test the applicability of Equation 4, and the two methods were in good agreement. Conclusions The ratios shown in Table 1 demonstrate the regional variability of the ratio, Q500/Ql00. The arithmetic average of the ratios is 1.6. However, regional variation in the ratio is substantial. The average ratio for mountainous regions is only about 1.3, while the ratio for nonmountainous areas averages about 1.8-1.9. The mountainous northwest part of Montana is an exception to this general rule and is an area known to be affected by mixed population flooding (Omang et al., 1986). The variation in the ratio for the six states is typical of the variation for other western states. This variation in the ratio reflects, in part, changes in the factors causing floods. Becker, L.D. 1974 1980 Crosby,O.A. 1975 References A Methodfor Estimating Magnitude and Frequency of Floods in South Dakota. U.S. Geological Survey Water-Resources Investi gations 35-74. Techniques for Estimating Flood Peaks, Volumes, and Hydro graphs on Small Streams in South Dakota. U.S. Geological Survey Water-Resources Investigations 80-80. Magnitude and Frequency of Floods in Small Drainage Basins in North Dakota. U.S. Geological Survey Water-Resources Investi gations 19-75 (pB-248 480/AS). Druse, S.A., Lowham, H.W., Cooley, M.E., and Wacker, A.M. 1988 Floodjlow Characteristics of Wyoming Streams-Compilation of Previous Investigations. Wyoming State Report. Kircher, J.E., Choquette, A.F., and Richter, B.D. 1985 Estimation of Natural Streamjlow Olaracteristics in Western Colorado. U.S. Geological Survey Water-Resources Investigations Report 85-4086.

PAGE 166

142 Regional Flood-Frequency Relations Livingston, R.K., and Minges, D.R. 1987 Techniquesfor Estimating Regional Flood Characteristics of Small Rural Watersheds in the Plains of Eastern Colorado. U.S. Geological Survey Water-Resources Investigations Report 874094. Lowham, H.W. 1988 Strearriflows in Wyoming. U.S. Geological Survey Water Resour ces Investigations Report 88-4045. McCain, J.R., and Jarrett, R.D. 1976 Manual for Estimating Flood Characteristics of Natural-Flow Streams in Colorado. Colorado Water Conservation Board, Technical Manual No. 1. Omang, R.J., Parrett, C., and Hull, J.A. 1986 Methods for Estimating Magnitude and Frequency of Floods in Montana Based on Data Through 1983. U.S. Geological Survey Water-Resources Investigations Report 86-4027. Thomas, B.E., and Lindskov, K.L. 1983 Methodsfor Estimating Peak Discharge and Flood Boundaries of Streams in Utah. U.S. Geological Survey Water-Resources Investigations Report 83-4129. U.S. Water Resources Council 1981 Guidelines for Determining Flood Flow Frequency. Hydrology Committee Bulletin 17B.

PAGE 167

GEOMORPIDC AND SEDIMENT TRANSPORT ANALYSES FOR THE NAPA RIVER FLOOD CONTROL PROJECT* C. Gary Wolff and Mark R. Peterson Water Engineering and Technology Introduction The Napa River, located in Napa County, California (Figure 1), has a long history of flooding that has resulted in severe damage to agricultural and urban developments. The Napa River Flood Control Project, as described in the 1975 General Design Memorandum (GDM) (USACE, 1975), is designed to provide 100year flood protection for the City of Napa extending from RM 6 to RM 17.3. To achieve the desired level of flood protection, channel improvements are proposed from John F. Kennedy Memorial Park (RM 11.8) to the Trancas Road bridge (RM 17.3). The bendway between RM 14.7 and RM 15.6 will be cut off. Geomorphic and sediment transport analyses of the Napa River were performed to address: 1) the impact of sediment on project performance and 2) the impact of the project on the behavior of the stream system and the limits of the project's influence on the morphology of the stream system. Geomorphic Analysis The Napa River has been subjected to a number of perturbations that include dredging of a navigation channel, bend cutoff, sand and gravel mining, channel maintenance excavation, and significant changes in land use in the watershed. A navigation channel originally was dredged in 1950 and was followed by maintenance dredging in 1962-63, 1981-82, and 1988. The upstream limit of the dredging was RM 14.7. Aerial photographs taken in 1958 show that during that time, sand and gravel were removed from the Napa River, and channel maintenance work that included clearing and snagging of the channel was conducted. Accelerated develop ment of the watershed for grape production has resulted in the use of steeper valley margin lands that have the potential to significantly increase the sediment yield from the watershed. Increased sediment delivery from the watershed could increase the need for maintenance dredging both to permit navigation and to maintain flood protection. This paper is based on work perfonned by Water Engineering and Technology (WET) for the U.S. Anny Corps of Engineers, Sacramento District, under Contract No. DACW05-88-D-0044, Delivery orders 0007 and 00012. Details of the work are contained in two reports (WET, 1990a and 1990b). The analyses presented in this paper are those of the authors and do not necessarily reflect the view of the U.S. Anny Corps of Engineers.

PAGE 168

144 Transport Analyses for the Napa River Figure 1. General location map for the Napa River Flood Control Project. Field evidence that includes: 1) strath terraces, 2) knickzones in the bed of the channel, and 3) exposed bridge footings indicates that the Napa River upstream of Oak Knoll bridge is degrading. Field evidence also indicates that some degradation has progressed upstream as far as Zinfandel Avenue bridge (RM 34). Total amounts of recent channel incision range from 4.5 feet to 6 feet. The presence of several knickzones in the channel bed between Oak Knoll Avenue bridge and Zinfandel Avenue bridge suggests that the degradation has been punctuated and may not be attributed to any single event. The cause of the degradation could be channel dredging, sand and gravel mining, channel maintenance, tectonic uplift of the valley, or a combination of all of these factors. Tributaries to the Napa River that include Soda Creek, Dry Creek, and Conn Creek have incised in response to degradation of the main channel. Degradation generally predisposes channels to bank failure and channel widening (Schumm et al., 1984; Harvey and Watson, 1986; Harvey and Schumm, 1987). However, field observation of the Napa River indicates that bank erosion is not a serious problem in the study reach. The lack of bank erosion is probably due to the resistance of the bank materials. Bay and marsh sediments, cohesive fan margin sediments, and cemented fanglomerates form the lower channel banks and provide considerable stability to the banks. In the multi channeled reach of the Napa River upstream of the project reach (RM 21.5 to RM 23), cohesive bank materials are locally absent. Continued degradation in this area could provide a significant source of sediment to the project reach downstream.

PAGE 169

Wolff and Peterson 145 Sediment Transport Analysis A sediment transport analysis of the project reach (RM 6 to RM 17.3) was performed to establish baseline aggradation/degradation trends under preproject conditions and to predict changes under project conditions. The analysis was performed using a sediment continuity model based on joint application of the Meyer-Peter, Mueller (MPM) bed load and Einstein suspended load equations. Armoring calculations were used to establish the limit of degradation potential. Sediment transport in the project reach of the Napa River is complicated by tidal backwater, which results in the deposition of fine material (wash load) derived from upstream sources. A separate analysis based on an algorithm in the HEC-6 model (USACE, 1977) was performed to determine fine sediment deposition volumes. The sediment transport analysis of preproject conditions showed that deposition downstream of RM 14.7 is the most significant problem in the project reach. The average annual bed material aggradation volume computed for the reach downstream of RM 14.7, the reach with historical maintenance dredging, was 16,800 cubic yards. The computed average annual deposition volume for fine sediment in this same reach was 19,900 cubic yards. The deposition problem is due largely to the effect of tidal backwater. The project reach upstream of RM 14.7 was shown to be near equilibrium. The sediment transport analysis of project conditions showed that degradation potential upstream of the proposed bendway cutoff (RM 14.7 to RM 15.6) is the most significant problem associated with the proposed flood control plan. The degradation potential upstream of the proposed cutoff results in a significant increase in aggradation potential downstream. The average annual bed material aggradation volume computed for the area within and downstream of the cutoff was equal to 48,000 cubic yards, much greater than the 16,800 cubic yards computed under existing conditions. This result illustrates the need to control degradation in the project reach upstream of the cutoff. Armoring calculations showed that armoring does not appear to limit computed degradation for any case analyzed. The fine sediment deposition volume was computed as 13,900 cubic yards per year under project conditions, approximately 30% less than the 19,900 cubic yards per year computed under preproject conditions. This result should be used to indicate only a direction of change under project conditions, since the fine sediment model is a simple conceptualization of the actual physical processes and has been calibrated to only one data point. The bed material sediment transport analysis on the Napa River was repeated assuming complete stabilization of the channel bed upstream of the proposed cutoff (RM 15.6). The average annual aggradation potential of the project reach, including the cutoff and bendway, was computed as 22,100 cubic yards, significantly less than the 48,000 cubic yards computed with no upstream stabilization. To control degradation upstream of the cutoff, a single bed stabilization structure located at RM 16.2 is recommended. This location was determined using equilibrium slope calculations. Assuming the channel bed downstream of the structure degrades

PAGE 170

146 Transport Analyses for the Napa River to the equilibrium slope, the bed stabilization structure will be exposed on the downstream end by approximately two feet (this figure ignores local scour). Using design charts for a sheet pile bed stabilization structure (Linder, 1963), it was shown that the degradation downstream of the structure will have only minor effects upstream. Since the computed water surface elevation at the upstream end of the project reach is greater than that computed for existing conditions, the project should have no significant upstream effects if the bed stabilization structure is included in the design. Conclusions The geomorphic analysis has shown that the Napa River is presently degrading upstream of the project reach. The cause of the degradation is unknown, but it could be due to any of five factors: a) dredging of the navigation channel, b) sand and gravel mining, c) channel clearing and excavation, d) tectonic uplift, or e) any combination of the factors. Degradation of the Napa River has lowered base level for its tributaries. The resistance of the channel perimeter sediments along the degraded reach has prevented the occurrence of severe bank erosion as a result of channel degradation. The sediment transport analysis of the project reach (RM 6 to RM 17.3) has shown that the most significant problem under preproject conditions is aggradation downstream of RM 14.7. The problem is the result of tidal backwater and a high sediment yield from the watershed and upstream channel degradation. The most significant problem under project conditions is degradation upstream of the proposed cutoff and a resultant increase in aggradation downstream. A single grade stabilization structure located at RM 16.2 would provide adequate stabilization of the project reach upstream of the cutoff and mitigate project impacts upstream of the project. References Harvey, M.D., and S.A. Schumm 1987 Response of Dry Creek, California, to Land Use Change, Gravel Mining, and Dam Gosure: Erosion and Sedimentation in the Pacific Rim. IAHS Publication No. 165, p. 281-282. Harvey, M.D. and C.C. Watson 1986 "Fluvial Processes and Morphological Thresholds in Incised Channel Restoration." Water Resources Bulletin 3, No.3: 359368.

PAGE 171

Wolff and Peterson 147 Linder, W.M. 1963 "Stabilization of Stream Beds with Sheet Piling and Rock Sills. Prepared for Federal Interagency Sedimentation Conference of the Subcommittee on Sedimentation, ICWR. Schumm, S.A., D.I. Gregory, L.R. Lattman, and C.C. Watson 1984 River Response to Active Tectonics: Final Report, Phase I. National Science Foundation. United States Army Corps of Engineers (USACE) 1975 General Design Memorandum and Appendices. Napa River Flood Control Project, Napa County, California. 1977 HEC-6, Scour and Deposition in Rivers and Reservoirs, User's Manual. Water Engineering and Technology, Inc. (WET) 1990a Napa River Sediment Engineering Study, Phases I and II. 1990b Prepared for U.S. Army Corps of Engineers, Sacramento District, Contract No. DACW05-88-D-004. Napa River Sediment Engineering Investigation, Phase I. Prepared for U.S. Army Corps of Engineers, Sacramento District, Contract No. DACW05-88-D-004.

PAGE 172

TWO-DIMENSIONAL HYDRODYNAMIC MODEL OF THE COLUSA ** FLOOD OVERFLOW WEIR ON THE SACRAMENTO RIVER Lyle W. Zevenbergen, Love and Associates Mark R. Peterson, and Michael D. Harvey Water Engineering and Technology Introduction This paper describes the application of a two-dimensional hydraulic model (RMA-2V) used to simulate the Sacramento River at the Colusa Weir near Colusa, California. The project was conducted by Water Engineering and Technology (WEn under contract to the Sacramento District, U.S. Army Corps of Engineers (WET, 1991). As shown in Figure 1, Colusa Weir is located along the east levee at approxi mately RM 146 on the Sacramento River. The weir and accompanying gaging station are maintained and operated by the State of California Department of Water Resources. The weir is one of several flood overflow weirs that divert flow from the Sacramento River into Sutter Bypass to eliminate downstream levee overtopping. When constructed in the early 1930s, Colusa Weir was located along the outside bank of Cobbs Bend. As shown in Figure 1, Cobbs Bend has shifted and currently is at the downstream end of Colusa Weir. A pilot channel has been dredged along the 1930s Cobbs Bend alignment to maintain flow to the weir. The inside (right) bank of Cobbs Bend has been revetted to keep the bend from migrating further downstream. The RMA-2V model was used to determine weir performance under existing channel conditions and potential future channel conditions. The future conditions modeled were: 1) 10 years of channel migration, 2) 15 to 20 years of channel migration, and 3) the removal of 400 feet of the Cobbs Bend point. Upstream inflow rates of 135,000 cfs (high flow), 65,000 cfs (intermediate flow), and weir flow initiation (approximately 31,000 cfs) were modeled. Figure 2 shows the finite element network used to represent the existing river conditions. RMA-2V Model Description The following description is a summary of the information contained in the T ABS-2 User's Manual (Thomas and McAnally, 1985). The RMA-2V model (Two-** The views expressed in this paper are those of the authors and do not necessarily represent the opinions or conclusions of the U.S. Army Corps of Engineers. Assistance and support for this study, provided by members of the Hydraulics Design Section ofthe Sacramento District, U.S. Army Corps of Engineers, is gratefully acknowledged.

PAGE 173

Zevenbergen, Peterson, and Harvey r"' .. .. .. IIIV[.IOAO --.. __ !:.!v!..,_' / / -------+ LEvE[ ...... '" UCII I ""'". ""." "'---r'""" I,V[lIIltOFtO II .. seq:.,.[ ... nILES '" SCALI 1M nlllS .... .... .... .... .... .... .... d Figure 1. Project site. Figure 2. Finite element mesh-exlstlng conditions. 149 ., ...... ''''''''1 .m .. ,. ''''''''1

PAGE 174

150 Two-Dimensional Hydrodynamic Model Dimensional Model for Open-Channel Flows) computes water surface elevations and velocity components in the horizontal plane for subcritical, free-surface flows. The original model was developed by King, Norton, and Orlob (1973) and is supported by both the U.S. Army Corps of Engineers Hydrologic Engineering Center and Waterways Experiment Station. The program uses the finite element method to solve the Reynolds form of the Navier-Stokes equations for turbulent flows. A network of elements is used to describe the geometry of the channel and overbank area. Elements may be either three-or four-sided, and are described in space by nodes located at the element corners and midsides. Each node is assigned X and Y coordinates and an elevation. The model also requires Manning flow resistance factors and turbulence exchange coefficients for each element. Up to 10 different element types may be used. In general, channel elements are differentiated from overbank elements, and overbank elements are divided by land use. The model also requires boundary conditions where flow enters or exits the model. Boundary conditions may be velocity or unit discharge vectors, water surface elevations, or rating curves. Application of RMA-2V to the Colusa Weir Site Figure 2 shows the finite element mesh for the existing conditions runs. Note that the elements are concentrated along the main channel and pilot channel, while the overbank elements may be several acres large. The model was calibrated using a flood event occurring on February 26, 1986, when an upstream inflow of 87,000 cfs resulted in a gaged discharge of 43,300 at Colusa Weir and a recorded discharge of 43,700 cfs at the Colusa Bridge gage, respectively. The calibrated Manning coefficients were 0.025 and 0.030 for the main channel and pilot channel, respectively. The calibrated Manning coefficients ranged from 0.038 to 0.084 for overbank elements, which included cropland, orchards, and dense woods. The model was then validated using a flood event occurring on March 3, 1983, when an upstream inflow of 120,800 cfs resulted in a gaged discharge of 69,600 at Colusa Weir and a recorded discharge of51,200 cfs at the Colusa Bridge gage, respectively. The difference between the measured and predicted flow splits for this event was less than 100 cfs, which is well within the accuracy of the gage data. The model was then run to determine the weir overtopping discharge and the flow splits for upstream flow rates of 65,000 (intermediate flow) and 135,000 (high flow) cfs. These results (shown in Table 1) serve as a basis of comparison for potential future channel conditions. The predicted weir discharge of 81,400 cfs for high flow conditions is more than 10,000 cfs greater than previously believed. The model was then modified to simulate hydrodynamics corresponding to assumed river conditions resulting from: 1) 10 years of channel migration, 2) 15 to 20 years of channel migration, and 3) the removal of 400 feet of Cobbs Bend point. The channel migration runs were generated by extrapolating recent migration rates measured on aerial photography taken in 1981 and 1988 and consideration of other geomorphic characteristics in the study reach (WET, 1989; 1990). During this period, Arnold Bend point was being trimmed at 60 feet/year, Cobbs Bend outer

PAGE 175

Zevenbergen, Peterson, and Harvey 151 bank was migrating at 25 feet/year, and the other nonrevetted bends were migrating at rates between 10 and 32 feet/year. The results simulating these future conditions are also shown in Table 1. Fortunately, future channel conditions should not excessively reduce weir discharges. Table 1 Summary of Flow Distributions for All Model Runs Discharae {cfs} Condition Fl.In Colusa Colusa Total Weir Bridge (upstream) Calibration Observed 43,300 43,700 87,()()() Modeled 43,400 43,600 87,()()() Validation Observed 69,600 51,200 120,800 Modeled 69,500 51,300 120,800 Existing High flow 81,400 53,600 135,()()() Conditions Intermediate 26,700 38,300 65,()()() Low flow 0 31,300 31,300 No Action High flow 80,900 54,100 135,()()() Alt. 1 Low flow 0 31,700 31,700 (10 year) No Action High flow 80,700 54,300 135,()()() Alt. 2 Low flow 0 31,950 31,950 (15-20 yr) Removal of High flow 81,200 53,800 135,()()() Cobbs Bend Intermediate 26,300 38,700 65,()()() Point Low flow 0 31,600 31,600 Summary and Discussion The validation model was in excellent agreement with observed flow rates. Therefore, a high level of confidence was placed on the results of both the existing conditions runs and potential future conditions runs, especially for high flow conditions. The future conditions runs represent potential future channel alignments based on historic behavior and a thorough understanding of geomorphic response of the Sacramento River system. Past work and experience of WET geomorphologists and engineers in the study of the Sacramento River system provided the background needed to make such projections (WET 1989, 1990). For all high flow runs, over 0.5 feet of water surface drawdown occurred between the main channel and the weir. This result indicates that where a weir control section is set well back from the main flow, a two-dimensional model would be more accurate than the HEC-2 split flow option.

PAGE 176

152 Two-Dimensional Hydrodynamic Model For the Colusa Weir site, it is apparent that backwater from downstream of the study site plays a greater role in determining flood stages at the weir than local backwater from Arnold Bend and Cobbs Bend. Therefore, channel migration at the study site should not adversely impact weir performance. Also, if it became desirable to increase weir flow rates, reducing the hydraulic efficiency of Colusa Bridge might be an effective solution. Finally, maintenance of the pilot channel should be continued to limit the amount of water surface drawdown at the weir. This project demonstrated the usefulness of two-dimensional hydrodynamic modeling techniques for evaluating complex flow conditions developed in a reach of the Sacramento River. The RMA-2V model was applied to investigate river hydrodynamics for a wide range of both discharge and channel planform changes. References King, J.P, W.R. Norton, and G.T. Orlob 1973 A Finite Element Model for Lower Granite Reservoir. Water Resources Engineers. Thomas, W.A. and W.H. McAnally, Jr. 1985 User's Manual for the Generalized Computer Program System: Open-channel Flow and Sedimentation. T ABS-2. Water Engineering and Technology (WEn 1989 Final Phase II Report, Geomorphic Analysis of Reach from Colusa to Red Bluff Diversion Dam, River Mile 143 to River Mile 243. Report to Sacramento District, USACE, Contract No. DACW05-87-D-0094. 1990 Geomorphic Analysis and Bank Protection Alternatives Reportfor Sacramento River (RM 78-194) and Feather River (RM 0-28). Report for Sacramento District, USACE, Contract No. DACW0588-D-0044. 1991 1991 Sacramento River Bank Protection Project-Colusa Weir/ Cobbs Bend Study. Report to Sacramento District, USACE, Contract No. DACW05-88-D-0044.

PAGE 177

Part Five Local Flood Warning

PAGE 178

This page IS intentionally blank

PAGE 179

CAN WE HAVE TOO MUCH WARNING TIME? A STUDY OF ROCKHAMPTON, AUSTRALIA John Handmer Australian National University Introduction Flood warning studies are preoccupied with situations with short warning times. This is reasonable enough given the limited time available for the flood detection, warning, and response process, especially in view of the very substantial dollar and human losses often associated with such events. In contrast, longer warning times offer a number of important advantages, including opportunities to greatly reduce direct and indirect damages; time for a "rusty" or inexperienced system to learn as the flood progresses; and, where the time available is very long, major structural works can be undertaken in a crisis free atmosphere. However, long warning times are by no means problem free. A common problem experienced by authorities issuing warnings with long lead times is competition from "unofficial" warning sources. An active unofficial system also raises an important methodological issue for those examining system effectiveness in terms of damage reduction. It may be very difficult to distinguish the effects of the unofficial from the official systems. This paper examines the problem of unofficial, potentially competing forecasts. It does so through a case study of the 1988 flood in the Australian city of Rockhampton. Rockhampton Rockhampton is located on the Tropic of Capricorn, on the central Queensland coast. Just under 60,000 people live in the city, which straddles the Fitzroy River, some 50 km from the river mouth. At 52,900 square miles, the river is the largest coastal catchment in eastern Australia. If the record 1918 flood were to reoccur (10.11 meters), some 4 % of the city's population would be directly affected, as well as key commercial premises. The flood of March 11, 1988, reached 8.4 meters on the Rockhampton gage. The Rockhampton Flood Warning System The outstanding feature of flood warnings in Rockhampton is the amount of warning time available. Over a week before the flood peak reached the city, the Australian Bureau of Meteorology was able to specify the height and timing with reasonable accuracy. This is possible because of the size of the Fitzroy River *The research reported here was funded by the Australian Bureau of Meteor ology. The other members of the study team were David Ingle Smith and Mark Greenaway.

PAGE 180

156 Can We Have Too Much Warning Time? catchment, but also owes much to the bureau's ability to capture and process streamflow and rainfall data for the entire catchment and beyond. However, even without the bureau's involvement, accurate warnings of a few days are possible. Data collection. In the official flood warning system, the Bureau of Meteorology is the central collecting point for river height and rainfall data. The data come from gages owned and/or operated by the bureau and other agencies. In general, the gage readers are volunteers who, in 1988, sent their readings to the bureau by telegram. To maximize data quality, the bureau finds that it is often necessary to use personal contacts to supplement official sources. Of course, as a national agency the bureau also has access to data collected throughout Australia, and to satellite and radar material. Following the 1988 flood peak, a key gage malfunctioned and did not indicate that the water level was falling. Unfortunately, this incident led to rumors of a second flood peak. The initial flood warnings were based on radar data, synoptic reports, and meteorological advice. The first warnings were issued on Monday, February 29, for moderate flooding in the Connors River, the northern arm of the Fitzroy. As a result of data from the gage readers and synoptic reports, by 7:30 p.m. on March 2 the bureau was able to refine its forecast, predicting a flood peak at Rockhampton "about next Thursday, 10th March." Once the headwater tributaries started to flood there was about a week for Rockhampton residents and others to undertake unofficial data collection-to observe the flood and speculate on its likely height and timing. Observation and informal communication were important information sources for many people. Over 100 people checked the Rockhampton flood gage regularly. Of greater concern were those who were telephoning a key automatic gage. These unauthorized phone calls (the gage phone numbers are "secret") on occasion made it difficult for the official components of the warning system to get through to the gage and were a major drain on the gage's battery power supply. In addition to members of the public, the Bureau of Meteorology and other government and media organizations were collecting data directly from the upstream gages. Forecast preparation. The bureau prepares flood forecasts using its model of the Fitzroy catchment. As the flood wave approaches Rockhampton, these predictions will generally be discussed with officers from the Rockhampton Council prior to their release. The council's engineering section prepares its own flood forecast based on its extensive records of previous floods. The bureau aims to provide at least 60 hours warning of moderate or major flooding at Rockhampton and to forecast the peak with a week's notice of its approximate height. However, the flood eventually peaked on March 11, a day later than originally forecast. The exact timing and height of the peak is determined by tidal and associated hydraulic effects. In view of the length of time available and the variety of people collecting upstream river data, it is not surprising that the bureau's forecast was at times competing with unofficial predictions. On the whole, these came from local identities

PAGE 181

Handmer 157 or "gurus," long-time residents whose views received wide media publicity. They were frequently quoted as employing rules of thumb as the basis for their forecasts, a procedure likely to imprison them within their own experience. However, their rules of thumb would have been augmented by unofficial telephone access to the various rain and river flow gages. Of course, those at risk and their friends and relatives may also have access to the gages and may construct their own forecasts. Clearly, with active media encouragement the potential for rumors is real. Other government agencies may also issue their own forecasts. In recent floods in the Rockhampton area, these have concerned additional flood peaks and dam failuresall fictional. Warning dissemination. If floodplain occupiers act on their own observations, the process is relatively direct a1Jd has only to go through one set of filters and constraints. However, the official process can be quite complicated. For example, forecasts are telexed from the bureau's Brisbane office to State Emergency Services (SES) headquarters in Brisbane; these are faxed to the regional SES office in Rockhampton and then passed by hand to the local SES-the people actually combating the flood. This appears clumsy, but apparently works well in Rock hampton, where time is not critical. In general, the local media are very cooperative. However, interviewees from the warning and response system felt that the flood had been blown out of all proportion by the "southern media"-basically, broadcast media from the state capitals. This sort of intense media interest can place added strain on the emergency services, especially as some media organizations apparently telephoned the local SES every hour from 3:00 a.m. on. The absurdity of this is evident when we remind ourselves that the flood has over a week's lead time and a flat peak lasting a day or two. By association, the hype may have reduced the credibility of local media. The reports also caused anxiety to out-of-town people with friends and relatives in Rockhampton, many of whom telephoned the city to check on the situation, adding additional load on emergency workers and the telephone system. Another unfortunate effect of the media hype was that some Rockhampton people felt that important information was being concealed by local authorities. Multiple sources of information. Officials emphasized that a major advantage of the present official flood warning system was that there was only one source of information. However, as far as the flood-prone public are concerned, there are numerous sources, although the situation has improved considerably since the 1983 flood. While many of these sources, such as the print and broadcast media, the SES, and the police, may pass on the bureau's warning verbatim, they will also add material that may modify the context of the message. In addition to these official warning disseminators, people in the floodplain would have been receiving advice from friends and relatives living both inside and outside Rockhampton who had seen the media reports of "severe" flooding. They would also have been exposed to the southern media themselves through networked TV and radio programs and out-of town newspapers. In addition, certain organizations provided specific flood-related

PAGE 182

158 Can We Have Too Much Warning Time? information, such as the RACQ (road conditions) and the Harbor Master (boating conditions). These in no way conflicted with the bureau's warnings. However, the opinions of various local gurus received considerable publicity in the local media, as did the predictions of certain organizations. Table 1 lists the different sources of flood predictions for Rockhampton as quoted in The Morning Bulletin from March 2 to March 11, 1988. Where the bureau is active in providing flood warnings, it has tried to eliminate other warning sources. An important reason for this is that the more sources, the more likely it is that conflicting advice will be offered to the public. The flood warning literature emphasizes the importance of avoiding this situation. But, it also appears that where multiple sources present the same story, the result is powerful persuasion. The questions therefore are: did the other information sources conflict with the bureaus's advice, and did they affect the bureau's credibility? Table 1 shows that, on the whole, unofficial predictions published in the local newspaper did not conflict with the bureau's forecasts. On the occasion where there was a substantial difference, the official forecast dominated as it occupied all of page one of the newspaper. However, hype by network media may have been more of a problem. Credibility. It is almost inevitable that the bureau would need to revise very early flood-height predictions. During the last flood these revisions were relatively minor. If they were major they could damage the bureau's credibility. A related problem concerns doubts about the bureau's ability to issue reliable early warnings. One official commented that as the bureau was in Brisbane, hundreds of miles away, "How do they know what's happening so far in advance?" Conclusions As warning times increase, the official flood warning system is increasingly likely to find itself working in tandem with an unofficial system. At its simplest, this could be individuals making their own assessments based on environmental indicators such as heavy rain. However, it is more likely that those at risk will receive warning messages from many different sources. The Rockhampton study showed that, on the whole, unofficial forecasts with a clear local origin supported the bureau's official warnings. In terms of persuasion, the mUltiple nonconflicting sources were probably an advantage in that they would have satisfied the need for confirmation and reinforcement. However, although not specifically investigated by the study, it is clear that there was considerable media hype by sections of the national broadcast networks. This had a number of effects, including possibly reducing the credibility of the official warnings and raising anxiety among local residents who would see and hear these reports and be contacted by concerned friends and relatives from outside Rockhampton. One way around the problem of multiple sources is to provide a central, credible, accessible information source. In Rockhampton the council and emergency services attempted to do this.

PAGE 183

Handmer Table 1 Unofficial Flood Forecasts for Rockhampton in The Morning Bulletin, March 3-9, 1988 159 (Note that this may not be a complete listing. In addition, The Morning Bulletin carried leading articles with official forecasts.) Date 3 March 4 March 4 March 4 March 7 March 7 March 9 March 9 March BOM = Bureau of Meteorology Source quoted William Grigg, Civil Engineer, Capricornia Institute Rockhampton Mayor and Deputy Mayor Mrs. McCamley of Tartru s Stati 0 n Cec lIand, "long time resident" Spokesman for the Queensland WRC Deputy Mayor Rockhampton Council Jack Bredhauer and Basil Weisse, "bushmen" Prediction General comments reinforcing BOM. To assess risk, remember what happened in 1983 (8.25 m). This was in line with BOM predictions of 8.25 m. Water level at the station higher than 1954. Flood will not be as big as 1954 (9.4 m). Suggests a flood slightly higher than BOM. Peak would arrive on March 9 or 10 and be slightly higher than 1983. Peak would be 40.6 cm higher than 1983 (8.7 m). Both the WAC and Deputy Mayor are in line with BOM. although by 7 March, BOM was revising its forecast downwards. 8.7 m late 9 March or 10 March. A map of the city was published showing the extent of an 8.7 m flood. BOM was predicting 8.6 m on March lO-not a major difference. Flood will be smaller than 1983 (less than 8.25 m). This varies substantially from BOM prediction of 8.6 m, but the paper's front page carried the official prediction and council flood map. On a technical note, the fact that the one day error in flood peak arrival time persisted until late in the forecasting period, in a well instrumented catchment, suggests that improvement does not lie in further upstream instrumentation. Better understanding of estuarine river behavior is required.

PAGE 184

FLOOD WARNING SERVICES MARKET SURVEY STUDY FOR MARICOPA COUNTY, ARIZONA Laurie T. Miller and Fred K. Duren, Jr., James M. Montgomery Consulting Engineers J.M. Rumann Flood Control District of Maricopa County Introduction Maricopa County is located in central Arizona and encompasses more than 9,200 square miles, including the Phoenix metropolitan area. The Flood Control District of Maricopa County provides flood control facilities in the unincorporated portions of the county and participates in storm drainage projects that traverse multiple jurisdictional boundaries. The district also operates a network of 119 precipitation and 44 stage gages, some of which are implemented through intergovernmental agreements with local communities, the U.S. Geological Survey (USGS), and the U.S. Army Corps of Engineers. The telemetered gages transmit data to the district, which shares information with the National Weather Service (NWS). Climatic conditions in central Arizona produce intense storms that can cause flash flooding, associated property damage, and serious injury or loss of life. The district is considering more fully developing the data collection system and improving flood warning services as a means of nonstructural flood control. As a result, the district retained James M. Montgomery, Consulting Engineers, (IMM) to perform a Flood Warning Services Market Survey Study to evaluate the feasibility of establishing a comprehensive flood warning network within Maricopa County. The study includes the following major elements: 1. A system evaluation to assess current and future flood warning technology and other technologies applied to flood warning; 2. A market survey within Maricopa County to evaluate the local understanding of flooding and flood warning services and the perceived need for such a system; 3. A benefit/cost analysis of up to six flood warning scenarios; 4. A recommendation of three alternative programs to meet the flood warning needs of the county. Project completion is scheduled for October 1991. This paper discusses some of the interim findings of the study. System Evaluation Several different types of technology were investigated for their applicability to flood warning in Maricopa County. The system evaluation of current technology addressed precipitation and stage gages, weather forecasting, NWS radar, and Geostationary Operational Environmental Satellite (GOES) data. Future technology

PAGE 185

Miller, Duren, and Rumann 161 evaluation included Doppler Next-Generation Weather Radar (NEXRAD), mesonets, GOES-NEXT data, and wind profilers. Geographical information systems (GIS), hydrologic modeling, and lightning detection were other technologies investigated for application to flood warning. Flood warning systems currently in use across the U.S. were also reviewed. It was found that many systems had not been fully operational long enough to have been tested by a major flooding event. Further, many systems acted as flood detection systems and fell short of providing warning dissemination. Of those that did provide dissemination, flood information was provided to local officials to be used at their discretion. Market Survey The objective of the market survey was to assess the communities' understanding of and perceived needs for flood warning services. To obtain the desired informa tion, JMM prepared three separate questionnaires for distribution to 250 individuals from three categories of participants based in Maricopa County. The categories and number of participants included managers and administrators (50), technical staff and users/implementors of flood warning services (100), and homeowners (100). The managerial group was composed of mayors, city council members, heads of agencies, and others in decision-making or policy-setting roles. The technical group included city engineers, police and fire chiefs, public works directors, and others who would be involved in flood warning services within the county. Also part of the technical group were potential users of flood warning services, such as utilities, transportation companies, and the media. The homeowners' group was based on population density and on geographic representation throughout the county, with each community receiving at least one questionnaire. Participants were selected from membership lists of homeowners' associations where available and by consulting officials within the community for recommendations of homeowners to participate in the survey. The questionnaires for the market survey were distributed in April 1991. As of this writing, approximately 50% of the 250 questionnaires have been returned and are being evaluated according to selected criteria. Each question was categorized into one or more of the evaluation criteria. Group response to each question is being evaluated for each of the three survey groups. The responses will then be averaged by category to estimate the total response level for each of the three groups. To illustrate the data collected, an initial evaluation of level of interest in improved flood warning services is presented in Figure 1. As shown in the figure, there is a high level of interest in improving flood warning services within Maricopa County. To check the reliability of this response, certain factors, such as perceived flood threat and adequacy of the existing services, were compared to measure the consistency of response for interest in improved services. For example, strong agreement that flooding is a threat in the county and that current services are

PAGE 186

162 Interest in Irrproved FWS Perceived Urban Flood Threat Perceived Rural Flood Threat Flood Warning Market Survey end Managerial rsI Technical Horreowners Perceived Perceived No Opinion of Exist. FWS Exist FWS Exist. FWS Adequate Inadequate Adequacy Figure 1. Interest in improved flood warning services (compared to other responses). inadequate would substantiate a high level of interest in improving services. Results from these categories are also shown in Figure 1 for comparison purposes. A consensus exists that Maricopa County is threatened by flood potential, which would support interest in improved services. The response level of perceived adequacy does not show as strong a correlation. The response was much more evenly distributed among perceived adequacy, inadequacy, and no opinion selections. However, the perception of inadequacy of the three survey groups, ranging from 28% to 35%, is substantial. It was also found that there is an interest in providing more geographically specific flood warnings. Further, a desire was identified to integrate the data collection systems operated by the district, NWS, USGS, and Salt River Project, among others. The questionnaires and responses have been entered into a data base that is being used for evaluation of responses. Response summaries can be generated for numerous subgroups within each target survey group to compare responses. Data obtained from the market survey are being augmented by in-person interviews with representatives of selected organizations. Benefit/Cost AnalysiS The benefit/cost analysis will be performed for several flood warning program alternatives selected as a result of the system evaluation and market survey findings.

PAGE 187

Miller, Duren, and Rumann 163 The direct benefit approach and development of Day curves are two methods to be used in performing the benefit/cost analysis. Recommendations Three levels of flood warning services will be recommended to the district for potential implementation. The recommendations will be based on the communities' needs as identified in the market survey, on the practical application of available technology, and on economic considerations. The alternatives will reflect varying levels of funding, services provided (e.g., geographical extent, lead time, accuracy), staffing requirements, and feasibility. Summary Preliminary findings of the Flood Warning Services Market Survey Study indicate that there is strong support to improve flood warning services within Maricopa County. The district wishes to increase the usefulness of its existing data collection system and is committed to assuming an active role in flood warning. That role and the means to assume it will be defined from the results of this study.

PAGE 188

USING APPROPRIATE FLOOD WARNING TECHNOLOGY FOR COMMUNITIES AT RISK Mark E. Nelson U.S. Army Corps of Engineers The Scribner Experience An alternative to sophisticated electronic flood warning systems was needed in order to meet the design requirements of the Scribner Flood Control Project in Nebraska. A market search located no system that would provide the combination of flood warning time, reliability, simplicity, and economics that we needed. We then developed a flood warning system to satisfy the project requirements. The major drawback of the commercially available systems, for our application, is that they are too difficult and costly for small rural communities, such as Scribner, to operate and maintain. The project required a simple flood detection system that would provide enough warning time for Scribner to erect a levee closure structure across a highway before flood water from Pebble Creek reached that location. Many of the features of the commercially available systems, such as ALERT, were not needed in order to provide the required warning time. We had several objectives in mind as development began on an alternative flood warning system. First among the objectives was that the flood warning solution had to be at an appropriate technology level. Once turned over to Scribner, it had to be affordable, durable, and easy to operate and maintain. Low initial cost and local availability of spare parts were also important objectives. The design that evolved featured a combination of equipment, a hydrologic model of the stream, and a plan for community involvement. Equipment The equipment consists of stage warning devices that are simple by comparison to an ALERT stage sensing gage. A stage warning gage consists of a telephone alarm dialer in a shelter atop a stilling well, which contains float switches mounted on a vertical rod. The telephone alarm dialer plays prerecorded flood warning messages to individuals designated to receive them. The alarm dialer is activated by the float switches. Two separate message channels are available on the alarm dialer: one for burglary and one for fire. The lower float is connected to the burglary circuit and the upper float is connected to the fire circuit. The use of two floats, generating separate warning messages, permits a rate of rise to be determined. The computed rate of rise can be compared with rates of rise characteristic of approaching floods, as defined in the hydrologic model. The middle float opens the lower float circuit, enabling the upper float alarm to transmit. Rechargeable batteries power the alarm dialer and buried phone lines carry the warning messages to preassigned city and law enforcement personnel via the local phone system. A staff gage, mounted near the stilling well, provides visual

PAGE 189

Nelson Float CabJe$ 1 1 1 Ie' CMP----l 1 Initial AJcrrn Shur OF. FJDDr 5witcn lruba! Alarm AdJva!uy, Floar S,.,ifr;h 1 1 'I 1 1 I Will! Waf I J InitIal Alarm I I 1 ... J ;----rCleDnOLlI Door J: r r: ScI5t:.;:IQIiIII I u____ I Sk:1qr! : U I lJoIr I OMAHA DISTRICT FLOOD WARNING GAGE Figure 1. Stage warning gage. 165 confirmation of stream stage and can be used in between alarm transmissions to estimate the rate of rise. A drawing of a stage warning gage is shown in Figure 1. Hydrologic Model The hydrologic model employed in the flood control project design was used to analyze flood warning time. With the model, it was determined that two stage warning gages in the Pebble Creek Basin would provide adequate warning time to close the levee and that automated rainfall detectors were not necessary. The hydrologic model was also used to develop characteristic hydrographs for the watershed, using different storm intensities and orientations. The hydrographs were used to define critical rates of rise that would be used to identify the approach of serious flooding. The rate of rise information was then consolidated into simple lookup tables that were listed in the project operation and maintenance manual. No computer is required to operate the flood warning system, once installed.

PAGE 190

166 Appropriate Warning Technology Community Involvement Although mentioned last, the community involvement component is extremely important to the long-term success of the flood warning system. Community involvement includes the flood response plan and a plan to test, operate, and maintain the system. Community involvement should start early in the project development phase. It should involve community leaders, those who will maintain the system, and county and state officials involved in civil defense. A prototype flood warning system was built by employees of the Omaha District and Scribner in the spring of 1989. All parts for the system were purchased from suppliers in the Omaha area. By late spring 1989, the prototype system was on line awaiting its first test. The test came early on the morning of September 7, 1989, as heavy rains crossed the northern third of the watershed. Both the lower and upper floats of the upstream warning gage were engaged by the rising waters. Key people in the community were alerted to the upstream flooding and they drove out to monitor the situation and close low roads before they went under water. Since the heaviest rain was concentrated in the upper part of the watershed, the flood attenuated before it reached Scribner and it was not necessary to close the levee. As a result of that timely test, modifications were made to both the equipment and the community flood response plan. The following June, the system was given another operational test during widespread flooding. Portions of the basin experi enced up to a 50-year flood, and both flood warning gages delivered alarms. This flood resulted in a full scale test of the flood response plan. Following the first alarm from the upstream gage, contact was made with the National Weather Service in Omaha. They took the report of flooding and gave the town a forecast of more heavy rain. The town then activated its emergency operations center and made preparations to close the levee. At its peak, flood water rose to within 1.5 feet of the stage necessary for closure and surrounded the town. Other Communities A second low cost flood warning system was installed near Emerson, Iowa, as a replacement for an ALERT system that had failed several years earlier. The replacement system consists of a single automated flood stage detection gage and an observer network. The flood warning gage was installed in June 1990 and supplements the observer network. An operation and maintenance manual, with a hydrologic model and a community response plan, was developed for the Iowa town. The flood detection gage used for the Iowa community was a modified version of the one used in Nebraska. The gage was banded to a concrete bridge pier without drilling into the prestressed concrete. Barely six weeks after it was built, it received its first operational test. On July 25, 1990, heavy rains battered southwest Iowa, with over six inches of rain falling on the basin in less than 24 hours. City officials were alerted by the device and monitored the flood as it rose to within six feet of overtopping a bridge adjacent to

PAGE 191

Nelson 167 the levee. The flood provided a good test at a time when the community flood response plan was still under development. An additional flood warning system of this type was built at Cold Brook Reservoir, a Corps facility upstream of Hot Springs, South Dakota. Its purpose is to warn of uncontrolled releases through the principal spillway. The flood detection gage was built along the reservoir shore and integrated with the existing reservoir gaging station. The lower float was set so that the lower alarm would trigger before any water was released from the dam. The upper alarm was set to activate before the discharge was capable of causing property damage in the housing area immediately downstream of the dam. So far, the basin has not experienced flooding since the warning gage was built. Lessons The Omaha District's experience with both types of warning systems has provided important lessons. Among those is when to substitute simpler technology in place of sophisticated systems, such as ALERT. Factors that should signal the designer not to use a system requiring a computer base station include: 1. The absence of a facility with an uninterruptable power supply and 24 hour per day staffing. 2. The lack of city staff capable of maintaining the base station computer and software. 3. The lack of an adequate budget for continued operation, maintenance, and replacement of expensive critical components. 4. The lack of community support for a complex system. The process for selecting equipment and designing the system must be subjected to the same methodology as other engineering decisions. In order to be a useful solution, the flood warning system should prove itself cost effective, durable, and appropriate for the community.

PAGE 192

MANAGING FLOOD WARNING SYSTEMS: THE UNITED KINGDOM EXPERIENCE Dennis I. Parker and Sylvia Tunstall Flood Hazard Research Centre Middlesex Polytechnic, United Kingdom Introduction The potential benefits of flood forecasting and warning systems have long been recognized, but the United Kingdom (U.K.) is facing up to the problems of managing flood warning systems to perform effectively to convert potential benefits into actual ones. This paper examines recent U.K. evidence on the performance of flood warning dissemination systems and briefly addresses related issues. The problem ofunderperformance is also likely to exist in the U.S.: comparison may be fruitful. U.K. Flood Defense Policy and Flood Warning Systems The U.K. 's flood defense policy is both stmctural and nonstructural. The major threat to life is from sea flooding. Since the devastating 1953 floods, reliance has been on extensive sea defenses coupled with a storm tide warning service. Many sea defenses are reaching the end of their design life and are in a poor state of repair. Some sea defenses are being replaced, but increasing reliance is being placed upon flood warning systems. During 1990, in coastal Towyn in north Wales, 6,500 people were evacuated following the breach of the sea wall. Unfortunately the flood warning system showed major failings (Welsh Affairs Committee, 1990). However, the most frequent flood damage occurs in river floodplains where city, town, and village encroachment is centuries' old (penning-Rowsell et al., 1986). The U.K. is densely populated; land use pressures are intense, especially near major cities; and, while there is a universal development control system, riverine floodplains have become developed. The choice is often between developing greenbelt or floodplain land, or between higher density flood-free developments or floodplain developments. In many cases, therefore, planning authorities have permitted floodplain development, especially as the U.K. 's rivers are more docile and relatively easily embanked. The ratio of normal river channel capacity to the discharge of large floods is generally greater in the U.S. than the U.K. The U.K.'s rivers are generally short and flashy, but nevertheless flood forecasting and warning systems have been developed. The new National Rivers Authority (NRA) is seeking ways of improving the geographical coverage and effectiveness of these systems. In the U.K. the NRA is the flood forecasting agency. Usually flood warnings are passed by the NRA to the police, who disseminate the warnings to local authorities and to floodplain users. Rainfall-runoff models are the basis of most riverine flood forecasting systems, but during the past 20 years potential warning lead times have improved dramatically by the development of rain radar and surge modelling for coasts and estuaries. Satellite imagery offers the possibility of further advance.

PAGE 193

Parker and Tunstall 169 Assessing Performance Measures of flood warning system performance are discussed in detail elsewhere (Neal and Parker, 1989; Parker and Neal, 1990). A crude measure that permits no measure of the quality of system performance is the proportion of floodplain properties in a given area served by a flood warning service. A more sensitive indicator might be estimated flood damage avoided by flood warnings, but this is difficult to estimate. The approach used in our recent research has been to identify by interview survey the proportions of floodplain users who recognized that they received flood warnings, to identify flood warning lead times experienced by those warned, and to measure the level of satisfaction felt by potential and actual flood warning recipients with the service. This approach measures the total performance of the entire flood warning dissemination system as viewed by the end user. Surveys During the mid 1980s one U.K. Water Authority (since 1989 part of the NRA) commissioned the Flood Hazard Research Centre (FHRC) to investigate the effectiveness of its flood warning service through interview surveys. The initial study focused upon the water authority's operational staff. This was followed with further investigations focusing upon potential and actual warning recipients. In 1991 the NRA commissioned the FHRC to undertake a national investigation of flood warning levels of service. This paper focuses upon some of the results from the completed surveys in riverine floodplains. First, some 650 interviews were completed between September 1986 and September 1987 in floodplains within three catchments of the Severn Trent region of the NRA. There were 12 floods and numerous warnings during this period, including a 20-year event and a l00-year event. Second, nearly 200 interviews were completed in the River Thames Valley to the west of London following a 5-year event in January 1990. Finally, 343 interviews were completed in the Severn Valley after a 15to 40-year event there in January and February 1990. Survey Evidence Full details of the research are published (Neal and Parker, 1989; Tunstall et al., 1991). The key results are for floodplain inhabitants and businesses either served by or, in a minority of cases, might be served by NRA flood warnings (Figures 1 and 2). During 1986-87 the warning system displayed variable performance in the Severn Trent region. The Upper Severn catchment system had the most satisfactory performance, while the Upper Trent catchment received a geographically limited lOO-year flood, but without a flood warning being issued. In the 1990 floods the proportion not warned but flooded was lowest in the Middle Severn, whereas in both the Lower Severn and Estuary approximately 50% were flooded and not warned. The equivalent percentage in the Thames Valley was much lower.

PAGE 194

170 '------3n .IlION "88-7 ..... Managing Flood Warning Systems FIGURE . PERFORMANCE OF FLOOD WARNING SYSTEMS .. ... MODt.E SVERN ... m .' lPPER SEVERN me-7 ..., .. .. .. '"' ESTUARY 1WO ... JO uPPR TREJtfT "88-7 ..."'" ." -_/ ." 'THAN(S WJ.EY 11180 ...... NOT 'M.RHED FLOODED ""RHEa FLooceD MRNED NOT FLOOCEO II NOT _RHED NOT FLOODED FIGURE 2. LEVEL OF SATISFACTION WITH THE PERFORMANCE OF THE FLOOD WARNING SYSTEM BY AREA NOT AT ALL SATISFIED fi.iJ QUITE SATISFIED NOT VERY SATISFIED CJ COMPLETELY SATISFIED THAMes ""LLEY,,,,,O Figure 2 shows the levels of customer satisfaction with the flood warning service in each area. With the exception of the 1986-87 Upper Severn catchment floods, all other areas show a relatively high degree of customer dissatisfaction with the flood warning service. Most of this dissatisfaction is due to customers not receiving a warning at all. The number of interviewees requesting longer flood warning lead times was high. Key Lessons Learned and Management Issues The interagency flood warning dissemination process is cumbersome: our expectations of what constitutes its effective performance have lowered. With notable exceptions, the NRA's flood forecasting performance is sound, as is their passage

PAGE 195

Parker and Tunstall 171 of flood warnings to the police, though warning lead times sometimes remain short. Dissemination decays thereafter and valuable lead time is lost. Direct warning dissemination from the NRA to floodplain users is preferable, but so far infeasible on resourcing grounds. Now the interagency warning system is breaking down because none of the U.K. warning dissemination agencies has a statutory duty for providing flood warnings. The NRA has a permissive power only. Under tightening resource constraints, the police and local authorities are less willing than previously to accept a role in flood warning dissemination. Legal liability problems for warning failures are of growing concern. Faced with these problems, the NRA is considering what levels of flood warning service (including none) it should provide in different situations. References Neal, Jeremy, and Dennis J. Parker 1989 Flood Warnings in the Severn-Trent Water Authority Area: An Investigation of Standards of Service, Effectiveness and Customer Satisfaction. Geography and Planning Paper No. 23. Enfield: Middlesex Polytechnic. Penning-Rowsell, Edmund C., Dennis J. Parker, and Don M. Harding 1986 Floods and Drainage: British Polides for Hazard Reduction, Agricultural Improvement and Wetland Conservation. London: Allen and Unwin. Parker, Dennis J., and Jeremy Neal 1990 Evaluating the Performance of Flood Warning Systems. In John Handmer and Edmund Penning-Rowsell (eds.) Hazards and the Communication of Risk. Brookfield, Vermont: Gower Publishing Co. Tunstall, Sylvia M., Dennis J. Parker, Greg Twomey, Ann Doizy, and Emma Sangster 1991 The Severn Floods of January and February 1990. Enfield: Middlesex Polytechnic, Flood Hazard Research Centre. Welsh Affairs Committee 1990 The Breach of the Sea Defences of 26/27 February Along the North Wales Coast Report together with the Proceedings of the Committee, Minutes of Evidence and Appendices. London: HMSO.

PAGE 196

LOCAL FLOOD WARNING SYSTEMS IN NEW JERSEY Robert D. Schopp and Rebecca J. Burns U.S. Geological Survey Introduction Local flood warning systems in New Jersey have changed significantly since 1988. Record-breaking floods in the 1970s and early 1980s encouraged the National Weather Service (NWS) and local officials to find ways to improve flood monitoring and warning. As a result of this interest, telemetry systems have progressed beyond the older, telephone-based systems to include UHF and VHF terrestrial radio and government and commercial satellite. Telemetry in New Jersey Before 1988 Flood warning telemetry in New Jersey was first used in 1960 and was installed and operated mainly by federal agencies, such as the NWS and the U.S. Geological Survey (USGS). The first local flood warning system in New Jersey was installed in Somerset County in 1978 by the USGS. The system was telephone-based and included eight streamflow-gaging stations equipped with rain gages. Passaic River Flood Warning System The second local flood warning system in the state was the Passaic River Flood Warning System (Figure 1), which was completed in 1988. A flood in April 1984 in the Passaic River basin caused more than $350 million damage in the basin and displaced approximately 6,000 people (World Water, 1988). The Passaic River Flood Warning System uses VHF radio to transmit data from 31 self-reporting rain gages. Streamflow-gaging stations at 18 sites report by use of data collection platforms (DCP) through the Geostationary Operational Environmental Satellite (GOES) and telephone. A total of 11 computer base stations were established at four county, two state, and five federal offices. The base-station computers are linked by UHF terres trial radio and/or a commercial satellite (Figure 2). Hydrologic and meteorologic data, as well as weather forecasts, watches, and warnings, are exchanged among all base stations. Flood watches and warnings are automatically transferred to 15 high risk municipalities by the county base-station computers. Flood inundation maps were prepared for 26 communities in the basin to aid in understanding what areas would be affected by flood warnings transmitted by the system. S M ** lerraisco ,Incorporated's Enhanced ALERT with IFLOWS Interface software package is used at the base stations. (ALERT stands for Automated Local Evaluation in Real Time, and IFLOWS stands for Integrated Flood Observation and ** Use of brand or company names in this report is for identification purposes only and does not constitute endorsement by the U.S. Geological Survey.

PAGE 197

Schopp and Burns .,. 39' EXPLANATION PrlncipaJ city 10 20 30 MILES o 10 20 :30 KILOMETERS FlQure 1. Map of the areas covered by local floocHwaming systems in New Jersey. 173 Warning System; both of which are NWS computer packages). The rain gages, soft ware, and communications equipment were installed by Sierra-Misco under contract with the U.S. Army Corps of Engineers (USACE). The New Jersey Department of Environmental Protection acted as the local sponsor of the project. The USGS purchased and installed Handar DCPs for 18 of the streamflow-gaging stations included in the project and Sierra-Misco radio-reporting telemetry for three flashy urban streams. Somerset County Flood Monitoring System Somerset County upgraded its flood warning system and tied or integrated into the Passaic system in February 1990. The upgrade included a conversion of the eight previously established gages to VHF radio telemetry and the addition of 11 rain gages and six streamflow gages. Because of the topography of the county, two VHF repeaters were needed to transmit the sensor readings into the county base station.

PAGE 198

174 Local Flood-Warning Systems Commerical salellit. o-------------+ Be,oen Counly f\ ""'" / 1111\ Mwl. tSi]\@ I 1\ 11\1 NWS I II P .... i. U '\ Now Yorl< I I County I II I II En .. @ @ I I I County U NWS I \ \ Newark I I \ \ /@som""Olcounty NWS I I Harrisburg I \ \ I I I I I I I I USGS Trenton I NJ 1"1 I Police \ I U Me",o' County I i I SU NJOEP EXPLANATION I I --Satellite-radio path I --UHF-radio path +1 (';:\S Satellite-equipped computer \V bue station f5'\ NWS ti} UHF-radlo-and satellite-\V Philadelphia eQuipped computer base station fu\ computer \:=.J base station ill. UHF-radio repeater Figure 2. Communications networks used in New Jersey fIood-waming systems. The base station is linked with the other base stations in New Jersey by UHF radio. The upgrade was designed, installed, and is operated by the USGS in cooperation with Somerset County. Assunpink Creek Flood Warning System Assunpink Creek and its tributaries are located near Trenton and have caused considerable flood damage during the last 100 years. The USACE, asked to study flood problems in the Trenton area and propose solutions, recommended a flood warning system. The USACE prepared a preliminary design (USACE, 1990) and the USGS prepared the final system design. Installation of the system is scheduled to take place in 1992. The system will include a base station, eight new rain gages, one weather station, and two new river-level sensors in Mercer County. The base station will be tied to the other base stations in the area by UHF radio (Figure 2).

PAGE 199

Schopp and Burns 175 Proposed New Jersey Tidal Telemetry System The USACE is planning to study the feasibility of a tidal flood monitoring system along the Atlantic Coast of New Jersey. This system would include radio reporting tide-level sensors at critical bridges along the Intercoastal Waterway. Critical bridges are those used in the evacuation of the tourist areas on the barrier islands along the coast. Several radio-reporting weather stations also will be placed along the coast to improve the NWS's ability to predict tidal flood levels and issue timely calls for emergency evacuations. Five additional base stations are planned that will interconnect with the present UHF radio "backbone" communication network. Problems Encountered The principal problem encountered in the operation of the various systems has been maintenance. The cost of maintaining the equipment has been higher than initial estimates. Running computers 24 hours per day resulted in the failures of all of the color video monitors and 40 % of the hard disk drives in the first three years. Another problem is the difficulty involved in setting up a contract that will cover maintenance of the entire system. The software used in New Jersey flood warning systems is a modified version of the standard NWS ALERT software. Only the firm that modified the software is in a position to optimally maintain it. Splitting the software and hardware maintenance contracts would be one solution but could lead to "fingerpointing" when a difficult problem is encountered. The satellite data link, however, has proven to be very reliable. The contractor providing the service monitors the system continuously and frequently notes problems and solves them without being called. The one disadvantage of the satellite service is its high cost, which may amount to tens of thousands of dollars annually. Reducing the amount of redundant data sent over the network would greatly reduce the transmission costs. Conclusions Local flood warning systems in New Jersey have improved the ability of emergency management agencies to respond to floods. Rectification of maintenance problems would improve the reliability of the systems. References U.S. Army Corps of Engineers (USACE) 1990 Flood Control Study, Reconnaissance Report and Environmental Assessment-Assunpink Creek Basin, New Jersey. Philadelphia District. World Water 1988 Flood Warning Gives New Role to Corps. June.

PAGE 200

COMMUNITY ALERT NETWORK: DAM SAFETY AND FLOOD WARNING Steven E. Smead Community Alert Network Introduction This paper will present the latest advances in telephone notification technology, the capabilities of Community Alert Network (CAN), and the potential applications for flood warnings in your jurisdictions. There is no question that releases, spills, or breaches of local dams represent very large and very real potential dangers for residents of the floodplains throughout the United States. Historical evidence of spills or releases at dam sites, particularly during heavy rains, indicates the amount of people that must be notified of impending or potential flooding that could occur in their neighborhoods is very large. This problem has increased recently with industrial and residential developments taking place so close to known floodplains. Problem Solution Public safety officials and the operating organizations that manage local dam sites must have a quick and reliable method of notifying response teams, utility officials, public safety officers, and the media, as well as informing residents of floodplains, with particular emphasis on special facilities, in the event of a flooding emergency. The manner of notification must both enhance the crisis management procedures utilized by officials and promote public confidence in the overall safety of their communities. CAN is strategically positioned to play an active role in assisting both public safety and dam-site officials in meeting today's comprehensive community alerting and notification needs for a flooding emergency. CAN provides a unique computer driven telecommunications service designed to assist government agencies, organizations, and communities when emergency notifications must be disseminated to large numbers of people in a short period of time. The services of CAN are a tool to be added to the emergency warning capabilities of emergency-management and public-safety officials for delivering critical information by telephone to individuals in affected areas, and has saved lives in the past. CAN not only saves lives, but valuable time, personnel, equipment, and money, while allowing officials to concentrate on the emergency without having to tie up personnel, equipment, and their own phone lines. Operation Overview CAN contracts with companies or communities to provide emergency notification services to specified individuals or residents of a specific floodplain area. Once the contract has been signed, it takes CAN approximately 60 days to receive, process, and store the data and be ready to serve the needs of the client.

PAGE 201

Smead 177 The client provides CAN with the names, addresses, and telephone numbers that are to be installed in the CAN data base.-The data base can be grouped as one or divided into many different sublists or unique calling groups. For example, list number one may be the local media, list number two could be the residents of floodplain number one, list number three could be public safety officials, list number four could be additional affected residents of the floodplain area, and so on. CAN maintains the confidentiality of all such information. An overview of the operation of CAN is the following: an incident occurs, an authorized official is informed and calls CAN to detail the emergency. CAN computers then dial the phone numbers in the requested targeted areas and/or of specially identified individuals to deliver the emergency message. Detailed Operation Once CAN is in place, the services are available to help inform people about a wide range of impending emergency situations that could stem from any flooding or dam site emergency. A message could be issued with instructions for residents to begin to take precautionary measures, including moving personal property to safer locations. A second message could be issued that would instruct residents or emergency response personnel to begin to put flood shields in place and place sand bags in appropriate locations. And finally the message to evacuate to higher ground can be delivered. CAN is capable of delivering any unique, detailed message to those who need to know about an impending or actual emergency. These series of messages to residents and individuals in floodplain areas could help reduce the potential for lost lives and property claims against flood insurance programs. Let's take a step back to the beginning. After the appropriate ofticial is informed of the emergency, a predesignated person calls the CAN 24-hour emergency phone number and dictates a password or access code to the CAN operator. The CAN operator is trained to receive and input all information as well as to ask questions to ensure that optimum emergency preparedness and public awareness are achieved. The authorized caller identifies which of the specially predetermined list(s) of telephone numbers are to be called and dictates the message(s) to be delivered. Some messages may have been prerecorded and simply need to be identified, others are recorded in real time and different messages can be delivered to different groups simultaneously. While it is important to make sure that all affected individuals are notified, the services of Community Action Network can be focused to alert only those people who have been or will be affected. The computers at CAN begin dialing the requested phone numbers and delivering the designated message(s) via the fastest long distance transmission and delivery system available from our operating centers. The phone rings at a designated number and, when answered, receives the recorded message. The CAN computers will make three initial attempts to contact every identified phone number to deliver the emergency message. At the end of the calling session a report is printed and the summary information regarding the number of calls made, messages delivered, busy signals, and no answers is shared with the

PAGE 202

178 Community Alert Network client. Any client may request that additional calls be made to previously busy or unanswered phone numbers or, as is frequently the case, to issue an updated message to people in the affected areas. The final report of all calling activity is delivered to the client by first class mail or fax, if previously arranged. Summary This service has been successfully used in times of floods, hurricanes, hazardous material accidents, and many other emergency situations. The system has saved lives, response teams are able to be contacted immediately, missing persons have been located, and critical information has been disseminated to thousands of affected individuals and residents during emergency situations. The services of Community Alert Network can be part of your jurisdiction's comprehensive emergency notification plan for as little as $3,500 during the first contract year and $2,500 in the following years. The services of Community Action Network can help make your emergency response plans more comprehensive and help you execute your flood warnings more effectively, thus helping to make your community a safer place to both live and work.

PAGE 203

SUMMERTIME PRECIPITATION IN THE lllGHER ELEVATIONS OF COLORADO Larry Tunnell National Weather Service There is currently considerable controversy concerning high elevation summer precipitation in Colorado. For the purposes of this paper, high elevation means above around 8,000 feet and summer means from June 1 through September 30. This is not an entirely academic question since various design criteria (for dams, culverts, etc.) involve the maximum rainfall expected for a given time in a certain location. Many people take offense at paying taxes for constructing a culvert that will handle a rainfall event that only happens every 1,000 years or so, or a dam that will handle a precipitation event that only occurs every 10,000 years or so. After the Big Thompson flood, there was considerable talk along the lines that this could have happened in any drainage and that we should adjust our design criteria to reflect this. Bob Jarrett of the U.S. Geological Survey (USGS) and others have indicated that, based on paleohydrologic evidence, this was an event that had not occurred in this area for between 10,000 and 40,000 years. Can we afford the price we must pay to design for all possibilities? The purpose of this paper is to give some idea of the historical basis for summertime rainfall in Colorado above around 8,000 feet. This rain is produced mainly by showers and thunderstorms that begin forming in the mountain areas in middle or late morning or early afternoon and continue in some areas until after dark. The National Weather Service operates around 12,000 Cooperative Weather Stations throughout the United States. These provide various information to supplement that from the offices of the National Weather Service. These Cooperative Observers, generally unpaid, furnish various information. Most furnish 24-hour precipitation at the site, along with maximum and minimum daily temperatures. There are currently over 200 of these observers in Colorado. Around 65 Cooperative Stations in Colorado furnish hourly precipitation. Other precipitation information is obtained from research sites, snow stations operated by the U.S. Soil Conservation Service, and stations operated by other entities, such as the Colorado State Engineer, the USGS, and the Bureau of Land Management. This paper will address data from some of these sources. No one seriously disputes the fact that high elevation stations in Colorado get less summer precipitation than those at lower elevations. One of the reasons for this is that there are only two basic moisture sources for summertime precipitation in Colorado: the Gulf of Mexico and the Pacific Ocean. The few air masses that enter the state from the west and north during the summer are generally dry and do not contain sufficient moisture to produce heavy rains. Air from the Gulf of Mexico, which is the source of much of the moisture in Colorado, loses up to three-quarters of its moisture by the time it moves up into the higher elevations in Colorado. Thus, we know that air masses that affect the Colorado mountains in the summer are drier than those at lower elevations. Another thing we know is that the storms

PAGE 204

180 Summertime Precipitation that do form are smaller than those at lower elevations. For instance, during the summer of 1972, upper air soundings were taken at Leadville, at an elevation of around 10,000 feet, each morning at 11:00 a.m. The study also included measure ments of cloud size and height of cloud bases. Results from this study indicated that the clouds were almost always totally above the freezing level, signifying that the ice phase process was the main process involved. It is very difficult to produce heavy rain in these circumstances. Another finding from this study was that the average precipitable water in the air was only around a third of an inch, compared to over an inch at Denver for the same period. Also, average thunderstorm cloud size was only around one square mile, indicating that rain produced by the thunderstorms would not cover a very large area. This contrasts with lower elevation thunder storms, which tend to be much larger. By the time summer air masses, from whatever origin, move into the north western portion of Colorado, they are almost always tired and uninspiring. This is borne out by a study of 33 years of precipitation from 13 stations in northwestern Colorado during the period from June 1 through September 30. Elevations of the stations involved in this study varied from 5,290 to 8,010 feet. Findings of this study indicated the following: 1. There were 52,338 days of 24-hour precipitation available. 2. On only 99 days (0.2 %) did any location receive 24-hour precipitation equal to or greater than one inch. Many of these cases were in general storms after September 10, which would produce mainly snow at higher elevations. Otherwise, the highest 24-hour amount was 1.98 inches, and only seven times was 24-hour precipitation equal to or greater than 1.5 inches. The Fraser Experimental Forest and Range Experiment Station is located approximately five miles southwest of Fraser at an elevation of9,070 feet. A 4O-year study of precipitation at this station during the months of July and August was done in part to determine the maximum daily precipitation at this location during these months. The maximum daily precipitation in July varied from around a quarter of an inch to a maximum of less than 1.5 inches. For August, it varied from less than a quarter of an inch to around 1.2 inches. Maximum monthly totals for these 40 years was around 4.3 inches for July and 4.4 inches for August. The Soil Conservation Service operates a network of over 150 SNOTEL sites in Colorado. These are locations at which various meteorological parameters are sampled automatically, including 24-hour precipitation. Data from this system was analyzed for 37 of these sites during the period June I through September 30 for the years 1979 through 1988. Elevations of these sites varied from 8,250 feet to 11,550 feet. Results of this study can be summed up as follows: 1. There were a total of 295 summers at 37 SNOTEL sites involved in the study. 2. There were two instances of peak 24-hour precipitation equal to 2.6 inches. All others were less, with 12 instances of peak 24-hour precipitation of two inches or more (12 out of 295, or 4%).

PAGE 205

Tunnell 181 3. The median peak 24-hour precipitation at the 295 stations was 0.9 inches. During a nine-year period (1966 through 1974), Colorado State University supported a study of hourly precipitation on profiles of four mountain pass areas in central Colorado. All but one were above 8,000 feet and four were above 10,000 feet. For our purposes, the results can be summarized as follows: 1. The data in the study covered 10,850 sampling days. 2. There were only 152 cases where 24-hour precipitation at a station equaled or exceeded .3 inch. This amounts to 1.6 cases per station per year. 3. There were only 19 cases where six-hour precipitation was equal to or greater than one inch. 4. Of these 19 cases, four occurred in a general storm in late September. Of the other 15, the highest six-hour amount was 2.26 inches. All other six hour amounts were 1.44 inches or less. The above may come as a surprise to those who are familiar with HMR 55 and 55A, Probable Maximum Predpitation Estimates Between the Continental Divide and the 103rd Meridian, in which such figures are given as a 24-hour PMP of over 12 inches for the Leadville area, with greater than seven inches in six hours. There is evidence that these figures are too high. If so, why? One possibility that comes to mind is that some of the estimates that are produced by transposing storms occurring in, for example, Montana to Colorado and assuming similar precipitation rates may not be justified. Another may be that some of the historical data used may be sus pect. There is not space in this paper to go into the first possibility. We can, however, discuss the second. In July of 1937, 4.34 inches of rain was reported at Leadville in a 24-hour period. This was an interesting and somewhat humorous case that has been studied in detail by Loren Crow, and this paper will only summarize some of the results of his research.: 1. The precipitation measured during this storm was done with an "experi mental" gauge that resembled a front-end loader turned on its end. 2. This type of gauge was used to measure precipitation at Leadville during the period August 18, 1918, through February 8, 1939, probably to determine if a better gauge for measuring snow could be developed. 3. Unfortunately, the effect of this gauge during hailstorms, which occur frequently during the summer in Leadville, was to collect the hailstones in the manner of a front-end loader, which held them until they melted, funneling the moisture into the weighing apparatus at the bottom. This tended to greatly overestimate the actual water equivalent in many instances. The above is an indication of the care that must be exercised in drawing conclusions based on precipitation data. The data from the Colorado Cooperative Network comes through the office of the author before being forwarded to the National Climatic Center, and he can vouch for the fact that some is very questionable, whether from misplacing decimal points, making illegible entries, or

PAGE 206

182 Summertime Precipitation other factors. Much of this faulty data goes into the Climatic Records and is used in various studies without the careful screening that is necessary if the studies are to be of value. The point of the above is not to attempt to convince the reader that heavy rains do not occur in the mountains of Colorado. Such is not the case. Anyone who spends much time in the Colorado mountains in the summer is familiar with the brief heavy showers that occur from time to time. It does, however, point to the conclusion that heavy rainfalls of the duration of those that occur in the plains are very infrequent in mountain areas, and ones that do occur are almost always small in area, producing insufficient rain for significant flooding of streams and rivers, in the absence of a melting snow pack.

PAGE 207

Part Six Stormwater Management

PAGE 208

This page IS intentionally blank

PAGE 209

INNOVATIVE STORMW ATER MANAGEMENT TECHNIQUES IN NORTHEASTERN ILLINOIS-A CASE STUDY Gerald J. Kauffman and Anwer Ahmed Donohue and Associates Background A Stormwater Management Plan (SWMP) was prepared for a rural 680-acre site in Hoffman Estates in northeastern lllinois. An unnamed tributary to Poplar Creek flows through the site, and a 4O-acre regional wetland is located on the tributary adjacent to the railroad tracks that divide the site. Three existing 36-inch CMP culverts under the tollway south of the regional wetland serve as the only outlet from the watershed. The SWMP was developed in accordance with the following objectives: 1. Utilize the 40-acre wetland for regional stormwater detention. 2. Provide regional detention for flood control, water quality purposes, and cost-effective construction of the stormwater conveyance system by reducing the flow from each upland parcel to the regional wetland. 3. Prevent increased downstream flood damages due to urbanization and paving of the watershed. Minimize the discharge from the site to meet the Metropolitan Water Reclamation District of Greater Chicago (MWRDG), Village and Northeastern lllinois Planning Commissions detention ordinance requirements. 4. Minimize the impact of ponded water adjacent to the railroad right-of-way and insure that the existing stages and duration of ponding will not be exceeded. 5. Integrate regional detention with preservation and enhancement of the wetland areas. Preserve natural drainage and overland flow corridors within the watershed. 6. Develop unit peak release and unit storage volume criteria for providing detention on individual parcels adjacent to the wetland areas. Hydrologic Model Development The hydrologic analysis for the watershed was performed using the U.S. Army Corps of Engineers HEC-l computer program. The 680-acre watershed was divided into 24 sub-basins according to land use and future parcel boundaries. The development of the sub-basin parameters assumed that the parcels would be 50% impervious, including the area up to the centerline of the adjacent perimeter roadways. Huffs rainfall distribution with lllinois State Water Survey (ISWS) Bulletin 70 rainfall volumes was used to estimate 2-, 10-, 25-, 50-, and lOO-year flows and runoff volumes. Results were verified using SCS Type IT rainfall distribution and National Weather Service TP40 rainfall depth data.

PAGE 210

186 Stormwater Management in Northeastern lllinois Three future drainage scenarios were modeled with varying detention require ments at individual parcels. The first scenario determined peak storage volume in the regional wetland based on regional wetland storage only and no upland detention storage on individual adjacent parcels. The second and third scenarios evaluated the impact of providing upland detention or adjacent parcels based on a release of 0.45 cfs per acre and 0.35 cfs per acre, respectively, along with detention storage in the regional wetland. The results of the HEC-l model for existing conditions and the future drainage scenario No. 2 based on 0.45 cfs per acre release from upland detention parcels are presented in the next section and summarized in table form. Peak Flow, Stage, and Ponding Duration Results The existing conditions analysis indicated that the peak l00-year inflow to the regional wetland area is approximately 300 cfs, with an estimated peak outflow from the site of 100 cfs. The future condition scenario with on-site detention based on a 0.45 cfs per acre release results in a peak inflow of approximately 268 cfs with a peak outflow of lOO cfs. The lOO-year flood high water elevation for both the existing and proposed condition is 821 feet. This analysis is unique because future peak flows, stages, and duration of ponding will not exceed existing values. Normally, peak flows and stages only are not exceeded as required by local storm water detention regulations and state water law. The duration of ponding criteria requires duplication of existing hydrologic conditions and flow hydrographs wherever feasible. This is done by providing additional storage volume at wetland mitigation areas and providing upland parcel detention storage. The HEC-l analysis confirmed that the existing duration of ponding along the railroad right-of-way would not be significantly increased for future conditions. The results of the analysis are summarized in Table 1. Future ponding duration exceeds the existing ponding duration at depths exceeding one foot for only three to five hours for the 10-year event and one to five hours for the lOO-year event. The existing and future peak outflows and the peak stages will remain similar. Upland Detention Requirements Of the three drainage scenarios considered, a maximum unit release rate of 0.45 cfs per acre for upland parcel regional detention basins was selected. The detention volumes obtained from the HEC-l analysis are summarized in Table 2. The upland parcel unit storage requirements range from approximately 0.09 to 0.17 acre feet per acre. An average unit detention volume of 0.12 acre feet per acre was selected to regulate the detention requirements on individual lots based on a 0.45 cfs/acre release rate.

PAGE 211

Kauffman and Ahmed Frequency 1G-year 1OG-year Table 1 Ponding Duration, Peak Stage, and Peak Outflow: Regional Wetland Adjacent to Railroad Embankment Elevation Existing 817 40 hrs. 79 hrs. 817.5 22 hrs. 32 hrs. 818 16.5 hrs. 24 hrs. 818.5 10.5 hrs. 15 hrs. 819 2.5 hrs. 5.5 hrs. Peak Outflow 62 cfs 62 cfs Peak Stage 818.98' 819.08' 817 54.5 hrs. 96 hrs. 818 32.5 hrs. 43 hrs. 819 23.5 hrs. 28 hrs. 820 15.5 hrs. 17 hrs. 821 4 hrs. 4 hrs. Peak Outflow 100 cfs 100 cfs Peak Stage 821.04' 821.02' Table 2 Summary of Individual Sub-Basin Flows, Release Rates, and Detention Volumes for Upland Detention Basins Lot Outflow Detention Unit Sub-basin Area Rate Volume Det. Vol. 187 (acres) -.!f&.. (acre ft) {acre ftLacre 1 172.0 n 29 0.17 3 18.4 8 2 0.11 4 31.2 14 5 0.15 5 25.4 11 3 0.12 6 21.7 10 2 0.10 7 13.9 6 2 0.14 8 14.3 6 2 0.14 9 16.9 8 1 0.06 10 20.9 9 2 0.10 11 21.5 10 2 0.09 lla 6.3 3 7 0.12 12 42.4 19 2 0.09 13 21.1 9 4 0.15 16 25.8 12 2 0.11 17 17.6 8 2 0.14 18 14.3 8 2 0.12 19 16.8 6 2 0.09 20 29.5 13 3 0.10 22 24.0 11 3 0.13 Average = 0.12 acre ft/acre

PAGE 212

188 Stonnwater Management in Northeastern lllinois The future overall release from the 680-acre watershed will be 101 cfs or 0.15 cfs per acre, which is consistent with NIPC recommended criteria. Annexation Agreement The criteria for the Regional Stonnwater Management Plan were incorporated into an annexation agreement. This will insure that parcels, as they are converted from rural to suburban use, will have regional stormwater detention basins designed in accordance with a coordinated watershed-wide approach. The main provisions of the annexation agreement are: 1. Regional upland wet detention ponds should have a maximum peak release of 0.45 cfs per acre. 2. Overall release from the 680 acre watershed should not exceed 0.15 cfs per acre. 3. Minimum unit detention volume on upland parcels shall be 0.12 acre feet/acre. The regional HEC-l model is available to individual parcel owners for updating as the area is converted from rural to suburban use over the next 10 years. Or individual property owners may choose to size the regional detention ponds in accordance with the 0.45 cfs per acre release and 0.12 acre feet/acre detention volume criteria, thus saving substantial remodeling costs. Comparison of Hydrologic Methods An interesting portion of the analysis included a comparison of required release rates and storage volumes using the SCS dimensionless hydrograph methodology and the modified rational method (MRM). This analysis was conducted to select a unit detention volume and confirm that, as parcels are converted from rural to suburban land use, future downstream flows, stages, and ponding durations will not exceed existing conditions. Table 3 summarizes the comparison for a 15-acre parcel. Comparison of Rainfall Data and Time Distribution A component of hydrologic analysis requiring considerable engineeringjudgment is selection of rainfall data, methodology (i.e., design event or historical rainfall simulation), and time distribution of rainfall. We examined the effects of these parameters on flows, stages, and volumes for various events. Table 4 summarize the comparison for the overall 680 acre watershed and a typical 30 acre parcel.

PAGE 213

Kauffman and Ahmed Table 3 Comparison of Release Rates and Storage Volumes Parameters: Parcel Size = 15 acres Peak Unit Release = 0.45 cis/acre Future Land Use = Suburban (50% impervious) C = 0.55 (modified rational method), CN = 80 (SCS-TR55 method) Time of Concentration (Tc) = 36 min. = 0.6 hour Design Event = 200-yr, 24-hr, TP40 = 5.80", Bul. 70 = 7.58" Storage Unit Surface Calculation Rainfall Volume Volume Area Method Data acre ft {acre ftLacre} @2' Fluctuation MRM TP40 1.4 0.09 0.7 MRM x 125% TP40 1.7 0.11 0.85 Donohue HEC-1 Bulletin 70 1.8 0.12 0.90 MRM Bulletin 70 1.9 0.13 0.95 SCS-Type II TP40 2.2 0.15 1.10 Hoffman Estates Bulletin 70 2.4 0.16 1.20 SMG (HEC-1) Bulletin 70 29.2/170 0.17 14.6 SCS-Type II Bulletin 70 3.4 0.23 1.7 Summary and Conclusions The regional stormwater master plan has been designed to: 189 % of Pareel Area 4.7% 5.7% 6.0% 6.3% 7.3% 8.0% 8.4% 11.5% I. Prevent downstream flooding as a result of development in the watershed. 2. Maintain wetland areas. 3. Preserve open space. 4. Provide soil erosion and sediment control. 5. Maintain aesthetics. 6. Meet and exceed criteria set by -MWRDGC Village -NIPC The features of this regional watershed-wide approach to stormwater management are: 1. Future stormwater detention engineering costs will be saved. The existing watershed model can be easily updated and revised as parcels are developed. 2. Unit peak release and detention criteria for upland parcels and incorporation of stormwater detention provisions into annexation agreement provides for specific, easy to understand, and uniform standards for design.

PAGE 214

190 Stormwater Management in Northeastern lllinois 3. Use of a regional model allows the village to assess cumulative effects of development on downstream flows, stages, and ponding durations rather than on a parcel-by-parcel basis, which has so often been done in the past. 4. Allows comparison of various rainfall depth, time distribution, and historic storm data to optimize stormwater detention design criteria. Table 4 Overall 68O-Acre and Typical 30-Acre Parcel Comparison Rainfall/Distribution (Prop. 100-yr 24-hr duration) 1. TP4O/Huff's 3rd Quartile 2. TP4O/SCS Type II Watershed = 680 Acres Peak Runoff (efs) 175 237 3. Bulletin 70, Huff's 3rd Quartile 239 292 4. Bulletin 70, SCS Type II 5. Aug. 13-14, 1987 Event (Greater than 100-yr. event) Rainfall/Distribution (Prop. 100-yr 24-hr duration 1. TP4O/Huff's 3rd Quartile 2. TP4O/SCS Type II 3. Bulletin 70, Huff's 3rd Quartile 4. Bulletin 70, SCS Type II 5. Aug. 13-14, 1987 Event 296 Watershed = 30 Acres Peak Runoff (cfs) 13 30 19 43 (Greater than 100-yr event-9.35" over 18 hrs.)42 Notes: U.S.N.W.S., TP40 = 5.80" (100-yr, 24-hr) ISWS, Bulletin 70 = 7.58" (100-yr, 24-hr) Peak Stage (cfs) 819.88 820.00 821.02 821.07 821.62 Rowand stage information excerpted from HEC-1 runoff summary. For the 680-aere watershed, peak inflow for the August, 1987 and Bulletin 70, SCS Type II events are also similar (292-296 cfs). Also, TP4O/SCS Type II and Bulletin 70/Huff's 3rd quartile peak inflows are similar (237-239 cfs). The TP4O/Huff's 3rd quartile flow at 175 efs is 40% lower than the Bulletin 70/Huff's 3rd Quartile flow. For the 3O-aere parcel, note that flows from the August, 1987 and Bulletin 70, SCS Type II design events are similar (42-43 cfs). SCS, Type II distribution generates higher peak runoff than Huff's 3rd quartile distribution.

PAGE 215

NAVIGATING POLITICAL WATERS: SUCCESSFUL IMPLEMENTATION OF URBAN STORMW ATER MANAGEMENT PROGRAMS Sharon L. Oakes Disaster Recovery Resources Introduction Often, the most sensitive aspect of program development is the "political," or public hearing process. Once the technical work has been completed, the dreaded public hearing looms on the horizon. Between the initial authorization of the program development efforts and the scheduled public hearing date, an elected or appointed governing board's membership, priorities, and political nature will most likely have changed. There is a real possibility that this "new" board, in response to both legitimate and emotional issues raised by the public, will delay, deny, or signifi cantly modify the program. How can program managers become more effective in navigating urban storm water management programs successfully through this political and public hearing process? Comparing the difficulties encountered in the city of Gainesville, Florida's recent implementation of two different stormwater management programs will illustrate the importance of early public involvement in the program development process. Two Case Studies (' .. a,se Study I. In 1984 the city needed to undertake a flood restudy program to update existing floodplain maps and the related flood control ordinance. A consultant completed the technical analysis. City staff were to map results, make ordinance amendment recommendations, and coordinate the recommendations of the flood restudy program through the required public hearings for adoption by the city commission. The consultant's research and analysis indicated that predicted flood levels were higher than the levels defined in the existing ordinance. After the maps and ordinance revisions were prepared, the public hearing process was started. In addition to the mandated public notice requirements, the city published a half-page newspaper advertisement, with the floodplain limits delineated in color, and sent letters to owners of property inside the IO-year floodplain. Within two days after the advertisement appeared, staff received more than 200 calls from citizens requesting flood zone information on their property. Although nearly all the calls were from single family homeowners, staff still anticipated that any voiced objections would be focused on the methodologies and assumptions applied in the flood restudy analysis. At the public hearing, homeowners raised questions regarding impacts on property values, flood insurance requirements, and the validity of the proposed maps. Their reasoning was based on a severe storm event several months earlier that

PAGE 216

192 Navigating Political Waters had not produced the flood levels predicted by the study. Staff were not prepared to respond adequately to these concerns. The commission directed staff to address these issues prior to their taking any action. This posed a monumental task for staff. Effective communication channels had to be established with a public that now questioned the credibility of the flood restudy program. The damage repair process included meetings with neighborhood representatives, more thorough staff and public education on the National Flood Insurance Program, restudy of certain areas, informal workshops, presentations to the Board of Realtors, and a comparison of the earlier storm event with the storm assumptions used in the flood restudy analysis. These efforts and final ordinance adoption took 18 months to complete. The final work product was certainly superior to that which had originally been presented. Determining how the past storm event compared with the assumptions made in the flood restudy analysis resulted in a higher level of confidence in the work completed. The Board of Realtors started requiring flood zone information on home purchase contracts. The city's participation in the National Flood Insurance Program was better understood. But these results could had been achieved more quickly and in a more constructive atmosphere had public involvement been sought at the commencement of and throughout the flood restudy program. Case Study II. More recently, a need to provide more protection to the city's natural creek systems was identified. In certain creek areas, the regulated flood zones were located well within ravine banks. Loss of native vegetation and creek bank erosion had occurred when well meaning citizens constructed homes on the creek banks. City staff were directed to prepare recommendations for a creek protection program. The first phase of this program was to develop criteria for selecting creeks to be regulated and to inventory the city's creek system. Creek protection goals and methodologies were drafted. Meetings were held with environmental regulatory agency representatives to determine how the creek protection program would interface with existing environmental regulations. A recommended list of regulated creeks and setback zones was prepared. Before any official action was scheduled, letters were sent to local environmental and homebuilder groups, the Board of Realtors, and the Chamber of Commerce inviting them to an informal meeting with city staff on the proposed creek protection program. At this meeting, several legitimate concerns were raised. Further research by staff and additional meetings with special interest groups resulted. These public involvement efforts paid off when the public hearing was held. Although there was not consensus on every point, no "surprise issues" were raised. With only minor revisions to staff recommendations, the creek protection program ordinance was adopted. Program development and implementation took about one year. When the city decided later to develop a public information brochure on the creek protection program, a private citizen volunteered to write it and a local land development firm and environmental group participated in its publication. This level of cooperation would have been unlikely if the creek protection program had been adopted in an atmosphere of controversy and mistrust.

PAGE 217

Oakes 193 Public Involvement Strategies Three strategies are recommended for the program manager to ensure successful public involvement. First, the commitment to public involvement should begin at the precontract or concept stage. Public involvement experience should be included as a criterion for consultant selection. Time and funding to include public involvement should be incorporated into budget and deadline decisions. Second, it must be recognized that the public is not a single homogeneous entity. Generic newspaper notices will not always reach all individuals and groups who may have a vested and vocal interest in the program. Regulatory agencies are good sources for obtaining information on special interest groups and making individual contacts. Networking with local neighborhood groups, the Chamber of Commerce, the Board of Realtors, and environmental groups will also help in identirying and reaching appropriate public sectors. Last, public involvement should be part of all aspects of program development. During data collection, input from groups and individuals should be sought. In this stage, it is inappropriate to "hard sell" the program. Set initial goals of explaining the project's scope, providing information on anticipated timelines, listening to concerns, and making a commitment to continue communication. The level of mutual trust and credibility established now will be invaluable when more volatile issues are presented to the public. When draft recommendations have been formulated, informal workshops should be held to discuss findings, potential impacts, and alternatives. The focus at these informal workshops should be on issues, not positions. More than one workshop is recommended, because it will take time for all affected members of the public to become aware of the proposed program and to generate a full spectrum of their concerns. The public hearing is the formal, usually legally mandated opportunity for the public to voice its concerns. Although it would be unreasonable to expect no objections to be made at the public hearing, all relevant issues should have been identified at prior meetings and workshops. This early communication should significantly decrease the possibility that surprise issues will be raised. Advance discussion of public concerns allows the program manager to be well prepared to i>rovide rational responses to all legitimate and emotional public issues. Conclusion Public involvement should be an important element during the development of all storm water management programs. Merely following the minimum noticing requirements mandated by law is not sufficient if program managers wish to 1) quickly identiry affected special interest groups, 2) adequately understand public concerns and issues prior to the public hearing, and 3) establish credibility for the staff and the program. In stormwater management, as in other aspects of life, change makes people uncomfortable. By making the public part of the program development process, both a better quality program and public acceptance of it should result.

PAGE 218

STORMW ATER QUALITY MONITORING AS A PLANNING TOOL-A CASE STUDY Cynthia L. Paulson Brown and Caldwell Consultants Dottie Nazarenus City of Fort Collins, Colorado Introduction Final National Pollutant Discharge Elimination System (NPDES) stormwater permitting regulations include requirements to monitor urban runoff during wet-and dry-weather conditions to support municipal permit applications. The application monitoring requirements, however, are only the beginning of a much larger need for long-term urban runoff quality monitoring as part of the water quality planning process. The city of Fort Collins, Colorado, is one municipality that has recognized this need and has implemented a stormwater quality monitoring program to provide a basis for their long-term stormwater program planning. Overview of EPA Application Monitoring Requirements For background, an overview of the EPA regulatory requirements for NPDES stormwater permitting is presented. Dry-weather field screening. Dry-weather field screening is required for all outfalls or field screening points (determined by a grid system) in Part 1 of the permit application. The field screening consists of two grab samples separated by at least four hours during a 24-hour period. The constituents to be tested with a field test kit are shown in Table 1. Wet-weather outfall monitoring. Wet-weather monitoring is required for three storm events at five to 10 major outfalls representative of various land uses. The storm event monitoring is to include a flow-weighted composite sample consisting of sample aliquots taken at least every 20 minutes for the duration of a storm event Of for the first three hours, whichever is shorter. The constituents to be monitored number more than 140, as shown in Table 2. Long-term monitoring. The regulations also require that a long-term monitoring program be proposed in Part 2 of the application and that monitoring be performed at outfalls and instream locations. Rather than provide specific definition of the long term monitoring requirements, the EPA has allowed flexibility for municipalities and regulatory agencies to work together in developing a monitoring program appropriate for site-specific conditions.

PAGE 219

Paulson and Nazarenus 195 Table 1 Dry-Weather Sampling Constituents pH aty of Fort Collins Phase 1 Total chlorine Total chromium Surfactants Fecal coliform Oil and grease City of Fort Collins Phase 1 Ttef' 1 -Basic Constituents Row Specific conductance pH Temperature Dissolved oxygen Nkalinity Dissolved solids Suspended solids EPA Regulation pH Total chlorine Surfactants Total phenol Total copper Comments swimming pool connections industrial facilities, auto engines, hot water heat/boilers, steel fabri cators carwashes, laundromats, some industry direct sanitary connections restaurants, auto repair and service stations, industry oil and grease and chromium sub stituted chromium a better indicator, less prevalent than copper Table 2 Wet-Weather Constituents EPA regulation Row pH Temperature Dissolved solids Suspended solids Comments load estimates estimate hardness to evaluate impact of metals standard water quality constituent standard water quality constituent standard water quality constituent standard water quality constituent common stormwater constituent major stormwater impact

PAGE 220

196 City of Fort Collins Phase 1 Nitrate + nitrite Ammonia Total Kjeldahl N Phosphorus dissolved and total Copper dissolved total recoverable Lead dissolved total recoverable Zinc dissolved total recoverable Fecal coliform Fecal streptococcus COD BOD ultimate Oil and grease Tier 2 Selected Organics Pesticide scan Herbicide scan Volatile scan Semi-volatile scan Stormwater Quality Monitoring Table 2 Wet-Weather Constituents EPA regulation Comments Nitrate + nitrite drinking water impacts, nutrient Ammonia aquatic life impacts, nutrient Total Kjeldahl N nutrient Phosphorus nutrient causing eutrophication dissolved and total Copper automotive industry and traffic, household plumbing common total stormwater constituent, Lead automotive industry and traffic, common stormwater constituent, total Zinc total Fecal coliform Fecal streptococ cus COD BOD 5-day all priority pollu tants automotive industry and traffic, common stormwater constituent sanitary sewer connections animal wastes industrial impacts organic loadings restaurants, auto repair and service stations, industry residential use, spring and summer residential use, spring and summer gas additives (Benzene, Toluene), solvents, microelectronics produc tion (Methylene chloride), dry cleaning (Carbon tetrachloride) heavier solvents, industrial cleaning agents, non-polar compounds, paint additives, varnish

PAGE 221

Paulson and Nazarenus City of Fort Collins ANISe 1 Base neutral TPH as waste oil & diesel TPH as gasoline 197 Table 2 Wet-Weather Constituents EPA regulation heavier solvents, some pesticides and herbicides, PCBs, paint addi tives, varnish, typically sulfuror nitrogen-containing gasoline products, lubricants, die sel, heavier molecular aliphatic compounds gasoline products lower molecular aliphatic com pounds Pentane, Hexane *Combined scan analyses provide measurements for all priority pollutants. An Alternative Approach to Monitoring Municipalities across the country will be spending hundreds of thousands of dollars on stormwater quality monitoring programs to meet EPA permitting require ments in the next five to 10 years. A comprehensive stormwater monitoring program framework is recommended to ensure that the information collected can be fully utilized as a water quality planning tool. The comprehensive monitoring program can be designed to meet EPA application requirements as well as long-term information needs. The State of Colorado Stormwater Task Force has developed an alternative ap proach to urban runoff quality monitoring that incorporates EPA requirements into a more comprehensive program. An overview of the approach is presented in Figure 1. Greater detail on the approach has been provided in the state's Stormwater Management Program (Colorado Department of Health, 1991) and other papers (Paulson, 1991). The City of Fort Collins Monitoring Program The City of Fort Collins initiated a storm water quality monitoring program in winter 1990 under the city Stormwater Utility. Some background on the stormwater utility is presented, followed by a description of the monitoring program. City of Fort Collins Stormwater Utility. The city of Fort Collins Stormwater Utility will mark its 10th year of operation in 1991. Highlights of this past decade include: a master plan and fee structure for nine basins; erosion control standards;

PAGE 222

198 Stormwater Quality Monitoring a detailed inventory of the drainage system in most of the 42-square-mile area; design and construction of a city park and bike path to replace a mobile home park in the floodplain; design and construction of a major outfall channel to the Poudre River; and, a balance sheet showing fixed assets of $9.1 million. The 1990s will see the focus turn more to environmental issues, to monitoring stormwater runoff, reducing negative impacts, permitting, and public education on storm water qUality. Emphasis on soft improvements such as detention ponds and grasslined channels has been the preference of the citizens of Fort Collins since the beginning of the stormwater management program. Although the original purpose was to slow floodwaters, detain flows, and reduce the peak, those preferable soft improvements may now have a more important function of reducing pollutant levels and decreasing the environmental impact of stormwater runoff in the Poudre River. Most citizens are concerned about pollution and will take responsibility when they know what to do. The city's stormwater inlet stencilling program that was initiated several years ago, has the message, "Dump No Waste-Drains to the Poudre River." This has been very effective in educating both children and adults to the fact that whatever is in the street and running into area inlets and catch basins will eventually go untreated into the river. It is essential that attention is given to educating other city departments and government agencies to ensure that their planning activities coordinate with the stormwater utility when such things as a new project, change to the city code, or revision of duties and responsibilities in a department might have a potential impact on the drainage system. Initiation of the city's monitoring program. One of the key objectives of the Fort Collins Stormwater Utility is to be proactive. Examples of recent activities related to the city's stormwater quality monitoring program include: staff has participated on the state of Colorado special task force to develop and initiate NPDES permitting; $100,000 was budgeted in 1990-91 to initiate the stormwater monitoring program, including purchase of two complete monitoring stations for wet-weather sampling; in late fall of 1990, Brown and Caldwell Consultants was selected to help design and develop the program and to begin both dryand wet weather sampling at 12 locations throughout the city; and, staff is in the process of selecting a full-time stormwater quality coordinator to begin in June or July. The preliminary work done to date has been well worth the limited investment of resources. The consultant, Brown and Caldwell, has provided an excellent field

PAGE 223

Paulson and Nazarenus r-----' Define Monitoring LOb jectives J --1--Review Existing Information Preliminary Monitoring .".., ", ........ N < Problem? __ ..... 199 ........ ....... """ ", Yes Storm water Long-term Management Monitoring Program Detailed Follow-up Monitoring 1 ",.......-...-........ < Problem? .... -v:s Source Identification: .... BC r-, L _.J Components added to EPA approach Figure 1. Stormwater quality monitoring approach.

PAGE 224

200 Stormwater Quality Monitoring report of each sampling event as well as a consolidated statistical report and a brief written report to highlight results and significant findings for each test. Staff has sorted the data in order to look strictly at constituents, sampling sites, and areas of the city that are problematic. This has been extremely helpful for education as well as assisting in future decisions about what should be solved first and what and where are the priorities. Program objectives. The objectives of the city of Fort Collins monitoring program include the following: Identification of problem areas Detection of illicit discharges Preliminary evaluation of: dry-weather runoff quality impacts of stormwater runoff (wet-weather) on receiving waters wet-weather runoff pollutant concentrations (Event Mean Concentrations) and projected loads reported water quality problems Development of benchmark data for future trend analysis Preparation for NPDES permit application Achievement of these objectives will require a comprehensive and phased program approach. The city of Fort Collins is taking an approach that includes the following phases: Phase I-characterization of urban runoff quality and identification of problem areas at high priority sites Phase 2-continuation of the characterization and problem identification to evaluate the areas not addressed in Phase I Phase 3-preliminary tracking of problems to their sources, particularly for dry-weather flows, and better characterization of wet-weather flows Phase 4-elimination of problem sources and development of a basis for implementation of Best Management Practices (BMPs) The Phase I monitoring program has not been designed to fully meet final EPA permit application monitoring requirements, primarily because of the high cost to meet these requirements. Instead, the program has been designed to collect preliminary information in support of current stormwater quality planning efforts and will be adapted in the future to meet Colorado Department of Health (CDH) permitting requirements. Because the city falls below the 100,000 population level, it will not be required to collect data for the application process until after October 1992. In the meantime, the city has presented its current monitoring program plan to the CDH for their review and has incorporated their comments to help ensure that the data will be viable for the permitting process in the future.

PAGE 225

Paulson and Nazarenus 201 Sampling sites. Phase 1 of the monitoring program addresses a total of three wet weather (two instream and one outfall) and 10 dry-weather (10 to 12 outfall or manhole) sampling sites. The sites have been selected on the basis of several criteria, including: land uses specific industrial activities potential for illicit discharges history of the area, including known problems and public complaints hydrological conditions drainage basin area available existing data importance of receiving water access and safety concerns Sampling constituents. Sampling constituents were selected to provide the most information possible while still maintaining budget constraints. The Phase 1 constituents, shown in Tables 1 and 2, are a combination of a list developed by the city, final EPA permit application monitoring requirements, and the approach included in the technical section of the Colorado State Stormwater Management Program Plan. Preliminary results. To date, the city has monitored a total of three dry-weather events and three wet-weather events. The laboratory data and observations from the dry-weather sampling have shown evidence of illicit connections/discharges to the storm drainage system. Specifically, high levels of fecal coliform, oil and grease, surfactants, temperature, and variable pH have been recorded. Data from the wet weather events are less conclusive, but have shown relatively fewer water quality problems than the dry-weather events. Paulson, Cynthia 1991 References Stonnwater Quality Monitoring-One Approach. Presented at the Colorado Engineering and Management Conference, February. Colorado Department of Health 1991 Colorado State Stonnwater Management Plan. Draft, January.

PAGE 226

SUCCESSFUL MUNICIPAL STORMW ATER MANAGEMENT: KEY ELEMENTS Andrew J. Reese ERC Environmental and Energy Services Company Introduction A number of cities are experiencing immense new demands on their stormwater management programs brought on by the impacts of the EPA stormwater NPDES regulations. For a number of these cities no coherent stormwater management program was in existence prior to the publication of the final regulations in November 1990. Rather, they had a loosely defined "drainage" program. These cities must now build simultaneously a stormwater quantity and quality management program-often from scratch. ERC Environmental and Energy Services Company (ERCE) has conducted a national survey of successful storm water management programs in an attempt to isolate those key elements that make for an effective and efficient stormwater management program. Organizing Stormwater Management Programs The term "stormwater management" can take on as many different meanings as there are stormwater programs. Some are narrow, single purpose. Some are broad, basin or political entity-wide. And there are a number of general and specific ways to group the various functions and components of stormwater management for consideration. A stormwater management program may have many or few of these components. And each of these components may be well or poorly developed. A way of looking at stormwater management is to divide it into specific components that are either defined duties or physical products. These components are: 1) long range planning guidelines or goals; 2) the legal, technical, and financial underpinnings required to support items 1 and 3; and 3) day-to-day management tools, responsibilities, and procedures. Long-range aspects include development of basic policies and goals as well as, more technically, stormwater master planning. Legal support is provided by written and, often, unwritten ordinances, regulations, and policy statements. Technical support is provided by design criteria manuals, data bases, on-line computer models, and geographic information systems or automated mapping support. Stormwater financing has become complex, going well beyond basic tax-based methods to stormwater utility charges, with various types of modifying factors and secondary funding methods. The basis for charges is normally relative impact on, or contribution of runoff to, a drainage system. The advantage of a utility form of financing is that it provides stable, adequate, consistent, and equitable funding. The day-to-day aspects of stormwater management, supported by the legal, technical, and financial underpinnings, carry the stormwater management program to accomplishment of long-term goals defined by the storm water master plans and

PAGE 227

Reese 203 goal and policy statements. Day-to-day management tools and procedures include all application of tools and procedures to carry out the five basic functions of stormwater management: 1) administration, financial management, and program development; 2) planning, design, and engineering; 3) operations and maintenance; 4) regulation and enforcement; and 5) capital improvements. ERCE Survey of Stormwater Management Programs During 1987 and again, less formally, in 1990, ERCE and its subconsultants conducted a selected nationwide inventory on various aspects of stormwater management in leading cities and associated counties. Because of the great diversity in government types, drainage environments, and general approach to stormwater management, a quantitative comparison or tabular summary of information for the cities and counties surveyed would be misleading. Rather, a listing of problems observed in most cities and of characteristics of those cities and counties that appeared to have successful stormwater programs are presented here. Characteristic Problems Certain characteristic problems are, or had been, the experience of most surveyed. Those cities that developed successful programs exhibited fewer of these traits. No city exhibited all of them. All experienced different levels of success and failure in each area. Stormwater management was often diffused among many staff elements and seen as a low priority. It was often managed by someone as an additional duty and was poorly financed and unplanned. Specific goals, objectives, or policies had never been developed or articu lated in writing and formally adopted. There was little knowledge of the state of the drainage infrastructure. If stormwater master plans or other types of stormwater assessment were available, they were often difficult or cumbersome to interpret, out-of-date, and not directly usable as analysis tools. Many local flooding problems existed that were beyond the financial and/or technical ability of homeowners to repair, but were not under the jurisdiction of the city or county. Comprehensive and usable planning, policy, or design documents backed by effective ordinances and enforcement powers rarely existed or were enforced. Many runoff problems were caused by flow from areas outside the jurisdic tion of the city or county. Interjurisdictional cooperation was lacking. Local, on-site detention was in a poor state of repair. Funding was inadequate and poorly targeted to meet needs.

PAGE 228

204 Municipal Storm Water Management Engineering methods used by designers were often outdated and not uniform in method, degree of accuracy, or presentation. Little data existed in directly usable form. Thus, many design engineers used inferior methods rather than develop the necessary information. There was little knowledge of or concern for the environmental aspects of urban runoff. Planners had an inability to predict the impacts of potential developments or the impacts of required stormwater controls such as detention. Day-to-day maintenance was driven by complaints and political pressures, rather than being professionally prioritized based on need. Use and enforcement of proper erosion control measures was seen as unimportant or not possible. Those responsible for stormwater management spent a perceived dispro portionate amount of time in plans review, often recalculating by hand various drainage calculations. The development process was not defined well enough to insure compliance with technical requirements, timely inspection, and as-built certifications. Many development control policies were simply "understood" by local engineers but never written, thought through in detail, or coordinated. Characteristics Of Successful Programs Those cities that appeared to have the most successful programs exhibited some common characteristics. These are listed below: Almost every successful program in a larger urban area had developed a stormwater utility incorporating a user fee as the main source of funding. To help overcome multijurisdictional problems, several cities/counties established a regional entity that performed some stormwater work. In many successful programs the total stormwater management function was consolidated under one authority. Each successful program exhibited a tight control on development; not in the sense of restricting development, but in guiding it to insure new problems were not created and that it bore its proper share of cost. Each successful program had well developed permitting procedures, thorough ordinances, detailed submittal requirements, adequate inspection staffs, complete design criteria manuals, and streamlined enforcement capabilities. Each had an efficient sense of purpose, job descriptions with well-defined goals, and resources to achieve these goals.

PAGE 229

Reese 205 Most had stormwater master plans that they were actively using and keeping updated through on-line computer models. Most had major and minor capital improvement programs that dealt logically with priority problems. Many had data collection programs and automated mapping, and used GIS systems for planning and facility management. Successful programs had ongoing public relations programs to encourage the support and participation of the general public. Most successful programs were moving toward maintenance of the whole public drainage system, including areas where public waters significantly entered the system. Most had well-developed levels of service. Successful programs depended on key individuals or groups to champion the cause of stormwater management in both the political and private sectors, and in key staff positions.

PAGE 230

NPDES STORMWATER PERMITTING IS THERE SOMETHING MISSING? William P. Ruzzo Brown and Caldwell Consultants Introduction By now there are not many environmental professionals, either in private practice or in public service, who have not heard about the final EPA National Pollutant Discharge Elimination System (NPDES) stormwater permitting regulations. The regulations, which took effect on November 16, 1990, will require 185 cities, 47 counties, and over 100,000 industrial facilities to apply for a stormwater discharge permit over the next year to year and a half. The reaction of many individuals is to either take a wait-and-see attitude or become proactive and develop an aggressive program to deal with stormwater quality issues. In either case, these individuals may be overlooking, or may not be aware of, an inherent requirement in the regulations: to integrate storm drainage management with water quality management. It can be argued that the EPA assumes that many communities and industrial facilities already have drainage management plans, which is not the case. By requiring water quality planning, the EPA may be putting the cart before the horse. For the purposes of this paper, storm drainage management refers to the process of controlling urban runoff quantity (i.e., drainage and flood control requirements), whereas water quality management refers to the process of controlling urban runoff quality. For consistency, the author is suggesting that stormwater management then be used to refer to both quantity and quality. Since urban runoff quantity and quality issues are interrelated and essentially inseparable, water quality management, as required by the NPDES permits, must abide by the same principles required for storm drainage management, such as: allocation of space for facilities, multiple use of facilities, preventive measures instead of corrective measures, provisions for long-term maintenance, and provisions for water pollution control features. These five urban drainage management principles are reviewed and reasons presented as to why they are also important to urban runoff water quality manage ment. Stormwater Management Principles The dictionary defines principles as "a general truth or law . that which is inherent in anything . essence." Many of the following principles have been identified by authorities in the field of stormwater management and have been adopted by communities throughout the country {DRCOG, 1969; ASCE, 1990;

PAGE 231

Ruzzo 207 Scheaffer, 1982; APWA, 1982; Tulsa, 1990). These principles are considered basic understandings in the urban drainage (and now, water quality) field. Even though these principles are inherent, our knowledge in stormwater management continues to grow, and newer principles have been fairly recently recognized. One example is that water pollution control measures are essential features in drainage management (APWA, 1981; Scheaffer, 1982). In addition, a recommendation of this author is that a new principle be adopted to provide for long term maintenance needs, as they are also essential. Space allocation. Stormwater management needs must be included in the urbanization process by incorporating stormwater planning with all regional and local urban master plans and site facility planning. Storm drainage management facilities, such as channels and storm sewers, serve both conveyance and storage functions requiring space. However, the space requirements for adequate drainage may compete for space with other land uses. If adequate provisions for drainage requirements are not included in urban master plans, urban runoff will conflict with other land uses, result in flood damages, and impair or even disrupt the functioning of other urban systems. Many of the best management practices for urban runoff quality control also require adequate space for implementation. For instance, wet and dry detention basins, infiltration basins, and constructed wetlands all require extensive amounts of dedicated land. If the land requirements are not currently available or are not included during the planning stages for future development, then opportunities for controlling urban runoff quality are seriously diminished. Multiple use of facilities. Stormwater runoff and the facilities to accommodate the runoff can be an urban resource. Drainageways can provide environments for various life forms such as aquatic life, animals, birds, and vegetation. Drainage facilities can also provide areas for active and passive recreation. Competing uses for available land and the improved economics of combining facilities makes the multiple use of drainage facilities, for all practical purposes, mandatory. Previously, multiple uses for drainage facilities included aesthetic, environmental, and recreational benefits, which were for the most part optional. Now, mUltiple use of facilities must include water quality enhancement. When drainage quantity control dictates the need for a detention site, then the facility must be designed to enhance water quality as well, otherwise opportunities will be lost. In many instances, the increased cost for water quality enhancement will be within the range of 10% to 20% (Ruzzo, 1987), not including land values. One reason water quality costs are not higher is that water quality detention will also reduce the size of downstream drainage facilities and, therefore, their cost. Another reason for the minimal increase in detention costs is that modifications to detention facilities for water quality are not that complex and, in fact, are inherent in well designed quantity control facilities. Preventive measures. If urbanization takes place without adequate stormwater facilities, the cost to retrofit the system in the future may be prohibitive, if not

PAGE 232

208 NPDES Stormwater Permitting physically impossible. Therefore, it becomes important to plan and implement the required stormwater system as urbanization takes place in order to minimize costs. Based on the author's experience with many drainage master plans, the costs per mile may range from $500,000 to $1 million when drainage facilities are planned ahead. When drainage facilities are retrofitted, however, the costs per mile can typically exceed $2 million. A dramatic example of a cost-effective preventive measure to control urban runoff water quality is the control of erosion and sedimentation. In the report America's Qean Water (ASIWPCA, 1985), sediment from all sources reportedly "affects more river miles than any other pollutant. The cost to correct the impact of sediment is prohibitive, but actions to minimize sediment at the source can be very cost-effective. The EPA has recognized the need to control sediment at the source and has included requirements in the NPDES regulations for the development and implementation of management plans to control erosion and sedimentation from construction activities, which is the main source of sediment in urban areas. Provisions for long-term maintenance. There is considerable concern about the deterioration of infrastructure throughout the United States. Whereas drainage system deterioration is due in part to lack of funds, many systems were simply not designed to minimize maintenance requirements and costs. For all future facilities we must recognize that stormwater systems are part of the overall infrastructure, and the consequences of system failures due to lack of maintenance are unacceptable. This is why the author has recommended that provisions for long-term maintenance be included as a basic stormwater principle. Without maintenance of the drainage facilities, we can eventually expect (ASCE, 1990; Tulsa, 1990): partial or total loss of facility capacity, total replacement of the facility at a higher cost, flood damages to public and private property (i.e., streets, bridges, utilities, housing, public service facilities), loss of economic base (i.e., jobs and business income), impacts on beneficial uses of receiving waters, higher health risks, and loss of life. Like other services provided to the public, maintenance is not inexpensive. The annual cost of operations and maintenance can be from 15-25 % of the total cost of drainageway facilities (capital, administrative, legal, engineering, operations, and maintenance) The performance requirements for water quality best management practices will amplify the importance of adequate facility maintenance. For detention ponds to maintain design pollutant removal efficiencies, the inlet, outlet, and basin areas must be preserved in good working condition. The performance of infiltration and filtration facilities (i.e., swales, trenches, basins, and porous pavement) can be seriously diminished by sediment, which must be periodically removed to maintain

PAGE 233

Ruzzo 209 effectiveness. The use of wetlands for water quality treatment may require extensive maintenance in the form of plant species management, harvesting of dead plants, and sediment control. Water pollution control features. The essence of this paper is that water quality facilities must be integrated with drainage quantity facilities. Recognition of this importance was identified early in the 1 970s by the Urban Water Resources Research Council and more recently in stormwater management publications (APWA, 1981; Scheaffer, 1982). However, this message has not been well integrated into drainage management and, in fact, required congressional action (i.e., the Clean Water Act of 1987) to force water quality controls to be incorporated into storm drainage systems at a national level. Subsequently, the EPA regulations have specifically identified that certain measures be included, such as: For munidpalities: proposed management programs to reduce pollutants using "management practices, control techniques and system design and engineering methods" [at 40 CFR 122.26(d)(2)(iv)], and For industrial fadlities: the draft general permit requires that a "storm water pollution prevention plan" be developed that includes the description and implementation of storm water management controls. Conclusions There are many communities and industrial facilities that have not yet adequately addressed drainage requirements. The Water Quality Act of 1987 and the EPA stormwater regulations are now forcing communities and industries to include both quantity and quality features in the drainage system. This will be a difficult task even for those communities and industries that have well-established drainage management programs, but will be an overwhelming task for those now struggling simply with issues. Since communities and industrial plants must submit NPDES applications within one to one and a half years, there will not be a lot of time for them to become educated in stormwater management. This is why understanding the basic principles of stormwater management is vital to the success of the NPDES program. Our best chance to succeed with this program is to keep to the basics. If we abide by these stormwater principles, then the programs developed during the NPDES process will be greatly enhanced. However, if these basic stormwater management principles are not included in the NPDES process, and quantity and quality requirements are not integrated, then there will be something missing. References American Public Works Association (APWA) 1981 Urban Stormwater Management. Chicago, lllinois.

PAGE 234

210 NPDES Stormwater Permitting American Society of Civil Engineers (ASCE) and Water Pollution Control Federation 1990 Manual of Practice for the Design and Construction of Storm Drainage Systems. New York (draft). Association of State and Interstate Water Pollution Control Administrators (ASIWPCA) 1985 America's aean Water. Washington, D.C. Denver Regional Council of Governments {DRCOG) 1969 Urban Storm Drainage Criteria Manual. Denver, Colorado. Ruzzo, William P., and William H. Mitzelfeld 1987 Basin Wide Water Quality Planning for Nonpoint Source Pollu tion. Presented at 1987 Denver Regional Meeting on Water Management, sponsored by the U.S. Committee on Irrigation and Drainage, September 2-4. Scheaffer, John R., Kenneth R. Wright, William C. Taggart, and Ruth M. Wright 1982 Urban Storm Drainage Management. New York: Marcel Dekker. City of Tulsa 1990 Stormwater Management Criteria Manual. Department of Public Works, City of Tulsa, Oklahoma (draft).

PAGE 235

A COUNTYWIDE STORMW ATER AND FLOODPLAIN ORDINANCE Jonathon P. Steffen DuPage County, illinois, Department of Environmental Concerns Joseph E. Stuber CH2M HILL Introduction The Salt Creek watershed of DuPage County, illinois, experienced significant flood damage in 1987 and 1988. The cause of the flooding may be attributed to several factors, the most significant being the extent and nature of development within the watershed. In recognition of the need to resolve existing flooding problems and reduce the potential for flooding due to new development, DuPage County adopted a county wide Stormwater Management Plan and is developing a countywide Stormwater Ordinance. The plan and ordinance cover a broad range of stormwater management issues related to flood control projects, development practices, and floodplain and riparian land management. Storm water Management Plan The Stormwater Management Plan responds to the opportunity inherent in state of llIinois legislation (P A 85-905) that authorizes regional stormwater management in northeastern llIinois counties. It also recognizes the integrated nature of watershed systems and the need to consider watershed planning on a watershed-specific basis. The plan consolidates the stormwater management framework throughout the multijurisdictional county into a united, countywide structure. It sets minimum standards for floodplain and stormwater management and provides for countywide coordination of stormwater management in both natural and constructed drainage ways and storage facilities. Objectives and policies. Six primary objectives were developed to establish the direction of the Stormwater Management Plan: 1. Reduce existing potential for stormwater damage to public health, safety, life, and property. 2. Control future increases in stormwater damage within and adjacent to DuPage County. 3. Protect and enhance the quality, quantity, and availability of surface and groundwater resources. 4. Preserve and enhance existing aquatic and riparian environments and encourage restoration of degraded areas. 5. Control sediment and erosion in and from drainageways, developments, and construction sites.

PAGE 236

212 Stormwater and Floodplain Ordinance 6. Promote equitable, acceptable, and legal stormwater management measures. These objectives are supported by 15 policies that define physical and institu tional characteristics of stormwater management: 1. Require appropriate and adequate provision for site runoff control, emphasizing site runoff control. 2. Encourage use of stormwater storage in preference to conveyance. 3. Require design and evaluation of each site runoff control plan consistent with watershed capacities. 4. Restrict future development in the floodplain to facilities that will not adversely affect flood damage potential or wetland environments, and prohibit development in the floodway unless it involves facilities that enhance flood protection. 5. Require preservation of wetlands. 6. Incorporate water quality and habitat protection measures. 7. Require regular, planned maintenance of storm water management facilities. 8. Encourage control of stormwater quantity and quality at the most sitespecific or local level. 9. Define clearly the responsibilities and authorities of government entities 10. Require cooperation and consistency in stormwater management. 11. Promote delegation of authority to the most appropriate level. 12. Require strict compliance and enforcement of the stormwater management policies and their implementing regulations. 13. Foster the use of simple technologies wherever appropriate and realistic, but demand use of more sophisticated techniques where necessary. 14. Select cost-effective methods of achieving objectives. 15. Estimate costs of stormwater management recommendations and identify appropriate revenue sources before their adoption. These objectives and policies provided the framework for DuPage County storm water management standards discussed in the plan. They also provided direction for criteria and guidelines and for development of the stormwater and floodplain ordinance. Plan standards established based on the objectives and policies are supported by specific criteria and guidance documented in a series of 17 technical appendixes. Stormwater Ordinance General provisions. The comprehensive DuPage County Stormwater Ordinance includes provisions for the rate, quantity, and quality of runoff, and applies county wide to all activities that affect stormwater runoff including development, redevelopment, and substantial improvement of property.

PAGE 237

Steffen and Stuber 213 The ordinance is designed to be overseen by DuPage County and administered and enforced by local jurisdictions or a combination oflocal and county jurisdictions. Local administration and enforcement authority is maintained at the discretion of individual municipalities. Permitting requirements for development vary with the location and extent of activity. General requirements. Development within DuPage County must meet the criteria and guidelines established in the plan, meet the requirements of watershed plans, and meet the requirements established for special management areas. The ordinance also stipulates that responsibility for long-term maintenance must be established. In addition, the ordinance: prohibits additional threats to public health or safety, beneficial stream uses, or functions of aquatic habitat; prohibits activity that would increase flood heights, velocities, floodplains, or flood damage; prohibits degradation of surface or groundwater quality; requires that stormwater control measures are in place before other construction begins. Site runoff requirements. The site runoff requirements in the ordinance: prohibit damage to adjacent property, require that drainageways and storage facilities are located within easements or rights-of-way, require best management practices pursuant to the Clean Water Act, require that peak rate of runoff be maintained at predevelopment conditions, require that the volume of depressional storage be maintained, restrict the use of floodplains for site runoff storage, establish responsibility for maintenance required on subdivision plats. Sediment and erosion control requirements. Sediment and erosion control measures must be in place before construction begins and until permanent control measures are in place. Special Management Areas-Floodplain requirements. This section of the ordinance: restricts increases in flood elevations or decreases in conveyance capacities either upstream or downstream of the new development, requires 2-foot freeboard for new development along the floodplain, allows floodproofing for existing structures only, sets compensatory flood storage (h ydraulicall y equivalent to original storage) at 1.5 to 1, requires replacement of flood storage lost due to channel improvement, prohibits increases in release rates from existing storm sewer outfalls.

PAGE 238

214 Stormwater and Floodplain Ordinance Special Management AreasWetland requirements. This section of the ordinance: requires that wetlands be delineated according to the Federal Wetland Delineation Manual, Tightly restricts development in wetlands classified as critical, allowing 3 to 1 mitigation in special cases, restricts development in wetlands classified as regulatory, allowing 1.5 to 1 mitigation in special cases. Special Management Areas---Riparian land requirements. This section of the ordinance: generally defines riparian land as those within 50 feet of the normal shoreline of critical wetlands, lakes, streams, and rivers; discourages development on riparian lands, but allows it with functional mitigation; Requires that buildings have 50-foot setbacks from the ordinary water line. References DuPage Stormwater Management Committee 1989 DuPage County Stormwater Management Plan. Prepared with the DuPage County Stormwater Management Division and CH2M HILL. DuPage Stormwater Management Committee 1989 Countywide Stormwater and Flood Plain Ordinance. An appendix to the DuPage County Stormwater Management Plan.

PAGE 239

FROM LIABILITY TO RESOURCE: THE CITY OF AURORA, COLORADO'S CHANGING APPROACH TO DRAINAGEWAY DESIGN William E. Wenk Landscape Architect Introduction The city of Aurora is typical of many suburban communities throughout the country. It has grown from a small town, incorporated in the 1930s, into a city through an explosive growth cycle in the late 1970s and early 1980s. Comprehensive planning for growth was often preceded by development, leaving the questions of civic goals and quality of life to be answered later. Historically, the city viewed drainageways as having a singular function, to move stormwaters quickly and efficiently through populated areas. In the early 1980s, the public grew more vocal about the loss of trees, open space, and rural character of many areas, and community residents and regulatory agencies pressured the city into considering a broader range of functions for drainageways and to preserve and enhance natural and recreational resources related to drainageways. Simultaneous expansion of the Federal Clean Water Act, Section 404 regulations, and increased awareness of multiple use and preservation oflimited resources prompted the city to examine alternative approaches to traditional drainageway design. The city responded by developing a new approach to drainageway design, integrating environmental, aesthetic, and regulatory concerns into the planning and design process. The result has been the construction of drainageways that are dramatically different in function and appearance from past projects. All meet required channel capacity and stability requirements, but go beyond these limited engineering objectives to create a series of open spaces, trail corridors, and wildlife habitat areas that enrich the lives of residents in surrounding areas and that meet or exceed federal regulatory requirements for the preservation, protection, and enhancement of environmentally sensitive areas. This new approach requires a significant departure from past procedures for design, construction, and maintenance of stream corridors, where improvements were designed primarily by engineers without recognition of impacts on natural areas, recreation values, or visual impacts on neighborhoods. Development of conceptual and detailed designs that meet a broad range of diverse, and possibly conflicting, goals requires a collaborative effort, including a broad range of city agencies, community members, and design disciplines, including engineering, ecology, and landscape architecture. The following projects illustrate creative responses to a range of design issues, in addition to providing flood control and channel stability. Each project meets engineering requirements for lOO-year flood protection, but also addresses design issues that serve to integrate each project into its surroundings.

PAGE 240

216 A Changing Approach to Drainageway Design Horseshoe Park Horseshoe Park is a Y -shaped, 70-acre combined active and passive open space park, formed by the confluence of West Tollgate Creek and the emergency overflow spillway of Cherry Creek Dam. The park encompasses approximately 1.6 miles of creek, with an open space width up to roughly 800 feet. The l00-year flood event at the downstream end of Horseshoe Park is estimated to be approximately 15,000 cfs, with the normal flow estimated to be less than 10 cfs. Roadway embankments constructed across the lower end of the park created meandering, braided channels, ox bows, small ponds, and extensive wetlands. Increased storm water flows caused potential flooding problems for several adjacent homes. In 1980, the city channelized the streams to move flood waters through the park and to protect adjacent homes. Straightening and channelization destabilized the streams and caused them to cut even more deeply and to expose utility lines. As the streams dropped, the water table supporting the wetlands dropped as well, and by 1986 most of the wetlands had dried up. Deterioration of existing trails, flooding problems, exposed utilities, and intense lobbying efforts by adjacent neighborhood groups prompted the city to retain a multidisciplinary team of engineers, ecologists, and landscape architects. The city also assembled an in-house multidisciplinary team, with representatives from Utilities, Public Works, and Parks and Open Space. A joint effort between the city, neighborhood representatives, and the consultant team produced site design concepts that comprehensively address engineering issues, recreational opportunities, and natural area restoration. Extensive excavation of a portion of the upland area in the park was required to remove houses from the floodplain. A series of drop structures provides channel stability, and a special weir structure diverts daily stream flows from the incised channel into a network of braided stream channels to nourish the extensive park wetlands. A connected system of trails runs throughout the park, linking key points of adjacent neighborhoods as well as the citywide trail system. Engineering requirements for flood control were met, and the full potential of the site as a recreational amenity and open space was developed. Revegetation of upland areas included a diverse mix of prairie grasses and flowers. A mix of tree and shrub plantings enhanced the natural regeneration of riparian and wetland plants, and provided more diverse and extensive wetland habitat. To prevent further erosion from street runoff from adjacent areas and to improve water quality of the storm runoff, a series of low maintenance "parklets" were constructed at the end of each cul-de-sac. The parklets are small, quarter-acre, gently sloping turf grass areas irrigated by storm runoff, intended for neighborhood picnics or informal field games. Flood irrigation provided by the "first flush" of storm runoff waters the turf and prevents pollutants from reaching the stream. Construction was completed in 1988. The trail system is heavily used, and a diverse ecology has been restored. The channel has stabilized and l00-year flood protection extends to adjacent homes. The project is not, however, without problems. Maintenance requirements for parklets and the upland prairie in the park are not well

PAGE 241

Wenk 217 understood, and the park suffers from inadequate and misguided maintenance efforts. Also, standard channel mowing schedules have not allowed the mix of prairie grass and flower species to develop fully. Sand Creek Sand Creek is a sandy, alluvial channel, formed by a large drainage basin, approximately 32 miles long. Increased urbanization created larger flows into the stream resulting in higher stream velocities, more sediment transport, erosion, and channel bed degradation. Significant numbers of adjacent homes sat in the 100-year floodplain, and the channel was unstable. Roads lined both sides of the channel, which precluded easy widening of the channel without creating major impacts on the existing riparian and wetlands vegetation. Increasing the steepness of the channel side slopes precluded the opportunity to develop landscape plantings along the stream edge to remain compatible with the residential area. The surrounding neighborhoods have matured over their 20-year history, and a number of problems have become apparent. The lack of pedestrian and street channel crossings required a long detour for children to reach schools on the opposite side of the channel. The city's extensive network of trails along other drainageways could not be linked along this portion of Sand Creek because of its narrow channel right-of-way, steep banks, and lack of space for trails. As with Horseshoe Park, a consultant team consisting of engineers, ecologists, and landscape architects was formed. Project goals included providing 100-year flood protection and stabilizing the channel, extending the trail system along the creek, developing a planted parkway edge between the channel and the roads, creating pedestrian crossings to link the neighborhoods, and mitigating wetlands lost by channel construction. A previous drainage master plan for this area proposed lining the channel with riprap. New guidelines for implementation of a federal 404 permit, requiring mitigation of wetlands lost during channel construction, caused re evaluation of this proposal. Other alternatives were explored to more appropriately meet a broader range of programmatic requirements. Final design concepts proposed substituting soil cement erosion bank protection for the riprap. Soil cement, constructed primarily from the sands of the channel, visually integrates with the natural materials of the channel bottom, precludes unwanted plant growth, and allows landscaped parkways and pedestrian trails along both sides of the channel. Drop structures of sheet pilings were clad with architectural pre-cast concrete to create a more visually acceptable structure throughout the residential area. At several structures, the concrete spillway cap on the sheet piling was augmented with a low water pedestrian crossing and stairways to provide pedestrian access across Sand Creek and between neighborhoods. To satisfy the requirements of the U.S. Army Corps of Engineers Section 404 permit, a mitigation plan was prepared, and wetlands lost were mitigated by onand off-site wetlands reconstruction. Channel engineering improvements were constructed in 1988. Construction of wetlands plantings required as part of the 404 permit and the upland tree and grass

PAGE 242

218 A Changing Approach to Orainageway Design plantings were completed in 1989. Wetlands have successfully been established in the channel, and pedestrian crossings are heavily used by the area's residents. Shop Creek Cherry Creek Reservoir is a major flood control and recreation resource for the Denver metropolitan area. Concern over the potential for accelerated entrophication of the reservoir led to the investigation of sources of pollution from urban runoff. The rapidly developing Shop Creek basin upstream of the reservoir was identified as a key source of phosphorus, a critical nutrient with respect to in-lake algae growth. The Shop Creek channel was also experiencing severe erosion as runoff from upstream development increased. Significant sedimentation was creating a substantial maintenance problem and sedimentation of the lake's edge. In 1987, the city of Aurora contracted with water quality and hydraulic engineers and landscape architects to design a system of improvements to stabilize Shop Creek and to reduce by at least 50% the phosphorus load it conveyed to Cherry Creek Reservoir. In addition to channel stability and water quality issues, the project sponsors and Colorado State Parks Department required that the recreation experience of the area's visitors not be compromised by the channel improvements, and that the improve ments must visually integrate into the gently undulating prairie landscape of the recreation area. The channel is highly visible from an adjacent state highway and roads and active use areas within the recreation area. Original channel improvement concepts, prepared prior to the consultant team's involvement, proposed a straight, trapezoidal channel with several baffle-chute drop structures to control the elevation difference of 40 feet between the upper and lower limits of the project area. To control phosphorus, treatment facilities consisting of a detention pond with a permanent pool for settling and biological uptake, followed by a series of combined wetland/infiltration areas, were recommended to satisfy the project objectives of developing a nonmechanical, low maintenance improvement system that could be integrated into the natural landscape of the recreation area. It was determined that a straight, trapezoidal improvement concept would have negative visual and aesthetic impacts on the area. Alternative channel alignment and drop structure types were explored to more effectively integrate the channel improvements into the existing landscape. To avoid imposition of visually incongruent elements such as a straight, trapezoidal channel, the existing meander of the stream was retained and side slopes were graded to blend with the surrounding topography. Channel drop structures, constructed of soil cement, were designed that use on-site soils as the primary structural material. The crescent shape of the drop structures creates a form that minimizes the visibility of structural elements and recreates the gently sloping diagonal lines of the surrounding landscape. Extensive wetlands along the channel mitigate wetlands lost as part of channel construction, provide a more diverse ecology and wildlife habitat, and supplement the phosphorus removal function of the upstream water quality retention pond. Channel slopes were designed to minimize storm-flow velocities, and to provide a stable channel bottom to minimize washout from the wetlands of the phosphorus ridge organic humus.

PAGE 243

Wenk 219 Construction of the project was completed in 1989. Water quality monitoring of the creek downstream of the project area is currently assessing phosphorus levels to determine if removal goals are being met. Wetlands created as part of the project are thriving, and the project, when viewed from roads and recreation facilities, is minimally, if at all, visible. Conclusions Horseshoe Park, Shop Creek, and Sand Creek each illustrate highly individual and innovative approaches to drainageway design. The resultant projects create assets whose values extend beyond those of flood control to provide recreation, open space, and wildlife areas for the community. The benefits realized in these projects only hint at what is possible. The projects described in this chapter suggest the following conclusions. Public officials are often unaware of the multiple use potential of stream corridors or feel that hydraulic requirements preclude development of recreation, open space, and habitat areas within the corridors. An awareness program should be developed to point out successful projects where a variety of issues have been addressed as part of the channel design. Assembly of a multidisciplinary design team is essential. All issues related to the project, especially those not directly related to hydraulic design considerations, should be identified prior to the assembly of the consultant team, and an appropriate team assembled. Each discipline on the design team is equally important. The team should be composed of members who have mutual respect for, and knowledge of, the other member's capabilities. All disciplines should be involved from the onset of the project. Significant opportunities will be missed if professionals such as landscape architects or ecologists are brought in late in the design process. Often, the most creative responses are a result of ongoing collaborations where each team member contributes significantly in formulating basic conceptual approaches. Municipalities should be open to innovative and new approaches to channel design and should encourage limited experimentation to address emerging issues, such as water harvesting and the use of alternative construction materials, for example. Ongoing maintenance of areas other than channel structures may be important and may require ongoing collaboration between city departments and agencies. Higher levels of specialized maintenance during the initial establishment of natural areas can result in lower long-term maintenance costs. Minimal increases in project budgets can often result in significant increases in the recreation and open space value of urban channel improvement projects. Initial project budgeting should consider the potential benefits,

PAGE 244

220 A Changing Approach to Drainageway Design beyond stormwater control, allowing for a greatly enhanced value of the project for the community. References Fisk, David, et aI. 1989 Wetlands: Concerns and Successes. Symposium Proceedings; American Water Resources Association. WulIiman, James T., et. aI. 1988 Multiple Treatment Systemjor Phosphorus Removal.

PAGE 245

INCORPORATING STORMW ATER QUALITY ENHANCEMENT FEATURES INTO URBAN FLOOD CONTROL PROJECTS James T. Wulliman CH2MHILL Introduction Streams and drainageways associated with urban flood control projects can be key components of a strategy to manage stormwater quality. The channel conveyance network is pivotal because it often contributes to water quality problems, but can be a cost-effective part of the solution with the right design approach. The NPDES stormwater regulations underestimate the role that drainageways play in both water quality problems and solutions by implying that as "waters of the U.S.," they are to be protected and undisturbed. The presumption is made that leaving streams undisturbed and implementing source controls in the upstream basin will preserve the natural integrity of the channels. However, in the semiarid West, the combination of increased runoff because of urbanization and decreased sediment load due to basin controls leads to bed and bank erosion, degrading aquatic habitat, water quality, and channel vegetation. Instead of adopting a policy of leaving streams undisturbed, it is better to consider stream stabilization and the selective incorporation of water quality enhancement facilities into the channel network as components of the overall stormwater management system. This paper presents design considerations for incorporating water quality enhancement features into urban flood control projects. Also included is a discussion of the influence of pollutant inflow load on the design of enhancement facilities, guidelines for selecting pollutant removal goals for a project, and options for configuring water quality facilities. Pollutant Inflow Load Stormwater quality enhancement facilities are primarily designed to control sediment and nutrients. Although other pollutants would be trapped in such facilities, the objective is to keep pollutants such as paints, solvents, automotive fluids. industrial chemicals, and sanitary sewage out of the stormwater system via effective source control programs. Sediment and nutrients in storm water can degrade aquatic habitat and accelerate the eutrophication of ponds and lakes. Source controls alone may not be enough to reduce sediment and nutrient loads to acceptable levels. For instance, erosion rates from construction sites can be as high as 200 tons per acre per year (Chen, 1975). A sound construction erosion plan (one that is, for example, 80% effective) could reduce the rate to less than 40 tons per acre per year. The reduced rate, however, would still generate a sizable sediment load, perhaps 40 times greater than from nonconstruction areas. In addition, sediment and nutrients entering stormwater as a result of drainageway erosion may also be difficult to control. Stabilizing the many miles of streams and drainageways susceptible to erosion in the

PAGE 246

222 Stormwater Quality Enhancement Features Denver metropolitan area through the construction of channel improvements, for instance, will require years. Nutrients are more difficult to remove from stormwater than sediment because nutrients tend to be adsorbed into fine-grained silts and clays and are difficult to settle out. Nutrients are also present in stormwater in dissolved forms and cannot be settled out. The proportion of suspended and dissolved forms of pollutants, the size range of particles associated with suspended forms, and the tendency for suspended forms of pollutants to become soluble and dissolved impact the design of enhance ment facilities. Another important design consideration pertains to the relative contribution of pollutant loads during base flows; small, frequent storms; and large storms. Large storms can contribute greater loads of sediment and nutrients than a number of small storms or long periods of base flows. Trap Efficiency Objectives Once the characteristics of the inflowing pollutant load are identified, it is important to determine how much removal of the load, if any, is desired in a project. The adoption of specific trap efficiency objectives for a project depends on the following considerations. Applicable regulations. Trap efficiency objectives must satisfy performance or sizing criteria that may be specified by local, regional, state, or federal authorities. For instance, the Cherry Creek Basin Authority in the Denver metropolitan area has established an objective of reducing the stormwater phosphorus load entering Cherry Creek Reservoir by 50% (Cherry Creek Basin Authority, 1989). The prevalence of performance or sizing criteria is expected to increase in the next several years as communities implement comprehensive stormwater management planning. Impairment of receiving waters. Impairment of beneficial uses in downstream receiving waters from pollutant loads influences the adoption of trap efficiency objectives. The greater the impairment from specific pollutants, the higher the removal goals would generally be for those pollutants. Removal goals may become most stringent upstream of valued lake resources because the high natural trap efficiencies of lakes make them susceptible to impairment from stormwater pollutants. Sufficiency of source control measures. Even if regulators interpret pollutant removal "to the maximum extent practicable" to be achievable solely with effective source control programs, communities may consider the selective use of structural controls as backup systems for source control programs, especially to reduce sediment and nutrient loads being conveyed to critical receiving waters. Small incremental cost. In locations where storm conveyance or detention facilities are to be constructed for purposes of flood control, water quality enhancement features may often be incorporated into such projects for a small, incremental cost. For instance, a "slow flow" wetland channel may be able to be constructed for about the same cost as a conventional grass-lined channel. In addition, a detention pond

PAGE 247

Wulliman 223 with a water quality outlet and a tOO-year outlet may require no more storage volume than a pond with 10and l00-year outlets. Maintenance respon81"ilities. A substantial responsibility, in addition to trapping sediment and other stormwater pollutants, is removing and disposing of the accu mulating material. Such maintenance operations are typically difficult and expensive, and locating disposal sites for removed sediments can be additionally difficult. If pollutant trap efficiency goals are to be incorporated into flood control projects, regional and publicly owned and maintained facilities are more dependably maintained than numerous on-site facilities. Availability of right-of-way. Trap efficiency objectives are most readily incorporated into projects having adequate right-of-way, such as in newly developing areas or in urban areas where ample stream right-of-way exists. Configuration of Enhancement Facilities The configuration of water quality enhancement features, which includes extended detention ponds (with or without permanent pools), wetlands, infiltration facilities, and land application systems, is influenced by the characteristics of the inflowing pollutant load, the specific trap efficiency objectives adopted for the project, and an assessment of the best location to trap sediment and associated pollutants. Sediment entrapment locations. Detention ponds with permanent pools are effective in trapping sediment and are able to "hide" accumulated sediments under the water surface. However, floatable debris can detract from the pond's appearance, and the processes of dewatering wet ponds during sediment removal operations and drying the accumulated material for transport tends to be difficult. Dry ponds with a nonvegetated bottom surface, on the other hand, may facilitate sediment removal operations, but would not hide the material from view. A dry pond with a vegetated bottom surface has the capacity to hide accumulating material and to "rise" with the sediment deposition. However, if the vegetation comprises a wetland, regulatory agencies responsible for enforcing Section 404 of the Clean Water Act may object to the disturbance caused by sediment removal operations. None of the above configurations fits all situations; rather, the layout of water quality ponds and other enhancement facilities must be based on sitespecific and project-specific objectives and constraints. Locating wetlands and infiltration or land application facilities (if used) downstream of water quality ponds would guard against harmful effects of heavy sediment accumulation. Detention pond opportunities. Opportunities exist to incorporate water quality ponds into detention ponds used for peak flow attenuation. Based on a review of pond sizing calculations, a detention pond with a water quality outlet and a lOO-year outlet may require no more storage volume than a pond with 10and lOO-year outlets. Alternatively, including a water quality pond in a facility with 10and tOO-year outlets may add approximately 25% to the required lOO-year volume.

PAGE 248

224 Stormwater Quality Enhancement Features Open channel opportunities. Opportunities also exist to incorporate water quality enhancement measures into open channel design. If inflowing sediment loads are relatively low (because of source control practices or an upstream pond), a slow flow wetland channel may be used. The flat longitudinal slope, hydraulically "rough" wetland vegetation, and ample cross-sectional area of this type of channel produces low flow-through velocities. Low velocities promote settling, increase biological contact time and infiltration potential, and minimize washout of soil and humus. Conclusions Instead of adopting a policy of leaving streams undisturbed, which can lead to their degradation, stream stabilization and the selective incorporation of water quality enhancement facilities into the channel network must be considered as components of a total system of stormwater management. References Chen, Charng-Ning 1975 "Design of Sediment Retention Basins." In Proceedings of National Symposium on Urban Hydrology and Sediment Control. Lexington, Kentucky: University of Kentucky. Cherry Creek Basin Authority 1989 Cherry Creek Basin Water Quality Management Plan, in coopera tion with Denver Regional Council of Governments and Colorado Department of Health.

PAGE 249

Part Seven Arid Region Issues

PAGE 250

This page IS intentionally blank

PAGE 251

FLUVIAL GEOMORPHOLOGY PRINCIPLES APPLIED TO A STREAM REGULATING PROGRAM Joseph V. Borgione and Chad Gourley Utah Department of Natural Resources Water Rights Division Introduction The dynamics of rivers and streams and the formation of channel patterns are governed by eight variables, as described by Leopold and Wolman (1957). These variables include width, depth, discharge, velocity, slope, roughness, amount of sediment, and grain size (caliber) of sediment. Each individual variable is subject to continuous variation, resulting in continuous adjustment of the rest. These adjustments may be only at a micro-scale or in some cases, such as extreme discharge events, the adjustments are clearly noticeable and even devastating. In nature, three basic channel patterns are observed: straight, braided, and meandering (Leopold and Wolman, 1957). These patterns represent the morphologic response to the eight variables. Natural channel stability or lack thereof is a dynamic process, one that is modified through time (Schumm 1977). Migration of stream channels, meander cutoff, bar deposition, and floodplain adjustment can and should be expected in nature. The State of Utah Stream Alteration Program In 1971 the Stream Alteration Act was passed by the Utah legislature. It was updated in 1985, and current rules and regulations pertaining to this legislation are administered by the state engineer. Effective October 1987 and expiring October 1992, the state of Utah is granted a General Permit 040 (OP 40) by the U.S. Army Corps of Engineers (USACE) under section 404 of the Clean Water Act. With this unique agreement, the USACE "authorizes discharges of dredged or fill material into certain streams in the State of Utah provided a State Stream Alteration Permit has been issued" (Scholl, 1987). In effect, the state of Utah acts as an agent of the USACE through the administration of the Stream Alteration Permit Program. It is important to note that the general permit does have specific limitations-the USACE remains in jurisdictional control of lakes and wetlands, and any and all work in these areas must be authorized by the USACE. The general permit does not apply if the proposed project will affect an endangered or threatened species and/or habitats of such species or properties that are eligible for listing on the National Register of Historic Places. If the proposed project includes pushing of streambed materials, as opposed to lifting them, obstructs navigation of waterways, or relocates a stream channel (without specific criteria), the general permit does not apply, and an individual permit from the USACE is required.

PAGE 252

228 Fluvial Geomorphology Principles The exclusions of endangered or threatened species and habitat, as well as historic places, are fairly self-explanatory. However, the other exclusions may need clarification. In the case of lifting versus pushing of streambed material, the state may authorize the lifting out of such material. However, pushing material from one reach of a channel to another is an activity the USACE has elected to oversee under specific Clean Water Act guidelines. If, in the case of a channel relocation, the proposed project is to return river or stream from a channel created during a flood event(s) to the channel configuration prior to those events, the general permit will apply, provided that bank stabilization is not a viable option. A timeframe of five years after the flooding event(s) is generally the limit for the general permit. The GP 40 should not be confused with the USACE nationwide permit. Under the nationwide permit program, certain activities are viewed by the USACE as not needing an individual permit. However, in the state of Utah, if work does fall under the nationwide permit, this does not exempt the need for a state Stream Alteration Permit. It is important to note that the state of Utah does have legal responsibilities as covered by the Stream Alteration Act, regardless of USACE intervention. Despite the complexity of the GP 40, the program does offer a number of advantages. The state engineer has seven area offices across the state. The area engineer of a given region is charged with the duty of administering water rights of the area. Through the normal course of business, the area engineer becomes familiar with local water users and their needs. An integral part of the stream alteration process is input from the area engineer; this includes water rights research, local conditions, and site inspection. The GP 40 program allows water users to deal directly with the agency that administers water use regulation. In most cases, the GP 40 eliminates the need for two permit applications-the state permit application is usually all that is needed. The permit application is designed to facilitate and minimize paperwork for the applicant. A handbook covering the rules and regulations, as well as suggested construction techniques, is available. The Permitting Process To initiate the process, the applicant submits the completed permit application to the appropriate area engineer's office or directly to the Stream Alteration Section Office. The permit application describes the proposed project, lists anticipated detrimental effects, and provides the names and addresses of property owners adjacent to the proposed project. Plans and drawings of the project are also required. The GP 40 requires circulation of the permit application to a variety of state and federal agencies for review, such as U.S. Fish and Wildlife, the state Division of Wildlife, the EPA, and the adjacent property owners. These agencies and individuals are given 20 days to respond to the proposal. If more time is needed it can be requested. When the comment period is closed, or when all concerned parties have submitted comments, the application is approved or rejected based on the merit and integrity of the proposal as well as the comments received from the various agencies. More information may be needed from the applicant, and in some cases a formal

PAGE 253

Borgione and Gourley 229 Environmental Assessment is deemed necessary. However, the majority of applications are approved with standard conditions attached. Annual maintenance applications are accepted and follow the same procedural guidelines already mentioned. These can be renewed annually by written request and must meet the conditions set forth by the original permit. Emergency permits to protect life and/or property may be issued orally or in writing. The state engineer must be notified within 72 hours after emergency work is initiated. Following this notification, a site inspection is made to ensure the work is needed and to see if additional mitigation may be required. Any work performed to alter a natural stream in the state of Utah without an authorized permit, or contrary to permit conditions, is a violation of state law and a Class B misdemeanor. Any violation of state law concerning stream alterations may be in violation of federal law under the Clean Water Act and would be subject to federal prosecution. Geomorphology and Stream Alterations As indicated earlier, before an application to alter a natural stream is approved, careful evaluation by various agencies is required. The state engineer's office provides geomorphic analysis of the proposed alteration. Depending on the project, an in-depth review of the channel geometry, flow characteristics, sediment regime, and bank stability may be necessary. Construction techniques that utilize the natural channel morphology, capacities, and revegetation are encouraged and often required. Dredging of unstable channels to increase capacity is discouraged and is allowed only in extreme cases. This is true with total stream channel relocation as well. If dredging is permitted, material removal should not exceed the original channel depth prior to aggradation. For a stream relocation, sinuosity, channel geometry, and gradient of the new channel should be consistent with those geomorphic parameters of the abandoned channel. All activities should minimize bank vegetation distur bance, and revegetation is normally required. Bank stabilization is a common application request. Erosion of banks in agricultural settings, wild land, and even urban districts makes up a large number of permit applications. The state engineer's office works closely with the state Department of Agriculture and the U.s. Soil Conservation Service in nonpoint source pollution problem solving. Traditional bank protection techniques such as dumping broken concrete or rip rap of entire reaches are discouraged in favor of buried armament and revegetation. The use of automobiles as riprap is not allowed. Jetties placed along bends in a river are an approach that, when implemented correctly, directs the high flows away from the bank, keeps the setting more natural (thus enhancing wildlife habitat and recreational values), and can even minimize construction costs by reducing the amount of costly riprap. Construction of these rock structures is relatively simple; an appropriately sized footing is dug into the channel bottom and into the bank, and rocks of appropriate size are then put in place. By anchoring the jetties into the channel bottom and into the bank, the

PAGE 254

230 Fluvial Geomorphology Principles probability of loss to erosion or downcutting is minimized. The bank is sloped to accommodate revegetation, usually two horizontal to one vertical is recommended. The finished product should give the appearance of nothing more than a few rock piles extending slightly into the channel, with natural vegetation on and in between to anchor the banks, providing stability. Drop structures or grade control structures are another approach that, when correctly utilized, keeps channel velocities from becoming excessively erosive, maintains a consistent base level, and provides fish habitat. As with jetties, appropriately sized rock is anchored into the channel bottom. It is recommended that two parallel rows of rock be used to reduce the chance of undercutting. The rocks should be buried so that they protrude 20-25% of the bankfull depth above the channel bottom (Rosgen and Fittante, 1986). A variation may include a slightly v shaped structure with the apex upstream. As the stream flows over this shape, the current is directed toward the center. Rock gabion baskets are oftentimes used in bank armament. As with any structure used for erosion control, gabions should be well anchored into the channel bottom and bank to prevent underand back-cutting at high flows. If rip rap is used, the same anchoring technique is recommended. The use of concrete pieces is discouraged. If it must be used, then it should be clean (without asphalt) and angular, and any rebar should be cut flush to the slab to eliminate the possibility of injury to stream users. Conclusions Through its agreement with the U.S. Army Corps of Engineers, the state of Utah regulates and maintains the waterways of the state. Water users, recreationists, and wildlife benefit by such a program. The system of permit approval is based on need and merit of the proposed project. Working with the fluvial system and the natural processes that govern stream channel formation and function allows for naturally productive and stable streams and rivers. In most cases construction techniques do not have to be expensive or elaborately engineered to control erosion and maintain capacity of a river or stream. References Leopold, Luna B., and M. Gordan Wolman 1957 River Channel Patterns: Braided, Meandering and Straight. United States Geological Survey, Professional Paper 282b. Rosgen, Dave, and Brenda L. Fittante 1986 "Fish Habitat Structures-A Selection Guide Using Stream Classification. In Fifth Trout Stream Habitat Improvement Work shop. Proceedingsofa Conference HeldAugust 11-14,1986, Lock Haven, Pennsylvania.

PAGE 255

Borgione and Gourley 231 Scholl, Colonel Wayne J. 1987 Written communication. Schumm, Stanley A. 1977 The Fluvial System. New York: John Wiley and Sons.

PAGE 256

CORRELATION BETWEEN FLOW, SLOPE, ROUGHNESS COEFFICIENT, AND REGIME WIDTH FOR UPPER INCISED CHANNELS ON ALLUVIAL APRON OF SUMMERLIN AREA, LAS VEGAS, NEVADA Donald W. Davis Boyle Engineering Corporation Summerlin Description Summerlin consists of a proposed 20,OOO-acre master-planned community on the west side of the Las Vegas Valley. Drainage characteristics consist of flows beginning in high steep rocky terrain. From these flows several small and large washes outlet onto an alluvial apron. Most of the washes are deep, well-defined, and incised into alluvial fan remnants. The washes remain well-defined downstream to a bifurcation point where flows divide into multiple braided channels on active fan surfaces. The planners for Summerlin desired a drainage plan that allowed flood flows to continue in a seminatural regime, utilizing existing natural flow paths as much as possible. To comply with drainage plan objectives, there was a need to define a stable regime width to which flows could be confined. Regime Theory Current methods of determining flood hazard assessments and defining design criteria for alluvial fans primarily involve flow regime theory to predict width CW), depth (y) and velocity (v) of flood flows based on the discharge (Q). The most familiar regime equations are probably those included in the FEMA guidelines, which are based on the following common assumptions: 1) flood events will form their own new channels since the materials deposited from previous floods are highly erodible, 2) channels widen to establish a stable width dependent on the discharge, and 3) the flows tend to flow at critical depth and velocity. FEMA guidelines include a single channel regime on the upper fan and a wider braided mUltiple channel regime on the lower portion of the fan. The braided condition is assumed to be due a change from critical depth to normal depth (Crampton, 1989; Dawdy, 1979, 1981, 1986; French, 1987; Leopold and Maddock, 1953). Another regime equation developed for the community of Cabazon, California, assumes the same width-to-depth ratio adopted by FEMA (stabilizes at 200); however, it is based on the assumption that Manning's equation is more appropriate than the critical depth assumption. This method is dependent on the selection of a roughness coefficient, typically assumed fairly constant, and generally results in a supercritical flow condition (Crampton, 1989; French, 1987; PRC Toups, 1980). While critical flow is generally assumed, studies and observations have indicated that supercritical flows may occur on the upper portions of fans. It may be noted that supercritical flows rarely occur in perennial streams, in part because the roughness coefficient has been found to be directly proportional to a power of the slope (S) and inversely proportional to a power of the hydraulic radius. It is reasonable to assume

PAGE 257

Davis 233 that a similar increase in roughness coefficient with increase in slope would occur on an alluvial fan. French has indicated that if the FEMA and Cabazon regime equations are both considered to hold true, then the parameter of the roughness coefficient over the square root of the slope must be equal to a constant (French, 1986, 1987; Jarret, 1984). Assumptions With consideration to regime theories, the active washes in the Summerlin area were examined. The upper incised washes exhibit characteristics of having confined the flood flows for hundreds of years or more, as evidenced by the formation of caliche (cemented materials) and desert pavement on the fan remnants. These washes have capacity to convey estimated 1OO-year flood flows with reasonable average depths of approximately one to three feet. A hydrologic analysis of the Summerlin area provided 1oo-year flood flow values at numerous locations in an effort to identify the major considerations in a flood control master plan (Boyle, 1990). It was assumed that the major rainfall events that established the channel configurations in the vicinity were proportional to the tOO-year storm. It was also assumed that flood flows approach critical flow conditions. The reach of an incised channel just upstream of the bifurcation point was assumed to be the most likely location for a critical flow regime to occur. The channel has had its greatest opportunity to stabilize before the sudden expansion when the channel slope intersects the fan surface. Results Fifteen locations where single channels divided into multiple braided channels were evaluated. A roughness coefficient (Manning's n) was calculated for each channel reach just upstream of the bifurcation point using the estimated tOO-year discharge, the measured active wash width, the hydraulic radius equal to critical depth, and the measured slope. The results demonstrated a uniform relationship between the calculated n value and the channel slope. The value of n versus slope is plotted for each data point in Figure 1. The results confirm an expected increase in n with increase in slope. Alluvial flows are assumed to approach minimum energy and adjust their sediment load and configuration accordingly. Thus, the more energy a flow has, the more natural energy dissipation occurs. A steeper slope produces greater sediment transport capacity and erosion potential. Erosion and movement of materials consume energy or increase apparent roughness. A steeper slope produces higher velocities, which subsequently produces higher energy losses from disturbances in the flow. Additionally, since finer materials are transported downstream, an upper steeper

PAGE 258

234 Flow, Slope, Roughness, and Regime Width .D9 .os .-........................ -........... ........... ; ............ ............ ............ -:.. -......... j 1 1 ? 07 c U1 C!J 25 .os z z cc L: .os ......................... ....................... 1 ............ .1...... c. . I: :r'" .1 .......... ............ j .......... L ......... : ....... r!!j
PAGE 259

Davis 235 upper steeper slope tends to have larger and coarser materials directly increasing the channel roughness. The value of K, for the expression K = (I) was determined for each data point. French has suggested that this value may be a constant for alluvial fan flow regimes. The value ofK ranged from 0.273 to 0.320. It was found that there was a trend toward a decrease in K with an increase in slope. The value of K versus slope is plotted in Figure 2 for each data point. A regression analysis of the data points yields the following equation for K: K = -0.6625 S + 0.334 (2) This equation is plotted as the line in Figure 2. With values of K determined from the regression equation, the corresponding values of n (n = K S 1h with K from Equation 2) are plotted as the line in Figure 1. An equation for regime width was derived by combining Manning's equation, the expression for n, including K, and equations for critical flow. This reduced to the following equation in English units: W = Q / 956300 Kg (3) where K is dependent on the slope and given by Equation 2. This equation closely approximates the actual measured widths of the existing channels. The widths were compared to the widths determined by the regime equation developed by Dawdy, adopted by FEMA, and given as: W = 9.5 Q.4 (4) A summary of the data, including a comparison of calculated regime widths, is shown in Table 1. Conclusion The data indicates that there is a consistent and uniform relationship for the incised channels in the Summerlin area. Floodways can be designed to maintain a uniform sediment transport capacity and confine flows to a stable width, wide enough to allow for natural energy dissipation based on the developed relations for flow, slope, roughness coefficient, and regime width. The data support that there is a general correlation between slope and roughness coefficient. The data also indicate that regime width may be influenced by slope, although it is generally estimated by methods that are independent of slope. The methodology of calibrating an expression for K and using Equation 3 for regime width may have use for other areas with uniform characteristics.

PAGE 260

236 Flow, Slope, Roughness, and Regime Width Table 1 Data Summary HYDRO-100-YR MANNING MEASURED WIDTH WIDTH LOGICAL FLOW SLOPE DEPTH COEF. WIDTH EQu.3 EQU.4 RATIO RATIO NODE Q S Y n K 10M W3 W4 W3/1oM W4/1oM (cfs) (ft/ft) (ft) (ft) (ft) (ft) RED 15,500 56F 3,300 62F 1,300 46F 3,400 9G 2,650 l1A 540 25A 1,000 lOA 3,500 12A 860 lB 650 lR 3,300 4R 3,800 3A 2,700 21G 1,550 6A 600 Crampton, Walt 1989 .018 3.102 .043 .317 500 434 451 .87 .90 .020 3.234 .045 .319 100 96 243 .96 2.43 .020 3.201 .045 .319 40 38 167 .94 4.18 .024 3.299 .050 .320 100 106 246 1.06 2.46 .033 2.794 .057 .312 100 98 222 .98 2.22 .036 2.829 .059 .312 20 21 118 1.06 .5.88 .038 2.687 .060 .310 40 41 151 1.02 3.76 .045 2.567 .065 .307 150 164 249 1.09 1.66 .050 2.430 .068 .304 40 44 142 1.11 3.54 .059 2.017 .072 .295 40 40 127 1.00 3.17 .060 2.227 .074 .300 175 208 243 1.19 1.39 .067 1.541 .073 .282 350 276 257 .79 .73 .067 1.782 .075 .289 200 196 224 .98 1.12 .070 1.954 .078 .294 100 120 179 1.20 1.79 .083 1.257 .079 .273 75 61 123 .81 1.64 References Appendix III: Technical Basis for Design of Structural Improve ments for the Borego Valley Flood Control Master Plan. Group Delta Consultants. Boyle Engineering Corporation 1990 Summerlin Stormwater Management Plan. Dawdy, David R. 1979 "Flood Frequency Estimates on Alluvial Fans." Journal of the Hydraulics Division, Proceedings of the American Society of Civil Engineers 105 (HYll, November): 1047-1413. 1981 "Flood Frequency Estimates on Alluvial Fans." Journal of the Hydraulics Division, Proceedings of the American Society of Civil Engineers 107 (HY3, March): 379-380.

PAGE 261

Davis 1986 237 "New FEMA Guidelines for Alluvial Fan Flooding." In Proceedings of a Western State High Risk Flood Areas Sympo sium: Improving the Effectiveness of Floodplain Management in Arid and Semi-Arid Regions, pp. 20-23. Association of State Floodplain Managers. French, Richard H. 1986 "Needed Fundamental Hydraulic Research. In Proceedings of a Western State High Risk Flood Areas Symposium: Improving the Effectiveness of Floodplain Management in Arid and Semi-Arid Regions, pp. 184-185. Association of State Floodplain Managers. 1987 Jarret, R.D. 1984 Hydraulic Processes on Alluvial Fans. Developments in Water Science, 31. Amsterdam, The Netherlands: Elsevier Science Publishers B.V. "Hydraulics of High-Gradient Steams." Journal of the Hydraulics Division, Proceedings of the American Sodety of Civil Engineers 110 (11): 1519-1539. Leopold, Luna B., and Thomas Maddock 1953 The Hydraulic Geometry of Stream Channels and Some Physio graphic Implications. Geological Survey Professional Paper 252. PRe Toups, Thielmann, J. 1980 Cabazon Flood Study. June.

PAGE 262

ESTIMATING SEDIMENT DELIVERY AND YIELD ON ALLUVIAL FANS Michael D. Harvey and Robert C. MacArthur Water Engineering & Technology Introduction The Caliente Creek drainage basin (470 square miles) is located at the southern end of the San Joaquin Valley, near Bakersfield, California. A large flood detention reservoir is planned for Caliente Creek, which is an ephemeral stream. The proposed structure is to be located in the medial region of the Caliente Creek alluvial fan, about two miles downstream of the Highway 58 crossing. The highway embankment behaves as a detention structure during flood flows. The present fan surface is incised about 40 feet below the surface of a Pleistocene-age fan and, therefore, the modem fan in the project reach is confined laterally. Feasibility studies showed that the economics of the project were very sensitive to the watershed sediment yield at the proposed damsite. A three-phased Sediment Engineering Investigation (SEI) was conducted in 1990 by Water Engineering & Technology (WET) for the Sacramento District U.S. Army Corps of Engineers to: 1) identify specific geomorphic characteristics of the stream channels and watersheds upstream from the proposed flood control reservoir that would affect the sediment yield at the damsite, 2) estimate sediment production and yield for various frequency precipitation and flood flow events, and 3) circumstantiate estimated annual and single event sediment yields with stratigraphic analyses of the fan sediments (WET, 1990; 1991). Average annual precipitation varies from about six inches at lower elevations to about 45 inches in the higher elevations. Approximately 90% of the precipitation occurs during winter frontal rainstorms from November to April. Very intense localized thunderstorms can occur during the remainder of the year. Floods are generated by both types of precipitation events. Significant floods occurred in 1969, 1978, and 1983. Watershed and Channel Morphology The morphology of the drainage basin is controlled by geology and structural setting. The lower elevation portions of the basin are composed of Pleistocene and recent-age alluvial fans. Tertiary-age nonmarine rocks separate the fans from the majority of the basin that is underlain by quartz diorites. Basin slope and, to a lesser degree, erosion potential relate to the rock type, age, and amount of weathering. Slopes vary from 10 to 15 degrees on the Tertiary-age formations and often exceed 35 degrees on the diorite. Structurally controlled grabens in the upper reaches of the two principal tributaries (Walker Basin and Tehachapi creeks) are depositional centers that impede sediment delivery to the lower basin. Active faults (the Edison, Breckenridge, and White Wolf) traverse the basin. Major earthquakes as recent as 1952 have caused mass wasting of slopes in the watershed.

PAGE 263

Harvey and MacArthur 239 Sediment sources within the basin can be grouped broadly into hillside and channel sources. Dry ravel of the weathered diorite produces sediment from the steep, poorly vegetated slopes in the upper watershed. These sediments accumulate on valley floors and are episodically transported downstream by water floods and debris flows. Individual floods may transport several decades worth of accumulated materials from the valley floor to locations downstream. Sediment delivery to the alluvial fan area and dam site is controlled primarily by the sediment storage potential of various segments of the basin. Storage potential is a function of lithological control as expressed by valley width and channel slope. Depositional reaches characterized by wide Valleys and gentle channel slopes are located along each of the three main creeks and are interspersed between canyon sections that have very low sediment storage potential because they are very narrow and steep. The boundaries separating depositional reaches from transport reaches are lithologically controlled. Channel morphology is also indirectly controlled by the basin lithology. In the canyon sections, bedrock control causes the flows to be perennial, but the morphol ogy of the channel is a function of the recent flood history. Within depositional reaches, flows are ephemeral and channel morphology reflects the last flow that was experienced. Significant infiltration losses occur in the depositional reaches and fan areas that reinforce sediment storage during lower magnitude events. The episodic nature of the sediment delivery process makes traditional estimates of annual sediment delivery rates somewhat meaningless. Sediment delivery from the Caliente drainage basin to the proposed damsite is dependent on the geomorphology of the basin and the episodic occurrence of flood flows. Sediment delivery is also somewhat independent, or at least lagged, from actual sediment production and delivery processes that occur in the upper two-thirds of the watershed. Sediment Yield Estimates To determine the volume of sediment that could enter the reservoir during its design life (100 years), both the average annual sediment yield and single event sediment yields were estimated using a variety of sediment engineering procedures. Morphometric data for the alluvial fan in the vicinity of the proposed reservoir site were obtained from 2-foot contour mapping. Sixteen bed and bank material samples were collected and two Wolman Counts were conducted at representative locations throughout the drainage basin. Stratigraphic analyses of the fan sediments were made in a 3,OOO-foot-long and 6-foot-deep trench dug across the fan and in 10 backhoe pits that were dug on the fan upstream of the highway crossing. Average Annual Sediment Yield Eight different sources of data and/or computational methods were used to estimate the possible range of average annual sediment yields at the proposed damsite. The following sources of data and procedures were used: 1) previous reports and publications of sediment data, 2) Soil Conservation Service (SCS) reservoir sedimentation data, 3) Corps of Engineers reservoir sedimentation survey

PAGE 264

240 Sediment Delivery and Yield on Alluvial Fans data, 4) SCS sediment yield maps for the western United States, 5) average annual sediment yield was computed from the total event sediment volumes for single events ranging from the 2-year event up to the Probable Maximum Flood (PMF) based on channel transport capacity rather that watershed sediment production and delivery, 6) a Corps of Engineers flow duration and sediment load curve integration method was used to estimate the average annual sediment production and yield to the reservoir site, 7) the Pacific Southwest Inter-Agency Committee (pSIAC) method was used to estimate basin-wide sediment yield from the entire watershed, and 8) the Dendy and Bolton Regional Analysis Method for sediment yield was applied. Table 1 presents the estimated sediment yields from the various sources of data and computational procedures. Based on measured sediment accumulation rates recorded in the six Tulare, Kings, and Kern County reservoirs, the approximate range of observed sediment yields is 0.2 AF/sq mi/yr to 2.2 AF/sq mi/yr, with an average of approximately 1.0 AF/sq mi/yr. WET (1990) determined that the sediment delivery and yield at the dam site depends on the channel transport capacity on the valley floor upstream from the reservoir, rather than the watershed production of sediment during a given event. The broad (3,000 to 6,600 feet wide) valley floor contains an almost unlimited supply of easily mobilized sediment. Utilizing this information, the sediment yield to the dam site was estimated from computations of the total event sediment volumes for single events ranging from the two-year event up to the PMF based on channel transport capacity rather than watershed sediment production and delivery. Using area weighing methods on all of the values reported in Table 1, an average annual sediment yield of 0.75 AF/sq mi/yr was obtained. It is important to note that in arid and semi-arid climates, basin sediment yield is very episodic in nature. The annual yield during a dry year may be small, while in an excessively wet year it can be very high. Therefore, the presentation of a single average annual yield value may be misleading. For planning purposes, the consideration of a range of possible annual yields is more meaningful. Single Event Analyses It is important to estimate the sediment production and delivery from single events, because one or more single events (50-year event or greater) during the design life of the project can significantly affect the operation and maintenance of the reservoir. The study reach upstream from the proposed dam site was partitioned into four different zones or subreaches based on distinct hydraulic and geomorphic character istics. The reach-averaged sediment grain size data and averaged channel hydraulic conditions for a range of discharges were used to compute representative water discharge versus total bed material load relationships for each of the subreaches and flow conditions. For discharges with recurrence intervals between five and 100 years, the computed sediment volumes were sufficient to eliminate from about 1.5 to 44 % of the reservoir volume, respectively, which suggests that the reservoir storage capacity may have been under-designed.

PAGE 265

Harvey and MacArthur 241 Table 1 Summary of Estimated Sediment Yields for Caliente Creek Fan Data Drainage Basin, Reservoir or Drainage Source Computational Method Used Area Yield (sq mi) (AF/sq milyr) SCS Blackburn Canyon Reservoir 7.1 2.20 SCS Antelope Canyon Reservoir 4.4 1.50 COE Lake Isabella 2,074 0.37 COE Pine Flat Lake 1,542 0.20 COE Success Lake 393 0.76 COE Terminus Lake 560 0.75 SCS SCS Yield Map of Western US 470 0.47 Computed Integration of the Event Volume vs. 470 0.55 Frequency Curve Computed Flow Duration Method 469 0.90 Computed Dendy & Bolton Method 470 0.71 Computed PSIAC Method 470 0.75 Computed Kern County Water Agency Study 470 0.97 No measured sediment data were available to circumstantiate the computed volumes. However, the stratigraphic analysis of the trench and pits did provide data to check the computed volumes in the reach upstream of the Highway 58 crossing. Deposits from the 1983 and 1978 floods were recognized and measured. Area weighted volumes of sediment were computed, and these provide single event sediment yields for the 1983 and 1978 events of 1.7 AF/sq mi and 1.6 AF/sq mi, respectively. For the 50-year modeled event (1983 flood) the computed sediment yield was 1.2 AF/sq mi, which is less than the value determined from the stratigraphic analysis (1.7 AF/sq mi), but the values are of the same order of magnitude. The discrepancy between the values can be attributed to the complex nature of sediment transport during a flood event. The computed values are based on hydraulic conditions only, but the stratigraphic investigation identified mudflow and hyperconcentrated flow deposits, as well as fluvial deposits, in both the 1983 and 1978 sediments. Conclusions This sediment engineering investigation of both the average annual and single event sediment yields to a proposed dam site on Caliente Creek alluvial fan utilized

PAGE 266

242 Sediment Delivery and Yield on Alluvial Fans both geomorphic and engineering analyses. Eight different sources of sediment yield data produced an average value of 0.75 AF/sq mi/yr, but the range was from 0.2 to 2.2 AF/sq mi/yr (Table 1). Computed single event sediment yields for discharges with recurrence intervals between five and 100 years have the potential to eliminate from about 1.5% to 44% of the reservoir storage, which suggests that the storage capacity is under-designed. Stratigraphic analyses of sediments deposited on the fan during a 50-year flood in 1983 suggest that the hydraulically driven computational methods underestimate watershed sediment yield, because sediment is transported by mudflows and hyperconcentrated flows as well as by Newtonian flows during a flood event. References Water Engineering & Technology 1990 Caliente Creek, California, Project: Geomorphic Analysis. Report, Contract No. DACW05-88-D-0044. Sacramento District, U.S. Army Corps of Engineers. December. 1991 Caliente Creek, California, Project: Field Circumstantiation of Sediment Engineering Investigation. Contract No. DACW05-88D-0044. March.

PAGE 267

CONSIDERING STORAGE EFFECTS ON ALLUVIAL FAN FLOODING Stephen G. King and Edward R. Mifflin Michael Baker, Jr., Inc. Introduction The methodology developed by Dawdy and adopted by the Federal Emergency Management Agency (FEMA) for assessing the risk of alluvial fan flooding considers a flood following an unpredictable path. The cross-sectional dimensions of that path, however, are dependent on the magnitude (Le., discharge value associated with) of the flood. Furthermore, the peak discharge is taken to be constant through any cross-section of the flood path. Thus, if a flow of 10,000 cubic feet per second (cfs) is realized at the apex of an alluvial fan, 10,000 cfs is assumed to be the peak flow of that storm somewhere downfan of the apex, perhaps five or 10 miles downfan. Because alluvial fan flooding is "flashy" by nature, some loss to the peak flow value reSUlting from storage in the path would be expected. This paper investigates the magnitude of such a loss in a channel using the discharge-depth-width relationship for single-channel regions that is found in the FEMA methodology. Summary of Investigation The investigation was performed using the U.S. Army Corps of Engineers HEC1 hydrologic computer model. The losses resulting from storage within the flood path were modeled using the reservoir routing routine in HEC-l. Specifically, a flood path that is 50,000 feet long was simulated as 10 consecutive reaches (reservoirs), each 5,000 feet in length, with a cross-sectional area proportional to the 0.8 power ofthe outflow. That is, the storage volume-discharge relationship assumed was v = 5000 A = (0.11478)(0.6638) QO.8 = 0.0762 QO.8 where V is the storage volume in acre-feet, A is the cross-sectional area in square feet, and Q is the outflow in cfs. The input data for the reservoir routing are provided in Table 1. Figure 1 is the (input) hydro graph at the apex of the fan. Note that the hydrograph depicts a flash flood for appreciable flows. Also note that the flows are given as a percentage of the peak flow. This hydrograph was developed using the Soil Conservation Service dimensionless Unit Hydrograph option ofHEC-l (with a one-hour lag) and a uniform runoff with a one-hour duration. Five peak discharge values, ranging from 100 to 10,000 cfs, were investigated. The reduction in peak discharge after 50,000 feet ranged from 65 % to 93 %. The results are presented in Table 2 and shown in Figure 2.

PAGE 268

244 Distance Down Fan o 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 50,000 Storage Effects on Alluvial Fans Qutflow 0 3 5 10 15 20 40 60 80 100 200 300 400 500 1,000 2,000 4,000 8,000 10,000 20,000 Table 1 Input Data Depth 0.000 0.109 0.134 0.177 0.208 0.234 0.309 0.363 0.407 0.445 0.587 0.691 0.775 0.848 1.118 1.476 1.947 2.569 2.809 3.707 Table 2 Results Percent of Peak Discharge Retained Volume 0.00 0.18 0.28 0.48 0.67 0.84 1.46 2.02 2.54 30.30 5.28 7.31 9.20 10.99 19.14 33.32 58.02 101.02 120.76 210.26 100 cfs 500 cfs 1,000 cfs 5,000 cfs Peak Peak Peak Peak 100 100 100 100 96 98 98 99 92 96 96 98 88 93 95 97 84 91 93 97 80 89 91 96 77 87 90 95 74 85 88 94 70 83 87 93 68 80 85 92 65 78 84 92 10,000 cfs Peak 100 99 98 98 97 97 96 95 94 94 93

PAGE 269

King and Mifflin 245 Figure 2 indicates that the reduction in peak discharge is proportional to the length of the flood path (Le., the relationship is approximately linear). Note that the higher the peak discharge value, the smaller the proportionality constant. That is, the percent reduction at a given distance downfan for a discharge having a peak of 10,000 cfs is less than the percent reduction of a discharge that has a peak of 100 cfs. Conclusion This investigation indicates that the losses in alluvial fan flooding due to storage are less than would be expected. However, there are other factors, such as infiltration, that should be considered in determining the losses in alluvial fan flooding. Dawdy, D.R. 1979 References "Flood Frequency Estimates on Alluvial Fans." Journal of the Hydraulics Division, ASCE Proceedings 105 (HYll): 1407-1413. Federal Emergency Management Agency 1990 FAN: An Alluvial Fan Flooding Computer Program, September. u.S. Army Corps of Engineers, Hydrologic Engineering Center 1981 HEC-J Flood Hydrograph Package, September. Revised March 1987.

PAGE 270

246 Storage Effects on Alluvial Fans 100 90 80 0. 70 ... 0 I-60 z w U 0: W 50 0. VI w 40 0: 30 J: U VI i5 20 10 0 0 100 200 300 400 TIME, minutes Figure 1. Hydrograph at apex of fan. 0.95 w 0: 0.9 J: u VI i5 0.85 0. ... 0 w O.B z w u 0: 0.75 w 0. 0.7 0.65 +----;,..---,----r----,.---r---,..---,----r----,.---='l!J o 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 DISTANCE DOWN FAN o 100 cfa + 500 cfa o 1,000 eta 11 5,000 ets x 10,000 cfa Figure 2. Results.

PAGE 271

OPTIMIZATION MODELING FOR FLOOD CONTROL ON DISTRIBUTARY FLOW AREAS IN THE SOUTHWEST James R. Morris Arizona Department of Water Resources Introduction The arid and semi-arid desert environments of the southwestern United States are distinctly different and more complex fluvial systems than those found in the more humid areas of the country. In a recent publication, Ward (1988) describes the differences as "a function of such factors as short duration, high intensity rainfall, abrupt changes in topography, and a sparse vegetation community which creates the relatively bare surface conditions of desert soils." One of the unique landforms found in the southwest is the alluvial fan. Most alluvial fans have areas of distributary flow where the channels may split, recombine, and split again as they progress downslope. Distributary flow areas may be stable, where the distributary channels are separated by ridges of considerable relief, or they may be unstable, with little level difference between channels. Flows from the apex of the fan move downslope in varying directions, depending on the age, type, and geomorpholic characteristics of the apex area. At some point, usually a change in slope, the flow will typically enter a distributary pattern. Development of alluvial fans that have distributary flow areas requires that the whole fans be considered in any future development planning. This chapter explores the use of an optimization model to determine a least-cost flood control solution for such an area. Distributary Flow Pattern Distributary flow channel networks can occur on alluvial fans of any age. When entrenched, they are an indication that the alluvial fan has moved from a depositional landform to an erosional one. On very old fans, the distributary pattern is usually replaced by a single, or several separate, well-incised channels. Human activity on the distributary flow area may cause changes greatly in excess of the size of the construction, filling, or grading. The construction of roads and ditches and the installation of culvert structures can result in a change of flow patterns and flood depths in parts of the distributary system far from the actual site of the activity. Flood hazard determination is difficult at best because the pattern and number of channels may change after each storm. Figure 1 illustrates the hypothetical channel network used in this study. It consists of 25 channel segments of various lengths and directions of flow. Actual distributary flow areas may cover tens of square miles from the apex, or entry point of the flow, to the discharge point at the toe of the slope.

PAGE 272

248 Optimization Modeling for Flood Control Figure 1. The hypothetical distributary flow network. Model Development The model developed for this study was conceived as a reconnaissance-level tool to be used to generalize the best flood control channel alignment and to be used in conjunction with studies of detention/retention and nonstructural solutions to develop a total area plan. Several simplifying assumptions were made to aid in model formulation. These assumptions and the form of the model are discussed in this section. A nonlinear programming model was developed for this study. The possible flow paths for a single channel flood control solution can be determined from Figure 1. One possible path is from 1 to 3 to 5 to 20, where the numbers designate flow split locations. The following parameters are required for each channel segment: 1. Channel length in feet 2. Average channel slope in feet per foot 3. Construction costs per square foot of channel in dollars 4. Land acquisition costs per square foot of channel in dollars. Additional parameters are: 1. Flow rate at the apex in cfs. 2. Manning's n value for the channel type. Channel width (W) and flow depth (Y) are solved as a nonlinear constraint using Equation 1, a formulation of Manning's equation: Q = 1.49(WY)exp(5/3)(S1I2n(W+2Y)exp(2/3 (1) where Q is the flow, n is Manning's n value for the channel, Y is the depth of flow, S is the average channel slope, and W is the channel width.

PAGE 273

Morris 249 The total cost of the channel is minimized using the General Algebraic Modeling System (GAMS) Version 2.04. The major components of the objective equation are listed in Equations 2 through 4. Land Acquisition Costs: LC = L(W + l(0)Lc (2) where L is the channel length, W is the channel width, and Lc is the land cost per square foot. W + 100 includes a 100 foot maintenance right-of-way. Construction Costs: CC = LWCc (3) where L is the channel length, W is the channel width, and Cc is the construction costs per square foot. Design Cost: DC = .15CC (4) where Cc is the total channel construction costs. The objective function takes on the form of Equation 5 using the components listed above: Minimize Cost = SUM(LC + CC + DC LEB ) (5) where LEB is the land enhancement benefits obtained by not building the channel in the other potential locations. Equation 5 considers only potential single paths by the inclusion of the constraint outlined in Equation 6. Q = SUM(Qd) (6) where Q is the inflow to the system and Qd's are the individual flows at the end of each path. Results The assumed input parameters for the study were selected at random to model the variability of land costs and slopes that might occur in a developing area. The use of unequal geometric parameters and varying land costs would be expected in any real application of this kind of procedure. The highlighted path shown in Figure 1 was selected by the model as the least cost alternative. The total cost of the channel system selected was approximately $15 million dollars, indicating that the net land enhancement benefits outweigh the cost of channel construction. Summary and Conclusions The model developed in this study is a tool for evaluating one of the alternatives that should be a part of a total management plan for a distributary flow area. Optimization techniques allow for a complete exploration of factors and conditions that might affect the single channel structural flood control solution. The model might be improved in a number of ways: 1. The model assumes that delivery of the apex flows to the toe at any of the channel segment ends is acceptable. In actuality, this flow may exceed the flow normally found at the toe.

PAGE 274

250 Optimization Modeling for Flood Control 2. It would be possible to add maximum slopes or velocities as a constraint and allow the inclusion of grade control structures and their relative cost. 3. Another option not explored in this study might be the relationship between construction of multiple channels down all alignments. Land enhancement benefits would still be derived from the removal of the area from the FEMA alluvial fans designation, and local drainage concerns on the fan might be easier to mitigate. In summary, the use of optimization techniques could provide significant benefits in the design of flood control solutions on alluvial fan areas of the Southwest. Additional research into the problem from this viewpoint would be warranted. References Meeraus, A. and Brooke, A. 1987 PC-GAMS Program and User's Manual. Ward, R.L. 1988 Present Status of Management and Technical Practices on Alluvial Fan Areas in Arizona. Report Number FHW A-AZ88-278. Arizona Department of Transportation.

PAGE 275

CONTINUING RESEARCH ON TWO-DIMENSIONAL MODELING OF FLOOD HAZARDS ON ALLUVIAL FANS I.S. O'Brien and W.T. Fullerton FLO Engineering Introduction The application of two-dimensional models to simulate alluvial fan and river flood hazards is becoming widespread and in some cases, a necessity. Over the last two years a two-dimensional flood routing model, FLO-2D, has been applied on numerous projects to investigate both clear water flooding and mud flow hazards on urbanized alluvial fans and to predict hydraulics for flood mitigation design. It is a flexible tool that enables the engineer to examine a number of unconfined flooding scenarios and flood mitigation schemes in rapid succession. FLO-2D simulates flow over complex topography and roughness, accounting for the effects of buildings and obstructions. Recent advances in the model include the simulation of fan rainfall, interception, infiltration, and channel transmission losses. The effects of these hydrologic processes and a comparison with the FEMA alluvial fan flood prediction methodology are presented in an example application to a large fan in southern Nevada. Description of the FL0-2D Model The focus of modeling alluvial fan and river floodplain flows is to predict the area of inundation and a general range of flow hydraulics, such as velocity and depth. FLO-2D predicts flow depth and velocity by solving the continuity equation and the two-dimensional momentum equations. The momentum equations are approximated by diffusive wave equations. The model uses a finite, central difference routing scheme and uniform grid elements. For a derivation of the constitutive equations, see Lenzotti & Fullerton (1990a). FLO-2D can simulate channel flow, street flow, two-dimensional overland flow, and the interflow between these, including return flow to the channels and streets. When routing a flood hydrograph with FLO-2D, the flow regime in channels, streets, or on overland surfaces can vary between supercritical or subcritical. Main channels are represented by variable geometry cross-sections. Street flow, modeled as shallow rectangular channels, can be simulated along two perpendicular axes. FLO-2D can simulate complex urban flood problems with flow in streets and around buildings (O'Brien and Fullerton, 1990). The effects of flow obstructions, such as buildings or walls, that limit floodplain storage or constrict flow paths are simulated with grid element area and width reduction factors. The outflow from bridges and culverts can be simulated, as well as their backwater effects, which may include flow over adverse slopes. Mud and debris flows can also be simulated with FLO-2D. Routing these hyperconcentrated sediment flows requires supplemental equations to define flow motion, including relationships between viscosity and sediment concentration and

PAGE 276

252 Two-Dimensional Modeling on Alluvial Fans yield stress and sediment concentration. When simulating mudflows, not only can peak flow attenuation be predicted, but the bulking and slowing of the frontal wave as it progresses over a dry fan surface can be modeled. This paper will focus on clear water floods, a discussion of mud flow theory, and results are presented in O'Brien and Fullerton (1990). A Case Study-Flooding On A Semi-Arid, Large Alluvial Fan In 1990, FLO Engineering conducted a flood hazard delineation analysis of a very large, semi-arid alluvial fan in southern Nevada. The purpose of this study was to evaluate the flood hazard near a proposed development at the fan terminus. A railroad embankment upstream of the development cut laterally across the bottom third of the fan. Discharge through several culverts in the railroad embankment could impact the proposed development site and needed to be analyzed. Both FLO-2D and the FEMA alluvial fan methodology were applied. This fan, shown in Figure 1, was relatively uniform in shape and nearly five miles in length from fan apex to fan terminus. Its conical shape and lack of development made it an ideal fan for the application of the FEMA methodology for predicting hydraulic boundaries (Figure 1). There were no major incised channels or other topographic features to otherwise alter the flow path on the fan. Fan slope and roughness values were represented by a single value as dictated by the FEMA method. RESORT DEVELOPMENT 3.5 FPS 4.5 FPS 7.5 FPS 8.5 FPS FAN APEX Figure 1. Alluvial fan topography and FEMA hydrauliC boundaries.

PAGE 277

O'Brien and Fullerton 253 The 2-,5-, 10-,25-,50-, and loo-year return period flood hydrographs resulting from six-hour storms in the upstream watershed were determined at the fan apex using the U.S. Army Corps of Engineers HEC-l program. The predicted peak discharges for each event were used in applying the FEMA method. With the FLO-20 model, the loo-year flood hydrograph was routed over the alluvial fan using a 15 % bulking factor for sediment loads. In addition, fan rainfall and the runoff from seven small contributing watersheds were routed over the fan. Results FLO-20 predicts flood depths in excess of one to two feet over most of the fan, with a maximum depth of six feet and a maximum velocity of 13 iPs near the fan apex. Ponding upstream of the culverts is predicted. The main body of the flood follows a path to the northwest where the fan is steeper. The FEMA method predicts shallow flooding after midfan not to exceed 0.5 foot depth and 3.5 iPs velocity with maximum depth and velocity of3.5 feet and 8.5 iPs, respectively, near the fan apex. The influence of the railroad embankment, the ponding at the culverts, and the concentration of flow downstream of the culverts could not be predicted by the FEMA method. The FEMA method results are compared with the FLO-20 results in Figure 2. COMPARISON OF FLO-2D AND FEMA PREDICTED FLOW DEPTHS II HlrY 1-15 r 1 t 1 RESORT lC' -to PREIDICTE H DEVELOPMENT-1\ L W D pm o T U -.. I -U1V[ ACNIIA RI I If 1 1 I 1 0 FEMA DEPTH J BOUNDARIES I 0.5 IT II 0, 1;-1.5 IT 1 Ir-2.5 IT ., 1/3.5 IT 1 I I lrFAN APEX FLOW DEPTII CONTOUR INTERVAL = 1 IT J.,.'f 6 1 I'--POINT OF 1 UNION PACIFIC I BIFURCATION RAILROAD I Q 1 I 1 I ; I GRAPHIC SCALE I I ---!"ii"Ii '--I '\ Figure 2. Comparison of FLQ-2D and FEMA predicted flow depths.

PAGE 278

254 Two-Dimensional Modeling on Alluvial Fans The FEMA methodology is not conservative when compared to a physically based model such as FLO-2D. The limitations of the FEMA methodology are discussed in a companion paper in this volume (O'Brien and Fullerton, 1991). In 1990 a flood event estimated to be less than the 2-year return period flood occurred on the fan. Evidence of flood depths of approximately one foot were observed at the culverts and near the development. These depths exceeded the FEMA predicted shallow depth boundary .5 foot) for the tOO-year flood more than two miles further downfan. Recent FL0-2D Advances Improve Flood Simulation The alluvial fan in question was so long that it was impossible to disregard fan rainfall, interception, and infiltration and the possible effects that these might have on the flood hydrograph as the storm moved over the fan and into the upstream watershed. Infiltration and channel transmission losses on a dry fan can be significant and slow the advance of the flood wave as it progresses downfan. These losses were predicted using the Green-Ampt infiltration model. Details concerning the choice of hydrologic and infiltration parameters can be found in Lenzotti & Fullerton (1990b). Rainfall on the fan reached the railroad embankment culverts prior to the arrival of the flood wave from the upstream watershed and contributed to the aerial flood distribution on the lower terminus of the fan. Fan infiltration and channel trans mission losses reduced the magnitude and slowed the arrival of the frontal wave, enabling the peak flood wave to overtake the frontal wave. As the flood progressed over the fan, it was revealed that most of the predicted infiltration occurs during the initial rainfall on the fan, and then during the rising limb of the hydrograph, until the soil is saturated. The peak discharge is reduced by about 14% because of overland infiltration and channel transmission losses. Often fan hydraulics are governed by a network of small channels. Aerial photo graphs and field observations of the project site indicated the presence of innumer able small braided channels on the lower portion of the fan. A multiple channel component was designed to replace overland sheet flow with flow in a series of multiple shallow channels. The number of channels for a given portion of the grid system is specified together with a width/depth ratio. The multiple channel routine distinguishes between overland sheet flow originating as rainfall and the flow in the channels. The local overland sheet flow enters the channels for conveyance to a contiguous grid element. The temporally varying depth is used to calculate the channel widths in each grid element. Conclusions Simulating fan rainfall and infiltration is important to the accuracy of modeling alluvial fan flooding. On large alluvial fans, the fan rainfall can initiate flooding prior to the arrival of the flood wave from the upstream watersheds. Fan infiltration and channel transmission losses can attenuate the peak discharge. When combined with the new mUltiple channel routine, simulating fan rainfall and infiltration enables

PAGE 279

O'Brien and Fullerton 255 FLO-2D to more accurately model the complete physical system of alluvial fan flooding. References O'Brien, J.S., and W.T. Fullerton 1990 "Urban Floodplain and Alluvial Fan Stormwater Modeling." In Transferring Models to Users and Urban Hydrology: Proceedings of the 26th Annual AWRA Conference. Denver, Colorado. 1991 "FEMA Method for Predicting Flood Hydraulic Boundaries on Alluvial Fans Requires Verification." In Inspiration: Come to the Headwaters, Proceedings of the 15th Annual ASFPM Conference. Denver, Colorado. Lenzotti & Fullerton Consulting Engineers 1990a Flood Hazard Delineation on Alluvial Fans and Urban flood plains. Breckenridge, Colorado. 1990b Off-Site Alluvial Fan Flood Hazard Delineation, Southern Nevada. Breckenridge, Colorado.

PAGE 280

FEMA METHOD FOR PREDICTING FLOOD HYDRAULIC BOUNDARIES ON ALLUVIAL FANS REQUIRES VERIFICATION I.S. O'Brien and W.T. Fullerton FLO Engineering Introduction The Federal Emergency Management Agency (FEMA) has published a user's manual for its probabilistic approach to delineating hydraulic boundaries on alluvial fans. The manual describes the application of the computer program FAN, an encoded version of the FEMA method. The FEMA method has been the subject of much scrutiny (Fuller, 1990; Baker et al., 1990; French, 1987; Grindeland et al., 1990). The inherent assumptions in the method do not apply to most alluvial fans, and it should not be used on urbanized fans. The limited available field data do not support its continued application. FEMA's Alluvial Fan Method for Predicting Hydraulic Boundaries Most alluvial fan flood studies are conducted because of underlying interest in flood insurance or flood mitigation for existing or proposed development. FEMA's method for assessing flood hazard zones on alluvial fans combines a single channel method (referred to as the Dawdy method [Dawdy, 1979]) and a multiple channel method that is applied downfan of the single channel point of bifurcation (DMA, 1985). The depth and velocity boundaries are determined in one-foot or 1 fps increments, respectively. Flood insurance rates are based on predicted hydraulic boundaries with a flow depth exceeding 0.5 foot. These boundaries have a risk assessment of 1 % corresponding to the loo-year return period flood. The only data required to apply the FEMA single channel model are several peak discharges from the upstream watershed and their associated return periods. The hydraulic boundaries are expressly dependent on the skew coefficient of the flood frequency curve of un gaged watersheds. For the multiple channel method an average fan slope and Manning n roughness value are also required. The resulting velocity and depth boundaries, when superimposed on a topographic map, are intended to define the physical extent of flooding. The data base required by the method is insufficient to adequately address the complex physical processes of unconfined flow over an alluvial fan. Discussion of the Assumptions and Limitations of the FEMA Method An in-depth investigation of the FEMA method reveals that it is based on a set of assumptions implicit in the use of channel regime equations (Dawdy, 1979; DMA, 1985; French, 1987). The method also incorporates various assumptions regarding channel avulsion and critical flow. Concern over these assumptions and the effect of the FEMA method results on floodplain management have been expressed by others

PAGE 281

O'Brien and Fullerton 257 (Fuller, 1990; Baker et al., 1990). A brief discussion of several of the assumptions in the FEMA method follows. FEMA's method assumes an equal probability and random distribution of a single channel across a given fan contour. The fan is assumed to have a uniform suiface at locations equidistant from the fan apex. The application of the method should be limited to undeveloped, uniform topography alluvial fans (fans with an idealized, plain conical surface). When applied to a fan with complex topography, local relief is ignored. Within a predicted FEMA depth boundary, large variation in actual flow depth will occur. FEMA elevation requirements for structures within a given boundary are the same regardless of whether the structures are designed for a channel bottom or a higher ridge. The portions of a fan that are steeper or have existing channels have a higher probability of flow occurrence. The flow is assumed to be confined to an identifiable channel that migrates over the fan during the flood event. Channel migration is based on an assumed probability of avulsion of 0.5 reSUlting in a channel avulsion factor of 1.5. This assumes that there is a 50% chance that the channel will avulse upstream of a given hydraulic boundary during the peak flow. The avulsion factor has the consequence of increasing the fan width corresponding to a particular depth or velocity boundary and extends the boundary further downfan. A 1.5 avulsion factor results in a 50 % larger boundary width than if no avulsion was considered. Considering the actual process of channel avulsion, avulsion increases neither the physical width of the potential flow surface nor the downfan flow hydraulics. Although any value can be assumed, the avulsion factor has not been verified and there is no basis for choosing a specific value. Channel migration does not occur during every flood event and depends on debris loading, channel incision, and potential for channel erosion. Channel avulsion is more likely to occur on small braided channels than on large incised channels. DMA (1985) reports that avulsions occur on a geologic time scale rather than over a "planning horizon." The concept of channel avulsion is less important for large flood events, where flow depths overwhelm small obstructions and tend to follow the steepest route downfan. FEMA's method does not consider development and flood control measures that affect the flow path and discharge. The assessment of fan flooding is generally associated with development where the flow paths have been altered by buildings, levees, or debris basins. FEMA's method assumes the flow is unobstructed or unconstrained. FEMA addresses this problem: In portions of alluvial fans in which natural alluvial fan processes may not occur, such as in areas of entrenched channels, areas protected by flood control works and heavily developed areas, the Study Contractor should

PAGE 282

258 FEMA Alluvial Fan Method exercise good engineering judgement in determining the most appropriate methodology (FEMA Guidelines, 1985). Nevertheless, FEMA's method is routinely applied to developed fans. Fan hydraulics predicted with the FEMA method are based on empirical channel geometry relationships not representative of self-Jorming channels on alluvial fans. The expressions for channel width W and flow depth d as function of jeak discharge Q, as based on the Dawdy method, are (Dawdy, 1979): W = 9.5 QO. and d = 0.07 QO.4. These equations result in a constantwidth-to-depth ratio WId = 136 regardless offan slope, channel roughness, and bed and bank material, all of which cause variation in channel geometry in the downfan direction or between different fans. FEMA's single channel method assumes critical flow regime. Supercritical flows on steep alluvial fans are not uncommon. Variable slope and roughness contribute to the transient nature of fan flows and an assumption of critical flow may result in underestimated flow velocities that might impinge on a flood structure. FEMA 's method assumes there is no attenuation or surging of the peak discharge as the flow progresses over the fan. Flood wave attenuation and surging cause large variation in the flow depth and velocity as the flood progresses over the fan. Flood storage, variable roughness and topography, fan rainfall, and infiltration all affect the peak discharge. On large alluvial fans the increase in discharge from fan rainfall can be significant and flood wave attenuation can be affected by infiltration and channel transmission losses. Fan hydraulics are unaffected by sediment concentration. The FEMA method predicts flow hydraulics for water only. A sediment concentration of 40% by volume would bulk the peak discharge by 1.67 times that for water. Besides discharge bulking, high sediment concentration can result in a significant variation in the fluid properties of viscosity and yield stress, which will effect flow depth and velocity. Verification of the FEMA Methodology There is a paucity of alluvial fan flood data. In 1985 FEMA contracted with DMA to conduct a study to verify the assumptions of the FEMA method. The DMA report is cited in support of the FEMA method. Selected flood data were analyzed on 11 fans in Nevada and California. No hydrologic analyses were conducted on the watersheds to assign flood frequencies. Unit runoff for the storms used to calculate channel widths varied from 40 to 10,000 cfs/mi2. Most of 11 storms analyzed were significantly less the 1OO-year event. The channel width predicted by the FEMA method was compared with "observed" widths. Observed channel widths were estimated from aerial photos

PAGE 283

O'Brien and Fullerton 259 taken, in some cases, years after the flood had occurred. A comparison of the observed width and FEMA method predicted single channel width resulted in an 18% average difference for 10 of the 11 fan floods-apparently, a good correlation. Would the channel width/depth ratios compare as well? The FEMA predicted width/depth ratio of a single channel is constant 136. From the DMA (1985) report, W /d ratios can be computed from reported discharges, observed widths, and using the previously described depth-discharge equation for the 11 alluvial fan floods. The Wid ratio ranged from 19 to 289, a poor comparison with the FEMA value of 136. Nevertheless, these results led DMA to conclude, "the width of a single channel can be reasonably determined by the present FEMA method." DMA was unable to substantiate an avulsion coefficient of 1.5. There was insufficient data to even attempt a statistical correlation. DMA concluded that, "the present data base is insufficient to better define the avulsion coefficient. Therefore, the present value of the avulsion coeffident should continue to be used." A questionable conclusion! The conclusion should have been drawn that there was no data to justify an avulsion coefficient. Finally, DMA plotted channel position near the fan apex and concluded that the evidence supported the FEMA method assumption of random channel locations on the fan. The channel direction, however, was not correlated with the fan slope. If the channel path did not correlate with the steepest portion of the fan, then the randomness conclusion would have been justified. The FEMA method requires verification. The DMA study accomplished nothing toward this goal. There are no published results to verify the method with measured flood depths and velocities. Conclusions Application of the FEMA method is often requested on alluvial fan studies by either the floodplain manager or FEMA's technical representative. Although intended for the purpose of assessing flood insurance rates, the FEMA method has been used to evaluate the design of mitigation measures. Consequently, there is a concern that the FEMA method results may be incorporated into the design of flood containment structures. Several conclusions can be drawn from this review: I. The theoretical derivation of the FEMA model is flawed. The FEMA method for the prediction of flood hydraulics on alluvial fans does not present a realistic analysis of the physical processes on alluvial fans. 2. There are technical deficiencies in its application, and there are no published guidelines on the limits of its applicability. Highlighted in text was the lack of verification of its assumptions, empirical equations, or results. 3. The method does not err on the conservative side and it should not be used for flood mitigation design.

PAGE 284

260 FEMA Alluvial Fan Method 4. Finally, the risk assessment based on the FEMA method does not reflect the potential of the flood hazard. References Baker, V.R., K.A. Demsey, L.L. Ely, I.E. Fuller, P.K. House, I.E. O'Conner, I.A. Onken, P.A. Pearthree, and K.R. Vincent 1990 "Application of Geological Information to Arizona Flood Hazard Assessment." In Proceedings ASCE Symposium on Hydraul icslHydrology of Arid Lands. San Diego, California, pp. 621-626. Dawdy, D.R. 1979 "Flood Frequency Estimates on Alluvial Fans." J. Hydr. Div., ASCE, Vol. 105, pp. 1407-1413. Dawdy, D.R., I.C. Hill, and K.C. Hanson 1989 "Implementation of FEMA Guidelines on Alluvial Fans." In Proceedings of the 1989 ASCE National Conference on Hydraulic Engineering. New Orleans, Louisiana, pp. 64-69. DMA 1985 Alluvial Fan Flooding Methodology and Analysis. Prepared for FEMA by DMA Consulting Engineers. Rey, California. Federal Emergency Management Agency 1985 Guidelines and Specifications for Study Contractors. Federal Insurance Administration, Washington, D.C. French, R.H. 1987 Fuller, I.E. 1990 Hydraulic Process on Alluvial Fans. New York: Elsevier Science Publishing. "Misapplication of the FEMA Alluvial Fan Model: A Case History." In Proceedings ASCE Symposium on Hydraulics/ Hydrology of Arid Lands. San Diego, California. pp. 367-372. Grindeland, T.R., I.S. O'Brien, and R.M. Li 1990 "Flood Hazard Delineation on Alluvial Fans." In Proceedings of the ASCE Symposium on Hydraulics/Hydrology of Arid Lands, San Diego, California. pp. 268-373.

PAGE 285

FLOODPROOFING DEVELOPMENT ON ALLUVIAL F ANS* Vassilios A. Tsihrintzis and Blake N. Murillo Psomas and Associates Michael E. Mulvihill and William 1. Trott Loyola-Marymount University Abstract Techniques used to calculate flow rates, debris productions, and sediment transport within alluvial fan zones located upstream of proposed development projects are presented. Computed parameters are used in the design of a network of open channels that collect the flood flows on the upstream side of the project, carry it safely through, and disperse it on the downstream side. Introduction The project boundary encompasses 1,574 acres located in Coachella Valley, Riverside County, California and consists of development of a community of approximately 5,800 homes and several golf courses. It is located within the lower reaches of a 225-square-mile watershed and is within the twoto three-mile wide, twoto three-foot deep floodplain of Coachella Valley. The watershed contains five canyons: Long Canyon, Eastwide Canyon, Thousand Palms Canyon, Pushawalla Canyon, and Edom Hills. The project site, located within the braided channel zone in the coalescent alluvial fan of these canyons, extends nearly perpendicular to the flowpath. Existing Conditions Based on Riverside County ordinances, the proposed housing pads have to be protected from the lOO-year storm runoff. A hydrology study was undertaken utilizing HEC-I to determine design flows approaching the upstream (west and north) sides of the property. Since rainfall-runoff data were nonexistent, model parameters for HEC-I were established from previous similar studies in the project area. The uncertainty of the direction of flows in alluvial fans required consideration of two alternative flowpaths (see Figure I): Thousand Palms Flowpath No. I assumes that the flow from Thousand Palms Canyon hits the northern part of the west side of the project; Thousand PalmslEdom Hill flowpath assumes that the flow from Thousand Palms Canyon travels southerly (Thousand Palms Flowpath No.2) and combines with the flow from Edom Hills. Pushawalla Flowpath hits the northern side of the property. Computed flood peaks were 23,000 cfs for the Thousand Palms *The project was funded by Del Webb Corporation, California.

PAGE 286

262 } .. -. Floodproofing Development on Alluvial Fans \ \' 1 I' I I I I I I I I I I I .I I I : c:l. 'en '.' 00 xc 1-1<1 .; Figure 1. Alternative Flood Flow Approaches to the Project Boundaries

PAGE 287

Tsihrintzis, Murillo, Mulvihill, and Trott 263 Flowpath No.1, 30,000 cfs for the Thousand PalmslEdom Hill Flowpath, and 3,200 cfs for the Pushawalla Flowpath. Several debris computation methods (the Pacific Southwest Inter-Agency Committee Method, the Tatum Method, the new U.S. Army Corps of Engineers Los Angeles District Method, the Universal Soil Loss Equation, and the Modified Universal Soil Loss Equation) were used to estimate l00-year debris productions from the drainage area. Results of these methods were found to vary widely. How ever, the Los Angeles District Corps of Engineers Method was selected for this project because conditions used to derive it were closer to the project's conditions. This method predicted 1,585,000 cubic yards of debris from Thousand Palms, 1,660,000 cubic yards from Thousand PalmslEdom Hill, and 240,000 cubic yards from Pushawalla Canyon. Based on soil samples throughout the drainage basin and considerations of the topography and soil type and condition, it was felt that debris production methods give very conservative results. Most of the debris produced in the canyons is expected to settle in the upper areas before reaching the project. Sediment loads entering the project will result from channel degradation within the lower reaches of the fans. In evaluating channel sediment transport, a channel was assumed to be eroding within the alluvial fan. Flow hydraulics were based on the Edwards-Thielmann (1982) method in the single channel reach of the fan and on the DMA (1985) method in the braided channel reach of the fan. Three methods were considered for sediment transport analysis: the Meyer-Peter, Muller bed load equation in combination with Einstein's suspended load computation procedure, the Yang total load formula, and the Ackers and White total load formula. In a given channel, these methods were applied for a series of flows and the corresponding sediment transport capacities were computed. Regression curves of the form Q s = m Qn were then fit through the calculated data to correlate sediment transport capacity Q s to flow rate Q. These regression equations were used in conjunction with the l00-year storm runoff hydrographs to derive sediment hydrographs and compute sediment volumes. Several channel reaches were identified based on slope and bed material characteristics. Sediment load into the project site results from the transport capacity of the Thousand Palms Flowpath No.1, Thousand Palms/Edom Hill Flowpath, and Pushawalla Flowpath. Computed sediment transport volumes were 95,000 cubic yards for Thousand Palms Flowpath No.1, 446,000 cubic yards for Thousand Palms/Edom Hill Flowpath, and 13 ,000 cubic yards for Pushawalla Canyon.

PAGE 288

264 Floodproofing Development on Alluvial Fans Flood Control Channel Design In order to provide the required level of flood protection to the proposed on-site properties, the flood flows are intercepted on the western project boundary in collection basins and conveyed through the project area by utilizing the golf course fairways and greenbelt open spaces as grass-lined channels (Figure 1); then flows are dispersed through dispersal weirs on the east side of the project. Because of the uncertainty of the direction of flows out of Thousand Palms Canyon, the collection basins and the fairway channels are designed to have adequate capacity in any event, whether Thousand Palms Canyon flows approach directly from the northwest or combine with the Edom Hill flood flows from the west. A total of five channels were used (Figure 1, Fairway Channels 1, Ib, 2, 3, and 4). Channel capacities increase from 8,000 to 30,000 cfs as one moves from the northern channel (Fairway 4) to the southern (Fairways 1 and IB). This is due to the slope of the fan that tends to concentrate the flows toward the southern portion of the project. Design of the fairway channels was based on HEC-2 backwater analysis. Sediment transport analysis through the fairway channels was done by utilizing HEC-6. In general, due to sediment deposition within the channels, HEC-6 predicted more conservative water surfaces than HEC-2. Summary and Conclusions The methods applied in the design of this project reflect current knowledge on the flow behavior and sediment transport on alluvial fans. The design con cept-floodproofing the property by collecting, concentrating, and spreading back the flows-proved to be the most economic alternative. However. due to the uncertainty in the flowpath of alluvial floods, the provided total flow capacity through the fairway channel system is approximately 75,000 cfs, compared to the 30,000 cfs of total flow. This conservatism in the design is necessary with the present knowledge of flow behavior on alluvial fans. References Edwards, K., and J. Thielmann 1982 FloodplainManagement, Cabazon, California, ASCE, Las Vegas Nevada, April 26. DMA Consulting Engineers 1985 Alluvial Fan Flooding Methodology Analysis, FEMA, Contract EMW-84-C-1488. July.

PAGE 289

Part Eight Innovative Software Applications

PAGE 290

This page IS intentionally blank

PAGE 291

VISUALIZATION TECHNIQUES FOR FLOODING MODELS G. Braschi, S. Braschi, and L. Natale University of Pavia, Italy Introduction The hydraulic calculations of floods supply the input for socioeconomic methods that evaluate the damage risks. This idea is well expressed in the procedures suggested by U.S. Army Corp of Engineers. At present, floodplain zoning is often based on mapping done with simple methods. These methods are mostly accepted and largely used by agencies for floodplain management. Nevertheless, when the topography of the floodplain and the inundation dynamics preclude the use of those simple methods, mapping is based on results obtained from mathematical models. Recently, two-dimensional models for flood wave propagation have been presented in the scientific literature. Some of these models simulate the complete dynamics of the event and take into account convective inertia effects. These models are still too complex and unwieldy and can be utilized only for local event description. Simpler mathematical models are more suitable for practical floodplain mapping. Such models integrate simplified equations of the hydraulic phenomenon using different numerical integration schemes. Two-dimensional propagation models are always expensive to use, so when they are applied, it is economically convenient to request a more extensive exploitation of the model's features in order to analyze all the aspects that influence the evaluation of the flood risk. Moreover, since mathematical models can simulate the dynamics of the flooding phenomena and so can be utilized to obtain a realistic visualization of the flooding development, the use of the mathematical tool to supply correct information to the public is not to be neglected. The computer simulation can be dynamically displayed with a tri dimensional visualization as a motion picture so that nontechnical people can understand the practical uses of the mapping procedures. The research team of the Department of Hydraulic and Environmental Engi neering of the University of Pavia has worked on these arguments for many years in the framework of the research of national interest promoted by GNDCI ofC.N.R. This broad-based project, which gathers researchers from several universities and Italian research institutes, faces in a global and systematical way the problem of the protection of the national territory from hydrogeological catastrophes. Use of Two-dimensional Flood Models As mentioned in the introduction, some of the two-dimensional mathematical models for flood wave propagation on an initially dry soil are generally applicable for the solution of interesting practical situations. Two of the authors of this article have developed, independently of each other, this kind of mathematical model (Braschi and Gallati, 1989; Natale and Savi, 1991): the results presented here are obtained with the model proposed by Braschi and Gallati.

PAGE 292

268 Visualization Techniques for Flooding Models The mathematical description of the flow is based on the well-known shallow water equations, neglecting the convective inertia term: div Q + ablat = q (1) av/at + g[grad(h+zO)+I] = 0 (2) where: x,y,t are the space and time coordinates; zO em] is the ground level on a fixed horizontal reference; h em] is the water depth; V(u,v) [m/s] is the flow velocity averaged over h; Q=hV [m2/s] is the unit width discharge; q [m/s] is a local discharge source per unit area; g [m/s2] is the acceleration due to gravity; J(jxjy) is the friction slope given by the Manning formula. Neglecting the convective terms simplifies the problem of properly defining the boundary conditions: since the supercritical flow cannot numerically occur, a single condition needs to be given at every boundary point. The differential equations are discretized on a staggered FD grid with implicit scheme: the integration domain is approximated by a system of nonhomogeneous, rectangular, contiguous cells over which the value of water height and bottom level are identified by the centeroid value. The flow unit discharges relative to the cell sides are then nonlinear functions of the water height in the central cell and in the four surrounding cells. Combining the equations for each cell and properly incorporating the boundary conditions, a system of N nonlinear equations in the N cell unknown water depths is written. The solution algorithm, based on a nonlinear Gauss-Seidel method, applies to the whole field, whether it is wetted or not, so that the numerical solution of the problem automatically provides also the position of the wetted front at each time. A relaxation coefficient could eventually be applied to improve the conver gence. The mathematical model was used to simulate the dynamics of hypothetical flood ing phenomena in different places and to recreate the inundation of a rather large area (pian della Selvetta) in Valtellina that occurred during the extraordinary flood of 1987 when a dike on the river Adda broke and poured 14 million cubic meters of water into an area of about 18 square km. The flood, which caused considerable damage and interrupted communications along the Adda valley for several days, lasted about 48 hours. The information collected during the flood (maximum levels reached by the water in the different parts of the flooded area, water levels of the Adda River near the breach, discharges of the Adda River) were sufficient to calibrate the mathematical model rather carefully.

PAGE 293

Braschi, Braschi, and Natale 269 The two-dimensional mathematical model allowed calculation of the time evolution of the hydraulic variables (water levels, flow velocity, unit discharges) in the nodes of the computation grid superimposed on the flooded area. New computation modules were added to the hydraulic mathematical model to calculate other quantities of practical interest: for instance hydrodynamic forces exerted on structures, thickness of sediment aggradation and degradation. A transport model based on the results obtained with the hydraulic model is able to describe the path of the floating bodies transported by the flow and to define the areas at risk from pollution coming from different possible sources. Identification of Flood Effects The inundation of a populated and cultivated area causes several types of economic damage. Indirect damage of importance is caused by hydrodynamic forces against the structures and by the consequent movement of mobile structures due to the flow (ACER, 1988). The flow action is computable by means of the module of total force of the flow over a unit width of wall. Also, the deposit of sediments carried by the flow through the dike breach can considerably damage civil and industrial buildings; sediment of cultivated fields can cause more loss to the crops than temporary submersion by water. The damage due to local erosion caused by the flow is also not negligible. In the evaluation of hydraulic risk of inundation of a floodplain it is useful to compare the different kinds of direct and indirect damage caused by the flood. The set of results obtained with the mathematical model can be organized in a table that compares, by damage classes, the loss extent due to different factors. The comparison table must be based on the results of socioeconomic studies (HEC, 1983). An example of a comparison table has been built in order to draw attention to the visualization techniques used to represent the results of the mathematical model. Figure I gives a description of the vulnerability of the different zones of the floodplain. Along with the visualization of the damage distribution, the map of the different kinds of damage is shown in Figure 2. These maps synthesize all the information obtained with the computer simulation and can be utilized for the following socioeconomic evaluations. The damage maps in Figures 1 and 2 are static pictures of the vulnerability of the floodplain. Sometimes the dynamic outlook of the effects caused by the flood can be useful; for example, when the possible pollution of an area is studied, the identification of the zones affected by the path of pollutants and the evaluation of pollutant concentrations and persistence in the various parts of the flooded area is needed. A dynamic simulation can easily show how the development of the pollution phenomena changes as a function of the position of the source of pollution. In the present study two cases of pollution have been examined. In the first study the pollution source is the river itself, because it is supposed that the polluted water of the river flows through the dike breach. In the second study it is supposed the pollution originated with the flooding of a village sewer system.

PAGE 294

270 classes of damage severity above 4.0 3.0 t.O CJ 2.0 3.0 B 1.0 2.0 below 1.0 Visualization Techniques for Flooding Models Pian della Selvetta Figure 1. Areal distribution of damages. Most severe damage due to: D sediment deposition hydrodlnamlc forces :Yillapilta I. waler depth \ .. Pedemonte ........ -\ \ .... '-.. Pian della Selvefta Figure 2. Sources of damage.

PAGE 295

Braschi, Braschi, and Natale 271 In the first case the dynamic simulation shows that a large part of the flooded area is polluted. The pollution cloud propagates together with the flood wave; at the end a wide zone contains a considerable concentration of pollutants, as shown in Figure 3. In the second case the small transport velocities of the flow that reaches the villages greatly limits the diffusion of the pollutants. Likewise, the dynamic visualization of the transport of floating bodies, such as timbers, can bring out the zones where the passage or deposit of floating debris can cause damage. above 4.5 4.0 -4.5 3.5 -4.0 !! 3.0 3.5 2.5 -3.0 Ej = 1.0 -loS 0.5 -1.0 below 0.5 ViQapnta .. ./ I ... 't.,; Pian della Selvetta Figure 3. Concentration of pollutants. DynamiC Visualization of Inundation The visualization techniques described in the previous paragraph are addressed to engineers or technicians experienced in floodplain mapping; the results are represented on computer display or plotted on technical maps. The illustration for nontechnical people has to be more accessible and has to be spread by means of the media (TV broadcasting, videotape, etc.). For this aim it is very useful to represent results with tridimensional visualization. This kind of representation does not have to be considered an exact description of the results of the computational model if this hampers understanding. In order to present these ideas, we prepared videotapes that show the flooding development, how the water levels change in time, and also how pollution propagates from the dike breach. References Assistant Commissioner-Engineering and Research (ACER) 1988 ACER Technical Memorandum No. 11, u.s. Department of Interior.

PAGE 296

272 Visualization Techniques for Flooding Models Hydrologic Engineering Center (HEC) 1988 Flood Damage Analysis. Davis, California: U.S. Army Corps of Engineers, Water Resources Support Center. Braschi G., and M. Gallati 1989 "Simulation of Levee Breaking Submersion of Planes and Urban Areas" in Hydrocomp '89, Elsevier. Natale L., and F. Savi 1991 "Propagazione di onde di sommersione su terreno inizialmente asciutto," (to appear on Idrotecnica.)

PAGE 297

FLOODPLAIN MANAGEMENT WITH MAC PROJECT II Albert H. Halff Albert Halff Associates Henry M. Halff Halff Resources Halff Associates, a midsize engineering firm, is often engaged to produce plans such as those needed for floodplain management. Producing a floodplain manage ment plan on time and within budget can be exceedingly difficult. This paper describes how a low-end project planning program, MacProject II (1989), can help to determine the costs and schedule of a typical floodplain management project. MacProject II and other similar programs are ideally suited to planning and managing smallto medium-sized projects. We recommend their use for any project with 10 or more tasks that must be accomplished by: two or more individuals or groups, and where two or more tasks can be concurrent. The Turtle Creek Floodplain Management Project Our work on Turtle Creek in Dallas County, Texas, provides a good example of floodplain management project planning. The creek is perhaps the best-known urban park in the Southwest. It flows through three communities, the last being the city of Dallas, and empties into the Elm Fork of the Trinity River. Despite a variety of flood control measures, increasing development has rendered the creek subject to significant flood control problems. Roads and bridges are overtopped during floods, and the 1oo-year rain would flood many houses along the creek. Silt and pollution also constitute problems along the creek. In the spring of 1989, the city of Dallas asked us for a floodplain management study of Turtle Creek. The city's objectives for the effort are to control problems associated with flooding, erosion, silt, and water quality. Needed in this first phase of the project are: a set of alternative solutions, a delineation of the floodplain for each solution, and an analysis to determine the best solution. Planning a Floodplain Management Study Effective project planning on the scale of the Turtle Creek floodplain manage ment effort can be broken down into five steps. 1. Identify the objective(s) of the project. 2. List the tasks and subtasks required to meet the objectives.

PAGE 298

274 MacProject IT 3. Establish task durations and interdependencies. 4. Identify the resources needed for each task. 5. Check and revise the plan to meet known constraints. Project objectives. Project objectives are captured in three questions. 1) What is to be delivered to the client? In the case of Turtle Creek, the city of Dallas requires a report describing alternative solutions for Turtle Creek floodplain management and a recommendation of a single solution. In addi tion, the city called for regular interim status reports and for reviews of intermediate products such as the current floodplain delineation. 2) When is the product needed? Scheduling is a matter of time and money. In the case of Turtle Creek, time is not a severe constraint. We informally established that a period of performance on the order of a year would be adequate for this job. 3) What other constraints apply to the project? As in many projects, particularly for government, cost is a critical constraint. Another is the availability of resources for the work. Additional constraints also apply to this and to other projects. Other commitments and holidays restrict the availability of key personnel. Customer-furnished resources must also be assessed. Tasks and subtasks. Once the main objective has been established, it can be broken down into a hierarchy of main tasks or phases and sub tasks or subprojects. The breakdown is usually accomplished on the basis of experience. In the case of the Turtle Creek floodplain management project, we identified seven main tasks, shown in Figure 1. Many of these tasks can be broken into subtasks. Figure 2, for example lists the tasks in the SUbproject for floodplain delineation. Task durations and interdependencies. MacProject constructs a schedule from two types of data provided by the planner: the duration of each task, and the inter dependencies among tasks. Figure 1 is a display from MacProject II. It shows not only the tasks but also, in standard PERT -chart format, the interdependencies among them. Annotations above each task indicate the task's duration and MacProject II's calculation of its earliest start date. The "shadowed" tasks and the heavy black lines indicate the critical path, the set of tasks whose durations determine the total project duration. Figure 2 shows the schedule for the floodplain delineation SUbproject in the form of a Gantt chart. Evident from this chart are the task durations and slack, the extent to which a task's duration could increase without affecting overall project duration. MacProject II provides displays of both PERT and Gantt charts. Costs and resources. Estimating the costs of a project is a matter of identifying the resources required for each task and then adding up the costs of those resources. MacProject II allows the planner to identify two types of costs: fixed and resource. Fixed costs, such as the cost of maps and plans, are independent of the level of effort required by the task. Resource costs, usually those of labor, are determined

PAGE 299

Halff and Halff 5/1/91 4/3/91 145 ------.23 12/26/91 Evaluationl Recommendation Earliest Start Duration Legend Figure 1. Turtle Creek FPM PERT chart 5/lf)1 10/1/90 1/1/91 4/1/91 7/1/91 10/1/91 Start Flo Delineation Control Surveys I Invento lv and coliect data; Ii Id reconnaisance "" Bridge Surveys :;>'::" P rchase Base Maps I I Model Hy rolegy I r" Reid Veri ation 0 H draulics c= Progress Rep rt with City Staff 0 C mplete Delineation 0 Floodplai Delineation Com pie e f Figure 2. Floodplain delineation Gantt chart 275

PAGE 300

276 MacProject n by level of effort. Planners using MacProject can designate the daily costs of each resource and the assignment of resources to each task. Using these data and fixed cost figures, MacProject can estimate rate of expenditure throughout the project. PIan evaluation and revision. Plans should be evaluated on three major criteria. Schedule-will the project be completed on time? Cost-will the costs meet budgetary constraints? Feasibility-are sufficient resources available to complete the project? MacProject II provides the tools for evaluating project plans on each of these criteria. When plans do not meet one or more criteria, the planner must evaluate alternative plans. Generally, the three criteria trade off among each other. For example, making consecutive tasks parallel to solve a schedule problem may also have the effect of endangering project feasibility. MacProject has some capabilities, such as automatic resource leveling, that can be used to solve planning problems. More importantly, it allows the planner to easily evaluate a large number of alternative plans and solutions to planning problems. Conclusion MacProject II and other low-end planning programs are useful devices for planning typical engineering efforts such as floodplain management planning. These programs work with a planner to provide immediate assessment of scheduling, costs, and resources for a project. Used judiciously, planning programs such as MacProject II can provide the engineer with a realistic picture of his or her project before the proposal is submitted and work begins. References Claris Corporation 1989 MacProject II (Computer Program). Santa Clara, California: Claris Corporation.

PAGE 301

PRACTICAL APPLICATIONS OF THE HEC FLOOD DAMAGE ANALYSIS PACKAGE Troy Lynn Lovell, Walter E. Skipwith, and Michael A. Moya Albert H. Halff Associates Introduction Procedures for flood damage analysis have evolved since the late 1960s from hand computations to the use ofindividual and linked computer programs to perform the hydrologic, hydraulic, and economic calculations. The U.S. Army Corps of Engineers' Flood Damage Analysis (FDA) Package links hydrologic (HEC-l), hydraulic (HEC-2) and economic (SID and EAD) computer programs developed by the Hydrologic Engineering Center (HEC). These individual programs are linked through a data management system, which is called the HEC Data Storage System (HECDSS). This paper includes a review of the basic concepts involved in flood damage analyses, a discussion of the FDA package, and two actual case studies of applications to flooding problems in Texas. Case Study 1 is a small, urbanized watershed (Briar Creek, Bryan, Texas), with residential areas subject to flooding. Case Study 2 includes two proposed levee projects on a large river, one to protect an existing treatment plant and industrial land, and one to protect a large residential, commercial, and industrial area. Flood Damage Analysis Concepts and Procedures The general concept of analysis used for the Briar Creek Case Study is illustrated in Figure 1. The Trinity River Case Study utilized several of the same programs, but did not use a "linked" approach. The basic principle upon which flood damage calculations are based is that the flood damage (in dollars) to an individual structure can be calculated by determining the flood stage (depth of flooding) at the specific location under consideration and by knowing the relationship between flood depth and damage potential of the structure. Another way of expressing flood damages is by means of "average annual damages," which is the frequency-weighted sum of damage for the full range of damaging flood events and can be viewed as what might be expected to occur in any present or future year (annual average over a long period of time). It represents the annual damage for a particular set of hydrologic (rainfall-runoft), hydraulic (depth of flooding), and damage (dollars-depth) conditions. When a calculated average annual damage (AAD) value is desired, then the damage corresponding to each depth of flooding is weighted by the probability of that depth occurring, and these weighted damage values are added, with the sum representing the average flood damages.

PAGE 302


PAGE 303

Lovell, Skipwith, and Moya 279 Basic and Derived Relationships The basic relationships that must be developed to determine either single event (Le., the lOO-year frequency flood) or average annual damages are: Discharge to frequency of recurrence (rainfall-runoft), (HEC-l) Stage (depth of flow) to discharge (flow of water), (HEC-2) Stage to damage (flood damage for each depth of flow), (SID) After these relationships are developed, they can be combined into derived relationships, such as the discharge-damage, stage-frequency, and damage-frequency functions. The damage-frequency relationship is derived by combining certain basic and derived relationships, using the common parameters of stage and discharge. The "frequency weighting" process to derive the average annual damage value consists of computing the total area under the damage-frequency curve. Basic Procedures An economic model (SID) to process information about floodplain structures is used to develop elevation-damage curves and to compute single eventflood damages. SID is encoded with pertinent data such as structure identification, damage reaches and index locations, finished floor elevations, estimated values of structure and contents, stage-damage curves, and the combined hydrologic-hydraulic data from HEC-1 and HEC-2 (elevation vs. frequency). Single event flood damages are computed at the designated locations for each damage reach. The output from SID is stored in a data file for use by the EAD program. The EAD model integrates the frequency versus damage curve, producing average annual damages by land use category and damage reach. This procedure is repeated with revised data to reflect proposed improvements. Case Study 1: Small Creek Application (Briar Creek) In 1989, Albert H. Halff Associates prepared a Stormwater Management Plan on Briar Creek for the city of Bryan, Texas. Detailed hydrologic (HEC-1) and hydraulic (HEC-2) models of the Briar Creek watershed were developed, and floods with a frequency of recurrence of 2, 5, 10, 25, 50, 100, and 500 years were computed. Surveyed finished floor elevations were obtained, and structure values were determined from the Brazos County Appraisal District. Utilizing the program SID, estimated flood damages were computed for each storm frequency, and data files for EAD were created. Estimated Average Annual flood damages were then computed using EAD. Four alternative improvement plans were analyzed to reduce potential flood damages. Each plan was modeled to determine the reduction in water surface elevations and average annual flood damages. The cost of proposed improvement plans was amortized over a 50-year period at an 8.875% interest rate.

PAGE 304

280 Flood Damage Analysis Package With a total of70 structures within the Briar Creek l00-year floodplain, the 100year flood could potentially cause damages of about $2,333,000. Table 1 summarizes the estimated single event damages. Table 1 Estimates of Single Occurrence Flood Losses Frequency Damages 5QO.Yr $2,752 Values are in 1,OOOS 100-Yr 5Q.Yr $2,333 $1,894 25-Yr 10-Yr $1,293 $310 5-Yr $0 Estimates of average annual damages under existing (1990) channel and bridge conditions with future (2000) flood discharges were calculated. The total existing average annual flood losses estimated in the Briar Creek study area are about $130,800, based on 1990 prices. The recommended improvements, with positive benefit-to-cost ratios included: Selective channel clearing and routine maintenance Estimated Cost = $9,200 per year; Damages Reduced = $26,500 (EAD) Benefit-to-cost ratio = 2.9: 1 Enlargement of inadequate culverts and selective channel clearing Estimated Cost = $17,500 per year; Damages Reduced = $64,600 Benefit-to-cost ratio = 3.7:1 Case Study 2: Proposed Trinity River Levee Projects In 1989-90, Halff Associates prepared flood damage and design studies for two major flood control projects along the Trinity River: the Rochester Park area levee and the Central Waste Water Treatment Plant (WWTP) flood protection project. HEC procedures were generally used for this economic analysis (SID and EAD). The economic analysis study area included all properties lying within the 100year and SPF floodplain limits for the Trinity River between South Loop 12 and the existing Dallas Floodway. Finished flood elevations were estimated from the topographic maps and residential areas were grouped by blocks. Major industrial facilities were individually referenced. The water surface profile elevations for 2-, 5-, 10-,25-,50-, 100-, and 500-year flood events, based on existing (1991) channel floodplain and bridge conditions, were used to evaluate flood damages and in determining the relationship of damageable properties to both elevation and frequency of flood occurrence. Existing damageable properties were classified into the major damage categories shown in Table 2. Typically, the value of existing residential contents was estimated to be 40% of the value of the structure.

PAGE 305

Lovell, Skipwith, and Moya Table 2 Property Damages for Existing Conditions 100-Year and Standard Project Floods-Trinity River, Dallas 100-Year SPF Central WV'ITP $72,300,000 73,700,000 Cadillac Heights (Homes/Businesses) 2,500,000 5,700,000 Industries Along MLK 500,000 1,700,000 Rochester Park (Homes/Businesses) 2,100,000 3,900,000 Southeast Service Center 3,900,000 11,500,000 281 The benefit/cost ratio for the Rochester Levee project was calculated as shown in Table 3. The Central WWTP Flood Protection Study is ongoing and benefit/cost analysis is not complete. Table 3 Rochester Park Levee Benefit-Cost Analysis Benefits:Average Annual Damages Reduced = $1,164,000 Costs:Project Total Investment Project Ufe and Discount Rate Project Annual P&I Cost O&M Cost Total Project Annual Cost B/C Ratio: $10,400,000 50 years at 8-7/8% $940,000 $50,000 $990,000 $1,164,000 / $990,000 = 1.2 Conclusions and Recommendations After utilizing the HEC family of hydrologic-hydraulic-economic programs for the case studies, the following conclusions and recommendations are made: 1. The FDA package is relatively easy to use, has been documented by many studies, and is a very flexible tool to use in flood studies. 2. An analyst should utilize graphics and printout for review to assist in visualization of the FDA process, to assess reasonableness of results, and to document study. 3. For small watersheds with a limited number of damage reaches and few alternatives, the linked analysis process may not be efficient.

PAGE 306

282 Flood Damage Analysis Package 4. Relatively large study areas, with numerous damage reaches and a multitude of alternatives, lend themselves to the linked process. 5. Using linked programs and models, without reviewing the output/input files, tends to create a "black box" effect, which can be dangerous (no matter what the project size). References U.s. Army Corps of Engineers, Hydrologic Engineering Center 1988 Flood DamtZge Analysis Package. Davis, California. April. 1989 EAD, Expected Annual Flood DamtZge Computation. Davis, California. March. 1989 SID, Structure Inventory for Damage Analysis. Davis, California. March. 1990 HEC-J Flood Hydrograph Package. Davis, California. Septem ber. 1990 HEC-2 Water Surface Profiles. Davis, California. September. 1990 HECDSS User's Guide and Utility Program Manuals. Davis, California. July. Albert H. Halff Associates, Inc. & Garret Engineering 1990 Stormwater Management Plan, Phase II, Briar Creek and Tributaries, for the City of Bryan, Texas. April.

PAGE 307

THE ALTERNATIVE FUTURES ASSESSMENT PROCESS: BUILDING A CONSENSUS FOR COASTAL MANAGEMENT IN TEXAS Andrew Mangan Texas General Land Office Thomas Bonnicksen Texas A&M University Jim LeGrotte Federal Emergency Management Agency Introduction In 1989, the Texas Legislature passed a bill directing the Texas General Land Office to coordinate development of a long-range comprehensive management plan for the state's coastal public lands. The commissioner of the General Land Office appointed a citizens advisory committee, a state agency task force, and a federal agency task force to help formulate the plan. In the spring of 1990, the agency and these advisory groups held five public meetings along the coast to promote public participation in the planning process. The meetings revealed that three issues were of immediate coastwide concern: shoreline erosion and dune protection, wetland loss, and beach access. These were made the initial focus of the planning effort. It was clear that the coastal management plan would be successful only if it incorporated the concerns of all segments of the coastal community and received broad public endorsement. The General Land Office sought a planning method that would enable such diverse interests as the oil and gas industry, real estate development, commercial fishing, recreation and tourism, conservation, and government to work together compatibly and productively. In the summer of 1990, the Land Office engaged the Office for Strategic Studies in Resource Policy at Texas A&M University to apply its "Alternative Futures Assessment Process" (AFAP) to the coastal planning effort. The AFAP is a computer-assisted workshop process. It is designed to help groups with different perspectives on an issue of common concern to concentrate on areas of agreement and develop an action plan to meet shared objectives. Participants in the process construct a flexible computer model that permits systematic analysis of subjective information. They pool their knowledge, experience, and opinions; develop a common understanding of the issue under consideration; and explore alternative courses of action to solve identified problems. The computer model helps them to compare the expected outcomes of current policies or practices with those of the alternatives they propose. This enables them to reach a consensus on preferred strategies.

PAGE 308

284 Alternative Futures Assessment Process The computer software used in the AFAP, which runs on an mM-compatible personal computer, is a cross-impact simulation language that shows how variables interact over time. Use of the AFAP for Coastal Management Planning Five workshops were held on each of the three target issues. Three Foundation Workshops-one on the lower, one on the middle, and one on the upper coast-were followed by a Strategy Workshop and a Capstone Workshop (Figure 1). The regional Foundation Workshops were intensive information gathering sessions. Each was limited to 28 participants representing a wide range of interests within the region. (A few individuals representing coastwide interests participated in more than one regional workshop.) The product of each workshop was a list of 30 variables to be considered for inclusion in the computer model. CAPSTONE WORKSHOPS STRATEGY WORKSHOPS FOUNDATION WORKSHOPS Recommended policy and research priorities for each issue for the entire Texas Gulf Coast. Interests, concerns, trends, objectives, interactions, and strategies for each issue for the entire Texas Gulf Coast. Interest, concerns, and strategies for each issue for the subregions of the Texas Gulf Coast. Figure 1. The Alternative Futures Assessment Process. The variables defined the participants' interests in and concerns about the issue under discussion and the principal related problems affecting their region of the coast. In the workshops for shoreline erosion and due protection, for example, variables included erosion rates, structures impeding sand transport, vehicle density, public funds appropriated for erosion response, and property loss (Table 1). Each variable was assigned a unit of measure to describe its current status and help predict

PAGE 309

Mangan, Bonnicksen, and LeGrotte 285 change over time. Participants also recommended courses of action to solve identified problems or improve upon current management. In the Strategy Workshops, the computer model was constructed and proposed courses of action were evaluated. The General Land Office divided participants in each Strategy Workshop into 15 stakeholder groups whose members had similar interests and concerns. From the combined list of 90 variables generated in the Foundation Workshops, participants in the Strategy Workshop selected a total of30 key variables to represent the principal concerns about the issue coastwide. Each stakeholder group was allowed to select one peremptory variable that could not be eliminated from the final list. Then the stakeholder groups recorded their objectives for each of the 30 key variables. Objectives were stated in terms of desired degree of change for each variable. Groups could select from eight objectives: No change, Not Up, Not Down, Up %, Down %, Up Max., Down Max., and Don't Care. The computer software converts objectives into a form that can be used to evaluate policies. Next, participants filled out questionnaires and estimated two kinds of long-term trends in the key variables: the maximum increase for each variable, and the expected change (increase or decrease) over a 20-year period under current policies. These were assigned numerical values for the computer model. Trends were classified as desirable or undesirable, and new policies were proposed to reverse or retard undesirable trends. The trend analysis also took into account external forces, such as erosion rate, that cannot be entirely controlled by policies. Participants used a cross-impact matrix (a chart with the variables listed across the top and down the left side, creating 900 cells) to indicate interactions among the 30 key variables; that is, how many other variables affect each variable and in what way. This information was entered into the computer along with the trend estimates. The computer program linked the trends and variable interactions, enabling workshop participants to compare the possible long-term consequences of proposed new coastal management policies with the likely consequences of continuing current policies. The analysis of probable effects of proposed coastal management policies helped participants to refine their recommendations to compensate for undesirable interactions and side effects. Policy recommendations were formulated by teams, then discussed and revised by all participants. The Capstone Workshop for each issue produced final policy recommendations. The recommendations included not only preferred management strategies, but recommendations for future research to fill gaps in knowledge about the issue and proposed sources of funding to implement the strategies.

PAGE 310

286 Alternative Futures Assessment Process Table 1 EZ-Impact Shoreline Erosion/Dune Protection PROJECT: EROSION Variable List Variable Variable Unit of No. Name Description Measure 1 AVAILDAI Available Date Scl DayslYr 2 RESEARCH Gulf Coast Res. Fund $lYr 3 BAY-EROS Bay Shoreline Erosion Ft LostIYr 4 BAY-VEG Bay Shoreline Veg. Acs Cov/Sh M 5 BEA-NOUR Beach Nourishment Cu YdslMilYr 6 BEA-REPS Beach Replenish. Prg. $lYr 7 COASTRUC Coast Cnstruct. Impd. Cu Y ds BlkIY r 8 DRGREUSE Dredged Spoil Reuse Cu YdslYr 9 DUNE-DIM Dune Dimensions % ProtlMilY r 10 DUNE-SYB Dune Stability % Covrd by Veg 11 ECO-INTG Ecological Integrity AcslMi Undistb 12 APPROPR Gen. Fed/State Apprp $/Yr 13 GULFEROS Gulf Shoreline Erosion Ft LostlYr 14 HWYLOSS Highway Losses Days ClosedlYr 15 HUM-EROS Hum. Induced Erosion Ft/Yr 16 IMPCOMM Impact on Commerce $/Yr 17 INLSTRUC Inland S. Imped. Ged Cu Y ds BlkIY r 18 PLANNING Planning # Plans/Yr 19 PUB-EDUC Public Education Hrs/Yr 20 RIV-SAND Riverine Supp. Sand Cu YdslYr 21 SAND-BUD Sand Budget Cu Y ds AvaillYr 22 SETBACKS Set Backs Ft Nn High Tide 23 SHIPTRAF Ship Traffic #/Yr 24 ST-COORD State Inter. Coord. MOUs/Yr 25 SUBSIDEN Subsidence InIYr 26 TOURISMS Tourism $ GeneratedlY r 27 TRASH Trash TonslMilYr 28 VEH-BEAC Vehicle Beach Use # on BeachlYr 29 WETLANDS Wetlands AcslYr 30 HAB-LOST Wildlife Hab. Lost Acs LostlYr

PAGE 311

Mangan, Bonnicksen, and LeGrotte 287 Conclusion The AFAP proved highly successful for development of a management plan to address coastal erosion and dune protection, wetland loss, and beach access on the Texas Gulf Coast. The recommendations generated by the workshop process were compiled in a report, the Texas Coastal Management Plan, published in the fall of 1990. This report, with minor revisions after a second series of public meetings to obtain public comment on the draft plan, served as the basis for legislative proposals submitted to the Texas Legislature in the spring of 1991. Two bills, S.B. 1053 (H.B. 1622) and S.B. 1054 (H.B. 1623), incorporating most of the recommendations made by AFAP workshop participants, were passed in May of 1991 and now await the governor's signature.

PAGE 312

COMPUTER AIDED EVALUATION OF FLOODPLAIN DEVELOPMENT RISK Mark R. Peterson, C. Gary Wolff, and Jay D. Schug Water Engineering and Technology Introduction Accommodation of the risks of development adjacent to alluvial stream channels has posed a difficult problem for design engineers. In years past, the basic approach in many areas was total channel protection and conveyance of flood flows in supercritical concrete-lined channels. This approach enables development to occur right up to the channel's edge, as long as channel capacity is adequate and the channel operates as designed. Although this approach has the advantage of maximizing the amount of developable land, disadvantages include human safety issues (such channels generally must be totally fenced oft), loss of riparian habitat and wildlife values, and conversion of a dynamic natural stream corridor into a single purpose, flood conveyance element. All too often, debris loading and the sediment transport characteristics of the natural stream system have been overlooked or ignored. Consequently, flow bulking prevents the improved channel system from providing the required level of flood protection or leads to failure altogether. In recent years, the multiple values of urban stream corridors have been recognized. These include water quality enhancement, aesthetics, wildlife habitat, riparian vegetation, wetlands, recreation, and flood control values. Efforts are now being made to manage these riparian corridors as a community resource. Although progress and changes have been made, one of the major items that still must be addressed in floodplain management is the determination of an acccptable level of risk for development adjacent to the unmodified stream channel. The FEMA floodplainlfloodway delineation and mapping efforts have provided good information regarding potential impacts of development on flooding risks. However, such analyses assume a fixed channel boundary condition. A fixed-boundary approach does not necessarily provide an assessment of the risks associated with development adjacent to natural channel systems that may experience drastic changes in both profile and/or planform during a single flood event. Arroyos that traverse urban developments in the arid to semiarid southwestern United States exemplify this type of flood-driven behavior and associated risk. Floodplain Development Risk The logical first step either in identifying the need for improvements or in developing a flood control plan is to acquire an understanding of system behavior both temporally and spatially. An understanding of past and present geomorphic behavior of the system provides a basis for predicting future behavior (Schumm et al., 1984). Information requirements include morphometric variables (width, depth, slope, bed materials, etc.); bank stability characteristics (bank height, angle, and material properties); sediment transport characteristics; and past, present, and

PAGE 313

Peterson. Wolff, and Schug 289 expected future land use. Lagasse et al. (1985) developed the prudent line approach to risk determination for development adjacent to arroyos for the Albuquerque Metropolitan Arroyo Flood Control Authority (AMAFCA). More recently, Water Engineering and Technology (WET) developed a geomorphic/engineering-based, computer-aided expert system (CADWET) for assessing watershed and channel stabilization requirements. CADWET originally was developed from an extensive knowledge and data base that had been amassed in the incised channel watersheds of the southeastern United States. The basic processes controlling channel stability, however, are the same for any incised channel or arroyo system (Schumm et al., 1984). Model Description CADWET integrates consideration of readily identifiable channel characteristics with numerical techniques for assessing both vertical (bed aggradation/degradation) and lateral (bank erosion) stability. The geomorphic evolution of an incised channel from a condition of disequilibrium to a new state of dynamic equilibrium can be described by a five-phase geomorphic model of channel evolution in which degradation leads to channel widening as a result of bank failure once a critical bank height threshold has been exceeded (Harvey and Watson, 1986). This geomorphic model, which integrates hydraulic and geotechnical parameters with the morphologic characteristics of the channel, provides the process basis for the CADWET model. Planform change due to meander translation and/or cutoff must be evaluated separately. The CADWET system consists of an integrated set of computer programs. The channel stability assessment was developed using the Production Rule Language (PRL) and the LEVELSIPC expert system development software. Data input programs were coded in Fortran and combined with spreadsheet models developed using Lotus 123 and compiled using the Baler spreadsheet compiler. The CADWET system was given a consistent user interface using HiScreen XL software. CADWET consists of six major modules shown schematically in Figure 1. In the Front End module, the user identifies watershed problems that may include aggradation, degradation, bank erosion, flooding, infrastructure damage, and environmental problems. Depending on the types of problems identified and the desired level of information, the CADWET system provides the user with information on how and where to obtain watershed data, what type of data to obtain, how to conduct field investigations, general descriptions of what CADWET does and how it does it, and several other help functions. The control module allows the user to bypass several of the CADWET modules. It therefore facilitates execution of data entry routines, allows direct execution of the stability assessment knowledge bases, or allows direct access to the preliminary design spreadsheet. The data input module facilitates data input to the CADWET system. The data processed in this module consist primarily of geomorphic and hydraulic information. Morphometric properties of the channel, such as bank heights and angles and bed

PAGE 314

290 Computer Aided Risk Evaluation Figure 1. Overall flow chart of CADWET. slopes, can be either entered manually or computed using data obtained directly from an existing HEC-2 data file. When appropriate, sub-reaches can be delineated using a data entry spreadsheet. Once channel reaches have been identified, channel stability is evaluated on a reach-by-reach basis within the expert system stability assessment module of the CADWET system. Evaluation criteria are internal to the program and are based on user responses to interrogation by the CADWET system. Factors used in deter mination of channel stability include: system characteristics determined from comparative aerial photography, specific gage analysis, comparative channel profile information, presence or absence of bars, depths of moveable bed sediments, presence or absence of eroding banks, and presence of head cuts. Evaluation of bank stability can be the result of site-specific geotechnical analyses, application of a regional curve delineating threshold conditions for bank stability (Thorne et al., 1981), field observations of bank failure within a reach, or the presence or absence of mature vegetation along the channel banks. For each attribute, the user is required to assign a confidence factor ranging from 0 to 100, which expresses a level of certainty in the fact or conclusion. The CADWET system weights the user assigned probabilities in combination with internal weighting factors that describe the relative importance of these attributes in reaching a conclusion regarding channel bed and bank stability. In each case, minimum thresholds are identified and incorporated into the expert system. These thresholds must be reached before a conclusion regarding channel stability is

PAGE 315

Peterson, Wolff, and Schug 291 achieved. Insufficient information or confidence levels will prevent generation of a conclusion. Similarly, checks are made for conflicting information. Flexibility has been incorporated into the CADWET system so that responses to questions can be reviewed and, if appropriate, changed. Once system stability has been evaluated, an assessment of the need for bed stabilization can be conducted using the preliminary design spreadsheet. This module allows the user to explore several channel stabilization options. Current and potential future bed stability conditions that reflect grade control emplaced at various locations can be evaluated and graphically compared by using a user-defined relationship for equilibrium slope. Bank stability for both existing and future channel conditions can be evaluated using the Culmann method or the computational procedures described by Thorne et al. (1981). Comparisons can be made between existing and future stream conditions for various alternatives. Both hydraulic and geotechnical stability are computed using a dimensionless stability number procedure described by Watson et al. (1988). Capabilities With CADWET, the engineer can evaluate the existing conditions in the stream system and identify potential mitigating measures in each reach. Used in combination with engineering analysis tools such as HEC-2, the CADWET system can provide valuable assistance for developing flood control and channel stabilization require ments that adequately account for future trends in system behavior. Conclusion The CADWET system provides a working tool to assist engineers and planners in both problem definition and preliminary design of mitigation measures in watersheds experiencing problems related to flooding, erosion, and sedimentation. Although originally developed for watersheds in the southeast, the CADWET system approach is applicable to any watershed system. When applied correctly, the CADWET system can help identify risks to development adjacent to the channel system and provide assistance in determining mitigation measures. In particular, the linkage between the process-based geomorphic understanding of system behavior and the engineering analysis methods for bed and bank stability that are incorporated into CADWET provides a powerful tool for assessing the risks associated with development adjacent to arroyos in the southwest. References Lagasse, Peter F., James D. Schall, and Mark R. Peterson 1985 "Erosion Risk Analysis for A Southwestern Arroyo." Journal of Urban Planning and Development 3, No.1 (November). ASCE, Paper No. 20165

PAGE 316

292 Computer Aided Risk Evaluation Thome, C.R., J.B. Murphey, and W.C. Little 1981 Stream Channel Stability. Appendix D, USDA Sedimentation Laboratory, Oxford, Mississippi. Prepared for U.S. Army Corps of Engineers, Vicksburg District, Vicksburg, Mississippi. Watson, C.C., M.R. Peterson, M.D. Harvey, D.S. Biedenham, and P. Combs 1988 "Geotechnical and Hydraulic Stability Numbers for Channel Rehabilitation: Part IT, Application." Hydraulic Engineering, Proceedings of the 1988 National Conference, American Society of Civil Engineering, Colorado Springs, Colorado. Schumm, S.A., M.D. Harvey, and C.C. Watson 1984 Incised Channels: Morphology, Dynamics and Control. Littleton, Colorado: Water Resources Publications. Harvey, M.D., and C.C. Watson 1986 "Fluvial Processes and Morphologic Thresholds in Stream Channel Restoration." Water Resources Bulletin 22, No.3: 359368.

PAGE 317

Part Nine Multi-Objective Planning

PAGE 318

This page IS intentionally blank

PAGE 319

REDUCED FLOOD LOSSES IN OREGON AND OREGON'S STATEWIDE PLANNING PROGRAM George M. Currin Federal Emergency Management Agency During the last 20 years, Oregon has had three presidentially declared flood disasters, while Washington, Oregon's neighbor to the north, has had 18 presi dentially declared flood disasters. One of the reasons for Oregon having fewer flood disasters may be the direct result of their Statewide Planning Program. While Oregon communities have implemented this statewide program for conservation and development purposes, they have also realized a secondary benefit-flood loss reduction. An important part of Oregon's planning program is its emphasis on coordinated planning. Oregon has 19 statewide planning goals administered by the Oregon Land Conservation and Development Commission. Communities in Oregon must comply with these goals when adopting comprehensive plans and zoning. Each county and the cities within it are required to adopt plans that are consistent with each other and the statewide goals. The programs of state agencies also are required to be consistent with statewide planning goals and with acknowledged local comprehensive plans. The statewide planning program has many aims and objectives, the most important of which are expressed in 19 statewide planning goals. Oregon's Statewide Planning Goals are quite detailed, mandatory, and have the force oflaw. Generally, the program is intended to: conserve farm and forest land, coastal resources, and other important natural resources; encourage efficient development; coordinate the planning activities oflocal governments and state and federal agencies; enhance the state's economy; and reduce the public costs that result from poorly planned development. Although each of the goals addresses a different topic, they can be grouped into four broad categories. The first set comprises those that deal with the planning process (citizen involvement, land use planning). A second group, the conservation goals, deals with topics such as farm lands, forest lands, and natural resources. The third group is made up of goals that relate to development (housing, transportation, and public facilities and services). The fourth category contains the goals that deal with Oregon's coastal resources. Oregon's floodplain management program owes much of its success to statewide planning. A statewide program offers the opportunity of mitigating flood hazards through comprehensive land use planning. Although Goal 7 is responsible for protecting life and property from natural disasters and hazards, this paper will look at five additional conservation and development goals that complement this effort. They are: Goal3-Agricultural Lands, Goal4-Forest Lands, GoalS-Open Space,

PAGE 320

296 Reduced Flood Losses in Oregon Goal 14-Urbanization, and Goal lS-Willamette River Greenway. These goals overlie many of the same areas that are subject to flood hazard and the requirements of Goal 7. Statewide Goals Goal3-Agricultural Lands. This goal requires that agricultural lands be preserved and maintained for farm use and be inventoried and preserved by adopting exclusive farm-use zoning. Agricultural lands in western Oregon are those with U.S. Soil and Conservation Service soil classification types I, II, m, and IV. This is most significant because most of the Special Flood Hazard Areas (SFHA) have class I, II, and m soils. Goal 4-Forest Lands. This goal requires that forest lands be retained for the production of wood fibre and other forest uses. Lands suitable for forest uses are inventoried and designated as forest lands according to forest site classes of the U.S. Forest Service. Many of these lands are also class I through IV soils that are preserved by Goal 3 and are also located in the SFHA. Goal S-Open Space. This goal provides for programs to conserve open space and protect natural and scenic resources, such as important habitats for plant, animal, or marine life, wetlands, public and private golf courses, neighborhood parks, wildlife preserves, and nature reservations. Goa114-Urbanization. Provides for the orderly and efficient transition from rural to urban land use through urban growth boundaries established to identify and separate urbanizable land from rural land. Goa115-Willamette River Greenway. A legislative policy directing development and maintenance of a natural, scenic, historical, and recreational greenway along the Willamette River. Implementation The above five goals are implemented by a community's comprehensive plan and zoning ordinance, which must comply with the statewide planning goals and be approved by the state. Urban growth boundaries. Every city is required to establish an urban growth boundary (UGB). The primary concern is that the growth and management of urban areas be accomplished in a rational, coordinated manner, where both city and county jurisdictions agree on planning proposals, implementing ordinances, such as zoning, and processing to enhance information flow between the two government levels. This requires that cities and counties adopt consistent plans and implementing ordinances affecting lands outside of city limits but still under county jurisdiction and that formal procedures be adopted between the jurisdictions that permit joint management of those lands and allow one jurisdiction to respond effectively to the planning proposals of the other.

PAGE 321

Currin 297 Exclusive jann use (EFU) zoning implements Goal 3 on Class I, II, III, and IV soils through farm parcel sizes that continue the commercial farm use in the area. These parcel sizes vary from 20 acres for horticultural specialties to 320 acres for primary crop lands. A residence is permitted in conjunction with a commercial farm use; nonfarm dwellings are prohibited. Forest zoning for both nonimpacted and impacted (FI and F2) forest lands implements Goal 4, after inventory and identification of those soils that have forest capabilities, as defined by forest site classes of the u.s. Forest Service. FI forest lands have a minimum parcel size of 80 acres and prohibit any residences. F2 forest land parcels vary in size from a minimum of 20 acres for a woodlot to 80 acres if adjacent to FI zoned lands. A residence may be allowed conditionally in conjunction with the propagation or harvesting of a forest product. Park and recreation (PR) zoning is one method of implementing Goal 5 and is applied to parks and golf courses to preserve their uses and limit their conversion to non-open space uses. Another and just as significant method of implementing Goal 5 is through the property development standards that are found in each zone. For example, in Class I Stream Setbacks, a residence shall not be located closer than 100 feet from the ordinary high water of a Class I stream in any resources zone (EFU, FI, F2) nor closer than 50 feet in any other zones, such as residential zones. Also, a maximum of 25 % of existing natural vegetation may be removed from the setback area. This in essence establishes a 100to 200-foot corridor for the passage of water on streams without a designated floodway or, in many cases, establishes a corridor much wider than some designated floodways. Willamette River Greenway. A Greenway Development Permit is required for new intensifications, change in use or developments-including public improvements and subdivisions-which are proposed within the boundaries of the Willamette River Greenway. The approval of a Greenway Development Permit requires findings of conformance with specific criteria and setback requirements. Some of those criteria include the preservation of areas of annual flooding, floodplains, and wetlands, and conservation or preservation of areas along the alluvial bottom lands and lands with severe soil limitations from intensive development. Additional criteria include preserving and maintaining land inventoried as agricultural and minimizing interference with the long-term capacity of lands for farm use. New intensification, developments, and changes of use shall be set back 100 feet from the ordinary high water line of the river. Conclusion Oregon has spent the past two decades on statewide land-use legislation, adopted mandatory Statewide Planning Goals, and sought acknowledgment of local comprehensive plans and zoning to comply with those statewide goals. This has resulted in flood loss reduction through the mitigating effect of preserving a vast amount of the Special Flood Hazard Areas for farm, forest, and open space uses.

PAGE 322

298 Reduced Flood Losses in Oregon The traditional method is to take a piecemeal approach, layering environmental safeguards on top of one other and requiring local officials and citizens to react to proposals, with the result being a complicated, time consuming, and costly review process. The state of Oregon took a comprehensive approach, and spent that time up front to identify those areas that could be developed and the constraints that would apply to the development.

PAGE 323

COMPREHENSIVE FLOODPLAIN MANAGEMENT: THE DENVER AREA EXPERIENCE Bill DeGroot Urban Drainage and Flood Control District Introduction While annual flood losses nationwide continue to rise, the flood damage potential in the Denver metropolitan area, up through the loo-year event, has actually decreased. The decrease has been the result of a cooperative effort between local governments and the Urban Drainage and Flood Control District in the implemen tation of a two-pronged approach: 1) fix past mistakes of development in floodplains while, 2) preventing new mistakes from being made. All of this occurred during the time when the population of the metro area increased by about 500,000 people. The Urban Drainage and Flood Control District was established by the Colorado Legislature in 1969 to assist local governments in the metro area with multi jurisdictional drainage and flood control problems. The district covers an area of 1,608 square miles, including Denver, parts of the five surrounding counties, and 29 cities and towns. The present popUlation of the district is about 1.8 million people. The district is an independent agency governed by an appointed 17 -member board of directors. Fifteen members are locally elected officials (mayors, county commissioners, city council members) and two are registered professional engineers. The district maintains a small staff, utilizing private consultants and contractors as much as possible. As a result, the district operates a $10 million annual program with only 17 full-time employees and five part-time college student interns. The staff is responsible for management of all project funds; supervision of all work done by consulting engineers; and coordination of all planning, design, construction, and floodplain management efforts with local governments. Responsible Growth The legislation that established the district gave it fairly broad powers but very little money to implement those powers. Initially, the district was authorized to levy 0.1 mil for planning and operations, which amounted to approximately $400,000. The first major activity of the district was to inventory drainage basins and sub basins to determine the extent of problems and to develop a plan to attack those problems. The initial study indicated that approximately 25 % of the major drainageway miles within the district were developed, with the remaining 75% undeveloped and amenable to preventive approaches. It was logical to consider that, if effective preventive measures could be undertaken on the undeveloped drainage ways, significant savings in future remedial needs could be realized. The district board therefore made a commitment to develop a comprehensive floodplain management program to prevent new problems from being created by new development.

PAGE 324

300 Comprehensive Management in Denver The board also realized that the South Platte River, the backbone of the drainage system for the entire metro area, was so large and had so many problems that it could absorb all of the district's time, effort, and money. Therefore the board decided to emphasize work on tributaries to the river. In 1973, following four years of problem identification and planning, the board requested authority to levy an additional 0.4 mil for a design and construction program. The legislature granted the request, beginning in 1974. Also in 1974, the board established the floodplain management program, to be funded out of the original 0.1 mil. In 1979, the board requested a 0.4 mil increase for maintenance and preservation of floodplains and floodways. The legislature approved the request beginning in 1980. Finally, in 1985, the district turned to the South Platte River. A master plan study for the river was completed in late 1985. Using the master plan, the board sought and received an additional 0.1 mil authorization (excluding Boulder County) for funds to be used for the South Platte River, and that request was approved in 1986. The district now had a comprehensive program addressing all aspects of flood management, a set of tried and proven policies and procedures, and a reasonable and reliable level of funding. Details of the individual district programs are provided in greater detail in the following sections. District Programs The Master Planning Program is funded out of the original 0.1 mil authorization. Key policy decisions that guide the program are: 1) each master planning effort must be requested by the local governments and must be multijurisdictional; 2) master plans are completed by consultants acceptable to all local sponsors and the district; 3) the district pays 50% of the study costs, with the local sponsors sharing the other 50%; and 4) the master plan must be acceptable to all the affected local govern ments. The program has evolved into four major areas of interest: I) major drainageway master planning; 2) outfall systems planning; 3) drainage criteria manuals; and 4) special projects, such as criteria for channels and structures on sandy soils, wetland issues, and gravel mining guidelines. Most recently the district has become involved in the coordination of NPDES permit applications from metro area local govern ments. The master plans provide the basis for input to the district's Five Year Capital Improvement Program. Forty-one major drainageway and 19 outfall system master plans have been completed, and several more are in progress. The Floodplain Management Program was established in 1974 to prevent new flood damage potential from being introduced into l00-year floodplains, while encouraging the utilization of nonstructural methods of flood damage mitigation. The major activities of the program are: 1) the National Flood Insurance Program (NFIP), 2) floodplain regulation, 3) flood hazard area delineation, 4) flood warning,

PAGE 325

DeGroot 301 5) flood damage surveys, 6) development reviews, 7) maintenance eligibility, and 8) public information. The district works with local governments to assure that they remain in the NFIP and keep flood insurance available for their citizens, and also with the Federal Emergency Management Agency to assure consistency between district studies and Flood Insurance Rate Maps. The district has the authority to regulate floodplains, but has chosen not to do so as long as the local governments implement their own regulations. The district assists the local governments with their floodplain regulations, including the requirements of the NFIP. The district continues to identify and publish 100-year floodplains through its flood hazard area delineation program. The district assists local governments in the development of flood warning plans and the installation of flood detection networks. In addition, the district hires a private meteorological service to provide daily forecasts of flood-producing events, which are made available to all local governments. The district has a program to notify occupants of floodplains of the flood potential they face through the annual mailing of over 26,000 informational brochures to each address in or adjacent to each district-identified floodplain. The district reviews and provides comments on proposed developments in or near floodplains at the request oflocal governments. Drainage and flood control facilities constructed by, or approved for, construction by local governments must be approved by the district in order for those facilities to be eligible for assistance from the district's maintenance program. The Design and Construction Program is responsible for the implementation of master-planned projects. The board has established these key policies: 1) proposed improvements must be requested by local governments, 2) proposed improvements must have been master-planned, 3) district funds must be matched by local governments, 4) local governments must agree to own completed facilities and must accept primary responsibility for their maintenance, 5) tax revenue received from each county will be spent for improvements benefiting that county over a period from 1974 to five years into the future, and 6) the district will not develop a public works department, but will rely on local government public works departments. Generally, the district coordinates final designs prepared by consulting engineers. The local governments are involved in all aspects of the design process; they generally acquire project rights-of-way and serve as the construction contracting agency. Each year the board adopts a Five Year Capital Improvement Program which lists projects and district participation by county from 1974 to five years into the future. This plan forms the basis for district participation in design and construction projects. The program has been involved in over $90 million of construction projects, including $39 million in district funds. Key policy decisions for the maintenance program include: 1) maintenance of facilities funded by the district shall be the primary responsibility of the local governments; 2) to the extent the funds are available, the district will assist local governments with maintenance and preservation of floodplains and floodways; 3) the

PAGE 326

302 Comprehensive Management in Denver order of priority for the expenditure of district maintenance funds is district-owned facilities, district-funded projects, projects funded by others, unimproved urban drainageways, and unimproved rural drainageways; 4) funds derived from the maintenance mil levy are returned to each county annually in the same proportion as they are received; 5) local matching funds are not required; and 6) the district will not create a public works department. The annual maintenance work program is developed for each county based on the funds available for that county and maintenance requests from the local governments in that county. The work is divided into three types of activities: routine, restoration, and rehabilitation. Routine maintenance consists of mowing, trash and debris cleanup, weed control, and revegetation efforts. Restoration projects include detention pond mucking, trash rack cleaning, tree thinning, repairing local erosion problems, and local channel grading and shaping. Rehabilitation projects are major construction efforts intended to reclaim and rejuvenate existing facilities that have been neglected until serious problems have developed. Examples include rebuilding or replacing drop structures, building low-flow channels, establishing maintenance access, and providing protection for existing structures. All maintenance activities are done by private contractors. The South Platte River Program was established to provide special attention to the river. The district shares in the cost of capital improvement projects with a minimum contribution of 25 % from the participating local government. Maintenance is also a primary activity. The district may contribute up to 100% of the cost of maintenance activities. Other efforts include cooperative projects with property owners to stabilize river banks, acquisition of rights-of-way, detailed inventories of facilities and properties along the river, and periodic surveys of the river to track and assess horizontal and vertical movement of the riverbed. Summary and Conclusions The reduced flood damage potential referred to earlier is the result of: 1) the cooperative planning, design, construction, and maintenance of more than $90 million of remedial drainage and flood control projects; and 2) the cooperative utilization of floodplain regulations and other land-use policies that have prevented essentially any new development in defined lOO-year floodplains. When comparing the total reduction in flood damages resulting from remedial construction with the small number of new structures in the floodplain, and considering that 500,000 new residents and their attendant homes and job locations were developed in this period of analysis, it is clear that, in the Denver metropolitan area, flood damage potential is being and has been significantly reduced. That does not mean that the job is over, but it does demonstrate that the area is on the right track.

PAGE 327

MANAGEMENT OF NATURAL AND BENEFICIAL FLOODPLAIN VALVES: THE COMMONWEALTH OF VIRGINIA'S STRATEGY Douglas J. Plasencia Virginia Division of Soil and Water Conservation Integrating the management of natural and beneficial floodplain values into existing floodplain management programs can be difficult and confusing. However, the inclusion of natural and beneficial values in floodplain management is a logical step in the progression of flood protection policies for many states. With renewed and hopefully sustained interest in the environment, the management of floodplain natural values is receiving a great deal of attention. Multiobjective management of river corridors is gaining national attention. The promulgation of environmental regulations and policies quite often impacts land uses in the floodplain environment. With all of this interest in natural and beneficial floodplain values, what is the role of floodplain managers in their management? To answer that question requires an understanding of who floodplain managers are and the programs they represent. Floodplain managers come from a variety of backgrounds. They work in an environment that requires an understanding or appreciation of hydrology, hydraulics, land-use planning and zoning, legal issues, insurance, governmental programs, engineering design of structures and controls, and now apparently natural resource management. The one common thread between all these specialties, however, is a goal to seek ways to reduce or eliminate existing and future flood damages. The programs that floodplain managers represent generally are authorized for this same goal, the goal of flood protection. With the recent inclusion of natural resource issues and the management of natural and beneficial floodplain values, there has been a tendency for some floodplain managers to lose sight of the goal of flood-loss reduction. Other floodplain managers have rejected the notion of managing natural and beneficial values because they feel these types of issues interfere with their program goals. And, finally, others may be restrained by their agencies from incorporating the management of natural and beneficial floodplain values because of lack of program authority. For example, the development of floodplain open space for recreation is not a normal program goal of floodplain management. The protection of rare species indigenous to a floodplain is not a normal program goal for floodplain management. The acquisition of historical properties in the floodplain is not a normal program goal for floodplain management. The Virginia Bureau of Flood Protection struggled with the same issues, and developed a perspective and strategies that will support the management of natural and beneficial floodplain values within the confines of a "traditional" floodplain management program. These issues were discussed in the recently published document The Floodplain Management Plan for the Commonwealth of Virginia. A chapter and several appendices were devoted to the management of natural and beneficial values in the floodplain. The plan attempts to recognize the value of floodplain management activities that lead to improved water quality, better habitat for wildlife and fisheries,

PAGE 328

304 Natural and Beneficial Floodplain Values enhanced recreational opportunities, and other related benefits. However, it is also evident in the plan that the promotion of these benefits is important to a floodplain management program for one reason: the benefit of flood protection. Virginia's program, like most other floodplain programs, has been established to mitigate flood losses. While the program staff may have individual views that embrace the management of natural and beneficial floodplain values, the program's support for the management of these values still rests on the potential for flood protection. The primary benefit is the creation of floodplain open space, or floodplain uses other than human occupants and businesses. Any effort that leads to the advancement of a natural and beneficial floodplain value normally translates into floodplain open space. Floodplain open space is important to Virginia's floodplain management program for several reasons. The first and most obvious is that if new structures are located out of flood hazard areas, their chances for flood or stormwater damage are reduced or eliminated. The second reason is that floodplain open space allows a river to continue to store excess waters in the overbanks and attenuate flood flows to control damages to downstream inhabitants. However, it must be cautioned that promotion of floodplain open space may not be in the overall interest of every community. Certain communities have water dependent industries. Others, as typified in western Virginia, are in mountainous regions where the only land suitable for development occurs in the floodplain, and at times open space policies may actually lead to the demise of some other unique feature or value in the area. But in general most communities have ample room for development outside of the floodplain environment. Implementation of this type of strategy, however, is more of an effort in searching out and creating opportunities rather than establishing a fixed program. In Virginia there are notable programs in place, public and private, that are attempting to acquire, by easement or fee title, land rights for suitable natural lands. These include the cooperative efforts of the Nature Conservancy and the Virginia Department of Conservation and Recreation, where critical sites are purchased using private resources and then donated to the state for management of rare or threatened ecosystems. Also, the Virginia Outdoor Foundation established by the Virginia General Assembly is authorized to acquire gifts ofland, money, and other resources to protect critical land resources. One effort the floodplain management program could pursue would be to promote the acquisition of critical flood sites within the context of these programs. Another effort will be in the education of local officials. Local policies directly impact the manner in which a floodplain is developed. Several Virginia localities have very restrictive floodplain land-use standards, but most adopt the minimum floodplain management standards of the National Flood Insurance Program. The Community Rating System may offer incentives for additional open space for these communities, but these may be more readily facilitated through the development of comprehensive plans for the floodplain. The plans allow the community to develop an appropriate blend of land uses while preserving certain natural and beneficial

PAGE 329

Plasencia 305 floodplain values. A secondary issue is a review of the community's development policies within the floodplain. A message that local officials need to hear more often is that not only can floodplain development lead to the demise of certain natural and beneficial values, but that floodplain development oftentimes leads to increased local expenditures as compared to nonfloodplain sites. Placement of waterlines, sewerlines, and roads in the floodplain often are accompanied by increased maintenance costs because of the damages induced by high water tables and flooding. If evacuation of floodplain residents occurs during times of flooding, increased expenditure for housing, meals, transport, and police and fire protection are normal to those efforts. Public employees are placed at increased levels of danger when having to deal with evacuation and protection of areas that are flooded. A message in Virginia's floodplain management plan is that communities should seriously review their public assistance policies for new development in the floodplain. Another strategy is supporting efforts that lead to the development of multi objective projects in the floodplain. These efforts, while noted for being arduous, can lead to a very effective blend of talents and resources. This blend of talent may be able to develop a feasible project and management plan that may not have been attainable as a single purpose project. Again, many times these projects develop management strategies for recreational corridors, fisheries, wildlife habitat, and other suitable uses that promote low density or open space in the floodplain. Currently the Virginia Bureau of Flood Protection is attempting to develop working relationships and to establish quarterly meetings with state agencies more involved with the management of natural resource values in the floodplain. It is hoped that these initial meetings will provide the setting that will promote multiobjective efforts in Virginia. Regulations that promote beneficial values also can be prominent avenues that promote flood protection. A recent act promulgated by the Virginia General Assembly is the Chesapeake Bay Act. This law is exclusively a water quality act focused at decreasing nonpoint pollutants into the bay and covers a significant portion of the state. Within the act are standards for setback from water bodies or mitigation of impacts for structures in certain water courses. In essence these standards support floodplain management objectives by providing additional open space along the water body. Supporting the adoption and enforcement of these types of regulations can be quite beneficial to flood protection efforts. While the above listed efforts are not inclusive of all strategies that will be attempted in Virginia, they should provide evidence that the management of natural and beneficial floodplain values is good for flood protection. Floodplain managers may wish to incorporate the management of natural and beneficial values into their programs, but they also must be careful to not lose sight of their overall program goals and authorities.

PAGE 330

THE MISSOURI RIVER CORRIDOR PROJECT, OR THE BEGINNINGS OF A RIVER RENAISSANCE John H. Sowl National Park Service Midwest Regional Office Introduction The valley of the Missouri River between the states of Nebraska and Iowa (or, more specifically, the Middle Missouri River Sub-basin) represents an area that has experienced drastic and significant changes to the landscape during its human occupation. From the presence of the first Native Americans to the relatively recent influx of Europeans and other ethnic groups, the region has developed a diverse cultural heritage and history. It has also experienced a multitude of human designs on the land, some having dramatic negative effects on the natural environment, such as the channelization of the Missouri River. The resulting losses to that ecosystem's once rich biological diversity include more than 50% of the river's biomass, more than 80% of its riverine wetlands, and approximately 100% of its sandbar habitat (Hallberg et al., 1979). After being "straightened and tamed," cleared of much of its rich riverine ecosystem, fouled by insensitive development along its length, and, in many ways, ignored and left for dead by floodplain managers, the Missouri River-the longest river in North America-is now on the threshold of a renaissance. This rebirth is being facilitated by the Missouri River Corridor Project in an attempt to rehabilitate the river into a renewed and vibrant resource for nature and humankind. Purpose of the Project The Missouri River Corridor Project's primary goal is to renovate, where feasible, the decreasingly viable oxbow lakes and backwater areas along the river. These oxbow lakes and the corridor in which they reside along the river represent one of the last bastions for the native plants and animals that inhabit what remains of the once extensive floodplain ecosystem in this region (Hallberg et al., 1979). Since the river was channelized by the U.S. Army Corps of Engineers to promote navigation and to provide flood protection, there has been a steady loss of flora, fauna, wetland, riparian, and forest habitat (Funk and Robinson, 1974; Fredrickson, 1979; Hallberg et aI., 1979). This is primarily due to such factors as accelerated riverbed degradation rates and detrimental land-use patterns resulting from the channelization process (Bragg and Tatschl, 1977; SIMPCO, 1978), thus making it impossible for new oxbows to form to replace the old ones (Hallberg et al., 1979). The project's secondary goal is to identify historic structures and districts; sites associated with Native American culture and history; sites related to early Euro American exploration and settlement, such as the Lewis and Clark Trail; and other sites of contemporary cultural interest.

PAGE 331

Sowl 307 The project's third and final goal is to identify and assess public recreation sites within the corridor to facilitate increased river, oxbow, and backwater access for canoeing, hunting, and fishing; areas for hiking, camping, nature study, and bicycling; and points of interest for sightseeing, to name only a few examples of possible activities. By renovating backwater areas and identifying cultural and recreational sites within the corridor area, it is hoped that the project will also help provide increased tourism and new avenues of economic development for the communities and inhabitants within the corridor boundaries. Project Authorization and Regional Scope As of 1991, the Missouri River Corridor Project involves a corridor through portions of ten counties in eastern Nebraska and extends from South Sioux City in the north to the Nebraska-Kansas border (Figure 1). More than one-third of the state's population resides in or near this 387-kilometer (242-mile) section of the Missouri in Nebraska. The project, which is based on A Management Plan for Oxbow Lakes on the Middle Missouri River, by Sowl (1986), began in earnest in 1987, and was conducted in two phases. Phase I addressed the 219-kilometer (137mile) portion of the river from South Sioux City to the mouth of the Platte River. This inventory and analysis was completed by the U.S. Army Corps of Engineers in 1989. The Phase II portion extends 168 kilometers (105 miles) from the Platte River to the Nebraska-Kansas border and was completed by the National Park Service through its Rivers, Trails, and Conservation Assistance Program in 1990. Local sponsors of this effort in Nebraska were three Natural Resources Districts (NRDs), which are tax-levying subdivisions of state government whose respon sibilities include soil and water conservation, wildlife enhancement, and public recre ation. They are the Papio-Missouri River NRD, the Lower Platte South NRD, and the Nemaha NRD. These NRDs will ultimately facilitate the implementation of recommendations that stem from the corridor study by working independently or with federal, state, local, and private groups in any appropriate combination. Findings As of 1991, the Missouri River Corridor Project has identified 84 major fish and wildlife habitat, culturallhistorical, and public recreation sites and a multitude of smaller, similar sites within the Phase I and Phase II project areas. This plan calls for the simultaneous enhancement, over time, of elements within each of these three resource categories, with the three interacting together to achieve the region's fullest possible enhancement potential. The key foundation to the eventual success of this endeavor is the rehabilitation of the riparian wetlands and adjacent habitats for native plants, fish, and wildlife. Without these scenic natural attractions and the re-creation of a sustainable balance between the region's human population and the natural resource base, it appears to be quite unlikely that the project will meet the NRDs' advanced goals of increasing

PAGE 332

308 Missouri River Corridor Project KANSAS Figure 1. Missouri River Corridor Project, Phase I and Phase II.

PAGE 333

Sowl 309 tourism and economic development within the corridor. The reason for this is that the natural environment within the corridor seems to offer the most promise of being the principal attractive feature in the study area. In addition to offering one of the world's major rivers, the broad valley is bordered by dramatic, rolling bluffs; patches of both bottomland and upland forests are scattered through the landscape, as are remnants of once-extensive native prairie, riparian wetlands, and oxbow lakes. Breathtaking scenic vistas can be experienced from atop many of the bluffs. There are many areas that offer excellent hunting and fishing opportunities as well as a chance for nature enthusiasts to enjoy the out-of-doors, away from the confines of the cities. Soils in the area are, for the most part, deep and rich and form the basis for a strong agricultural economy. Five of the highest quality oxbows or backwaters out of the 26 identified are having designs created by the U.S. Army Corps of Engineers, NRDs, and private consultants for the rehabilitation and restoration of fish and wildlife habitat and public recreation sites into the Missouri River ecosystem. The Middle Missouri River corridor is filled with both cultural and historical points of interest. These include sites of prehistoric habitation by Native Americans, the Lewis and Clark Trail, jumping-off points for the westward Euro-American migrations of the early nineteenth century, old fort sites, period architecture, Indian reservations, and towns and cities, to name a few. The study area seems to lend itself to four general subject categories: Native Americans, Early Exploration, Westward Expansion, and Life on the Plains. While the region is not limited to these themes, they do appear to offer the greatest potential to facilitate the preservation and development of the area's culturallhistorical resources and summarize the essence of the region'S character. As of 1991, two culturallhistorical sites are being developed based on conceptual designs offered by the National Park Service: one is the Blackbird Wayside Area and Scenic Overlook-an interpretive shelter in the shape of a stylized Native American earth lodge. Situated on a prominent bluff over looking the river valley, the shelter will have panels telling of the natural and cultural history of the area and its inhabitants through time. The second site is a 136year old farm called Golden Spring. Again, an interpretive shelter will describe the historical aspects and significance of the site. Much of the existing and potential public recreation opportunities within the Missouri River corridor have some relationship to the area's natural, cultural, and historical resources. There are several clusters of recreation resources located within the corridor with other sites scattered along the river. These clusters are areas that offer a special diversity of public recreation opportunities, such as boating, fishing, hunting, the enjoyment of nature, excursion trains, parks, riverboats, museums, festivals, bicycling, and auto touring routes. The Missouri River itself figures heavily in any public recreation equation by providing a much-needed water recreation base in a region starved for adequate open-water recreation opportunities.

PAGE 334

310 Missouri River Corridor Project Conclusions The Missouri River Corridor Project offers a great variety of opportunities to preserve and restore significant portions of the Missouri River ecosystem, and to enhance the area's cultural, historical, and recreation resources; in short, the beginnings of a renaissance for the long-abused river resource. If action is taken now, it seems possible that most major projects within the corridor could be completed by the year 2000. As oflate 1991, two additional NRDs have applied to the National Park Service to extend the corridor project along the river another 229 kilometers (143 miles) to the north and west up to the Nebraska-South Dakota border, for a total corridor length of 616 kilometers (385 miles). Application has also been made by a Resource Conservation and Development Project (RC&D) in South Dakota to extend the project another 475 kilometers (297 miles) through a portion of that state along the Missouri. According to Brown et al. (1990), this decade may be critical for deciding the fate of our natural resource heritage. For the river itself, this next decade could represent the last real chance to save those fragments that remain of what was once one of the continent's richest biological ecosystems. For the people, this project offers an enhanced awareness of their heritage within the Middle Missouri River Valley and provides opportunities to experience the unique resources of this region. References Bragg, Thomas B., and Annehara K. Tatschl 1977 "Changes in Flood-Plain Vegetation and Land Use Along the Missouri River from 1826 to 1972." Environmental Management 1 (4): 343-348. Brown, Lester R., et al. 1990 State of the World. New York. W.W. Norton and Company. Fredrickson, Leigh H. 1979 Floral and Faunal Changes in Lowland Hardwood Forests in Missouri Resultingfrom Channelization, Drainage, and Impound ment. Washington, D.C.: Eastern Energy and Land Use Team, U.S. Fish and Wildlife Service. Funk, John L., and John W. Robinson 1974 Changes in the Channel of the Missouri River and Effects on Fish and Wildlife. Jefferson City, Missouri: Missouri Department of Conservation.

PAGE 335

Sowl 311 Hallberg, George R., Jayne M. Harbaugh, and Patricia M. Witinok 1979 Changes in the Channel Area of the Missouri River In Iowa; 1879-1976. Iowa City, Iowa: Iowa Geological Survey. Siouxland Interstate Metropolitan Planning Council (SIMPCO) 1978 Missouri River Woodlands and Wetlands Study. Sioux City, Iowa: Sowl, John H. 1986 Siouxland Interstate Metropolitan Planning Council. A Management Plan for Oxbow Lakes on the Middle Missouri River. Madison, Wisconsin: Master's Thesis in Landscape Architecture from University of Wisconsin-Madison.

PAGE 336

SANTA ROSA CREEK RESTORATION PROJECT Linda Stonier National Park Service Rivers, Trails and Conservation Assistance Program The city of Santa Rosa is located in rural Sonoma County, 60 miles north of San Francisco. With a population of 160,000, it is one of the country's fastest growing urban areas. It was settled in the mid-1800s as a farming and commerce center. In those days, Santa Rosa Creek supplied water to a local brewery. New infrastructure in 1873 piped its waters to city residents. Steelhead spawned in the creek into this century. Over time, the creek became a dump for dairy manure and tannery waste and a conduit for human sewage effluent. With development of the watershed, it was increasingly faced with sudden bursts of runoff and property damage became increasingly expensive. The flood of 1955 took a human life. In the 1960s the Soil Conservation Service funded "drainage improvements" on Santa Rosa Creek. The local match was the cost of right-of-way acquisitions. The engineers designed a straight, trapezoidal, gunnite channel for the 1 % flood with maximum watershed development. A box culvert was used downtown. Santa Rosa Creek flows 25 miles westward from its headwaters on Hood Mountain to the Laguna de Santa Rosa, a unique, low-lying basin. An upstream diversion and floodwater impoundment controls flood waters in the upper reaches where the channel is still soft. Downtown, the creek emerges from underground into a grouted rock channel, which is owned and maintained by the Sonoma County Water Agency (under authority of the Sonoma County Board of Supervisors). The channel is flanked on both sides by shaled service roads. Adjacent land uses include redevelopment and new commercial areas, subdivisions, agriculture, and parks. As they approach the Laguna, the service roads rise into levees and the stream ceases to be a flood control channel. A compromise between floodwater conveyance and the natural order has developed here. While streamside vegetation is maintained at a two-year successional stage, some remnant riparian forest and remnant creek channel remains. Agricultural fields flood. The channel is attempting to re-establish its natural morphology and re-ereate its meanders. In 1969, long before the term "multi-objective corridor planning" was coined, the city of Santa Rosa began to think about the health of its creek system again. With the drainage improvements almost completed, the city adopted a Natural Waterways Study as part of its General Plan. Its conclusions included the recognition of retaining all waterways in a natural condition as a worthy goal. Its recommendations included a set of priorities for recognizing outstanding natural qualities and multiple use potential of the area's waterways. Meanwhile, channelization of the creek system has continued. Three years ago, a coalition of business people, architects, planners, political activists, politicians, ecologists, and property owners formed the Committee for the Restoration of Santa Rosa Creek. The success of their creek restoration project,

PAGE 337

Stonier 313 which is receiving technical assistance from the Rivers, Trails and Conservation Assistance Program, is attributable to the following elements. Element #1: A diverse, representative, well-connected core citizens' coalition. The broad-based makeup of the committee exemplifies the multiobjective planning model. Committee goals identify the creek as a connector and catalyst for downtown revitalization, while seeking to protect and enhance the natural qualities. Element #2: Early, broad public participation in creating a vision. To increase awareness and interest in creek values, the committee set up a speakers' bureau and led creek walks. In "T AKEP ART" workshops, citizens built models of their dreams for a restored Santa Rosa Creek. Creating access with creekside trails, enhancing community identity with water features in downtown development, and protecting creek ecology were among the citizens' multiobjectives. Element #3: Attention to education. The committee is actively engaged in informing itself, the political and agency establishment, and the community-at-large about creek restoration possibilities. In March 1991, the committee and the National Park Service sponsored a two day In-Service Forum. The purposes of the forum were 1) to introduce innovative creek restoration techniques to local public officials, agency personnel, and citizens' task force members who are developing the master plan and who will be overseeing its implementation; and 2) to demonstrate and underscore the interdisciplinary nature of a good stream restoration project. A team of visiting experts in hydrology, riparian ecology, native fisheries, urban design, and interagency cooperation were invited to Santa Rosa to brainstorm solutions, with local participants and with each other, for a restored Santa Rosa Creek. Participants took field trips, then captured their ideas into restoration recommendations and priorities. Element #4: Small, simple pilot projects demonstrating restoration techniques, providing public access to the creek, and creating opportunities for agencies, landowners, and volunteers to work together. A Brush Creek restoration project will use terraces to expand the cross-sectional area of 1,200 feet of tributary channel, making room for riparian vegetation to be planted. This project is the first attempt to undo the creek channelization. The planning, design, and cost-sharing team involves public, private, and nonprofit players. The project has been awarded a $45,000 grant from the California Urban Stream Restoration Program. The creekside trail, already popular with local residents, is a simple, low-cost footpath. Element #5: Buy-in of the public agencies. As previously mentioned, the Water Agency owns and operates the channelized portion of the creek. A Creek Restoration Master Plan is currently under way, jointly funded by the city, county, and Water Agency, and guided by a planning team of agency personnel and citizens. The success of this endeavor is as dependent upon the willingness of the agencies to be team players as it is upon their financial support. Element #6: A combination of creative thinking and concrete goals for the short and long-term, as the following recommendations from the In-Service Forum illustrate:

PAGE 338

314 Santa Rosa Creek Restoration 1. Protect the headwaters. As they are still in good condition, this is your restoration model; 2. Work out the agreements between multiple responsible agencies early; include recreation in the initial planning; 3. Biologists, landscape architects and engineers should interact in the design stage; 4. Think about ultimate corridor management responsibilities at the design stage. The managing agency should have multiobjective goals. Think about its mission, personnel, funding, and training up front; 5. Design the corridor so that you can break it into pieces, and take them one at a time. Practical design suggestions included the following: 1. Revisit the 1 DO-year flood calculation. New and better computer models can give a much more detailed water surface profile. A 10% difference in the predicted flood level can provide significant capacity to work with in restoration. Improving conveyance downstream improves things upstream. 2. Look at changing maintenance practices. Concentrate maintenance activities to one side of the creek. 3. Allow for flexible channel design. For example: a. Cut and fill benches in the grouted rip-rap for pedestrian walkways or plantings, or both; b. Pull maintenance roads down the slope to create room for trees next to a low-flow channel; c. Excavate part of a slope and vegetate above a retaining wall; d. Put a vertical wall on the north side to allow planting on the south bank. e. To pick up freeboard, use floodwalls or replace bridge footings. 4. Where right-of-way is available, re-create a two-stage channel. Keep land use intensity low and allow the floodplain to flood. 5. Require developers to retain runoff from new or incremental impervious surfaces on site. This will improve water qUality. Moreover, not taxing the channel with this runoff will provide more flexibility for reconstruction to accommodate peak flows. 6. At appropriate sites, especially downtown, scallop out the channel and allow development to cantilever over it in compensation; use the area underneath when it isn't flooded. 7. Recognize that you can't do everything everywhere in the short term, so set some priorities. Over the long term, however, there are no limitations.

PAGE 339

Stonier 315 References Bums, Jim, Committee for the Restoration of Santa Rosa Creek, and National Park Service, Rivers, Trails and Conservation Assistance Program. 1990 Creek Dreams Revealed: An Idea Book. Draft. Santa Rosa, California. City of Santa Rosa 1969 Natural Waterways Study, Santa Rosa Area General Plan. Santa Rosa, California: City of Santa Rosa. 1991 Brush Creek Restoration Grant Application to the California Department of Water Resources Division of Local Assistance Stream Restoration Program. Santa Rosa, California. Dodds, Robert 1976 The Piteous Tale of an Urban Creek: A Very Interpretive History of Santa Rosa Creek. The 200 Series, No. 11. Santa Rosa, California: Santa Rosa Junior College. Elizabeth S. Andrews, Phillip Williams and Associates, San Francisco, California; John L. Barnett, Greenway Coordinator, Boulder, Colorado; Earle Cummings, Program Manager, California Department of Water Resources, Urban Stream Restoration Program; William J. Hoeft, Senior Civil Engineer, Santa Clara Valley Water District, San Jose, California; Robert A. Leidy, Fish and Wildlife Biologist, U.S. Environmental Protection Agency, San Francisco, California; Robert M. Searns, Urban Edges, Inc., Denver, Colorado; J. Theodore Stanley, President, The Habitat Restoration Group, Scotts Valley, California; Phillip B. Williams, Phillip Williams and Associates, San Francisco, California. 1991 Visiting Experts to Santa Rosa Creek In-Service Forum.

PAGE 340

SOUTH PLATTE RIVER IMPROVEMENTS: FLOODPLAIN MANAGEMENT ACHIEVES MULTIPLE OBJECTIVES Steven R. Williams Hydro-Triad Ben Urbonas Urban Drainage and Flood Control District R. Jay Nelson Dames and Moore Introduction An alternatives evaluation and preliminary design study was prepared for a reach of the South Platte River near Denver, Colorado, identified as the Globeville and North Areas. The project required the coordination of sponsor, agency, and citizen goals to achieve a mUltipurpose flood control project. History This reach of South Platte River over the last 75 years was urbanized with considerable fill and narrowing of the natural channel. Its banks were covered with concrete rubble and other materials to fight bank erosion and to sometimes dispose of such materials. Parts of the river bottom were dredged for gravel, resulting in considerable channel degradation. As the area urbanized and became an industrial zone, its importance to the Denver area economy grew and became a very important employment center. This reach of the South Platte River was flooded on several occasions in the past, most recently in 1965 and 1973. The 1965 flood was judged to exceed the predicted 100-year peak flow. Since the area is primarily industrial in land use, flooding not only causes direct damage to property, but more importantly disrupts the economy and employment within this area. Thus, increasing flood carrying capacity of the river in this reach has significant economic benefit to Denver, Adams County, and Commerce City. Project Goals Goals were established between the project sponsors and funding agreements were formulated. Goals included: Total containment of the lOO-year flood within river banks so as to meet FEMA guidelines. Maintain valuable riparian river habitat and improve on it where possible to make the river corridor as "natural" and aesthetically pleasing in character as the limited right-of-way would permit.

PAGE 341

Williamr, Urbonas, and Nelson 317 Protect and enhance when possible public access to the river, including the addition of hiker and biker trails. Integrate all of the needed improvements with other improvements planned by public and private organizations along the river corridor, including new parks, highway improvements, expansion on the National Western Facilities, etc. AHernative Development and Evaluation As indicated previously, the overall project goal was to identify alternative solutions to flooding and river bank overtopping within the study area. A wide variety of alternative floodplain improvements could satisfy the project goals. To maintain a manageable group of alternatives for consideration, a three-phased approach was used. The Phase 1 goal was to develop a series of alternative concepts for the entire study reach. Phase 2 work focused on defining sub reaches within the study reach and developing alternatives for each subreach. Phase 3 consisted of combining sub reach alternatives identified in Phase 2 within the framework of the entire study reach project goals. Four concepts were evaluated as part of Phase 1. These consisted of lowering the water surface elevation, filling the floodplain outside of the floodway, confining the flood flows within the existing channel right-of-way, and floodproofing structures within the existing floodplain. As part of Phase 2, twelve sub reaches were defined, based on structural, hydraulic, and utility constraints. Following evaluation of alternatives for each sub reach, four base alternatives were established for the entire project reach. The four base alternatives included modification/relocation of existing structures, construction of floodwalls or levees, excavation of the channel invert, and widening of the river channel. A series of maps, graphics, and evaluation matrices were prepared for project meetings. Project sponsors, agencies, and other interested community groups were invited to attend these meetings and comment on the various alternatives. Each alternative was evaluated based on engineering opportunities and constraints and also on a qualitative basis with regard to environmental, recreational, and aesthetic opportunities and limitations. From this review process 18 alternatives were selected for further, more detailed consideration. The final alternative evaluation process, Phase 3, focused on combining portions of the 18 alternatives into a smaller, more manageable group of alternatives. Phase 3 evaluations would include economic considerations in addition to environmental, aesthetic, and recreational goals. Six best available alternatives were identified. Plan and profile maps were prepared to illustrate the six alternatives and both qualitative and quantitative matrices were prepared to illustrate opportunities and constraints.

PAGE 342

318 South Platte River Improvements Preliminary Design Several sub reach improvements were found to be common to all alternatives. The selected preliminary design alternative consisted of a combination of the final six proposed alternatives. The combined alternative included extensive use of floodwalls and flood levees to limit the purchase of right-of-way. Boater passage chutes were added to all existing drop structures and the Burlington diversion dam. The Franklin Street Bridge and adjacent railroad bridge and water line will also be raised. Recreation opportunities were enhanced through the use of pedestrian and bicycle trails, increased open areas, and pocket parks. Existing areas with significant vegetation and wildlife habitat will be preserved and protected. Gravel point bars and disturbed areas will be revegetated using upland grass mixtures and wetland plant species. Conclusion Upon review of the alternative evaluation and preliminary design process, several items stand out that had both positive and negative impacts. Perhaps the foremost is the need to quickly develop and evaluate all available alternatives on a conceptual basis. The intent of this process is to reduce the number of alternatives to a more manageable number early in the project. A second important item is the need for accurate channel cross-section data and structure measurements. We found reliance on prior construction drawings and survey data resulted in errors in the hydraulic analysis. Third, the need for up-to-date and accurate right-of-way information is extremely important. Without these data, accurate evaluation and economic analysis cannot be done. Finally, the importance of the softer improvement features cannot be overlooked. These include landscape design, environmental mitigation, recreational needs, and blending with adjacent land-use plans.

PAGE 343

Part Ten Geographic Information Systems and Flood Hazard Mapping

PAGE 344

This page IS intentionally blank

PAGE 345

GUIDELINES AND SPECIFICATIONS FOR EROSION STUDIES Mark Crowell and Michael K. Buckley Federal Emergency Management Agency Federal Insurance Administration Introduction The Federal Emergency Management Agency (FEMA) is preparing for a potential nationwide study that will involve the collection, analysis, and computation of coastal erosion rates. This study will provide the basis for administering Section 544 of the Housing and Community Development Act of 1987 (commonly known as the Upton-Jones Amendment) as well as other potential programs of land use management that deal directly with shoreline recession and erosion. To plan and administer programs of this nature, FEMA is developing a comprehensive set of "Guidelines and Specifications for Erosion Studies. The purpose of the guidelines is to: 1) provide standards for the review of existing erosion rate data, and 2) standardize data collection techniques and erosion analysis methodologies for the development and compilation of new erosion rate data (Crowell et al., 1991). The guidelines will be used by FEMA study contractors to insure that erosion data for the nation's coastal and Great Lakes shorelines are compiled in a consistent, accurate, and reliable format and are incorporated into a geographic information system (GIS) database. The bulk of the guidelines contains specifications for the compilation and development of new erosion rate data. This is essential because accurate computer ized methodologies for collecting data of this complexity have only been developed within the past 10 years, and virtually all of the researchers engaged in such studies use a variety of disparate methodologies. Thus, acceptable erosion rate data (that is, data commensurate with the basic requirements of our program) are available for only 10-40% of our nation's coast. Nonetheless, the first step in the study procedure will be for the FEMA study contractor to review and assess existing erosion rate data to determine their accuracy and potential for direct use in the erosion program. Existing data may be used in their entirety or modified and/or updated to meet program standards. If the existing data fail to meet the specifications listed in the guidelines and cannot be updated or enhanced, or if the data can only be partially salvaged, then the study contractor shall initiate new erosion rate studies. The methodology presented in these study contractor guidelines for compiling new erosion rate data was recommended by the National Academy of Sciences for use under the National Flood Insurance Program. This methodology requires the digitization of shorelines from four to eight historical maps and current air photos. These shorelines are combined and stored in a GIS database. Subsequent statistical and empirical analyses of the multiple shorelines shall enable FEMA to compute and predict future erosion rates for use in the determination of coastal setback lines.

PAGE 346

322 Guidelines for Erosion Studies The principle steps and/or components involved in the compilation of new erosion rate data for coastal communities are detailed in the guidelines and are as follows: 1. Review Existing Erosion Rate Data. The FEMA study contractor reviews and assesses existing regional, county, or statewide erosion rate data to determine their potential for direct use in the erosion program. If satisfac tory data do not exist for a particular area, then new erosion rate studies will be performed. 2. New Erosion Rate Studies. A. Compile a Historical Shoreline Location Database (HSLD). The study contractor creates a GIS database that contains the digitized position of four to eight historical and current shorelines. The shorelines are digitized from maps dating from the 1840s to the present and air photos dating from the 1930s to the present. B. Create a Historical Shoreline Positional Change Database (HSPCD). This database is created by running a transect program on the HSLD. The program produces a series of shoreline-perpendicular transects, spaced at ISO-foot intervals, and overlays them on the multiple historical and recent shorelines contained in the HSLD. The program measures and tabulates the position where the historical and current shorelines intersect each transect and stores the data in a matrix. This database is the source from which subsequent statistical computations (see below) are per formed. c. Create a Blocked Erosion Rate Database (BERD). The study contractor develops this database by applying statistical, geomorphic, and qualitative analyses to the transect data contained in the HSPCD. This allows for the determination of a single erosion rate for each transect. Next, the study contractor blocks the erosion rate data by combining segments of the coast having similar rates of erosion into wider regions for which a single representative erosion rate has been determined. This converts the data into a more usable and manageable format. D. Produce Draft Erosion-Flood Insurance Rate Maps (E-FlRMS). The study contractor supplies FEMA with draft Erosion-Flood Insurance Rate Maps (E-FIRMS) that show the blocked erosion rates and the location of blocked erosion rate borders. E. Prepare an Erosion Rate Study Report (ERSR). The study contractor prepares a report that provides a background of the procedures and methodologies used to compile the erosion data and also provides a justification for the statistical and qualitative methodologies applied in the studies. Technical guidance is also given in a series of appendices that provide the following:

PAGE 347

Crowell and Buckley 323 Detailed instructions for raw data preparation, digitization, and transfor mation (modified from Leatherman et al., 1991). Discussion of source data error analysis and mapping accuracy (modified from Crowell et al. 1991). Techniques of statistical analyses of erosion rate data (modified from Dolan et al., 1991). References Crowell, Mark, A. Todd Davison, and Michael K. Buckley 1991 "Comprehensive Erosion Hazard Identification Through the National Flood Insurance Program." In Geological Sodety of America, Abstracts with Programs, Volume 23. In press. Crowell, M., S.P. Leatherman, and M.K. Buckley 1991 "Historical Shoreline Change: Error Analysis and Mapping Accuracy, Journal of Coastal Research. In press. Dolan, R., M.S. Fenster, and S. Holme 1991 Analysis of Shoreline Erosion and Accretion: Spatial and Temporal Sampling. Department of Environmental Sciences, Univer sity of Virginia, Charlottesville, Virginia. Unpublished Report. Leatherman, S.P., G. French, and L. Downs 1991 Historical Shoreline Mapping: Procedure and Accuracy Assess ment. Laboratory for Coastal Research, University of Maryland, College Park Maryland. UnpUblished Report.

PAGE 348

MULTI-OBJECTIVE RIVER CORRIDOR PLANNING USING GEOGRAPHIC INFORMATION SYSTEM METHODS: A CASE STUDY OF THE CACHE LA POUDRE RIVER CORRIDOR Duane A. Holmes River and Trail Conservation Assistance Program National Park Service Abstract In 1987, the National Park Service (NPS) was asked by the city of Greeley, Colorado, to assist the city and Weld County in establishing a geographic infor mation system (GIS) data base for multiple objective river corridor planning applications. After receiving approval from our Washington, D.C., office, we agreed to proceed and selected the Geographic Resources Support System (GRASS) as the GIS software for this effort. Digital thematic map layers used in this effort included: U.S. Geological Survey hydrographic, transportation, and public land boundary digital line graphs (OLGs) and U.S. Geological Survey digital elevation models (OEMs) for the 71h minute Greeley, Bracewell, and Windsor quadrangles; custom digitized zoning, land use, wildlife habitat, historic features, land ownership, farm road, mineral extraction, floodplain, and developed recreation site layers; and a 10meter resolution, panchromatic satellite image. The first use for this digital data base was the identification of an optimal greenway corridor linking the town of Windsor and the city of Greeley, which are about 15 miles apart along the Cache la Poudre River. It is proposed that this greenway eventually link with one being built along the Cache la Poudre in Larimer County and the city of Fort Collins. GRASS GIS tools used in the conduct of our analysis included: weight (which allows one to produce combined overlay maps in which various categories represented on a single overlay map are given a range of values and then summed), Gcost (which evaluates the cost of traversing a grid cell surface in all directions from a starting location), and Gdrain (which traces a flow through a lowest cost path in an elevation model [USACE, 1988]). In this effort, "elevation" is actually an imaginary surface, overlaying the planning area, in which a greater cost (eg., economic, social, environmental) of achieving a planning objective is represented by a higher elevation. Background The NPS maintains a small but highly motivated nationwide staff that is dedicated to using innovative means to leverage the efforts of their limited resources Gary S. Waggoner in the Geographic Information Systems Division of the National Park Service provided most of the inspiration for applying the methods used in conducting this corridor analysis. Gary, Dave Duran, Sharon Shin, Nancy Thorwardson. and Susan Stitt (all in our GIS Division) made this possible.

PAGE 349

Holmes 325 to assist as many state and local governments as possible in the local development of trail opportunities and river or other conservation and enhancement plans. The city of Greeley, Colorado, became aware of the efforts of this program-the River, Trail and Conservation Assistance Program-and requested our assistance. Greeley's idea for linking two counties and up to four communities by a greenway, and using a GIS in planning for the effort, was a bold idea and one we were proud with which to work. This project afforded us the opportunity to evaluate the effectiveness and efficiency of using GIS tools and methodologies for corridor planning. The city's plans to set up their own GIS and incorporate our cooperatively developed digital data base for use in the multi-year greenway project implementa tion phase as well as water, sewer, zoning, and other planning lent strength to the position that this would be an appropriate investment in digital data base construction (which is the main cost and labor consumptive part of any GIS oriented effort) (Burrough, 1986). GIS analyses conducted as part of this planning effort differ little if any, from typical McHargian planning methods (McHarg, 1969). The advantages of the GIS to this effort included the large number of planning constraints and opportunities susceptible to simultaneous analysis, the rapidity with which a great variety of alternative plans were generated and analyzed, the highly evolved map generating capabilities of the GIS, and opportunities for future plan modifications based on "as-implemented" changes in some corridor segments. Also, an as-yet untested (by us) GIS capability to generate a realistic-appearing, as-it-would-look-if constructed, videotape of project proposals could prove to be an extremely powerful communication tool in project approval and implementation stages. Two major difficulties were presented in choosing to use GIS tools and methodologies in developing our plan. The first was the time and personal commitment required to learn about how to set up and use this highly complex and often obscure system of analysis, then communicating related needs for time, funding, and cartographic accuracy to others involved in the planning effort or overseeing our program management. The second difficulty was the great deal of time required for data generation, collection, editing, and compilation into a comprehensive data base capable of sustaining analyses for a wide range of ultimate planning analyses and multiple community objectives. Other than the employment of a GIS, this planning effort was quite traditional in its scope and execution. Meetings were held with project proponents to identify initial interests and tasks for the effort. Public meetings were held to identify issues, concerns, and desires for the plan. Data were collected on the corridor's resource opportunities and constraints. Additional meetings were held with corridor landowners to identify their specific concerns. Goals were refined; methods of resolving concerns and satisfying desires were developed; alternative plans were generated, reviewed, and modified; and a final plan was developed. The plan is currently undergoing minor modifications prior to printing and distribution.

PAGE 350

326 Cache La Poudre River Corridor System The computer system used for the analyses is a Dell System 310, which is an 80386-based PC running at 20 MHz. It has 4 MB of RAM, a 322 MB hard disk drive, and a 20 MHz 80387 math coprocessor chip. An NEC MutiSync 2a with 640 X 480 pixel resolution serves as our graphics monitor and a WYSE-60 monochrome monitor with an enhanced 10 1 keyboard serves as our command input device while operating in graphics mode and a Tektronix 4696 color inkjet plotter serves for draft output. We also have a mouse and other output devices (printer and pen plotter) hooked to our system. Thirty-two MB of our hard disk are partitioned off for normal DOS operations while the remainder of our hard disk contains our GIS and the UNIX operating system on which it runs. Total system cost, excluding all output devices except for the pen plotter, was just under $15,000. Data acquisition cost approximately $3,500. Our costs for hardware, software, and data acquisition should be viewed as very low. Anyone attempting a similar effort should anticipate expending roughly three times the above amounts. Analysis As a result of our public hearings, data gathering, and corridor inspection, we concluded that historic and cultural features, hydrology, wildlife habitat, existing land use, transportation, mining, zoning, slope, aspect, transportation, existing developed recreation site locations, and the lOO-year floodplain were important to our analyses. The NPS acquired and coded the DLGs and DEMS available for the area from the U.S. Geological Survey and the satellite imagery, while the city of Greeley drafted the other thematic layers on stable base material (mylar) for digitizing under an NPS contract. Map layers were digitized in normal SAGIS import format (a vector or line file format), then converted to GRASS format (which is a raster-based or cell file format). These activities made up our data base construction effort. Following data base construction, the GIS was (or is currently being) used to identify: a variety of potential trail corridor locations based on differing assumptions of what constituted optimal siting; priority areas for conservation or enhancement based on existing and/or potential environmental or cultural values; optimal areas for development based on existing zoning and land uses and recognizing those areas that should be conserved or that are considered necessary for trail development; the best areas for recreation node development based on trail siting, development areas, zoning, land ownership, nature of the activity the site would be developed for, suitability of all corridor sites for that activity, etc. and; prime areas for interpretive or educational activities.

PAGE 351

Holmes 327 Summary Acquiring hardware, software, a comprehensive data base for multi-objective planning applications, and staff skilled in GIS-oriented planning requires a large financial and time investment. However, if complex analyses are to be made of a large number of factors and if a commitment has been made to maintain and continue use of the data base, a digital data base and GIS system could be an excellent investment. Burrough, P.A. 1986 McHarg, Ian L. 1969 References Prindples of Geographical Information Systems for Land Re sources Assessment. Clarendon Press. Design with Nature. Doubleday and Company. u.S. Army Corps of Engineers (US ACE) 1988 Users and Programmers Manualfor the Geographical Resources Analysis Support System. U.S. Army Corps of Engineers, Construction Engineering Research Laboratory.

PAGE 352

NFIP MAP REVISION PROCESS MADE SIMPLE Mary Anne Lyle Federal Emergency Management Agency Vince DiCamillo and Mary Jean Pajak Greenhome and O'Mara Obtaining a revision to a National Flood Insurance Program (NFIP) map is not a complex task if one has well-defined concepts of the following: 1) the revision basis, 2) the applicable NFIP regulations, and 3) the data required by the Federal Emergency Management Agency (FEMA) to support the revision. The procedures for revising and amending NFIP maps are contained in Parts 65, 70, and 72, Title 44, Chapter 1, of the U.S. Code of Federal Regulations. Documents exist that provide assistance to persons seeking revisions to NFIP maps. These documents, which may be obtained from FEMA, are titled Appeals, Revisions, and Amendments to Flood Insurance Maps, A Guide for Community Officials, dated January 1990, and Revisions and Amendments to National Flood Insurance Program Maps, A Technical Guidefor Engineers and Hydrologists, dated January 1991 (hereafter referred to as the Technical Guide.) This paper does not detail the specific data requirements for each type of revision; however, detailed information can be found in the documents listed above. The Technical Guide contains information about data requirements specific to revision requests involving coastal and alluvial fan flooding, and both documents address data requirements for requests involving riverine flooding and floodways. Typical bases for revisions or amendments, the NFIP regulations that apply, and the mechanisms for the revisions or amendments are as follows: 1. A property or structure has been inadvertently included within the Special Flood Hazard Area (SFHA). No fill is involved. Inadvertent inclusions are usually due to map scale or topographic limitations-Part 70 of the NFIP regulations applies, and the mechanism for the revision is a Letter of Map Amendment (LOMA). 2. A property or structure has been elevated to or above the base flood elevation (BFE) by the placement of fill so that it is no longer within the SFHA-Part 65 (Section 65.5) of the NFIP regulations applies, and the mechanism for the revision is a Letter of Map Revision (LOMR) based on fill. 3. A refinement of the SFHA boundary shown on the effective NFIP map is warranted by more detailed or more accurate topographic data (no change to BFEs)-Part 65 (Section 65.5) of the NFIP regulations applies. The mechanism for the revision may be a LOMR or repUblication of the map; FEMA makes the determination once the request is received and reviewed. 4. A revision to the BFEs shown on the effective NFIP map is warranted by a more accurate hydraulic model. This situation can occur when the

PAGE 353

Lyle, DiCamillo, and Pajak 329 requestor adds new cross sections at the site of the revision-Part 65 (Section 65.6) of the NFIP regulations applies, and the mechanism for the revision may be a LOMR or republication of the map. 5. A revision to the floodway boundary shown on the effective NFIP map is warranted by a more accurate hydraulic model. This situation also may result when the requestor adds new cross sections at the site of the revision-Part 65 (Section 65.7) of the NFIP regulations applies, and the mechanism for the revision may be a LOMR or republication of the map. 6. A proposed project may result in a revision based on one or more of the situations described in items I through 5-Part 72 of the NFIP regulations applies, and the mechanism for the revision may be a conditional LOMA, a conditional LOMR based on fill, or a conditional LOMR. Revision requests should be sent to the FEMA regional office. Some requests are handled at the regional office, others are forwarded by the regional office to FEMA headquarters in Washington, D.C., and on to the Technical Evaluation Contractor (TEC) for review. The first step in FEMA's processing is to determine the type of revision request and to ascertain that all necessary supporting data have been received. If additional data are required, a checklist is sent to the community and/or requestor (if other than the community) requesting the required data. No technical review of any data is initiated until all necessary data are received. The requestor is responsible for supplying the data needed to support the revision. Once all data are received, the technical review is initiated. The TEC reviews the supporting data to verify that: 1. The community has concurred on the request. 2. The state has approved the request, if applicable. 3. The topographic mapping supports the desired SFHA boundary. 4. The hydrologic methods utilized are appropriate and have been applied correctly. 5. The appropriate hydraulic models have been supplied for review and the modeling has been performed correctly. 6. Documentation has been provided that indicates that affected property owners have been notified of the changes. 7. The annotated maps, tables, and profiles correctly depict the modified SFHA boundaries and/or floodway. 8. Any additional issues of concern to FEMA regarding the request have been resolved. If, during the technical review, errors or inconsistencies are found in the data provided, a letter is sent to the community and/or requestor requesting that the errors or inconsistencies be resolved. No further action is taken on the request until a response to this letter is received.

PAGE 354

330 NFIP Map Revision Process After all such errors and inconsistencies are resolved, the TEe reviews the data submitted in support of the revision request with FEMA, and FEMA determines the appropriate mechanism for resolution of the request. The most expeditious method is a revision by letter. Typically, a letter will be issued within four weeks of FEMA's final determination concerning the revision request. The total time required to complete a revision request resolved by letter is approximately three months from the date all data are received. Some revisions, however, may require republication of the map. A republished map could take up to 15 months to become effective. Figure 1 summarizes the basic data required for SFHA boundary revisions and the additional data required for BFE revisions, floodway revisions, and conditional revisions. All materials submitted in support of a revision request should be neatly organized and securely bound together. The hydrologic and hydraulic methodologies should be presented clearly, all assumptions should be stated, and all methods used should be described. Enough data must be presented to enable the reviewer to verify the results. All changes desired must be documented. It is important to include a copy of the effective NFIP map annotated to show the desired changes. An annotated map is required because it illustrates for FEMA the specific changes the requestor expects FEMA to make. In addition, all topographic and other technical data must be certified by a licensed land surveyor or registered professional engineer, as appropriate. Typical problems encountered by the reviewer include the following: 1. Inconsistencies between the hydraulic calculations, workmap, and annotated NFIP map (e.g., the calculated floodway width does not agree with the floodway plotted on the workmap and/or annotated NFIP map, the distances between cross sections in the hydraulic model do not agree with those on the workmap and/or annotated NFIP map, the floodplain boundary delineated on the workmap does not match the boundary shown on the annotated NFIP map). 2. Hydraulic modeling problems (e.g., error messages that are not resolved or explained, inappropriate treatment of unusual flow patterns, use of unusual parameters, incorrect bridge modeling). 3. Floodway surcharge violations. 4. Revised BFEs or revised floodplainlfloodway boundaries that do not tie in to those shown on the effective NFIP map in areas not affected by the revision. If the requestor performs the basic checks necessary to ensure that such problems do not appear in the supporting data to be submitted, the revision request should proceed smoothly through the technical review stage.

PAGE 355

Lyle, DiCamillo, and Pajak BASIC DATA REQUIREMENTS FOR SPECIAL FLOOD HAZARD AREA (SFHA) WITH NO CHANGE TO BASE FLOOD ELEVATIONS (BFEs) Leiter from community requesting revision Description 01 request Annotated copy of the effective NFIP mop Certified topographic dotal Certification that fill placed meets the requirement for 95% compoclion2 Slate approval of Ihe revision request, if applicable I Revisions to approximate SFHA boundaries will be based on Ihe new certilied topographic dolo and FEMA's best estimate 01 the BFE. Reques" based on reanalyses 01 Ihe Ilaod hazards in approximate SFHAs must be accomponied by Ihe supporting dolo lor Ihose analyses. 2Required lor revisions based on lill areas larger than a single residennallol. BASIC DATA REQUIREMENTS FOR BFE REVISIONS Leller from community requesting revision Description 01 request Annotated copy of the effective NFIP mop Hydraulic analyses for 10-, 50-, 100-, and 50D-yeor Roods Annotated Flood Profile State approval of the revision request, if applicable HydrologiC analyses3 State approval of revised discharges, if applicable 3 Required lor revisions based on revised lIood discharges. BASIC DATA REQUIREMENTS FOR FLOODWAY REVISIONS Leller from community requesting reVision Description of request Annotated copy of the effective NFIP map Hydraulic analyses for 1 OD-year flood and floodway Annotated Floodway Dota Table 331 Copy of public notice stating the community's intent to revise the Roodway or a statement by the community that it has notified all offected property ownersA State approval of the revision request, if applicable Dota required for BFE revisions, if applicable Dota required for SHFA boundary revisions, If applicable ARequired when the lIoodway is Increasing or shiNing to areas previously not localed wilhin the Iloodway. ADDITIONAL DATA REQUIREMENTS FOR CONDITIONAL REVISIONS Initial fee Icontact Regional Office for current schedulel Plans 01 the proposed project For proposed projects of 5 acres or of 50 lots or greater in approximate study areas, a delailec study of the 1 OD-year flood potential is to be performed to determine the pre-project 100-year flood elevations and the effects of the proposed project on the 1 OD-year flood elevations Flood elevation increases caused by the prolect must not exceed 1 foot or the State's ollowable surcharge. Figure 1. Basic and additional data required for NFIP map revisions.

PAGE 356

FLOOD MAPPING STUDIES IN NORTHERN ITALY Luigi Natale and Fabrizio Savi Universita di Pavia Paolo Bonaldi ISMES spa Pier Giorgio Manciola Universita di Perugia Introduction In July and August of 1987, the floodplain of the Adda Valley, in the region of Italy called Valtellina, was inundated. The flood and resulting landslides inflicted casualties and caused severe damage to private and public property. In the course of the flood event a huge landslide completely obstructed the Adda riverbed near Valpola. The city of Sondrio was also seriously threatened by the flooding of the Mallero stream. The overflow was induced by significant over-
PAGE 357

Natale. Savio Bonaldi. and Manaola 333 Scope of the Study As noted in the introduction, the purpose of the study was to furnish maps of flood-prone areas of part of the Adda River and the Mallero Stream. These maps specify zones at different levels of risk and constitute the fundamental technical methodology necessary for design of structural and nonstructural measures to protect against floods. Two scenarios have been considered with respect to the mapping of flood-prone areas adjacent to the Adda: inundation by natural floods and inundation by collapse of the natural dam produced by the Valpola landslide. From the operational point of view, it is necessary to clearly distinguish the origin of catastrophic events since natural floods and dam-break waves require different techniques of mathematical simulation. Flood-risk maps for natural floods are statistically based. The maps are prepared by assigning to every section of the river a value of peak flood discharge with a given recurrence time. The propagation of the flash wave due to the collapse of the natural dike of the Valpola Lake can be analyzed in a deterministic fashion. In this case the area at risk can be identified as the area whose altitude is lower than the altitude of the wave peak. The Valmalenco study focused on a particular objective. It aimed at preparing flood-risk maps and evacuation plans for the city of Sondrio threatened by the overflow of the Mallero. The study examined thoroughly the spectrum of probable flood events. Hydrologic Studies and Flood-Simulation Models The flood planning studies required a detailed updating of the topographic surveys of the Valtellina and the Valmalenco. Air photogrammetry has been used to map previously neglected areas and to improve resolution for zones of particular significance. In the course of these surveys, 230 sections have been drawn across water bodies and accompanying structures. The preparation of flood-risk maps for natural flood scenarios required that the peak discharge for given time of recurrence be calculated for the Adda and Mallero. These calculations were performed for a sufficient number of section lines suitably distributed along the length of each water course. The study has been conducted at a regional scale in order to enlarge the data base used in the hydrologic calculations. In order to determine the discharges with assigned recurrence times, a study of floods in measured sections was first conducted. The results were then transferred to sections of interest. For the statistical analysis or maximum peak discharges in the measured sections, the exceedance method has been adopted so that all information could be completely exploited. The application of the method required an exhaustive search of original records of relevant peak discharges stored in the archive of the Italian Hydrographic Service. At the end of the hydrologic studies, the reliability of the statistical estimates has been evaluated through measuring the consistency and precision of the estimators.

PAGE 358

334 Flood Mapping in Northern Italy The delineation of areas subjected to natural flooding has been performed by assigning to every significant section of the rivers the value of peak discharge having a time of recurrence of 200 and 1,000 years. The hydraulic calculations have been performed with a steady-state, one-dimensional model with variable spatial discharge. The model takes account of obstructions presented by bridges and other river structures. By means of sediment transport calculation routine, the model also considers the extent to which over-sedimentation during flooding reduces the wetted section. The model assumes that flood waters pool in the floodplain, unless currents form outside the river channels. In the latter case it identifies the location of new channels and simulates the separation and junction of the currents. With respect to the risk scenarios having different occurrence probabilities (f =200 years and T = I ,000 years), the simulation results have been elaborated with detailed maps that can be directly used for the preparation of flood-risk plans. The results of the simulation runs have been illustrated with maps of flood-prone areas on which are indicated the average predicted velocity of the current and the predicted elevation of the free surface. Other maps show water depth and maximum elevation of submersion (Figure 1). The study distinguishes between: areas directly flooded by the overflow from the river, areas affected by floodplain currents that originated upstream, zones protected by structures but having elevation below the flood level and so being inundated when the embankment either collapses or is pierced by underpasses, zones below the flood level protected by hydraulic structures that do not guarantee an adequate freeboard (in this case, 20 cm). In order to understand the formation of flash waves downstream of the Valpola lake, five different hydraulic scenarios with variable peak discharges (ranging from about 600 m 3 /s to more than 19,000 m3/ s ) have been studied. The maximum values are approximately 50 times greater than the peak discharge of the I,OOO-year flood wave. The simulation of the downstream propagation of a flash wave has been conducted employing two distinct unsteady flow models. The first, which integrates the equations of De Saint Venant by means of a numerical scheme based on the method of characteristics, assumes a simplified geometry of the stream bed. The second, which utilizes a finite-difference scheme, resolves a simplified form of the equations with the assumption of kinematic wave propagation. In both cases the simulation of the downstream impact is accurate. The simulation is suspended at the section in which the peak discharge of the flash wave is less than the discharge of the 200-year flood. The solid transport critically influences the amount of overflow of the embank ment of the Mallero at Sondrio. Therefore, the mapping of the flood-prone areas in the city of Sondrio has been carried out by considering different situations of overbank deposition. The mathematical model utilized for the simulation of possible outcomes of flooding at Sondrio is composed of two parts: a steady-state model extended along the entire stream course of the Mallero in Sondrio and a transient model of the progress of a flash wave in an urban environment.

PAGE 359

Natale, Savi, Bonaldi, and Mandola 335 REACH OF STUDvJ Figure 1. Study area. The transient model simulates urban flooding at a scale equivalent to the length of streets. It assigns the total effects of storage to nodal points and the effects of transport to canals that connect the nodes. The canals duplicate the street network (Braschi, Gallati and Natale, 1990). Historical and Socioeconomic Investigations Upon completing the engineering studies, a historical investigation has been conducted aimed at specifying the most dangerous event, the season of the year in which this event is most probable, and the locations most often and most strongly struck by calamities. This analysis allows the evaluation of the predisposition of a particular area to suffer calamitous events. In addition to data on the physical character of calamitous events, information on the historical unfolding of floods and on the types of damage inflicted is useful to the authors of Civil Protection Plans. The information gathered through archival

PAGE 360

336 Flood Mapping in Northern Italy research has been indexed in a data bank. The data have been subjected to a statistical analysis which reveals that the severity of events has increased with time, owing to the ever-increasing development of the valleys by humans. The risk of natural calamity is not negligible in as much as eight disasters are probable every 50 years. In Valtellina, flooding constitutes the most frequent and general risk to large areas. Flooding as well as landslides are for practical purposes limited to the second half of the summer and the beginning of autumn. The socioeconomic investigation, calibrated on natural events of equal risk, has had as its objective the quantitative estimate of social goods at risk from flooding. It has employed several indicators, including the area and land use of flood-prone zones, the number of residents and the number employed, the responsive capacity of public agencies, the concentration of students of each age group, the number of livestock, and the principal activities of socioeconomic interest. Information has also been collected to organize and manage plans for spreading the alarm by means of lifelines and communication pathways. These emergency response plans indicate at least in part the key nodes and service centers called upon in the event of a flood. The survey has identified possible civilian shelters, the streets, bridges, and access points available for civil defense during periods of alarm, and evacuation plans and preferential routes for supplying aid during catastrophic events. Conclusions The effects of a chain of events that leads to a wave that overwhelms and empties a barrier lake could be catastrophic for no more than 20 km downstream from the collapsed dam. Further downstream the effects of natural floods are worse than those of dam break floods, consequently no provisions for evacuation are recommended. The evacuation procedures of the civil protection plan, drafted in the period immediately following the formation of the Valpola Lake, can be considered obsolete inasmuch as efforts to consolidate the landslide material render most unlikely the hydraulic scenarios that led to the evaluation of very high flash waves generated by the lake. The study indicates that the major consequences to be feared from natural flooding are economic in nature. The risk to human life can be reduced to very low values by means of moderate precautions. The study lists engineering structures (essentially dikes in the proximity of houses and roads) and other nonstructural measures needed to mitigate possible economic damages. References Braschi G., M. Gallati, and L. Natale 1990 "La simulazione dell'inondazioni in ambiente urbano." Rapporto CNR-GNDCI, Linea 3, Genova.

PAGE 361

Natale, Sa vi, Bonaldi, and Mandola 337 Giglioni G., and P.G. Manciola 1987 "Azioni non strutturali per la mitigazione degli effetti degli eventi estremi: copertura assicurativa del rischio." Rapporto CNRGNDel, Linea 3, Genova. Regione Lombardia 1988 Piani di allarme connessi alia situazione di rischio idro meteorologico della Val di Pola. CAE-ISMES, Milano.

PAGE 362

CHANNEL MIGRATION ON THE TOLT AND RAGING RIVERS, KING COUNTY, WASHINGTON Susan 1. Perkins Shannon and Wilson Introduction The Tolt and Raging rivers are rapidly migrating, gravel-bedded rivers that drain the western foothills of the Cascade Range in King County, Washington. Data on historic rates and patterns of channel migration on the rivers were used to estimate probable future limits of channel migration and to develop maps showing channel migration hazard zones. Description of the Study Areas The Tolt and Raging rivers study areas extend six and eight miles, respectively, upstream from their mouths on the Snoqualmie River. Both rivers flow through steeply sloping, forested, V-shaped Valleys in their headwaters. Within the study areas, channel gradients are lower and the river valleys broaden to widths of 800 to 2,400 feet. The rivers flow primarily in gravelly alluvium, but till and other glacial sediments are exposed in the banks where the rivers abut terraces or valley walls. The Tolt River drains an area of 101 square miles, and channel gradients within the study area decrease downstream from 0.7% to 0.4%. The Raging River drains a 33square-mile basin and study area channel gradients range from 0.9% to 1.6%. Methods Features recorded during field studies included levees and revetments, eroding banks, floodplain channels, depositional zones, river bank height and composition, vegetation type and age, and descriptions of river and floodplain morphology. Tracings of historic river channel positions from aerial photographs and maps dating from 1936 to 1991 were superimposed, using a projector equipped with an adjustable focal length, to produce base maps such as the one shown in Figure 1. Each river was divided into reaches based on type and rate of channel migration. Historic rates of channel migration were calculated for each reach by dividing the distances between successive river positions by the elapsed time between positions. Historic Rates and Patterns of Channel Migration Channel migration rates have varied dramatically during the past 55 years (fable 1). The highest rates generally correspond to periods with moderate to large floods (return periods of 10 to greater than 60 years). However, formation and rapid growth of an avulsion channel in Reach C of the Tolt River occurred between 1983 and 1989, a period when the largest flood had a return period of six years or less. The highest measured rates on the Raging River occurred between 1985 and 1991,

PAGE 363

Perkins LEGEND 1936 1964 _._ @ -1970 River Mile .----:.:..l:.= 1970 Valley Wall o 800 1600 H H H H I I Scale in Feet Figure 1. Mapped 1936-1970 river positions for Reach C of the Tolt River Study Area. Downstream is to the left. o 800 1600 H H H H I I HAZARD CLASS Scale in Feet Extreme c=J Low (mH@@U High HI!;:'::!] Potential Landslide Hazard Lil Moderate Protected by Revetment Figure 2. Channel Migration Hazard Map for Reach C of the Toll River 339

PAGE 364

340 Reach A B C o E Reach A B C 0 E F G *1958-85 Channel Migration Table 1 Average Channel Migration Rates (ftjyr) ToR River 1936-1964 1964-1977 1977-1989 1989-1991 0 0 0 0 7.54 0.28 2.66 67.40 5.8 1.62 7.29 54.60 3.43 0.96 0.72 4.25 0.98 0.13 0 0 Raging River 1936-64 1964-85 1985-91 0 0 0 0.33 0.29 0.83 0.26 1.46 4.39 1.11 2.21 2.06 2.57 1.48 10.00 0.92* 4.10 0.09* 1.47 a period that included the four largest floods of the 1946-91 period of record. Severe channel migration also occurred in the large 1932 flood, based on historical accounts and the appearance of the Raging River in the 1936 aerial photographs. Rapid channel migration on the Tolt River occurred during the 1959 flood (the largest in the 62-year record), and the November 24, 1990, flood (the sixth largest flood of record). These periods of rapid channel migration corresponded to increases in reach averaged active channel widths of up to 77% and formation of multiple channels in some reaches. The intervening periods of slow or no migration corresponded to decreases in channel width and a return to a single-thread channel (e.g., compare the Tolt River 1964 and 1970 channels in Figure 1). Many of the houses along the banks of the two rivers were built during the period of relative quiescence prior to the late 1980s.

PAGE 365

Perkins 341 Historic channel migration rates vary greatly between river reaches (Table 1). Levees have successfully prevented channel migration in the downstream reach of each river (Reach A) since about 1940. Channel migration rates generally decrease upstream on the Tolt River, but are highest in the central portion of the Raging River study area. Differences between reaches reflect patterns of shear stress and sediment deposition, erodibility of bank materials, destabilization by upstream channel changes, and the type of channel migration. The highest channel migration rates occurred in Reaches Band C of the Tolt River, where new channels formed abruptly by avulsions and then widened rapidly. Sites where avulsions are likely to occur, such as floodplain channels, exist in most reaches of the rivers and were mapped for use in predicting sites of future channel migration. Predictions of Hazards From Future Channel Migration The historic channel migration rates were used to make conservative predictions of the probable future limits of channel migration over both lO-year and l00-year time periods. Based on these predictions, land in the study areas was classified and mapped according to its relative level of hazard from channel migration (Figure 2). To predict channel migration distances, the historic average annual migration rate for a particular reach was multiplied by the number of years in the prediction period. For the lO-year prediction, morphologic criteria were used to identify areas likely to erode. Relatively high, short-term historic rates of erosion calculated for eroding areas (i.e., noneroding areas were excluded) were applied to these areas. For the I OO-yearprediction, average long-term historic migration rates (including noneroding areas) were applied throughout each reach. Avulsions were assumed to occur at potential avulsion sites, then grow laterally at the higher historic rates calculated for avulsion channels. Modifications were made to these procedures where the migrating river encountered a valley wall or high terrace. Conclusions Historic channel migration rates on the Tolt and Raging rivers have varied dramatically over time, with decades-long periods of relative stability interspersed with periods of rapid channel migration. Probable limits of future channel migration were estimated based upon historic migration rates and river morphology. These predictions provide a quantitative basis for classifying land in these river Valleys according to its relative level of hazard from channel migration.

PAGE 366

AUTOMATED DIGITAL LINE GRAPH DATA CAPTURE PROCEDURES Iohn M. Taylor Michael Baker, Ir., Inc. Background On Digital Mapping Since its inception in 1968, a primary emphasis of the National Flood Insurance Program (NFIP) has been to assess the flood risk within flood-prone communities nationwide. The results of these assessments, which have been conducted for more than 18,000 communities, are presented on Flood Insurance Rate Maps (FIRMs) and in collateral Flood Insurance Study reports. To improve service to map users, the Federal Emergency Management Agency (FEMA) investigated Geographic Information Systems (GISs) that could be used to automate the compilation of FIRMs. A FEMA report entitled Flood Hazard Mapping Geographic Information System Pilot Project (FEMA, 1986) presents details of the investigation. FEMA is using GIS technology to convert information from FIRMs to a digital format. One application of digitized information is the development of Flood Risk Directories (FRlDs) which provide flood risk information in a directory or telephone book format. FRIDs are prepared by combining digital FIRM data with digital street address information and will enable users to make approximate determinations of which street addresses are subject to flood risk without consulting hard-copy FIRMs. Selecting a Digital Data Format and Processing System Under FEMA's established procedures, all NFIP mapping was to be prepared manually. Therefore, before any digital data could be created, a standard digital format had to be identified. To ensure that the digital FIRMs would be accessible nationwide, the U.S. Geological Survey (USGS) digital format standard known as Digital Line Graph Level 3 (DLG-3) was selected. Additional information on the DLG-3 format is provided in the USGS publication entitled Digital Line Graphs From I:24,OOO-Scale Maps (U.S. Department of the Interior, Geological Survey, 1986). The Data Capture Process Figure 1 presents the steps involved in the data capture process. These steps, accomplished at a CADD work station, can be divided into three major processing categories: 1. Information Gathering and Preparation 2. Digitization and File Creation 3. Quality Control (QC)

PAGE 367

InformBtion Glltherlng lind Prepllrlltion Process Q' ... ObtBln ObtBln Crellte Estllbllsh Prepare FIRM USGS PBnel CoordlnBte FIRM 0 PBnels QUlldrllngles Index System Pllnels for 0 0 0 (;) OIgltizlltio OIgltizlltion lind File CreBtion Process Digitize Crellte Cop, H PI". H .d'q' H .... FIRM USGS DlltB to QUlldrllngle Zone LBbel Qulldrllngle Pllnels QUBdrBngle QUlldrllngle Nelltllnes Text Corner Flies 0 Flies 0 0 e Text e II II QUBlIty Control Process Execute Execute Perform No Output --+I LINECHECK -+ COMPLEXER DLG VlsulIl DLG Proce911 Process QUBllty QUlIlIty File e e Control Control e Process Check .. o Plpre I Datil Capture PI'OCeIIS Plowchart e

PAGE 368

344 Digital Line Graph Data Capture Information gathering and preparation. The following activities are performed during this first step: 1. FIRM panels and USGS quadrangles are obtained. 2. Panel Index is created. 3. Coordinate system is established. 4. FIRM panels are prepared for digitization. Digitization and file creation. FIRMs are digitized panel by panel. The graphics operator starts by collecting aIllinework features (i.e., corporate limits, floodplain boundaries, limits of study, base flood elevation (BFE) lines) and then collects the labeling. All items are to be digitized using a customized menu (Figure 2), which represents a series of user commands to initialize internal element features. Each user command consists of a series of steps that is performed each time a map FEMA FLOOD STUDY DIGITIZING MENU lIAS( /tjJ.p IICl.tIIoIIfr nAllJI[$ II)Ij! 'EXT []I] OO-P UIS V 1, D ..... NE.tTUtS LAXE:S BIJlH)t.RT AE w:; .til .t99 I x Not flOOD X RIVERS WRlCER ELEVATlDN OOUNDARY "r. hcL flGod -. Flow -..... E".." B 00 I lOl1 I IlTALBI 00"''''' MARl( STUI)'( I ROTATE I SIflIII( I TEXT V, TEXT I I t (Ri. I I IF LOCATION S11DT /H) COllI< ..,.,. [E] IIUAD 1.1.\1' fEA TLIIfS []i] ] n""""" R."",",,' ...,"""" NEAnHS TOT LEVELS I I UPUN1 I llWlT Of R.1lCDOIG I lUI' f'Q/ocAL I C01S1AL I R.[)()[H; AREA ""' ..,., Figure 1. Customized digitizing menu.

PAGE 369

Taylor 345 feature is collected. By selecting the appropriate user command from the menu, the user defines the following five attributes: 1) level, 2) annotation, 3) color, 4) line code and 5) line weight. The unique combination of these attributes represents a single feature on the FIRM. Quality control. The following four QC checks (three automated and one manual) are performed to ensure the accuracy of the linework and labeling within each polygon: 1. Initial Linework Check 2. Second Digital Check 3. Final Digital Check 4. Visual Check The digital QC subroutine LINE CHECK, a batch process, is executed first. LINE CHECK locates all free endpoints, all lines that are crossing without endpoints, and any duplicate lines, and places a small circle at the offending location. The second check is for duplicate labels within a given polygon area, errors in closure on the polygons, and errors in legal area names and BFE labels. (For example, a Zone X area cannot have an elevation associated with it, whereas a Zone AE area must have an elevation.) If errors are identified, the CENTROID COMPLEXER will place an error message in the file at the text string location. After all errors are identified and addressed, a final digital QC check is performed using the DLG OUT routine. If an area without a label is located, the DLG OUT routine places an indicator at the minimum x, minimum y location for that area. For the final, visual QC check, a paper plot of each panel is made from the quadrangle files that have been used throughout the digital QC process. The plots are overlayed on the mylar copy of the FIRM panel and compared using a light table. The Final Digital Product The format of the digital FIRM files that are to be delivered is defined in National Flood Insurance Program, Standards for Digital Flood Insurance Rate Maps (Federal Emergency Management Agency, 1990). During the output process, four DLG files are to be created. Each DLG file contains the features collected for one of the following categories: 1. Flood Zones 2. Political Areas 3. Map Panel Areas 4. FEMA Hydrography

PAGE 370

346 Digital Line Graph Data Capture The contents of each file are described in greater detail in the above-mentioned digital FIRM standards. Control of each output file is in the DLG OUT parameter file. References Federal Emergency Management Agency 1986 Flood Hazard Mapping Geographic Information System Pilot Project. September 30. Federal Emergency Management Agency, Federal Insurance Administration 1990 Standards for Digital Flood Insurance Rate Maps. Revised February. u.S. Department of the Interior, Geological Survey 1986 Data Users Guide I: Digital Line Graphs From 1:24,OOO-Scale Maps.

PAGE 371

Part Eleven New Techniques in Ice Jam Control

PAGE 372

This page IS intentionally blank

PAGE 373

ISRAEL RIVER ICE CONTROL STRUCTURE Kathleen D. Axelson U.S. Army Cold Regions Research and Engineering Laboratory Introduction Ice jam-related flooding along the Israel River in Lancaster, New Hampshire, occurred on an almost annual basis during the period from 1950 to 1979. Serious flooding in March 1968, resulting from a break-up ice jam event, prompted the town to seek a method of alleviating the ice jam flooding. The U.S. Army Corps of Engineers, New England Division, and the Cold Regions Research and Engineering Laboratory were involved in the design of an ice control project at Lancaster. The purpose of the project was to control the production and transport of frazil ice, thereby decreasing the volume of ice available to feed ice jams downstream. While minor flooding caused by backed up drains and shallow overbank flow along a low-lying area has occurred since the completion of this innovative, inexpensive project in September 1981, there has been no major or damaging flooding. This paper reviews the performance of the project and assesses its impact on ice jam flooding in Lancaster. Site Characteristics The Israel River is approximately 21 miles long and has a drainage area of about 136 square miles at its confluence with the Connecticut River in Lancaster, New Hampshire. A mildly sloping reach often affected by backwater from the Connecticut River extends about 1.5 miles upstream from the confluence of the two rivers. Otherwise, the river is generally shallow and relatively steep, with a rough bed. The drainage basin is mostly forested and has little development. A small amount of runoff storage is provided by marshy areas. To determine peak flows from this ungaged river basin, direct drainage area transposition from a neighboring gaged watershed was used. Daily maximum and minimum temperatures for December, January, and February were averaged for each year available. No correlation was found between the season-averaged temperature and the occurrence of damaging ice jam floods in Lancaster. Ice Jam Flood History The town of Lancaster has experienced a number of floods resulting from break up ice jams on the Israel River, caused ultimately by frazil ice. The steep slope and turbulence of the Israel River, combined with the low winter temperatures, are ideal conditions for the formation of large quantities of frazil ice. The ice is transported downstream until it reaches the ice-covered backwater reach, where it tends to deposit beneath the ice cover. The deposited frazil ice increases the thickness and strength of the ice cover in this location, so that large forces are necessary to break

PAGE 374

350 Israel River Ice Control up the ice. Ice jams are caused when broken-up ice, transported along the river, reaches the strong, intact ice cover formed in the backwater. The ice begins jamming at the upper end of the backwater reach, about 1.5 miles upstream from the confluence with the Connecticut River, and progresses upstream toward the center of town (2,800 feet). The major flood damages occur in the area of the Main Street bridge. Few ice jam floods on the Israel River are reported prior to 1936. Ice jam flooding appears to have increased in frequency and severity after the removal of four dams on the river sometime before 1950. It is probable that the pools formed by these dams provided frazil ice storage and decreased the downstream transport of frazil ice. Sheet ice cover growth on these pools also decreased the production of frazil ice. Therefore, the frazil ice accumulation in the backwater reach was decreased. The overall result was that the thickness and strength of the ice cover in the backwater reach were significantly less than after the dams were removed. Known ice jam floods and high water marks since 1950 are shown in Table 1 (USACE, 1973; 1980). The Detailed Project Report (USACE, 1973) notes that 15 ice jam floods occurred during the period from 1940 to 1973. As is often the case, the ice-related flooding at Lancaster is characterized by higher stages at lower discharges than for open water floods. The ice jam flood of record, described in detail by Frankenstein and Assur (1972), occurred in March 1968. The peak stage was about 866.6 ft NGVD, and the estimated discharge was 2,600 cfs. In compari son, the open water flood of record, which occurred in 1927, had a stage of 863.5 NGVD associated with an estimated peak discharge of 8,840 cfs. Date April 1950 February 1953 April 1960 March 1964 March 1968 March 1970 February 1973 March 1974 March 1977 January 1978 Table 1 Known Damaging Ice Jam Floods High Water Mark (ft NGVD) (if available) 863.8 858.7 859.4 859.5 866.6 863.6 NjA NjA 6" in Police Station NjA Estimated Discharge (cfs) 3570 700 3180 2050 2640 3600 1240 1550 3090 2330

PAGE 375

Axelson 351 Ice jams have often formed at Lancaster with no known flooding, generally when there has been an early season break-up of relatively thin ice. The most damaging ice jam flood events have occurred when an early season break-up jam has been followed by a period of cold weather, freezing the jam in place. The frozen-in-place jam would require large forces and therefore high water levels to cause it to break up and move. The combination of open water reaches and low midwinter tempera tures causes the formation of substantial amounts of frazil ice. This frazil ice further thickens the ice cover in the backwater reach. Continued ice growth in the river also increases the volume of ice available to jam later in the season. Ice Control Project The Israel River flood control project includes a 160-foot long, 9-foot high ice control structure (ICS) located about 2,500 feet upstream from the Main Street Bridge (US Routes 2 and 3). The ICS (Figure 1) is an earth and rockfill embankment protected by a layer of gab ion mattresses and a three-inch thick concrete cap. Four concrete sluiceways, each four feet wide by 7.5 feet deep, allow fish passage during the summer and fall (USACE, 1973; 1982). The project also includes the seasonal placement of a submarine net across the river about one mile upstream from the ICS. The submarine net, which has been in use since 1968, collects ice pieces from the upper river, thus decreasing the ice available to the jam. The Israel River ICS is located approximately 2.5 miles upstream from the confluence with the Connecticut River and one mile upstream from the mildly sloping backwater reach. Figure 1. Israel River Ice control structure.

PAGE 376

352 Israel River Ice Control Effectiveness of the Ice Control Project While damaging ice jams were almost an annual event on the Israel River prior to the completion of the Israel River Ice Control Structure in September 1981, only minor flooding incidents have occurred since the completion of the ICS. The cases involving high water levels since the construction of the ICS (1983-84 and 1989-90) were associated with early jams that froze in place, followed by further ice growth and later jam formation, except the January 1991 event. In this event, an early season break-up jam coincided with high levels on the Connecticut River. Differences in climatic, hydrologic, and operating conditions during the first nine years of the project's operation make it difficult to quantitatively determine its effect on ice jam flooding at Lancaster. The construction of a new Main Street Bridge in 1990 also affects river hydrology and ice jam characteristics. However, a comparison between two similar winter seasons can be used as a qualitative measure of the effectiveness of the ice control project. The 1989-90 winter season was similar to 1967-08 in that the early season temperatures were lower than normal and a runoff event caused an early season ice jam that then froze in place in Lancaster. During both years, continued low temperatures induced further ice growth on the river. The discharge during the March 1990 break-up event, estimated to be 5,470 cfs, was larger than that experienced in March 1968 (2,640 cfs). Even with a larger discharge, no flood damages resulted from the ice jam of March 1990, while the flood of March 1968 remains the flood of record. This qualitative comparison indicates that the ICS, in combination with the submarine net, has had a positive impact on ice jam flooding in Lancaster. References U.S. Army Corps of Engineers (USACE) 1973 Lancaster Local Flood Protection-Detailed Project Report. New England Division, Waltham, Massachusetts. 1980 Historical Ice Jam Flooding in Maine, New Hampshire and Vermont. Section 206 Report, New England Division, Waltham, Massachusetts. 1982 Lancaster, New Hampshire, Israel River Operation and Maintenance Manual. New England Division, Waltham, Massa chusetts. Frankenstein, G., and A. Assur 1972 "Israel River Ice Jam." Proceedings, lAHR Ice Symposium, p. 153-157.

PAGE 377

ICE JAM FLOOD FREQUENCY ANALYSIS TECHNIQUES Jon E. Zufelt and James L. Wuebben U.S. Army Cold Regions Research and Engineering Laboratory Introduction Flood frequency analyses are used to predict the likelihood of flooding events and are useful in estimating the benefit-to-cost ratio of proposed flood control projects. An open water flood frequency analysis estimates the probability distribution of the magnitude of discharge in the river in question. Once this is obtained, the stages associated with these discharges can be calculated with a rating curve or with the assistance of numerical backwater models. Equating stages to levels of damage results in a damage-frequency curve that can then be used for calculating the average annual damages and benefits of a particular project. During the winter, however, an ice cover, accumulation, or jam can form on the surface of a river, restricting flow and raising water levels. The stage experienced during an ice cover or jam can be much higher than that associated with the same discharge during open water periods. Ice jams are also very site specific; they may only influence the stage over a short reach. For an open water flooding event, the effects of a flood discharge can be easily determined over long reaches of the river simply by referring to the rating curve of the location desired. For an ice jam event, however, the stage depends on the location, extent, and thickness of the jam as well as the discharge. For these reasons it may be necessary to calculate the stage frequency distribution at several locations within the study reach. Necessary Data for Analysis As stated above, the desired output of a flood frequency analysis is a single stage-frequency distribution that covers both open water and ice jamming periods. When the dominant cause of the annual peak stages is ice jams, a mixed-population frequency curve is developed in which the open water and ice jam annual peak stages are analyzed together. For most rivers, however, flooding events are usually divided into two separate populations: open water and ice jams. In this case, a combined population frequency curve is derived from the frequency curves developed from the two separate populations. There are two methods for determining the ice jam stage-frequency distributions, both of which are described in the next section. The direct method relies on the historic stage data available for the location of interest. The indirect method is used when little or no stage information exists for a site and must therefore be synthesized from other information that may be available. This information may include the discharge in the river or in similar nearby river basins and stages upstream or downstream from the ice jam, as well as ice jam location, extent, thickness, and roughness.

PAGE 378

354 Ice Jam Flood Frequency Analysis There are many sources of historic data, some of which are usually overlooked because they may be incomplete or unreliable. Ice jam data are often incomplete or nonexistent, and it is usually necessary to resort to some of the more unreliable data sources to obtain any information at all. Stage and discharge data are typically obtained from a U.S. Geological Survey (USGS) gaging station and are highly reliable for open water periods. As mentioned above, ice jam stages are highly site specific, and it is rare that a gaging station is located within the study reach. Even when one is fortuitously located, there are several reasons for the data to be suspect. The gage can be frozen or damaged by moving ice. The rating curve may not be corrected for the effects of ice jams, resulting in reported discharges being too high. The timing of the gage data is also very important. Most discharge data are given as average daily values. The high stages of an ice jam event can be very short in duration, possibly only a couple of hours. During a break -up ice jam, the discharge is constantly increasing, which can result in two peak stages, one associated with the ice jam, and a second (up to a few days later) associated with the actual peak discharge following the failure of the jam. These two peak stages are in different populations and must be treated as such. It can also be difficult to determine the discharge associated with the peak stage due to the ice jam. Other data sources that are highly reliable include the reports of ice jamming events filed by federal, state, and local water resources agencies, as well as those of civil defense organizations (USACE, 1991). Prior flood insurance studies may contain valuable information on peak stages or discharges. Somewhat less reliable are historic photos or newspaper articles. Town reports or museums may contain information on the more dramatic events. Interviews with local residents can provide some useful information if the events in question are not too distant in time. In sparsely populated areas, one may have to resort to such sources as tree scars, vegetation trim lines, or structural damage in determining stage levels. Direct Analysis Method If a reliable record of ice jam stages exists at a location, the calculation of the frequency distribution by the direct method is fairly easy. The annual ice jam peaks are tabulated by rank, and the plotting position is calculated using the Weibull formula: P = m/(N+ 1) where P is the exceedance probability of the event of rank m, and N is the number of events. In cases where there are years in the record for which there was no ice jam, N is replaced by the total number of years of continuous stage record. The resulting points are plotted on probability paper, and the best-fit line is drawn by eye, rather than by using regression equations, due to the usually small number of data points. There is a chance that over the period of the stage record, several data sources exist. Good hydrometric stage data for winter periods are usually limited and could be supplemented by other data such as historic photos or news accounts. A method

PAGE 379

Zlifelt and Wuebben 355 of analyzing the stage data of multiple sources of varying length of record and varying reliability is given by Gerard and Karpuk (1979). Indirect Analysis Method When no historical record of the ice jam stages exists at a location, it is necessary to use the indirect method of calculating a frequency distribution. The ice jam stages are calculated using estimates of the discharge and ice conditions. From the record of a nearby gaging station, the historical discharges for the ice jamming period are tabulated along with any information concerning the ice jam itself, such as location, extent, thickness, and roughness. In the absence of a gaging station near the ice jam location, estimates of the discharge history can be made from other gaging stations along the river or in similar river basins as long as they are not affected by the backwater from ice jams. Care must also be taken to obtain the discharge values during the ice jam event and not those following the failure of the jam (which are usually higher). The historic ice jam data will provide information on the location of the toe of the jam and its extent upstream, which is necessary for properly synthesizing the ice-affected stages. Several of the tabulated ice jam discharges are selected and the ice jam stages calculated using an ice-affected backwater model (such as HEC-2 with the ice cover option). The stages are then plotted on probability paper as in the direct method. In the event that no information on past ice jams exists, stages can be calculated for two extreme cases of ice cover formation, smooth ice covers, and fully developed ice jams throughout the study reach. The smooth ice cover represents the lower bound of stages during periods of ice, while the fully developed jam represents the highest possible stages. This results in two stage-frequency curves representing the upper and lower bounds of ice conditions. The actual stage-frequency distribution usually lies between these two curves. It may be possible to further define the ice jam stage frequency distribution. Below a certain threshold discharge, break-up (and thus ice jamming) does not occur. Similarly, above a certain discharge, an ice jam physically cannot remain in place. The identification of these two additional discharges from historic data allows further refinement of the ice jam stage-frequency distribution. Combined Frequencies Once the ice jam stage-frequency distribution is developed, it can be combined with the open water stage-frequency distribution rather easily. Several stages are chosen and tabulated, along with their exceedance probability, for both the ice jam and open water conditions. The exceedance probability of the combined-population frequency curve for a given stage is thus calculated as Pc = Pi + Po (Pi) (PO> where Pi and Po are the exceedance probabilities of the given stage for ice jam and open water conditions.

PAGE 380

356 Ice Jam Flood Frequency Analysis Summary A method was presented for developing the combined stage-frequency distri bution for ice jam and open water events. Two methods of calculating the ice jam stage-frequency distribution were described. The direct method is used when a reliable record of historic ice jam stages is available. The indirect method is used when no stage record exists and the historic stages must be synthesized. Possible data sources were identified as well as sources of data error. References U.S. Army Corps of Engineers (USACE) 1991 Ice-Influenced Flood Stage Frequency Analysis Engineering. Technical Letter No. 1110-2-321, Washington, D.C.: Department of the Army. Gerard, R.L., and E.W. Karpuk 1979 "Probability Analysis of Historic Flood Data." Journal of the Hydraulics Division, ASCE, No. 105(HY9) pp. 1153-1165.

PAGE 381

Part Twelve Lessons from Recent Floods

PAGE 382

This page IS intentionally blank

PAGE 383

LESSONS FROM THE 1990 TAIWAN TYPHOON SEASON Spenser W. Havlick Boulder City Council and University of Colorado at Boulder The island of Taiwan is the second most densely populated area in the world, with 564 people per square kilometer, which is topped only by Bangladesh, with a population density of 616. In 1990 six fierce typhoons struck Taiwan. Superkiller Typhoon Dot made landfall September 6, 7, and 8 and ravaged villages and infrastructure with floods, landslides, and storm surges, especially along the east central portion of the country. With expected increasing frequency and magnitude of tropical cycles or typhoons, dramatic lessons become available from the Taiwan experience for other rapidly developing countries. Third world nations face greater risks than Taiwan due to insufficient capital and hazard information (Havlick, 1986). The 1990 Taiwan typhoon season was the most severe in the memory of villagers and officials of this rapidly industrializing and urbanizing nation. The lack of floodplain mapping and the absence of geologic overlay zones or an appropriate postdisaster reconstruction model serve as agonizing reminders that risk reduction must parallel economic development in areas of frequent hazard events. Taiwan, with its frequent earthquake, landslide, debris flow, mountain flash flood, and tsunami events, would logically be in a state-of-the-art condition to prevent loss of life and property from floods and other natural disasters. This presentation with graphics and interview narrative shows that without enforcement of proper flood hazard reduction measures, a popUlation is placed at considerable risk. Aboriginal villagers who were interviewed by the author could not understand why the government of Taiwan forced them out of their high mountain subsistence living only to be placed in government housing in high flood hazard areas. The survivors from the eastern Taiwan towns of Tungmen and Hoping stop in disbelief and anguish to realize that the dozens of lives and hundreds of homes that were lost could have been prevented (Iu, 1990). The structural mitigation measures of sea walls, levees, check dams, flood diversion channels, and other flood-proofing strategies were proven to be most inadequate (Havlick, 1991). Although a sophisticated typhoon warning system has been developed to alert urban residents who have radio, TV, or newspapers, the notification in rural areas is ineffective. Furthermore, land use practices throughout, including floodplain development, are putting large numbers of people at risk in the event of future typhoons and subsequent flooding. Lessons from the 1990 Taiwan typhoon season come in several forms for floodplain managers and others interested in flood damage mitigation. High marks can be given for the typhoon and subsequent flood warning systems that are in place and used judiciously by citizens and government officials in large urban centers. For example, typhoons that spawn in South Pacific waters warmer than 26C sweep northwestward from the Philippine Islands and Okinawa and head toward Thailand, Hong Kong, or Taiwan before dissipating their energy as they come upon land masses in mainland China or Japan.

PAGE 384

360 Lessons from the 1990 Taiwan Typhoon Season The path, magnitude, velocity, and approximate landfall times are carefully given in the news media at least four to five days prior to the major event (China Post, 1990). Continued updates of the storm track are provided via television, radio, and daily newspapers. The mountainous and hinterland populations are not notified with the same thoroughness as major urban centers. Communication links are not as well established in the central and southern portions of Taiwan. That is why villages like Tungmen and Hoping in Hualien County of east-central Taiwan were caught unprepared for the floods produced by Typhoon Ophelia and Typhoon Dot. In Tungmen the singular rod of homes and markets was totally torn asunder by the raging boulder-filled flood waters. Eighty-three deaths were reported. Water supply, roads, utilities, and shelter had not been restored three months after the typhoon hit the village. Then, to compound the tragedy of this event, it was learned that over the years government policy was to force these aboriginal tribes out of their high mountain subsistence living and house them in towns accessible to modern amenities like schools, labor markets, transportation, and conveniences. However, apparently no one bothered to determine the high hazard floodways and villages. Tungmen, Hoping, and many others were constructed in the center of floodplains. It was only a matter of time before a large typhoon would produce the fatal flooding, landslides, debris flows, and destruction-1990 was that time. I saw grave sites where 12 out of 14 family members were killed in a typhoon-triggered mudslide. East-central Taiwan was essentially cut off from the rest of the country for several weeks. In addition to senseless building in the high hazard flood zones, there seems to be no restrictions enforced about removal of native vegetation from steep slopes in river valley regions. There is a very noticeable conversion from the semitropical rainforest vegetation to cash crops. The euphemism "agriforestry" is used to describe the clearcutting of the indigenous species. Bamboo (for the edible tips and construction uses), betel nut (for use as a narcotic chewing gum), mango, papaya, etc., are seen growing on very steep hillsides. Their inability to retain heavy runoff increases soil erosion, which in turn triggers landslides, rockslides, and devastating debris flows (Chang, 1989). Unregulated mining practices for limestone and cement production, ferrous alloy metals, coal, and gravel create aggravated scarring of the eastern Taiwan mountain landscape after heavy rains. If mine reclamation practices are recognized, there is certainly no evidence of application. The sediment load in the flooded rivers of Taiwan could only be believed by visual documentation or personal observation. Stream channels were silt-filled early in the typhoon season in 1990. As the fourth, fifth, and sixth typhoons occurred there was little channel capacity left in central Taiwan. Consequently, I observed streams that were at least 15 to 18 meters (45 to 55 feet) above the normal flow stream level at flood stage. Alluvial fans at the river discharge points covered tens of hectares with newly deposited gravel, silt, and organic debris from inland watersheds. Railroads, bridges, and highways that parallel the eastern coastline were quickly severed by the torrents flowing from the mountains directly to the Pacific Ocean.

PAGE 385

Havlick 361 Taroko National Park, located 26 kilometers north of Hualien, was another area hit hard by Typhoon Dot. Several hundred road washouts occurred within the park boundary. The police station and former park headquarters were demolished by a landslide. The hot springs above Tienhsiang were buried and the hydropower dam at Chipan was overtopped, tearing out the only east-west highway at the same time. About 30 kilometers north of Taroko Gorge (failuko Hsia), the Hoping River at flood stage changed its channel, cut into roadside cliffs, and isolated interior villages. The downtown streets of Hualien, the large east-central metropolitan area of 350,000, were over four meters (13 feet) deep in typhoon-driven flood waters. To the south between Shoufeng and Lintien, the east coast express railroad track tunnel was buried by flood silt and streambed boulders. Ironically, the railroad track ran underground via a "subfloodplain" tunnel. The railroad tunnel was constructed because the former track, which crossed the alluvial fan on bridges, was knocked out almost annually. As a strategy to avoid flood disruptions to the north-south main line (express train service), the track was placed under the floodplain. The river channel was widened, deepened, and straightened. Furthermore, substantial rock levees exceeding 10 meters were constructed to contain any future discharge from flooding the new tunnel and track. During the 1990 typhoon episodes, the levee was breached in several places and the flash flood waters proceeded to cut away federal highway number 9, scour hundreds of hectares of prime agricultural cropland, and then flood the railroad tunnel, filling it with typical floodplain debris. Needless to say, the track right-of-way was out of service for several weeks until the excavation and cleanup had been accomplished. The lessons from the 1990 Taiwan typhoon season may sound like a stale old lecture from a natural hazard handbook (or a state floodplain manager's manual). The structural approaches to flash flood events may cope with floods of small magnitude-lO-, 25-, or even 30-year events. Dikes, levees, dams, spillways, stream channelization, even tunnels under the high hazard zone may work some of the time. However, when Taiwan received six to eight typhoons with winds hitting 165 kmlhour and 10to l5-inch rainfalls, the ability of the engineering solutions to provide protection was exceeded in almost every situation. Typhoons Marian, Ophelia, Percy, Yancy, Abe, and Dot produced a cumulative impact of catastrophe between May 19 and September 10. The instruction from this especially powerful onslaught (the annual average is 3.5 typhoons per year, according to Taiwan's Central Weather Bureau) is that nationwide flood zone mapping must be completed as soon as possible and the enforcement of no new construction of high-human occupancy buildings must be upheld. Instead of bringing aboriginal people onto high hazard flood zones, disincentives and incentives must be devised to empty the high risk areas of human occupancy. Mountainside agricultural practices, roads, and other infrastructure must not be placed on slopes greater than 15% or 20%, depending on soil conditions and vegetative cover. Warning systems must be improved for rural poor and remote populations. Floodproofing should be considered only as a last resort option where removal of structures such as temples, shrines, graves, or other human artifacts is

PAGE 386

362 Lessons from the 1990 Taiwan Typhoon Season not feasible-or not "auspicious," as the Taiwanese would say. With the rapid urbanization of sprawling cities like Taipei (5.5 million), Kaohsiung (1.7 million), Taichung (800,000), and Tainan (400,000), natural drainageways, unstable slopes, and major high hazard risk zones must be mapped and future development curtailed. The tendency is to try to provide engineering works or structural modifications to withstand future floods or slope failures. The lesson of the 1990 typhoons showed this approach is inadequate. In the second most densely populated county in the world, with the capital city at 10,160 people per square kilometer (New York is 9,050 and Tokyo is 5,308), extraordinary efforts seem justified to safeguard lives and property in future typhoons. Chang (1988) has shown that in the Taipei basin from 1950 to 1984,53% of the floods were caused by typhoons and that from 1968-1986,68% of the serious slope failures were touched off during the typhoon season. The historical tracks of Taiwan's typhoons have been mapped by Chen (1981) between 1897 and 1979. Eighty typhoons (31 % during the study period) swept across southern Taiwan, heading northwesterly over Kaohsiung and Tainan. The second most frequent historical path (80 times during the same 82-year period) was in northeastern Taiwan, hitting Keelung and Taipei 27% of the time. This northern tip of Taiwan contains 43 % of Taiwan's 21 million people, according to the 1990 census. Even the most conscientious floodplain managers, equipped with state-of-the-art flood reduction tools, are helpless without an informed citizenry to urge the political forces into the newer nonstructural methods of flood hazard management. What we have not yet fully implemented in the West seems to be still further from being learned in Taiwan. This is all the more tragic when one considers a) what segment of the population is at greatest risk, and b) how global warming promises to intensify the magnitude, severity, and frequency of typhoons given birth in the already warm 26 C (78 oF) oceanic waters of southern latitudes. The megakiller cyclone (typhoon) which struck Bangladesh in the spring of 1991 with 139,000 deaths is a harbinger of events to come to the Pacific Rim countries in the years ahead. In Chinese literature there is a proverb that says it takes 10 years to plant a tree but it takes 100 years to educate a person (Chyi, 1991). It appears that Taiwan does not have 100 years to take the necessary corrective actions that would produce a liveable human environment for the future. References Chang, Shih-Chiao 1988 Landslides and Their Environmental Impacts in Northern Taiwan (1968-1986). Report to the National Science Council, Republic of China, 34 pp.

PAGE 387

Havlick 1989 Chen, K.Y. 1981 China Post 1990 Chyi, Shyh-Jeng 1991 363 The Study of the Relationship between Urbanization and Natural Hazards around Taipei Basin III. Report to the National Science Council, Republic of China, 34 pp. "The Tracks and Intensities of Typhoons in Taiwan." Geograph ical Research Bulletin of National Taiwan Normal University 7, pp.61-74. "Typhoons' 165 KPH Winds Lash Taiwan." September 8, p. 12. The Typhoon Hazards of Taiwan. Unpublished research paper, National Taiwan University Institute of Geography, January. Havlick, Spenser W. 1986 "Third World Cities at Risk: Building for Calamity." Environ ment 28 (9, November): 6-11,41-45. 1991 Ju, Lui Sz Shiau 1990 "Urbanization and Resource Management: Does Impact Analysis Tell the Full Story of Rapid Urban Growth?" The Chinese Architect 17 (3, March): 70-75. Taiwan Environmental Protection Union "Green Team" field trip coordinator. Personal interview. September 20.

PAGE 388

SNOHOMISH COUNTY WASHINGTON, THANKSGIVING DAY FLOOD, 1990 Sky Miller Snohomish County Public Works Introduction Over the Thanksgiving Day weekend, 1990, Snohomish County, Washington, experienced the flood of record on Skykomish and Snohomish rivers. Floods of this magnitude graphically demonstrate the need for floodplain management strategies. Snohomish County Public Works is developing comprehensive floodplain manage ment plans for the Skykomish and Snohomish river systems. This paper describes the flood on the Skykomish and Snohomish rivers and how ongoing floodplain planning efforts are affecting repair and recovery efforts. Thanksgiving Weekend, 1990 Two large cold fronts dumped up to three feet of snow in the Cascade Mountains during the week preceding the flood. Then, two huge warm fronts dumped up to 15 inches of rain, melting the snow, and creating a single runoff event. The emergency response was very good. The National Weather Service had announced on Monday the potential for flooding on Thursday. County-wide, several hundred floodplain residents were evacuated before the peak of the flood. Thousands of cattle and other livestock were also moved to high ground; 175 people were rescued by helicopter. One hundred and forty-two homes were damaged beyond repair; 23 homes/cabins have completely disappeared. Fourteen levee breaks flooded 17,500 acres up to 20 feet deep, 42 roads were washed out, and more than 20 bridges were damaged. Major highway inundation and damage cut off access to five towns and countless residents. Not one life was lost. The Skykomish River The Skykomish River, designated by the state as a wild and scenic river, is large, steep, and very dynamic. Channel migration, bank erosion, and high velocities are the primary cause of major damage. One location experienced main channel lateral migration of 500 feet. The river now occupies the location of Skyko 2, a subdivision that lost seven cabins overnight. Elsewhere in the system, several homes were left hanging precariously over the river's edge, some high enough to be completely out of the regulatory floodplain. Two training dikes (Hansen and Sultan) intended to halt the channel migration were washed out. Numerous bank failures contributed to extensive private farmland damage. Main stem spawning species of salmon, the Chinook and Chum, are expected to have heavy losses due to scouring of eggs in the gravel. Destruction of habitat is expected to have an impact on Coho salmon and Steelhead, which spawn in the smaller tributaries.

PAGE 389

Miller 1"" Diking DistrictsiFlood Control Districts !;\ E:::3100 Year Flood Plain Boundary Figure 1. Snohomish River diking and flood control districts. 365

PAGE 390

366 Thanksgiving Day Flood The Snohomish River The Snohomish River, into which the Skykomish flows, is a mildly sloped, tidally influenced river ringed by 44 miles of levees. These levees protect 25,000 acres of farmed bottomland and are maintained by nine separate diking districts. The dike heights vary, but all districts are at the 1O-year level of protection or less. Decades of disputes, secretive raising of levee heights, and alternating flooding impacts have fueled what has been termed "dike wars" by the local media. The French Slough levee in the upper Snohomish began overtopping along its entire three miles oflevee. Floodwater spilled into this district for 20 hours, filling 6,000 acres up to 20 feet deep. The levee sustained considerable damage, but did not fail. This levee system functioned as designed: the levees overtopped, filled the district without failure, then flood gates and a pumphouse discharge stored flood water back into the river after the flood. The Marshland Flood Control District maintains 8.5 miles of levee that is generally higher than neighboring districts. The upper Marshland levee withstood most of the flood, mostly due to the French Slough district across the river taking such a large volume of water that the river level remained constant for 20 hours. Then, when the French Slough District had completely filled, the river rose quickly, and the added pressure burst the Marshland levee in two locations. A wall of water roared through the district, moving houses off their foundation and destroying roads and railroads. Several families were removed from rooftops by helicopter. This surge continued down through the district, filling 6,000 acres up to 10 feet deep. The lower levee and the Lowell-Snohomish Road were destroyed from the inside out. Diking District 13 (562 acres) was the first to begin taking water over the levees. Several partial repairs to levees damaged in a flood three weeks prior were washed out. Flood water filled District 13, and flowed into District 6, which was taking substantial amounts of water over its own levees. District 6 filled 8 to 10 feet deep to the top of its levees, which broke from the inside out at the lower end. Diking District 1 levees also broke early in the flood, which filled 3,800 acres 10 to 12 feet deep and inundated about 40 homes to the eaves for the fifth time in 16 years. Water stayed ponded behind the levees for several weeks. Floodplain Management Plans Before the flood, the county had initiated a comprehensive flood control planning effort (funded, in part, by the Washington State Department of Ecology). A committee was assembled consisting of representatives from the diking districts, Indian tribes, agencies, and environmental groups to help develop floodplain management strategies for the Snohomish River. An unsteady state hydraulic model was used by the group as an analytical tool to determine the upstream and downstream effects of various levee configurations. Long-term, specific recommendations were made by the group, including maximum and equitable heights for all levees in the Snohomish system. Pre flood committee consensus had centered on a maximum height for all levees, acknowledging that all

PAGE 391

Miller 367 floods could not be contained by the levees. It was agreed that all levee systems need to be designed to overtop, with spillway sections directing high velocity water away from homes and structures. This does two things: 1) it halts the continual raising of the levees, which effectively stacks water in the channel until one or more levees fail catastrophically; and 2) by having neighboring levee systems at comparable levels, the water is not concentrated into one district, but spread out such that no levee overtops by more than a foot or so. The tradeoff with overtopping levees is that they are much more expensive, therefore they need to be built lower. This sacrifices the protection from smaller floods in order to better withstand the forces of large floods and avoid the potential for catastrophic failure. Debate had centered on how much of this tradeoff the individual districts were willing to make. Before any recommendations could be implemented, the flood hit. This overtopping concept prevailed after the flood, when all parties agreed to a lower levee profile in light of the tremendous destruction caused by levee failures. Ultimate levee profiles are approximately two feet lower than originally suggested. Reconstructed levees will be built to this agreed upon height. Levees currently higher than the maximum allowed will be lowered. This amounts to about three miles of levee in the Marshland District alone. Another recommendation is the buyout of flood-prone Diking District 6. This district consists of 460 acres, primarily beef pasture, that is protected by 2.2 miles of levee. This levee requires governmental funding of repairs every two to five years. Improving the levee to a level comparable with neighboring districts will cost nearly 60% of the worth of the land it protects, without substantially changing the land use or value. In fact, merely maintaining the levee at its existing level has cost taxpayers twice what the land behind the levee is worth. Purchase of the land and removal of the levee would eliminate repeated governmental repair costs, provide enough off-channel storage to decrease peak stages, reestablish productive tidally influenced freshwater wetlands, and provide additional recreational area. The Washington state legislature has initiated budget proposals to fund $350,000 of a total $990,000 cost. Lessons Planning efforts, technical studies, and direct observation of the effects of a large flood solidified floodplain management strategies even before repair work began. 1. All levees in the system need to be at equitable heights. 2. The levees cannot hold out all floods and must be designed to overtop. 3. Economics of some flood-prone areas justify public purchase to remove the protection and let it flood.

PAGE 392

THE SHADYSIDE, OIDO, FLASH FLOODS: JUNE 14, 1990 William L. Read National Weather Service Description of the Event On the evening ofJune 14, 1990, severe flash flooding occurred near Shadyside, a town in southeast Ohio. The most severe flash flooding occurred on Wegee and Pipe Creeks, two small tributaries to the Ohio River, resulting in 26 fatalities and considerable property damage. Although the topography of the watershed was typical of other flood-prone valleys in the central Appalachians, there had been no recorded history of flash flooding on either one of these two creeks. Consequently, public awareness of the possibility of a flood of this magnitude was essentially nil. A meteorological synoptic situation developed over the Ohio Valley that was typical for June, but a series of mesoscale events focused over the headwaters of these creeks and produced catastrophic consequences. Real-time radar estimates of rainfall underestimated the event by a factor of more than two, most likely due to the tropical nature of the airmass in place over Ohio at the time of the flood. Postanalysis of radar rainfall estimates from the WSR-57S radar at Pittsburgh indicated that three to four inches of rain fell on the headwaters between about 8 :30 and 9:45p.m. EDT on June 14. The watersheds were small, with each creek draining about 12 square miles. Topographically, the area consists of rather low hills with narrow, steeply sided valleys. The soils in the area, which were 40 to 70 inches deep, were nearly saturated due to above normal (200%) rainfall from the previous month. The brief torrential rain, combined with the saturated soils, resulted in rapid runoff and flash flooding, which began at the headwaters between 9: 15 and 9:30p.m. EDT and reached the Ohio River around 1O:00p.m. EDT. Residents in the area reported the flood as a "wall of water" ranging from 10 to 30 feet above bankfull. While some debris damming may have contributed to the flash flood, hydraulic calculations performed after the event using the above rainfall estimates and soil conditions produced flooding on the order of the magnitude observed. Warning Service The extremely wet soil conditions and the movement of rather heavy thunder storms into southeast Ohio during the early evening hours on June 14 prompted the Nationitl Weather Service (NWS) Forecast Office in Cleveland to issue a Flood Watch about two hours prior to the onset of flash flooding near Shadyside. However, the lack of rainfall or stream gage measurements and the underestimation of rainfall in real time by radar techniques precluded the issuance of any flash flood warnings for this event. The technology in place over southeast Ohio (and most of the United States) is simply not sufficient to detect rainfall events of this scale in a timely fashion in order to issue effective warnings.

PAGE 393

Read 369 An important part of the detection and warning operations of the NWS is having a knowledge of flood-prone streams in an office's warning area. A further complication to the detection and warning problem on June 14 was the lack of history of flash flooding anywhere near this magnitude on these creeks. The headwaters of these streams were not included in FEMA 100-year floodplain studies. None of the residents interviewed could recall any flood events even close to this magnitude on either creek. The NWS disseminates warning and forecast information through several methods. Primary methods are the National Oceanographic and Atmospheric Administration (NOAA) Weather Wire Service (NWWS), provided under contract with CONTEL, NOAA Weather Radio (NWR), or through vendors who access NOAA's Family of Services database. Media in southeast Ohio receive NWS warnings and forecasts through Associated Press and NWWS. The NWR coverage in the area in and near Shadyside was quite poor, due to the distance to the transmitter and terrain blocking from surrounding hills. The state of Ohio retransmits NWS watches and warnings automatically through a link to the Ohio law enforcement telecommunications system to at least one office in each county. The county office must then fan out warning information by telephone. Given the many other duties of dispatchers at these law enforcement offices, there is potential for delay or gaps in this fan-out procedure. The media in southeast Ohio appear to be very proactive concerning hazardous weather and flooding. Most residents along Pipe and Wegee creeks, as well as officials in Shadyside, were aware that a Flood Watch had been issued for their area through the television station or radio station they were tuned into. These stations broke into scheduled programming frequently and the TV stations ran periodic crawls. Many of the residents and local officials interviewed stated they knew that a Flood Watch was in effect but did nothing to prepare for possible flooding. They also stated that had a warning been issued, they probably would not have done anything either. The actions people took once the flash flood was under way varied. Many were caught by surprise by the fast-moving flood waters and were unable to take any escape actions. A few residents reported they ran up the hillsides near their homes. Others chose to "ride out" the flood in their homes even if they heard the flood waters coming. Rather surprisingly, few tried to escape by car and none of the fatalities were attributed to people trying to escape in cars. Very little in the way of flash flood preparedness had been conducted in Shadyside. On the other hand, preparedness for tornadoes and other hazards, such as fire, seems to be quite good. The state of Ohio requires teaching of tornado safety in the public schools. This is due, in part, to recent tornado outbreaks (1974, 1985). School children in the Shadyside area had received this education but could not recall any flood safety education.

PAGE 394

370 The Shadyside Flash Floods Conclusions and Recommendations The NWS's current capability to detect and warn for flash floods is not sufficient for events on the scale of Shadyside. Furthermore, even if detection and warning capability were sufficient, methods of communication of the warning will have to become automated down to communities the size of Shadyside to assure rapid, organized response. Local preparedness for flash floods is not nearly as great as for other hazards. The NWS is embarking on a massive modernization of the technology it uses to forecast, detect, and warn for severe weather and flash flooding. Through use of the Next Generation Weather Radar (NEXRAD), new satellite systems, new surface observing systems, and an upgraded computer system to integrate the data and expedite warning issuance, forecasters will begin to have the necessary tools to attempt to detect and warn for these small-scale events. Considerable research will be needed to use the new data for events the scale of Shadyside. In order to develop community preparedness, efforts need to be taken nationwide to identify similar flash-flood-prone communities. This could be a monumental task, given that there may be thousands of these communities in the Appalachians alone. The Shadyside event should be used as an example for teaching the need for preparedness in other flash-flood-prone communities. In addition, flash flood preparedness needs to be incorporated in our public education system as is tornado safety in Ohio. For the Shadyside flash floods, due to the rapid onset of the flood, an automated local flood warning system would have had to have been in place to provide sufficient lead time for residents to take life-saving actions. In order to have an automated flood warning system there in the first place, there would have had to have been an awareness of the potential for devastating flash flooding. Therein lies the dilemma-how do we convince other flash-flood-prone communities to invest in the system and develop a local action plan if there is no history of flash flooding? It is hoped that, using the Shadyside experience as an example, the NWS and concerned local officials can begin to tackle this problem.

PAGE 395

Publications of Interest from the Natural Hazards Research and Applications Information Center Campus Box 482 University of Colorado Boulder, Colorado 80309-0482 (303) 492-6819 Special Publications SP02 Regulation of Flood Hawro Areas to Reduce Flood Losses, Volume 3. Jon Kusler. 1982. 300 pp. $8.00. SP03 Strengthening State Floodplain Management, Appendix A to Volume 3. Patricia A. Bloomgren. 1982. 123 pp. $8.00. SP04 Innovation in Local Floodplain Management, Appendix B to Volume 3. Jon Kusler. 1982. 262 pp. $8.00. SP05 Floodplain Regulations and the Courts: 1970-81. Jon Kusler. 1982.51 pp. $5.00. SP09 Improving the Effectiveness of Floodplain Management in Western State High Risk Areas: Alluvial Fans, Mudflows, Mud Floods. Proceedings of a Workshop, Palm Springs, California, February 15-16, 1984.97 pp. $7.00. SPI0 Evaluating the Effectiveness of Floodplain Management Techniques and Community Programs. Proceedings of a Seminar, April 1984. 1985. 143 pp. $8.00. SPll Managing High Risk Flood Areas: 1985 and Beyond. Proceedings of the Eighth Annual Conference of the Association of State Floodplain Managers, June 1984. 1985. 326 pp. $8.00. SP15 Strengthening Local Flood Protection Programs. Proceedings of the Tenth Annual Conference of the Association of State Floodplain Managers, June 1986. 1987. 320 pp. $10.00.

PAGE 396

SP16 What We Have uamed Since the Big Thompson Rood. Eve Gruntfest, ed. Proceedings of the Tenth Anniversary Conference, July 17-19,1986. 1987. 290 pp. $4.00. SP20 The Anny Corp 0/ Engineers and the Evolution 0/ Federal Rood Plain Management Policy. Jamie W. Moore and Dorothy P. Moore. 1989. 184 pp. $15.00. SP21 The Natural Hazards Data Resources Directory. Leaura M. Hennig. 1990. 247 pp. $15.00. SP22 Partnerships: Effective Rood Hazard Management. Proceedings of the Thirteenth Annual Conference of the Association of State Floodplain Managers, May 1989. 1990. 300 pp. $10.00. SP23 Challenges Ahead: Rood Loss Reduction Strategies/or the '90s. Proceed ings of the Fourteenth Annual Conference of the Association of State Floodplain Managers, June 1990. 1991. 304 pp. $10.00. Monograph Series MG29 The Rood Breakers: Citizens Band Radio Use During the 1978 Rood in the Grand Forks Region .. Thomas E. Drabek, et al. 1979. 129 pp. $8.00. MG40 When the Ground Fails: Planning and Engineering Response to Debris Rows. Martha Blair, et al. 1985. 114 pp. $8.00. MG41 Community Recovery From a Major Disaster. Claire B. Rubin, et al. 1985. 295 pp. $10.00. MG45 Regional Management 0/ Metropolitan Roodplains: Experiences in the United States and Abroad. Rutherford H. Platt, ed. 1987. 320 pp. $10.00. MG47 Cities Under Water: Ten Cities' Efforts to Manage Roodplain lAnd Use. Raymond J. Burby, et al. 1988. 240 pp. $10.00. MG49 Roodproo/ Retrofitting: Homeowner Self-Protective Behavior. Shirley B. Laska, et al. 1991. 280 pp. $10.00. MG52 The Feasibility 0/ Vertical Evacuation: Behavioral, ugal, Political, and Structural Considerations. Carlton Ruch, et al. 1991. 262 pp. $10.00.

PAGE 397

Working Paper Series WP04 A Selected Bibliography of Coastal Erosion, Protection and ReWed Hwnan Activity in Nonh America and the British Isles. J.K. Mitchell. 1968. 70 pp. $4.50. WP12 Technical Servicesfor the Urban Floodplain Property Manager: OrganizJJlion of the Design Problem. Kenneth Cypra and George Peterson. 1969.25 pp. $4.50. WP29 Flood Insurance and Community Planning. N. Baumann and R. Emmer. 1976. 83 pp. $4.50. WP31 Warning for Flash Roods in Boulder, Colorado. Thomas E. Downing. 1977. 80 pp. $4.50. WP32 What People Did During the Big Tlwmpson Rood. Eve C. Gruntfest. 1977. 62 pp. $4.50. WP34 Human Response to Hurricanes in Texas-Two Studies. Sally Davenport. 1978. 55 pp. $4.50. WP35 Hazard Mitigation Behavior of Urban Rood Plain Residents. Marvin Waterstone. 1978. 60 pp. $4.50. WP39 Effects of a Natural Disaster on Local Mongage Markets: The Pearl River Rood in Jackson, Mississippi-April 1979. Dan R. Anderson and Maurice Weinrobe. 1980. 48 pp. $4.50. WP52 The Effects of Rood Hazard Information Disclosure by Realtors: The Case of the Lower Rorida Keys. John Cross. 1985. 85 pp. $4.50. WP61 The Local Economic Effects of Natural Disasters. Anthony M. Yezer and Claire B. Rubin. 1987. 75 pp. WP68 Rood Insurance and Relief in the U.S. and Britain. John V. Handmer. 1990. 32 pp. $4.50. Topical Bibliographies TB02 Bibliography on Rood Proofing. Anita Cochran. 1977. 9 pp. $1.00. TB03 Rash Rood Warning Bibliography. Kathleen Torres and Anita Cochran. 1977. 10 pp. $1.00.

PAGE 398

Quick R.esponse Research Reports QR02 The 1986 California Floods. Robert Bolin. 1986. 32 pp. $3.00. QR17 Spatial and Temporal Variability in Residential Land Values FoUowing Catastrophic Flooding. Burrell Montz and Graham Tobin. 1986. 14 pp. $1.75. QR38 Managing Reconstruction Along the South Carolina Coast: Preliminary Observations on the Implemenlotion 0/ the Beach/ront Management Act FoUowing Hurricane Hugo. Timothy Beatley. 1990.27 pp. $2.75.


Download Options

Choose Size
Choose file type
Cite this item close


Cras ut cursus ante, a fringilla nunc. Mauris lorem nunc, cursus sit amet enim ac, vehicula vestibulum mi. Mauris viverra nisl vel enim faucibus porta. Praesent sit amet ornare diam, non finibus nulla.


Cras efficitur magna et sapien varius, luctus ullamcorper dolor convallis. Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Fusce sit amet justo ut erat laoreet congue sed a ante.


Phasellus ornare in augue eu imperdiet. Donec malesuada sapien ante, at vehicula orci tempor molestie. Proin vitae urna elit. Pellentesque vitae nisi et diam euismod malesuada aliquet non erat.


Nunc fringilla dolor ut dictum placerat. Proin ac neque rutrum, consectetur ligula id, laoreet ligula. Nulla lorem massa, consectetur vitae consequat in, lobortis at dolor. Nunc sed leo odio.