USF Libraries
USF Digital Collections

Coast to coast

MISSING IMAGE

Material Information

Title:
Coast to coast 20 years of progress : proceedings, Twentieth Annual Conference of the Association of State Floodplain Managers : June 10-14, 1996, San Diego, California
Series Title:
Special publication (University of Colorado, Boulder. Natural Hazards Research and Applications Information Center) ;
Portion of title:
Proceedings, Twentieth Annual Conference of the Association of State Floodplain Managers : June 10-14, 1996, San Diego, California
Physical Description:
1 online resource (xiv, 391 p.) : ill. ;
Language:
English
Creator:
Association of State Floodplain Managers -- Conference, 1996
University of Colorado, Boulder -- Natural Hazards Research and Applications Information Center
Publisher:
Natural Hazards Research and Applications Information Center, Institute of Behavioral Science, University of Colorado
The Association
Place of Publication:
Boulder, Colo
Madison, WI
Publication Date:

Subjects

Subjects / Keywords:
Floodplain management -- Congresses -- United States   ( lcsh )
Flood damage prevention -- Congresses -- United States   ( lcsh )
Genre:
bibliography   ( marcgt )
conference publication   ( marcgt )
non-fiction   ( marcgt )

Notes

Bibliography:
Includes bibliographical 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:
aleph - 002021658
oclc - 428807945
usfldc doi - F57-00084
usfldc handle - f57.84
System ID:
SFS0001165:00001


This item is only available as 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 2200313Ia 4500
controlfield tag 001 002021658
005 20090928142449.0
006 m d
007 cr bn|||||||||
008 090731s1996 coua sb 100 0 eng d
datafield ind1 8 ind2 024
subfield code a F57-00084
035
(OCoLC)428807945
040
FHM
c FHM
043
n-us---
049
FHMM
090
TC423 (ONLINE)
2 110
Association of State Floodplain Managers.
b Conference
n (20th :
d 1996 :
San Diego, Calif.)
1 0 245
Coast to coast
h [electronic resource] :
20 years of progress : proceedings, Twentieth Annual Conference of the Association of State Floodplain Managers : June 10-14, 1996, San Diego, California.
3 246
Proceedings, Twentieth Annual Conference of the Association of State Floodplain Managers : June 10-14, 1996, San Diego, California
260
Boulder, Colo. :
Natural Hazards Research and Applications Information Center, Institute of Behavioral Science, University of Colorado ;
Madison, WI :
The Association,
1996.
300
1 online resource (xiv, 391 p.) :
ill.
440
Special publication (University of Colorado, Boulder. Natural Hazards Research and Applications Information Center) ;
v 32
500
Description based on print version record.
504
Includes bibliographical references.
650
Floodplain management
z United States
Congresses.
Flood damage prevention
United States
Congresses.
710
University of Colorado, Boulder.
Natural Hazards Research and Applications Information Center.
773
t Natural Hazards Center Collection
4 856
u http://digital.lib.usf.edu/?f57.84



PAGE 3

The opinions contained in this volume are those of the authors and do not necessarily represent the views of the funding or sponsoring organizations or those of the Association of State Floodplain Managers. The use of trademarks or brand names in these technical papers is not intended as an endorsement of the products. Published 1996. 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 (303) 492-6819; fax: (303) 492-2151; e-mail: hazctr@colorado.edu and The Associ a tion of State Floodplain Managers, Inc. 4233 West Beltline Highway Madison, WI 53711 (608) 274-0123; fax: (608) 274-0696; e-mail: asfpm@execpc.com

PAGE 4

Preface "Coast to Coast: 20 Years of Progress" was the theme for the Association of State Floodplain Managers' 20th Annual Conference held in San Diego, California, from June 10 to 14, 1996. That theme is reflective of the growth of the Association as well as its impact on floodplain management practices in the United States. The progress seen over the past 20 years includes the growth the Association has enjoyed: 20 years ago the conference was held in one hotel room. Attendance at San Diego was well over 400 people, including representation from foreign countries. Plenary sessions featured outstanding speakers that are, in fact, national leaders in floodplain management. In the first session, Richard Krimm, Acting Associate Director for Mitigation for the Federal Emergency Management Agency, outlined goals for the National Mitigation Strategy; and Michael Davis, Deputy Assistant Secretary for Planning, Policy and Legislation, U.S. Arn1y, updated the attendees on new directions for the Corps of Engineers. The conference started on a high note and somehow continued to grow in energy. The closing plenary session had a fast-paced, thought-provoking commentary on national flood policy since the 1993 floods, by General Gerald E. Galloway, and an inspirational wrapup by Frank H. Thomas. Workshops offered training in mitigation planning; basic floodplain management; the National Flood Insurance Progran1; HEC/RAS for managers; risk analysis; and computer modeling programs, Check-2 and Quick-2. The workshops were well-attended and further diversified the conference's educational opportunities. While this volume of proceedings captures the words and teclmical content of the conference, it cannot convey the chemistry and interchange an10ng more than 400 floodplain managers simultaneously seeking and providing solutions to floodplain issues and sharing state-of-the-art thinking. The profession will continue to evolve in response to issues and events; you can be a part of it by contributing your experience and point of view. The Association is greatly indebted to the conference team, our host city of San Diego, and the exhibitors-all their efforts were extraordinary. Finally, we applaud the participants for their energy and desire to contribute to the exchange of ideas. If you missed the 20th annual conference, be assured that this level of information and energy is becoming nonnal and will reach critical mass again next year, in Little Rock, Arkansas. Hope to see you there! George Hosek Chair, Association of State Floodplain Managers

PAGE 5

This page IS intentionally blank

PAGE 6

Acknowledgements The 20th annual conference of the Association celebrated a significant milestone, but it could not have taken place without the help of a lot of people who were willing to work hard and contribute their time and talents towards that success. We thank the Board of Directors for their support and encouragement; Chair George Hosek and Executive Director Larry Larson for their leadership and support; and Diane Watson for her freely given talent, advice, knowledge, and heart! Super "kudos" to the many speakers and moderators who provided the technical basis of the conference, and to the attendees who, in the long run, are the most important part of all. Special thanks to David Kennedy, Director, California Department of Water Resources; and George Qualley, Chief, Division of Flood Management, for allowing staff time to work on the conference. Our thanks to all those who contributed to the organization and development of this conference. We know that many unnamed people assisted and we acknowledge their help, along with those identified below-for it was truly a teanl effort. Andy Lee and Mike Parker Jan Farmer Dante Accurti Exhibits Chair Conference Directors Program Chair Program Chair's Team Jan Fanner, Program Chair Arizona Floodplain Management Association (AFMA) Chuck Williams, County Engineer, Navajo County Public Works David Williams, West Consultants Tom Loomis, AFMA Cathy Register, AFMA Sandra Steichen, AFMA George Hosek, Michigan Dept. of Natural Resources (DNR) Larry Larson, Wisconsin DNR Exhibits Chair's Team Dan Accurti, Exhibits Chair Pennsylvania Dept. of Transportation Diane Watson, ASFPM Ruth Duelley, Floodplain Management Program, California Dept. of Water Resources (FPM, CDWR) Conference Director's Team Andy Lee, Conference Director FPM, CDWR Mike Parker, Conference Co-Director, Santa Barbara Flood Control and Water Conservation District (SBFC& WCD) and Floodplain Management Association A. Jean Brown, FPM, CDWR Joe Hill, Flood Control Division, San Diego Public Works Dept. (FCD,SDCPWD) Sue Josheff. Wisconsin DNR Wendy Malamut, Federal Emergency Management Agency Region IX Shawnie Gookin, Berryman & Henigar

PAGE 7

Making it Happen Theme Andy Lee, FPM, CDWR Joe Hill, FCD, SDCPWD Jerry LOllthain, Washington Dept. of Ecology Registration Mike Parker, SBFC& WCD Jeff Paley, SBFC& WCD Georgia Navarro, SBFC& WCD Antoinette Daniel, FPM, CDWR Sally Boria, FPM, CDWR Logo Design Andre Carii, Wiley Design Mike Miller, Office of Water Education, Graphic Services, CDWR Andy Lee, FPM, CDWR Conference Brochure Antoinette Daniel, FPM, CDWR Ucla Soltani-FlyIUl, FPM, CDWR Diane Watson, ASFPM Ken NalUllan, FEMA Region IX General Support Ruth Dudley, FPM, CDWR I-Ming Cheng, FPM, CDWR Antoinette Daniel, FPM, CDWR Uda Soltani-Flynn, FPM, CDWR Don YOllng, FPM, CDWR Sally Boria, FPM, CDWR Local Arrangements Joe Hill, FCD, SDCPWD Anders Egense, Boyle Engineering Allen Cooper, Cooper Engineering Carole Forrest, Woodwa rd-Clyde Arsalan Dadkhah, Consultant Dennis Bowling, Rick Engineering Shawnie Gookin, Berryman & Henigar Publications Table Sue Josheff, Wisconsin DNR VI Golf Tournament Clancy Philipsborn, The Mitigation Assistance Corporation Dennis Bowling, Rick Engineering Technical Field Tour Jon Walters, Nolte & Associates Dennis Bowling, Rick Engineering Conference Signs Maggie Mathis, Dewberry & Davis Food Festival Wendy MalanlUt, FEMA Region IX Shawnie Gookin, Berryman & Henigar Jane Lockwood, New Hampshire State Planning Office Dennis Bowling, Rick Engineering Anders Egense, Boyle Engineering Hospitality Room Cay Brown, FPM Shawnie Gookin, Berryman & Henigar Workshops David Williams, West Consultants John F. Magnotti, FEMA Headquarters Zekrollah Momeni, Dewberry & Davis Andy Corry, FPM, CDWR Bob Durring, FEMA Region IX Diana Herrera, Nationa I Flood Insurance Program French Wetmore, French & Associates Clancy Philipsborn, The Mitigation Assistance Corporation Berry Williams, Berry Williams Associates David Stroud, ISO, Commercial Risk Services Mike Burnham, u.s. Army Corps of Engineers Family Tours/Children's Program Shawnie Gookin, Berryman & Henigar Demus Bowling, Rick Engineering

PAGE 8

Videotaping and Monitoring Howard Change, San Diego State University I-Ming Cheng, FPM, CDWR Students from San Diego State Fund Raising Chuck Spinks, Shawnie Gookin, Berryman & Henigar Dan Accurti, Pennsylvania Dept. of Transportat ion Diane Watson, ASFPM Andy Lee, FPM, CDWR Luau Sue Josheff, Wisconsin DNR Shawnie Gookin, Berryman & Henigar Sterling Yong, Hawaii DLNR, NFIP Program Mcmorabilia Sto.-c Licla Soltani-Flynn, Sally Borla, FPM, CDWR LytUl Phillips, ASFPM Accounting Ruth Dudley, Andy Lee, FPM, CDWR Surfing Safari Mike Parker, SBFC& WCD Steve Wagner, SBFC& WCD Eric Pearson, SBFC& WCD Notcbook & Portfolio A. Jean Brown, Ruth Dudley, Licla Soltalli-Flyt111, FPM, CDWR John Sibilsky, U.S. Army Corps of Engineers Door Prizcs Shawnie Gookin, Berryman & Henigar Jaytle Janda-Timba, Rick Engineering Joe Hill, FCS,SDCPWD Association Support Scrvices Diane Watson, Executive Office Larry Larson, Executive Director George Hosek, Chair Jerry Loud13.ill, Past Chair Jack Page, Conference Committee Chair LytlIl Phillips, Executive Office Key Support Elements Confcrcncc Hotcls lisa Butler, Kim Gaines, Catamaran Resort & Bahia Hotels Dana Inn Blue Sea uJdge Pacific Shores Inn Printing California Department of Water Resource s, Reprographics Spccial Food Luau Roast Pig: John Lee, San Choy Restaurant AIUliversary Cake: Ralph's Graphic Arts CDWR, Office of Water Education, Graphic Services Entc.-tainmcnt Rill Calhoun Mobile Music Kauli Brown, Na ALI-I Productions Local Bank Bank of America San Dicgo Visitors & Convcntion Bureau S:Uldra Butler, Diana Forcier Pipc and Drapc, Drayagc Gary Mueller, Carden Convention Services

PAGE 9

Audio Visual Phil Thomas, Anelw Audio Visual Name Tags Innovative Engineering Portfolios & Binders Dave Adams, McBee-Information Packaging Products Memorabilia Store Items New Directions, Metroform, Carson Integrated Marketing Personal Computer Rental Electro Rent, Inc. Conference Supporters We gratefully acknowledge the following federal, state, and local agencies, corporations, and associations who provided additional support for the conference. Platinum (SI,OOO or more) Michael Baker, Jr. Dewberry & Davis Woodward-Clyde Arizona Floodplain Management Association Floodplain Management Association (of California) Arizona Dept. of Water Resources Califomia Dept. of Water Resources Clark County, Nevada, Regional Flood Control District U.S. Anny Corps of Engineers Federal Emergency Management Agency Natural Resources Conservation Service U.S. Environmental Protection Agency National Park Service Gold (S750 or more) Transamerica Flood Hazard Certification Silver (S500 or more) Wisconsin Dept. of Natural Resources Bronze (S250 or more) State of Hawaii, Dept. of Land and Natural Resources, NFIP The Mitigation Assistance Corp. French & Associates Greenhorne & OMara West Consultants Berryman & Henigar Borcalli & Associates, Inc. Boy Ie Engineering Nimbus Engineers Nolte & Associates Rick Engineering Granite (Donor's choice) Leslie A. Bond & Associates viii

PAGE 10

Contents Part 1. National Policy and Programs New Directions for the Corps of Engineers Water Resources Programs Michael Davis ........................................ .3 Mitigation and Partnerships for Floodplain Management Shirley Mattingly ....................................... 9 Review of Literature on Federal Hazard Mitigation Efforts (1979-1995) Claire B. Rubin ....................................... 14 Part 2. Multi-hazard Mitigation v V Mitigating against Flood and Earthquake Hazards Michael Mahoney ..................................... 25 Promoting a Multi-hazard Approach when Retrofitting Floodprone Structures Clifford Oliver ........................................ 30.......-Malibu/Las Flores Canyon Watershed Hazard Mitigation Plan (Floods, Fires, Landslide) Bruce M. Phillips ..................................... 35 Part 3. Multi-use Watercourses Planning a Future for the Salt-Gila Rivers: A Case Study in Designing a Master Plan Process for a Multi-use, Multi-purpose Watercourse Catherine A. Tice and R. Keith Julian ....................... 45 Reclaiming Denver's Central South Platte River Leo Eisel, Brian Kolstad, Ben Urbonas, and Nick Skifalides ........ 51 t.--Floodplain Management in Urban Redevelopment: A Case Study in Multiple Objective Management Bernard B. Sheff and Kenneth A. Nacci ...................... 57 ix

PAGE 11

Methodical Mitigation-A Deliberate Approach to Floodplain Management Jan Horton .......................................... 69 Floodplain Management in Mecklenburg COlmty / Stephen R. Sands and William R. Tingle ................... V 85 Part 5. Coastal Issues Correlation of Hurricane Magnitude, Percentage Chance of Exceedance, .... Analyzing and Mapping Coastal Flood Hazards along the Open Coasts of the Atlantic Ocean and Gulf of Mexico Jerry Sparks, Darryl./. Hatheway, and Doug A. Bellomo ......... 107 An Evaluation of the Costs Associated with Managing Delaware's At.lantic Ocean Coast through a Policy of Retreat MIchael S. Powell .................................... 113 Part 6. Hydrology and Hydraulics Preparation of a Hydrology Manual for Imperial County, California Steve R. Knell, Theodore V. Hromadka, Johannes J. DeVries, ./ and Anders K. Egellse ............................... V.121 Flood Frequency Analysis in the Presence of Outliers, Historic Data, Varying Generalized Skews Wilbert 0. Thomas, Jr. and David P. Prellsch ................ 127 Evaluating Stonn Water Rlmoff from Steep Slope Arid Lands Clifford E. Anderson and Richard J. Heggen ................. 135 x

PAGE 12

jApproaches to In-Situ Calculation of Floodplain Roughness ./ Barry Hecht and Jonathan Owens ....................... VI42 !Notes on Translatory Waves in Natural Channels .. IH. W. Hjalmarson and J. V. Phillips ........................ 149 "'Recent Flood Damages and Bank/Scour Protection Measures at Bridge in South.east Arizona Zblgmew Osmolski and Fazle Karim ....................... 156 Part 7. Modelling and Computer Programs A Discussion of the U.S. Anny Corps of Engineers HEC-RAS .. Computer Program Wilbert 0. Thomas, Jr., Chris D. Krebs, and Gary W. Brunner .. Appr?ximate Floodplain Delineation Using WinXSPRO Martin J. Teal ....................................... 173 NFIP-Accepted Computer Models: Proprietary Issues vs. Public's Right to Appeal Jerry W. Sparks ...................................... 179 The NEXGEN Floodplain Hydraulics Program HEC-RAS Troy Lynn Lovell, Michael A. Moya, and Emilia Salcido ......... 184 Using the UNET Model to Estimate a 100Year Flood in Designated Floodway I-M1IIg Cheng ....................................... 190 Bridge Hydraulic Analysis with HEC-RAS Vernon BOllner and Gary Brunner ......................... 195 Part 8. Mapping Use of ARC/INFO for Floodplain Generation and Mapping in Jefferson COlmty, Kentucky Mark A. Sites, James A. Harned, Louis T. Greenwell, '.lI!d Alan M. Castaneda ................................ 205 A Substitute For Floodplain Delineations Gregory Rodzellko and Julie Lemmon .................... xi

PAGE 13

The Zone A Cnmch ./ David R Knowles and Peter A. Richardson ................ V215 Flood Hazard Mapping of the Bridge Canyon Fan Donald W. Davis and Gale Wm. Fraser II ................... 221 Part 9. Precipitation, Gaging, Forecasting, and Warning Flood Threat Recognition for Tangipahoa, St. Tammany, and Washington Parishes, Southeast Louisiana Mark R Wingate ..................................... 231 Guidelines for Developing Comprehensive Flood Warning Laurie T. Miller, Tom Donaldson, Dallas Reigle, Jesus Romero, Ste h D. Waters, Patricia Q. Deschamps, Sam A. Arrowood, and Wayne oflft7 Spatially Distributed Rainfall: the Use of Volunteer Gaging / Richard J. Heggen, Clifford E. Anderson, and Steve Hemphill ... V243 New Precipitation Frequency Studies for the United States / Lesley T. Julian and John L. Vogel ...................... v. 250 ................ b6 Part 10. Stormwater Management Urban Stormwater Regulations: A Worthy Opponent to ......................... 665 .......... V<71 Multi-objective Planning and Design of Stomlwater Detention Facilities/" Ronald D. Flanagan ............................... V.277 Innovative Approach for Peak Discharge Reduction in an Urban Environment Using a Multipurpose Detention Basin Douglas Lantz, Zbigniew Osmolski, and Fazle Karim ........... 283 xii

PAGE 14

Management Planning for Redevelopment of enver's Stapleton Airport oim M. Pflaum, William Wenk, and Ben Urbonas ......... 66 tUpper Peaks Branch: Flooding in an Unmapped Area JAlbert H. Half!, Walter E. Skipwith, Kevin Shunk, and Ben Cernosek 302 An Expressway, Stomlwater Management, and the Environment: A Case Stu.dy Ward S. Miller and Richard L. Thompson ................... 309 Kyle Canyon Detention Basin: Conception to Construction Ken Gilbreth and Kevin Eubanks .......................... 316 Part 11. Construction Techniques, Building Performance, and Data Collection .--Home Foundations:. A StillUuary of Current Studies V_ William L. Coulbourne and Cecelia Rosenberg ................ 323 Step by Step-Hand in Hand: Bringing Slab Elevation and a Vi?eo to South Louis,i,ana /I Patricia M. Skinner and Fred E. Gene Baker ................ 329 Performance of Flood Proofed Structures Tested by Floodwater Larry S. Buss ....................................... 335 GPS Elevation Surveys-A Key to Proactive Floodplain Management L David F. Maune ..................................... 340 S,tr.eanllined Data Collection for Substantially Damaged Structures in Ohj,p../' tnc Berman and Donald W Glondys .................... n46 Part 12. Planning for, Using, and Maintaining Structures for Loss Reduction ; Urban Planning for an Area Protected by Levees: The Natomas Basin in Sacramento County, California James C. Campbell ................................... 355 Battelle's Levee Rehabilitation and Letter of Map Revision v<::: Daniel M. Hill ....................................... 362 Grade Stabilization Structures for Natural Rivers loseph C. Hill Kenneth C. Hanson, and Jon Walters .......... Y168 Xlll

PAGE 15

Flood Control Planning for the American River Watershed, California / Ricardo S. Pineda and George T. Qualley ................. V. 374 Environmental Management vs. Floodplain Management at Reelfoot Creek in Western Tennessee / David S. Smith, Donald R. Davenport, and Roger A. Gaines .... V. 380 Part 13. Conference Summary v< Coast to Coast: Twenty Years of Progress Frank H. Thomas ................................ .. 389 xiv

PAGE 16

Part 1 NATIONAL POLICY AND PROGRAMS

PAGE 17

This page IS intentionally blank

PAGE 18

New Directions for the Corps of Engineers Water Resources Programs Michael Davis Deputy Assistant Secretary of the Army for Planning, Policy, and Legislation It is a real pleasure to be here at the annual conference of the Association of State Floodplain Managers, and to help celebrate your 20th anniversary. You should be proud of your history because you certainly have effected positive changes in floodplain management in the last 20 years, in the local, state and federal level. I understand that the ASFPM's relationship with the Office of the Assistant Secretary of the Army for Civil Works has been very positive in the past, and we want to continue to enhance that. I have also personally worked very closely with your sister agency, the Association of State Wetlands Managers, and forged relationships with others in the development community and the environmental community. We in the Secretary's office, as well as the Corps of Engineers, share many of your objectives in floodplain management and we certainly respect your views regarding federal floodplain management progranls and their impact on the states. I want you to know that I have a open door policy and certainly Secretary Lancaster does. To be successful in our endeavors, we must work closely with you on floodplain management issues, and wetlands protection and other progranlS. What I would like to do today is to focus on three specific areas. First, I'll talk about the Water Resource Development Act of 1996 (WRDA) and some of the key policy initiatives that we have and that Congress is considering. Then I will discuss the specific programs the Corps has in the area of floodplain management and flood damage reduction, and tie that in to some of the proposed WRDA initiatives. And fmally, I will cover the Clinton Administration's wetlands initiatives because there is a very direct link between wetlands protection and floodplain management. Any time one has the opportunity give a presentation these days, if one is a federal representative or talking about federal programs, it is

PAGE 19

4 New Directions for Corps Water Resources Programs generally a good-news-and-bad-news-type scenario. I can get the bad news out of the way first, then we can talk about some of the more positive things. We are all facing pressures-fiscal pressures, government downsizing, reduced funding, and reduced staff. All agencies are being hit with this, some harder than others, and I understand that the same thing is going on at the state and local level as well. The Corps is aggressively seeking sustained funding so we can have programs for the future and make commitments to sponsors and others. We also need to have a clear road map for how we can proceed now and into the next century. We're trying to provide for our operation and maintenance needs for a set of aging projects we have around the country that will need increased attention and increased funding as we continue to rehabilitate, repair, and maintain those projects. At the same time, we don't want to become just an O&M agency within the Corps of Engineers. We do want to do construction, we do want new starts, and we want them planned and designed in an environmentally responsible way. I think that it is very important that we have a biannual Water Resources Development bill. We need the project authorizations, as well as some policy changes to ensure that we can adapt our programs to the increasing fiscal pressures. There are basically three bills in play right now. The Administration's bill, H.R. 3563, was introduced in April 1996. In our bill there are 13 projects requiring authorization, eight project modifications, and numerous policy initiatives that we would like to see added to the law. We characterize our bill as a lean and green Water Resources Development Act. Most of the policy provisions are needed for the Corps to respond to environmental restoration and other environmental issues. The House of Representatives has introduced a bill, H.R. 3592, which includes 22 projects for authorization, 12 of which are included in the Administration's bill. The House bill also includes 23 small flood control projects and studies and 71 project modifications. Obviously this is a significantly larger bill than ours. The Senate version of the WRDA bill, S.640, includes 22 project authorizations and about 20 project modifications. We expect Congress will act on these bills over the coming months, and hopefully a conference committee action will take place in late July or early August. Let me shift now and talk about some of the Corps floodplain management and flood dmnage reduction programs and relate back to the some of the provisions in these various WRDA bills that could affect our ability to work in these areas. In the FY 1996 budget the Administration sought some bold new initiatives for the Corps of Engineers in flood damage reduction. We supported greater roles for the states, local governments, and tribal governments in solving flood problems through comprehensive floodplain management. We also proposed major increases

PAGE 20

Davis 5 in technical and planning assistance to the state, local, and tribal governments. Unfortunately, Congress didn't agree with us on the FY1996 budget and rejected our approaches. In the FY1997 budget the Administration is again aggressively seeking to support comprehensive floodplain management programs at the state, local, and tribal levels. We're trying to use the best mix of strategies and tools to solve floodplain management problems, building on the strategies that are articulated in the 1994 Unified National Program for Floodplain Management. Our strategy is designed to reduce flood damage as well as future costs of flood emergencies. Again, we are promoting increased technical and planning assistance to state, local, and tribal governments. One of the cornerstones of the Administration proposal for this year's WRDA was to change the cost-sharing percentage for structural flood damage reduction projects. The current formula of 75% federal, 25% non federal would be changed to 50% federal, 50% non-federal for structural projects. Perhaps more importantly, we would require as a link to any structural or nonstructural project, the development of a comprehensive floodplain management plan for the community. The new cost-sharing provisions would apply to all projects that do not have signed Project Cost Sharing Agreement. The House WRDA provisions would take a different approach. The House version reconmlends a 65% federal, 35% non-federal formula for all projects, including nonstructural projects. In addition, the House version rejects the requirements for a comprehensive floodplain management plan, instead making the plan a voluntary aspect of the project. Also, the new policy would only apply to those projects that were authorized after the Water Resources Development Act was passed. This is very important because it really wouldn't allow us to realize the budget savings necessary to allow us to more equitably distribute the Corps of Engineers flood danlage reduction funds across the country. The Senate has not discussed the cost-sharing issue. We are concerned about this because due to decreased funding and the increase in the nmnber of projects it is becoming more and more difficult to function tmder the 75-25 cost-sharing approach. Also, more importantly, we believe that our approach with the 50/50 structural approach will encourage our local sponsors to look more seriously at nonstructural solutions to flood damage reduction. Another area that is included in our WRDA is authorization for comprehensive watershed initiatives. Traditionally, the Corps has focussed on one or more projects separately within a watershed. We are now looking for a more targeted approach by trying to look more broadly at watersheds. Corps customers, in fact, are requesting assistance in a wider range of water resources problems, such as water supply, water quality, and recreation. We are reviewing our watershed philosophy and

PAGE 21

6 New Directions for Corps Water Resources Programs investigating our potential involvement in these areas. We do have several watershed studies underway right now. The Corps has developed guidance for these studies and we're also looking very strongly at more collaborative interagency approaches. We are interested in building better state, local, tribal, private, and other federal agency partnerships. For example, one of our major priorities right now is the Everglades restoration in South Florida. That is a major interagency initiative between the Corps of Engineers, the Departnlent of the Interior, the Environmental Protection Agency, the state of Florida, and local governments. We need to be more aggressive in this area. In Section 6 of our bill we are asking for additional targets for ecosystem restoration. Section 5 includes expanded authority for the Section 1135 program to allow us to construct environmental restoration projects on lands outside of Corps project areas. We believe both these sections are very important to the Corps watershed initiatives. We have in the past and we wiII continue to provide technical support and assistance to state and local governments in the area of floodplain management. Two of the most important programs the Corps has are the Flood Plain Management Services and Planning Assistance to States Programs. We use Corps expertise to help state, local, and tribal governments prepare their own plans and initiate their own actions to reduce future flooding. These programs deal with a wide range of activities and services and are very flexible. They generally do not have to follow the Principles and Guidelines that other Corps studies must follow. Both programs are extremely customer oriented, in that we can perfornl studies for what is needed and requested, and provide results directly to our customers. Customer satisfaction, as in all Corps programs, is our highest priority. The Corps Flood Plain Management Services Program was established to assist state, local, and tribal governments in planning for the wise use of the floodplain. The Administration increased flIDding to $10 million for FY1997, up from the $6.5 million appropriated in FY1996. Technical services and planning guidance are provided upon request to state, local, and tribal governments without cost. It's a very quick turnaround, a very quick response program. Many of tlle teclmical services are provided in one day, most of the special study efforts are finished within one year. We believe this is a very effective, highly efficient, and successful program. For exanlple, we perfonn joint Hurricane Evacuation Studies with the Federal Emergency Management Agency and the National Weather Service. We have completed 50 studies, and have 16 under way. We also perform flood warning/preparedness studies for state, local, and ribal govenunents, and have completed 60, with 20 under way. Other tudies include dam break flooding studies, flood hazard analyses, flood

PAGE 22

Davis 7 hazard mapping, and other studies related to floods and flooding. Since 1970 we have responded to over 1.25 million requests under this program for either general information on flooding and floodplains or more specific requests for information on floodplain management. The Planning Assistance to States Program is a very broad authority under which almost anything pertaining to water resources can be investigated, such as floods, droughts, wetlands issues, water supply and distribution, and floodplain management measures. This program has a great potential to have a strong linkage with our watershed planning initiatives. The Administration increased flmding for the Planning Assistance to States Program to $3 million in the FY 1997 budget, up from the $1.5 million that was appropriated in FY 1996. This w,ill allow us to increase technical and planning support to state, local, and tribal governments. The studies under this program are relatively small, so the program has limited political and fiscal support, although it has been a very effective program. These studies are cost shared 50% federal, 50% non-federal. The program plays a major role in many state, local, and tribal government water resources planning activities and decisions. Since cost sharing was authorized, 314 studies in 47 states and 15 tribes have been completed. In the Administration's WRDA bill, and in the House and Senate versions, we are expanding the program to look more broadly at watersheds and ecosystems. In January 1993 the President established a task force of nine federal agencies that developed a comprehensive 40-point plan for dealing with wetlands issues. The plan was designed to make wetlands programs more fair, more flexible, and more effective. There are really two parts to that equation: more fair and flexible to landowners, and more effective in protecting wetlands resources. Of the 40 initiatives, the lmderpinning of the plan was the formal adoption of the goals of achieving no overall net loss of wetlands in the short term and the achieving an increase in wetlands in the long tenn. TIle regulatory programs that we have in the federal and state government have kept the wetlands patient on life support. We have slowed the wetlands losses substantially. Since the mid 1970s, we have reduced wetlands losses from 200,000 to 400,000 acres annually down to perhaps 60,000 or 70,000 acres of annual losses now. We feel that it is time to take the patient frolll the emergency room into the operating room. To continue to slow the losses, we are looking more comprehensively at non-regulatory approaches. To do this, we must look at programs such as the Wetlands Reserve Program of the U.S. Department of Agriculture to preserve and increase the wetlands in the country. We have completed a number of the 40 initiatives of the President's plan, have several others underway, and a few more that we will start in

PAGE 23

8 New Directions for Corps Water Resources Programs the coming year. There are a couple of things that we have completed that are very important in floodplain management. First of all, we closed a loophole by regulation that basically should stop all ditching and draining of wetlands without a permit. Before this, there was a loophole in the Corps of Engineers regulations that allowed certain activities to go forward without permits. Another important area is that in December 1995 we issued the first definitive federal guidance on wetlands mitigation banking, and we feel this was very important to wetlands protection and floodplain management. This allows us to be more strategic in siting wetlands mitigation projects. This way we can target wetlands mitigation in the watershed where it needs to be to restore floodplains and lost wetlands functions. And, finally the third thing I will mention as part of the President's wetlands plan is a real initiative to more fully engage the states. In the next 3 or 4 weeks we hope to issue guidance on what we call programmatic general permits. We have worked very closely with the Association of State Wetlands Managers in developing the guidance. The guidance would encourage states to take a more active regulatory role allowing the Corps to divest some of its responsibilities to the states where they do a good job, where they maintain at least the same amount of protection the Corps has or maybe more. This would then allow the Corps to free up its resources to do other things for the environment, such as more comprehensive planning approaches in watersheds. Success for the Corps depends on developing strong partnerships in the state, local, and tribal governments, and the private sector. I think we have an excellent start, and will continue to do that. We must focus on nonstructural approaches. This will require some cultural change at the state and local level, but it will require some cultural change in my organization as well. In our decision-making and policy-making we must focus on the human environment and the natural environment as well. It must be an equal part of our decision-making process. And, finally, we've got to look comprehensively, moving away from the small-decision-by small-decision approach in order to make significant progress on a watershed or ecosystems level.

PAGE 24

Mitigation and Partnerships for Floodplain Management Shirley Mattingly Director, Region IX Federal Emergency Management Agency Although I had originally planned to be sitting with you and eating lunch during this part of the day, I do consider it a great honor to be representing the Federal Emergency Management Agency's Director, James Lee Witt, in delivering this address. I know you are disappointed that he is not here. Just be reassured that your disappointment is matched by his own. James Lee has asked me to offer his sincere regrets. Natural events, as you know all too well, do not always conform to our calendars. The fires in Alaska have been dominating his time since this past weekend, and I believe he has also had the honor of a 16-hour plane trip on Monday. Still, I know that he would like very much to be here today. I know that, because I know how strongly James Lee feels about hazard mitigation in general and floodplain management in particular. He wanted me to emphasize to you just how important FEMA's partnership with the Association of State Floodplain Managers has been in the past and continues to be. As most of you know, our history goes back to your beginnings as an association, back when the Department of Housing and Urban Development ran the National Flood Insurance Program (NFIP). But the fact that we go back a long way is not the reason the Association is important to FEMA. It is important for what we are going to do together in the future. Partnerships, like the one between the Association and FEMA, are what this conference is all about. I think James Lee's favorite themes are mitigation and partnerships. And it is very clear to me, as well as to James Lee, that the key to making mitigation happen is partnership: bringing together science, engineers, government, insurance and financial institutions, and communities to labor collaboratively for protecting their investment in the future.

PAGE 25

10 Mitigation and Partnerships for Floodplain Management Earlier today you heard Dick Krimm, our acting Associate Director for Mitigation. FEMA did not have a Mitigation Directorate until James Lee Witt created one. James Lee is the first FEMA director to make mitigation the cornerstone of emergency management and, perhaps not coincidentally, the first to be awarded cabinet status by the President. James Lee Witt brought to FEMA a vision of a safer future, and he has set out to make it tme, by forging new partnerships and strengthening existing alliances for mitigation. Because of Director Witt we now have a national mitigation strategy and we are focusing on mitigation in our current round of funding negotiations with the states. But we realize that it is always a lot easier to talk about mitigation than to make it happen. It is easier to comprehend how a good mitigation strategy makes sense, than to turn that strategy or policy into action. So later this afternoon, with Doug Plasencia and Claire Rubin, I'll talk about a few cases from FEMA Region IX where partnerships have resulted in good mitigation deeds, where millions of dollars-and much heartache-will be saved in future floods. Of course, we need to realize that it is a worldwide trend we are trying to reverse-a lmiversal trend of mOlmting disaster losses. If we study demographic trends for the next 15 years we see that more and more Americans will live and work in regions with significant risk from one or more natural hazards. Clearly, one of the most effective tools we need to make mitigation happen is an infonned public. People accepting personal responsibility. A public that demands safer communities in which to live and work. And governmental and corporate citizens who set a good example, applying the best mitigation practices to our own facilities and activities, to reduce our vuinerability. We can achieve these goals only through collaborative efforts with our partners. Working together is becoming even more critical as we attempt to get the skyrocketing cost of disasters under control. This growing burden of disaster relief helps explain why hazard reduction policies have come to the front burner in emergency management and before Congress. Since 1989, U.S. taxpayers have financed an unprecedented $ 20.7 billion (total federal bill) in basic hunlan assistance, response, recovery, and reconstruction activities after federally declared natural disasters throughout the United States. Over $20 billion in under seven years. Mitigation has got to be the solution. Without reducing vulnerability, society can only shift the economic burden of disasters, not lessen it. In the past, Congress has always voted for supplemental disaster relief bills by simply adding to the federal deficit. This changed when Congress approved an $8.6-billion disaster supplemental appropriation after the Northridge earthquake; that $8.6 billion came at the expense of other

PAGE 26

Mattingly 11 programs that had to be cut by Congress. From now on, disaster relief dollars must compete on a national level with every other federal budget item. As disaster relief receives increasing Congressional scrutiny, hazard reduction must begin to playa more prominent role in managing disaster losses. We do not want to deny necessary disaster assistance to any family, or to any commlmity for that matter, victimized by flooding or by any disaster. What we want to do is to keep people and communities from becoming victims in the ftrst place. 11mt is risk reduction, and that is mitigation. James Lee tells a story about being with President Clinton in Woodland, Washington, after the flooding there. In one part of town, they saw a street where one side was flood danlaged, while the other side was not danlaged because of flood mitigation measures that had been implemented. President Clinton noted to Janles Lee that that sort of prudent action needed to be done wherever the threat of flood truly exists. So President Clinton feels the same way about this subject as you and I do. We need to keep people from becoming disaster victims in the ftrst place. Our partnership with the Association is cmcial to FEMA's ability to achieve its emergency management goals. One of the most visible signs of this partnership has been with the Conmmnity Assistance Program (CAP). TIle CAP helps us bring the NFIP to local communities. The CAP also helps enhance, or even in some instances create, your partnership with the local jurisdictions and it enriches the NFIP's relationship to each community. There are many of you who have done outstanding work in mitigating floodplain hazards, but I would like to take a moment to recognize one individual who has done a tremendous job in fostering this FEMA Association partnership while building relationships with local officials. Andy Lee, the CAP coordinator for California's Department of Water Resources (DWR) and our host for this Association conference, was an early il111ovator in active conmllmity education for floodplain management awareness. His organization has published the consistently excellent newsletter, Floodlight, for years and Andy has personally led the development of an outstanding floodplain management display booth during the two-week state fair in Sacramento. The booth brought together Doth FEMA and state DWR people to staff it, and Andy effectively coordinated with the Sacramento area Corps of Engineers to co-locate flood infornlation displays at the fair in order to maximize the visual impact. As a result of Andy's hard work, thousands of state fair visitors have learned the value of floodplain management.

PAGE 27

12 Mitigation and Partnerships for Floodplain Management These are exactly the kind of partnering activities that we must continue to replicate and build upon. One element of partnering that has made me, as a regional director, particularly proud of state floodplain managers has been the way in which state floodplain managers have been working with and partnering with the state emergency management agencies. As you are probably already aware, the performance partnership agreements (PPA) that are being implemented in every state are an attempt to more accurately address the specific emergency management needs of each state. The intent of the PPA is to place responsibility for risk management in a state with that state's governor. It is our hope that the state governors will use a collaborative approach and tap the most appropriate state agencies for accomplishing that state's goals. Which brings me to the second main point that I wanted to related to you today. If we are asking the governor of each state to be responsible for an appropriate level of risk management in his or her state, then it follows that we must find a way to create objective accountability for that effort. With a basic tenet of the PPA being the goal of reducing the administrative and project monitoring burden on the recipients of grant funds through increased "freedom" by the states, a natural question arises as to what consequences wiIl be faced by states that do not reach their goals. This philosophical change is viewed against a backdrop and national mood calling for more accountability in the way government spends money and an ever-shrinking funding pool resulting from efforts to reduce the deficit. Several years ago, Congress passed the Government Performance Results Act (GPRA). The GPRA requires a strict scrutiny of how money is being spent in relation to tangible results in an agency's performance. Concern about post-disaster payouts will continue to increase and the Director has already committed to defining criteria for the declaration of disasters by 1997. Although the NFIP funds do not come from the general revenue fund of the federal government, the current mood of the American public will not allow Congress to ignore NFIP funding levels if it feels too much is being spent in this area without sufficient results. So as you can see, as the concern for expenditures and corresponding results grows, so will the need to define appropriate performance standards. The Association wiIl undoubtedly playa large role in helping us to define observable, demonstrable results for the NFIP and will consequently influence our definition of performance measures under the PPA and the CAP.

PAGE 28

Mattingly 13 As our profession-floodplain management-gains respect, credibility, and support in the field, we must all rise to the challenge to keep growing, keep improving. We must become full participating partners in all levels of emergency management-before, during, and after a disaster. Take advantage of the training programs provided by your state and FEMA. TIrrough our PPA we have developed and offered workshops and instructional material on floodplain management. Also, please take advantage of our regional staffs that are always available to work with you, to provide technical expertise, equipment, and other resources. As FEMA negotiates mitigation memoranda of understanding with the states, I challenge each and every floodplain manager to play an active role in helping to formulate your state's mitigation plan. I urge you to work in partnership with your local and state emergency management directors in making sOlmd floodplain management a critical part of the state's comprehensive mitigation plan. You can help reduce community flood risks by continuing to monitor NFIP compliance. Encourage your local elected officials not just to adopt but also to enforce building and zoning codes and floodplain ordinances and to meet and even exceed the minimum NFIP requirements. Educate the public about the importance of floodproofing their properties, and maintaining their flood insurance policies. Take advantage of community awareness being built through the Cover America campaign of the NFIP: "When the flood waters washed away everything we had, I didn't think it could get any worse. But when the waters receded, THAT was the worst." With the creation of the Flood Mitigation Assistance Program that Dick Krimm talked about this morning, FEMA can now establish a pre disaster mitigation program. You can help make this a success by working with your conununities to identify worthwhile projects, and applying for mitigation funds through the state. The Association has been an important partner of FEMA and emergency management for the last 20 years. Your efforts have earned the respect and admiration of everyone who has had the chance to work with you and to experience your energy and dedication to flood risk reduction. We look forward to another 20 years of productive partnership with the Association.

PAGE 29

Review of Literature on Federal Hazard Mitigation Efforts (1979-1995)* Claire B. Rubin Claire B. Rubin and Associates Since the Federal Emergency Management Agency (FEMA) was formed, it has been the focal point of many mitigation policies, programs, and activities that have been initiated and/or implemented. In the past year or so, FEMA has given the topic of mitigation a high priority: it has established a Mitigation Directorate, in which mitigation for all progranls is housed, and devised a new National Mitigation Strategy, by means of a process with extensive public participation. Mitigation is not a new concept or program area, however; much has been known about mitigation for a long time. What were the experiences with mitigation progranls and projects, and what were their outcomes? To what extent has experience wilh policies and implementation over the last two decades been analyzed, evaluated, and incorporated into present actions? What are the positive or negative results of progranls in place? To answer that question, I llldertook a review of existing literature on mitigation during the past 20 years. A review of some of the major documents completed during the past two decades reveals the following: (1) Many excellent studies docllllent problems and issues, especially on mitigation implementation. The knowledge base is good. What stands out is the repetition of basic problems and issues over the years. This essay gives persO/U/1 observations alUt interpretations of a review of existing literature on federal mitigation efforts. It is based on research done for Rutherford H. Platt of the University of Massachusells, in conjunction with his research supported by the National Science Foundation, which will be published in a forthcoming Study of Hazard Mitigatioll. The observations and opinions expressed here are my own, mul not necessarily those of any of the supporting organizations or individuals.

PAGE 30

Rubin (2) Congress has initiated most of the major studies, the majority of which are critical of federal mitigation efforts to date. FEMA has initiated very few assessments of mitigation programs, plans, or implementation efforts. In other words, there is no shortage of problem and issue identification regarding mitigation of hazards; but there has been a shortage of corrective actions, particularly regarding implementation of mitigation. 15 Some of the key studies completed since 1979 are briefly described in the following sections. These studies, for the most part, have effectively identified needs regarding mitigation planning and implementation. Over the years, some of the recommendations have been carried out, but many flmdanlental needs have remained unmet. A key finding of my research: after about 25 years of federal involvement in floodplain management program, after almost 20 years since the formation of FEMA, and after almost 10 years since the Stafford Act was passed, federal agencies, and particularly FEMA, are still struggling to define, achieve, and evaluate their efforts and investments in natural hazard mitigation. THE APPROACH This review of mitigation as a strategy of hazard loss reduction from 1979 to date covers mainly floods and earthquakes. While a number of reviews, studies, and analyses have been done at the general level-affecting all of FEMA or concerning all types of hazards/disasters-most of these reviews and analyses tend to focus on either floods or earthquakes. Experience with these two major disaster agents should be indicative of experience with mitigation of hazards generally. Furthennore, each of these two has a major constituency or clientele and has a related research conmllmity that regularly follows and documents the changes, needs, issues, and research needs in each area of emergency management in connection with large national disasters. A great deal of hazards/disaster literature deals with (1) flood events (riverine, coastal and hurricanes), and (2) earthquake and related ground failure events. Not many studies have compared the mitigation efforts across two or more program areas; and in fact, the FEMA staff have in past tended to fall into one category or the other and not ranged across programs. Actually, this tendency to specialize in one hazard area is true of both public practitioners and researchers. TIle fonnation of the Mitigation Directorate at FEMA was supposed to help foster a broader perspective. This literature review documents some of the major public policy milestones along the federal mitigation highway that has been under construction since the late 1970s. It tries to examine how various

PAGE 31

16 Review of Literature on Federal Hazard Mitigation Efforts programs incorporated mitigation into their policies, programs, and activities through the years; and what assessments of the effectiveness of mitigation programs, projects, and educational efforts have been done. Problems with mitigation remain, though. As was noted in a recent national report (National Academy of Sciences, 1994, p. 3), "Mitigation has been an underlying requirement of federal emergency management policy for about 30 years, beginning with floodplain management requirements in the 1960s. In actual practice, however, only a fraction of the mitigation measures known to be effective have been implemented." Another experienced mitigation researcher, Raymond Burby, in reflecting on the aftermath of the 1993 Midwest floods, has stated that Federal agencies have not effectively used existing knowledge on private-sector decisions related to hazard mitigation and, except for the National Science Foundation, have not known enough to invest in building knowledge about floodplain management that would enable them to deliver programs more effectively. As a result, some federal programs have not penetrated private markets adequately (flood insurance, for example ... and many opportunities to foster private retrofitting ... are lost due to the absence of infomlation about to act effectively (1994, p. 44-47). BRIEF HISTORY OF MITIGATION IN FEDERAL POLICIES AND PROGRAMS Beginning in 1966, a mitigation policy was referenced in an Executive Order, which called for reduced development in floodplains to cut flood losses. Then in 1973, the federal emphasis on mitigation was increased in the Flood Disaster Protection Act, which said federally insured loans in communities that did not meet the requirements of the National Flood Insurance Program (NFIP) would be cut off. In 1974 the Disaster Relief Act made mitigation a prerequisite for receiving federal disaster aid. Specifically, public jurisdictions receiving aid were required to agree that "the natural hazards in the areas in which the proceeds of the grants of loans are to be used shall be evaluated and appropriate action shall be taken to mitigate such hazards, including safe land-use and construction practices ... (Federal Emergency Management Agency, 1994, p. 7). The intent was to prevent the recurrence of the disaster or to reduce its impact. In the case of flood disasters, it has further been required since 1980 that an "interagency hazard mitigation team" be assembled to prepare a report within 15 days after the disaster declaration "recommending specific recovery actions to be taken by each federal agency and each nonfederal level of governnlent. Federal agencies shall conform their re-

PAGE 32

Rubin covery actions to the recommendations of the report to the fullest extent practicable" (Office of Management and Budget, 1980). 17 In 1977, the National Earthquake Hazards Reduction Act (pL-95-124, as amended) involved the federal government in earthquake mitigation and provided resources for developing and implementing measures to mitigate earthquake hazards. Then in 1979, FEMA was created, with the intention of coordinating federal disaster programs within one agency. While mitigation has been receiving a great deal of public attention lately, it is not a new concern for many members of the hazards and disaster community. Many significant mitigation efforts were underway before the formation of FEMA. Two programs that included major mitigation components, the National Earthquake Hazards Reduction :Program (NEHRP) and the NFIP, predated the creation of FEMA. MILESTONES FOR ACHIEVING MITIGATION Some of the key requirements, actions, and experiences that significantly influenced federal level disaster mitigation policies and programs fall into three broad categories. I characterized the three major types as milestones. They include the following: (1) Organizationai/lnstitutionai Milestones-enabiing legislation, executive orders, program or organizational developments in FEMA; (2) Disaster Event Milestones-major or catastrophic disaster events that, because of their size, problems, or consequences, have become defining events for the emergency management conununity nationally; (3) Mitigation Assessment Milestones-such assessments occur in a wide variety of disaster evaluations, Congressional studies and reports, and other significant analyses or evaluation docunlents. Some of these have significantly influenced, either directly or indirectly, policy makers and legislators responsible for emergency management. In reviewing mitigation needs, accomplishments, and problems since FEMA was formed in 1979, it is my opinion that most of the assessments, evaluations, or other efforts to deternline effectiveness of policies and programs were taken in response to events and external pressures rather than internally initiated by FEMA program office staff. My impression is that mitigation efforts at the national level have been shaped by significant disasters, certain major studies and reports, and progranunatic and organizational changes in response to events. Table 1 shows the three types of activities that have been identified for the past 18 years. It depicts the major legislative/executive measures

PAGE 33

18 Review of Literature on Federal Hazard Mitigation Efforts Table l. Major milestones on the federal mitigation highway. YEAR ORGANIZATIONAL FLOODS! EARTH-CHANGES HURRICANES QUAKES MAJOR REPORTS 1978 NEHRP enacted Northeast BliZZllrd OSTP "Earthquake Hazard Issues for Implementation Plan" 1979 FEMA created Hurricane Frederic 1980 Memorandum of AgreeHurricane David NSF. A Report on Flood Hazard ment on Interagency Mitigation" Hazard Mitigation Task Burby el. al.. "Evaluation of Local Force (Floods) Experiences with Flood Plain Management" 1981 FEMA, "Evaluation of Economic, Social and Environment Effects of Flood Plain Regulations" 1983 Burhy & Cigler, "Effectiveness of State Programs for Floodplains" 1986 FEMA. "A Unified National Program for Floodplain Management" 1988 Stafford Act enacted 1989 Hurricane Hugo Loma NHRAIC, "Report of the Colorado (USVI, PRo SC, Prieta (CA) Workshop on Hazard Mitigation in NC) 1990's" 1990 Executive Order #12699 (seismic safety) 1991 GAO, "Federal, State and Local Response to Natural Disasters Needs Improvement. .. FEMA, "Financial Incentives for Seismic Rehab. of Hazardous Buildings. 1992 Hurricane Andrew (FL, LA) Hurricane Iniki (HI) 1993 Stafford Act amended Great Midwest FEMA, Report to Congress, (change in percentage Floods (9 states) "Improving Earthquake Mitigation" allocated for mitigation) NAPA, Report to Congress, "Coping Formation of Mitigation With Catastrophe" Directorate at FEMA GAO & CRS Reports to Congress NEHRP Advisory Committee Report 1994 Executive Orders Northridge Galloway Report on Midwest Floods #12941 and # 12699 (CA) FEMA IG, "Audit of FEMA's (seismic safety) Mitigation Programs" National Flood Insurance Reform Act 1995 U.S. Senate, Bipartisan Task Force on Funding Disaster Rolief, Federal Disaster Assistance. A hihliography of reports mentionc:u in (hi:.; tlhle is avaiJaol!! from the author.

PAGE 34

Rubin 19 that require mitigative actions; key disaster events; and some major, influential reports. While one cannot ascribe direct causal relationships, it is interesting to note what seem to be the influential effects of the major studies, each of which was critical of the federal response to one or more recent, major disaster events. For example, deficiencies in federal response to Hurricane Hugo (1989) and Lorna Prieta (1989) were noted by the U.S. General Accounting Office (GAO) in its 1991 report. Then in 1992, very serious problems with federal response to Hurricane Andrew generated several studies in 1993. The criticisms contained in the 1993 reports seem to have influenced both legislative changes and organizational changes at FEMA in late 1993 and in 1994. What followed was a discussion of the major milestones for the late 1970s, the decade of the 1980s, and the first five years of the 1990s. (More details will be available in Platt's forthcoming report.) OBSERVATIONS Here are some observations on, first, the literature and second, the relevance of the results of that effort to this audience. A review of various mitigation assessment reports from the last 20 years gives the impression of "deja vu all over again." The various reports seem to offer similar reconm1endations, even though their publication dates vary by almost 20 years. Some older reports (pre-FEMA) on mitigation are quite definitive about already identified needs regarding mitigation and its implementation. Similarly, the "Report of the Colorado Workshop in Hazard Mitigation in the 1990's," which was completed in 1989, also has a confident tone and emm1erated in a definitive way many constraints to mitigation. By contrast, the most recent mitigation efforts by FEMA-the 10 national fonill1s held during 1994, a new National Mitigation Initiative, and plans for a public conference-the First Biennial National Mitigation Conference in 1995-seem more tentative and lmcertain about what should be done next. This less confident stance may be due either to the uncertainty that relatively new appointees to FEMA feel or reflect a loss of confidence about achieving mitigation on the part of the agency's professional staff. The dilemma remains: it may take decades to see the results of mitigation measures that have been implemented and are successful, yet during those decades, several political administrations and several appointed administrators for the key mitigation programs will have come and gone at each level of governn1ent. How can we retain the institutional memory and capacity to achieve mitigation?

PAGE 35

20 Review of Literature on Federal Hazard Mitigation Efforts Since its inception, FEMA has given little attention to conducting or using research results or to conducting and using evaluation results. The National Academy of Public Administration study addressed the fonner, and the GAO has criticized FEMA for the latter. The NAPA team said, "FEMA's attitude toward sponsoring applied research, using outside research, and incorporating research results into opera tion, training, and educational efforts ought to be reviewed. FEMA has made little effort to use emergency management research results to improve state and local capacities .... This lack of a long-term plan for research and development as well as any systematic plan for the inclusion of new research results and findings into operational and training progran1s, are additional reasons why the agency is not at the cutting edge of its mission" (NAPA, 1993, p. 97). A secondary effect of the fact that the agency does little of its own research is that agency staff may not find the results produced by others of relevance and use to them. The role of the disaster research conununity in developing the National Mitigation Strategy is not known at this time. The most recent reports, by the National Academy of Sciences (1994) and the U.S. Senate (1995), indicate there is a lot of unfinished business regarding mitigation, especially if one is planning to undertake more precise forn1s of analysis, such as cost/benefit analyses and risk assessments. Constraints to hazard mitigation have been identified and listed many times, e.g., Natlll'al Hazards Research and Applications Information Center (1989); seven NEHRP documents done in the 1990s; and various flood reports. Why are there so many identification analyses and so few implementation efforts and impact analyses? Many major problems regarding mitigation implementation remain: There is a lack of strategic thinking and planning at each level of government-not only for mitigation but also for recovery. We are coming closer to the block grant approach to mitigation, via the Hazard Mitigation Grant Program, but lack broad-gauge planning and implementation capacity to effectively proceed. There is too much micro-level thinking and not enough macro-level thinking. Some of the root causes for this are: (a) Financial and fiscal problems persist; there are inadequate incentives for state and local officials to assmne responsibility for mitigation. (b) The needed institutional fran1ework is not in place for long-tern1 planning and implementation of mitigation.

PAGE 36

1ubin (c) Too many efforts are personality-driven and not program driven. (d) Often politics dominates, not rationality or conunon sense about local initiatives and priorities for mitigation. REFERENCES 3urby, Raymond J. 21 1994 "Floodplain Planning and Management: Research Needed for the 21st Century," Water Resources Update 97 (Autunm):44-47. Federal Emergency Management Agency 1994 Inspector General's Report. Reference to Sec. 406 of PL 93-288, as restated in Sec. 409 of the Stafford Act of 1988. Washington, D.C.: FEMA. National Academy of Public Administration 1993 Coping with Catastrophe. Washington, D.C: NAPA. National Academy of Sciences 1994 "Facing the Challenge." The U.S. National Report to the IDNDR World Conference on Natural Disaster Reduction, Yokohanla, Japan, May 23-27. Natural Hazards Research and Applications Infommtion Center 1989 "Report of the Colorado Workshop on Hazard Mitigation in the 1990s." Boulder, CO: NHRAIC. Office of Management and Budget 1980 Memo of July 10, as reflected in Interagency Agreement for Nonstructural Damage Reduction Measures, dated December 15. u.S. Senate 1995 "Federal Disaster Assistance." Report of the Bipartisan Task Force on Funding Disaster Relief. Senate Doc. No 104-4. Washington, D.C.

PAGE 37

This page IS intentionally blank

PAGE 38

Part 2 MULTI-HAZARD MITIGATION

PAGE 39

This page IS intentionally blank

PAGE 40

Mitigating against Flood and Earthquake Hazards Michael Mahoney Federal Emergency Management Agency INTRODUCTION One of the primary goals of the Federal Emergency Management Agency (FEMA) is reducing the ever-increasing cost of natural disasters. The need to reduce this cost was one of the reasons behind the development of our National Mitigation Strategy. In fact, the main goal of the strategy is to reduce losses resulting from natural hazards, such as earthquake, winds, and flooding, by at least half in the next 15 years. In order to be able to reach that goal, we as a nation will need to examine how we are designing and building stmctures. To help encourage building practices that can reduce the threat presented by different natural hazards, FEMA is committed to working with the design and engineering conununities and the nation's model code organizations to encourage the use of adequate loss reduction design standards. To date, FEMA's efforts in this field have taken several forms. I To address the flood hazard, Congress fonned the National Flood Insurance Program (NFIP) in 1968. The NFIP is based on the adoption and enforcement of specific flood-resistant design criteria by local conununities as part of the quid pro quo for the availability of federal backed flood insurance. To assist in the use of that design criteria, FEMA has funded the development and publication of teclmical guidance materials. We have also worked directly with the model code organizations to have most of tllis material incorporated into the nation's model building codes. To address the seismic hazard, Congress fom1ed the National Earthquake Hazards Reduction Program (NEHRP) in 1977 to coordinate the federal govenm1ent's role in addressing the earthquake hazard. The NEHRP is made up of four federal agencies (FEMA, National Institute of Standards and Technology, National Science Foundation, and the U.S. Geological Survey). FEMA, which is the lead agency for NEHRP, has worked with an outside organization, the Building Seismic Safety Council

PAGE 41

26 Mitigating against Flood and Earthquake Hazards (BSSC) to develop a resource docmnent for use by building regulatory organizations. This docmnent, the NEHRP Recommended Provisions for Seismic Regulations for New Buildings, was developed using a consensus procedure, and has now been either adopted or utilized by all three of the model code organizations. FEMA has also developed technical resource docmnents to address other aspects of the earthquake hazard, including several documents on how to identify and rehabilitate existing structures and addressing seismic issues specific to critical occupancies. Congress has recently called for the organization of a National Earthquake Loss Reduction Program, or NEP, and again tasked FEMA to take the lead. This new program is based on a report prepared by the President's Office of Science and Technology Policy that called for a new program with increased emphasis on earthquake mitigation. To do this, the NEP will add over a dozen additional federal agencies that also have earthquake-related responsibilities to the original four NEHRP agencies. MULTI-HAZARD APPROACH TO MITIGATION While these approaches to addressing the flood and earthquake hazards have generally been effective in their own way, they are not as successful as they could be. One reason why is that, until now, each approach only dealt with its own hazard. Until recently, both of these programs generally developed design criteria that only examined the impact on the hazard being addressed, and ignored what the impact might be on other hazards. There is a growing awareness that mitigation needs to move beyond a single-hazard focus and look at the impact of all hazards. For mitigation to be effective, coordination among the hazards is needed to avoid conflicts where action taken to reduce the threat of damage by one hazard may increase damage from another. A conunon example of a potential conflict between the flood and earthquake hazards would be any structure built in a flood hazard area and a high seismic area. Such a structure would have to be built on an elevated foundation to raise the building's lowest floor above flood levels; yet this type of foundation is probably the most susceptible to being damaged in an earthquake. The reason for this is that seismic loads could cause an elevated fOlmdation to act as a moment arm, and greatly increase the shaking of the structure. However, it should also be noted that if such a structure and its elevated foundation were properly constmcted and reinforced to resist larger coastal wind forces, the additional steps needed to provide a seismically resistant foundation may well only consist of a strengthened pile or column-to-beam connection, such as the addition of knee bracing, a common coastal construction technique.

PAGE 42

Mahoney 27 J To be effective in addressing this problem of coordination among the different hazards, coordination needs to be done during the development 6f these individual standards. This coordination is one of the key f>mponents of a multi-hazard approach to mitigation. TECHNICAL GUIDANCE FOR A MULTI-HAZARD APPROACH FEMA has recently undertaken several projects to provide technical guidance for the design and construction community that address specific technical needs using a multi-hazard approach. The first project is the development of a series of publications designed to present construction guidance in a manner that can be used by home builders and other non engineers to mitigate the effects of specific natural hazards for oneand two-family dwellings. The docwnents will also include prescriptive building plans that comply with the model codes and are resistant to damage from natural hazards. While each volwne will address a specific hazard, related material on other hazards will also be provided. To date, the development of two docmnents is underway. The fust volmne will be the ''Home Builder's Guide to Seismic Resistant Construction." This volwne is actually an update of an existing FEMA manual; the Home Builder's Guide to Earthquake Design (FEMA-232). That manual was originally published by the Department of Housing and Urban Development (HUD) in 1980 based on data from the San Fernando earthquake and, while the information is still valid, it needs to be updated. The second volunle will be the "Home Builder's Guide to Wind Resistant Construction" and will address high wind and flood hazards in a coastal environment. This second document is meant to complement, not replace, the current Coastal Construction Manual. The development process of both documents will include a peer review to ensure that they contain material that is technically correct and complete. Reviewers will include home builders, contractors, engineers, architects, building code officials, and knowledgeable homeowners. This outside review will serve as a "reality check" for the final product. Both docunlents will make it very clear that they are not a substitute for using the local building code or working with the local code enforcement office, but instead are meant to be a resource to provide several techniques by which the home builder can meet the local building code in an manner that addresses the risk presented by that particular hazard. Both documents are scheduled to be published next year. The second project addresses residential structures that are exposed to flood and earthquake hazards. As described above, the NFIP flood design criteria require that all new constmction as well as all structures that have been substantially danlaged or improved, be elevated above anticipated flood levels. In a seismic area, such an

PAGE 43

28 Mitigating against Flood and Earthquake Hazards elevated foundation can increase seismic loads on the entire structure. To resolve the problem of how to elevate a floodprone building in a manner that still complies with current seismic building codes, FEMA has contracted for the development of a guidance document that will present basic designs for a series of sample building foundations that can be used to elevate a building above flood levels while being capable of resisting seismic and wind loads. The sample foundation designs provided in this publication will be in compliance with the seismic requirements of the NEHRP Recommended Provisions as well as all three of the model building codes. The designs will be for areas of both moderate and high seismicity and will be applicable for different flood conditions. The designs used will be representative of elevated foundations used throughout the country. The publication will present the foundation designs and supporting backgrOlmd information in a manner understandable by the non-engineer, since the target audience is the homeowner and his or her contractor. This effort is being performed under FEMA's National Earthquake Technical Assistance Contract (NETAC). Due to contracting problems, this project has been delayed, and we do not have a completion date at this time. The previously mentioned Coastal Construction Manual (FEMA-55) is another project that will be taking more of a multi-hazard approach. FEMA is planning to update this document starting next year, and the update will include some earthquake funding to include seismic considerations in the design process. Other projects include recent improvements to the American Society of Civil Engineers "Minimum Design Loads for Buildings and Other Structures." COORDINATION AMONG THE DIFFERENT HAZARDS The coordination of efforts among hazards will require more than a few guidance documents that take a coordinated look at several different hazards. To effectively coordinate these activities, the actual design criteria or standards need to be addressed. In late 1994, FEMA sponsored a two-day workshop that brought together experts from different hazards to discuss this issue. What came out of that group was a recommendation for an independent, non-governmental body capable of involving and coordinating the present standards-writing organizations as well as the various outside interests, such as the architects, engineers, materials interests, contractors and other similar groups. Such a body would need to recognize and utilize the existing consensus bodies that now address specific hazards, not to replace these established groups, but instead serve as a coordinating body. Possible models for this body included NIBS, which was formed to act as a coordinating body for government activities.

PAGE 44

Mahoney 29 As a first step, FEMA has funded the first phase of a proposed project, the goal of which is fonnation of a National Multihazard Mitigation Council. The first phase will consist of fonning a Multihazard Project Committee, developing an organizational structure, mission statement and procedures, and conducting a forum on the proposed cOlmcil. That committee will be meeting for the first time later this stUnmer. Assuming the project goes forward, the council that would ultimately result will be charged with helping to examine and coordinate the improvement of existing building standards, technical resource materials, and model code language that address wind, flood, and seismic hazards. CONCLUSION If we as a nation are going to reduce the ever-increasing cost of natural disasters, and meet the goal of the National Mitigation Strategy, we will need to examine how we are designing and building structures and better uccount for the different hazards that may be present. To do this, we ultimately will need to coordinate the activities of the existing groups that address these hazards. To help encourage this, FEMA is committed to working with the design and engineering communities and the nation's model code organizations to encourage a coordinated approach to the use of adequate loss reduction design standards.

PAGE 45

Promoting a Multi-hazard Approach when Retrofitting Floodprone Structures Clifford Oliver Federal Emergency Management Agency INTRODUCTION Often damaged structures are retrofitted in the wake of a major flood. Retrofitting often occurs when states and communities enforce floodplain management and building code requirements or as a result of voluntary action on the part of property owners. The focus of property owners, building officials, and the media is often on avoiding the recurrence of similar damage from future floods. This uni-hazard focus can lead to damaged structures being retrofitted to address flood hazards while ignoring the threat that other natural hazards such as earthquakes, high winds, and erosion can present. This paper will explore what needs to be done to address this issue and what activities the Federal Emergency Management Agency (FEMA) currently has underway to ensure that retrofitted building are designed and constructed to reduce damage from all natural hazards. THE FEDERAL EMERGENCY MANAGEMENT AGENCY With the 1994 reorganization of FEMA, the Mitigation Directorate was founded. This brought the mitigation component of the National Flood Insurance Program (NFIP), the National Earthquake Hazards Reduction Program (NEHRP), the National Hurricane Program (NHP), the National Dam Safety Program (NDSP), and sections 404 and 409 of the Stafford Disaster Relief Act together within one organizational lmit. The synergy brought on by the reorganization has allowed the Mitigation Directorate to focus on promoting multi-hazard mitigation strategies. These strategies are outlined in the recently completed National Mitigation Strategy (FEMA, 1996). The insurance component of the NFIP is the responsibility of FEMA's Federal Insurance Administration (FIA). The Mitigation Directorate works extremely closely with the FIA to maximize the promotion of mitigation through the insurance aspects of the NFIP.

PAGE 46

Oliver NEW PUBLICATIONS AND GUIDANCE DOCUMENTS Several new guidance documents have been prepared or are under development that provide information on how to deal with other hazards when retrofitting a floodprone structure. New Retrofitting Manual 31 FEMA completed development of a new manual, Engineering Principles and Practices for Retrofitting Flood Prone Residential Structures (FEMA, 1995). This manual focuses on identifying all natural hazards that will impact a structure and provides detailed guidance on designing retrofitting that will account for all known natural hazards. In conjunction with the manual, FEMA developed two training vehicles. FEMA is now offering a one-week resident course at the Emergency Management Institute at the National Emergency Training Center, Emmitsburg, Maryland, and makes available a two-day short course for use in the field. American Society of Civil Engineers Standard 7-95 FEMA funded a recently completed effort by the American Society of Civil Engineers (ASCE) to include flood loads in the national load standard entitled "Minimum Design Loads for Building and Other Structures," ASCE 7-95 (ASCE, 1995a). This recently completed standard includes, for the first time, detailed guidance on determining applicable Dood loads and how to compute load combinations. Model Building Code organizations are currently being asked to incorporate ASCE 7-95 into U1eir model building codes. New ASCE Standard on Design and Construction FEMA funded an effort by ASCE to develop a new standard on how to design and construct buildings and other structures to resist flood damage (ASCE, 1995b). One important aspect of this standard will be to provide detailed on guidance on how to design and construct buildings and other stmctures to resist loads determined through the application of ASCE 7-95. This standard is presently in a pre standard format and should be completed by 1997.

PAGE 47

32 Promoting a Multi-hazard Approach when Retrofitting New Manual on Elevating Structures FEMA is completing work on a new publication entitled ''Elevating Residential Structures to Resist Multiple Hazards." This new manual will provide a non-technical step-by-step process for the design of standardized elevated foundations. This project is well underway and should be completed in 1996. Disaster-Specific Guidance Documents Since 1992, FEMA has prepared disaster-specific documents to provide guidance on elevating damaged stmctures located in flood prone areas. These documents are intended to educate owners of damaged buildings and other structures and state and local officials responsible for overseeing and regulating reconstruction on how to ensure that the design of the retrofitting complies with applicable NFIP, local, and state floodplain management requirements as well as building codes and standards. Such documents have been prepared in response to Hurricanes Andrew and Iniki, the Midwest Flood of 1993, flooding in the Houston, Texas, and Albany, Georgia areas, and the Northridge, California, earthquake. POST-DISASTER ACTIVITIES Since 1992, mitigation staff have become an integral part of FEMA's disaster response activities. The creation of a Deputy Federal Coordinating Officer for Mitigation (DFCO-M) position within FEMA's Disaster Field Offices (DFO) resulted from the fonnation of the Mitigation Directorate. The DFCO-M is empowered to constitute a mitigation staff within DFOs. The mission of this staff is to ensure that reconstmction activities incorporate multi-hazard mitigation to the maximmn extent reasonable. The DFO mitigation staff coordinate mitigation activities that are carried out under tlle NFIP, NEHRP, NHP, NDSP, and the Stafford Disaster Relief Act. After federally declared disasters, mitigation staff work with individuals impacted by the disasters, state and local officials, and other federal agencies to promote multi-hazard mitigation. NEW REGULATIONS Pending revisions to the Hazard Mitigation Grant Program regulations (Section 404 of the Stafford Act) and the proposed Flood Mitigation Assistance Grant Program regulations (created under the National Flood Insurance Reform Act of 1994) both require that known natural hazards be taken into consideration when retrofitting floodprone structures. It is proposed that retrofitting projects will not only need to conform to the requirements of the NFIP, but also to a comprehensive building code that

PAGE 48

Oliver conSiders the effects of other natural hazards. This will be an important step forward in ensuring that federal mitigation funds are applied to projects that truly offer to reduce future losses from all known natural hazards. IMPLEMENTATION OF MULTI-HAZARD MITIGATION STRATEGIES BY STATES AND COMMUNITIES 33 Many states and communities have both a floodplain ordinance that complies with the NFIP requirements and a comprehensive building code. Enforced together, a NFIP-compliant floodplain ordinance and a comprehensive building code will result in retrofitted structures being designed and constructed to resist danlage from various natural hazards. Some states and communities that participate in the NFIP enforce a floodplain ordinance and either enforce an insufficient building code or have no building code at all. Further complicating matters, some states and communities have a comprehensive building code that is inadequately enforced. This is often due to states and communities either waiving floodplain management or building code requirements after a major disaster and/or a lack of state and commlmity commitment to enforce the building code. After major disasters, tremendous pressure can be brought to bear on state and community officials to relax rebuilding requirements. 111ese states and conummities present the greatest challenge to FEMA. A primary goal of FEMA is to affect change in attitudes in these states and comIl1lmities to promote multi-hazard mitigation during pre-disaster and post-disaster situations. CONCLUSIONS FEMA is working hard to change attitudes by promoting multi-hazard mitigation when retrofitting floodprone structures. This is being done both iIi the pre and post-disaster settings. FEMA is working with such groups a\ the American Society of Civil Engineers, model building code organizations, American Institute of Architects, Association of State Floodplain Managers, and the National Association of Home Builders to achieve this goal. REFERENCES American Society of Civil Engineers (ASCE) 1995a ASCE Standard 7-95; Minimum Design Loads for Buildings and Other Structures. Washington, D.C.: ASCE.

PAGE 49

34 Promoting a Multi-hazard Approach when Retrofitting 1995b "ASCE Prestandard: Flood-Resistant Design and Construction Practices for Buildings and other Structures." Washington, D.C.: ASCE (unpublished). Federal Emergency Management Agency 1989 Engineering Principles and Practices for Retrofitting Flood Prone Residential Structures. FEMA-259. Washington D.C: FEMA, Mitigation Directorate. 1996 National Mitigation Strategy. Washington, D.C.: FEMA.

PAGE 50

Malibu/Las Flores Canyon Watershed Hazard Mitigation Plan (Floods, Fires, Landslide) Bruce M. Phillips Robert Bein, William Frost & Associates INTRODUCTION AND PROJECT BACKGROUND Las Flores Canyon in the City of Malibu, California, has received national media attention over the last several years from experiencing several natural disasters including fire, a massive landslide, flooding, and mud and debris flows, which have threatened public safety and resulted in significant economic losses. TIle city recognized the need to address these hazards and the associated annual maintenance costs through the : development of a "Hazard Mitigation Program" for Las Flores Canyon, iwhich received public assistance flUlding from the Federal Emergency lManagement Agency (FEMA). i The 4.2-square-mile Las Flores Canyon watershed consists of a very : steeply shaped canyon that rises from sea level to more than 2,500 feet in elevation in less than five miles. TIle nature of the Las Flores Canyon watershed generates a high potential for large quantities of debris and : sediment production, also an extremely rapid nmoff from precipitation. : TIle floodplain of the lower canyon develops significant overflow flooding resulting frolll the limited hydraulic capacity of the creek. Quantifiable danlage has been associated with recent federally declared disasters during three separate stonns of significant rainfall. TIle 17-acre Ran1bla Pacifico ; landslide is located within the canyon, approximately 1/4-miles inland from Pacific Coast Highway and directly adjacent to the Las Flores Creek (Figure 1). The interaction of both the landslide and dynamics of the watershed results in a natural hazard potential that is significantly magnified. Recent accelerated movement in the landslide was influenced by the 1993 wildfire and large amounts of rainfall during the following winter seasons, which resulted in the landslide's continued encroachment i into the existing Las Flores Creek floodplain and significant erosion of the i slide mass increasing the state of the landslide instability.

PAGE 51

Yea, Floodplain InundA!IOf' Arf'a === I i\!; t Inres CAnyon Drtve 6/Ji:.o., ...::: Figure 1. Lower portion of Las Flores Canyon, inclicating the relative location of the Rambla Pacifico Landslide and the impacted properties from the floodplain and landslide. The Pacific Coast Highway at the mouth of the canyon is a major hydraulic restriction for the storm flows and a constant maintenance problem. Ci.) (j) 3:: !!?. 6"' c: ? 0" CD III o l ::J ::J CD iil ::r CD a. I Rl l a 3:: ;:;: cC o ::J iJ iii" ::J

PAGE 52

Phillips 37 Disaster History On November 2, 1993, a fire began six miles north of Malibu and raced southward towards the city over the next 30 hours, ultimately destroying 15,600 acres. All the vegetation within the entire 2,700-acre Las Flores Canyon watershed was burned, including all vegetation on the Rambla Pacifico landslide face. One of the effects associated with fire in a watershed is the significant increase in debris production potential inunediately after the fire, which can increase by a factor of five. Three significant rainstorms occurred in Febmary 1994, in the immediate aftermath of the wildfire, and the Las Flores Canyon Creek streambed was raised approximately 10 feet, with debris and sediment deposited on the Pacific Coast Highway. The cost of the cleanup for only a 2-year event was $1.5 million. The following winter reason resulted in several significant storms in January 1995, which also resulted in flooding and debris acclllmlation within the creek, along with significant maintenance cleanup costs. Also during March 1995 several high intensity storms caused road failures and debris deposition completely blocking the Pacific Coast Highway bridge. LANDSLIDE GEOTECHNICAL CONDITIONS TIle Rarnbla Pacifico Landslide occupies about 17 acres on the western wall of Las Flores Canyon, approximately 1/4 to 1h mile inland from the Pacific Ocean. On the basis of surface expression and surface and subsurface movement vectors, the landslide is composed of two different lobes, commonly referred to as the north and south lobes. These lobes possess different geotechnical characteristics and move at different rates. In cross section the slide is approximately 110 to 120 feet deep in the deepest central portion of each lobe. The denudation of the vegetation frolll the fire resulted in an exposed ground surface that was riddled with fissmes and provided a more direct path for infiltration into the slide mass, increasing grollldwater levels and accelerating landslide movements. Slump failures occurred along the toe of the landslide from scour and bank erosion from Las Flores Creek, resulting in increased sediment in the floodplain. These failures resulted in approximately 50,000 cubic yards being eroded from the slide toe in a single storm. An important concern of erosion from this portion of the landslide is the reduction in the resisting forces at the toe since the landslide generally attempts to naturally self-buttress and slow acceleration. EXISTING FLOOD HAZARDS Severe flooding generally occurs in Southern California watersheds during most rainfall events of any significant magnitude and the Las Flores

PAGE 53

38 Malibu/Las Flores Canyon Watershed Hazard Mitigation Plan watershed's unique dynamics magnify the runoff response and the impacts of sedimentation. The existing creek has the hydraulic conveyance capacity for smaller, more frequent rainfall events. Detailed floodplain mapping revealed that Las Flores Canyon Road would become a secondary flowpath, and this was verified through actual flooding. Hydraulic analysis also indicated that the bridge for the Pacific Coast Highway at the mouth of the canyon is a severe hydraulic restriction, with only hydraulic capacity for 30% of an estimated 50-year peak flow rate and results in significant deposition of sediment at this location. HAZAR D MITIGATION PLAN DEVELOPM ENT Developing an implementable hazard mitigation progranl for Las Flores Canyon required comprehensive planning techniques that not only addressed the specific natural hazards, but also the multiple issues from the diverse interest groups affected in the watershed. The planning process for the development of this specialized mitigation program relied upon establishing a solid technical fOlmdation by defining the baseline condition and developing a thorough understanding of the natural processes occurring in the watershed. The teclmical evaluation performed for the plan separated the geotechnical and the hydraulicjhydrologic processes into two independent engineering investigations. Project Objectives Long-term stabilization and/or control of the Ranlbla-Pacifico landslide, flood protection, and channel stability of Las Flores Creek with regard to sediment transport and erosion are the primary needs to be addressed by the proposed control measures identified in the hazard mitigation program. The mitigation measures identified for consideration enhance the level of public safety, while attempting to be compatible with and preserving the existing valuable natural resources. Constraints and Design Considerations Numerous design considerations were integrated in guiding the plan fonnulation, including (l) regulatory pennitting, (2) emergency access requirements, (3) traffic and circulation, (4) property acquisition and relocation, (5) coastal resources, (6) water quality, (7) riparian habitat, and (8) tile California Department of Transportation. ALTERNATIVE FORMULATION FornlUlation, evaluation, and selection of conceptual design alternatives considered the basic needs and constraints within this portion of Las Flores Canyon, in addition to most effectively meeting the project

PAGE 54

39 objectives. The fonnulation process also used consensus building among the agencies to address the diverse environmental and land use issues or constraints. A primary consideration in the development of the recommended mitigation program is funding. The project must be economically feasible and the improvements must be accomplished with the most economic means available, which is consistent with federal guidelines. A feasibility evaluation was performed for the various identified control measures. Essential technical basis; economic viability; environmental suitability; and legal, administrative, political, and other features of each alternative control measure were exanlined in this process. A primary concern that greatly influences the feasibility is the enviromnental acceptability of the plan because of the sensitivity of the canyon and coastal resources. Landslide Hazard Mitigation Alternatives TIle specific geotechnical alternatives evaluated for the landslide mitigation (Table 1) were measured by their potential to cost-effectively raise the "safety factor" to 1.25. That was considered to be the lowest acceptable factor of safety that would be suitable for long ternl stability. alternatives are the most cost-effective, long ternl, landslide control measures from a construction cost perspective and the "buttress option with off-loading" offered the lowest relative construction cost. Table 1. Geotec1mical mitigation alternatives for landslide. Do Nothing Dewatering or Partial Dewatering Structural Restraint Systems Retaining wall Shear pins or soldier piles -Tie backs Chemical Stabilization Grading Alternative Lay-back of entire slide face Buttress option Buttress option with off-loading Flood Control Alternatives The primary approach selected for the watershed management measures in Las Flores Canyon focused on conveyance-oriented measures, rather than

PAGE 55

40 Malibu/Las Flores Canyon Watershed Hazard Mitigation Plan a storage-oriented approach, because of the extreme physical constraints. Initial plan fonnulation reviewed several conventional flood control techniques, and also developed several innovative measures to address the unique multi-objective requirements and constraints within the floodplain (Table 2). The extreme physical characteristics of the watershed, including the hydraulic regime of the creek with high velocities and the significant an10unt of bed material transported, are primary contributing factors influencing the suitability of the control measures. Table 2. Alternative flood protection and sediment control measures. Loose rock rip-rap channel revetment Elevate PCH & reconstruct bridge Concrete line trapezoidal channel Natural floodplain Channelization with invert control Engineered eardlen clwmel structures Las Flores Canyon Road vertical Vegetative streambank realigmnent stabilization Ocean oudet reconfiguration Gabion channel revetment Reinforced box culvert Regional stornlwater detention Annorflex trapezoidal channel Composite/tiered channel section Reinforced concrete rectangular channel Multiple in-line debris basins IDENTIFIED HAZARD MITIGATION SYSTEMS Effective and implementable control measures for an acceptable hazard mitigation program were generated through a "systems" approach. It involved (I) a thorough evaluation of individual control measures for either geotechnical or flood protection, (2) perfonning a feasibility investigation of these control measures, and then (3) combining the feasible individual elements into systems that satisfied the various objectives, with varying degrees of hazard mitigation. The alternative hazard mitigation "systems" (Table 3) represent composite alternatives of

PAGE 56

Phillips 41 Table 3. Smnmary of alternative hazard mitigation system elements. All. Improved Debris Landslide Raise Las Replace Alternate Reconstruct Slide Floudwav No Channel Basin Buttress! Flores PCH RlUTlbla Rambla Dewater At:.quistuon Luyback Road Bridge Access A X 13 X X X C X X X X X X II X X X X X X X X X X X .-X X X Panlal X the various feasible control measures that would achieve a technically improved solution by satisfying several of the issues. The total costs associated with the implementation of the various system alternatives ranged from approximately $9.2 million to $27.2 million. The total costs included both constmction and property acquisition for the private land encumbered by the landslide and the floodplain. The property acquisition costs ranged from $9.2 million to $16.6 million and represent a significant component of the system costs.

PAGE 57

This page IS intentionally blank

PAGE 58

Part 3 MULTI-USE WATERCOURSES

PAGE 59

This page IS intentionally blank

PAGE 60

Planning a Future for the Salt-Gila Rivers: A Case Study in Designing a Master Plan Process for a Multi-use, Multi-purpose Watercourse Catherine A. Tice R. Keith Julian Woodward-C Iyde BACKGROUND Planning for and managing such a vital resource as an entire state's major watershed on a sustained-yield basis becomes especially challenging for multiple-use objectives. This paper describes a process developed by a consulting team (Woodward-Clyde) working in conjunction with 21 land use and resource management agencies (the Master Plan Participants) to design, plan, cost, and schedule the creation of a multi-purpose master plan for the Salt-Gila River watercourse between Granite Reef Dam and Painted Rock Dam in Central Arizona-a distance of approximately 100 miles. Can conmlOn agreement on watercourse management be reached when the multiple interests involved include such diverse water users and owners as Native American comrmmities, local governments, aggregate mining industries, a flood control district, and the U.S. Army Corps of Engineers? Nmuerous other federal and state management agencies were also resource stakeholders in the outcome. Maricopa COlmty is the urban center of Arizona with a population of 2.5 million (more than 60% of the entire state's residents) and a variety of land and water use objectives. However, the Salt-Gila watercourse begins and ends in environmentally sensitive riparian wetlands, wilderness, and wildlife refuge areas that require different management values and goals. The river system also varies physically, both seasonally and armually: in the heat of the Arizona summer the rivers are a trickle, but during the winter rainy season they have the power to knock out bridges and breach dams.

PAGE 61

46 Planning a Future for the Salt-Gila Rivers THE MASTER PLANNING PROCESS The Flood Control District of Maricopa County, along with repre sentatives of the local jurisdictions, the rock products industry, and a munber of federal and state agencies, had long recognized the need to develop a watercourse master plan to guide future land use, regulatory permitting control, and development along the rivers. All these diverse agencies and land use jurisdictions joined together in conmlOn purpose as the Master Plan Participants in 1991. After a previous estimate that the cost of producing a master plan would be close to $6 million, the Master Plan Participants decided to scope (i.e., design) a range of master plan options. The objective of the plan design process was to identify alternative approaches to master planning that could achieve some or all of the goals of the Master Plan Participants, and to estimate costs for these approaches. Five approaches-comprehensive, extensive, moderate, and limited master plans, as well as no project-were developed as part of the scoping effort. In addition, an effort was mounted to build a shared conmmnity vision for the project. A proactive public involvement program called for early identification of interested citizens, parties, and agencies and solicitation of their input to establish local issues and concerns. This pape,' presents the Master Plan design process, outcomes, present status, and likelihood of future implementation. The identification of alternative approaches for preparing a Master Plan and the selection of a preferred approach was accomplished over a 14-month period during which the Flood Control District, the Master Plan Management Conunittee, and the Master Plan Executive Committees worked together to develop a mission statement for the Master Plan and to identify and refine the following master planning goals and objectives: To develop a hydraulic master plan that evaluates and manages the risks of loss of life and damage to property within the 100-year floodplain. To identify existing conditions and assess future impacts of development on the natural and hlUnan-made environments. To strive to develop consensus among participants on river management issues and plans. To maintain, protect, and enhance environmental quality and integrity. To streanlline and coordinate regulatory policies and procedures. To produce a master plan that may be adopted by the Flood Control District Board of Directors and other jurisdictions, and to adopt uniform plan-based land use ordinances and/or regulations for enforcement.

PAGE 62

!Tice and Julian 47 Design of a Preferred Approach to Master Planning t. arly in the process a public involvement program was developed, with ,an implementation schedule based on the major milestones of the eclmical work tasks. Another important early task identified existing sources of technical, institutional, and environmental information on the master plan area. Evaluation of the quality, volume, and currentness of this data would be used to help determine whether new studies and additional infornlation gathering would be needed before the Master Plan could be prepared. Identification of those institutional and regulatory issues that presented potential opportunities or constraints to Master Plan development was undertaken following completion of the annotated bibliography. The preferred content of the Master Plan was identified through discussions with the Management Committee and from feedback provided at the public meetings. The specific areas of interest and concern identified were: implementation of an enforceable Master Plan; .. flood control/floodplain management; o streamlined pernlitting process (National Environmental Policy Act (NEPA), Section 404, Clean Water Act (CW A)); identification of cunmlative impacts; water quality/water resources management; reclamation of aggregate mining facilities; G cleanup of landfills; o habitat management; .. envirorunental enhancement; economic benefits; and e recreational uses. Creating a Range of Master Plan Approaches and Alternatives Using findings of early tasks (annotated bibliography and identification of institutional, regulatory, technical, and social issues), as well as the guidance provided by the Management Conmlittee in the mission statement and goals and objectives, a report was prepared that presented five master plan options, each of them varying in degree of completeness anj breadth of coverage, as well as potential costs. The range of options were designated: comprehensive, extensive, moderate, and limited master plan, and no project. The report also contained a sunmlary of master plan option features, estimated costs, and schedule for development and implementation (see Table O. After careful deliberations, a consensus was

PAGE 63

48 Planning a Future for the Salt-G ila Rivers Table 1. Comparative summary of Master Plan approaches. MASTER PLAN TYPE ESTIMATED COST ESTIMATED COST TO IMPLEMENT (TO PLAN) SCHEDULE OVER 20YEAR(TO PLAN) PERIOD Comprehensive Master Plan $ 15 -$ 20 Million 80 + Months $ 75 Million + (minimum) Extensive Master Plan $ 8 -$ 10 Million 60 80 Months $ 30 $40 Million (minimum) Moderate Master Plan $ 4 -$ 7 Million 36 60 Months $ 10 $15 Million Limited Master Plan $ 2 -$ 3 Million 18 36 Months $ 7 $10 Million No Project $ 0 --Many millions In future damage to environment and lost economic opportunities reached that the "moderate" approach offered the most attractive combination of features to the greatest munber of stakeholders. The consultant was then directed to proceed with preparing a detailed scope of work, estimated cost, and schedule. Features of the Moderate Master Plan Option The intent of the various elements of the moderate Master Plan was to achieve the maximum benefits from a master plan within a reasonably short (3 to 6 years) time frame and at a fundable cost. The moderate Master Plan included "something for everyone" in that it addressed all the areas of concern that were identified; however, no one was able to achieve all their objectives. The moderate Master Plan called for the fonnation of a management entity early in the process. The management entity would coordinate and direct the plan development and implementation. The entity could be an existing Maricopa COlmty agency or jurisdiction or a new entity composed of representatives from the Master Plan Participants. There was an assumption that the management entity would coordinate the local Clean Water Act Section 404 permitting and monitoring on behalf of the Master Plan Participants. Enforcement of pennit conditions and land use plans would remain with existing agencies (e.g., the Corps, the U.S. Environmental Protection Agency). I I

PAGE 64

rice and Julian 49 Fonnal NEPA review, possibly in the fonn of a Progranunatic Environmental Impact Statement, would be completed during the Master flan development process, thereby allowing for future plan modifications without major new environmental studies. The NEPA process would also thy the groundwork for the development of site-specific, detailed plan elements that could be modified and adopted by individual jurisdictions. The moderate Master Plan would also create a framework for close coordination of land use planning between jurisdictions within the Master Plan area. The management entity would have coordination and oversight responsibility for plan implementation, but no authority to mandate compliance. TIle outcome of the planning process and adoption of a Master Plan would involve producing a series of detailed, coordinated technical sub plans, called plan elements. Specific objectives of each participant could be achieved through the refinement to meet local needs of these elements, which would be subsequently adapted as part of the general plans of Master Plan Participants. Examples of possible elements are: comprehensive flood control and hydraulic management guidelines, including maintenance of a watercourse hydraulic model; lmifoml water quality guidelines for discharge, recharge, and withdrawals; unifonn guidelines for sedimentation control and aggregate management; lmifonn recreation management guidelines; natural resource management guidelines; and sharing planning and environmental infom1ation anlOng jurisdictions and agencies and creating a repository for relevant infom1ation. Based on the expressed interest and support of resource management agencies and the interested public, the Master Plan Participants would designate key actions to be carried out as part of planned improvements or enhancements of watercourse environmental characteristics. These enhancements would be identified during phase II of the planning process so that the required actions would be incorporated into the Master Plan concept and subsequently considered in the regulatory review process. The moderate Master Plan would encourage (though not mandate) a number of environmental enhancements to the Master Plan area, such as: acceleration of cleanup of landfills, hazardous wastes, and other environmentally degrading "hot spots"; creation, restoration, and management of habitat where feasible;

PAGE 65

50 Planning a Future for the Salt-Gila Rivers identification, creation, and management of offsite environmental mitigation opportunities; identification of clUllulative environmental impacts and benefits; establishment of a streamlined environmental review process for any plan-conforming development proposals; identification of opportunities for wildlife enhancement and recreational "eco tourism"; reclamation of aggregate mining areas; and creation of an awareness of environmental impacts of upstream users/jurisdictions on downstream users/jurisdictions. Institutional Constraints and Concerns Various planning issues and considerations needed to be resolved under the moderate Master Plan approach. While a moderate plan would be more institutionally feasible than a comprehensive or extensive one (since less local authority would be relinquished), there would be no enforcement authority granted to the management entity to implement the plan. Any enforcement would result from requiring proposed developments or land use change within the Master Plan area to conform to all plan conditions in order to obtain expedited pemlits and environmental approvals. The moderate Master Plan would also require a smaller funding commitment and agency support than the comprehensive or extensive plans. Adopting a moderate plan would involve lower up front costs before benefits could be determined. The moderate Master Plan, including development of plan elements and environmental enhancement guidance, could be completed in less than five years. Based on the assunlptions and scope in the final report, developing a moderate Master Plan would cost about $7 million (planning and approval costs only). The estimate assumed that phase II would be primarily funded by Master Plan Participants (although funding might be available through other sources) and that phases III and IV could be funded by state and federal grants or special appropriations. Only about 10% of the total cost would have to be borne by the local participants. Once the final Master Plan scoping report was submitted to the Master Plan Participants, they opted to take a much more conservative approach to implementation. A Master Plan Task Force was established under the auspices of the Maricopa County Association of Governments. The Task Force members are largely representatives of local government and for the past two years have worked on developing lmifoml land use elements as a precursor to a Master Plan. No plans or funding to implement the moderate Master Plan concept have emerged to date.

PAGE 66

Reclaiming Denver's Central South Platte River Leo Eisel Brian Kolstad McLaughlin Water Engineers, Ltd. Ben Urbonas Urban Drainage and Flood Control District Nick Skifalides Wastewater Management Division, City and County of Denver INTRODUCTION TIlt: South Platte River reaches from the 14,000-foot peaks west of Denver more than 300 miles east to its confluence with the North Platte River at Norih Platte, Nebraska, As the South Platte River passes the Denver metropolitan area it flows through approximately 10,5 miles of the City m,d County of Denver. This reach has been totally modified as the city has grown and no longer resembles the South Platte River of the past. The reach provides an opportunity for implementation of multi-purpose water resource projects and policies that, at least in part, reclaim the central South Platte River by providing improved aquatic and terrestrial habitat, recreational opportunities, and increased flood carrying capacity, The lO,5-mile reach of the South Platte River through Denver is totully urban and includes: 19 parks adjacent to the river; 12 miles of recreational trails immediately adjacent to the river and connecting to another 6 trails with an additional 50 miles of trails; A wastewater treatment plant and a decommissioned wastewater treatment plant; Two electric generating power plants; 10,5 miles of riparian habitat; and

PAGE 67

52 Reclaiming Denver's Central South Platte River 10.5 miles of wann water aquatic habitat. Surrounding this reach is a metropolitan area with more than 2 million people who demand numerous goods and services dependent on the South Platte River and its corridor including: Flood hazard mitigation; Municipal, industrial, and irrigation water supply; Recreation and aesthetic values; and Riparian and aquatic wildlife habitat. MULTIPLE-PU RPOSE PLANNING REQUIRED In order to meet these goals and provide the numerous goods and services dependent on the South Platte River, the Urban Drainage and Flood Control District has cooperated with the City and County of Denver to conduct multiple-purpose water resource planning for the South Platte River. Providing flood hazard mitigation and reclaiming Denver's central South Platte River requires responding to the many needs and demands while insuring that existing stakeholders including communities, industries, a power company, recreational users, fish and wildlife agencies, environmental regulatory agencies, and existing residents are kept whole in the process. At the san1e time, constraints and limitations exist that restrict the reclaiming process, including limited funding, conflicts anlOng goals, institutional conflict, and a finite South Platte River water supply. The multiple-purpose water resources planning employed by Denver and the District has produced a series of feasible projects, programs, and policies acceptable to the many stakeholders. Multiple-purpose planning has long been recommended and employed for water resources planning and development at various levels of government. The Urban Drainage and Flood Control District has used multiple-purpose planning for urban flood hazard mitigation projects (Grigg et al., 1975; Urban Drainage and Flood Control District, 1977). Recent investigations (Association of State Floodplain Managers, 1995; Association of State Wetland Managers, 1991; Stewart and Scott, 1995; Federal Emergency Management Agency; National Park Service, 1995) provide theory and practical procedures for application of multi purpose planning to flood hazard mitigation projects involving other goals and objectives. PROJECT IMPLEMENTATION The recently completed $7.5-million project for the Confluence Park to 1-25 reach demonstrates a successful flood hazard mitigation project that

PAGE 68

Eisel, Kolstad, Urbonas, and Skifalides also provides important recreational opportunities for boating and recreational access to the river together with addition of 5 acres of wetlands, reconstruction of a 24-cfs diversion structure for year-round electric generating plant cooling water, and wildlife and aquatic habitat. 53 The ongoing Upper Central South Platte Valley at Zuni design project expands the purposes further to include expansion and improvement to the riparian and wildlife habitat, recovery of the South Platte fishery in the reach, providing recreational access to the river, provision of a safe boating drop structure, and diversion of cooling water for a power plant, along with meeting all flood control objectives. These projects demonstrate that urban flood hazard mitigation projects can produce significant benefits for a wide range of urban stakeholders and downstreanl reaches as well. Confluence Park to 1-25 Project The multiple-purpose project in this reach consists of: (1) Removing the old diversion danl, intake, and boat chute. (2) Constructing a new diversion dam with a crest approximately 3 feet lower, which is also expected to improve aquatic habitat downstream. (3) Constructing a new and more efficient power plant cooling water diversion structure. (4) Widening the boat chute and improving the drops for an improved boating experience and adding fish passage. Photographs of this project are presented in Figures 1 and 2. Due to the complex hydraulics, a physical model study was completed to detemline hydraulics through the area while providing a starting water surface for the analysis between Speer Boulevard and 1-25 adjacent to the new Elitch Gardens amusement park. Other components in this multiple purpose project include widening of the pedestrian bridge, new ADA approved ramps on both sides, an east side plaza and access to the river, connections to the new trailfmaintenance road, and landscaping using primarily native species. The channel in the reach upstream of Speer Boulevard to 1-25 was lowered an average of 3 feet and widened approximately 100 feet. The toes of the slopes were lined with riprap to prevent the typical scour and bank sloughing previously experienced. A maintenance road/trail was added to the east bank and landscaping was added using different zones. The area at the water's edge was planted with 10,000 live staked willows and the riparian zone inmlediately up the bank with a water table about 2 feet below tile surface was planted with wethmd-type plants such as arctic willows. TIle steeper slopes of the east bank were planted with a variety of native trees and shrubs including choke cherry and rabbitbrush. Approximately 300 trees and over 3,500 slmlbs were planted along the east bank and Confluence Park. Trees

PAGE 69

54 Reclaiming Denvers Central South Platte River Figure 1. Some of the key improvements made at Confluence Park, including a remodeled whitewater boat chute flanked by new pedestrian plazas and walkways and a new stepped concrete/grouted boulder cL'Ull. included primarily native species-hackcherry, hawthorn, jlmiper, cottonwood, and choke cherry. Shrubs included rabbitbrush, dogwood, juniper, western sandcherry, coyote willow, and dwarf arctic willow. The Colorado Division of Wildlife's list of trees and plants was used to select plantings. The 1995 spring stonns affected the project by depositing silt behind the dam and creating sandbars in the channel. Due to the wet spring and late snows during 1995, the South Platte River sustained flow for April, May, and June varied between 2,000 and 4,000 cfs; the more typical monthly flow for the Platte in this reach is approximately 200 cfs. Four hydrograph peaks of nearly the 10-year flow event (approximately 9,000 cfs) occurred from April to Jlme. These 1995 spring flows provided a good test of the structural integrity of this multiple-purpose project.

PAGE 70

Eisel, Kolstad, Urbonas, and Skifalides 55 Figure 2. TIllS view looking east toward downtown Denver shows the pedestrian bridge that was widened for better access from the existing west bank plaza to the new east bank improvements. Light fIxtures were also added to dle bridge and surroundi ng area. South Platte River Zuni Reach Project Thc Zuni reach has a diversion just downstream of Thirteenth Avenue to provide cooling water for the Public Service Company's Zuni Power Plant. TIle existing diversion depends on a rubberized inflatable dam. The dam height fluctuates depending upon the anlOunt of flow in the river. There is presently no boat chute or fish passage for this obstruction. Boaters must now portage around the dam. Due to constrictions of the Thirteenth and Fourteenth A venue bridges and low banks upstrean1, there is widespread flooding in this reach. ll1e current plan is to increase flood conveyance capacity of d1e channel by reshaping the channel from 1-25 to Eighth A venue and eliminating the existing inflatable dam. Once implemented, improvements will result to the river's fish habitat, boating, landscape, aesthetics, and wildlife habitat.

PAGE 71

56 Reclaiming Denver's Central South Platte River CONCLUSIONS hnplementation of the Confluence Park to 1-25 reach and the design process for the Zuni reach of the South Platte River through the central Denver flood hazard mitigation project indicate the absolute necessity for multiple-purpose flood hazard mitigation projects in urban areas. Without designing and constructing these projects to incorporate nun1erous features for a wide variety of stakeholders, successful design, funding, and construction of these projects would not be possible. REFERENCES Association of State Floodplain Managers 1995 "Multi-Objective Watershed Management National Assessment." Madison, WI: ASFPM. Federal Emergency Management Agency and National Park Service 1995 A Multi-Objective Planning Process For Mitigating Natural Hazards. Denver, CO: National Park Service. Grigg, N.S., L. H. Bothan, L. Rice, W.J. Shoemaker, and L.S. Tucker 1975 Urban Drainage and Flood Control Projects. Economic, Legal and Financial Aspects. Colorado State University, Environmental Resources Center Report No. 65. National Park Service 1991 A Casebook In Managing Rivers For MUltiple Uses. Association of State Wetland Managers, Association of State Flood Plain Managers. Philadelphia, PA: National Park Service. Stewart, Theodor J., and Leanne Scott 1995 "A Scenario-Based Framework For Multicriteria Decision Analysis and Water Resources Planning." Water Resources Research 31 :2835-2843. Urban Drainage and Flood Control District 1977 Methodology for Evaluation of Feasibility: Multijurisdictional Urban Drainage and Flood Control Projects.

PAGE 72

Floodplain Management in Urban Redevelopment: A Case Study in Multiple Objective Management Bernard B. Sheff Kenneth A. Nacci STS Consultants Ltd. INTRODUCTION The first settlers in what is now the City of Kalamazoo built their homes and industry along the natural, free-flowing Arcadia Creek to use its water As commerce and industry prospered arOlmd this site, it became the heart of the conummity, and remains so today as the north central business district (CBD) of downtown Kalamazoo, Michigan. As the city grew, the creek was charmeled undergrOlmd as it flowed through do\vntown and incorporated into the city's stomlwater drainage system. Buried and built upon, the creek lay forgotten. As development continued on the west side of downtown, stonnwater runoff increased dramatically. Thl.' creek, in its natural setting to the west, retained the capacity to accollunodate the new demand. Its capacity downtown, restricted by its enclosure, was not sufficient, resulting in the creation of a 100-year fh)(lplain throughout the north CBD. In 1982, the floodplain condition, cO\lpled with declining property values, vacant buildings, and high criminal activity in the north CBD prompted the plarming for stoml sy:.:tem improvements and economic strategies to revitalize this once prosperous section of downtown Kalanlazoo. The challenge was to develop a flood control project that served as a water anlenity and ultimately an enticement to urban redevelopment. FLOODPLAIN MANAGEMENT Arcadia Creek Flood Profile The original Arcadia Creek had been channelized, covered, and restricted over the years so that the only time the creek was observed was during stomls when portions of downtown were flooded. In addition, downtown buildings constructed prior to 1930 had used the covered creek as

PAGE 73

58 Floodplain Management in Urban Redevelopment structural foundations and walls in basements. Therefore, while many along the original alignment recognized the creek for its flooding, they did not realize the creek existed on the other side of the wall in their basements. This lack of understanding greatly complicated the process of redevelopment in the floodplain. The average cross-sectional area of the original creek was 67.5 square feet, with a maximum capacity of 450 cubic feet per second Ccfs); however, flood flows of 620,890, 1,020 and 1,315 cfs for the 10-,50-, 100and 500year events were generated in the 7.4-square-mile watershed. These flood flows were calculated as part of the original Flood Insurance Study, and were checked and found to be applicable to the current conditions. Indeed, tlle original floodplain maps developed for the creek in downtown Kalanlazoo utilized the two main roads that traveled east and west through the city as floodways for 100-year events and the remaining portion of tlle north CBD was floodplain. The original creek had a highly variable configuration including concrete box sections, concrete arches, and fieldstone arches with concrete or cemented stone inverts. In addition, the concrete base separated the creek from natural groundwater elevations by six to eight feet, thereby removing base flow from the creek throughout most of downtown. Further complicating matters were the random and (to a certain extent) unknown connections of stoml and/or other discharge points along the creek's path. These connections required substantial evaluation so that all areas serviced by the creek for stonn drainage were included in the new construction. Arcadia Creek Flood Control The new design of tlle creek needed to not only constrict the floodway and floodplain, thereby removing tlle nortll CBD from potential flooding, but also provide an aesthetically pleasing water anlenity throughout a six-block area. This included a pond and munerous walkouts over the channel. The new flood control structure is a 12 by 20 concrete open charmel fitted with weirs at various locations in order to provide the appearance of high stream flows. Specifically, notwithstanding the fact that there is no base flow to the creek for over half of tile creek's length, the upper portions of tlle creek watershed are flashy, witil average daily flows substantially less than even minor stonns. One of tlle lmique concepts regarding the new creek construction is tlle use of the existing creek downstrearn of the newly created pond. As a cost saving measure, the existing creek and newly constructed creek jointly outlet from the pond witil the old creek collecting stonnwater from the northern portion of the CBD and recombining with the newly constructed creek at a junction box 540 feet downstream of the pond. This approach required extensive analysis using the split flow options with HEC-2. Other lillique

PAGE 74

Sheff and Nacci 59 features of the structure include the surface water intake structures in the road flood ways at the west end of the corridor. These structures collect all overland flow and combine it with the creek flow in the channel, which then passes through the north CBD. The newly constructed creek is designed to retain all flood flow up to the 500-year event with adequate freeboard to accommodate all connected stonn sewers that are not in surcharged condition. The channel has perfomled well to date. During a recent storm water event of an approximate 25-year recurrence interval, the peak water surface elevation along the creek was .25 to .5 feet less than that predicted by the HEC-2 model. URBAN REDEVELOPMENT Arcadia Commons With the completion of preliminary engineering for an open air flood control project, the City of Kalamazoo and the Kalanlazoo Downtown Development Authority were able to negotiate with local public and private institutions (participants) for development rights within a six-block aren of the mile-long Arcadia Creek stormwater corridor. This area and tile participants of this multiple use urban redevelopment project becanle known as Arcadia Commons. Development in Arcadia ConmlOns includes the combined renovation of tile nationally registered historic Lawrence and Chapin Building (80,000 sq. ft.) and a new 50,000 sq. ft. administrative building by First of America Bank Corporation; a 35,000 sq. ft. campus facility constructed by Kalamazoo ValLy Conummity College; a new public muselml of science and industry also constructed by Kalamazoo Valley Conmmnity College, which is expected to draw over 200,000 visitors annually; a refurbished Radisson Plaz,l Suite Hotel completed by the Upjohn Company; a 40,000 sq. ft. regi(lI1al oncology facility jointly constructed by the Borgess and Bronson Hospitals with expectations of 60,000 patient visits annually; rehabilitation of the Visiting Nurses Headquarters with over 200 employees; a new 20,000 sq. ft. Michigan National Bank building; renovation of the historically significant Salvation Army Citadel by a local insurance/investment fiml; and parking facilities that include a renovated parking structure with a new 600-:;pace parking structure. Environmental Assessment and Remediation When the Arcadia Creek project was initiated, tile intent was that once the new flood control structure was constructed, the downtown area would be safely protected from floods up to a 500-year event, and properties would illilllediately be put on the market for redevelopment.

PAGE 75

60 Floodplain Management in Urban Redevelopment Unfortunately, as the project progressed, so did the national awareness of environmental issues. A great portion of the project was perfonned during a time when it was feIt sites must be returned to pre-use or pristine conditions. The downtown Kalamazoo area was highly industrialized at the turn of the century, and a good portion of the waste residuals from these industrial operations, including foundries, forges, blacksmithing, and tool and die machine operations, were dunlped along the banks of Arcadia Creek. Indeed, entire areas had been brought to grade for construction of parking lots adjacent to Arcadia Creek on 8 to 10 feet of heavily metal-laden residual waste. This non-point source random fill throughout the corridor, in addition to the more traditional fonns of environmental contamination associated witll leaking underground storage tanks and a block-long dry cleaning facility, made development along the corridor problematic at best. Limited envirorunental assessments were perfonned throughout the six-block corridor to develop a basic understanding of the contanlination along each site proposed for development. These investigations were perfonned prior to construction of the creek so that remediation could occur in concert with other construction operations along the creek alignment and building demolition for the new development in the area. Due to a rather unfriendly set of environmental laws in Michigan at the time of the initial construction along the corridor, contamination at most sites was excavated and shipped to a landfill for disposal. Sites with organic contamination were handled in rather unique ways using land leases and such that the grOlmdwater could be monitored and remediated as necessary over the time while the project proceeded and development occurreJ. It should be noted that the environmental laws in Michigan changed substantially in 1992, allowing the final piece of the Arcadia ConmlOns development to be handltd in a very different marmer than the previous excavate-and-dispose options. Specifically, the final step in development was construction of a large parking stmcture on the north side of the creek corridor. Post-industrial uses of this city block left the entire block contaminated with heavy metals and organic compounds. Using the newly promulgated laws within the state, the materials were closed in-place using the new parking structure as a cap io limit infiltration, thereby encapsulating the materials on-site. This method saved approximately $1 million and enhanced the financial pro fonna necessary to complete this last step in the redevelopment. Historic Preservation Nunlerous historically significant buildings existed within the Arcadia Creek corridor. In many instances, the adaptive reuse of these buildings by the participants was not feasible. Demolition of several of these buildings was required to construct their facilities. The use of federal dollars to construct portions of Arcadia Creek required the completion of an environmental

PAGE 76

Sheff and Nacci 61 impact statement. The assessment indicated a negative impact on the historic nature of the north CBD. In order to remain eligible for federal dollars, the City of Kalamazoo, the historic community, and the participants developed a memorandum of understanding that detailed design standards for new constmction and created an advisory board to provide architectural and site plan review. The result has created an urban redevelopment area that includes a mix of renovated historically significant buildings and new construction that is sensitive to the Arcadia Creek corridor's past. Recreation Recreational facilities immediately west and east of Arcadia Commons were incorporated into the design of the Arcadia Creek flood control project. On tlle west end is a two-block linear park that provides walkways, seating, and scenic overlooks of the open channel. On the east end, a pond created as part of the flood control project serves as focal point for a festival site that is home to festivals and special events that bring vitality to the downtown area. Economic Impact TIle economic impact of urban redevelopment in the north CBD has been significant. The public sector expenditures of $18 million leveraged over $200 million in private development. Public funds include state and federal grants, tax increment financing within the downtown district, and private philanthropic contributions. It is important to note that the receipt and use of federal and state funds is rcstricted within floodplain areas. TIle reconstmction of the creek and ultimate confinement of the floodplain allowed the city to utilize these public ftmds. Before redevelopment, the state equalized value within Arcadia Conunons approximated $60,000. It is now estimated at $400,000. It is further estimated that festival activity in downtown Kalanlazoo adds nearly $12 million to the local economy. CONCLUSION rille floodplain created by the Arcadia Creek stonn system prohibited reinvestment within the north CBD. Traditional approaches to floodplain management were available to the City of Kalamazoo. The decision to utilize the stoml system as a magnet for redevelopment allowed the city to remediate environmentally contanlinated property, entice private development, and provide recreational opportunities within an urban setting. The successful management of the Arcadia creek floodplain has provided the catalyst for achieving the multiple objectives of urban redevelopment in downtown Kalanlazoo.

PAGE 77

This page IS intentionally blank

PAGE 78

Part 4 LOCAL PLANNING AND MANAGEMENT FOR FLOOD MITIGATION

PAGE 79

This page IS intentionally blank

PAGE 80

Floodplain Management in Los Angeles County Allen Ma Los Angeles County Department of Public Works INTRODUCTION Los Angeles is located in a geologically young area of the western United States. It is characterized by highly erodible, steep mountains and flat alluvial plains. Also, due to a semi-arid climate, undeveloped and developed mountainous areas are often subject to a severe fire hazard. The combination of high intensity stomlS from the Pacific Ocean and recently burned hillsides often results in catastrophic storm flows containing mud and debris inundating various lower floodplains. As Los Angeles developed, solutions to the recurrent flooding were usually in the form of structural improvements. Much of the urbanized area of Los Angeles today is served by danls, debris basins, extensive tlood control channels, and underground storm drains. When Los Angeles County joined the National Flood Insurance Program (NFIP) in 1980, the pressure was growing to develop in floodplain areas. The attraction of the natural surroundings also drew people to build along watercourses. The population increase of some two million people underscored the need for a proactive floodplain management progranl. FLOODPLAIN MANAGEMENT REGULATIONS TIle county's floodplain management regulations provide guidance in a number of ways. The County Building Code regulates construction within an area subject to flooding. Construction is prohibited unless the development provides for flood protection and for the appropriate mitigation of adverse impacts to adjoining properties. Also, the County Utilities Code regulates the maintenance of natural watercourses and identified tloodways to preclude the placement of obstructions by the respective property owners of watercourses. Progressive levels of enforcement, such as daily fines and recorded violation notices against the property, are proposed to increase compliance.

PAGE 81

66 Floodplain Management in Los Angeles The county adopts flood way maps that identify areas subject to flooding. The county floodway maps generally have characteristics similar to Flood Insurance Rate Maps (FIRMs). The intent of the floodway map is to regulate development by reserving an open space for the passage of storm flows. The county flood way maps also serve as a secondary source of information for developers. The county's delineation of floodways within designated flood Zone A areas saves developers from the arduous task of addressing the cumulative effect of existing and anticipated development. In addition, the county's floodway mapping efforts ensure proper identification of flood hazards meeting local community and NFlP standards. The county has adopted 61 floodway maps for those areas with the highest potential for growth with some 40 additional watercourses still to be mapped. MAPPING TECHNIQUES The cOlmty's floodway maps are based on a different hydrologic method than that used to delineate Special Flood Hazard Areas on the FIRMs. Typically in rural areas, the resultant county "capital flood Qs" are signifi cantly greater than flow rates used for FIRMs. However, in urban areas the two rates are roughly equal. Consistent with NFIP regulations, the county uses the higher (stricter) standards to regulate development. The conversion of floodway maps and FIRM information into an electronic format is currently underway. The county's computer-aided design system will provide quick and accurate identification of parcels within areas subject to flooding as a service to property owners and lenders. Other electronic databases, such as property owner information and current topography, may be correlated to assist in the evaluation of viable locations for development or the enforcement of floodplain management regulations. The system will also facilitate development of community outreach mailing lists to properties within flood hazard areas. PUBLIC OUTREACH Regulations and mapping are often not enough to ensure effective floodplain management. Property owners often do not fully comprehend the extent of a flood hazard in their conmllmity. Also, major storms are infrequent and the bOlmdaries of a flood hazard are often poorly defined. In the intervening years between these major stonns, property owners construct fences, sheds, corrals, and other improvements affecting the flow of flood waters. These changes are difficult to regulate and have the potential to exacerbate the flooding. Also, property owners do not often adequately prepare for potential flooding. Mitigation of these problems requires education through public outreach programs.

PAGE 82

Ma 67 To assist in flood preparedness, the Department of Public Works developed a booklet entitled "Homeowners Guide to Flood, Debris and Erosion Control" and provided copies to over 36,000 residents in the last five years. The booklet suggests temporary measures in preparing for flooding in Los Angeles County and is most often given out by the County Fire Department while distributing sand bags to residents. The booklet has proven to be an effective tool for flood preparedness. After fires in mountainous and hillside areas near developed areas, the county estimates the debris and flooding potential. When a significant potential for damage exists, the Department of Public Works provides engineering advice as it did after the disastrous fires in the San Gabriel and Santa Monica Mountains in 1993. By providing over 450 residents with suggestions to mitigate flooding and debris problems, the effort has contributed significantly toward flood preparedness. Engineering advice is also provided to residents in the unincorporated county area upon request. The Department of Public Works recently developed a panlphlet entitled "Be Flood Aware" to warn of the dangers of flood waters. As part of the county's Community Rating System effort, the pamphlet was sent to over 1400 property owners in floodplain areas in the unincorporated comIty. These publications, including topics such as flood insurance, floodplain management, and flood hazard mitigation, are available to the public through the county public library. COMMUNITY RATING SYSTEM As part of the county's Community Rating System efforts each year, the Department of Public Works provides outreach infornlation to property owners subject to repetitive flood losses. The department also informs owners in flood hazard areas of the availability of flood insurance and advises lenders and flood insurance carriers of maps identifying areas of flooding. Record keeping and docunlentation systems are being modernized to help in these outreach efforts and to maintain a Class 8 rating under the Conmlunity Rating System. OTHER RELATED EFFORTS As land is developed in rural areas, changes to special flood hazard areas can occur by the encroachment of flood control improvements into natural watercourses. At times, these improvements protect land that was previously inundated by water. However, an approved FIRM revision is required prior to the elimination or reduction of the flood hazard. To ensure the map revision is submitted, the county includes the processing of these map revisions with the Federal Emergency Management Agency (FEMA) within the permit approval process.

PAGE 83

68 Floodplain Management in Los Angeles County The Department of Public Works requires that developers obtain FEMA's conditional approval before issuing a construction pennit for any improvement. The department reviews and forwards the developer's applications to FEMA. FEMA's final approval of the revisions to the FIRMs is also required before the transfer of facilities for operation and maintenance. In this way, the county ensures that homeowners receive the full benefits of any proposed flood control improvement. Often flood control improvements in rural areas provide residual benefits such as water conservation when levees with natural bottoms or unlined detention basins are built. These improvements provide for flood protection while allowing for the recharge of local groundwater basins. However, vegetation often develops and regulatory agencies place additional burdens on the Department of Public Works in maintaining these improvements. Also, these regulatory agencies often have conflicting or overlapping regulations and requirements. The department is working with these regulatory agencies to develop a feasible solution. The county and the u.s. Army Corps of Engineers are proceeding with improvements to the lower portion of the Los Angeles County Drainage Area (LACDA) basin. When completed, these improvements would protect over 80 square miles from overflows of the lower Los Angeles River and other key tributary channels. At the same time, FEMA proposes to implement new Zone AR regulations requiring flood insurance and building regulations until the improvements are completed. At stake with the new regulations is a potential loss of over $30 billion in economic activity over a lO-year period for the region. In preparation for the final Zone AR regulations, the Department of Public Works will be distributing pamphlets on flood awareness, prepared ness, and flood insurance availability to residents in the proposed flood Zone AR area. Also, presentations will be given to local cOtmnunities to explain the potential flood hazard and the proposed improvements. Finally, the department is refining an emergency response plan in the event an overflow occurs before the LACDA Project is completed. CONCLUSION Los Angeles County remains prone to some of the highest rainfall intensities in the continental United States. Because of continued competition for funding, regulatory mandates, and environmental constraints, floodplain management remains the only viable option for many areas of the county. Serving public needs in Los Angeles County with increasing regulatory responsibilities under the NFIP is more challenging than ever.

PAGE 84

Methodical MitigationA Deliberate Approach to Floodplain Management Jan Horton Illinois Emergency Management Agency One might think of "Methodical Mitigation" as a sequel to last year's production of "Clearing the Floodplain," presented at the Association of State Floodplain Managers' Annual Conference in Portland. Now that the Illinois Buyout Program has acquired more than 1800 privately owned parcels and removed nearly 1200 structures from the floodplain, communities must deal with the ownership of extensive public land and, in some cases where floodplain dwellers have relocated out of town, a smaller tax base. It is apparent that local governments will need some well thought-out mitigation plans before embarking on additional acquisitions. In other words, the Interagency Mitigation Advisory Group (IMAG), which implements the Hazard Mitigation Grant Program (HMGP), is tired of "writing the manual" as we go, a situation that developed after the Tvlidwest flood of 1993 during our efforts to assist individuals with a buyout as quickly as possible. Now that many of the 1993 HMGP projects are closing, the federal/state mitigation team is seeing the aftermath of the voluntary buyout. Clearing the floodplain and using HMGP funds for acquisition are still our primary objectives; however, one must be certain that local governments go about it methodically, rather than sympathetically. Many conununities have comprehensive land use plans, but these are often prepared to spark new development for economic reasons, which at times is cotmter-productive to risk reduction. As a result of the major floods and the acquisition progranl, previous land use plans may be undesirable because of unacceptable risks. For nearly two decades, communities have prepared emergency operations plans, spelling out exactly who is in charge and how the conununity will respond to and recover from a disaster. These plans are important; however, they do not

PAGE 85

70 Methodical Mitigation-A Deliberate Approach go far enough. They do not provide for long-tenn, or even short-tenn, mitigation measures in the recovery effort. lllinois is experiencing its fourth annual major flood at this time and is once again trying to recover. Between 1993 and 1996, 61 of lllinois' 102 counties had a major flood declaration; 29 had multiple floods. Some counties spared from flooding during the last three years are flooding this year. Combine this with the 30 tornadoes documented on April 18 and 19, 1996, and the ever-present risk of earthquakes on the New Madrid fault, and one might see the need to have in place a pre-disaster all-hazard mitigation plan. All levels of government are becoming increasingly aware that mitigation plans are not only necessary; they are essential! When the television reporter sticks a microphone in the mayor's face, he or she can say with conviction how the community will handle the situation as outlined in the local Hazard Mitigation Plan. "Breaking the cycle" of flood-repair-recover and flood again is the goal of every mitigation minded community. As a result of the emphasis on the National Flood Insurance Program and the new Flood Mitigation Assistance Program (FMAP), the Illinois Emergency Management Agency (lEMA) teanled up with counterparts at the lllinois Department of Natural Resources' Office of Water Resources and the Federal Emergency Management Agency (FEMA) to form joint planning teams to work with over 40 local jurisdictions involved in HMGP projects. Since the partnership was already in place with agency representatiw.s working together in the FEMA/lEMA Disaster Recovery Office, the planning teams evolved within the realm of standard mitigation activities. In other words, the planning teams were established without fanfare, without involvement from upper management, and without a fonnal agreement between agencies. Having a Hazard Mitigation Plan is a prerequisite for any new local mitigation project reviewed and evaluated by the IMAG before it is recommended for approval. Since the 1993 flood, mitigation has been given greater visibility, especially with local officials who in the past were more concerned about extraordinary response efforts. With increased visibility comes accountability, which we look at favorably. In our two 1996 disasters, we have noticed that the media are still following our efforts. Their interest did not wane once the hazard event was over, and both television and newspaper reporters have junlped on the mitigation bandwagon to follow the continuing progress. "Methodical mitigation" was already being pursued statewide in a proactive way before this current flooding. This spring's flood simply reinforced the necessity for local communities to begin the process and prepare mitigation plans. The local hazard mitigation plan will also

PAGE 86

Horton 71 provide the Interagency Mitigation Advisory Group with the material needed to make informed decisions for awarding HMGP funds to local governments. If "clearing the floodplain" is the goal, the IMAG needs to know whether the proposed project will be in the best interest of both the conummity and the taxpayers whose money we are spending. So what exactly is methodical mitigation? It can best be explained by telling what it is not. Methodical mitigation is not done haphazardly or at the spur of the moment. Methodical mitigation is not done when the river is about to crest. Methodical mitigation is not done by a single individual. Methodical mitigation is not politically inspired. Methodical mitigation is not done in a sympathetic mode. Methodical mitigation does not get a lot of media hype; but just wait-When the forces of Mother Nature lay claim to one's conmlunity, the local mitigation plan may get a great deal of exposure. At the very least, it will allow the community to be eligible for federal HMGP and FMAP funds as well as non-federal mitigation ftmds authorized in the state budget. The formula for methodical mitigation is simple: MM = LT+(PH,V,G+O,AM+PI)+A. If U1e LOCAL TEAM (LT) of conm1unity leaders, agency heads, and interested citizens has done pro-active mitigation planning, it will know exactly what its members need to do at the time of a disaster because they 'NiH have been incorporating mitigation measures into their regular routines. The City of Tulsa demonstrated this when they were doing buyouts within a week of their devastating flood. It is imperative that counties and incorporated communities: (1) identify their POTENTIAL HAZARDS (PH) and VULNERABILITY (V); (2) determine their GOAL (G) and OBJECTIVES (0) and make sure they are compatible with other local planning efforts; (3) assess their mitigation activities and evaluate ALTERNATE MEASURES (AM), selecting the most appropriate and affordable strategy spelled out in the HMP; and

PAGE 87

72 Methodical Mitigation-A Deliberate Approach (4) request PUBLIC INPUT (PI) to the HMP draft. By perfomling these tasks, the community will be able to prepare and ADOPT (A) a mitigation plan that addresses their goal and outlines their implementation strategy. None of this is new. In fact, planning as a concept should not be new to local governments, but in Illinois we work with many small communities that have a part-time mayor and a population of less than 1,000. In situations like this, one must be understanding about what we are asking them to produce because, unlike in the Chicago area, these downstate jurisdictions have limited resources. It is important that we "keep it simple or we will scare them away." Therefore, the hazard mitigation plan is very easy to complete. It is what emergency managers in the late seventies referred to as a boilerplate (fill in the blanks) plan. The procedure includes meeting with locals; discussing their hazard history in the context of the plan; looking at maps-lots of maps; and providing the technical assistance in a non technical way. Methodical mitigation may not be sophisticated, but it is effective, adequate and well-received in our communities most in need of mitigation. Note: copies of the Hazard Mitigation Plan Model are available from the author.

PAGE 88

Flood Hazard Mitigation: Planning and Implementation Matthew G. Wahl Peoria County Planning and Zoning FLOOD HISTORY OF PEORIA COUNTY The Peoria, Illinois, area contains over 100 square miles of Special Flood Hazard Area (SFHA), the majority of which is located along the TIlinois River. Much of this area has been built upon, containing industrial, commercial, and residential development. Industrial development began due to the access of transporting raw materials to local companies by the use of barges. Commercial development was fostered with the increasing popularity of such water-related activities as boating and fishing. Residential development occurred as homes were constructed as sununer cottages. Many houses were located at the river's edge to take advantage of the various amenities associated with the close proximity to the river. During a relatively flood free period (1943-1979) many of these structures were expanded and became permanent residences. Severe flooding in Peoria has been an ongoing phenomenon for a munber of years. The Illinois River has experienced numerous floods in the past two decades. The river itself is fed by seven major streams in unincorporated Peoria County, the largest of which is the Kickapoo Creek, which nms through Akron, Radnor, Limestone, and Kickapoo townships. Many of the streams, especially the aforementioned Kickapoo Creek, have experienced the same flooding problems as the Illinois River. Since 1979, the residents of Peoria County have suffered through four catastrophic floods (1979, 1982 (2), and 1983). More recently in 1995, the Illinois River and Kickapoo Creek inundated numerous structures causing extensive damage and flood losses in our area. The National Flood Insurance Program (NFIP) has paid out millions of dollars in flood insurance claims to individuals in the SFHA. In the 1979 flood, 41 % of the structures located in the SFHA had the first floor completely inundated by flood waters.

PAGE 89

74 Flood Hazard Mitigation: Planning and Implementation RECENT FLOODS: 1979, 1982 (2), 1983, 1995 Illinois River 428.4-Zero Elevation for the Gage 455.5-Flood Crest 03/23/82 451.1-Damage Begins to Occur 455.8-Flood Crest 12/10/92 454. I-Flood Crest 05/30/95 457.1-Flood Crest 03/24/79 454.1-Flood Crest 04/17/83 DEVELOPMENT OF THE HAZARD MITIGATION PLAN(S) The Peoria County Hazard Mitigation Plan was completed in 1985, while the City of Peoria/Village of Peoria Heights Hazard Mitigation Plan was completed in 1988. A key component of both plans was a floodplain survey in which buildings are listed by the following criteria: Building nunlber Address Tax identification number Ground elevation Depth (structure's elevation below or above the base flood level) Mitigation alternatives. PLAN IMPLEMENTATION: FLOODPLAIN ACQUISITION PROGRAM Acquisition (1) Preliminary acquisition tasks. (2) Prepare database to include project area property information as well as project accounting systems. (3) Appraisals. (4) Comparable sales approach. In 1985 interest rates were still relatively high. Numerous foreclosures and repossessions were stiJI taking place due to the economic instability of the area.

PAGE 90

Wahl (5) Negotiations with property owners. a) Importance of availability of negotiator to property owner during the negotiation and/or relocation process. b) "Real" property questions. c) Family heirlooms. d) Real property removal (policies and guidelines). e) Scheduling of negotiations (preferably away from holidays). (6) Property closing process and required documentation. Title and Abstract Company-Company responsible for conducting the 75 title search and preparation of the written title commitment. It is important to select a reliable company to conduct this work in order to ensure the agencies right to "free and clear" title to the property. Propeliy Acquisition Documents ("Tools of the Trade") a) Just Compensation Form-Document illustrates owners' name(s), property address, parcel identification number, structure style, purchasing agency, and offer for parcel(s) and/or improvements. b) Title Commitment-Document outlines owner's foruml name and legal description of property to be acquired. In addition, any outstanding mortgages, liens, etc., also should be indicated in the commi tment. c) Warranty Deed-Document that legally conveys property from property owner to purchasing agency. d) Quit-Claim Deed-Document that conveys any improvements on property considered to be under tenant ownership. e) Disclosure of Ownership or Beneficial Interest-Document used to indicate any outside parties' interest in the property. f) Disclaimer Affidavit for Tenant Owned Improvement-Document used to disclaim in legal ownership of improvements on property (Le., land owner/improvement owner).

PAGE 91

76 Flood Hazard Mitigation: Planning and Implementation g) Lease Agreement-Document used to allow former property owner to occupy structure after the scheduled date of closing (90-day limit). Requires former owner to provide liability insurance for property until the structure is permanently vacated. Written proof of insurance is required at closing. h) House Repurchase Agreement-Document used to allow acquired property to be repurchased for 6% of the appraised value of the structure. This value was determined to cover the demolition costs of the foundation, septic system, well, and any related ground cover. Agreement also stipulates bonding requirements for the installation of foundation work, sewage system, water system, and roadside culvert. There is a 30-day time limit to remove structure from the SFHA. This process allows for an inexpensive alternative to purchasing a new structure. Relocation (1) Relocation assistance with regard to acquisition and relocation for low-to moderate-income individuals. (2) Replacement Housing Costs-Additional funds have been made available to aid homeowners in acquiring a new structure. Replacement costs are based on the difference between the cost of relocation unit less the negotiated price for the acquired structure. (3) Rental Assistance Payment-Provided to qualified tenants. Based on an increased rent payment (difference between current rent and future rental costs) over 42 months. (4) Moving Expense Payment-Provided to both tenants and home owners. Based on a fixed expense or on actual moving/storage costs. (5) Relocation of Structure-Dwelling is treated as personal property when it is detached from the ground. Demolition (1) Demolition contract qualification requirements. Contractor must have in his or her possession a valid Certificate of Eligibility from the State of Illinois. The certificate will indicate a

PAGE 92

Wahl financial rating for the contractor in regard to the type of work and contract amount that he or she is certified to accept. The state takes numerous factors into consideration when rating contractors, including number of years in business, practical experience (in regard to previous demolition contracts), available demolition equipment, etc. The certification is an extremely useful rating tool when bidding out large demolition projects. It behooves bidding by responsible and experienced contractors. (2) Compensation policy. Lmnp StUll bid versus square footage bid. (3) "Real" property answers. 77 Contractors can decide after the initial work order for the demolition has been issued to sell articles located on the property. Restoration and Development Plans (1) Floodplain restoration-Properties are required by agreement to remain as open space and in public ownership for eternity. Parcels are graded during the demolition process and then seeded with prairie grass in order to stabilize the soil and mitigate erosion problems. (2) Future open space development plans for acquired properties-Parks, picnic areas, bicycle/walking paths, athletic fields, etc., are encouraged uses for property acquired through the floodplain acquisition programs. Water-related uses, such as boat ramps and parking areas, are also excellent uses for the property. (3) Park districts and townships provide an excellent resource for both development and maintenance programs for these properties. These entities may also consider taking ownership of properties. FLOODPLAIN MANAGEMENT Local Floodplain Ordinance ll1e local floodplain management ordinance requires higher standards for development in the SFHA. The ordinance specifies freeboard requirements ancl material types for construction in the SFHA as well as the regulations regarding filling in the floodplain.

PAGE 93

78 Flood Hazard Mitigation: Planning and Implementation Enforcement Enforcement of the floodplain ordinance is provided by the Planning and Zoning Administrator and related support personnel. Flood Insurance Rate Maps and Floodway and Flood Boundary Maps Flood msurance Rate Maps (FIRMs) and the Floodway and Flood Boundary Maps are deciphered for the public by the Planning and Zoning Department. Flood zone designations are determined and released to the public for their use. Community Rating System (CRS) The Community Rating System (CRS) is a FEMA/NFIP program that rewards a community's efforts to reduce flood damage to existing buildings and to protect any new or substantial construction to minimum NFIP standards. Peoria County is a participating CRS community and continues to strive for additional CRS credit through the implementation of sound flood mitigation principles. Partnerships Successful floodplain management requires a good working relationship with state and federal flood mitigation agencies.

PAGE 94

Drainage Master Planning for the Largest Irrigation District in the United States Steve R. Knell Imperial Irrigation District Anders K. Egense Eugene F. Shank Theodore V. Hromadka Boyle Engineering Corporation INTRODUCTION Within southeastern California's Imperial Valley lies an area that has one of the highest agricultural production rates in the world (see Figure 1). This arid region generates this level of production as a result of yearround growing conditions and an extensive network of irrigation canals operated by the Imperial Irrigation District (lID). In concert with the canal system, the lID maintains a corresponding network of "drains," 1,430 miles of open channels (and some pipes) that were primarily designed to convey surface and tile drain runoff from irrigation of the cultivated tlelds. The drains discharge into the region's two major rivers (the New and Alamo rivers) and drainage sink (Salton Sea). To the extent that the individual drains have capacity, they also convey storm water runoff. With the North American Free Trade Agreement (NAFTA) and the expanding role of cross-border trade, the farn1ing towns that dot the valley have grown at record rates in response to industrial, commercial, and residential needs. Considering the need for proactive stormwater management in conjunction with this development, lID initiated the preparation of a PDMP spanning both the lID service area and the sUITOlmding tributary drainage basin. In addition to describing drainage basin characteristics and unique aspects of the lID drains, this paper The autlwrs thank the llD for the support of this project and for pemlission to publish this pape r.

PAGE 95

III t }.) )}" Offsite Drainage Catchment Boundaries (typ,) Drainage Basin and Study Area Boundary Figure I. hnperial Valley drainage basin_
PAGE 96

/(nell, Egense, Shank, and Hromadka 81 the approach and results of the PDMP, including initial concepts Ifor improving the drains to provide prescribed levels of flood protection. DRAINAGE BASIN CHARACTERISTICS The Imperial Valley is located within the southern portion of the larger Salton Sea basin, which encompasses approximately 3,380 square miles. Tne central portion of the drainage basin is characterized by very flat terrain. Within the central area of the basin is the 860-square mile area that is irrigated by lID (lID service area) and is home to several urban centers and scattered clusters of individual homes. Surrounding the central area of the basin are expanses of largely undeveloped terrain that include :sparsely vegetated desert, sand dunes, and steep, rocky mountains. Drainage from the mountainous areas flows into broad alluvial washes hhat impinge on the perimeter of the lID service area. The 61 "offsite" ;catchments range from 170 acres to 260 square miles. The lID drains are interconnected into 160 individual drainage l"systems," where each system is a separate watershed that has one ldischarge point to a river or the Salton Sea. The earthen drains vary from lsmall trapezoidal channels at the upstream ends (as small as 10 square to large multi-channel cross-sections with total areas on the order of '11,000 square feet at the downstream ends. A typical drain has a ltrapezoidal cross-section with 1.25H: 1 V side slopes, a 3-foot bottom width land an 8-foot depth. Invert slopes of the drains are typically flatter than prevailing slopes of the land surface. The average invert slope of the i
PAGE 97

82 Drainage Planning for the Largest Irrigation Districl required and available storage and flow rate. Offsite catchments were analyzed for peak flow rate and runoff vohune at the point where the drainage path intersects the perimeter of the lID service area. A more detailed analysis was utilized to evaluate improvement needs for each drain system within the lID service area. The watershed for each drain system was segregated into subbasins down to the cell level. The concentration point of each cell was represented by a node located along the respective drain. Cells (nodes) were interconnected in a link-node system where each of the 1,630 links are described by detailed drain system data provided by lID. Peak flow rates and runoff volumes were determined at each node. MASTER PLANNING PARAMETERS lID staff worked with the Drainage Committee of the lID Board of Directors to assess the relationship between improvement costs and three important parameters: design storm levels, level of confidence in the hydrologic analyses, and the use of floodplain management. These efforts lead to selection of the parameters described below. Design Storm Level The design storm level defines the magnitude of the peak runoff quantities that are to be used for analyzing the existing storm drainage facilities and sizing potential improvements. Two storm levels were selected to determine peak runoff quantities: a 2-year design storm for the agricultural areas within the lID service area and all areas outside lID, and a 25-year design storm for the urban areas within the lID service area. Level of Confidence Hydrologic analysis involves the application of statistical methods to rainfall data in order to develop estimates of various return frequency storms. The level of confidence is a measure of the statistical reliability of the results of these analyses. Different agencies select different levels. For example, the Federal Emergency Management Agency (FEMA) uses the 50% confidence level in defining its floodplain maps. Local agencies often select higher levels because they are involved in the design and construction of flood control facilities. An 85% confidence level was selected for the lID PDMP (see also Knell et aI., 1996). Floodplai n Management Floodplain management is an approach that can be utilized to reduce the size of a drainage system by detaining some of the runoff in a distributed fashion throughout the catchment before it enters the drainage system. By

PAGE 98

Knell, Egense, Shank, and Hromadka 83 reducing peak runoff quantities, this approach can lead to a reduction in the size of the drainage conveyance facilities, and a concomitant reduction in capital facility and improvement costs. lID policy currently allows each quarter-section field (160 acres) to have at most one 12-inch diameter tail water outlet to discharge surface runoff. One application of floodplain management would be to use berms around each field such that storm flows are detained and only allowed to discharge via the single 12-inch outlet. These berms already exist around many of the fields, although some berms may not withstand the pressure of ponded storm water runoff. Estimates were made to determine the influence of various degrees of on faml floodplain management on runoff volumes and improvement costs. Different runoff curve numbers (CN) in the hydrologic analyses were used to represent the degree of floodplain management. The construction of bemlS around each field is not presently required by any regulatory policy, and it was the opinion of the Board Drainage Committee that instituting such a policy could be a burden on the agricultural industry. Based on the perspecti ve that existing bemls will provide some measure of floodplain management, the analysis approach selected for the PDMP assumed that one-quarter of the design storm nmoff would be detained on tlle fields. FLOOD CONTROL IMPROVEMENT CONCEPTS To accommodate nmoff from the offsite areas, earth embankment levees along the perimeter of the lID service area were selected because of their simplicity and low cost. The levees would be constructed with locally available materials, sized for 2-year stoml runoff detention, and include emergency spillways sized for 100-year flow rates. For the drains within the lID service area, two approaches were evaluated. Each approach is described below. Free-Flowing System Approach All drains would be sized to convey the peak design stonn flow and all constrictions in the drains would be removed. Peak runoff flow rates were compared to the existing conveyance capacity to determine the deficiency ill tenns of cross-sectional flow area. The improvement for a particular drain segment (link) is the volunle of excavation necessary to provide conveyance. Road crossings would be replaced by various structures depending on size and site-specific requirements: larger pipe culverts, reinforced concrete box culverts, or bridges. For each link, the existing drain geometry and an estimated munber of road crossings fomled the basis for estimates of construction costs for installing free-flowing road crossings.

PAGE 99

84 Drainage Planning for the Largest Irrigation District Total Storage System Approach All the drains would be sized to store all the runoff from the design stoml and all existing road crossings remain in place. Runoff volumes were compared to the existing storage capacity to determine the deficiency in terms of channel volume. The analysis was performed on a link-by-link basis such that all runoff would be contained in the drain segment into which it discharges. The volume deficiency was computed for each link, and the improvement for each link was the volume of excavation necessary to provide total storage. CONCLUSIONS The lID PDMP provides an initial evaluation of the hydrologic and hydraulic characteristics of the drainage systems within and tributary to the Imperial Valley area, and provides estimated costs for improving these systems to provide the selected levels of flood protection. The results indicate that the total storage approach is least costly in terms of construction cost, with an average cost of about $165 per acre (based on 500,000 irrigated acres within the lID service area). Future elements of lID's stormwater management plan include extending the PDMP efforts to} more detailed investigations aimed at developing "drain-specific" improvements. Future improvements may entail combinations of the free flowing and storage-based approaches depending upon actual conditions along each drain. REFERENCES Knell, Steve R., T.V. Hromadka, lI. DeVries, and A.K. Egense 1996 "Preparation of a Hydrology Manual For Imperial Valley, California," Association of State Floodplain Managers National Conference, San Diego, California.

PAGE 100

Floodplain Management in Mecklenburg County Stephen R. Sands Ogden Environmental and Engineeri ng Services, Inc. William R. Tingle Mecklenburg County Storm Water Services INTRODUCTION Stoml Water Services (SWS) is charged with the management of all floodplains in Mecklenburg COlmty and recognized a need to evaluate their current practices. SWS recognized that all levels of government, all businesses, and all citizens have a stake in properly managed floodplains. Therefore, SWS developed a floodplain management plan to examine and potentially re-focus SWS efforts, with the involvement of numerous stakeholder groups. This plan was formed to serve as the guideline to evaluate current operations and potential modifications to several agencies' involvement in floodplain management. In addition, this plan and the process by which it was developed allowed SWS to apply for a reduction of flood insurance rates for all citizens county-wide through the National Flood Insurance Program's Conummity Rating System (CRS). The potential of implementation of a multi-agency geographic information system (GIS) system was also evaluated during this effort. OBJECTIVE The floodplain management approach is not limited to the traditional flood control or land acquisition measures. In general, floodplain management aims to achieve two objectives: (1) To prevent or reduce the loss of life, disruption, and damage caused by floods, and (2) To preserve and restore the natural and beneficial functions of the floodplains.

PAGE 101

86 Floodplain Management in Mecklenburg County A community should use as many different and effective measures as it can to reach the two objectives. Examples of measures that work toward the first objective include channel improvements, floodplain construction building codes, early flood warning, and retrofitting buildings. Examples of measures that work toward the second objective include development regulations that protect wetlands and storm water management practices that filter or clean the runoff that enters the streams. The floodplain management approach involves a variety of organizations, not just the public works department traditionally concerned with channel maintenance and flooding. These can include planning and zoning offices, emergency managers, the Red Cross, parks departments, developers, and floodprone property owners themselves. The key to coordinating all activities and agencies to ensure that they support each other and other community goals and objectives is the preparation of a floodplain management plan. FLOODPLAIN MANAGEMENT PLAN DEVELOPMENT PROCESS In order to develop a plan that ensures that all agencies and interest groups will participate in the plan's action items, a seven-step process was used that reviewed the flood-related problems and developed a coordinated response among many agencies and stakeholders. By following this process, the credits available through the CRS program were maximized. However, the benefits of reduced flood insurance premiums are minor in comparison to the benefits generated by a multi-agency, multi-objective approach to improving floodplain management. A two-part workshop was conducted in the fall of 1995 and was attended by approximately 30 participants from various local, state, and federal agencies, environmental groups, homeowners, and developers. TIle following steps were followed to ensure accurate development of a comprehensive plan. Step 1. Describe the flooding problem and the natural and beneficial uses of the floodplain. Step 2. Review and compile all floodplain management measures tIlat impact flood damage and protect natural and beneficial floodplain functions. Step 3. Identify tile appropriate measures for use in the county. Step 4. Develop a draft action plan to evaluate and implement appropriate measures. Step 5. Circulate the draft to tile agencies and people most affected. Step 6. Adopt and implement the plan. Step 7. Monitor, evaluate, and revise the plan, as needed.

PAGE 102

Sands and Tingle FLOODPLAIN MANAGEMENT POLICIES AND RECOMMENDATIONS 87 The main result of the seven-step plan development process was a list of recommendations and/or floodplain management measures. The measures were developed through a consensus-building process among staff and stakeholders as described in step 4 of the planning process. The following general floodplain measures were recommended for further evaluation, consideration, and implementation, if appropriate. The first nine measure"s relate to activities that are appropriate throughout the entire county. These are defined as phase 1, policies and recommendations. The next four measures relate to activities that are applied specifically to subwatersheds. These are defined as phase 2, flood loss prevention and reduction in subwatersheds. In addition, the floodplain measures are grouped into general policy statements (italicized in the following list). Phase 1-Policies and Recommendations I. New development should be managed so flood problems are not increased. (1) Floodplains are needed to store and convey flood waters and to provide riparian habitat. The ideal way to do this is to maintain the floodplain as open space. (2) Where floodplain development is allowed, current regulatory programs should be evaluated to ensure that they provide adequate flood protection. (3) New developments throughout the watershed should account for the impact of their runoff on drainage, flooding, and water quality. II. The county's drainage system should be maintained to maximize its ability to carry and store water. (4) Procedures should be developed to ensure proper drainage system maintenance. (5) The design and maintenance of channels throughout the watershed should use natural features where practicable. Ill. The flood warning and response plan should be evaluated to determine its effectiveness to protect people and property during and after a flood. (6) The local flood warning program should be evaluated to determine its effectiveness to maximize the lead time available to respond to flooding. (7) The flood response plan should be evaluated to deternline its to protect life and property during and after a flood.

PAGE 103

88 Floodplain Management in Mecklenburg County IV. The public should be informed about and involved in floodplain management. (8) A public information program should be implemented to infonn, educate, and involve the public in floodplain management activities. V. Floodplain management agencies and organizations should coordinate their efforts. (9) Agencies and organizations involved in floodplain management should communicate and coordinate their efforts as much as possible. Phase 2-Policies and Recommendations VI. Subwatershed plans, or other studies, should be prepared to identify the best mix of floodplain management measures to solve local flooding and development concerns. (10) A systematic approach should be followed to reduce flood damage to existing development. (11) Guidelines for acquisition of flood prone areas. (12) Guidelines for flood control projects. (13) Guidelines for retrofitting projects. Details of each of these general reconunendations are provided in Chap(er 4 of the Floodplain Management Guidance Document. PLAN IMPLEMENTATION The implementation of the Floodplain Management Guidance Docume/lt begins with its adoption. implementation is dictated by a detailed action plan that specifies responsible agencies and the associated schedule for addressing the reconmlendations. Some of the recommendations will be investigated during the implementation phase and could involve revisions to ordinances or policies. An extensive stakeholder involvement and interaction process must be followed to ensure "buy-in" into any floodplain management reconullendation that involves revisions to ordinances or policies. Depending on the revision, the appropriate advisory committee (Mecklenburg-Charlotte Stornl Water Advisory Committee, Mecklenburg-Charlotte Planning COIlUllission, Building Development Committee, etc.) will review and comment on the revisions prior to consideration by the appropriate governing body. Therefore, the need for on-going interaction with all stakeholder groups is recognized. Reconmlendation 9 of the Floodplain Management Guidance Document includes the establishment of a Floodplain Management Coordinating Conunittee (FMC C) that is responsible for notifying all agencies and stakeholder groups that may be affected by a proposed change

PAGE 104

Sands and Tingle 89 in SWS direction. The FMCC is the mechanism by which this on-going interaction is guaranteed. Early notification of the stakeholder group is key. Without bringing all stakeholders into the process as soon as possible, consensus or "buy-in" can not be achieved. Therefore, it is recognized that this document serves as a "defined process to plan." The upcoming years during which the initial phases of this Floodplain Management Guidance Document are implemented will serve as the actual development of the plan to achieve the two objectives: (1) to prevent or reduce the loss of life, disruption, and danlage caused by floods, and (2) to preserve and restore the natural and beneficial functions of the floodplains. By ensuring that all levels of government, all businesses, and all citizens that have a stake in the floodplain are involved in the process, the success of achieving these two objectives is maximized. GIS PROGRAM DEVELOPMENT A major factor in expediting the implementation of the floodplain management plan is the automation of data gathering and analysis by utilizing GIS. These data sets can also be continually updated and developed into future coverages to meet many SWS floodplain management needs. A GIS action plan was developed to detemline strategies for the automation of floodplain related data. This action plan is described in the following list. (1) All GIS-related operations being perfomled in Mecklenburg County by all public and private agencies and utilities were proposed to be identified. (2) The extent of data collection and maintenance efforts by each agency as well as the format being utilized for data collection and storage was proposed to be identified. This effort includes a determination of the schedule for creation of any other databases or coverages. (3) Based on tmderstanding of the GIS resources in Mecklenburg County, a list of possible inter-agency GIS appl ications including a recommended schedule of development and estimate of the cost of development for each application was proposed to he developed. For exanlple, it was proposed that some applications may not need to be developed immediately because the data collection effort is incomplete. In these cases, the effort focuses on ensuring that the data being collected is

PAGE 105

90 Floodplain Management in Mecklenburg County being compiled in a fonnat that would be usable by SWS in future applications. It was suggested that other applications should be created immediately because their source data is available and the application has immediate merit. Problems were encountered in gathering the local inter-departmental GIS coverages. SWS decided to focus its efforts internally on researching and developing SWS-specific coverages and applications to meet SWS needs. Existing coverages and databases from other agencies will be used if available. The following list describes the approach. (1) Other cities were visited to interview GIS and engineering personnel to determine the types of applications that are successfully being used by their programs. In addition, the past problems that have occurred Witll the development of the applications and the proposed direction of the agencies regarding future applications were discussed. (2) A list of potential GIS applications to be created was developed based on the interviews and team knowledge of other possible applications, and SWS desires. (3) Research was perfonned using the coverage developed in a pilot area and other needed coverages to detennine the level of accuracy, required database fields, and forn1at of contributing coverages and databases. (4) Other agencies responsible for creating or maintaining these coverages or databases were contacted to detennine schedule of creation, fonnat of databases/coverage, level of accuracy, and level of maintenance. (5) A list of applications was recommended including a time of completion, cost of completion, and required databases/coverages. (6) The applications and database were developed in accordance to that schedule. CONCLUSIONS Many lessons were learned as SWS ventured through the process of developing the floodplain management plan. The most significant of these was involving the various interest groups to try and reach consensus on a variety of issues. After the third public meeting, SWS reached a point where the development community, floodplain residents, and environmentalists appeared to understand each others' viewpoints a!ld were willing to work towards the development of a comprehensive plan. The floodplain management plan will provide a systematic direction for the future of SWS's floodplain management activities and programs. It also creates a means to bring together all floodplain stakeholders to make decisions that may affect present and future management issues.

PAGE 106

One-third Century of Flood ManagementObservations and Suggestions Joseph C. Hill County of San Diego Flood management has been in effect forever. Over the last several thousand years, there is a recorded history of people managing their activities relative to floods. Structures have been constructed away from major flood areas and the floodplain has been used for agriculture or other purposes that can sustain periodic flooding. During the 1960s, while with a Michigan consulting firm and the Corps of Engineers in Los Angeles, I found that projects included identification of flood areas so that structures could be built in locations that would avoid adverse flooding impacts. Events within our society and our profession have changed the importance and the methods of flood management. For instance, when Congress established the National Flood Insurance Program (NFIP) in 1968, the emphasis on flood management shifted from structural flood control projects to regulation of floodplains. This paper gives a history of events that changed the importance, methods, and effectiveness of flood management in San Diego County. It is structured by identifying major events, or turning points and discussing the effect on flood management. SAN DIEGO COUNTY FLOODPLAIN MANAGEMENT PROGRAM State of California, Department of Water Resources Bulletin 112 The state published Bulletin 112 in 1964 and provided floodplain maps for the major rivers in San Diego. This publication provided the basis for restricting development in these rivers and the Corps of Engineers and the county followed the state lead by mapping other streams. This program preceded development on most major rivers and creeks so that houses and other structures could be directed to higher ground and the streambeds retained in a natural condition. This early program was the most effective component of the San Diego floodplain program.

PAGE 107

92 One-Third Century of Floodplain Managemenl City Proposes a Concrete Channel through Mission Valley This project, designed by the Corps of Engineers, had been in the planning stages for many years and the city all owed a large shopping center, Fashion VaIley, to construct stores, hotels, commercial buildings, etc., in the floodplain while relying on the proposed channel for flood protection. At a public hearing in 1971, a geography professor from San Diego State University led strong opposition to "ugly concrete channels" and the project was not approved. This set the stage for an anti-channel, pro-natural floodplain movement and launched the county program. County Initiates Floodplain Program Just after the city's Mission Valley channel hearings, the county began an expanded floodplain mapping program. The program (Hill and Brown, 1985) exceeds the criteria of the NFIP with 200 foot/inch orthophoto base maps that are re-scaled photographs showing houses, roads, and other features that allow citizens to accurately relate the location of floodplain limits to their property. County floodplain engineering criteria result in wider flood ways and identification of erosion and sedimentation that could destroy stmctures if ignored (Hill and Spalding, 1986). Corps of Engineers Responsible for Area NFIP As the NFIP was implemented, the Los Angeles District of the Corps was given responsibility for the San Diego region. The Corps provided floodplain mapping for several major streams and the staff was responsive to issues in San Diego. Although there were many differences of opinion in hydrology, consistent flood flows and mapping were accomplished. Planners Initiate Floodplain Zoning In response to the popular natural floodplain concept, the County Planning Department was directed to place an overlay zone over the floodplain as defined by the floodplain maps. As a result, thousands of notices were sent to people with property in floodplains. Property Rights Objection Several commtmities had several hundred houses in the floodplain overlay zone. Community groups were organized overnight and strenuous objections were filed with the Board of Supervisors, U.S. Congressmen, the President of the United States, etc. All aspects of the mapping were attacked, including the hydrology (it never rains in southern California) and the floodplain analysis. County and Corps engineers had to work closely with citizen groups in review of the technical aspects of the

PAGE 108

Hill mapping. Several of the floodplain studies were reviewed and revised to counter the assertions that they did not reflect a reasonable flood hazard and that the zoning was a basis for a inverse condemnation finding. Public Hearings for Floodplain Zoning 93 A series of hearings before the Board of Supervisors (many with capacity crowds) finally resulted in completion of floodplain zoning on most major rivers. Since the Flood Insurance Rate Maps were also being published, members of Congress were involved in the dispute. The process required years of extensive involvement of the Corps and the county staff. FEMA Replaces the Corps in San Diego As the NFIP evolved, a regional FEMA office was established in San Francisco, and it assumed responsibility for the Flood Insurance and Floodplain Management Progranl, relieving the Corps. FEMA contracted with the State of California, the Corps, and the County of San Diego for additional floodplain studies. Although the county completed its assigned work, FEMA never considered it for additional studies, but contracted with private fimlS. The county had no part in the selection process and minimal influence in review of the work. In spite of the fact that some rivers in San Diego are subject to major streambed erosion and sedimentation, the contractor was directed not to consider these factors in floodplain analysis. Many of the resulting floodplain maps made the county's job of regulating development more difficult. With a remote iocation and limited staff, FEMA was not as effective as the Corps. Events of 1978-1980 An initiative on the California ballot, Proposition 13, was approved in 1 Q78. Since the flood control budget had been structured to focus on planning rather than construction, the tax rate was low. After Proposition 13, taxes were proportioned on previous years' amOlmts, but at lower rates. TIle post-Proposition 13 budgets were reduced with no opportunity [01 an increase. The first major flood in 40 years occurred in 1978. A larger flood in 1980 focused attention on the fact that it can rain in Southern California. Although there was extensive danlage and concern, the effects of Proposition 13 were more important and there were no significant additional funds for flood management. One major impact of the floods was the realization that emergency operations needed to be improved. Robert Bl\TI1ash, director of the River Forecast Center in Sacramento, had developed a vastly improved program for obtaining real time precipitation and stream flow data directly from remote field stations with radio and computer systems. San Diego was the first to implement a

PAGE 109

94 One-Third Century of Floodplain Management county-wide system. The Automatic Local Evaluation in Real Time (ALERT) has been expanded and is used throughout California and in many areas of the world (Hill and Burnash, 1986). Alluvial Fans In the mid 1980s FEMA recognized potential flood hazards on alluvial fans and began to map them. A preliminary alluvial fan map had been prepared for the Borrego area of San Diego County. The county retained a consultant (Hill and Dawdy, 1987a) and reviewed the preliminary study. After extensive discussions with FEMA, a county/FEMA alluvial fan map and report were developed (Hill and Dawdy, 1987b) and the map was approved in 1987 by the Board of Supervisors as part of a Borrego Flood Management Report (Hill et aI., 1988). The report includes criteria for development on fans (Dawdy et aI., 1989; Hill and Spalding, 1989). Community Coordination Floodplain management is incorporated into the planning process by plotting the floodplain on the community plan land use maps. Typically, property within the floodplain is down zoned to reduce the incentive to develop in the floodplain. Most community plans restrict channel construction unless there are existing houses in the floodplain (Saipe et aI., 1988). Private projects can be coordinated with community plans, infrastructure plans, and floodplain management (Hill and Walker, 1986). Growth Initiative-Resource Protection Ordinance In the late 1980s San Diego experienced rapid growth and concern arose over its negative impact on the environment and quality of life. Initiatives were placed on the ballot to restrict the number of building permits issued and place environmental restrictions on them (Hill and Saipe, 1988). Floodplains were identified as an environmental resource and, although the initiatives did not pass, the county and the city did pass similar resource protection ordinances. Regulations in the county ordinance include wider floodways when rivers are remapped; floodway use limited to recreational, agricultural, and open space; the requirement that development be set back from the floodway 100 feet or 15% of the floodway width; a prohibition on channelization unless necessary to protect existing structures; a requirement that flood way bank construction be natural in appearance; and a limit on fill in the floodway fringe. Flood Insurance Study Update FEMA is completing a revised Flood Insurance Study (PIS). The new FIRM will include the entire county, so that floodplains will not stop at

PAGE 110

Hill 95 corporate boundaries. The county is including floodplain plots in its GIS database, providing for more accurate plotting of floodplains and for plots at remote locations. The FIS identifies erosion and sedimentation hazards (Hill and Mohr, 1988) and will add 43 miles of flood hazard streams. Flood Control Facility Construction-Removal of Structures As a result of the 1978 and 1980 floods, in which hundreds of houses were flooded and difficult emergency evacuations were necessary, it became obvious that flood control facilities were necessary to protect existing houses and other structures. Flood control channels were constructed by the Corps of Engineers and the county in Los Coches Creek and by the county in Spring Valley. These and smaller facilities have removed hundreds of structures from the flood hazard areas. STORMWATER MANAGEMENT Flood Control and Drainage Regulation FEMA recognizes the importance of private and public drainage and flood control facilities. The county has developed hydrology and design stillldards that provide a solid basis for design and construction (Hill, 1990). A floodplain ordinance, approved by FEMA, was adopted in 1988. Floodplain construction is coordinated with FEMA to provide consistent county floodplain maps and FIRMs. After several years of development, the county completed the flood control plans that identified streams, existing facilities, needed improvements, and construction costs. Methods of financing and fees on new development were included in some areas. Off site impact from urban areas (Hill, 1987) is a major concern and mitigation measures are needed for many projects. The importance of this aspect of storm water management cannot be overemphasized. Stormwater Quality The Clean Water Act identifies "non-point" source runoff as a major pollutant of downstream receiving waters. The responsibility for this program is delegated to the state in California. The state, in tum, places virtually all the responsibility on the municipalities. As a result, we are expected to accomplish monitoring, surveying, public relations, enforcement, construction of pollution reduction facilities, etc., equivalent in cost to the flood management program with no additional funds, while the state has transferred 44 % of the flood control district fund to the school systems. Obviously, the stonnwater quality program has had a major adverse impact on the flood management program.

PAGE 111

96 One-Third Century of Floodplain Management CONCLUSIONS AND SUGGESTIONS Flood management in San Diego County is a complex and effective program that protects life and property and the natural (environmental) aspects of rivers and creeks. It also reduces the need for costly construction of channels ($7 million per mile) to protect houses and other structures built in floodplains. Floodplain mapping ($10,000 per mile) prevents construction in floodplains. If floodplain mapping prevented the need to channel one-half of the 270 miles of mapped floodplains (Hill and Brown, 1985), the cost avoidance would be 135 x $7 million or $1 billion. However, the public and politicians do not recognize the value of this program. A strong public relations program (the City of San Diego spends $250,000 annually on stormwater public relations) is needed. An expanded federal/state/municipal program would also strengthen flood management. REFERENCES D.R. Dawdy, J.e. Hill, and K.e. Hanson 1989 "Implementation of FEMA Guidelines on Alluvial Fans," Proceedings of the American Society of Civil Engineers NatiOlwl Conference on Hydraulic Engineering. Hill, J.e. 1987 "Flood Plain Management and Urbanization," Proceedings, 11 th Annual Conference, Association of State Floodplain Managers. Hill, J.e. 1990 "Design and Construction of Urban Stomlwater Management Systems," Proceedings of the Water Pollution Control Federatiolt Annual Conference, Manual of Practice Program. Hill, J.C., and AJ. Brown 1985 "San Diego County Flood Plain Management Program," Proceedings of the Ninth Annual Conference, Association of State Floodplain Managers. Hill, J.e., and R. Bumash 1986 "Alert in San Diego-Quantum Leap in Storm Waming/Data Collection Systems," Western State High Risk Symposium, Association of State Floodplain Managers.

PAGE 112

Hill Hill,J.C., and D.R. Dawdy 1987a "Interpretation of FEMA Alluvial Fan Maps for Planning and Regulation," Proceedings of the Eleventh Annual Conference, Association of State Floodplain Managers. Hill, J.C., and D.R. Dawdy 97 1987b "Criteria for Alluvial Fan Management Based on FEMA Study," Proceedings of the Eleventh Annual Conference, Association of State Floodplain Managers. Hill, J.e., and K. Mohr 1988 "Floodplain Management with Highly Unstable Stream Beds," Proceedings, Arid West Conference, Association of State Floodplain Managers. Hill, J.e., and e. Saipe 1988 "Citizen Growth Control and Environmental Protection Ballot Initiatives," Proceedings of the Arid West Conference, Association of State Floodplain Managers. Hill, J.e., and W.P. Spalding 1986 "Floodways in Movable Bed Rivers," Western State High Risk Symposium, Association of State Floodplain Managers. Hill, J.e., and W.F. Spalding 1989 "Single Channel Alluvial Fan Hydraulics with Application to Design of Medium-Density Projects," Proceedings, American Society of Civil Engineers Conference on Hydraulic Engineering. Hill, J.e., and W.R. Walker 1986 "Private Sector Involvement in Flood Plain Management," Proceedings, Tenth Annual Conference, Association of State Floodplain Managers. Hill, J.e., D.R. Dawdy, and G.K. Lutes 1988 "Alluvial Fan Management," Proceedings, Arid West Conference, Association of State Floodplain Managers. Saipe, e., J.e. Hill, and A.J. Brown 1988 "Floodplain Management 'Through the Community Planning Process," Proceedings of the Twelfth Annual Conference, Association of State Floodplain Managers.

PAGE 113

This page IS intentionally blank

PAGE 114

Part 5 COASTAL ISSUES

PAGE 115

This page IS intentionally blank

PAGE 116

Correlation of Hurricane Magnitude, Percentage Chance of Exceedance, and Damage Potential for Coastal Counties from Texas to Maine Darryl J. Hatheway Dewberry & Davis INTRODUCTION The frequency of formation and landfall of hurricanes magnifies the risk for coastal commtmities. From 1886 to 1994, an average of five hurricanes has occurred each year in the North Atlantic basin (Hebert et aI., 1995). By tmderstanding the nature of damaging influences of the various intensity levels of hurricanes and the level of exposure (damage in dollars) of coastal cotmties, identification of areas with high vulnerability to hurricanes is possible. This can allow federal, state, and local officials to formulate and focus their hazard mitigation strategies on the most vulnerable cotmties and the surrotmding commtmities. HURRICANE DESCRIPTION A tropical cyclone is defined by an area of closed circulation over tropical wdters, in which the winds rotate cotmterclockwise in the northern hemisphere and clockwise in the southern. Tropical cyclones with wind speeds 74 mile per hour or greater are classified as hurricanes and conunonly affect the coastal cotmties of the United State's North Atlantic basin, which includes the coastal areas of the Atlantic Ocean and Gulf of Mexico. The various hazard components and risks associated with a hurricane can be subdivided into those related to storm surge, wind, and rain (Bryant, 1991). The tropical cyclone is identified by its stages of development and intensification, with associated wind speeds. It can grow from a tropical depression to a tropical stonn to a categorized rank of hurricane and then make the transition into its extra tropical stage. The Saffir-Simpson Hurricane Scale measures the intensity by numbered categories (1 to 5). Wind speed, storm-surge height, and

PAGE 117

102 Hurricane Magnitude, Chance of Exceedance, Damage Potential damage/destruction potential are factored into this rating system as follows (National Weather Service 1993). Category 1: Winds 74-95 mph-No real damage to building structures. Damage primarily to unanchored mobile homes, shrubbery and trees. Also some coastal road flooding and minor pier damage. Category 2: Winds 96 -110 mph-Some roofing material, door, and window damage to buildings. Considerable damage to vegetation, mobile homes, and piers. Coastal and low-lying escape routes flood 2-4 hours before arrival of the center. Small craft in unprotected anchorages break moorings. Category 3: Winds 111-130 mph-Some structural danlage to small residences and utility buildings with a minor amount of curtain wall failures. Mobile homes are destroyed. Flooding near the coast destroys smaller structures with larger structures damaged by floating debris. Terrain continuously lower than 5 feet msl may be flooded inland 8 miles or more. Category 4: Winds 131-155 mph-More extensive curtain wall failures with some complete roof structure failure on small residences. Major erosion of beach areas. Major damage to lower floors of structures near the shore. Terrain continuously lower than 10 feet msl may be flooded, requiring massive evacuation of residential areas inland as far as 6 miles. Category 5: Winds greater than 156 mph-Complete roof failure on many residences and industrial buildings. Some complete building failures with small utility buildings blown over or away. Major damage to lower floors of all structures below 15 feet msl and within 500 yards of the shoreline. Massive evacuation of residential areas on low ground within 5 to 10 miles of the shoreline may be required. COASTAL EXPOSURE Since 1980, coastal counties vulnerable to the severe coastal flooding and wind damage associated with hurricanes have increased their populations, thus increasing the exposure to the risk of natural disasters. Accompanying the increase in population is increased urban and commercial growth in the coastal zone. Properties in coastal areas vulnerable to severe flood and wind damage from hurricanes in the United States include virtually all of the coastal counties from Texas to Maine. The coastal growth trends are accompanied by higher property values and

PAGE 118

Hatheway 103 urban and commercial development to support the growth and expanding local and tourist populations. Hurricanes present one of the greatest potentials for substantial loss of life, properly damage, and economic impact, because more than 45 million U.S. residents live within the coastal areas vulnerable to hurricanes. In these areas with the highest growth rates, from Texas to Maine, there has been an estimated 15 % increase in population (more than 5 million people) from 1980 to 1993 and an estimated 65% increase in the value of insured coastal residential and commercial property from 1988 to 1993 (Insurance Research COlmcil and Insurance Institute of Property Loss Reduction, 1995). However, more than 85% of those residents have never experienced a direct-hit hurricane (Hebert et aI., 1995). HURRICANE EXPERIENCE In an analysis of hurricane experience levels of coastal county populations from Texas to Maine (Hebert et aI., 1984), the direct and indirect hurricane landfalls in each county were tabulated. Direct hits by hurricanes were considered to be StOrolS during which the eye passed directly over the coastal cOlmty. The indirect hits included the occurrence of hurricane force winds and/or stonn-surge tides of 4 to 5 feet in adjacent counties. In a 1995 update of the previous study, the assessment was expanded to include the nunlber of direct hits by landfalling hurricanes in coastal states from Texas to Maine from 1900 to 1994 (Hebert et aI., 1995). The assessment was further modified in this paper to include indirect hits. Of the 154 U.S. hurricanes originating in the North Atlantic Ocean, Caribbean Sea, and Gulf of Mexico since 1900, 55 of them struck Florida, making it the state most susceptible to hurricanes. Not only is Florida ranked the highest in overall nunlber of hurricanes, it has been hit by the greatest mmlber of hurricanes of Category 3 strength or higher. Behind Florida in frequency of occurrence of direct and indirect hits by hurricanes since 1900 are Texas, Louisiana, North Carolina, and South Carolina, in that order. Many of the 45 million people currently living in the coastal areas vulnerable to hurricanes have moved there during the past 25 years, a period when the activity level of hurricanes and direct hits by hurricanes bve been very low. Only about one-fifth (12) of the 62 direct hits by hurricanes of Category 3 or higher since 1900 have occurred in the last 25 years. In contrast, approximately 50% of the costliest (more than $25 million in damage) hurricanes have occurred during the past 25 years, with Hurricane Andrew in 1992 being the most expensive. This is a result of increasing growth trends along the coast, not increases in hurricane activity.

PAGE 119

104 Hurricane Magnitude, Chance of Exceedance, Damage Potential The trend of increasing development along the coastal cOlUlties will magnify the vulnerability of those areas to catastrophic losses from the impact of a hurricane. Although not every coastal county has experienced the effects of a hurricane, at least one coastal county in every state from Virginia to Maine has been affected by either a direct or indirect landfalling hurricane since 1900. Based on the data for experiencing direct and indirect hits, the probability of a coastal county experiencing a Category 1-5 hurricane was evaluated by determining the chance of exceedance in any given year, with the 20-year value or 5% chance of a Category 3 or greater hurricane being experienced in any given year. 1l1is determination provided a basis for evaluating the risk of exposure to an intense hurricane based on historical data from 1900 to 1994. The following tables give the damage potential and insured coastal property exposure for the top 10 coastal counties from Texas to Maine, based on 5% annual chance exceedance in any given year for Category 4 (Table 1) and Category 3 (Table 2) hurricanes. The damage correlation for the rankings is derived from the estimated percentage of damage to the exposed property expected from a Category 3 or 4 hurricane. There were no coastal cOlmties with a 5% annual chance exceedance for experiencing a Category 5 hurricane. The percentage ranges for danlage potential in the damage correlation colunm are derived from the estimate of coastal stonn damage (in dollars) from Hurricanes Opal, Andrew, Hugo, and Camille. These ranges are only estimates of the potential for losses and may fluctuate dramatically, depending upon whether the event has both severe storm surge and winds (like Hugo) or is primarily a surge (Opal) or wind (Andrew) event. SUMMARY The changing coastal environment has exposed the need for improved building design and construction standards. Severe flooding from hurricanes destroys and damages residential and cOnIDlercial properties; coastal and bay erosion takes valuable property away; and public and private transportation, water, sewer, and electrical services in the impacted communities are disrupted. The increased vulnerability of coastal development to natural hazards has revealed the need to establish critical erosion zones, high hazard areas, and improved construction standards. The key to the survival and continued economic health of communities in coastal disaster-prone areas is to improve hazard mitigation strategies, increase local awareness, strive for a greater understanding of the prevalent natural hazards and destructive forces of severe coastal floods and wind storms, and encourage the enhancement and enforcement of more stringent building regulations. These strategies

PAGE 120

Hatheway 105 Table 1. Damage correlation and insured property exposure for top 10 coastal counties for 5% annual chance exceedance of experiencing a Category 4 hurricane ($l,OOOs). Saf-Sim Insured Coastal Property Values Damage Potential HurT. Residential Commercial Total Category 4 estimated costs State Coun!}: M!!&!!. 1993 1993 22% to FL DADE cat 4 $ 62,564,084 $ 98,279,480 $160,843,568 $ 32,168,714 to 48,253,070 FL BROWARD cat 4 54,074,968 62,292,372 116,367,344 23,273,469 to 34,910,203 TX GALVESTON cat 4 7,135,426 12,174,506 19,309,932 3,861,986 to 5,792,980 FL COLLIER cat 4 10,885,182 8,071,747 18,956,928 3,791,386 to 5,687,078 TX BRAZORIA cat 4 6,212,878 11,647,470 17,860,348 3,572,070 to 5,358,104 MS HARRlSON cat 4 5,277,006 8,193,182 13,470,188 2,694,038 to 4,041,056 MS JACKSON cat 4 3,639,503 6,601,311 10,240,814 2,048,163 to 3,072,244 FL MONROE cat 4 4,908,742 3,156,032 8,064,774 1,612,955 to 2,419,432 LA LAFOURCHE cat 4 2,520,077 3,911,034 6,431,111 1,286,222 to 1,929,333 T;( MATAGORDA cat 4 1,265,195 3,760,481 5,025,676 1,005,135 to 1,507,703 Table 2. Damage correlation and insured property exposure for top 10 coastal cQlmties for 5% annual chance exceedance of experiencing a Category 3 hurricane ($l,OOOs). Saf-Sim Insured Coastal Property Values Damage Potential HurT. Residential Commercial Total Calegory 3 estimated costs State Coun!}: M!!&!!. 1993 1993 10% to 15% NY SUFFOLK cat 3 $ 87,789,496 $ 41,005,688 $128,795,184 $ 12,879,519 to 19,319,278 FL PALM BEACH cat 3 49,226,364 53,755,380 102,981,744 10,298,174 to 15,447,262 CT NEW HAVEN cat 3 43,616,276 43,525,444 87,141,720 8,714,172 to 13,07\,258 FL PINELLAS cat 3 34,718,072 36,564,468 71,282,544 7,128,254 to 10,692,382 FL HILLSBORO cat 3 29,279,844 40,688,232 69,968,080 6,996,808 to 10,495,212 LA ORLEANS cat 3 14,268,698 29,071,426 43,340,124 4,334,0\2 to 6,501,019 LA JEFFERSON cat 3 15,713,332 21,348,722 37,062,056 3,706,206 to 5,559,308 FL LEE cat 3 18,073,684 16,243,274 34,316,960 3,431,696 to 5,147,544 TX JEFFERSON cat 3 8,073,854 24,805,080 32,878,934 3,287,893 to 4,931,840 FL BREVARD cat 3 16,312,511 12,816,371 29,128,882 2,912,888 to 4,369,332

PAGE 121

106 Hurricane Magnitude, Chance of Exceedance, Damage Potential can be implemented through coastal hazard mitigation efforts, identification of coastal high hazard and flood-related erosion zones, establishing and demanding hurricane-resistant structure design and construction standards, and entering into partnerships between the private and public entities to educate the planners, engineers, and construction trades on the importance of hurricane hazard mitigation. REFERENCES Bryant, Edward 1991 Natural Disasters. University of Cambridge. Hebert, Paul J., Jerry D. Jarrell, and Max Mayfield 1995 "The Deadliest, Costliest, and Most Intense United States Hurricanes of This Century (and other Frequently Requested Hurricane Facts)." Proceedings for 17th Annual National Hurricane Conference. National Weather Service. Hebert, Paul J., Glenn Taylor, and Robert A. Case 1984 Hurricane Experience Levels of Coastal County Populations-Texas to Maine. Technical Memorandum NWS NRC 24. Miami, FL: National Hurricane Center. Insurance Research Council and Insurance Institute for Property Loss Reduction 1995 Coastal Exposure and Community Protection: Hurricane Andrew's Legacy. Wheaton, IL. National Weather Service 1993 Natural Disaster Survey Report: Hurricane Andrew-South Florida and Louisiana-August 23-26, 1992. U.S. Department of Commerce, National Weather Service.

PAGE 122

Analyzing and Mapping Coastal Flood Hazards along the Open Coasts of the Atlantic Ocean and Gulf of Mexico Jerry Sparks Darryl J. Hatheway Doug A. Bellomo Dewberry & Davis INTRODUCTION When the National Flood Insurance Program (NFIP) was established in 1968, there were widely used and accepted methodologies and computer models for analyzing riverine flood hazards. However, no such standard methods were available for analyzing coastal flood hazards. Therefore, JV:;l' the last 25 plus years, the NFIP has developed standardized methodologies and computer models for analyzing the unique processes, mechanics, and forces associated with coastal storm flood events. These tools for coastal flood hazard identification and mapping are documented in the Federal Emergency Management Agency's (FEMA's) March 1995 "Guidelines and Specifications for Wave Elevation Determination and V Zone Mapping-Final Draft. /I HISTORY OF THE COASTAL HIGH HAZARD AREA (V ZONE) TIle NFIP first began mapping coastal high hazard areas (V Zones) in the 1970s with technical guidance provided by the U.S. Anny Corps of Engineers (Corps). The Corps reconmlended that a wave height of 3 feet be considered critical in temlS of producing velocities and impact'> that may cause significant structural danlage. It also reconmlended procedures for mapping the inland limit of the 3-foot wave for both developed and Llndeveloped coastal sites (U.S. Army Corps of Engineers, 1973). In 1975, lhe Corps issued a follow-up report that further substantiated the critical Ilature of the 3-foot wave (U.S. Amly Corps of Engineers, 1975). In addition, it updated its previously recommended procedures to include fetch length analysis and expanded the discussion of V Zone mapping in :lensely developed areas.

PAGE 123

108 Analyzing and Mapping Coastal Flood Hazards Between 1975 and 1980, NFIP maps using the Corps guidance for delineating V Zone boundaries were published for approximately 270 communities. Although wave crest elevations were not shown, the 3-foot wave associated with the base (lOO-year) flood was used to delineate the inland limit of the V Zone boundaries. The stillwater elevations, which consisted of the astronomical tide and stonn surge, were published as the regulatory base flood elevations (BFEs). In 1976, FEMA contracted with the National Academy of Sciences (NAS) to ascertain whether wave heights should be included in the BFEs for coastal Flood Insurance Studies (FISs) and, if so, how those calculations should be perfonned. The NAS recommended including wave height analysis for open coasts, embayments, and estuaries and provided FEMA with a methodology for doing so (National Academy of Sciences, 1977). This methodology considered varying fetch lengths, barriers to wave transmission, and the regeneration of waves likely to occur over flooded land areas; however, the extent and elevation of wave runup, amount of barrier overtopping, and coastal erosion were not addressed at this time. In 1979, FEMA adopted the NAS methodology, making the Wave Height Analysis for Flood Insurance Studies (WHAFIS) computer model available for use, and initiated an intensive effort to incorporate the effects of wave action on the NFIP maps for coastal communities along the Atlantic Ocean and Gulf of Mexico. Thus, BFEs for coastal sites became a composition of both the stillwater elevation plus an estimated wave crest elevation. In the late 1970s and early 1980s, structures along the New England coast, designated as outside the flood hazard area according to the NAS methodology, experienced considerable wave damage from notable northeast stomlS. In 1981, FEMA recognized and approved a methodology that detennined the height of wave runup landward of the stillwater lille (Stone and Webster, 1981). The computer model developed was modified in 1987 and 1989 for increased convenience of input conditions and to improve the computational procedures. In 1986, in response to criticism indicating a significant underestimation of the extent of the V Zone, FEMA undertook an investigation to reevaluate V Zone identification and mapping procedures. On October 1, 1988, the definition of "coastal high hazard area" in Section 59.1 of the NFIP regulations (44 Code of Federal Regulations 69.1) was revised to read: "Coastal High Hazard Area" means an area of special flood hazard extending from offshore to the inland limit of a primary

PAGE 124

Sparks, Hatheway, and Bellomo frontal dWle along an open coast and any other area subject to high velocity wave action from storms or seismic sources. 109 To clarify this redefinition, a definition of primary frontal dWle was added to tile NFIP regulations as follows: "Primary Frontal Dune" means a continuous or nearly continuous mound or ridge of sand with relatively steep seaward and landward slopes immediately landward and adjacent to the beach and subject to erosion and overtopping from high tides and waves during major coastal storms. The inland limit of the primary frontal dune occurs at the point where there is a distinct change from a relatively steep slope to a relatively mild slope. The primary frontal dune is recognized as a transitory deposit of sediment, and in major storms it is subject to high energy wave action and erosion, and may possibly even be breached. Prior to the 1988 NFIP regulation modification, many dunes were designated as outside of the V Zone, which allowed the degradation of the dWle for construction or other purposes, thereby reducing the initial line of natural protection and increasing flood hazards. The 1988 modifications to the regulations were made to preserve the natural protection and reduce the flood hazards associated with human alterations of the dunes. These definitions are still used today and are a major component in any coastal PIS. During this same period, the Corps' Coastal Engineering Research Center (CERC) performed a study for FEMA outlining the various methods of assessing storm-induced erosion (Birkemeier et al., 1987). of this investigation, FEMA recommended considering the effects or ',[(Jml-induced erosion (as opposed to long-tem1 erosion or accretion); hO\vever, given the extreme limitations of the models available at the tim,, FEMA chose to employ a simplified procedure for quantifying the amoti11t and extent of erosion during severe coastal stom1S. These procedures were developed using historical data for 30 major coastal events, establishing a relationship between stoml surge elevation and cross-sectional area of the dune profile above iliat elevation. These criteria were codified in Section 65.11 of ilie NFIP regulations. GUIDE FOR ATLANTIC AND GULF OF MEXICO COASTAL FLOOD HAZARD ANALYSIS AND MAPPING As the result of the 25-plus-year evolution of coastal analysis and mapping, FEMA is preparing to publish a final version of the "Guidelines and Specifications for Wave Elevation Determination and V Zone Mapping." This docWl1ent explains how to analyze the various

PAGE 125

110 Analyzing and Mapping Coastal Flood Hazards accompanying hazard components of a coastal stonn event, such as beach and dune erosion, wave heights, wave runup and setup, and dune and barrier island overtopping. It also addresses the specific evaluation criteria for coastal protection structures such as seawalls, bulkheads, revetments, and levees. In addition, it explains how the results of these FEMA developed and accepted analyses are applied to coastal hazard mapping and includes the criteria for V Zone mapping. An appendix at the end of the document provides an example coastal flood hazard study illustrating the use and application of the various methodologies and models. As explained in the users' guide, mapping V Zones requires locating and mapping the most landward of the following three points: (1) the point at which a 3-foot wave height may occur; (2) the point at which the eroded grOlmd profile (or non-eroded profile, if applicable) is 3 feet below the computed wave runup elevation; and (3) the inland limit of the primary frontal dlme as defined in the NFIP regulations. To locate these points and accurately map wave crest elevations on the NFIP map, coastal engineers are needed to perfonn the associated analyses, which involve determining stillwater elevations and deepwater wave conditions, assessing coastal erosion and scour, perfonning wave runup and overtopping analyses, and computing propagating wave heights. Given the sheer length of shoreline for many coastal communities, economic constraints prohibit detailed representations of the entire shoreline. Therefore, transects, which can be thought of as "cross sections" or "profiles" perpendicular to the shoreline, are used to represent a lengrh of shoreline with similar physical and cultural characteristics. These may vary significantly along any given shoreline, and the conditions, such as the defined BFE and wave characteristics, may also vary. An understanding of where changes in these conditions may occur and of how the varying physical features may affect coastal flooding patterns and wave crest elevations is crucial to properly locate these transects. Once the appropriate conditions have been determined along each transect (water surface elevations, wave characteristics, and eroded ground conditions), wave height, wave nmup, and overtopping analyses are perfonned. The results of these analyses are then compared to the inland limit of the primary frontal dune, and the most landward of the three criteria previously mentioned is mapped as the inland limit of the V Zone. The flood elevations and ponding depths, resulting from the wave height analysis, runup computations, and overtopping assessment are then superimposed on the transects and transferred to topographic maps.

PAGE 126

Sparks, Hatheway, and Bellomo 111 Knowing the location and elevation of the zones along the transects, an engineer can then begin interpolating these results between transects. Knowing how the physical features and cultural characteristics may affect the propagation of waves is essential in this process, and a good understanding of typical coastal flood patterns is the best tool for the job. Once checked for completeness, accuracy, and reasonableness, the results shown on the topographic map are transferred to the NFIP map, which is then sent to the community for review. CURRENT STATUS TIle final draft report was distributed to the Association of State Floodplain Managers' Coastal Committee, the Corps, and FEMA Regions I, II, III, IV, and VI in the fall of 1995 for a peer review. FEMA is currently evaluating the comments received and preparing the fmal report. TIle document will then be published and implemented for conducting coastal PIS studies and restudies along the Atlantic and Gulf of Mexico coasts. At present, FEMA is preparing procedures for analyzing and mapping coastal flood hazards for the Great Lakes, which generally parallel the now standard Atlantic and Gulf seacoast procedures but are tailored to the Great Lakes and generally result in lesser wave hazards. However, there is no similar comprehensive document for coastal hazard assessments in the Pacific Ocean region. CONCLUSION Coastal processes are not easily defined by equations and computer programs. They require knowledge and understanding of the coastal engineering and oceanographic principles governing the dynamic forces of the oceans, and sound engineering judgments are necessary to assess individual coastal flood hazard components. Methodologies and computer models used by FEMA to perform coastal flood hazard assessments will continue to be refined and newer versions considered for use. For now, the guidelines and specifications for the Atlantic Ocean and Gulf of Mexico will be valuable tools for proper evaluation of the flood risks associated with coastal floodplain developments. REFERENCES Birkemeier, William A., Nicholas C. Kraus, Norman Scheffner, and Stephen C. Knowles 1987 Feasibility Study of Quantitative Erosion Models for Use by FEMA in the Prediction of Coastal Flooding. U.S. Army Corps of Engineers Technical Report CERC-S7-S. Vicksburg, MS: USACE.

PAGE 127

112 Analyzing and Mapping Coastal Flood Hazards National Academy of Sciences 1977 Methodology for Calculating Wave Action Effects Associated with Storm Surges. Washington, D.C.: NAS. Stone and Webster Engineering Corporation 1981 Manual for Wave Runup Analysis, Coastal Flood Insurance Studies. Boston, MA. U.S. Army Corps of Engineers, Coastal Engineering Research Center 1984 Shore Protection Manual. Volumes I and II, 4th Edition. Washington, D.C.: U.S. Government Printing Office. U.S. Army Corps of Engineers, Galveston District 1973 General Guidelines for Identifying Coastal High Hazard Zone, Flood Insurance Study-Texas Gulf Coast Case Study. u.S. Army Corps of Engineers, Galveston District 1975 Guidelines for Identifying Coastal High Hazard Zones.

PAGE 128

An Evaluation of the Costs Associated with Managing Delaware's Atlantic Ocean Coast through a Policy of Retreat Michael S. Powell Delaware Department of Natural Resources and Environmental Control INTRODUCTION Delaware currently mitigates against erosion and related storm damage on the Atlantic Ocean shoreline through a combination of engineered shore protection projects and development standards. Federal and state expenditures on Delaware's shore protection have risen dramatically since the 1980s as large-scale engineered beach nourishment projects have been COll',tructed and are being maintained on most of the developed portions of Delaware's ocean coastline. These costs are expected to increase further as additional communities require protection and as sea level rise and related shoreline migration necessitate larger projects with more frequent maintenance to achieve the same level of benefit. The high cost of Delaware's current beach nourishment program warrants an exploration of alternative means of achieving the state's goals: maintaining Delaware's quality recreational beaches, and protecting landward property and infrastructure. This study describes and estimates the costs of one alternative management strategy-strategic retreat. METHODOLOGY The underlying hypothesis of this study is that as shoreline migration reduces beach width in developed areas, adequate beach width could be restored and maintained by removing oceanfront structures to keep pace with shoreline migration. The study area covers the Atlantic Coast of Delaware-about 25 miles of coastline and 2053 structures within 600 feet of the ocean. The goal of the study is to estimate the cost of retreat for a five-decade period beginning in 1990, using a variety of assumed shoreline migration rates. In order to estimate this cost, a model was developed that simulates shoreline migration, identifies impacted

PAGE 129

114 Costs of Managing Delaware's Coast through a Policy of Retrea structures, and calculates the costs associated with the loss of those structures. The retreat model simulates future shoreline migration and identifies those structures that would be intercepted by the reference feature each decade. Specifically, the model simulates a landward migration of the summer (non-storm) location of the semi-monthly high tide line ("reference feature" hereafter) from its 1989 location. Given an assumed migration rate, the model calculates the decade in which the reference feature will first reach individual structures. The model outputs a total value of all such impacted structures in each community, by decade. In order to run the retreat model, every residential and commercial structure vlithin 600 feet of the reference feature, as detemIined through an analysis of 1989 aerial photographs with field verification, was incorporated into a data set. A total of 1845 single-fanlily residential structures and 208 multi-family residential and commercial structures were inventoried. Using data from a variety of sources, the following infonnation was obtained for each structure: ecommunity name eage etype of heat eoceanfront estructure type enumber of bathrooms egarage edistance from reference feature esquare footage enumber of fireplaces ecentral alc Two hundred and seventy-four property transactions that took place between 1987 and 1990 in and adjacent to the study area were also analyzed. For these structures, a second, smaller data set was created containing all of the above attributes and the purchase price. Using the sales data set, a value was estimated for each single-family residential structure in the study area using the following equation: Log (Price) = 130 + f31DISTANCE + f32BATHROOMS + f33BETHANY + f34SBETHANY + f35NBETHANY + f36FENWICK + + f3s0CEANFRNT + YFRONT + f3lOCANALFRNT + f311SQFT + f312FIREPLACE + f313CARPORT + f314AGE For the 208 multi-family residential and commercial structures in the study, values have been estimated using a commercial appraisal guide. This involved categorizing each structure, documenting its age, and using replacement cost minus depreciation as an estimate of value.

PAGE 130

I powell 115 RESULTS The retreat costs presented in Table 1 are the result of using a shoreline migration rate of three feet per year, a figure generally used in the study area (Maunneyer and Carey, 1985), and a long-term discount rate of 3%. Table 1. Estimated cost of retreat by community (x $1 million). Decade Fenwick South Sea Bethany North Dewey Rehoboth Island Bethany Colony Beach Bethany Beach Beach 1990-1999 0.00 17.44 0.00 4.85 0.00 2.20 0.00 2000-2009 0.98 0.91 0.00 4.71 3.54 2.64 0.00 2010-2019 7.89 0.33 58.13 1.51 11.36 3.28 13.09 2020-2029 4.78 5.39 0.00 1.33 3.80 3.16 2.73 2030-2039 3.39 0.41 0.00 1.81 7.20 2.66 2.41 Total 17.04 24.48 58.13 14.21 25.90 13.94 18.23 DISCUSSION Tbe retreat costs estimated in this study may be most useful as a comparison to Delaware's current beach nourishment projects, or to other management options. However, making the comparison to beach nourishment is complicated by at least two factors. First, the oldest nourishment projects in the study area commtmities have been constructed and maintained since 1988, a relatively short period from which to glean longtem1 costs (Table 2). Second, levels of storm protection and recreational space afforded by the hypothetical retreat policy are almost certainly not equal to those being maintained tmder the current nourishment projects. The retreat model developed for this study is at least superficially comparable to Delaware's current beach nourishment projects in three conm1Unities in the study area. Field inspections performed in South Beihany Beach, Bethany Beach, and Dewey Beach before initial construc-Table 2. Nourishment costs, by conununity (x $1 million). Fenwick South Sea Bethany North Dewey Rehoboth Island Bethany Colony Beach Bethany Beach Beach 1988-1996 3.4 3.2 0.8 3.6 0.0 2.8 0.0

PAGE 131

116 Costs of Managing Delaware's Coast through a Policy of Retreat tion of beach nourishment projects found the reference feature at or near the footprint of most seaward buildings. Thus, nourishment was first undertaken under conditions similar to those that would trigger removal of oceanfront structures under the retreat model used in this study. An examination of Tables 1 and 2 brings out several points: The fact that 58 residential structures are located on the active beachface in the Town of South Bethany shifts much of the cost of retreat to the first decade in that community, greatly increasing the overall cost of retreat. In the relatively new conununity of North Bethany, the largest in the study area, most oceanfront development postdates Delaware's coastal construction setback regulations, which were implemented in 1981. This setback of structures lowers retreat costs significantly by postponing the initial removal of structures. It is also significant that North Bethany has been able to forego the construction of beach nourishment projects, yet storm damage in that community has been the lowest in the study area. The presence of seven large high-rise buildings in the relatively small community of Sea Colony increases retreat costs tremendously. Presumably, nourishment costs for Sea Colony'S small segment of beach will continue to be fairly low. This suggests that dense development practices lead to nourishment being relatively economical as a management strategy compared to retreat (perhaps a common-sense conclusion). In Fenwick Island, Sea Colony, North Bethany, and Rehoboth Beach, retreat cost estimates for the first decade are zero because no structures are forecast to be removed. This "cost-free" result is deceptive in that the decrease in beach width as the reference approached oceanfront structures will cause real economic losses in recreational value and storm protection over 1990 levels. Retreat costs may be inflated if property values have been enhanced by existing beach nourishment projects that have generally been cost shared on a statewide basis in the study area. Black et al. (1988) found that in South Bethany oceanfront property values would drop to salvage rates by 2000 unless shoreline erosion was checked. CONCLUSION The goal of this project was not to detennine whether retreat is the optimal shoreline management policy for the study area, but rather to estimate the cost of a hypothetical retreat policy under a variety of

PAGE 132

powell 117 shoreline migration rates, discount rates, and other variables. The nourishment costs are provided as context, and not for direct comparison. The continuation of Delaware's current policy of nourishment will be affected by many political and scientific factors beyond the scope of this study. Such factors would likely include the availability of sand resources, willingness by local communities to cost-share, technological advances, sea level rise, and others. In addition to the high cost of property acquisition found in this study, other factors such as the price and availability of insurance and disaster assistance, "takings" and other political difficulties associated with acquisition, the availability of relocation sites, and sea level rise, would have to be part of any consideration of a policy based in retreat. As additional nourishment projects are constructed in the study area, and as maintenance is performed on existing ones, the cost history of nourishment should become more reliable as an indicator of future costs. As additional development occurs, the structure database will be updated and refined. This will enable a more rigorous comparison of those approaches. At the same time, the U.S. Army Corps of Engineers has proposed a 50-year shore protection project along Delaware's Atlantic coast. While the federal funding of this project is uncertain, it represents a management policy offering higher levels of recreational benefit and storm protection than the retreat policy considered in this study. REFERENCES Beaches 2000 Planning Group 1988 "Beaches 2000 Planning Group: Report to the Governor." Black, David E., Lawrence P. Donnelley, and Russell F. Settle 1990 "An Equitable Arrangement for Financing Beach Nourishment Projects," Ocean and Shoreline Management 14. Maimneyer, Evelyn, and Wendy L. Carey 1985 Striking a Balance: A Guide to Coastal Processes and Management in Delaware. Delaware Department of Natural Resources and Environn1ental Control. State of Delaware 1984 "Beach Preservation Act," Title 7, Delaware Code, Chapter 68. U.S. Army Corps of Engineers 1992 Delaware Coast From Cape Henlopen to Fenwick Island Reconnaissance Report. U.S. Government Printing Office.

PAGE 133

This page IS intentionally blank

PAGE 134

Part 6 HYDROLOGY AND HYDRAULICS

PAGE 135

This page IS intentionally blank

PAGE 136

Preparation of a Hydrology Manual for Imperial County, California Steve R. Knell Imperial Irrigation District Theodore V. Hromadka Johannes J. DeVries Anders K. Egense Boyle Engineering Corporation INTRODUCTION The hydrology manual prepared for the Imperial Irrigation District (lID), in southern Imperial COlmty, California (see Figure 1), provides guidelines for the deternlination of stornl runoff for the design of flood management facilities, floodplain analysis, and drainage system design. The manual has been written specifically as part of a drainage master plan for lID (see Knell et aI., 1996). However, since the master plan area encompasses a major portion of Imperial County, discussions with the county for adoption of the manual as the cOlmty hydrology manual are ongoing, with the expectation that the county will adopt the manual. 'TI1e hydrology manual provides procedures for computing runoff from rainfall for specific frequencies and duration. Appropriate loss rate procedures are based on land use and soil types. Runoff for small subbasins is computed using the Rational Method. For areas above one square mile in area, unit hydrograph calculations are used to compute runoff. Stream flow routing procedures are defined for routing of flows between subbasin node points. The alllhors thank the Imperial Irrigation District for their support of this project alld for their permission to publish this paper.

PAGE 137

...... I\) I\) Imperial County I K:::::::::::::::::::::\:: N 110 Service I '& a Area 0
PAGE 138

Knell, Hromadka, De Vries, and Egense 123 HYDROLOGY MANUAL DEVELOPMENT APPROACH The major focus of the hydrology manual is to provide policy standards and guidelines for stormwater management. The results from application of the manual procedures should be consistent and fair, scientifically defensible and dependable, reliable and reproducible when applied by different users, and fairly easy to use. The procedures should also provide a reasonable "standard of care" (see for example, Nestlinger, 1990). Due to the uncertainty in establishing an accurate value for the peak discharge associated with a specified frequency of a flood, the parameters in the manual were chosen to provide an 85% confidence level for the flood discharges. This allows uncertainty in estimating peak discharge to be accounted for since on a regional basis only 15% of the design discharges will be "too small" in contrast to the 50% that would result from using the expected value of the peak discharge. RAINFALL Annual rainfall throughout Imperial County is very low, and intense shortduration rainfall events are responsible for most floods. A large portion of the county is below sea level (Imperial Valley and Salton Sea), but these areas are surrounded by mountains. A significant flood hazard is posed by streams originating in the mountains and draining into the valley areas. Two fonns of rainfall data were developed for the manual. For hydrugraph applications, a design storm is used. The defining parameters are stoml duration, point rainfall depth, areal depth adjustment, storm intem:ity, and time distribution of the rainfall. For the Rational Method cak,;lations, rainfall intensity-duration-frequency curves were developed. ,\ design rainfall procedure similar to that used for other counties (see, for example, San Joaquin County, 1996), was initially proposed to be used for this manual. The development of design storm procedures of this type is described by DeVries and Hromadka (1994). A review of the available rainfall data indicated that only one gage in the Imperial Valley (at EI Centro, see Figure 1) had data for durations of less than 24 hours. Although this gage had incremental rainfall values for durations as short as 5 minutes, it was not sufficient for defining design storms. Fortunately, regional data were available from the newly extended database for the publication that will replace National Oceanic and Atmospheric Administration (NOAA) Atlas 2 for the southwestern United States. The design stoml concept was therefore used in this manual to develop hypothetical design stomlS of various durations and frequencies for calculating runoff based on the recently published rainfall maps (NOAA, 1995). For a given stonn frequency, rainfall values are determined for specific durations (s-minute, lO-minute, IS-minute, etc.). These data are

PAGE 139

124 A Hydrology Manual for Imperial County, California used to compute incremental rainfall amounts (say for each 5-minute interval), which are then arranged into a pattern to form a design stornl for hydro graph calculations. The design storm pattern is based on a single synthetic 24-hour critical storm pattern that includes peak rainfall intensities from 5 minutes up to 24 hours. For small watersheds (usually under 5 square miles), only the peak 3-hour period of the storm is needed. LOSS RATE COMPUTATIONS The two watershed loss components of initial abstraction and infiltration are related to the hydrologic soil groups in the subarea being analyzed, soil cover and condition, and extent of watershed development. The major factor affecting loss rates is the nature of the soil itself, including surface characteristics, ability to convey water to subsurface layers, and storage capacity. Soils classified into the commonly used four hydrologic soil groups as defined by the U.S. Soil Conservation Service are Group A (low runoff potential), B (soil with moderate infiltration rates), C (soil having slow infiltration rates), and D (high runoff potential). Detailed ,oil survey information from the Soil Conservation Service was used to prepare maps of hydrologic soil groups. Specific vegetation types and !!le condition of the cover (poor, fair, or good) are also used to calculate loss rate, initial abstraction, and stonn runoff yield. RUNOFF ANALYSIS METHODS Relatively simple procedures have been fOlmd to give accurate estimation of discharges for design of project components for flood management projects. For small areas (less than one square mile) the well-known Rational Method has been found to provide a good estimation of the peak discharge. For larger areas, unit hydrograph procedures provide accurate determination of the runoff hydrograph. Effective rainfall is determined by calculating time-dependent losses and subtracting the losses from the gross rainfall. The two watershed loss components of initial abstraction and infiltration are incorporated in procedures of this manual. Rational Method Calculations For this method, the rainfall is defined by an intensity-duration-frequency (IDF) relationship (as an equation or in tabular fonn), and the runoff coefficient C is based on vegetation, cover density, infiltration capacity of the soil, and slope of the drainage area. The manual gives a confluence analysis procedure for estimating the peak flow by the Rational Method at the junction of two or more stream charmels. In this procedure, the Rational Method is used to estimate peak

PAGE 140

Knell, Hromadka, De Vries, and Egense 125 now by adjusting the catchment area to give a more realistic estimate of the contributing catchment area based on the critical duration of rainfall. Unit Hydrograph Calculations Unit hydrographs are determined from dimensionless S-graphs representative of the type of watershed being analyzed (Hromadka et aI., 1993). Individual S-graphs are used for valley, foothill, or mountain watersheds. S-graphs also may reflect urbanization, so that the watershed may be represented by a "valley-developed" S-graph or by a "valley-undeveloped" S-graph. Combinations of S-graph types can also be used. Base flows seldom occur in Imperial County streams, and any subsmface flow components of the runoff hydrograph that may occur are incorporated in the unit hydrograph response. Stream flow routing is used where routing may affect the runoff hydrograph. Reservoir routing is used to analyze the effects of detention basins on reducing peak discharges. TIle dimensionless distribution graph (or dimensionless S-graph) is a foml of a tmit hydrograph whose ordinates are expressed in terms of percentage of ultimate discharge and the time at which these discharges occur are fractions of the "basin lag." "Lag" for a watershed is the time (in hlJurs) from the beginning of a continuous series of tmit period effective rainfall to the instant when the rate of the watershed runoff equals 50% of the ultimate rate of the resulting runoff. The lag relates time relationships of the hydrograph to physical characteristics of the watershed. Lag times determined from calibration in other California counties have shown that lag is related to the time of concentration (T J used in Rational Method analyses. Here, the relationship between lag and time of concentration is: lag = 0.8 T c' Because the time of concentration is also ,ill important parameter for tmit hydrograph analysis determination, the hydrology manual provides procedures for calculation of Tc that also take into account the return frequency of the event being modeled. When lags determined from slmlmation hydrographs for several gaged are correlated to the hydrologic characteristics of other watersheds, an empirical relationship can be detemlined. This relationship can then be used to detemline the lag for drainage areas for which the hydroiogic characteristics can be determined, but for which distribution graphs are not available because of inadequate hydrologic data. Given the absence of more extensive site-specific data for Imperial Valley, this is Ule approach that is used for the hydrology manual. CONCLUSIONS In Ule preparation of the new hydrology for Imperial Valley, the new NOAA rainfall maps were judged to be the best source of design storm

PAGE 141

126 A Hydrology Manual for Imperial County, California data. To account for uncertainty in establishing an accurate value for the peak discharge associated with a specified flood frequency, the rainfall and loss rate parameters in the manual were chosen to provide an 85% confidence level for the flood discharges. This philosophy, which is similar to that used in Orange County and other Southern California county hydrology manuals, provides the necessary "standard of care" for hydrologic analyses based on the procedures described in this manual. REFERENCES DeVries, 1.1. and T.V. Hromadka 1994 "Development of Design Stornl Procedures for San Joaquin County," In Predicting Heavy Rainfall Events in California: A Symposium to Share Weather Pattern Knowledge. C. Dailey, ed., Sierra College, Rocklin, California. Hromadka, T.V. 1995 "Updating The Rational Method For Peak Flow Estimation," presented at the 1995 Arid West Conference, Association of State Floodplain Managers, and Arizona Floodplain Management Association, San Diego, California. Hromadka, T.V., R.H. McCuen, J.J. DeVries, and T.J. Durbin 1993 Computer Methods in Environmental and Water Resources Engineering. Mission Viejo, CA: Lighthouse Publications. Knell, Steve R, A.K. Egense, E.F. Shank, and T.V. Hromadka 1996 "Drainage Master Planning for the Largest Irrigation District in the U.S.," Association of State Floodplain Managers National Conference, San Diego, California. Nestlinger, A.J. 1990 "What is a County Hydrology Manual?" In J.J. DeVries, ed., Proceedings of a Workshop on County Hydrology Manuals, Mission Viejo, CA: Lighthouse Publications. National Oceanic and Atmospheric Administration 1995 "Draft Precipitation Frequency Maps./I Southwest Semiarid Precipitation Study. Silver Spring, MD: National Weather Service. San Joaquin County 1996 Hydrology Manualfor Department of Public Works. Stockton, CA.

PAGE 142

Flood Frequency Analysis in the Presence of Outliers, Historic Data, and Varying Generalized Skews Wilbert O. Thomas, Jr. David P. Preusch Michael Baker Jr., Inc. INTRODUCTION Bulletin 17B, Guidelines For Determining Flood Flow Frequency, was published by the Hydrology Subcommittee of the Interagency Advisory Committee on Water Data (IACWD) in 1982. Federal agencies were requested to use these guidelines for flood frequency analysis for gaged streams in all planning activities involving water and related land resources. Accordingly, the Federal Emergency Management Agency (FEMA) requested that Bulletin 17B guidelines be used for flood frequency analyses of gaged streams conducted in support of the National Flood Insurance Program (NFIP) (FEMA, 1995). Bulletin 17B recommended fitting the Pearson Type III distribution to the logarithms of the annual peak discharges using the sample moments to estimate the distribution parameters and provided procedures for (1) outlier detection and adjustment, (2) adjustment for historic data, (3) development of generalized skew, and (4) weighting of station and generalized skews. As stated in Bulletin 17B, "there is no procedure or set of procedures that can be adopted which, when rigidly applied to the available data, will accurately define the flood potential of any given watershed." As illustrated in this paper, the use of historic data, the detection and adjustment of outliers, and the choice of the appropriate skew value require that engineering judgment be applied in computing flood frequency estimates such as the 1 % annual chance flood. Annual flood data for two gaging stations in central Texas are used to illustrate the engineering judgement required in flood frequency analyses conducted in support of the NFIP.

PAGE 143

128 Flood Frequency Analysis with Outliers, Historic Data, Skews ADJUSTMENT FOR HISTORIC DATA AND HIGH OUTLIERS Two periods of record used in the Bulletin 17B flood frequency analysis are (1) systematic period of record (N), defined as the period of time when annual peak discharges are observed systematically through standard stream-gaging procedures; and (2) historic period of record (H), defined as an extended period of time for which the largest floods, either systematic or historic, are known. Historic peak discharges are those that occurred before or after systematic data collection, or during a break in the record. Information on historic floods is usually obtained in published and unpublished reports, newspaper files, or from local residents. Systematic peak discharges known to be the largest in the extended period H are called high outliers in Bulletin 17B. All systematic and historic peak discharges above some threshold must be known in the period H to be used in the historic adjustment. The threshold level is defined by the availability of historic information. A high-outlier threshold is computed in the Bulletin 17B analysis using a test described by Grubbs and Beck (1972). A systematic peak discharge does not have to exceed the high-outlier threshold to be included in the historic analysis. The computed high-outlier threshold is usually much larger than any systematic or historic peak discharge and is intended only as a guide in identifying systematic peak discharges that are sufficiently large to be considered for the historic adjustment. The statistical treatment of high outliers and historic peaks in Bulletin 17B is the same. The historic adjustment procedure is discussed by IACWD in Appendix 6 of Bulletin 17B and by Thomas (1985). GENERALIZED AND WEIGHTED SKEW In the Bulletin 17B analysis, the sample moments (mean, standard deviation, and station skew) are used to compute flood frequency estimates such as the 1 % annual chance flood. Because there is large uncertainty in computing station skew for sample sizes commonly available in flood frequency analysis, the station skew is weighted with a regional or generalized skew to obtain an improved estimate. The generalized skew is based on a regional analysis of several long-term stations in a hydrologically homogeneous region (IACWD, 1982). An example of a regional analysis of skew for the southwestern United States is given by Beard and Chang (1978). A generalized skew map is provided as Plate I of Bulletin 17B for those who prefer not to develop their own generalized skew procedures. A weighted skew is determined by weighting the station and generalized skew in inverse proportion to their individual mean square errors (IACWD, 1982).

PAGE 144

Thomas and Preusch 129 EXAMPLE OF HISTORIC ADJUSTM ENT WITH VARYING GENERALIZED SKEWS Annual flood data for the Guadalupe River at Comfort, Texas, can be used to illustrate the historic adjustment for high outliers and historic floods and the effect of varying generalized skew. Systematic flood data are available at Comfort from 1919-32 and from 1939 to 1994 (Figure O. For some unknown reason, the systematic peak discharges for 1920, 1921, 1923, and 1924 are not available. Historic flood data are available for the floods that occurred in 1900, 1915, 1935, and 1936. Note that the 1935 and 1936 floods occurred during a break in systematic stream gaging from 1933 to 1938 and are therefore historic peaks. 2 5 .= 8 IJ. 300 ------------------------* Hlltorle Period \ '1 ... f-:-j f+\ ... ----+l.\ 250 Record I --200 : = Hiltorie Flood ... .5 150 1-"8* 0 ii: 8. 100 iii * ---. 1--J----t------li-------t-----j II 1Il040 1MO 1MO 1900 1Q20 11140 11180 11180 2000 1850 1870 1880 1Q10 1Q30 1Q5() 1Q70 1Q11() WaterY_ Figure 1. Annual maximum peak discharges for Guadalupe River at Comfort, Texas. The peak of record, 240,000 cubic feet per second (cfs), occurred in 1978 during systematic stream gaging and is considered the largest flood at Comfort since 1847 (H=147 years). The next highest floods are the 1900 and 1932 floods with discharge values of 182,000 cfs. Historic information is not sufficient to detennine if any flood exceeded 182,000 cfs from 1848 to 1899. However, the historic information indicates it is probable that all floods equal to or greater than 107,000 cfs are known from 1900 to 1994 (H =95 years). Given the available historic information, four different scenarios can be assunled for the historic adjustment: Scenario I-use only systematic record, do not utilize any historic

PAGE 145

130 Flood Frequency Analysis with Outliers, Historic Data, Skews infonnation; therefore, no threshold (provided for comparison purposes); Scenario 2-the 1978 flood is the largest from 1848 to 1994 (H=147 years); therefore, threshold is 240,000 cfs; Scenario 3-the 1900, 1932, and 1978 floods are the largest from 1900 to 1994 (H=95 years); therefore, threshold is 182,000 cfs; and Scenario 4-the 1900, 1915, 1919, 1932, 1935, 1936, 1960, 1978, and 1987 floods are the largest from 1900 to 1994 (H=95 years); therefore, threshold is 107,000 cfs. Note that the threshold is determined as the lowest discharge of all floods known in H years, not the high-outlier threshold (computed as 729,000 cfs). To illustrate the results of the different assumptions about historic data, the 1 % annual chance flood (Q.OI) is sununarized in Table 1 for the four scenarios and weighted/generalized skews (IACWD, 1982); Beard and Chang, 1978). The data is listed by scenario in order of the number of peak discharges adjusted for historic information. Table 1. Summary of 1 % annual chance flood discharges Q.01 for the Guadalupe River at Comfort, Texas, for different scenarios and weighted/generalized skl"w. Number of Bulletin 17B Skew Beard and Chang Skell Scenario H Iyears) Threshold Icfs) Adjusted Peaks QOIiill.l QOliill.l I 0 0 278,000 252,000 2 147 240,000 I 248,000 235.000 3 95 182,000 3 275,000 255.000 4 95 107,000 9 301,000 278,000 Two choices of generalized skew exist for computing the weighted skew. The runs using the Beard and Chang (1978) generalized skew are recommended because that study was based on data for more stations in the vicinity of the Guadalupe River basin than the Bulletin 17B nationwide analysis. Given the choice of generalized skew and fact that the 1978 flood (240,000 cfs) is probably the largest flood in 147 years, an estimate of 235,000 cfs appears more reasonable for Q.OI' Since only one threshold level (240,000 cfs) and historic period (H=147 years) are used in the historic adjustment analysis (Scenario 2), the historical floods of 1900,1915, 1935, and 1936 are not used in the analysis. DETECTION AND ADJUSTMENT FOR LOW OUTLIERS Low outliers are extreme peak discharges that depart from the trend of the rest of the data. They are usually assumed to be due to statistical sampling

PAGE 146

Thomas and Preusch 131 variability, not measurement error. The detection of low (and high) outliers in Bulletin 17B is based on a one-sided 10% level of significance test for a normal distribution (Grubbs and Beck, 1972). Thomas (1985) describes the basis for this statistical test. All peak discharges below the low-outlier threshold are identified as low outliers, are automatically censored, and a conditional probability adjustment is used to account for the fact that one or more peak discharges is below the threshold (Appendix 5, Bulletin 17B). EXAMPLE OF LOW OUTLIER DETECTION AND ADJUSTMENT Annual flood data for the North Llano River near Junction, Texas, can be used to illustrate the detection and adjustment of low outliers. Systematic flood data are available near Junction from 1916 to 1977. Historic data are also available for the 1889 flood (84,000 cfs) that is considered the largest flood from 1875 until the initiation of systematic stream gaging in 1916. A major flood occurred in 1936 (102,000 cfs) that is larger than the 1889 historic flood. Using Bulletin 17B guidelines, Q.Ol is estimated to be 231,000 cfs. The frequency curve for North Llano River near Junction using a historic threshold of 84,000 cfs, H= 103 years and a weighted based on the Beard and Chang (1978) generalized skew is shown in Figure 2. Given that the 1936 flood (102,000 cfs) is considered the largest Q Z = 10' rr---,---,--,---,---,.-----,--,---,--..,--..--.---.--.---r---. o + o + BUll. 11-8 frequency 5ys I emo tic ptok s His lor i co I ad jus I ed 231,000c15 __ ..., ......... 10' 10' NOTICE -Preliminary computalion User is responsible for ossessmenl Qnd intilprellliion. 10 99 5 98 95 90 80 70 50 JO 20 10 2 1 0.5 0.2 ANNUAL [XC[[OANC[ PROBAB IlllY, PERCENI Figure 2. Frequency curve for North Llano River near Jlmction, Texas, without censoring.

PAGE 147

132 Flood Frequency Analysis with Outliers, Historic Data, Skews flood from 1875 to 1977 (103 years), an estimate of 231,000 cfs appears unusually high for Q 01' As illustrated in Figure 2, the computed flood frequency curve at the 1 % annual chance flood (231,000 cfs) is higher than the corresponding plotting position. Although no low outliers were identified in the JlIDction analysis (low-outlier threshold was 50 cfs), there are a few low peak discharges that depart from the trend of the rest of the data (see Figure 2) and these values tend to inflate the standard deviation and the estimate of Q.OI' An iterative censoring of low peak discharges was one of the approaches evaluated by the federal interagency work group that selected the low-outlier test used in Bulletin 17B (Thomas, 1985). A similar approach was employed in the Junction analysis and the results are given in Table 2. The censoring level in Table 2 was chosen at noticeable breaks in the data. Table 2. Sununary of the 1 % atillual chance flood (Q.Ol) for North Llano River near Junction, Texas, with iterative censoring of low peak discharges. Low threshold Ccfs) o 80 120 400 800 1,700 3,000 4,200 5,000 NlIDlber of peaks censored o 1 2 6 8 12 18 22 24 Q.O I Ccfs) 216,000 222,000 230,000 218,000 213,000 195,000 167,000 150,000 141,000 Station skew, rather than weighted skew, was used in the iterative censoring analysis because it provides a frequency curve that is more consistent with the frequency estimates based on plotting positions and with the historic floods experienced by the community. Censoring the two. annual peak discharges (78 and 117 cfs) causes Q.OI to mcrease slIghtly as the skew becomes more positive. Continued censoring causes the standard deviation and Q.01 to decrease. The objective of the iterative censoring procedure is to continue censoring lIDtil there is a minimal change in Q.Ol' The recommended analysis is a low-outlier threshold of 5,?00 cfs and an estimate of Q.o I of 141,000 cfs. The choice of the low-outlIer threshold was based primarily on a visual comparison of the computed frequency curve with the Wei bull plotting positions of the data and comparison with an estimate of 136,000 cfs from published

PAGE 148

Thomas and Preusch regional regression equations of Schroeder and Massey (1977). An estimate of 141,000 cfs is also more consistent with the historic data available for the North Llano River near Junction, Texas. SUMMARY 133 The use of historic data, the detection and adjustment of outliers, and the choice of skew require judgement in computing frequency estimates as part of the NFIP. The difference between systematic and historic flood peaks was defined and illustrated using data for the Guadalupe River at Comfort, Texas. The adjustment for historic data was described and the effect on Q.01 of various assumptions about historic information was ill ustrated. The detection and adjustment of low outliers often has a significant impact on the computation of extreme flood frequency estimates such as the 1 % annual chance flood. Often, it is necessary to raise the low-outlier threshold above the value computed by the low-outlier test. An iterative procedure for choosing a low-outlier threshold, in conjunction with a graphical comparison of the computed flood frequency curve with plotting positions of the data, was described and illustrated for the North Llano River near Junction, Texas. Finally, the choice of skew can significantly impact the analysis. Different estimates of generalized skew are usually available and an individual must choose the one considered most appropriate. The choice of generalized skew in the Guadalupe River example made a difference of approximately 5 to 10% in Q.Ol' although the difference could be greater for other exanlples. Sometimes, station skew should be used rather than a weighted skew, as illustrated in the North Llano River exanlple. The Bulletin 17B method is well documented and tested. In general, reasonable results can be obtained with a straightforward application of this methodology. The examples provided in this paper are provided to iliustrate some of the instances when engineering judgement and v,!riations from the guidelines are needed to obtain reasonable frequency estimates. REFERENCES Beard, Leo R., and Shin Chang J 978 Generalized Skew Coefficients of Annual Maximum Streamflow Logarithms in Southwestern Division, Corps of Engineers. Technical Report of the Center for Research in Water Resources, Austin, TX: University of Texas.

PAGE 149

134 Flood Frequency Analysis with Outliers, Historic Data, Skews Federal Emergency Management Agency 1995 Flood Insurance Study Guidelines and Specifications for Study Contractors. FEMA 37. Washington, D.C.: FEMA. Grubbs, Frank E., and Glenn Beck 1972 "Extension of Sanlple Sizes and Percentage Points for Significance Tests of Outlying Observations," Technometrics 14 (4):847-854. Interagency Agency Advisory Committee on Water Data 1982 Guidelines For Determining Flood Flow Frequency. Bulletin 17B of the Hydrology Subcommittee. Reston, VA: U.S. Geological Survey, Office of Water Data Coordination. Schroeder, Elmer E., and Bernard C. Massey 1977 Techniques for Estimating the Magnitude and Frequency of Floods in Texas. U.S. Geological Survey Water-Resources Investigations Report 77-110. Thomas, Wilbert 0., Jr. 1985 "A Uniform Technique for Flood Frequency Analysis," ASCE Journal of Water Resources Planning and Management 111 (3): 321-337.

PAGE 150

Evaluating Storm Water Runoff from Steep Slope Arid Lands Clifford E. Anderson University of New Mexico Richard J. Heggen University of New Mexico INTRODUCTION Like many cities in the western United States, Albuquerque, New Mexico, receives significant runoff from adjacent steep mountain areas. The Sandia and Manzano mOlmtains east of Albuquerque have a crest that is 4000 to 5000 feet higher than the urbanized areas at the base of the mountain. Contributing basins four to six miles long with slopes of 10 to 20% are conmlOn. In the upper portions of the watersheds, slopes greater than 20% are encOlmtered. West of the mOlmtain front, alluvial arroyos with slopes of 2 to 4% carry water to the Rio Grande valley. While natural slopes between 2 and 4 % would be considered steep in many cOllmllmities, it is the slopes steeper than 4 % that provide the greatest uncertainty for analysis. The standard procedures for computing lag time and time of concentration are reasonable for slopes up to 4%. For the purposes of this discussion, steep slopes mean those greater than 4%. THE ISSUE OF FROUDE NUMBERS The lag time and time of concentration are related to the velocity of channel flow. In a steeply sloped watershed, some empirical equations for lag time and time of concentration may indicate high velocities and sllpercritical flow conditions. Trieste (1992) shows that supercritical flows do not occur for any extended reach of a natural charmel, and that supercritical flows occur only in reaches of 7.6 meters or less. A discussion of this paper by Wahl (1994), presents some gage data that contradicts this assunlption for infrequent large discharges at four natural charmels. Based on the bulk of information available, it may be reasonable to assume that flows near the critical condition, or only slightly

PAGE 151

136 Evaluating Storm Water Runoff from Steep Slope Arid Lands supercritical, represent a reasonable upper limit for flow velocities at the small, steep slopes commonly found in the Albuquerque area. When considering the full length of a watershed, use of Froude numbers between 0.9 and 1.0 may be a reasonable maximum. COMBINING CRITICAL DEPTH AND MANNING'S EQUATIONS If the critical velocity is used as an upper limit for natural channels, formula No. 84 from the DAMBRK Model (Fread, 1988) is applicable to define the critical slope, Dc = 77000 n2 / yl/3 Where: Dc n y channel slope in feet per mile Manning's roughness coefficient depth of flow in feet (1) This equation can be obtained by combining the critical depth equation with Manning's equation for a wide rectangular channel. For conditions where the Froude nunlber (Fr) is not equal to 1.0, and supercritical or subcritical flow exists, a similar equation of the following form can be established: S = 1.486-2 0 g n 2 O y-O.333 (2) Where: S slope in foot per foot Fr Froude nunlber g gravitational acceleration in feet per second per second This equation can be refonnulated to obtain equations for Manning's roughness coefficient and velocity as follows: n = 1.486 g-0.533 Qo.o667 U-O 0667 SO.5 Fr-I.0667 (3) v = g0.4 QO.2 U-O 2 FrO.8 (4) Where: Q average flow along channel length in cubic feet per second U channel width-to-depth ratio V velocity in feet per second Equations 3 and 4 can be used to generate information about the parameters for steep slope areas. If Froude numbers are assumed to have an upper limit near 1.0, and the width-to-depth factor (U) is nearly constant, then for any given flow rate (Q), equation 3 shows that n varies with the square root of the slope. If the slope is doubled, then n values

PAGE 152

Anderson and Heggen 137 would be expected to increase by a factor of 1.44. With a similar consideration of equation 4, the velocity at a given flow rate does not vary with the slope. This simple comparison is consistent with the results of some recent investigations of steep slope areas. SNYDER'S LAG TIME EQUATION When using Snyder's synthetic unit hydrograph method, the lag time (or time to peak) for the Albuquerque area has been computed by the Corps of Engineers, Lg = 24 Kn (M Mea / D.5) 0.36 (5) Where: Kn visually estimated mean of Manning's n-value for average channel D average channel slope in feet per mile M length of main channel in miles Mea = travel length to the centroid of the basin in miles Lg = lag time in hours A similar form of this equation with different exponents is used by the Corps of Engineers in other western states. An alternate form of this equation commonly used by the U.S. Department of the Interior Bureau of Reclamation is: (6) Since equation 6 uses Manning's n-value as the Kn value, it is possible to substitute the n-value from equation 3 into Kn and use this value to compute the lag time. For the Rocky MOlmtains (New Mexico, Colorado, Utah, Wyoming, Montana, Idaho, Oregon), Kn values between 0.056 and 0.339 are reported in Design of Small Dams (U.S. Bureau of Reclamation, 1987). Experience with using equations 3 and 6 shows that Kn values are commonly greater than 0.10 but are consistent with values reported by the Bureau of Reclamation. However, when this substitution of n for Kn is used, the lag times computed frequently exceed the lag times computed for flatter slope conditions, suggesting a problem with use of this approach. REGRESSION EQUATIONS FOR HIGH-GRADIENT STREAMS Jarrett (1984) presented equations to predict Manning's roughness coefficient and velocity using multiple-regression analysis from measured watersheds. Jarrett's work resulted in the following equation for velocity:

PAGE 153

138 Evaluating Storm Water Runoff from Steep Slope Arid Lands v = 3.81 R O .83 SO.12 (7) Where: R hydraulic radius in feet S friction slope in foot per foot Values for hydraulic radius and hydraulic depth were approximately the same, and could be used interchangeably. The difficulty with this approach is the ability to detemline an estimate of the hydraulic radius (R). Jarrett's equation for velocity can be combined with the Manning and Froude number equations to obtain: V = 3.81 g-0.166 QO.332 FrO 332 U-O 332 SO.12 (8) Because this equation is based on measured watersheds, this equation may be superior to equations 4 and 6, where the actual watershed is similar to the conditions of Jarrett's watersheds. Equation 8 replaces the uncertainty of detemlining the depth of flow with the uncertainty of determining the width-to-depth ratio. Ugarte and Madrid (1994) developed an equation to determine Manning's flow resistance based on studies of 19 rivers in Chile. Their equation has the following fonn (converted to US customary units): n = 1.485 [ 0.183+ In ((1.3014 SO.0785 (R/D84)0.0211) /FrO.2054) ] (84).1667 g-0.5 (0) Where: 0xx channel bed grain size for which xx percent b) weight is finer This equation is reported to achieve a mean error of 2.2% for the measured streanlS. One deficiency with the use of this method is the ability to determine 084 in ephemeral arroyos. This value may be readily established for an existing flow. It is a much more complex problem to estimate the average particle size for the projected flood flows in a normally dry arroyo. Rickenmann (1994) reported on equations developed from studies of steep watersheds in Switzerland. For slopes above 0.6%, Rickenmann proposed the following equation based on a regression analysis: V = 0.37 gO.33 QO.34 SO.20 -0 .35 (10) This equation is reported to produce accurate results using physically based input data. Again, one difficulty with ephemeral arroyos is the

PAGE 154

Anderson and Heggen 139 determination of 090, Rickenmann also presents a regression equation for the width of flow for a steeply sloped channel with the following form: W = 5.01 QO.32 090.21 g-0.16 S-O.3S (11) This equation may have particular value to the determination of geomorphologic parameters from steep channel flow, and for the width-todepth ratio in particular. WATERSHED SLOPE ADJUSTMENT An alternative procedure to the adjustment of Snyder's Kn factor and the of the regression equations for velocity or roughness coefficient is to adjust the effective watershed slope for the steep slope conditions. This method is presented in graphic form in Figure 4-1 of the Runoff Chapter of the Denver Urban Storm Drainage Criteria Manual. An equation for the curve in Figure 4-1 was developed for the Albuquerque Development Process Manual with the following form: S' = 0.052467 + (0.063627 S) 0.18197 e(-62.37S*S) (12) Where: S measured slope in foot per foot S' adjusted slope in foot per foot No extensive documentation has been located on the deviation of Figure 4-1 other than the following brief explanation in the Denver manual: "In natural and grass lined drainage ways, channels become unstable when a Froude nunlber of 1.0 is approached. There are natural processes at work that limit the time to peak of a unit hydrograph as the drainageway becomes steeper." One way of considering the adjustment of channel slopes is that the steeper watershed channels form a series of short cascading flow reaches with the effective channel velocity being related to the flatter sloped bench areas and not to any vertical drops. As represented by equation 12, for a given Manning's n value and flow rate, the adjusted slope may be only slightly increased, and further increasing the slope has only a small impact on the velocity. The application of the Denver procedure seems to give reasonable answers when used with commonly used lag time and velocity equations, and currently equation 12 is recolDDlended for steep slope use in the City of Albuquerque hydrology manual (1991). An additional check for the Froude number should also be made when using the slope adjustment, to confine the Froude number to a range (suggested at 0.9
PAGE 155

140 Evaluating Storm Water Runoff from Steep Slope Arid Lands v = 1.4860.75 n-O .75 QO.25 UO .25 S.375 (13) When the actual slope is flatter than the slope computed by equation 12, the actual slope should be used. SUMMARY AND CONCLUSIONS Use of any of the equations and procedures identified herein will require extensive amounts of judgment. Some elements, such as channel slopes, basin lengths, and lengths to the centroid, can be directly measured. Other elements, such as Froude number, width-to-depth ratio, average flow depth, and unadjusted Kn require a thorough understanding of the local watershed properties and are not directly measurable. These procedures do not represent the total extent of methods available to account for steep slope conditions and further refinements are expected as further analysis and field data become available. A wide range of results can be obtained with the various procedures available, even among the formulas based on field data. This suggests that there are many problems with accurate and representative data acquisition. Most of the data comes from measurements of normal flow levels in perennial streams. Only the limited data by Wahl (1984) appears to include flow events significantly above the median flow. Wahl's data suggests that Froude numbers higher than 1.0 are found. The regime of normal flows may establish a maximunl Froude nunlber of 1.0, but uncommon events may produce different flow conditions. In arid climates the sedimentation and debris flows that accompany major runoff can profoundly alter flow assumptions. The flow properties are much more complex when the channel bed is also flowing with the water. More measurement and analysis are required; it is expected that the best procedure will be constantly changing in the coming years. Meanwhile, engineers and hydrologists are faced with predicting events that can affect peoples lives with very little definitive data. REFERENCES The D.P.M. Drainage Design Criteria Committee 1993 Section 22.2, Hydrology of the Development Process Manual, Volume 2, Design Criteria. Albuquerque, NM: For the City of Albuquerque in cooperation with Bernalillo County and the Albuquerque Metropolitan Arroyo Flood Control Authority.

PAGE 156

Anderson and Heggen 141 Fread, D.L. 1988 The NWS DAMBRK Model: Theoretical Background/User Documentation. Hydrologic Research Laboratory. Silver Spring, MD: National Weather Service. Jarrett, RD. 1984 "Hydraulics of High-gradient Streams." Journal of Hydraulic Engineering 110 (11). Rickenmann, Dieter 1994 "An Alternative Equation for the Mean Velocity in Gravel-bed Rivers and Mountain Torrents." pp. 672-676 in Hydraulic Engineering '94, Proceedings of the 1994 Conference, G.V. Cotroneo and RR Rummer, eds. New York: American Society of Civil Engineers. Trieste, Douglas J. 1992 "Evaluation of Supercritical/Subcritical Flows in High-Gradient Channel." Journal of Hydraulic Engineering 118 (8). Ugarte, Alfonso and Manuel Madrid 1994 "Roughness Coefficient in Mountain Rivers." pp. 653-656 in Hydraulic Engineering '94, Proceedings of the 1994 Conference, G.V. Cotroneo and RR Rununcr, eds. New York: American Society of Civil Engineers. Urban Drainage and Flood Control District 1969 Urban Storm Drainage Criteria Manual, Volume 1. Denver, CO. Wahl, Kenneth L. 1994 "Evaluation of Supercritical/Subcritical Flows in High Gradient Channel, Discussion of Kenneth L. Wahl." Journal of Hydraulic Engineering 120 (2).

PAGE 157

Approaches to In-Situ Calculation of Floodplain Roughness Barry Hecht Jonathan Owens Balance Hydrologics, Inc. In-situ roughness calculations provide a defensible basis for estimating roughness coefficients in designing or maintaining complex, multi-objective floodways. Conventional methods using standard manuals can work well for straight channels containing little vegetation, but are not really amenable to more complicated conditions typical of many floodways or river corridors. For example, the incremental method, described in Chow (1959) and many regional variants, use additive values for surface irregularities, obstruction, and variations in shape and size of channel cross section. These values are picked from a table, not calculated, and are therefore open to interpretation. The methods of Barnes (1967) and Arcement and Schneider (1989), which use photographs of sites where roughness has been calculated, also employ subjective choices; this can be problematic if none of the channels matches exactly. A better option is to calculate the hydraulic roughness in the actual channel of interest. The key to an accurate roughness coefficient lies in calculating rouglmess based on high-water marks (HWMs) and local conditions (slope and channel geometry). In this paper, we outline five steps that allow field data to be applied easily and cost-effectively to channel-management decisions. We also present t1lree selected case studies from the San Francisco and Monterey Bay areas where site-specific calculations of roughness proved to be both effective and central in reaching a management decision. Several findings pertinent to all three studies are presented in the final section. METHOD We encourage the use of roughness coefficients that have been calculated locally based on HWMs. We input known values for all parameters into the Manning equation and then solve for the roughness coefficient ('n'). The calculation of roughness can be easily done for low-or medium-flow conditions, because the actual water level can be recorded and the

PAGE 158

Hecht and Owens 143 discharge measured. However, roughness often decreases as stage increases, so values obtained at lowor medium-flow conditions must be carefully applied to the large flows that are of the greatest concern. The local data can most effectively be applied in five steps: (1) Validating-or calculating-peak discharge for storms that correspond to identifiable HWMs. (2) Calculating roughness based on field measurements of wetted perimeter, cross-sectional area, and water surface slopes. These measurements are based on a survey of the channel and HWMs. (3) Estimating roughness for flows at the design levels, using one or more of the accepted techniques of extrapolating roughness values obtained in step 2. (4) Assessing effects of likely changes in the channel or overbank areas. (5) Evaluating whether anticipatable episodic events are likely to fundamentally change the assumptions of the calculations, and adjusting accordingly. CASE STUDIES We have applied this approach to three California streams: a leveed river with an inboard riparian-woodland fringe, a naturalized channel established within an over-wide leveed floodway, and a deeply incised natural stream in a narrow riparian corridor. Case Study 1: Pajaro River near Watsonville, California Floodplain managers must often choose among alternative approaches to bank protection. Sometimes, such choices have enormous cost, public safety, regulatory, and conmlUnity planning ramifications. If representative reaches of the stream (or a nearby channel) already have some of the bank-protection measures in place, the actual perfornlance of these measures in that stream can be assessed. We made measurements of this type on the Pajaro River near Watsonville. A federal flood control project was designed in the late 1940s and constructed in the early 1950s, with a design capacity of 22,000 cfs. Approximately 15 miles of the Pajaro River were leveed. Since the levees were built, a narrow band of riparian woodland had become established along the river, usually occupying about half of the floodplain "bench"

PAGE 159

144 In-Situ Calculation of Floodplain Roughness within the levees. The involved agencies sought to quantify loss of flood conveyance associated with this fringe of willow, cottonwood, box elder and elderberry woodland. Our approach was to compare the roughnesses of two consecutive straight reaches, the first with riparian woodland on both banks, and the second with one wooded bank and one cleared bank fully riprapped with angular 6-inch quarry rock. The longitudinal slopes, woodland densities, peak discharges, and bed conditions were observed to be very similar in these two consecutive reaches. We posited that the difference in conveyance could be computed by doubling the difference in measured roughnesses between the fully-wooded and half-riprapped reaches. We computed the observed hydraulic roughness at the peaks of three events by solving the Manning's equation for 'n.' Peak discharge was obtained from the nearby Chittenden gaging station, with minor adjustments for tributary inflow. Cross-sectional area, hydraulic radius, and hydraulic slope were measured from high-water marks of flood crests corresponding roughly to 75, 33, and 15% of the design capacity. Results (Table 1) indicate that whole-channel roughness was essentially identical in the two reaches at the highest stages measured in this study, except for one section containing two large snags and rootwads. Roughness of the March 2 crest could be estimated from the March 25 flows, adjusted by the -0.4 power of the peak flow, as would be predicted from the at-a-station hydraulic geometry for relatively wide streams. The wooded reach was distinctly rougher (n=0.068) than the half riprapped reach (n=0.053 to 0.058) during the March 25 storm. Differences increased at the lower-stage event on April 30, when flows were actually below the riparian woodland. Table l. Hydraulic roughness values for Case Study 1, Pajaro River near Watsonville, California. Method HWM HWM HWM Date 2-Mar-83 25-Mar-83 30-Apr-83 Flow 16,210 cfs 7,000 cfs 3,260 cfs Station 4+97 0.046 0.058 0.044 8+10 0.043 0.053 0.05 10+00 0.039 0.053 0.043 15+00 0.042 0.068 0.068 15+90 0.054 0.095 0.081 Notes: Design flow = 22,000 cfs. Peak discharges assumed be those gaged at Chittenden Channel Condition Half-riprapped Half-riprapped Half-riprapped Fully Wooded wlo fallen trees Fully Wooded wI fallen trees

PAGE 160

Hecht and Owens 145 The difference between the two reaches at the lower flows is probably attributable more to: (1) leafing out-the deciduous woodland is in bud during early March and comes into full leaf by early April, creating considerably more roughness-generating surface area, and bending lower branches into the flow; and (2) more uncleared woody debris along the lower banks in the fully-wooded area. Nonetheless, density of roughnessproducing vegetation along the bank at stages when the main channel is nearly at the design capacity does not seem to appreciably affect channel hydraulics at the higher flows of prime concern for flood protection. Case Study 2: Wildcat Creek, Contra Costa County, California The study reach in northern RichnlOnd, California, is a multi-level floodway constructed in the late 1980s by the U.S. Army Corps of Engineers to reduce local urban flooding. The constructed flood control channel is about four times wider and two times deeper than the preexisting natural channel. Drainage area above the study reach is approximately 8.5 square miles; mean annual precipitation is 24 inches. An unusual feature of the project is in-channel vegetation, planned to reduce roughness by shading out undergrowth. However, the planted vegetation has not yet fully matured and currently causes significant hydraulic resistance. The goal of this case study involved evaluating the flood protection provided, and determining maintenance requirements (vegetation and sediment removal). The key aspect of the evaluation concerned assigning roughness values at cross sections representative of charmel reaches. A U.S. Geological Survey gaging station approximately 1.5 miles upstream was our source of information for historic and recent flows. Orr and Owens (1994) applied numerous methods to estimate rouglmess, but found the conventional methods (mentioned in the introduction) difficult to apply because of dense vegetation directly in the channel. None of the locations presented in Barnes (1967) or Arcement arld Schneider (1989) looked at all like Wildcat Creek. In the fall of 1994 they found HWMs at two of four cross sections, but those marks corresponded to a flow less than 1/7 of the design flow. Flexible vegetation, cattails and young willows, in the channel presented difficulty in extrapolating from the low flow of 303 cfs to the design flow of 2300 cfs. Subsequent high flows of the winter of 1995 left fresh HWMs. Following up on the work Orr and Owens, we identified HWMs corresponding to 1310 cfs (January 1995), and personally marked water levels during a later-season storm at a stage corresponding to 916 cfs (March 1995). We found that sediment deposition, which occurred significantly at three of the four cross sections, caused a decrease in calculated roughness

PAGE 161

146 In-Situ Calculation of Floodplain Roughness from January to March 1995 (see Table 2). The fourth cross section (66+00, where little deposition occurred) displayed the normal pattern of decreased roughness at higher flows. Flood flows bent many of the cattails, which then became buried by sediment, reducing roughness. Although roughness values estimated using the incremental method are a reasonable approximation of the calculated values (Table 2), calculations based on measurements are more defensible. Table 2. Hydraulic roughness values for Case Study 2, Wildcat Creek, Contra Costa County, California. "Modified" Incremental Incremental Vegetation Method HWM HWM HWM Method Cattails Prone Density Flow 303 cfs 1310 cfs 916 cfs design design design Station 66+00 0.03-0.067 0.043 0.054 0.043 0.041 0.048 83+11 na lM!M 0.043 0.051 0.047 0.062 93+00 na l1l!@ 0.047 0.07 0.058 0.227 96+25 0.185 !U.42 0.047 0.088 0.069 0.254 Channel willows and willows and willows and willows and willows and willows and Condition thick cattails bent cattails no cattails thick cattails bent cattails thick cattails fresh sediment Notes: Design flow (Q I 00) = 2300 cfs. Flow of 303 cfs occurred Feb. 19, 1994. Channel geometry assumed as surveyed Oct. 1994. Flow of 1310 cfs occurred Jan. 9, 1995. Channel geometry assumed as surveyed Oct. 1994. Flow of916 cfs occurred March II, 1995. Channel geometry assumed as surveyed Nov. 1995. Considerable sediment accumulated between the Jan. and March 1995 storms, except at station 66-+00. Range of values at 66+00 indicates the thickness of the HWM (debris jam). Case Study 3: San Francisquito Creek at Webb Ranch, Stanford, California Hall and Freeman design 0.044 0.053 0.071 0.100 willows and thick caltails We were asked to calculate the level of the 10-year event on San Francisquito Creek to guide design of a service road bridge. The analysis was first done using the conventional incremental method, adjusting for channel irregularity, cross-sectional variability, obstructions, vegetation, and meandering. We used the Aldridge and Garrett (1973) adaptation of Chow's method, developed for streams in Arizona. Values for four cross sections in a 700-foot reach resulted in an estimated 'n' value of 0.098 for this perennial channel cut into cohesive banks, lined by a riparian woodland with alder, buckeye, cottonwood, willow, and bay laurel. We subsequently returned to develop actual crest of-event roughnesses based on high-water marks from the January 1982,

PAGE 162

Hecht and Owens 147 January 1995, and February 1996 stonns. Estimated recurrence intervals for these events are approximately 25, 7, and 2 years, respectively. Our calculated values of roughness decreased with increasing flow, from an average of 0.067 (2-year event), through 0.053 (7-year event) to 0.047 (25-year event). Roughness calculated from the flows that left the three HWMs varied with the -0.3 power of discharge in this narrow channel. Roughnesses based on the site-specific field data averaged 0.047 (for the highest HWM), or 48% of the 0.098 estimated using the incremental method. We ascribe much of the difference to the very sparse tmdergrowth along the charmel at stages below the elevation of the 10-year event, because the undergrowth has been shaded out by the tree canopy that extends completely across the charmel at most locations. Most variants of the Chow incremental method assume presence of weeds, bushy willows, or shrubs within the area inundated by moderate-recurrence events. Shading out occurs widely in western streams with low to moderate width:depth ratios, but is not recognized by this method. Also, we estimated roughness values while vegetation was fully leafed, and we may have overestimated roughness of flows that occurred before the vegetation was in leaf. CONCLUSIONS (1) In-situ measurements of roughness offer a valid, defensible alternative to standard 'cookbook' estimates of Marmings In'; measured values are particularly suited for complex multi-objective floodways. (2) If high-water marks can be identified and assigned to a particular flood crest, the 5-step approach outlined in this paper can speed in-situ roughness calculations and make them more valid and versatile. (3) Om data suggest that it may be feasible to estimate roughnesses of wooded riparian corridors at stages near design capacity from the ratio of the observed peak discharge to the design flow, raised to an exponent of about -0.3 (narrow channels) or -0.4 (wide channels), consistent with at-a station hydraulic geometries, provided that the observed peak flows were sufficiently high to be affected by the naturalized woody fringe. (4) In-situ measurements help adapt for changes in charmels or for episodic bed sedimentation. (5) In-situ measurements of roughness in woodland-lined charmels are often lower than might be calculated from manuals, perhaps because peak floods occur when the trees are not in leaf, or because the maturing

PAGE 163

148 In-Situ Calculation of Floodplain Roughness woodland has shaded out undergrowth that most standard manuals assWUe to occur in all channels. REFERENCES Aldridge, RN. and lM. Garrett 1973 Roughness Characteristics for Stream Channels in Arizona. US. Geological Survey Open-File Report. Arcement, George, J. Jr., and Verne R. Schneider 1989 Guide for Selecting Manning's Roughness Coefficients for Natural Channels and Flood Plains. US. Geological Survey Water Supply Paper 2339. Barnes. Harry, H. 1967 Roughness Characteristics of Natural Channels. US. Geological Survey Water-Supply Paper 1849. Cllow. Yen Te. 1959 Open-Channel Hydraulics. Tokyo: International Student Edition, Kogakusha. Orr, Michelle and Jonathan A. Owens 1994 "Integrating Riparian Vegetation in Flood Control Channel Design: An Evaluation of Wildcat Creek, California." Unpublished reported submitted to Contra Costa County Flood Control District.

PAGE 164

Notes on Translatory Waves in Natural Channels H.W. Hjalmarson Consulting Hydrologist J.V. Phillips u.s. Geological Survey INTRODUCTION On the afternoon of August 19, 1971, an intense thunderstonn near Wikieup, Arizona, deposited a reported total of 7.62 em of rain in about 45 minutes and produced an extreme flood flow that severely damaged the U.S. Highway 93 bridge near the mouth of Bronco Creek. Computations of the published peak discharge (Aldridge, 1972) assmned stable-flow conditions and estimated discharge at 2,082 m3/s (73,500 ft3/s), which m:.:kes this flood one of the largest known flood peaks for a 49.2 km2 (J 9.0 mi2 ) drainage basin in the United States and the world (Costa, 1987). A recently obtained eyewitness account of large pulsating translatory waves, however, has prompted a new analysis that the peak discharge could have been as much as 2,742 m3/s (96,800 ft Is) or 32% greater than the published peak discharge (Hjalmarson and Phillips, in press). Computations based on free-surface instability criteria indicate that gravitational forces exceeded boundary-retarding forces, and flow was unstable in the steep sand channel. Additional evidence presented in this report suggests that translatory waves may have produced the peak discharge of floods in several other natural channels. DESCRIPTION OF STUDY AREA The Bronco Creek basin is about 12.1 kn1 long and 9.7 km wide willi a general fan shape and a total relief of about 950 111. The sand channel in NOle.: The authors appreciate the support of the Flood Control District of Mancopa County, Arizona, and the account of the flood furnished by Mr. Ernest Fancher of Wikieup, ArizolUl.

PAGE 165

150 Notes on Translatory Waves in Natural Channels the 4.0-km reach above the U.S. Highway 93 bridge is wide, flat, and rectangular in shape. Two tributaries from the southwest-Bronco and Greenwood washes-drain about 20 and 30% of the basin, respectively. The beds of Bronco Creek and Bronco and Greenwood washes change from boulders and bedrock to course-grained sand about 4.8,4.0, and 3.0 km, respectively, upstream from the slope-area measurement site. A corresponding change occurs in the channel-aspect ratio (the width-depth ratio) from about 7-12 in the bedrock channels to about 15-50 in the sand channels. The slope of the boulder channels is about 5% and changes to a rather uniform 3 % where the stream beds become sand channels. The channel gradient also is about 3% near the mouth of Bronco Creek. PREVIOUS INVESTIGATIONS After the August 19, 1971 flood on Bronco Creek, a four-section slope area measurement was made by the U.S. Geological Survey (USGS) in a unifonn 365-meter-long reach that ended about 305 m upstream from the bridge. Roughness coefficients selected in the field for the main channel of the slope-area reach were about 0.030, which corresponds to a computed peak discharge of 2,739 m3/s and velocities of as much as 11.09 m/s (USGS, unpublished data, 1971). During a routine office review, however, the roughness coefficients were changed to about 0.040, and, assuming stable-flow conditions, the published four-section slope area solution for peak discharge yielded 2,082 m3/s (Aldridge, 1972). Because the magnitude of the peak discharge was so rare for a 49.2-square-kilometer basin, other investigations to estimate peak discharge for the Bronco Creek flood were conducted by H.W. Hjalmarson (hydrologist, USGS, personal conmlunication, 1971), Carmody (1980), and House and Pearthree (1995). These investigations included hydrologic analyses of reported rainfall rates and hydraulic measurements of flow in Bronco Creek and the two major tributaries. A wide range of discharges has been determined from the investigations for the estimated peak flow (Table I). Table l. Summary of estimated peak discharges for flood of August 19, 1971, in Bronco Creek. Source Method Discharge, in m'/s (fr:ls) USGS (unpublished dara, 1971) Slope-area (11=0.030) 2,739 (96,700) Aldridge (1972) Slopearea (n=O.040) 2,082 (73,500) H. W. Hjalmarson (USGS, written commun., 1971) Conveyance-slope 1,076 (38,000) Cannady (1980) Hydrologic 793 (28,000) House and Pearthree (1995) Paleoflood 800 (28,200)

PAGE 166

Hjalmarson and Phillips EYEWITNESS ACCOUNT The authors recently obtained a detailed recorded account of the flood from Mr. Ernest Fancher, an employee of the Arizona Department of Tran.'portation (ADO,!) facility at Wikieup, Arizona. His account is slillilllarized as follows. 1/ About every 4 to 5 minutes a wave extending bank to bank would move rapidly downstream. The largest waves were 151 4-5 ft. (1.22-1.52 m) high and would pound over the bridge. Waves 400 to 500 yards (366-457 m) upstream would take about 30-45 seconds to reach the bridge. The water passed under the bridge at a great velocity lmtil a wave would hit. The waves occurred for about 2 hours and wave heights decreased in size later in the flood." These observations, which were also recorded at the time of the flood (E.I. Jencsok, senior hydraulics engineer, ADOT, personal commlmication, 1971), obviously are estimates but are considered accurate mostly because many similar waves were observed by several people during the flood. ANALYSIS OF THE REPORTED WAVES Recent analyses conducted by Hjalmarson and Phillips (in press) used free-surface instability and celerity equations to estimate peak discharge of the reported waves. Application of free-surface instability criteria developed by Koloseus and Davidian (1966) for rectangular channels showed that, at n=0.030, roll waves were likely for a wide range of flow rates. Velocity was computed using the celerity equation by Brater and King (1954) for a large translatory wave. Peak discharge was estimated mostly on the basis of studies conducted by Thompson (1968). Although lillcertainties are associated with the method and computations, results indicate the instantaneous peak discharge produced by the largest waves could have been as much as 2,742 m3/s (Hjalmarson and Phillips, in press). FurthemlOre, passage of the largest waves would have decreased the charmel capacity by more than 100%. The U.S. Highway 93 bridge over Bronco Creek has a design capacity of 481 m3/s. Results of the study, however, suggest translatory wave formation in the reach just upstream of the bridge is possible for rates as low as 142 m3/s. The flood caused serious danlage to the bridge. The results of these computations, including the duration of the wave occurrences, the wave velocity, and the wave height, are in close agreement with Mr. Fancher's observations. VARIOUS DISCHARGE ESTIMATES AND POSSIBLE MECHANISMS FOR WAVE DEVELOPMENT Is it possible that the widely diverse estimates of peak discharge by the several investigators are in some sense correct? These estimates may have a CODm1on link. According to Koloseus and Davidian (1966), flow is

PAGE 167

152 Notes on Translatory Waves in Natural Channels classified as unstable if the Froude number (Fn) is greater than the stable Froude number (Fs). In general, for wide, flat, and steep channels, unstable flow conditions exist for Fn greater than about 1.6 (Koloseus and Davidian, 1966). Additionally, the value of Fs is related to the aspect ratio of the channel. The value of Fs increases as the channel-aspect ratio decreases. For hydrodynanlically rough channels, small decreases in the channel-aspect ratio produce relatively large increases in Fs. At Bronco Creek and Bronco and Greenwood washes, a large increase in the channel-aspect ratio occurs at the transition from the boulder and bedrock channel to the sand bed. The flow, therefore, tends to become unstable at the transition because, with the decrease of Fs, the value of Fn needed to produce a high degree of instability (Fn/ Fs) is lower. This finding suggests that the waves probably were formed in the smoother, wider reaches downstream from the boulder and bedrock channels. Some mechanisms in the literature can explain the observed large, swift waves and large period between waves in the reach just above the Highway 93 bridge on Bronco Creek. Koloseus and Davidian (1966) state that when flow is classified as tmstable, free-surface perturbations with the characteristics of shallow-water waves become larger as they move down stream and give rise to translatory waves. Kranenburg (1992) demon strated how a series of waves may be formed by free-surface instability, how larger waves overtake smaller ones, and how waves become longer and higher with channel length. Holmes (1936) and Keulegan (1949) state that as waves move downstreanl, some will overtake and absorb others and increase amplitude and velocity. The number, height, and velocity of waves varies with the frequency of the initial disturbances and the channel length, according to Keulegan (1949). Bronco Creek's steep sand channel extends about 4 km upstream from the slope-area measurement site and could accOlmt for the reported large, swift waves at 4-to-5-minute intervals in the slope-area reach and at the bridge. The estimates by Carmody (1980) and House and Pearthree (1995) were of the base discharge. The reaches used by Hjalmarson (Table 1) in the upper sections of the sand channels possibly included waves as they were increasing in size. The Hjalmarson and Phillips analysis (in press) used the original slope-area survey data in the reach above the bridge where the waves were fully developed. Wave development below the bedrock and boulder channels and growth downstream could explain much of the mystery sUITOlmding the various discharge estimates for the flood of August 19, 1971, in Bronco Creek. OTHER POTENTIAL SITES FOR TRANSLATORY WAVES Another exanlple of a questioned peak discharge estimate is the catastrophic flood of September 14, 1974, in Eldorado Canyon, Nevada,

PAGE 168

Hjalmarson and Phillips 153 which killed at least nine people and destroyed many homes, vehicles, and boats (Glancy and Hannsen, 1975). The flood had a computed peak discharge of 2,152 m3/s for a drainage area of 53.3 km2 and similar to the Bronco Creek flood, the highwater marks in the slope-area reach (and the peak discharge) may have been produced by translatory waves (Glancy and Hannsen, 1975). Following are some statements made by observers of flow in the steep sand channel of Eldorado Canyon: "a 1.83 to 2.44 m high approaching wall ... and "initial wave followed by several wavelike surges ... Although Glancy and Harmsen (1975) acknowledge that flow could have been highly unsteady in the slope-area reach, slope-area techniques and assumptions of steady, uniform, and stable flow were used for the peak discharge estimate. McGee (1897) documents his eyewitness account of a sheet flood wave on the western piedmont slopes of the Tortolita Mountains north of Tucson, Arizona. According to McGee, the flood water spread beyond the confines of a channel at "race-horse speeds" with a wall of water 15 to 30 em high, and within the flood, transverse waves formed breakers. Large traIlslatory waves leaving the confines of steep canyons could possibly spread over the lower piedmont surfaces as described by McGee. The steep incised channels of many alluvial fans with slopes of about 3% or greater in the southwestern United States are possible sites for the fonnation of potentially hazardous translatory waves. Another example is the catastrophic flood of September 10, 1976, in Meyers Canyon that nearly destroyed the retirement community of Ocotillo, California. According to one report, the town was nearly cut in half by the flood that sent a wall of water a half mile wide and nearly 6 feet high rolling over and through the town's 100 homes (Los Angeles Times, 1976). Eyewitness accOlmts described the flow as looking like oceall waves and stated they saw "wave after wave-like breakers 4 feet high" and reported a peak velocity of approximately 30 miles per hour (l H l11/s). The flood left behind two fatalities, 20 destroyed homes, and another 70 homes that were badly danlaged (Los Angeles Times, 1976). SUMMARY AND CONCLUSIONS Evidence presented suggests that translatory waves and pulsating flow may be more common than traditionally thought and may have been overlooked in determining peak flow rates at some sites. Instability criteria should be considered for hydraulic analysis of flood flow in high gradient alluvial and other smooth channels. Application of translatory wave techniques needs verification by additional experiments, observations, and research.

PAGE 169

154 Notes on Translatory Waves in Natural Channels REFERENCES Aldridge, B.N. 1972 "Investigation of Floods from Small Drainage Basins in Arizona." pp. 107-126 in University of Arizona, Transportation and Traffic Institute, Proceedings, 21st Conference on Roads and Streets. Brater, E.F. and King, H.W. 1954 Handbook of Hydraulics. New York: McGraw-Hill Book Co., Inc. Cannody, T. 1980 A Study to Advance the Methodology of Assessing the Vulnerability of Bridges to Floods for the Arizona Department of Transportation. General Report No.1. University of Arizona Department of Civil Engineering. Costa, J.E. 1987 "A Comparison of the Largest Rainfall-Runoff Floods in the United States with those of the People's Republic of China and the World." Journal of Hydrology 96:101-115. Glancy, P.A., and Harmsen, L. 1975 A Hydrologic Assessment of the September 14, 1974, Flood in Eldorado Canyon, Nevada. USGS Professional Paper 930. Hjalmarson, H.W., and Phillips, J.V. in press "Potential Effects of Translatory Waves on Estimation of Peak Flows: A Case Study." Journal of Hydr. Engineering. Holmes, W.R. 1936 "Traveling Waves in Steep Channels." Civil Engineering 6 (7):467-468. House, P.K., and Pearthree, P.A. 1995 "A Geomorphologic and Hydrologic Evaluation of an Extraordinary Flood Discharge Estimate: Bronco Creek, Arizona." Water Resources Research 31 (12):3059-3073. Keulegan, G.H. 1949 "Wave Motion." In Rouse, H., ed., Engineering Hydraulics. New York: John Wiley & Sons, Inc.

PAGE 170

Hjalmarson and Phillips Koloseus, H.J., and Davidian, 1. 1966 Free-Surface Instability Correlations. Water-Supply Paper 1992-C, U.S. Geological Survey. Kranenburg, C.K. 155 1992 "On the Evolution of Roll Waves." Fluid Mechanics 245: 249-261. Los Angeles Times 1976 "Ocotillo Digs Out-Flood Leaves Town Battered but Undefeated." September 12. McGee, W.J. 1897 "Sheetflood Erosion." Geological Society America Bulletin 8: 87-112. Thompson, T.H. 1968 Determination of Discharge During Pulsating Flow. Water-Supply Paper 1869-D, U.S. Geological Survey.

PAGE 171

Recent Flood Damages and Bank/Scour Protection Measures at Bridge Crossings in Southeast Arizona Zbigniew Osmolski Fazle Karim Pima County Department of Transportation and Flood Control District INTRODUCTION Flood damage to public infrastructure and private property in Pima County, southeast Arizona, has occurred more frequently in recent years. All watercourses in Pima County, including the Santa Cruz and the Rillito, the major rivers draining the area, are primarily ephemeral desert streams. The Rillito River joins the Santa Cruz near the City of Tucson, and flows northeast 90 miles to join the Gila River near Phoenix (Figure 1). Annual flood series for the Santa Cruz River at Tucson indicate an apparent increase in magnitudes of floods during the past three decades (Webb and Betancourt, 1992). This increase, as illustrated in Figure 2, is accompanied by a change in storm types causing the floods, i.e., more annual floods in fall and winter and fewer in summer. The October 1983 flood is the largest on record, and the second-largest was in January 1993. This paper compares flood danlage to public infrastructure in the 1983 and 1993 floods, and describes bank/scour protection measures constr1lcted at the major bridge crossings on the Santa Cruz River after the 1993 flood. A computation of pier scour depths at a bridge site is presented to examine reliability of present methods for predicting pier scour depths. FLOOD DAMAGE IN 1983 AND 1993 The October 1983 flood is the largest on record for the major rivers in Pima COlmty and the costliest in danlage to public infrastructure and private property. Persistent rainfall from September 27 through October 3 by tropical storm Octave off the coast of Baja California, caused record floods on all watercourses (Roeske et al., 1989). Massive damage occurred to public and private facilities due to severe bank erosion and

PAGE 172

Osmolski and Karim il '-----. 'o ,\ 1 >Uoo",",,_ Pima County Department of Transportation and Flood Control District RIVER SYSTEM AND BANK STABILIZATION PROJECTS IN PIMA COUNTY CORONADO NATIONAL FOREST BEFORE1N3 1114-1'3 (JAN.) 157 111111111111 &FTt.A ,"3 (JAN) UNDER CONSTRUCTION Figure 1. River system and bank stabilization projects in Pima County.

PAGE 173

158 Protection Measures at Bridge Crossings in Southeast Arizona "" ( ,..:. .. .ra > :J .... 00 '-' iii 01 01 I iii 01 .;, 1"1= ";:: .. VJ "8 -.!! .. !;; :i: (;i ;:I = @ = 0 c: c: c: OJ OJ ;:I f-< OJ Q) Q) Q) :::!; :::!; :::!; Cl Cl Cl c: c: "0 c: > "> 0 > 0 0 0 0 :::!; :::!; LL :::!; :. :. OJ OJ Q) Q) :J Q) >->c: c: >0 .;, .;, N .. :J > C2 N ;:I .. U I I I I[ j @ [J) N I:j) iE 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o. 0 0 0 0 0 0 .n 0 .n 0 on 0 .n 0 .n 0 .n '" on .. .. M M N N (Sp) a6JelpS!O lfead

PAGE 174

Osmolski and Karim overbank flooding on all major rivers. Infrastructure damage included washed-out or damaged roads and highways, bridges, and utilities. 159 Damage to private facilities included destruction or severe damage to hundreds of residential and business units (pima County, 1984). The total cost of damage during the 1983 flood was estimated at $105.7 million, including emergency and permanent repairs under the Flood Repair and Flood Hazard Mitigation Program developed after the flood (pima County, 1993). The 1993 flood was caused by a prolonged rainfall from January 5 through January 19 with two distinct peaks on January 8 and January 19. Peak discharges during the 1983 and 1993 floods are shown in Table 1. Table 1. Peak discharges, 1983 and 1993 floods. River 1983 Peak Discharge* (cfs) Santa Cmz, Continental Santa Cmz, Tucson Santa Cmz, Cortaro Rillito, Tucson Tanque Verde, Tucson 45,000 52,700 65,000 29,700 8,600 1993 Peak Discharge* (cfs) 32,400 37,400 24,100 24,500 USGS Water Resources Data, Arizona, Water Years 1984, 1993 Peak discharges during the 1993 flood are generally lower than those in the 1983 flood. However, prolonged rainfall over 15 days in 1993 caused longer floods and greater runoff volumes throughout the greater Tucson area. It is reported that the runoff volume for the 1993 flood on the Rillito is the largest on record (pima County, 1993). As both peak discharge and flow duration (longer during 1993 flood) are the main contributing factors to the extent of danlage, it is expected that damage from the 1993 flood will be comparable (perhaps a little lower because of lower peaks) to that in 1983. The estimated cost of emergency and permanent repairs for the 1993 flood is $13.9 million (compared to $105.7 million in 1983). The maill reason for this large reduction is the construction of extensive soil cement bank stabilization after the 1983 flood (Figure 1), demonstrating the effectiveness of bank stabilization projects in Pima County. OtlIer contributing factors for reduced damage in 1993 are (1) improved design standards (e.g., bridges to convey 100-year flow, bank stabilization to withstand l00-year flow) established after the 1983 flood; (2) Floodprone Land Acquisition Program under which residential units in vulnerable locations were acquired, structures demolished, and residents relocated; and (3) improved floodplain management due to Floodplain and Erosion Hazard Management Ordinance No. 1988-FC2 (pima County,

PAGE 175

160 Protection Measures at Bridge Crossings in Southeast Arizona 1988), which restricts construction in floodprone areas. All flood control structures, primarily soil cement bank stabilization, built along major watercourses after the 1983 flood, had relatively little damage in 1993. EROSION PROTECTION MEASURES AT BRIDGE CROSSINGS A significant part of the infrastructure damage during the 1993 flood was caused by bank erosion near bridge crossings and excessive scour at bridge piers and abutments. Five bridge crossings on the Santa Cruz River at Ina Road, Elephant Head Road, Trico Road, Sahuarita Road, and Trico Marana Road were damaged due to bank erosion immediately upstream and/or to scour at bridge piers and abutments. After the 1993 flood, Pima County Flood Control District (PCFCD) and the Federal Highway Administration (FHW A), developed bank stabilization and scour countermeasure plans to protect these five bridge crossings from future floods, and construction has recently been completed. Protection measures constructed include a combination of soil cement bank stabilization at one or both banks upstream of bridges and abutment protection with spur dike. Soil cement stabilization was used to protect upstream banks since it has been proven to be very effective in southeast Arizona. Spur dikes were designed as quarter ellipse with soil cement lining to guide the upstream flow and prevent scour at abutments. In addition, two innovative approaches were used, as described below. River Bed Pavement under Bridge Structure Soil cement pavement was installed tmder the Ina Road bridge to prevc.nt scour at the piers and abutments. A typical design of such pavement is shown in Figure 3. It also serves as a grade control structure, thus providing an efficient and cost-effective measure for both river bed stabilization and protection to bridge substructures. Flexible Spurs with Synthetic Nets The south bank upstream of the Trico-Marana Road bridge experienced significant erosion during the 1993 flood due to 200-300 feet of lateral migration, threatening the roadway approach and south abutment of the bridge. To protect the south bank and the bridge from further erosion, an experimental approach using flexible spurs consisting of series of pi iesupported permeable panels of synthetic nets was utilized. The permeable panels will reduce flow velocity near the protected bank and redirect Ole flow path away from the eroded bank and toward the center of the charmel, resulting in progressive sediment accretion near the eroded banks. Because of its unique features, perfomlance of this project will be evaluated for application to other areas in Arizona.

PAGE 176

Osmolski and Karim 161 EXIST. BRIOf.E PIER EXIST. BRIOf.E -c""\ -/ ;;r SECTION T BRIOGE TO CONTROL STRUCTURE O SPRR BRIDGE GRADE CONTROL STRUCTURE ,------OIMENSION TE PROPOSED TOE ELEVATION Figure 3. Bridge grade control structure. Pier Scour Prediction: An Example The floods of 1983 and 1993 destroyed, damaged, or threatened the safety of mmy bridges in the Pima County area, and served as a reminder of the need for reliable prediction of scour depths at bridge piers and abutments. The main shortcoming of the available equations for predicting scour depths is that they are based entirely on laboratory data and their perfOimance under field conditions is not known. An example computation illustrates the effects of angle of attack of flow and debris accumulation on computed scour depths using the equation recommended in HEC-18 (FHW A, 1993). The computed pier scour depths (Q = 32,000 cfs) at the Southern Pacific Railroad (SPRR) bridge on the Rillito bridge in Tucson, Arizona are srnrunarized in Table 2. As the river turns by almost 90 degrees upstream of the bridge, an angle (0) of attack of flow near 30 degrees is considered a reasonable estimate. Similarly, increase in effective pier width due to debris acculllulation, by 50 to 100%, is within the expected range. However, in view of the field observations after the floods of 1983 and 1993, these considerations apparently result in significant overestimate of the expected pier scour depth (compared to field observations) at the SPRR bridge,

PAGE 177

162 Protection Measures at Bridge Crossings in Southeast Arizona Table 2. SlUlllllary of pier scour depths for SPRR bridge. Pier Width Adjustment Pier Scour Depth (feet) (a' effective width) Nonh Piers South Piers B 30 e 15 e 30 No adjustment (a' a) 14.8 19.7 25.2 33.6 50% increase for debris (a' 1.5a) 19.2 25.6 32.8 43.7 100% increase for debris (a' 2a) 23.2 30.9 39.5 52.7 a pier width 5.5 feet (nonh piers) ; 12.5 feet (south piers) as can be seen from the results in Table 2. These results suggest the need to reexamine the correction factor K2 for the angle of attack of flow and in the equation given in HEC-18 (FHWA, 1993) and to develop working guidelines for debris accumulation factors in detennining effective pier width. REFERENCES Federal Highway Administration 1993 "Hydraulic Engineering Circular No. 18." Washington, D.C. Pima County Department of Transportation and Flood Control District 1984 "Final Documentation October 1983 Flood Damage Report." Pima County Department of Transportation and Flood Control District 1993 "January 1993 Floods, Pima County, Arizona Summary Report." Pima County Department of Transportation and Flood Control District 1988 "Floodplain and Erosion Hazard Management Ordinance No. 1988-FC2 For Pima County, Arizona." Roeske, R.H., Garret, lM., and Eychaner, lH. 1989 Floods of October 1983 in Southeastern Arizona. U.S. Geological Survey Water Resources Investigations Report 85-4225-C. Webb, R.H., and Betancourt, J.L. 1992 Climatic Variability and Flood Frequency of the Santa Cruz River, Pima County, Arizona. U.S. Geological Survey Water Supply Paper 2379. Washington, D.C.

PAGE 178

Part 7 MODELLING AND COMPUTER PROGRAMS

PAGE 179

This page IS intentionally blank

PAGE 180

A Discussion of the U.S. Army Corps of Engineers HEC-RAS Computer Program Wilbert O. Thomas, Jr. Chris D. Krebs Michael Baker Jr., Inc. Gary W. Brunner u.s. Army Corps of Engineers INTRODUCTION For over 25 years, the U.S. Amly Corps of Engineers (Corps) HEC-2 stepbackwater program (U.S. Army Corps of Engineers, 1991) has been widely used to compute water-surface profiles for floodplain management in the United States and arOlmd the world. In 1995, the Corps released the Hydrologic Engineering Center's River Analysis System (HEC-RAS), the successor to HEC-2 (U.S. Army Corps of Engineers, 1995a and 1995b). HECRAS is completely new software that operates in a Windows environment and has significant features and refinements over HEC-2. HEC-RAS consists of a graphical user interface, hydraulic analysis programs, data storage and management capabilities, and graphics and reporting facilities. When HEC-RAS was being developed, a significant effort was spent on improving the computational capabilities over those in the HEC-2 program. Thus there are computational differences between the two programs. This paper describes some of these differences. The main differences are in overbank-conveyance and critical-depth calculations, bridge and culvert hydraulic computations, and floodwayencroachment computations. This paper only addresses the differences in the two programs relative to conveyance and critical-depth computations. New computational features in HEC-RAS (Version 1) and future additions to the next release of HEC-RAS (Version 2) are also discussed. OVERBANK CONVEYANCE CALCULATIONS Both HEC-RAS and HEC-2 utilize the Standard Step method for balancing the energy equation to compute a water surface elevation at a

PAGE 181

166 A Discussion of the HEC-RAS Program given cross section location. A key element in the solution of the energy equation is the calculation of conveyance. The conveyance is used to determine the friction losses between cross sections, the flow distribution at a cross section, and the velocity weighting coefficient, a. The approach used in HEC-2 is to calculate conveyance between every coordinate point in the overbanks of each cross section (Figure 1). The HEC-2 program sums up all the incremental conveyances (Ki ) in each overbank to obtain the total conveyance for the left (lob) and right overbank (rob). This method of computing overbank conveyance can lead to varying anlOunts of total conveyance when additional coordinate points are added to the cross section, without actually changing the geometry. The HEC-2 method for computing overbank conveyance has been retained as an option within HEC-RAS in order to reproduce studies that were originally developed with HEC-2. However, the default method used in HEC-RAS is to subdivide the overbank areas at n-value break points (locations where n values change) for overbank conveyance calculations (Figure 2). In Figures 1 and 2, Pi are the incremental wetted perimeters, Ai are the incremental cross-sectional areas and ni are the incremental n values. The two methods for computing conveyance will produce different answers whenever portions of the overbanks have ground sections WiUl significant vertical slopes. In general, the HEC-RAS default approach will provide a lower total conveyance for the sanle elevation and, therefore, a higher computed water-surface elevation. To evaluate the difference between the two ways of computing conveyance, comparisons were performed using 97 data sets from the Corps profile accuracy study (U.S. Army Corps of Engineers, 1986). Water-surface profiles were computed for the 1 %-annual-chance flood using the two methods for computing K =K+K+K+K lob I 2 3 4 A P ch ch K =K+K+K+K rob 5 6 7 8 Figure 1. HEC-2 conveyance subdivision method.

PAGE 182

Thomas, Krebs, and Brunner 167 ill n 2 nCh n3 A2 P2 A P A P ch ch 3 3 K =IS+IS lob K =K rob 3 Kch Figure 2. HEC-RAS default conveyance subdivision method. conveyance in HEC-RAS. The results confirmed that the HEC-RAS default approach will generally produce a higher computed water-surface elevation. Of the 2,048 open-channel cross sections, 47.5% had computed water-surface elevations within 0.10 foot, 71 % within 0.20 foot, 94.4% within 0.40 foot, and 99.4% within 1.0 foot. Because the differences tend to be in the same direction (higher elevations with HEC-RAS), some effects can be attributed to propagation. 'flle HEC-2 style method subdivides the overbank sections in greater detail than the HEC-RAS default method when computing total conveyance. The observation that the HEC-2 style method yields a larger total conveyance is consistent with Davidian (1984), who cautioned against subdividing cross sections that had basic geometric shapes such as rectangles, trapezoids, semicircles, or triangles. Davidian (1984) notes that rouglmess coefficients (Manning's n) are based on unit cross sections that have complete or nearly complete wetted perimeters. If cross sections with these basic geometric shapes are subdivided, the total conveyance may be increased to the extent that the composite n value for the entire cross section could be less than either of the incremental n values used in the subdivision. Comparisons of HEC-RAS results with those from HEC-2 were perfonned using the sanle 97 data sets from the Corps profile accuracy study (U.S. Army Corps of Engineers, 1986). Water-surface profiles were computed for the 10-and 1 %-annual-chance floods using HEC-2 and HEC-RAS, with both programs using the HEC-2 approach for computing overbank conveyance. Table 1 shows the percentage of 2,048 cross sections within plus or minus 0.02 foot. For the 10and 1 %-annual-

PAGE 183

168 A Discussion of the HEC-RAS Program chance floods, 63 and 88 cross sections, respectively, had elevation differences greater than plus or minus 0.02 foot. For those cross sections with differences greater than plus or minus 0.02 foot, approximately 96% of the differences can be attributed to the critical-depth computation and the propagation of these differences upstream. Table 1. Percentage of 2,048 cross sections with the indicated difference in computed water-surface elevation (HEC-RAS-HEC-2). Difference -0.02 -0.01 0.0 0.01 0.02 Total (feet) 10%-AnnualChance Flood 0.8 11.2 73.1 11.2 0.6 96.9 1 %-AnnualChance Flood 2 11.6 70.1 10.8 1.3 95.8 The results of these comparisons do not show which method is more accurate; they only show the differences between the methods. In general, it is felt that the HEC-RAS default method is more commensurate with Manning's equation and the concept of separate flow elemenl,> and is based on the geometry rather than how many points are used in the cross section. Furthermore, this method is more consistent with the theories and methods in other hydraulic progranls such as HEC-6 (U.S. Army Corps of Engineers, 1993), UNET (Barkau, 1992), and WSPRO (Sheannan, 1990). Further research with observed water-surface profiles is needed to make any final conclusions about the accuracy of the two methods. CRITICAL-DEPTH COMPUTATIONS The HEC-RAS program has two methods for calculating critical depth: a parabolic method and a secant method. The HEC-2 program has one method, which is similar to the HEC-RAS parabolic method. The parabolic method is computationally faster, but only locates a single minimwn energy at each cross section. For most cross sections there will be only one minimwn on the total-energy curve; therefore, the parabolic method has been set as the default method for HEC-RAS. If the parabolic method is tried and does not converge, the HEC-RAS program will automatically try the secant method. The HEC-RAS version of the parabolic method calculates critical depth to a nwnerical accuracy of 0.01

PAGE 184

Thomas, Krebs, and Brunner 169 foot, while the HEC-2 version of the parabolic method calculates critical depth to a numerical accuracy of 2.5% of the flow depth. Furthermore, HEC-RAS uses a low starting elevation within the main channel for the search routines while the starting elevation in HEC-2 is the projected water-surface elevation from the downstream cross section. This alone can lead to minor differences in the calculation of critical depths between the two programs. In certain situations it is possible to have more than one minimmn 00" the total-energy curve. When the parabolic method is used on a cross section with multiple minimums, the method will converge on the first minimum it locates. This approach can lead to incorrect estimates of critical depth, in that the computed value for critical depth may be on the top of a levee or an ineffective flow elevation. When this occurs in the HEC-RAS progranl, the software automatically switches to the secant method. The HEC-RAS secant method is capable of finding up to three minimlUllS on the total-energy curve. Whenever more than one minimum energy is fOlUld, the program selects the lowest valid minimum energy (a top of a levee or ineffective flow elevation is not considered a valid critical-depth solution). Given that HEC-RAS has the capability to find multiple critical depths and detect possible invalid answers, the final critical-depth solutions between HEC-2 and HEC-RAS could be quite different. In general, the critical-depth solution from the HEC-RAS program is more accurate than that from HEC-2. NEW COMPUTATIONAL FEATURES IN HEC-RAS VERSION 1 The following is a list of new computational features found in HEC-RAS VersIOn 1 that are not available in HEC-2 (excluding the features for bridge hydraulics): (1) HEC-RAS can perform subcritical-, supercritical-, and mixed flow-regime calculations in a single execution of the program. The cross-section order does not have to be reversed (as in HEC-2); the user simply presses a single button to select the computational-flow regime. When in a mixed-flow-regime mode, HEC-RAS can also locate hydraulic jwnps. (2) HEC-RAS can perform hydraulic computations for additional culvert shapes beyond those used in HEC-2,and has the ability to mix culvert shapes at the same road crossing with the culverts having multiple slopes and invert elevations. (3) HEC-RAS can model single reaches, dendritic stream systems, or fully looped network systems. HEC-2 can model only single

PAGE 185

170 A Discussion of the HEC-RAS Program reaches and a limited number of tributaries (up to three stream orders). (4) At stream junctions, HEC-RAS has the ability to perform the calculations with either an energyor momentum-based method. HEC-2 uses only the energy-based method. (5) HEC-RAS has the following new cross-section properties not found in HEC-2: blocked ineffective flow areas; normal ineffective flow areas can be located at any station (in HEC-2 they are limited to the main channel bank stations); blocked obstructions; and specification of levees. (6) In HEC-RAS the user can enter up to 500 points in a cross section. HEC-2 has a limit of 100 points. (7) HEC-RAS has the ability to perform geometric cross-section interpolation. HEC-2 interpolation is based on a ratio of the current cross section and a linear elevation adjustment. (8) HEC-RAS has an improved flow-distribution calculation routine. The new routine can subdivide the main channel as well as the overbanks, and the user has control over how many subdivisions are used. The HEC-2 flow-distribution option is limited to the overbank areas and breaks at existing coordinate points. NEW FEATURES PLANNED FOR HEC-RAS VERSION 2 Version 2 of HEC-RAS is being developed and will be released by October 1996. The featmes and enhancements planned for the new reiease are as follows: (1) WSPRO bridge routines (Shearman, 1990) as an additional option for low flow through bridges; (2) Bridge scom computations using Federal Highway Administration (FHWA) Hydraulic Engineering Circular (HEC) 18 procedures (Richardson et aI., 1993); (3) Inline weirs and gated spillways option; (4) Channel-modification featmes (similar to tlle HEC-2 channel improvement option); (5) Additional culvert shapes Low-profile arch High-profile arch;

PAGE 186

Thomas, Krebs, and Brunner 171 (6) New culvert features Adverse sloping culverts Supercritical and mixed-flow regimes in culverts; (7) Links to geographic information systems (ARC-INFO) Ability to determine cross sections in ARC-INFO and input to HEC-RAS Ability to export water-surface profiles to ARC-INFO for plotting onto terrain; (8) Improved model schematic features Subdivide and combine existing reaches graphicaIly Utilize UTM or latitude-longitude coordinates for plotting the stream system so it will be geographicaIly correct and look like the actual stream; (9) Ability to import HEC-2 data into separate reaches of a multi reach model; (l0) Ability to export graphics to a DXF file format; The new features and enhancements include improved and additional hydraulic computations, improved data management capabilities, and improved links with geographic information systems. The incorporation of these features should provide a more useful tool for the computation of water-surface profiles and floodplain management. FINAL COMMENTS TI1e Windows environment and the graphical user interface make HECRAS a very user-friendly program. The graphics capability is a valuable tool for evaluating the quality of input and output data. The report generator added to Version 1.1 enables the user to generate a text file with a list of all the input and output data. Users have complete control over what data are summarized in the report. Input data are separated into plan infonnation, geometric data, and flow data. Users can obtain detailed output from one of the standard swumary tables, or any user-defined summary table. In the future, both the sediment transport model HEC-6 (U.S. Army Corps of Engineers, 1993) and the one-dimensional unsteady flow model UNET (Barkau, 1992) will be added to HEC-RAS to expand its applicability and utility. Given this increased capability, it is likely that HEC-RAS will become an indispensable tool for hydraulic analysis.

PAGE 187

172 A Discussion of the HEC-RAS Program REFERENCES Barkau, Robert L. 1992 ''UNET, One-dimensional unsteady flow through a full network of open channels," (computer program). St. Louis, MO. Davidian, Jacob 1984 "Computation of water-surface profiles in open channels." Chapter A15 in U.S. Geological Survey Techniques of WaterResources Investigations, Book 3. Reston, VA: U.S. Geological Survey. Richardson, E. V., L. J. Harrison, J. R. Richardson and S. R. Davis 1993 Evaluating scour at bridges. FHWA-IP-90-017, HEC-18. Washington, D.C.: Federal Highway Administration Sheannan, 1. O. 1990 User's manual for WSPRO A computer model for water surjace profile computations. FHW A-IP-89-027, Washington, D.C.: Federal Highway Administration. U.S. Army Corps of Engineers 1986 Accuracy of computed water surface profiles. Research Document 26. Davis, CA: Hydrologic Engineering Center. U.S. Army Corps of Engineers 1991 HEC-2, Water surface profiles, User's Manual. Davis, CA: Hydrologic Engineering Center. U.S. Anny Corps of Engineers 1993 HEC-6 Scour and deposition in rivers and reservoirs User's Manual. National Technical Information Service PB94141769. Davis, CA: Hydrologic Engineering Center. U.S. Anny Corps of Engineers 1995a HEC-RAS River Analysis System, User's Manual. Version l.0. Davis, CA: Hydrologic Engineering Center. U.S. Army Corps of Engineers 1995b HEC-RAS River Analysis System, Hydraulic Reference Manual. Version 1.0. Davis, CA: Hydrologic Engineering Center.

PAGE 188

Approximate Floodplain Delineation Using WinXSPRO Martin J. Teal WEST Consultants, Inc. INTRODUCTION In many areas where detailed floodplain analyses cannot be justified on economic or other bases, an approximate floodplain study will still provide a means of delineating the 100-year floodplain and determining flood elevations (the areas within the floodplain determined by approximate methods are designated Zone A on flood insurance maps). The Federal Emergency Management Agency (FEMA) guide for study contractors (FEMA, 1995) identifies two hydraulic methods for determining the approximate 100-year flood elevation: (1) Normal-depth calculations using Manning's equation, and (2) Highway culvert nomographs available from the Federal Highway Administration. Method 1, Manning's equation, is often the simplest method to use for channel/floodplain areas where normal depth can be approximated. However, computer programs designed for backwater computations using multiple cross sections (e.g., HEC-2) are often cumbersome to use for analysis of a single cross section. WinXSPRO is a computer program to analyze the geometric and hydraulic properties of a single section. The results from WinXSPRO can then be used to prepare appr:)ximate floodplain delineations. ORIGINAL PROGRAM WinXSPRO grew out of an earlier program, XSPRO, developed in the late 1980s by the U.S. Department of Agriculture, Forest Service (the Forest Service), in association with the U.S. Department of the Interior, Bureau of Land Management (BLM). Specifically, a tool was needed to help hydrologists, fishery biologists, geomorphologists, engineers, and

PAGE 189

174 Floodplain Delineation using WinXSPRO others in computing stream flow, describing instream-flow regimes, monitoring stream channel processes, perfornling hydraulic reconstructions, providing information on riparian habitats, and designing effective channels and riparian structures. XSPRO is an interactive menudriven software package capable of analyzing stream cross sectional data (Grant et aI., 1992) and was released for distribution in 1992. XSPRO is a menu-driven DOS program and was specifically developed for use in the high-gradient streams often encountered by Forest Service and BLM personnel. The program calculates stage discharge curves (rating curves) for a single channel transect. Changes in channel cross sectional parameters with variation of stage (e.g., area, wetted perimeter) are also calculated. The program allows the user to subdivide the channel cross section so that overbank areas, mid-channel islands, and high-water overflow channels can be analyzed separately. The program also allows input of variable water-surface slopes so that slopes can be varied with discharge to reflect natural conditions. Cross section geometry, generally in the form of (X,Y) ordered pairs, is the primary input to XSPRO. The Y coordinate can be elevation, stage, or depth from a datum to the channel bottom. An option is available in the program to correct the input data if it was obtained by using either tile sag tape or rod and level survey collection methods. Once the cross section geometry is input (either manually or read from a file), the user can choose the desired analysis procedure (Geometry Only, Hydraulics Only, or Hydraulics and Regression). With the Geometry Only option, the user inputs the high and low stage limits for which an analysis is desired, and the vertical increment between these limits where computations will be performed. Division of the cross section into subsections (up to five) can also be entered by the user. Only geometric variables are computed using this analysis option (e.g., flow area, water surface width). This option is useful for comparing changes in cross-section geometry through time. The Hydraulics Only option requires the same input as the Geometry Only option with the addition of the energy slopes for the high and lov: stage limits. Also, a flow resistance equation needs to be chosen from the options presented. As previously mentioned, the XSPRO package was designed to be able to examine geometric and hydraulic conditions for single transects in steep streams of greater than 1 % slope. XSPRO supports three alternatives for analyzing boundary roughness and resistance to flow. The user can choose the Manning or Jarrett (1984) flow resistance equations, or use the equations suggested by Thorne and Zevenbergen (1985); Bathurst's equation (1978) for streams with relative roughness values greater than one (i.e., R/dS4> 1, where R is the hydraulic radius and dS4 is the sediment grain size for which 84% are finer by

PAGE 190

Teal 175 weight); and Hey's equation (1979) for streams with relative roughness values less than one. The Jarrett, Bathurst, and Hey equations were specifically developed for application in large roughness channels, such as those often found in steep, mountainous areas. For the Hydraulics Only option, all the values computed in the geometric analysis are included in the output, along with slope, Manning's n, average velocity, and discharge. The data are organized from the low-stage value up to the high-stage value. The Hydraulics and Regression option directs XSPRO to perfonn the hydraulic analysis and a regression analysis on the discharge versus hydraulic radius relation. NEW PROGRAM ENHANCEMENTS Since its original release in August 1992, XSPRO has been widely used for a variety of conditions, resulting in numerous suggestions for improvement of its ease of use, enhancements to its computational capabilities, and development of an improved user's manual. In response to the requests for improvements and enhancements, the Forest Service contracted with WEST Consultants in 1994 to develop a new version of XSPRO. TIle new XSPRO is written for Windows (hence the name WinXSPRO) and includes many of the desirable user interface features found in other popular Windows applications. These include toolbars, context sensitive on-line help screens and menus, instant graphics while in input mode, ability to provide output in a variety of fomlats for use in other progratllS, and error trapping to prevent entry of obviously incorrect data or premature exiting of the program. Figure 1 shows the main input window of WinXSPRO. The ability to use a mouse with the program enhances the ease of use considerably over the previous version. WinXSPRO retains all of the abilities of the previous program (XSPRO) described previously. However, WinXSPRO is more user friendly and offers many new features. One of the new features is the inclusion of an additional flow resistance relation. A theoretical method proposed by Nelson et al. (1991) is incorporated into the program. This method requires as input, in addition to the cross section geometry, a file containing sediment data for the transect. The drag of particles in the cross section is calculated, which provides a measure of the hydraulic roughness for computing a stagedischarge relationship. Another new feature implemented is calculation of best-fit regression equation for stage versus discharge (the stage versus hydraulic radius regression is retained from XSPRO), and production of plots of the data when a plan is executed using the Hydraulics and Regression analysis option. Sediment transport calculations were also added so that the user

PAGE 191

176 Floodplain Delineation using WinXSPRO Figure 1. The main input window of WinXSPRO. may calculate bedload by the Parker (1982, 1990) and Meyer-Peter and Muller (1948) relations, and bed material load by the Ackers and White (1973) relation. Other new features that will be useful to resource professionals include improved plotting routines and file management options. The llser is now able to choose the scaling factors for cross section plots and export the cross section and regression equation plots to several different file formats (DXF, HPGL, and others). Plots can now be easily developed for parameter versus parameter, where the user chooses which parameters should appear on the x-and y-axes. The user also has a system within the program to organize the analyses by project and by trial runs under each project. FLOODPLAIN DELINEATION The output from a hydraulic or hydraulic and regression analysis with WinXSPRO will list the discharge estimate using one of the aforementioned resistance equations for each stage requested by the analyst (note that use of resistance methods other than user-supplied Marming's n may require the approval of the FEMA Project Officer). By adding the stage corresponding to the 100-year flow to the minimum

PAGE 192

Teal 177 elevation in the channel, the 100-year flood elevation can be obtained. 'Ibis elevation can be marked on maps in the vicinity of the cross section to delineate the 100-year floodplain. Caution should be used in picking "representative" cross sections for the reach(es) under study. Also, the delineation of the floodplain upand downstream of the cross section must take into account the water surface slope (change in water surface elevations with distance). A cornnlOn assumption is that the water surface slope is about equal to the channel bed slope (approximate uniform flow conditions). SUMMARY WinXSPRO is a powerful yet simple-to-use program that can aid in delineation of floodplains for approximate flood insurance studies. The versatility of the program will also make it a valuable tool to resource specialists who have a need for stream channel cross section analysis. REFERENCES Ackers, P., and White, W.R. 1973 "Sediment Transport: New Approach and Analysis." Journal of Hydraulic Engineering 99 (HYll). Bathurst, 1. C. 1978 "Flow Resistance of Large Scale Roughness." Journal of Hydraulic Engineering 104 (HY12). Federal Emergency Management Agency, Flood Insurance Administration 1995 Flood Insurance Study Guidelines and Specifications for Study Contractors. FEMA 37. Washington, D.C.: FEMA. Grant, G.E., Duval, J.E., Koerper, G.J., and Fogg, J.L. 1992 XSPRO: A Channel Cross-Section Analyzer. BLM/FS Technical Note 387. Denver, CO. Hey, R.D. 1979 "Flow Resistance in Gravel-bed Rivers." Journal of Hydraulic Engineering 105 (HY4). Jarrett, R.D. 1984 "Hydraulics of High-Gradient Streams." Journal of Hydraulic Engineering 105 (HY4).

PAGE 193

178 Floodplain Delineation using WinXSPRO Meyer-Peter, E., and Muller, R. 1948 "Fonnulas for Bed-Load Transport." Report on Second Meeting of International Association for Hydraulic Research, Stockholm, Sweden. Nelson, 1.M., Emmett, W.W., and Smith, J.D. 1991 "Flow and Sediment Transport in Rough Channels." Proceedings of the Fifth Federal Interagency Sedimentation Conference, Las Vegas, NV. Parker, G., Klingeman, P.e., and McLean, D.G. 1982 ''Bedload and Size Distribution in Paved Gravel Bed Streams," Journal of Hydraulic Engineering 108 (HY4). Parker, G. 1990 "Surface-based bedload transport relation for gravel rivers." Journal of Hydraulic Research 28 (4). Thome, e.R., and Zevenbergen, L.W. 1985 ''Estimating Mean Velocity in Mountain Rivers." Journal of Hydraulic Engineering 111 (4).

PAGE 194

NFIP-Accepted Computer Models: Proprietary Issues vs. Publics Right to Appeal Jerry W. Sparks Dewberry & Davis INTRODUCTION The vast majority of hydrologic and hydraulic analyses used to prepare, revise, or otherwise amend Flood Insurance Studies (FISs) for the over 18,000 communities participating in the National Flood Insurance Program (NFIP) are performed using computer models. In order to protect the interests and rights of appeal of conmllmities and property owners impacted by NFIP mapping, there are specific availability and distribution requirements for computer models used in the preparation or revision of NFIP maps. However, as personal computer (PC) technology has emerged, many new computer programs designed to model a wide variety of complex flooding situations have been developed, particularly in the private sector. The author's proprietary ownership of the program and its source code has led to inherent conflict with the NFIP availability and distribution requirements, which are intended to ensure national consistency and fairness. NFIP REGULATORY REQUIREMENTS Subparagraph 6S.6(a)(6) of the NFIP regulations requires that any computer model used to revise NFIP maps must be: Reviewed, tested, and accepted by a government agency responsible for the implementation of progranlS for flood control and/or regulations of floodplains; Well-documented, including source codes and user's manual; and Available to the Federal Emergency Management Agency (FEMA) and all present and future parties impacted by

PAGE 195

180 NFIP-Accepted Computer Models: Proprietary Issues flood insurance mapping developed or amended through the use of the program. For programs not generally available from a federal agency, the source code and user's manual must be sent to FEMA (which administers the NFIP) free of charge, with fully doclUllented permission from the owner that FEMA may release the code and user's manuals to impacted parties. Thus, NFIP regulations obligate FEMA to assure that the data and methodology doclUllentation are available to those impacted by NFIP mapping. In fact, Subparagraph 67.8(e) of the NFIP regulations states that liThe Administrator shall make available for public inspection the reports and other information used in making the final BFE determination. II This requires that, should a party impacted by a proposed NFIP map action appeal on the basis of a flaw or other technical incorrectness within the program itself, the program, and its source code and user's manuals, must be made available to the appellant. These requirements were put into place because of the real. world impacts of NFIP mapping on commtmities and property owners, and their land use practices. AUTHOR'S PROPRIETARY RIGHTS VS. NFIP REQUIREMENTS From the late 1960s through the mid 1980s, the computer capabilities required to perform complex calculations were generally available OIlly through large mainframe computers. Such resources were generally limited in small-to mid-size engineering firms. For this reason, the majority of hydrologic and hydraulic computer models were developed in the public sector. The U.S. Army Corps of Engineers' HEC-l and HEC-2 models, the U.S. Geological Survey WSPRO model, and the Soil Conservation Service's WSP-2 model are exan1ples of federally developed computer models that have been extensively used in the development of NFIP mapping. As federally developed computer models available in the public domain, these programs easily complied with NFIP availability and distribution requirements. However, as microcomputer technology, computing power, and speed have advanced and PCs have become more widely available and used, the nlUllber of sophisticated models capable of modelling a wide variety of complex flooding situations has increased dramatically in recent years. In many areas of the nation, a more complex model, such as an unsteady flow model, may be more appropriate to model existing conditions than a steady state flow model, such as HEC-2. Often, these new models have been developed by private entities and are required by local or state floodplain management agencies as part of the building permit process because of their applicability to local conditions. To avoid duplication of

PAGE 196

Sparks 181 effort and inconsistencies in results that would arise from using different models for different review and permit agencies, it is normally desirable from the standpoint of communities, developers, engineers, and property owners to use these same models to support NFIP map revision and amendment requests. However, many of these models do not meet the requirements of NFIP map usage because they have not been reviewed or accepted by a governmental agency and/or the program's author does not wish to make available the source code and user's manuals. The development of these programs is normally the result of a significant investment and risk on the part of the program's author, who spends thousands of hours and dollars in research, development, and marketing to produce unique and creative products. Because the source code is the core of these products, release of the proprietary information to third parties is tantamount to divulging trade secrets. Thus, we are left with quite a dilemma: how to uphold the rights and interests of communities and property owners directly impacted by NFIP mapping while still protecting the rights and interests of entrepreneurs who have struggled to fill a niche or void in the engineering community and develop a competitive edge. The NFIP regulatory requirements regarding the availability and distribution of computer programs are not intended to infringe on the rights of private program authors or otherwise retard entrepreneurial effort'> in the private sector nor are they intended to allow or promote distribution in the public domain. It is recognized that through the free market, many creative and revolutionary advances are made. However, the right of communities and property owners to appeal a proposed NFIP map action is a ftilldamental tenet of the NFIP and cannot be ignored or overlooked in the privacy interests of a particular program's author. adICPR-A SUCCESS STORY The advanced Intercormected Charmel and Pond Routing (adICPR) computer model is a privately developed, one-dimensional tillsteady flow model. This program, which was developed for use in the analysis and design of tail water dependent systems, is used extensively in the State of Florida and in some cases is required by Florida's Stormwater Management Districts. However, in 1994 the model did not meet NFIP requirements and, therefore, could not be used for NFIP mapping purposes. The particular sticking point was the program author's objection to making his source code available to third parties; to do so, he felt, would essentially put his proprietary trade secrets in the public domain. Because of its increased use and the validity of the program author's privacy concerns, FEMA and the author undertook a cooperative effort that included a confidentiality agreement that allowed FEMA to receive

PAGE 197

182 NFIP-Accepted Computer Models: Proprietary Issues all confidential items necessary for review, including complete source code and documentation, free of charge, for a predetermined period of time. In addition, the confidentiality agreement contained specific provisions for release of the program's source code and documentation to third parties. These criteria are designed to limit the distribution rights retained by FEMA while at the same time allowing the model to be provided to individuals who demonstrate a need to review it in Support of a valid appeal of a proposed NFIP map action. Specifically, this agreement defined "impacted party" to mean an agent of an owner or lessee of land in a community who has filed an appeal of a FIS and can demonstrate that the preparation of the appeal materials requires review of the source code and user's manual. Further, the agreement provides the author the opportunity to review all information used by FEMA to determine that a requestor qualifies as an "impacted party" prior to release of the infonnation. In addition, the requestor must sign a non-disclosure agreement with the program's author that provides specific confidentiality and time constraints on the requestor's review and use of the source code and user's manual. The agreement between the author and FEMA contains provisions allowing the progranl's author to take actions to protect his trade secrets and other rights contained in the source code and user's manual. FEMA subsequently conducted extensive review and testing of the program and provided the progranl's author with technical comments on the program. These comments were then evaluated by the program's author in consultation witt. FEMA, and a new version of adICPR (Version 2.0) incorporating FEMA's comments was released in September 1995. This version of the progranl meets the requirements of 65.6(a)(6) and is, thus, now accepted for NFIP mapping purposes. CONCLUSION With the continuing growth of privately developed hydrologic and hydraulic computer progranls in the engineering software marketplace, it is expected that more program authors wiII be interested in having their programs reviewed for NFIP acceptance. There are obvious economic benefits to the author of a software package that can be marketed and sold as a program for use from the local pennit review process through the NFIP map revision request. The NFIP also benefits from the addition of tools designed to accurately model certain flood conditions that may otherwise be modelled inappropriately if only public domain programs are utilized. This is vital because accmately mapping flood hazards assures that future development will be reasonably safe from flooding and existing development in danger of flooding will be protected by an mechanism. Thus, the flood insurance fund remains solvent.

PAGE 198

Sparks Through the review of adICPR discussed above, FEMA learned a great deal about protecting the private interests of the program's author while meeting the procedural and technical requirements of the NFIP. 183 This process helped establish the standards of other program authors interested in NFIP acceptance. However, given present fiscal and resource constraints on FEMA's already strained NFIP mapping budgets, it is likely that FEMA's review mechanism will continue to evolve as other programs enter the marketplace. Options that may be pursued include charging the program's author for FEMA's time to review and test the model or possibly developing review procedures and then turning the reviews themselves over to a research-oriented agency, such as the American Society of Civil Engineers' Civil Engineering Research Foundation or the National Academy of Sciences, with FEMA contributing funding. REFERENCES FederJJ Emergency Management Agency 1989 "National Flood Insurance Program." Federal Register 54 (43):9523. Federal Emergency Management Agency 1989 "National Flood Insurance Program." Federal Register 54 (156): 33541. Federal Emergency Management Agency 1994 "Emergency Management and Assistance." 44 Code of Federal Regulations.

PAGE 199

The NEXGEN Floodplain Hydraulics Program HEC-RAS Troy Lynn Lovell Michael A. Moya Emilia Salcido Halff Associates, Inc. INTRODUCTION This paper describes the initial experiences and results from the authors' application of the new river hydraulics program HEC-RAS (Hydrologic Engineering Center, 1995a). HEC-RAS, River Analysis System, was developed at the Hydrologic Engineering Center (HEC), and was released for preliminary testing in early 1994. HEC-RAS is the windows-based river hydraulics program of HEC's "next generation" (NEXGEN) of hydrologic engineering software. It is widely anticipated to be the replacement for the "classic" backwater program HEC-2 (HEC, 1990). Halff Associates was involved as a "BETA" tester for HEC's Fisc.tl Year 1994 BET A and Fiscal Year 1995 BET A2 versions of the software. The finn also conducted HEC-RAS training short courses (Lovell, 1995). Over 200 users provided comments to HEC on the BET A versions. HECRAS, Version 1.0, was released in August 1995. Version 1.1 (January 1996, included several new features and mm1erous corrections. Version 1.2 (April 1996) had a minor correction in the report generator module. This paper discusses the current strengths and weaknesses of HEC RAS, from a practitioner's viewpoint. There will be comparisons of HECRAS with the HEC-2 backwater program. Specific examples of creeks modeled with both programs and the results will be included. COnm1el1ts regarding future features of HEC-RAS will be made (Bonner, 1996). HEC-RAS, A NEW FLOODPLAIN MANAGEMENT TOOL The HEC-RAS software is an integrated package, designed for interactive use in a multi-tasking environment (Bonner, 1995). The system uses a graphical user interface (GUI) for file management, data entry and editing, program execution, and output display. HEC-RAS is designed to provide

PAGE 200

Lovell, Moya, and Salcido one-dimensional river modeling using steady-flow, lUlSteady-flow and sediment-transport computations based on a single geometric representation of the river network. However, the initial release only provides steady-flow, sub-critical, supercritical, and mixed-flow regime profile calculations for a river network. 185 Profile calculations are performed using the standard-step procedure. Overbank conveyance is computed incrementally at coordinate points (HEC-2 style) or at breaks in roughness (HEC-RAS default). Subcritical, supercritical, and mixed-flow profile calculations can be performed. Documentation includes a user's manual and a hydraulic reference manual (HEC, 1995b). The user's manual provides installation instructions, a program overview, an example application, file management, data entry, performing steady flow analysis, and viewing results. The hydraulic reference manual provides the theoretical basis for profile calculations; data requirements; optional capabilities; modeling bridges, culverts, and multiple openings; and floodway computations. APPLYING HEC-RAS TO THE REAL WORLD TIle initial applications of HEC-RAS were primarily to test the program and provide conm1ents to HEC during the BET A testing period. Since that time HEC-RAS has been used for mmlerous floodplain analyses and design applications. Initial reactions were less than positive until familiarity with the program brought guarded enthusiasm. Most of these first applications were existing HEC-2 files that were imported without any file sanitization (primarily bridges). Experiences have, on the most part, been very positive. When the HEC-2 style of convey,mce is used in HEC-RAS, the progranls produce identical answers in many cases. The HEC-RAS program does a good job of converting HEC-2 files into reasonable models of bridges. Some modification of the model is necessary to correctly represent the bridges, even though the direct conversion will produce similar answers in many cases. Floodways (Method 4) seemed more difficult to obtain exact results. Floodways using Metllod 1 will usually produce identical answers, using the same encroachment stations from HEC-2, and HEC-2 style conveyance. SOME HEC-RAS CASE STUDIES Case 1: Large River with 86 Miles of Stream, Six Sets of Bridges and 122 Cross Sections This example was a HEC-2 river routing study of the North Canadian River in Oklahoma, being prepared for the Corps of Engineers. Many of the cross sections were originally from lake sedimentation surveys and had over 200 points. These cross sections were modified to 95-100 points

PAGE 201

186 The NEXGEN Floodplain Hydraulics Program HEC-RAS to fit the HEC-2 limited fonnat. Note that HEC-RAS will allow up to 500 points. Additional cross sections were obtained from bridge plans, U.S. Geological Survey data, and from topographic mapping. Since the study was for a routing model, in conjunction with a real-time reservoir operation system, one of the technical problems was to block out non conveyance areas, but maintain them for storage purposes. This was handled by using extremely high 'n' values in the non-effective areas. If the HEC-RAS program had been available and approved for use, this problem could have been efficiently and correctly modeled with the non effective option, which accounts for storage. After calibration to two USGS gages, a large range of discharges was processed through the HEC-2 program. To test the HEC-RAS program, the HEC-2 files were imported and the new "imported" files executed without alteration of any of the data. Results and Comparisons The HEC-2 program executed the 7 profiles (discharges of 2,000 to 150,000 cfs) in 40 seconds (on a 486/66). The HEC-RAS program failed to complete the sixth or seventh profiles, after laboring over 2 minutes. A computational error had occurred and locked up the personal computer. The HEC-RAS model was re-executed with only 2 discharges, which took 45 seconds. HEC-RAS was run using the HEC-2 style of conveyance, JS well as the default HEC-RAS style, for comparison. Table 1 shows a comparison of the two different HEC-RAS runs, and an actual HEC-2 output for the same cross section. Case 2: Bridge Design Problem Using Metric Units Halff Associates is preparing hydraulic design of "off-system" bridges for the Texas Department of Transportation in several west Texas counties, All the models of these remote bridges are based on limited cross-sections and are in metric units. The creeks and bridges were originally modeled using HEC-2, but were later imported to HEC-RAS for the analysis. HEC RAS was used for the efficiency in a bridge design analysis mode, when quick and high quality graphics are desired and metric data is required, Case 3: Channel Improvements A major weakness of the current HEC-RAS progranl is the lack of a charmel improvement option (CHIMP in HEC-2). Although not as practical as a design tool now, charmel improvements can be perfomled by manipulating the HEC-RAS geometry to reflect the proposed improve ment. The notes for the HEC-RAS short course at the University of Texas

PAGE 202

Lovell, Moya, and Salcido 187 Table 1. Comparison of HEC-RAS with HEC-2, North Canadian River at cross section river mile 460.21. HEC-RAS Variable HEC-2 HEC-2 Style HEC-RAS Water Surface Elevation (ft-msl) 1844.74 1844.54 1844.58 Velocity Head (ft) 0.08 0.09 0.09 Ch:\lUlel Velocity (fps) 3.37 3.49 3.59 Conveyance (cfs) 835,446 794,926 775,526 Energy Slope (ft/ft) 0.000573 0.000633 0.000665 at Austin (Lovell, 1995) include an improved channel project workshop, requiring manual encoding of the channel improvements. For two channel improvement applications the process of creating an improved channel geometry file (CI Records) using HEC-2, producing a TAPE 16 file (replaces CI cards with GR points), and then importing the file to HEC-RAS for final results and report graphics, was used. In both cases, the imported HEC-2 to HEC-RAS model produced water surface profiles almost identical to the original HEC-2 model, using CI records. Within the year the HEC-RAS progranl should be upgraded with the hydraulic design module that will include a channel improvement option. Case 4: Floodway Application TIle HEC-RAS program has the full array of encroachment options and is an excellent tool for making Federal Emergency Management Agency tloodway determinations. HEC-RAS allows the user to automatically import tlle calculated encroachment stations from Method 4 directly into a Method 1 file. Setting up floodway files is much easier than in HEC-2, and the graphics enhance the visualization of the computed floodways. Early experiences with executing HEC-RAS (Method 4) from an imported HEC-2 file did not produce as close correlation to HEC-2 as was desired. When the HEC-2 encroachment stations were encoded into the HEC-RAS Method 1 file, identical answers were obtained, even upstream of a bridge that had not been altered after importation. One excellent HEC-RAS graphics feature is the "pseudo 3-D" perspective (Figure 1). CONCLUSIONS Initial reactions to HEC-RAS were slightly negative due primarily to unfamiliarity with the program logic and file organization, as well as differences from HEC-2 concepts on several issues (e.g., the default to

PAGE 203

188 The NEXGEN Floodplain Hydraulics Program HEC-RAS upstream to downstream modeling). Minor computational and operational inconsistencies and a lack of a total printout of input and output data (later corrected in January 1996 version) also were inconvenient. Creek F100dway Detenrmtlon Plan: FnoI Floodway Encroachnert Analysis Riv St 37200 to 36200 PFI: 1 2 Figure 1. HEC-RAS pseudo 3-D perspective graphics of floodway model. The most useful features of the HEC-RAS program are: Easy-to-Iearn menus and procedures. Excellent graphics, which are available during encoding, editing, and reviewing of results. These graphics and tables can be easily imported to word processing software. Options for conveyance calculations, comparative tables, plotted profiles Easy-to-use non-effective areas, levees, blocked obstruction, and encroachment options. Expanded bridge modeling options, with multiple openings and shapes Some frustrating features include: Difficulty in keeping up with the data files: projects, plans, etc. Lack of charmel improvement options.

PAGE 204

Lovell, Moya, and Salcido 189 Inability to transfer the HEC-RAS files back to a HEC-2 fonnat for other applications. An inconsistency on the profile plots that will not allow the correct stations to be plotted unless the first station is "0." This can be manually corrected, but is inconvenient. After investing considerable time in utilizing the HEC-RAS program for a number of applications, the vast potential of the NEXGEN programs becomes more evident and skepticism is usually replaced with unabashed enthusiasm. As hydraulic modelers become more proficient with the Windows envirorunent and as the HEC-RAS program evolves with the unsteady flow, sediment transport, interactive screen editor, and channel improvement options, the program should become as indispensable as the HEC-2 program has been for the past 25 years. Based on the rapid grasp of the HEC-RAS program by students at several workshops conducted this year, it seems that the software is truly "userfriendly." Floodplain hydraulic modeling will be much more fun in the future with the NEXGEN hydrologic and hydraulic programs. REFERENCES Bonner, Vernon and Gary Brunner 1995 HEC River Analysis System (HEC-RAS). Davis, CA: Hydrologic Engineering Center. Bonner, Vernon 1996 "HEC-RAS for Advanced HEC-2 Users," Lecture Notes for Short Course. Davis, CA: Hydrologic Engineering Center. Hydrologic Engineering Center 1990 "HEC-2 Water Surface Profiles," User's Manual. Davis, CA: Hydrologic Engineering Center. Hydrologic Engineering Center 1995:1 "HEC-RAS River Analysis System," User's Manual. Davis, CA: Hydrologic Engineering Center. Hydrologic Engineering Center 1995b "HEC-RAS River Analysis System," Hydraulic Reference Manual. Davis, CA: Hydrologic Engineering Center. Lovell, T.L. 1995 "River and Flood Plain Hydraulics Using HEC-RAS," Lecturer Notes. Austin, TX: University of Texas at Austin, Continuing Education.

PAGE 205

Using The UNET Model to Estimate a 100-Year Flood in the Designated Floodway I-Ming Cheng California Department of Water Resources INTRODUCTION The UNET model simulates a one-dimensional unsteady flow through a full network of open channels. This computer simulation model was originally developed by Dr. Robert L. Barkau and later adopted by tl1e u.s. Army Corps of Engineers Hydrologic Engineering Center at Davis, California. This model was applied extensively in the 1993 Midwest flood. The California Department of Water Resources has used the UNET model to estimate a 100-year flood flow for Cross Creek, a designated floodway in Kings County, California. This floodway was designated by the State Reclamation Board in September 1982. THE DESIGNATED FLOODWAY AND THE STATE RECLAMATION BOARD The Reclamation Board, created by the California Legislature in 1911, is the state agency that cooperates with the U.S. Army Corps of Engineers in controlling flooding along the Sacramento and San Joaquin rivers and their tributaries. The Board's efforts focus on controlling floodwater, reducing flood damage, protecting land from floodwater erosion that would affect project levees, and controlling encroachment into floodplains and upon flood control works, such as levees, channels, and pwnping plants. The Board also plans and adopts designated floodways, which is a nonstmctural means of ensuring the safe passage of flood flows through flood prone areas. CROSS CREEK DESIGNATED FLOODWAY The study area has a semiarid climate. Annual precipitation in the basin varies from about nine inches near Highway 99 to 13 inches near the foothills, and averages about 37 inches in the Kaweah River drainage

PAGE 206

Cheng 191 above Lake Kaweah. The drainage basin is about 1,300 square miles upstream from Highway 99. The historical record indicates that floods produced high flows that combined channel and overland flows in the area. Cross Creek, with a total length of 20 river miles, is located in the downstream end of the Kaweah River system. The Kaweah River system, under study, is quite complex. It includes Kaweah River at McKays Point where the river splits into north and south flows. The south flow becomes the Lower Kaweah River stream system. Lower Kaweah River flows into the Visalia Plain and is further divided into several branches, such as Mill, Packwood, Cameron, and Outside creeks. The north flow becomes the St. Johns River. To the north, Cottonwood Creek, flowing westerly, combines with Sand Creek inflow to join the St. Johns River flowing into Cross Creek. The study area is traversed by six major state highways and by the Southern Pacific and the Atchison, Topeka and Santa Fe Railroads, and the Friant-Kern Canal. Today, this area is a highly developed farming region devoted predominantly to the production of citrus fruits, grapes, walnuts, cotton, and grain. THE UNET MODEL A schematic diagram for the UNET model is shown in Figure 1. The inflow hydrographs used in the simulation are from the Kaweah River at McKays Point, Cottonwood Creek, and Sand Creek. The hydrographs were first routed using HEC-l to take into account the storage effect before applying the UNET model to the system. The local inflow to the system was not considered for this study. The UNET model routed inflow hydrographs for St. Johns River and Cottonwood, Sand, Mill, and Packwood creeks, in which each travels with a different time frame. The HEC-l model is used again at Highway 99 and the Southern Pacific Railroad for the storage effect. Boundary conditions for UNET can be input from any existing HEC-DSS data base. DSS (Data Storage System) is a very useful data base, which stores the input hydro graphs and output files generated for graphical display and for comparison with observed data. The cross sections are input in a modified HEC-2 reverse backwater fonnat. The floodplain under study is very flat, ranging from several hundred feet to more than two miles wide. It is necessary to create a pilot channel in each simulated floodway to minimize the instability resulting from the shallow flow in the floodplain. A special overflow weir was also created in the model at the St. Johns River to divert the flood water exceeding the channel capacity to the over-bank area.

PAGE 207

1 2300tv 9.573 cia l\ 51. Johns Bank 15.9 miles 51. Johns Channel u. 1.1\. 16.4 miles ,. .... . "" e,. Mill Creek North 0;;:' 11 miles 7,000 eta lL\ ----------------------------Highway 198 Mill Creek South 10 miles Packwood Creek 9 miles 0700hr 2.600 cts Figure L Cross Creek UNET model schematic with 3"/day infiltration. 'ffi a: u !E l;l IL E ., : o UJ .... CO I\) C IJI 5' ec C Z -o !r <1> III .... 8 -< <1> III .... "T1 0' &.

PAGE 208

I I I i I lSGUlniJ F 10000 L 0 W I / II 7 / N / C F !l0011 / ;' S li-/ .,,-/ 1----/' II -S000 aara 12aa IBJANU I 00ra SOUTH HY99 COMPUTED laa-YR FLOW _ _ .JACKSON AVE REV laa-YR FLOW 1299 19J'AN,9 aara /'-, / / /-p\ \ 12aa 2IJJANY'J \ I aara \ \ \ \ \ \ 12aa 2IJANU -\ \ \ \ , ;-aara Figure 2. Estimated lOO-year flood hydrograpbs, Cross Creek designated floodway with 3"/day infiltration. .... co t.)

PAGE 209

194 Using UNET to Estimate a 100-Year Flood RESULTS OF UNET SIMULATION The UNET model routed the complete hydrograph through the river channel and floodplain with different time frames. Two simulations were made: one without infiltration in the region and one with the assumption of three inches per day of infiltration. The results, shown in Figure 2, indicated the peak flood flow arrival time at different locations. The resulting 100-year flood flow at the site investigated ranged from 16,000 cfs with infiltration to 23,000 cfs without infiltration. This result is comparable with the 1988 FEMA Flood Insurance Study for a 100-year flood flow of 19,200 cfs at the East Branch of Cross Creek above the Tule River. The East Branch of Cross Creek is about nine miles downstream from the study area. An improper selection of a pilot channel in the floodplain would result in a simulation instability and distorted flood flow. CONCLUSIONS The UNET model was adequate for the level of detail required for this study to route time-dependent flood flow through the complicated open channels and floodplain. However, due to instability resulting from shallow overland flow in the floodplain, a special modification had to be made in the UNET model. REFERENCES U.S. Army Corps of Engineer, Sacramento District 1972 Flood Plain Informations-Sand and Cottonwood Creeks and the Lower Kaweah River, Visalia, California. 1986 "Hydrology-Kaweah & Tule River Reconnaissance Study, California" (office report). Davis, G.H., B.E. Lofgren and Seymour Mack 1964 Use of GroundWater Reservoirs for Storage of Surface Water in the San Joaquin Valley, California. U.S. Geological Survey Water-Supply Paper 1618 (with California Department of Water Resources) Federal Emergency Management Agency 1988 Flood Insurance Study, Kings County, California. Washington, D.C.: FEMA.

PAGE 210

Bridge Hydraulic Analysis with HEC-RAS Vernon Bonner Gary Brunner Hydrologic Engineeri ng Center, U.S. Army Corps of Engineers HEC-RAS OVERVIEW The HEC-RAS River Analysis System (Hydrologic Engineering Center, 1995a and b) is an integrated package designed for interactive use in a multi-tasking environment. The package is intended to be the successor to the current steady-flow HEC-2 Water Sur/ace Profiles progranl (Hydrologic Engineering Center, 1990). Version 1.1 provides steady-flow water surface profile calculations for a river network with sub-critical, or mixed-flow regime on computers with the MS Windows T operating system. The program has been developed based on a single definition of the geometric data for all modeling. River networks are defined by dIa\ving, with a mouse, a schematic of the river reaches from upstreanl to downstreanl. As reaches are connected together, junctions are automatically fomled by the program. After the network is defined, reach and jtmction input data can be entered. The data editors can be called by the appropriate icons in the Geometric Data Window; or reach can be imported from HEC-2 data sets. Cross-section data are defined by reach name and river station. Data are defined by station-elevation coordinates. Up to 500 coordinates are allowed. There is no maximlUll number of cross sections. Cross sections can be easily added or modified in any order. Cut, copy, and paste features are provided, along with separate expansion or contraction of the cross-section elements of overbanks and channel. Cross-section interpolation is provided using cross-section coordinates. The program connects adjacent cross sections with major chords, and the user can add chords graphically. The interpolated sections are marked in all displays to differentiate them from input data. Steady-flow data are defined for the reach at any cross-section location. Multiple-profile calculations can be performed. The boundary conditions are defined at downstreanl and/or upstream ends of reaches,

PAGE 211

196 Bridge Hydraulic Analysis with HEC-RAs depending on flow regime. Internal boundary conditions are dermed at the junctions. Options for starting profile calculations include: knovlll water surface elevation, energy slope (nornlal depth), rating curve, and critical depth. Profile calculations are perfomled using the standard-step procedure. Overbank conveyance is computed incrementally at coordinate points (HEC-2 style) or at breaks in rouglmess (HEC-RAS default). Method comparisons are provided in Technical Paper No. 147 (Hydrologic Engineering Center, 1994). Subcritical, supercritical, and mixed-flow profile calculations can be perfomled. The location of the transition between supercritical and subcritical flow is detemlined based on momentum calculations. Detailed hydraulic junlP location and losses are not computed; however, the jump location is defined between two adjacent cross sections. Tabular output is available using pre-defined and user-defined tables. Cross-section tables provide detailed hydraulic infomlation at a single location, for a profile. Profile tables provide summary infonnation for all locations and profiles. Pre-defined tables are available for the cross section, bridge, culvert and floodway computations. User-defined tables can be developed, from a menu of 170 output variables, and stored for use like pre-defined tables. Graphical displays are available for cross sections, profiles, rating curves, and a X-Y-Z perspective plot of the river reach, as shown in Figure 1. User control is provided for variables to plot, line color, width and type, plus axis labels and scales. TIle user can also zoom in on selected portions of the display, and zoom out to the original size. All graphics are in vector fornl using calls to the Windows Graphics Device Interface. Graphics can be sent to output devices through the Windows print manager, or they can be written to a meta file or sent to the Windows clip board. BRIDGE AND CULVERT ROUTINES The bridge routines in HEC-RAS enable analysis of bridge hydraulics by several different methods without changing the bridge geometry. The model utilizes four user-defined cross sections in the computations of energy losses due to the structure. An effective-area option is used with the bounding cross sections to define the ineffective flow areas, shown as cross-hatched area in Figure 1. Cross sections are fonnulated inside the bridge by combining the two bounding cross sections with bridge geometry, defined by the roadway/deck, piers, and abutments. Bridge data are entered through the editor, shown in Figure 2, along with the bridge modeling methods.

PAGE 212

Bonner and Brunner 197 Figure 1. XYZ plot of bridge sections. Low-flow Computations The program first uses the momentum equation to define the class of flow. For Class A low flow (completely subcritical), the modeler can select any or all of the following three methods to compute bridge energy losses: standard-step energy, momentum, or Yarnell equation. The U.S. Geological Survey-Federal Highway Administration WSPRO bridge routine (Federal Highway Administration, 1990) will be included in a later program release. If more than one method is selected, the user must choose a single method, or the highest energy solution, for the energy loss through the structure. For Class B low flow (passes through critical depth) tile program uses the momentwll equation. Class C low flow (completely supercritical) can be modeled with either the standard-step energy method or tlle momentwn equation. Pressure Flow When the flow comes into contact with the low cord of the bridge, pressure flow begins. The program uses energy-based (like HEC-2 Nornlal Bridge) or pressure-flow equations. It checks for the possibility of

PAGE 213

198 Bridge Hydraulic Analysis with HEC-RAS ClrOU'Id S35 ",11 5:lIl BIri: 610 2S SZIJ 515 SIC 540 S35 5:lIl 2S SZIJ 515 SID D 1[D] Figure 2. HEC-RAS Bridge Data Editor. pressure flow when the upstream energy-grade line exceeds the maxinnun low chord. The program will handle two cases of pressure flow: 1) tile sluice gate equation is used when the tail water is below the bridge, and 2) the full-flow orifice equation is used when the tail water is submerged. Weir Flow When water flows over the bridge and/or roadway, the overflow is calculated using a standard weir equation. For high tailwater conditions, the anlount of weir flow is reduced to account for the effects of submergence. If the weir becomes highly submerged, the progranl will switch to calculating energy losses by the standard-step energy method. The criterion for switching to energy-based calculations is user controllable. When combinations of low flow or pressure flow occur with

PAGE 214

Bonner and Brunner 199 weir flow, an iterative procedure is used to determine the amount of each type of flow. Culvert Hydraulics The modeling approach for culverts, cross-section layout, the use of ineffective areas, and contraction and expansion coefficients, is similar to that for bridges. For culvert hydraulics, the program uses the Federal Highway Administration's culvert equations (Federal Highway Administration, 1985) to model inlet control. Outlet control is analyzed by either direct-step backwater calculations or full-flow friction losses, plus entrance and exit losses. The culvert routines have the ability to model the following shapes: box, circular, arch, pipe arch, and elliptical. Multiple culverts, of different types, can be modeled for a single location. Multiple Bridge Openings Multiple openings can be modeled by two approaches, as divided flow in two reaches or by the multiple-opening approach. The multiple-opening approach can analyze combinations of three types of openings: bridges, culvert groups, and conveyance areas. Up to seven openings can be defined at anyone river crossing. Each opening is evaluated separately and the total flow is distributed such that the energy loss in each is equal. Bridge Testing The bridge routines of HEC-RAS, HEC-2, and WSPRO were tested using 21 L;SGS data sets from the Bay St. Louis Laboratory (Ming et aI., 1978). All the models were able to compute water surface profiles within the tolerance of the observed data, which varied on the order of 0.1 to 0.3 feet (Hydrologic Engineering Center, 1995c). For HEC-RAS and HEC-2, tile energy-based methods reproduced observed bridge low-flow losses better than the Yarnell method. Also, the apparent downstream expansion reach lengths were shorter than rates suggested in HEC guidelines (Hydrologic Engineering Center, 1990). Because all the prototype bridge data ;ame from similar wide, heavily vegetated floodplains with low velocities, additional research was conducted using the RMA-2V computer program (King, 1994). Application of the 2-D model to the prototype data demonstrated that RMA-2V could reproduce observed bridge-flow depths and trill1sitions. BRIDGE FLOW TRANSITIONS An M.S. thesis project (Hunt, 1995) was conducted to investigate bridge expansion and contraction reach lengths and coefficients. Twodimensional models of idealized bridge crossings were developed using

PAGE 215

200 Bridge Hydraulic Analysis with HEC-RAS RMA-2V (King, 1994). River slopes, bridge opening widths, overbank to channel n-value ratios, and abutment type were varied; a total of 76 cases was modeled. From the model results, regression analyses were perfonned to develop predictor equations for contraction and expansion reach lengths, ratios, and coefficients. The flow transitions through a bridge crossing that blocks a portion of the overbank area are typically modeled with four cross sections, shown in Figure 3. The downstream and upstream sections 1 and 4 represent the full floodplain conveyance. The bridge-bounding cross sections 2 and 3 represent the effective flow area just downstream and upstream from the bridge. The bridge interior is modeled with the bounding cross sections plus bridge data. The question is where to locate the full-flow downstream and upstream sections 1 and 4 to model the flow transition. Expansion Reach Lengths (Le) The expansion ratio (ER in Figure 3) was less than 4: 1 for all of the idealized cases. The mean and median values of the expansion ratio for the idealized cases were both around 1.5: 1. These observations indicate that the traditional 4: 1 rule of thun1b will overpredict the expansion reach length for most situations. Many independent variables and combinations Lc 1 ./1 2 J ,/',/' 11 "'------L typical flow ,/' ER 1 transition pattem Flow .. ): /: 1/ .. 1/' assumed flow transition pattern for 1dimensionaJ '" :"'09 \ ,. '-..1 .... '-......, I. Figure 3. Conceptual illustration of transition reaches.

PAGE 216

Bonner and Brunner 201 of variables were investigated to find a possible correlation with Le. The variable that showed the greatest correlation was the ratio of the main channel Froude number at the most constricted Section 2 to that at the normal flow Section 1. equation for Le had an adjusted determination coefficient equal to 0.84 :md a standard of estimate (Se) of 96 feet. Similarly, the equatIOn for ER had a R = 0.71 and Se = 0.26. Based on the model data, a table of expansion ratios was developed. The distance to the downstream end of the expansion reach is estimated by multiplying the expansion ratio by the average obstruction length (half the total floodplain reduction caused by the two bridge approach embankments). Contraction Reach Lengths (Lc> In contrast to the expansion results, the results for contraction reach lengths lend some support to the traditional rule of thumb that recommends a 1: 1 contraction ratio. The range of values for this ratio was from 0.7: 1 to 2.3: 1. The median and mean values were both around 1.1 to 1. The Froude number ratio in the previous two equations also proved to be significant in its relationship to the contraction reach length. The most significant independent variable for this parameter, however, was the per centage of the discharge conveyed by the two overbanks. The best-fit equation had a R = 0.87 and Se = 31 feet. None of the attempted regression relationships was a good predictor of the contraction ratio. Expansion Coefficients (Ce> TIlt transition coefficients did not lend themselves to strong regression relationships, partly due to the fact that the velocity head differences were so small. The calibrated expansion coefficients ranged from 0.1 to 0.65. The median value was 0.3, which is less than the traditional value of 0.5. It is reconm1ended that the modeler use an average value and conduct a sensitivity analysis using values of the coefficient that are 0.2 higher and 0.2 lower, which represent the 95% confidence band for the best predictor equation. Contraction Coefficients (Cc> Of the 76 cases used in the regression analysis, 69 had calibrated C c values of 0.10. The values for the contraction coefficient ranged from 0.10 to 0.50. The mean was 0.12 and the median value obviously was 0.10. TIle data of this study did not lend itself to regression of the contraction coefficient values. For nearly all of the cases the value that was determined was 0.1, which was considered to be the minimum acceptable value.

PAGE 217

202 Bridge Hydraulic Analysis with HEC-RAS REFERENCES Federal Highway Administration 1985 Hydraulic Design of Highway Culverts, Hydraulic Design Series No.5. Washington, D.C.: U.S. Department of Transportation. 1990 User's Manualfor WSPRO-A Computer Model for Water Surface Profile Computations. Publication No. FHWA-IP-89-027. Washington, D.C.: U.S. Department of Transportation. Hydrologic Engineering Center 1990 HEC-2 Water Surface Profiles, User's Manual. Davis, CA. 1994 HEC River Analysis System (HEC-RAS). Technical Paper No. 147. Davis, CA. 1995a HEC-RAS River Analysis System, User's Manual. Davis, CA. 1995b HEC-RAS River Analysis System, Hydraulic Reference Manual. Davis, CA. 1995c Comparison of the One-Dimensional Bridge Hydraulic Routines from: HEC-RAS, HEC-2 and WSPRO. Research Document No. 41. Davis, CA. Hunt, John 1995 "Flow Transitions in Bridge Backwater Analysis," M.S. thesis, University of California at Davis and HEC Research Docwnent No. 42. Davis, CA. King, Ian P. 1994 "RMA-2V Two-Dimensional Finite Element Hydrodynamic Model." Lafeyette, CA: Resource Management Associates. Ming, C.O., B.E. Colson, and G.J. Arcement 1978 Backwater at Bridges and Densely Wooded Flood Plains. Hydrologic Investigation Atlas Series. Reston, V A: U.S. Geological Survey.

PAGE 218

Part 8 MAPPING

PAGE 219

This page IS intentionally blank

PAGE 220

Use of ARC/INFO for Floodplain Generation and Mapping in Jefferson County, Kentucky Mark A. Sites Louisville and Jefferson County Metropolitan Sewer District James A. Harned Louis T. Greenwell Alan M. Castaneda Ogden Environmental and Energy Services INTRODUCTION The Louisville and Jefferson COlUlty Metropolitan Sewer District (MSD) has been the regulatory authority for storm water-related issues in Jefferson COWlty, Kentucky, since 1987. In the same time frame, the government agencies within the county embarked on a venture to collectively develop a geographic infonnation system (GIS), the Louisville and Jefferson COlmty Information Consortiunl (LOnC). Lonc data is maintained by the individual participants and made available for use by the consortiUOl members. LOnC participants include the City of Louisville, Jefferson County, MSD, and the Property Valuation Administrator (PVA). LOnC employs a staff of 13 personnel who maintain the digital map data, which includes topography, and are available to consortiUOl members to develop custom applications. MSD, in particular the stomlwater plarming and review staff, realize the power and potential of LOnC to improve the day-to-day development review and long-tenn planning processes. MSD has developed several custom ARC/INFO applications for the purpose of developing plarming level models and mapping the resultant floodplains. The LOnC libraries include data related to soils, land use, and digital contours, which are the building blocks for hydrologic (HEC-l) and hydraulic (HEC-2) models. Custom applications are in place that compute runoff curve nUOlbers (CNs), produce a skeletal HEC-l input deck, develop cross-sections, produce a skeletal HEC-2 input deck, and map the floodplain based on hydraulic model output. The power of these applications is the speed and

PAGE 221

206 ARC/INFO for Floodplain Generation and Mapping accuracy with which floodplain maps can be produced, and the speed at which the impacts of basin-wide or site specific changes can be assessed. MODEL DEVELOPMENT AND MAPPING USING ARCIINFO The ARC/INFO commands to develop the major input parameters for hydrologic and hydraulic models are relatively simple. Most of the programming functions intersect polygons or three-dimensional (3D) surfaces. CN calculations are based on the intersection of land use, soils, and drainage basin divides. This intersection produces a set of polygons that can be related to database or look-up table of CNs for a specific hydrologic soil group, antecedent moisture condition, and land use. Crosssections are created by slicing a 3D ground surface at the point of interest. A 3D ground surface can be developed from either digital contours or the individual mass points. Similarly, a 3D water surface can be developed from hydraulic model output. When the ground and water surfaces are intersected, the result is a delineated floodplain. The accuracy of model input parameters developed from a GIS is dependent upon the data. In the case of LOJI C, all data sets are developed consistently to national mapping standards. The digital topographic data is accurate to within one foot vertically and two feet horizontally. This type of accuracy is consistent with that of the models available today for water surface profile computation. The soil and land use layers in LOJIC were prepared by the Natural Resources Conservation Service (NRCS) and the Planning Commission, respectively. They represent the best available information for use in modeling. The use of a GIS provides the benefit of detailed and consistent calculation of appropriate factors, and removes some of the variability associated with the human element in traditional modeling practice. However, the system should not be treated as a black box. The use of good engineering judgement should never be discarded and the results of GIS-produced model input should always be verified. CUSTOM LOJIC APPLICATIONS HEC-1 Input In addition to the CN, ARC/INFO has been used to compute the drainage basin area and lag time. A separate application arranges the basin area, CN, and lag time in the appropriate sequence of HEC-l input records. TIle model developer is still required to add rainfall data and necessary channel or structure routings to complete the model.

PAGE 222

Sites, Harned, Greenwell, and Castaneda 207 HEC Input The custom HEC-2 input development application produced cross-section coordinate data and stream centerline and overbank distances. Cross sections can be produced at locations specified by the modeler or at specified uniform distances. The application also has the capability to produce a cross-section coverage based on an existing HEC-2 input file. The nwnber of potential points within a given cross-section can be large; however, HEC-2 is limited to 99 pairs of coordinates in a given cross section. Therefore, the application has some built-in logic to select the proper points to adequately describe the section. It should also be noted that the cross-sections must extend to the full limits of the expected floodplain. The program begins at the left endpoint of the section looking downstream and moves between potential cross-section points. If a slope of 1 % or more is detected, the point is recorded. If this routine produces too many points, the slope parameter is relaxed and the process repeated. A second application utilizes the cross-section coverage and the conti:1Uous stream centerline coverage to compute stream and overbank distances between cross-sections. A third and final application incorporates the results of the first two applications to produce the skeletal HEC-2 input file. The input records populated by the application include the X 1 and OR records. As with the HEC-l application, the modeler must add now and Manning's n data and code any necessary bridges or culverts. Since LOnC topographic data is aerial photography based, it cannot determine the cross-sectional geometry below the waterline. This infOImation must still be gathered from a field surveyor adapted from an existing model. Floodplai n Generation 111is application requires the cross-section coverage and 3D ground surface to be in place prior to execution. Creating a 3D grolmd surface can be a memory and computational time intensive process. For large watersheds it is suggested that the area within the expected floodplain (the limits of the sections) be clipped from the data set and used to generate the sudace. The application requires either a standard HEC-2 output table (slUllmary table 150) or a comma delimited ASCII file containing the cross-section identifier and the water surface. The application was built for HEC-2, but can accommodate any hydraulic model output, such as HEC-RAS, using the ASCII forumt. Water surfaces are assigned to the appropriate cross-section and a 3D water surface is generated. The ground and water surfaces are intersected and the ground surface is essentially "filled" with water. This approach produces a smooth transition between

PAGE 223

208 ARC/INFO for Floodplain Generation and sections. Since the ground and water surfaces are generated on a grid system, the edge of the floodplain is generated with a ragged edge. This is not usually visible on mapping at scales over one inch equals 50 feet. ARC/INFO includes smoothing operations that can be used to transfonn the grid based edge into a continuous line. ARCIINFO GENERATED FLOODPLAIN MAPPING Floodplain mapping has three primary purposes; the federal flood insurance program, development review, and watershed planning. The Digital Flood Insurance Rate Maps (DFIRMs) for Jefferson County were produced by Figure 1. A small section of the test watershed. LOnC and published by the Federal Emergency Managemen Agency (FEMA) in February 1994. Detailed study areas (AE Zones) were hand plotted on 400 scale LOnC topographic maps and hand digitized to produce the FEMA floodplain. Approximate study areas (A Zones) that were not updated were digitized from 1978 FEMA maps. MSD intends to capitalize upon the LOnC topographic and land use data and the custom applications that have been developed to produce consistent floodplain mapping for the county. A small test watershed was selected to compare the results of the GIS-based model development and floodplain generation process to a calibrated detailed study area for which the resultant floodplains were hand-drawn and digitized. Figure 1 shows a small section of the test watershed. The heavy dark line delineates the old detailed study and the shaded area represents the GIS-produced floodplain. The water surface elevations produced by each method were similar, but as seen in Figure 1 the resultant floodplains vary. This is primarily due to the interpolation between sections during hand-drawing of the original floodplain. The differences are significant enough to have impacts on the determination of whether a particular home is within the floodplain limits. The magnitude and frequency of the differences are more pronounced in the approximate study areas (A Zones).

PAGE 224

Sites, Harned, Greenwell, and Castaneda 209 A county-wide set of GIS-based models will have implications on the insurance program. Changes will occur in the status of individual homes. A consistent, accurate set of floodplains and subsequent updates will be fair and defensible. IMPACTS ON WATERSHED MANAGEMENT Jefferson County is currently undergoing an update of the Comprehensive Plan, Cornerstone 2020. As part of this process a stream corridor plan has been developed and adopted and the floodplain ordinance is being revised. The proposed stream corridors will be defined based on a fully-developed or future condition. The proposed floodplain ordinance also incorporates the fully-developed condition concept. Individual site review and watershed management decisions will be based on the floodplain generated under the fully-developed condition. The fully-developed condition floodplain can be significantly larger than the current existing floodplain depending on land use and topography. The fully-developed condition floodplain was generated for the test watershed and the differences are shown in Figure 2. Again the existing FEMA floodplain is shown as the solid d'lrk line 1l1d the fullyjeveloped floodplain as the ,haded area. Because of the .0 o 0 .0", [J B OPOCl m OJ-I;] Figure 2. Fully developed floodplain. arge difference and impact to structures and property, the need for controls for future development can be identified. This illustrates hat floodplain mapping for planning purposes is a dynamic process. For his reason separate floodplain covers will exist for watershed nanagement and flood insurance purposes until such time as a fmal lIatershed master plan is adopted.

PAGE 225

A Substitute For Floodplain Delineations Gregory Rodzenko Julie Lemmon Flood Control District of Maricopa County INTRODUCTION Administrative costs can be quite burdensome when attempting to process multiple Letter of Map Revision (LOMR) requests for a simple development project. Our concept involves a unique agreement between the Flood Control District of Maricopa County and a private developer/homebuilder that reduces the administrative costs associated with removing a property's flood hazard designation from federal Flood Insurance Rate Maps (FIRMs). Our agency, the Flood Control District of Maricopa County, identifies flood hazards in areas where development is ongoing or imminent. This information is submitted to the Federal Emergency Management Agency (FEMA), which publishes FIRMs. To safeguard people and property from flooding, the FIRMs are used by conununities nationwide to regulate development within flood hazard areas (lOO-year floodplains). They
PAGE 226

Rodzenko and Lemmon 211 Local agency Identifies floodplain Data submitted to FEMA s Published -Homeowners must pay flood Insurance Developer submits plans to local agency I Agency submHs data to FEMA for CLOMR I Developer submits "As-Built" plans to local agency I Homeowner no longer pays flood Insurance I Figure l. Nonna} sequence of the Letter of Map Revision process.

PAGE 227

212 A Substitute for Floodplain Delineations conswning for all parties concerned. We collectively wondered if an alternative process were available. To disregard the existing flood hazard until Del Webb completed their project would have been irresponsible. Additionally, there are no guarantees that developers will complete their projects as originally conceived or that, until flood control infrastructure is in place, interim flood hazards will be addressed. Yet, to submit our flood hazard information to FEMA so that new FIRM maps could be published seemed a waste of time, money, and effort when the developer would soon embark on the process of having these areas removed from the maps (albeit over a 7-year period). Also at issue were the homeowners residing in the flood hazard areas being required to pay flood insurance until the LOMRs were submitted to FEMA and approved (with each submittal taking months to complete), NEW PROCEDU RE Our solution involves an agreement negotiated between the District ,md the developer. The flood hazard information involving the developer's property will not be submitted to FEMA if the District can be assured that flood control measures will be incorporated into the project along the way (Figure 2). This, in effect, will eliminate the flood hazard areas as Sign agreement with bond Build phased flood control Figure 2. Modified sequence of the Letter of Map Amendment process.

PAGE 228

Rodzenko and Lemmon 213 development occurs. To protect the District's investment in the already completed flood hazard study, a performance bond is required of Del Webb in an amount equal to the cost of all the work elements in a flood hazard study, including surveying, mapping, hydrology, hydraulics, and FEMA review fees (a total of $180,900). The bond acts as "insurance" to cover the District's cost of re-studying the affected property if Del Webb defaults on its agreement. (Note that a re-study would be necessary because mass grading associated with the development would alter runoff quantities and drainage patterns.) In addition, the agreement contains language requiring the developer to address any imminent flood hazards that might arise before the flood control infrastructure is completed. Provisions allow the District to take such actions itself if the developer fails to do so, with the costs being borne by the developer. The "protection" in place for the District is that if the developer, for any reason, does not build the flood control structures according to mutually agreed upon specifications and schedule, the District may call in the bond money, generate the floodplain mapping according to the physical conditions at the time construction stopped, and submit it to FEMA. The Board of Directors of the Flood Control District has implemented a policy that allows District staff to use this new procedure with other large homebuilders/developments as long as local floodplain management guidelines are met. Also, large parcels generate savings, small parcels require only one LOMR. ADVANTAGES OF THE NEW PROCEDURE The newly negotiated agreement has eliminated the homebuilder's cost of hiring an engineering consultant to prepare the LOMRs to "un-do" the admmistrative flood hazard; eliminated the staffs technical review costs at the local level to "un-do" the flood hazard; and eliminated the staff tecb1ical review at the federal (FEMA) level to "un-do" the flood hazard. POTENTIAL DISADVANTAGES could go wrong and who would be liable caused great concern on the part of the homebuilder and the Flood Control District. What if only part of the development is constructed? What if Del Webb sells the project to another developer? What if Del Webb goes out of business? Many hours were spent writing a "tight" agreement with mmlerous safeguards and review points so that the District can monitor the homebuilder's/developer's activity. The District also imposes rigorous guidelines regarding who can qualify for the program. This limits the District's exposure and allows the

PAGE 229

214 A Substitute for Floodplain Delineations District the opportunity to closely observe and evaluate the program, and its participants at the outset, limiting the size of any unforseen problems. Only the larger homebuilders currently qualify for the program; the District's general operating assumption is that larger companies are more fmancially stable and can more readily qualify for the bond. The agreement places the District in the position of depending on the developer to remain in business to address any problems. While the bond exists as an insurance policy to take care of problems not addressed by the homebuilder, the District would prefer not to "call in" such a bond. If all goes well, the District plans to expand the program by allowing access to smaller homebuilders. CONCLUSION It is difficult to estimate the percentage of clients a program such as ours potentially serves. Tens of thousands of building permits are issued each year throughout the United States and a majority of the permits involve private homebuilders and developers. All developers must address drainage and flood control issues during construction and some may involve the delineation of flood hazard zones. If agreements similar to ours were established prior to construction, thousands of homeowners along with hundreds of homebuilders and jurisdictions could benefit. Points to be considered before embarking on this alternative process: Willingness of homebuilders to go along with the conditions of an agreement. The stability of the homebuilders/developers. The ability of the local agency to oversee the process, which is very staff-intensive and may overload jurisdictions without sufficient technical/legal staff support. The staff time spent by all parties developing the first agreement is estimated at 400-500 person-hours. Parties involved included the Flood Control District, the District's legal counsel, Del Webb's management, Del Webb's attorneys, and Del Webb's engineering consultant. The person hours were nearly equally split between public and private. Using our first agreement as a model, others could likely spend considerably less tin!e and money developing an agreement specific to their project.

PAGE 230

The Zone A Crunch David R. Knowles Federal Emergency Management Agency, Region I Peter A. Richardson Green International Affiliates, Inc. INTRODUCTION The Mortgage Portfolio Protection Program (MPPP), which went into effect in 1991, allowed lending institutions to "force place" flood insurance coverage on structures in Special Flood Hazard Areas (SFHAs) delineated on Flood Insurance Rate Maps (FIRMs) if the mortgagor did not voltmtarily purchase a policy. The National Flood Insurance Reform Act of 1994 contained a provision for fining lending institutions that did not maintain flood insurance coverage for federally backed mortgages on structures in SFHAs in conununities participating in the National Flood Insurance Program (NFIP). Even before 1991, the secondary mortgage market (Farmers Home Administration, Farmie Mae, Freddie Mac, etc.) would not purchase mortgages on structures in an SFHA unless the full market value of the structure was covered by flood insurance. Under these strict financially based requirements, lending institutions began to make sure that loans at risk from flooding were covered by flood insurance. In other words, lending institutions began strictly enforcing the flood insurance requirements of the Flood Disaster Protection Act of 1973. TIie FIRM and accompanying flood insurance study (FIS) are the only source of infom1ation that can be used by lending institutions when determining whether flood insurance is required for a structure. Lenders are not at liberty to utilize additional scientific and technical data that would refute the FIRM delineation. Before the MPPP and, perhaps just COincidentally, prior to the bank collapses of the late 1980s, lenders often accepted data other than the FIRM when determining the need for flood insurance. Professional engineers, licensed land surveyors, community officiais, and the structure owners themselves often provided information to validate the contention that the FIRM was inaccurate in certain areas. This was particularly true for structures located in or adjacent to SFHAs

PAGE 231

216 The Zone A Crunch designated as Zone A (studied by approximate methods). As banking regulations affecting the NFIP have tightened, lenders have begun to adhere more carefully to using the FIRM for flood zone determinations. Recently, the banking industry has also indirectly put pressure on community permitting officials to carefully review the FIRM and resolve discrepancies (which occur more frequently in Zone A areas) prior to allowing development. Even when a community was not participating in the NFIP, strict use of the FIRM by lenders has prompted structure owners to petition for commlmity participation in the NFIP and/or get Zone A delineations revised. The dramatic increase in the scrutiny of FIRM data over the past several years can be seen in Figure 1. Map revisions not processed by Region I are not included in the graph. ] J J!I - :=0 ID" d! c( :::Ii 9 0 -.JS E :=0 :z: 360 300 RaquaotD Received ................................ FEMA Rggion I 2!S0 +..................................................................................................... 200 150 100 !So 0+---1989 1990 1991 1992 1993 1994 1995 Fiaca. V.ar (October Through September) Figure 1. Letter of Map Amendment requests received by Federal Emergency Management Agency, Region I. Various methods are employed by FEMA to revise Zone A designations. The revision process may be done for a single structure as part of a LOMA request or as part of a physical map revision for an entire watercourse. When the LOMA process is used, various technical analyses, like those in FEMA's publication, Managing Floodplain Development in Approximate Zone A Areas, are adequate. When more accuracy is needed, the FEMA regional office often has the Limited Map Maintenance Program (LMMP) contractor perform the study. Several case studies from New England describe the cost-effective methods used (or to be used) to make NFIP map changes and alleviate the Zone A crunch.

PAGE 232

Knowles and Richardson 217 CASE STUDIES Orrington, Maine Orrington joined the regular phase of the NFIP in 1994 under pressure from SFHA residents who were being forced by banks to buy flood insurance. The town's FIRM has only Zone A delineations. Community officials asked FEMA to do a detailed study of floodprone areas. Two homeowners applying for LOMAs at approximately the same location along the Sedgeunkedunk Stream provided widely different data for hydrologic studies. The data for one LOMA application indicated a drainage basin of 18.2 mi2 and a peak IOO-year flow of approximately 1,025 cfs using the U.S. Geological Survey regional equation for Maine. The other application provided data showing a drainage area of 17.6 mi2 and a peak 100-year flow of 7,700 cfs using TR-55. This appeared to be a difference in the application of the methodology. This case indicates the necessity for a study using a FEMA-approved methodology appropriate to the basin. A-Zone homeowners are now subject to a wide range of flow values and must perform site-by-site hydraulic analyses. Orrington is scheduled to be revised using the LMMP. Newfound Lake, New Hampshire The four communities bordering NewfOlmd Lake (Alexandria, Bridge water, Bristol, and Hebron) had inconsistent flood zone designations on their respective FIRMs for the lake and its shoreline. The Alexandria and Bridgf,water FIRMs had the lake and shoreline designated as Zone C. The Town of Bristol, where the control dam for the lake is located, had the lake as Zone B (even though the PIS report stated that the lake would experience a 3-foot rise in a 100-year event). Finally, the Town of Hebron designated the lake and much of the shoreline as a Zone A on its FIRM. The New Hampshire NFIP Coordinator received complaints from the Hebron building official that property owners in Hebron were being subjected to the requirements of the NFIP while those in the other three conul1lmities were not-even though in some cases structures in Hebron were at the sanle elevation or higher than those in the other communities. Newfound Lake is about 6 miles long by I mile wide. Its contributing drainage area is about 96 square miles at its outlet. The New Hampshire Water Resources Division (NHWRO) controls the level of the lake at the dam in Bristol. The lake is a popular recreation area near the White Mowltain National Forest and there are hundreds of structures around the lake, both full-time and sun1ffier homes and commercial establishments. Through the LMMP, FEMA Region I directed Green International Affiliates, Inc. (Green) to study the lake by detailed methods to determine a consistent elevation to be used by all four communities. A HEC-1

PAGE 233

218 The Zone A Crunch model was developed for the watershed area to account for the large amount of flood storage in the lake. Flood routing was performed with HEC-1 by developing a HEC-2 model for the control dam and running a range of discharges to develop a stage-storage-discharge curve. Rainfall depths for the 10-, 50-, and 100-year events were taken from T.P. 40 and a 500-year event was constructed by extrapolation. The four events were run in the HEC-1 model to detemline the peak stages for the lake and corresponding peak discharges at the dam. The peak return period discharges were then reentered into the HEC-2 model to develop final flood profiles through the lake. Determination of the base flood elevation (BFE) for the lake assunled that the lake was at its sun1Il1ertime "high" level and that the NHWRD was unable to reach the dam in time to release water via sluice gates or stop logs when the l00-year flood occurs. Londons Brook, Fairfield, Connecticut In the upper portion of the Londons Brook watershed, the stream channel had been diverted into a closed piping system to aIlow for the construction of a residential subdivision. Although this condition had existed for several years before the publication of the FIS for Fairfield. the town's FIRM showed the original brook location as a Zone A. In the process of trying to buy and/or seIl property in the area, a number of property owners realized that their homes were located in an SFHA (i., .. the fonner brook bed). A nunlber of these people never even knew the brook existed and that it was now running in a pipe under their street. Residents complained to town officials who notified FEMA of the problem. The issue also received Congressional interest. In the effective PIS for Fairfield, Londons Brook is a detailed study area from its confluence with the Rooster River to the point where it discharges from the piped system. The total contributing drainage area to the outlet of the closed piping system is approximately 0.8 square miles. It appears that the brook had been diverted into a piped system during three different phases of constmction, all of which occurred before or about the time the original FIS was completed. The first two phases were done to facilitate residential constmction. The last phase was done by the town to aIleviate flooding problems in backyards of the few homes which still contained portions of open brook. Long-time residents reported that flooding occurred until the town completed the piping project. FEMA Region I used the LMMP to correct the mapping problems with Londons Brook by having Green perfonn a detailed study of the brook from its current upstream limit of detailed study to a point upstream of the residential development. Hydraulic grade line calculations for the piping system were perfonned to detemline its maximum carrying capacity. Green detennined that the system would not carry the peak 100-

PAGE 234

Knowles and Richardson 219 year flood discharge without surcharging and computed overland flooding by perfonning a HEC-2 analysis using the difference between the total peak discharge and the capacity of the piping system. The overland flooding was mapped as a Zone AE and profiles were developed. Because several portions of the SFHA were computed to be less than a foot deep and the drainage area is less than one square mile, there was some discussion during the technical review of Green's work that a significant portion of the SFHA be mapped as a Zone AO. Green indicated, however, and FEMA Region I agreed, that the Zone AO designation would be difficult to regulate in a densely developed area. Although some structures previously not affected by the Zone A delineation may now be shown in the SFHA, many residences will be removed. By providing BFEs on the FIRM, FEMA will significantly reduce the Zone A crunch for Fairfield. Tributary to Middle Branch Mousam River, Alfred, Maine A ZOlle A is delineated on the current Alfred FIRM around a large wetland in the Middle Branch Mousanl River watershed. This designation is inconsistent with the contour mapping for the area (USGS quad with a 20-foot contour interval) as it includes sections of "high ground" more than 10 feet above the wetland. Also, the Zone A does not include the low-I:-;ng area tluough which the actual brook runs. A moderate number of pre-FIRM homes exist within the Zone A. When the town adopted the FIS ill 1990, most residents did not attend public meetings describing the study. nley were unaware of the financial implications surrounding the Zone A designation. To eliminate the Zone A crunch being placed on the town (lJlJilding pennits), residents (flood insurance premiunls), and FEMA Region I (LOMA requests), the town floodplain coordinator requested that FEMA include the wetland in a LMMP restudy for Alfred. A LMMP restudy, which included aerial photogramettric mapping, was scheduled to be performed by Green for the Mousanl River in Alfred. It was difficult, however, to include the wetland area as part of the LMW task because the wetland, and the tributary of which it is a part, have a total drainage area of only about 0.5 square miles at the dowllstream end of the delineated Zone A. In addition, the surrounding area contains only moderate residential development, mostly along Middle Branch Drive, and is several miles by road from any established vertical contIOI points. Only a very limited study could be justified. Giobal positioning system (GPS) survey methods were being used to establish grolmd control for the aerial survey on the Mousam River. With one additional GPS setup (for one hour, vs. two days of level running) in the Middle Branch Drive area, Green was able to cost-effectively establish vertical control necessary for a hydraulic analysis of the remote wetland.

PAGE 235

220 The Zone A Crunch The USGS regional regression equations for Maine were used to establish peak flood discharges and a HEC-2 model was developed for the stream through the wetlands. Although the wetland appeared to be a large flat ponding area on the USGS quadrangle sheet, surveyed cross sections revealed that there was a change in elevation of 18 feet over the 2000 foot length of stream through the wetland. It was thus determined that the regional regression equations would give satisfactory results. With a limited number of field-surveyed cross sections, this remote area was restudied at a very reasonable cost to FEMA and will be delineated as a detailed study area. The FIRM will show a much more accurate delineation of the SFHA. When the new FIRM becomes effective, the Middle Branch Drive area will have established BFEs and bench marks. Residents will be able to acquire elevation certifications from local consultants if needed for building permits or flood insurance. CONCLUSIONS Several issues related to the passage of the National Flood Insurance Reform Act have created the Zone A cnmch: Now that banks are strictly enforcing the flood insurance purchase requirement for structures in SFHAs, the accuracy of FEMA's mapping is being scrutinized more closely. In many cases, people who feel their homes have been incorrect;y placed in a flood zone contact their elected officials when they realize that resolving the problem requires money for engineering work. Zone A SFHAs pose the most significant concerns to commlmity officials and FEMA because there is generally little or no backup data available for use in establishing BFEs. Several A Zones were mapped in watersheds of less than one square mile and do not pose the significant flood hazards shown on a commlillity's FIRM (in many cases, mapping inaccuracies originated from the scale and contour interval of the original mapping used to delineate the potentially flooded area). FEMA must try to meet the ever-increasing need for more accurate PIS data. The LMMP has proven to be a cost-effective way to utilize improved technical methods to meet that challenge.

PAGE 236

Flood Hazard Mapping of the Bridge Canyon Fan Donald W. Davis Post, Buckley, Schuh & Jernigan, Inc. Gale Wm. Fraser II Clark County Regional Flood Control District INTRODUCTION Detailed alluvial fan flood risk analysis and flood hazard mapping were perfonned as part of the Bridge Canyon Wash Flood Insurance Study (FIS) Restudy. The analysis applied the traditional Federal Emergency Management Agency (FEMA) methodology for flood hazard assessment determination, with modifications to incorporate the constraints and unique features of the fan surface. DESCRIPTION OF AREA The study area lies about 90 miles south and slightly east of Las Vegas near the town of Laughlin, Nevada. Laughlin consists primarily of resorts along the Colorado River, and a small residential and commercial area. There is also a coal-fired power plant and a water treatment facility on the lower portion of the Bridge Canyon Fan. Most of the study area is undeveloped. The Bridge Canyon Wash watershed is approximately 8.0 square miles and is characterized by a canyon wash emerging onto a broad alluvial fan from a desert mountain environment. Inunediately below the fan apex the lateral boundaries are clearly defmed. The southern boundary becomes less and less well defmed downstream. About 2,300 feet below the apex is an incised channel feature. Approximately 10,000 feet below the apex, the incised channel loses distinction and forms a secondary fan on the overall fan surface. The overall fan continues to expand until approximately 19,000 feet down the fan. At this point about half of the fan surface narrows and is directed though a pass toward the Colorado River. The other half is modified by COOling ponds and structures associated with the power plant.

PAGE 237

222 Flood Hazard Mapping of the Bridge Canyon Fan HYDROLOGY When using the alluvial fan method of defining flood hazards, it is standard practice to assrnne that flood events are best described by the Log-Pearson Type m probability distribution. The distribution has three parameters: mean, standard deviation, and skew. In Clark County, the skew coefficient is approximately zero (U.S. Water Resources Council, 1981). The standard deviation for the Bridge Canyon Wash watershed was estimated to be 0.8, based on a relationship of standard deviation and watershed area given in U.S. Army Corps of Engineers (1988). The flood frequency curve used in the alluvial fan analyses is based on the loo-year peak flow rate at the apex of 5,270 cfs developed by Coe and Van Loo Consulting Engineers (1990). Given the skew, standard deviation, and 100-year peak flow rate, the logarithmic mean was estimated using the equations for synthetic statistics given in U.S. Water Resources Council (1981). The mean for the Bridge Canyon Wash apex was estimated to be 1.86. FAN HYDRAULICS General In this study, the traditional approach to flood hazard analysis (Federal Emergency Management Agency, 1990) was used with some modification based on the methodology developed by Michael Baker, Jr., Inc. (1993), French (1992), and Flippin and French (1994). The modification was llsed to take into account that on dissected fan surfaces the potential for existing channels to divert flow is taken into consideration. The probabilistic nature of the original method (Federal Emergency Management Agency, 1990) is preserved and supplemented by topographic data and the results of detailed field investigations. FEMA's FAN progranl (1990) \vas used to compute the contour widths corresponding to flood insurance zone boundaries. The flood frequency data for the apex was applied directly between the apex and the 1344-foot contour elevation. Below this elevation a path analysis methodology was used to accOlmt for geologic constraints. Avulsion Coefficient The standard FEMA methodology requires input of an avulsion coefficient. An avulsion coefficient of 1.0 was used in the probability analysis, which represents no additional increase in the flood hazard probability due to avulsions. An avulsion is defined as the occurrence when, during a single flood event, the flow abandons the path it has been taking and follows a new one. Downstream areas on the fan may be inundated before the avulsion, and other areas of the fan may be

PAGE 238

Davis and Fraser 223 inundated after the avulsion, thus increasing the flood hazard probability of on the fan. The avulsion potential was considered negligible for the following six reasons: (1) A vulsions may be caused by debris blockage of the flow path resulting in a sudden change of course. The upstream watershed is very sparsely vegetated and there is no evidence of vegetal debris accwnulations inhibiting flow paths. (2) A vulsions may be caused by large boulders that could suddenly impede the flow. The materials near the apex are fairly uniformly graded sands and gravels. There is no evidence of boulders near the fan apex. (3) The historical and typical flash flood hydrographs of this area have a very sharp peak of short duration. It is not likely that flows subsequent to the peak would take a different course than the flow path established by the peak. (4) The wash bed upstream of the apex of the fan is wide, with an even bed of loose, previously deposited materials. The easily erodible materials would not likely obstruct the flow of a channel cutting its own path during the peak discharge. (5) The avulsion coefficient greater than 1 may not be appropriate because it increases the probability of all points on the fan; therefore, the avulsion would have to take place at the apex of the fan to be totally justifiable. The fan analysis already accounts for flows assuming a random path down the fan and may also account for flows dividing and spreading into multiple channels. (6) There is an existing wide incised channel feature on the Bridge Canyon Fan, which influenced the use of the Path Analysis Methodology. The method of establishing flow paths that account for geologic constraints different from the ideal fan situation may considered more appropriate than a more arbitrarily derived avulsion coefficient. The naturally occurring geologic features are used to establish narrower limits on the fan surface for which a probabilistic calculated flow is applicable. Flow Regime The FEMA methodology used includes two possible flow regimes occurring on the fan, a single-channel flow and a multiple-channel flow. Typically the upper portion of an alluvial fan has a single-channel zone and the lower portion of a fan forms a multiple-channel zone. The single-channel zone is characterized by flow emanating from the canyon with high energy and erosive power that easily erodes a new channel in previously deposited alluvial materials. The flow may be supercritical or critical, but by FEMA methodology is assumed critical. The already high sediment load is increased further as the channel cuts through loose alluvium. The increasing sediment load and the non-rigid boundary cause the flow to lose energy. At some point the energy is not sufficient to continue in its scour mode and there is an abrupt loss of

PAGE 239

224 Flood Hazard Mapping of the Bridge Canyon Fan energy as the flow, unable to carry its heavy sediment load, changes to a deposition mode. Deposition of materials is the characteristic that fonus the fan. As a depositional feature forms, the flow tends to split. The flow loses momentum and continues to divide as it moves down the fan, continually losing energy and approaching a sheet flow condition. The point where the flow splits is referred to as the bifurcation point. According to the methodology used in the analysis, the area downstream of the bifurcation point is the multiple-channel zone. Flow in the multiple channel zone has a lower energy level, and is assumed to flow at subcritical normal depth. The cumulative effective flow width is 3.8 times wider than the single-channel width. This width ratio is based on the analyses of several well-documented alluvial fan floods and is used by the FAN program (FEMA, 1990). The depth and velocity of flood flows in the multiple-channel region are estimated by using Manning's equation with the friction slope set equal to the slope of the alluvial fan. The characteristics of the two types of flow are important in considering what is most relevant in mapping the project area. Based on aerial photography, topographic maps, and field investigations, the Bridge Canyon Fan seems to have a multiple-channel characteristic beginning near the considered apex. Active flow paths seem to continually divide and lose momentum and eventually lose distinction. The upper portion of the incised channel feature has apparently been filled by deposition. The depositional characteristics, not far downstream of the apex, are more consistent with a multiple-channel regime. Where the fan width is narrow, due to geologic constraints the 100year flow probability has a high depth and high velocity. The probable 100-year flow depth and velocity become less and less as the fan expands. The higher depths and velocities would be very erosive, typical of the single-channel regime. The multiple-channel regime does not become relevant until the lOO-year probable depths are shallower and velocities less erosive. The flood hazard mapping analysis utilized the single-channel approach when the fan was narrow and probable depth and velocities were high. The multiple-channel approach was used when the fan was broad and probable depths were about 1 foot or less and probable velocities were approximately 4 feet per second or less. The incised channel feature caused the single-channel approach to control farther down the fan. Path Analysis Methodology The methodology described in Michael Baker, Jr. (1993) for defining the probability of a given discharge being exceeded between two point.') (Path Analysis Methodology), was applied to areas below the 1344 foot contour elevation. The frequency at which a given discharge is exceeded between two points is a function of the width of a given flow, the width between

PAGE 240

Davis and Fraser 225 the points, the width of the fan, and the frequency at which the flow is exceeded at the apex. At the 1344 contour, the flood flow can potentially follow three paths (Figure 1). The flood frequency curves at the entrance to each path were defined using the Path Analysis Methodology. The frequency curves were defined by using computations to determine flow values for several reccurrence intervals. Given these values, the FAN program was used to develop flood hazard zone boundaries below the path entrance through which the flow passes. Path 1 is associated with the cumulative effect of three breakouts on the north side of a well-defined boundary of the Bridge Canyon Fan. Path 2 is the center portion, which forms into a broad incised channel feature, and Path 3 is a portion of the fan south of the incised channel. The aerial photographs, topographic maps, and field investigations indicate flows associated with Path 3 may spread and enter the Path 2 incised channel area. The ridge, on the south side of the incised channel feature of Path 2, Path 3 Path 4 Through Pass to Colorado River Path 2 Incised Channel Coalescent Fans Figure 1. Bridge Canyon schematic.

PAGE 241

226 Flood Hazard Mapping of the Bridge Canyon Fan prevents flow from spreading out of Path 2, but does not prevent flows from Path 3 spreading into Path 2. Therefore, Path 3 is mapped as coalescing with Path 2. It is assumed that for a given flood event, the flood flow may take Path 2 or Path 3, and the points in the area of coalescence have a probability of being inundated by Path 2 flows or by Path 3 flows. Coalescent Areas The mapping of the study area included mapping of areas of the fan subject to more than one flooding source. Separate flooding sources may include flow from a separate canyon, flow from a separate upstreanl constrained path, and flow concentrated to foml an effectual new fan created by a flood protection stmcture. In alluvial fan areas subject to flooding from more than one flooding source, flood depths and velocities were computed by assuming that the event of inundation by a flood from one source is independent of an event from any other source. In accordance with FEMA guidelines, the lillian of such events, which has a probability of 0.01, was used to define depths and velocities in areas where multiple alluvial fans intersect. The method is described in Michael Baker, Jr. (1994). The probability analysis is related to the fan width of each fan considered independently. Probabilities are calculated for fan widths at various contour elevation intervals and interpolated between intervals. The lower portion of the Bridge Canyon Fan was divided into two additional paths at contour elevation 780. Path 4 represents an area of the fan that tends to narrow, re-collect flow, and follow a path east to the Colorado River. Path 5 represents an area on the fan that tends to spread and be intercepted by cooling ponds associated with the power plant. Ridges in the natural topography and a gravel pit cause a divide between the two paths. As the width of Path 4 decreases through the narrow pass to the river, the probability of inundation by a flood on this portion of the fan also increases. The Path 4 area is susceptible from flooding sources of Path 2 and Path 3. Changes in the downstream probability of a different path do not affect the probability of the path being evaluated; therefore, at a divide, the widths associated with the path not being considered remain constant. For example, to evaluate the flood hazard of Path 4, the portion of the fan widths associated with Path 5, at the divide (elevation 780), were held constant, and added to the portion of the fan widths associated with PaUl 4, which were calculated at each contour interval proceeding up frolll ilie river to the divide.

PAGE 242

Davis and Fraser 227 CONCLUSION Alluvial fan flood hazard mapping was developed by implementing several methods of evaluation and determination of depths and velocity zones on the fan surface. Additional approaches may be incorporated into an analysis when the traditional FEMA approach is not totally applicable. The methodology included consideration of geographical divides on the fan surface, path analysis, narrowing of fan surface, coalescence of several fan areas and flow paths, and flood hazard impacts of a flood control dike, not all of which are fully discussed in this paper. REFERENCES Coe & Van Loo Consulting Engineers 1990 Bridge Canyon Wash Flood Control Master Plan. Laughlin, NV. U.S. Army Corps of Engineers, Los Angeles District 1988 Hydrologic Documentation for Feasibility Study, Las Vegas Wash and Tributaries, Nevada. Federal Emergency Management Agency 1990 Fan: An Alluvial Fan Flooding Computer Program, User's Manual. Flippin, SJ. and French, R.H. 1994 "Comparison of results from alluvial fan design methodology with historical data." ASCE, Journal of Irrigation and Drainage Engineering 120 (1):195-210. French, R.H. 1992 "Design of flood protection for transportation alignments on alluvial fans." ASCE, Journal of Irrigation and Drainage Engineering 118 (2):320-330. Michael Baker Jr., Inc. 1993 Volume 1, Duck Creek Hydrologic Unit, Clark County, Nevada, Flood Insurance Study. 1994 Section 4, Coalescing Alluvial Fans, Draft Training Manual. u.s. Water Resources Council 1981 Guidelines for Determining Flood Flow Frequency. Bulletin 17B. Washington, D.C.

PAGE 243

This page IS intentionally blank

PAGE 244

Part 9 PRECIPITATION, GAGING, FORECASTING, AND WARNING

PAGE 245

This page IS intentionally blank

PAGE 246

Flood Threat Recognition for Tangipahoa, St. Tammany, and Washington Parishes, Southeast Louisiana Mark R. Wingate u.s. Army Corps of Engineers INTRODUCTION Like many areas in southern Louisiana, St. Helena, Livingston, Tangipahoa, St. Tammany, and Washington parishes have a long history of flooding. The state Office of Emergency Preparedness (OEP) teamed up with the U.S. Anny Corps of Engineers, New Orleans District, to reduce the impact of riverine flooding in the parishes by improving the current flood threat recognition system. The Pearl, Bogue Chitto, TcheftIDcte, Tangipahoa, Tickfaw, Bogue Falaya, and Natalbany River basins were scoped for consideration (Figure 1). The study was initiated in 1994 under the federal PI arming Assistance to States (PAS) program, which authorizes the Corps to help states, tribes, local governments, and other non-federal groups prepare plans for the development, utilization, and conservation of water and related land resources. This cost is shared on a 50% federal/50% non-federal basis. A kickoff meeting was held on January 20, 1995, to discuss the study. Participants included the Louisiana OEP (LOEP); OEP directors from Livingston, St. Helena, St. Tammany, Tangipahoa, and Washington parishes; and the New Orleans District of the Corps. The study scope and the level of involvement required from each group was discussed. Each parish was requested to provide the Corps with historical flood infonnation. OEP directors from Livingston and St. Helena parishes responded that improved flood warning was not necessary for the Tickfaw basin based upon current development trends. OEP directors from St. Tammany, Tangipahoa, and Washington parishes responded that accurate real time precipitation and stage data would improve the execution of flood measures currently in place. Based upon the OEP responses, the Corps concentrated its efforts on a flood threat recognition system for St. Tammany, Tangipahoa, and Washington parishes (the three-parish area).

PAGE 247

WASHINGTON PH. \ OJiT I ORLEANS I Figure L Existing and proposed flood threat recognition system. .... a ....... EXIS11NG DIGmAL COWcnON PlA11'ORMS D.C.P. a>-Peart 1Ivw. PearllIvw, La. @-Tonglpahoa IIvw 0 Robert, 1.0. @-Tct..runct. 1Ivw. Pohum, La. a EXISTING NON-DCP SITES (Propo!ed PCP Upgrade) j -Bogue ChIno. BuIll, 1.0. Peart RIver 0 1.0. Natalbany RIver 0 IIapIIst Bogue Chitto near franklinton, La. o rsoPOSt;D Dr.;' S[J'JS 1 Natalbany Riftr near Amite, La. B Tangl '-IIvw @ La. IIvw near Amite, 1.0. Tct..runc:te Rivw near Covingtan, La. E Bogue Chitto near Warnenan, 1.0. F Bogue Falaya near Camp CcrvIngton, 1.0 Bogue PaIaya @ CovIngton, La. I\) c.:I I\) TI 0o a. -i :T .., CD a JJ CD 8 co 0" ::::l ::i" en o c :T CD Ql !tl. r a c en" iii" D>

PAGE 248

Wingate 233 DEVELOPMENT OF THE FLOOD THREAT RECOGNITION SYSTEM Flood Hazard Areas Meetings between LOEP, National Weather Service (NWS), U.S. Geological Survey (USGS), and the Corps were conducted with OEP parish directors and parish officials to recommend improvements to the current flood threat recognition system. The first priority was to identify the flood hazard areas of the Pearl, Bogue Chitto, Tchefuncte, Tangipahoa, Tickfaw, Bogue Falaya, and Natalbany rivers. Meetings with all participating federal and state agencies were held at each parish, along with consultations with each parish OEP director to identify every conummity in the three-parish area with a history of flooding. In addition, participants identified potential sites for using stage and precipitation data. Existing Flood Threat Recognition System The flood threat recognition system is critical in providing timely, accurate, and reliable information to federal, state, and local officials, and others. Flood threat recognition is an essential step in issuing warnings and insuring that emergency response (road closures, search and rescue, etc.) is timely and appropriate. Rivers in the three-parish area have relatively short reaches and rapid response times due to intense rainfall conm10n along the Gulf coast. For theoe reasons, warning time and public reaction to flood and flash flood watches and warnings are critical in reducing the impact of flooding on human lives and property. NWS, located in Slidell, Louisiana, is authorized by Congress to disseminate flood forecasting. Ideally, NWS forecasts should be as timely, accurate, and reliable as current technology allows. To provide for reliable forecasts, a gauging network (stage and/or precipitation gauges) was installed throughout the three-parish area. In some places, state-of-the-art automated precipitation and stage gauges with satellite telemetry exist, and in others, manually read staff gauges are used. NWS uses data from both types of gauges to prepare flood forecast infonnation. Currently, the forecast information is based upon information collected along the Pearl, Bogue Chitto, Tangipahoa, and Tchefuncte rivers at the seven sites shown on Figure 1. The current network provides real time data to NWS and the USGS for three of the seven locations, but no real time data is provided the OEP parish directors. In fact, during floods, the departmental personnel must manually read staff gauges along the major rivers. Currently, NWS makes river forecasts for the Pearl and Bogue Chitto rivers based upon sophisticated computer modeling that provides a timely forecast with high accuracy. Based upon limited hydraulic and hydrologic

PAGE 249

234 Flood Threat Recognition in Southeast Louisiana data, river forecasts for the Bogue Falaya, Tchefuncte, and Tangipahoa rivers are based upon flood forecast tables that can be used to forecast the maximwn river stage and time to crest. The forecast is based upon NWS's daily flash flood guidance, which is defined as inches of rainfall for a given duration required to produce flash flooding. This provides a valid forecast, but is not as accurate as computer modeling forecasts. Currently, river forecasts are not given for the Natalbany River due to a lack of stage data. Parish OEP Director Perception Each parish OEP director suggested that their flood warnings could be improved if the existing gauging network and retrieval of information were modified. Suggested modifications included fully automating the existing flood threat recognition system, and adding automated gauges in new locations. Design modifications to the current system should also provide each OEP director with real time stage and precipitation data at the parish level. Information retrieval should be designed to eliminate the need to manually gather data during a storm. Design Modifications The modified flood threat recognition system is the result of a joint effort between LOEP, parish OEP directors, parish officials, NWS, the Corps, USGS, and other interested groups. The first design criterion was to incorporate the latest technology of automated data collection to provide real time stage and precipitation data to each parish OEP director, NWS, and USGS. The second criterion was to position gauges so that forecasts could be provided for each flood hazard area throughout the three-parish area while maximizing reliability, accuracy, and warning time. In order to meet the design criteria, two types of monitoring networks were considered: 1) an automated local evaluation in real time (ALERT) network and 2) a digital collection platfornl (DCP) network. Both types are feasible choices for real time data collection. From the two systems, USGS, NWS, and the Corps recommended the DCP network over the ALERT network primarily on past performances of both networks in southern Louisiana. The fact that the ALERT system operates on very high frequency (VHF) radio waves and requires numerous radio repeaters was the primary technical reason for choosing the DCP network. The DCP network transmits signals to the Geostationary Orbit Environmental Satellite (GOES8), which in turn relays the signal to the appropriate base stations. In addition, the USGS, LOEP, and the Corps plan to develop a statewide DCP network.

PAGE 250

Wingate 235 The proposed network consists of 14 DCP sites to collect stage and precipitation data along six rivers in the three-parish area (see Figure 1). The proposed network places upgraded equipment at seven sites where data is currently collected by DCP and non-DCP equipment, and seven sites where stage and precipitation data has never been collected. Each of the 14 gauges will be programmed for random and redundant transmissions thereby providing real time data. Each DCP also will be outfitted with phone modems and two complete sets of spare DCP equipment will be obtained for redundant measures. The total cost of the proposed network is estimated at $190,000. Annual operation and maintenance costs are estimated at $40,000. l1rree alternatives were considered to provide real time data to each OEP parish director, NWS, and USGS. Each considered the fact that NWS and USGS have the capability to receive real time data from the DCPs via the GOES8 satellite. However, the OEP parish offices cannot justify this capability, thus various alternatives were considered for providing real time data to each OEP parish director. The first alternative called for a HydroMet base station at each OEP parish office that would be tied directly to each DCP via phone modem to receive real time data. The HydroMet base station allows the user to view stage and precipitation data in either text or graphic format, as well as print, and archive retrieved data. The base station also provides a redundant means for NWS and USGS to receive data. This alternative requires that each parish dedicate a high-end computer as the base station. This alternative is estimated to cost $10,000 per parish plus costs of running phone lines to each gauge. For the second alternative, data retrieval at NWS and USGS would be by the same means, but retrieval at the parish OEP would be via the Internet. A HydroMet base station was not suggested as part of this alternative to lower costs. This alternative requires that the user have a DOSbased personal computer with modem and an Internet provider. This would allow the user to view stage and precipitation data in a limited graphic and text format. However, this is a very slow and unreliable means to receive "near" real time data. The cost of this alternative is estimated at $3,000 per parish. For the third option, data retrieval at NWS and USGS again would be the same, but retrieval at the parish OEP would be via modem to the USGS local area network, and would not utilize a HydroMet base station at the parish. TIlis alternative requires a DOS-based personal computer with modem, which would provide viewing capabilities. However, USGS limits the number of phone lines into its local network, and the parish Would need training on software and operating systems used by USGS. The cost of this alternative is estimated at $3,000 per parish.

PAGE 251

236 Flood Threat Recognition in Southeast Louisiana NETWORK RECOMMENDATIONS The participating groups discussed the proposed DCP network and the advantages and disadvantages of each alternative for USGS, NWS, and each OEP parish director to receive real time stage and precipitation data. Each group agreed with the selected gauging sites, which were a function of the identified flood hazard areas. Each group also agreed that real time data must be transmitted to each OEP parish office in the most reliable means available. Therefore, the alternative that specified installing phone lines directly from each OEP parish office to each DCP was selected. CONCLUSIONS The proposed network will enable the NWS to provide forecasts based upon numerical models for the Pearl, Bogue Chitto, Tangipahoa, Tchefuncte, and Bogue Falaya rivers. These river forecasts are expected to be timely, reliable, and accurate in comparison to the current forecasting. River forecasts will also be provided for the Natalbany River. Each OEP director will benefit by receiving real time stage and precipitation data at the computer base station in each parish. This will enable each director to react in a timely and proper fashion to the existing flood threat based upon real time data in lieu of current data collection that is obtained via untimely faxes. This process will eliminate the need for parish personnel to gather stage and precipitation data by manually inspecting each gauge during a flood threat condition. Under the new operating system, these personnel will be able to carry out other flood fight activities in lieu of their current inspections. In terms of funding, USGS offered to cost share, on a 50/50 basis, all annual operation and maintenance costs with each parish. LOEP offered to provide funding for installation contingent upon a commitment from each parish to cost share in the annual operation and maintenance costs. The Corps offered to provide flmding for installation contingent upon developing a favorable benefit-to-cost ratio via an additional study lmder the federal program Continuing Authorities Section 205. The completion of this study recognizes that the vast participation of groups including LOEP, OEP parish directors, parish officials, parish personnel, NWS, USGS, and the Corps was invaluable. Numerous meetings were held between federal, state, and parish officials and personnel, and all participants listened and responded to the needs of each OEP parish director. The system design is primarily based upon each OEP director's input. Every participant played an integral part in the overall design. Given this type of team effort, system acceptance, use, operation, and maintenance was never a critical issue.

PAGE 252

Guidelines for Developing Comprehensive Flood Warning Laurie T. Miller Montgomery Watson Tom Donaldson Lower Colorado River Authority Dallas Reigle Salt River Project Jesu s Romero Yavapai County Flood Control District Stephen D. Waters Flood Control District of Maricopa County Patricia Q. Deschamps Geo. V. Sabol Consulting Engineers Sam A. Arrowood u.S. Army Corps of Engineers Wayne Cooley Arizona Department of Water Resources INTRODUCTION In recent years, there has been much progress in the development and implementation of local flood warning programs as a viable means of nonstn!ctural flood control. Existing programs across the country offer a range of services and cover a variety of areas from very small to entire states. Despite vast differences in program components, there is one common frustration: the difficulty of progressing beyond collecting and monitoring data to actually removing people and property from a flood

PAGE 253

238 Guidelines for Developing Comprehensive Flood threat. These guidelines are intended to present the total commitment required to provide comprehensive flood warning services and offer suggestions on how to develop a customized, comprehensive program. FLOOD WARNING PROGRAM ELEMENTS A complete flood warning plan includes the development and coordination of three basic elements: (1) detection and evaluation of a flood threat; (2) dissemination of warnings; and (3) response to the warnings. In addition, successful flood warning requires coordination among federal, state, and local government agencies. Major components that must be addressed are organized in accordance with the credit evaluation criteria for Activity 610, Flood Warning, under the National Flood Insurance Program's (NFIP) Community Rating System (CRS). FLOOD THREAT RECOGNITION A flood threat recognition system (FTR) is any system used to identify flood threat. It can be as simple as 24-hour monitoring of NOAA Weather Radio, or can be a complex system of hardware and software that delivers real-time data to many locations. It is necessary to first identify local flooding characteristics so that appropriate equipment can be selected. Information should include: (1) a good physical description of the watershed; (2) the type(s) of flooding that occur, flash or riverine; and (3) maps of the areas affected by flooding. Needed/Avai lable Lead Time The amount of lead time needed greatly influences the type of system required. The available lead time for an area may be determined by hydrologic and hydraulic studies of observed records, supplemented b:; rainfall-runoff analysis of observed and hypothetical frequency events. Selection of Appropriate System Components Tools vary from simple to sophisticated: volunteer observers, automated precipitation and stage gages, base station hardware and software, radar and satellite data, meteorological support, and aids like maps, graphs, and computer models. Communications There are three types of data transmission communications available on the market today: telephone, radio, and satellite. The type(s) used will be determined by the characteristics such as topography, availability of equipment and funds, and lead time. Redundancy is always desirable and

PAGE 254

Miller, Donaldson, Reigle, Romero, Waters, Descharrps, Arrowood, Cooley 239 can be achieved by combining any of the data transmission methods. At very critical gage sites, it may be wise to install two sets of gage equipment (transmitters, sensors, batteries) and receive data from both. WARNING DISSEMINATION Dissemination of flood threat information is "getting the word out" before a flood occurs to reduce the risk to life and property. It includes notifying emergency management, public works, and other essential personnel so that preventative steps may be taken to minimize the impacts of flooding. Before warnings can be issued, information pathways to the end-users of the warnings must be identified and optimized. The process consists of three primary functions: (1) deciding whether to issue a warning (usually detennined by preset criteria), (2) formulating the warning message, and (3) identifying the appropriate audience and means (radio, television, sirens, bullhorns, and door-to-door) of distributing the warning message. The primary government agency responsible for flood warnings is the National Weather Service (NWS). Existing local flood warning agencies rely heavily on interaction with the NWS for disseminating warnings to the general public. FIR and warning data are generally shared with the NWS. Local agencies may provide FIR and warning data to state and local emergency management and public safety agencies. Public Education Public education should be part of any warning dissemination program. Typical elements include public service messages, videos, pamphlets, and children's materials. Materials can be distributed through schools, libraries, conmnmity centers, government offices, and special events like fairs. Just before the flood season is a good time for a public education campaign. EMERGENCY RESPONSE The conununity's emergency action plan (EAP) is its response to a flood threat. The goals of the existing EAP and flood warning program should be compared and the EAP modified if necessary. Flood hazards should be identified as well as any operational or response constraints, such as short response times, access to certain areas during flooding, or long distances between emergency resources and flood hazard. Specific flood hazards should be inventoried and the warning methods established. Lines of communication and actions to reach the warning program goals should be identified. Once conununication needs are established, a detailed plan can be developed to include names, telephone numbers, and duties of the appropriate staff, as well as methods of communication.

PAGE 255

240 Guidelines for Developing Comprehensive Flood Warning Maintenance of an EAP Emergency response requires periodic maintenance to verify that the components will work in a real emergency. Practice drills must be held at least annually when no significant flooding occurs. The EAP should be updated at least annually to include any changes in staff, telephone numbers, and responsibilities. After a drill is completed or flood occurs, it is important to hold debriefings and implement any necessary changes. OTHER RESPONSE EFFORTS Other response efforts (ORE) are efforts in a community's flood response plan that are not specifically tied to the flood warning program, but would significantly benefit flood fighting efforts in the event of a flood. Each major task should be assigned to an office or individual. In large organizations, an individual should be identified as the one responsible for communication with other departments, as well as carrying out the task. Summary Comparison of Resources For each task in the flood response plan, it is very helpful if a summary comparison of resources is kept on file. Data to be collected include a list of what resources are needed to complete each task, the time required to perform the task, and the source(s) available to complete each task. CRITICAL FACILITIES PLANNING Critical facilities planning is coordinating with facilities with special needs or that require special attention during a flood. They include police and fire stations, hazardous materials storage, public and private utilities, hospitals, nursing homes, and schools. It is important to identify critical facilities in order to provide timely evacuation. Obviously, it is important to maintain an up-to-date, accurate list of individuals to contact in case of an emergency, including names and phone numbers of back-up personnel. MAINTENANCE A commitment to regular maintenance is required for the successful operation of any flood warning progranl. Maintenance must be performed to minimize the occurrence of equipment failures during flood emergencies. Any gages and base stations should be checked daily for proper operation. A preventative maintenance schedule should be devised that will ensure proper operation of the gage during a flooding situation. Service maintenance contracts can provide some of the needed preventative maintenance. There should be enough spare parts readily available to create or repair one complete remote site and any radio-relay sites. The items that require

PAGE 256

Miller, Donaldson, Reigle, Romero, Waters, Descharrps, Arrowood, Cooley 241 replacement more frequently should be stocked in larger quantities. Standardization of components will reduce the size of spare parts inventory. COST CONSIDERATIONS Flood warning has been shown to be an inexpensive alternative when compared to structural solutions to flood threat. However, it should be recognized that there are significant start-up costs to implement a flood warning program. These costs vary widely according to the needs, size, and type of flood threat of the individual community. Initial Costs The first cost is to develop a comprehensive plan to evaluate the community's needs versus resources, to design a suitable system, and to develop funding strategies for implementation. Once the plan is formulated, costs will be incurred to purchase equipment and spare parts, install the system, provide training, obtain hardware/software technical services, establish conununications links, and develop decision-making tools. Finally, permit and/or licensing fees will most likely be required for field and base station equipment. Annual Costs Annual costs are incurred to operate, maintain, and upgrade the Additionally, required updates and/or improvements to any component(s) should be identified as the system is used and tested during simulated or actual flood emergencies. Event-Driven Costs During a flood, additional costs will be incurred to monitor the flood threat and provide technical support to emergency services personnel. Potential Sources of Funding and Technical Assistance Fllllding is nearly always through cost-share agreements where the local community must nmd a portion of the costs and also agree to operate and maintain the system once it is installed. Fllllding and/or technical support at the federal level can be obtained from the U.S. Arnly Corps of Engineers, NWS, and the Natural Resources Conservation Service. Support on the state and local levels varies with location, but typically includes state departments such as water resources and emergency management, county flood control districts, and cities with established flood warning programs. Maintenance agreements may also be available.

PAGE 257

242 Guidelines for Developing Comprehensive Flood Warning Teclmical assistance can also be obtained from private consultants and from professional organizations. Organizations such as of the Southwest Association of ALERT Systems (SAAS), the California ALERT Users Group (AUG), and the Arizona Floodplain Management Association (AFMA) are excellent resources. PERMITS Installation of flood warning equipment will likely require a permit to allow permission to install, maintain, and operate a flood detection station. Agencies installing flood detection stations should seek legally binding permits because they guarantee long-term use of and access to the site. Types of Permits A land use pern1it is granted by private property owners or by agencies. If equipment is located within a designated floodplain, then a floodplain development or use permit may be required. When applying, it is helpful to provide a brief description of the overall flood warning system and its purpose, a detailed drawing or picture of the equipment being installed and a description of its function, a map of the station location and access routes, the expected length of time the equipment will be in place, and any expected operation and maintenance activities, their duration, and their frequency. Another permit is the licensing of radio equipment in the system, and the assignment of a radio frequency(ies). Licenses and frequencies are granted by the Federal Conu11lmications Conunissioll through a federal sponsor. Time Requirements The permit process can take a year or more for some federal agencies. A government agency may require inspection by utilities, an archeologist, a botanist, and/or an environmental engineer. SUMMARY A complete flood warning plan includes the recognition of a flood threat, dissemination of warnings, and response to those warnings. It is hoped that those considering flood warning might have a better understanding of the steps involved in implementing a system, and those already involved may discover some ways to improve their existing system in order to meet the goal of saving lives and property though flood warning. For a complete copy of these guidelines or for more information, please contact Laurie Miller at (602)954-6781.

PAGE 258

Spatially Distributed Rainfall: the Use of Volunteer Gaging Richard J. Heggen University of New Mexico Clifford E. Anderson Smith Engineering Steve Hemphill Jemez Engineering BACKGROUND Mean annual precipitation in Albuquerque, New Mexico, is 300 tnm. Winter precipitation, generally derived from frontal disturbances, tends to be protracted and of mild intensity. Summer precipitation, typically convective with orographic accentuation, is of short duration and higher rate. Runoff is ephemeral. Sununer precipitation constitutes the basis for flood design. In 1980, the Albuquerque Metropolitan Arroyo Flood Control Authority (AMAFCA) and the National Weather Service, Albuquerque Office (NWS), initiated a program of precipitation recording done by volunteers. Volunteers are solicited through personal contacts by both agencies. The NWS furnishes AMAFCA with pads of standard daily reporting forms, a sample form format, and prepaid-postage return envelopes. AMAFCA purchases inexpensive plastic raingages, distributes the materials, assigns each volunteer an identification number, and determines the gage location on a vicinity map. Approximately 60 active volunteers are spread over roughly 100 square miles of urban area. The volunteers read the gages and record the precipitation at the same time every day, preferably in the morning. When the volunteers are away, they are asked to find a substitute. At the end of each month, the volunteers mail the precipitation forms to NWS, with a copy retained for their records. The volunteer data supplements the official 90-year rainfall record at Albuquerque International Airport. Volunteer-derived data is not sought to alter NOAA estimates, the latter having the benefit of larger sample size and advanced meteorological analyses. Volunteer-derived data provides synoptic

PAGE 259

244 Spatially Distributed Rainfall-the Use of Volunteer Gaging stonn descriptions fitting within the broader climatological monitoring. Voltmteer monitoring and inunediate input on severe rainfall events is of great significance to AMAFCA's planning and response responsibilities. With only a few recording raingages in the Albuquerque area, volunteers provide the primary source of the magnitude and spatial distribution of many convective stonns. The volunteers inunediately report precipitation events of over 0.25 inch in one hour or 0.50 inch in one day by calling in their identification number, the time, and amount. After a severe event, NWS frequently prepares an isopluvial map faxed to AMAFCA and other local agencies. The map is used to identify potential problem areas and to facilitate perfornlance evaluation, cleanup, and maintenance. In 1980 and 1988, severe stornlS occurred within urban Albuquerque. The phone reports and the monthly written reports established that portions of the city received substantially greater than a 100-year storm. As the hydraulic capacity of some constructed facilities was exceeded, it was useful to know that the design hydrology was also exceeded. Albuquerque, Bernalillo County, and AMAFCA are all participants in the National Pollution Discharge Evaluation System (NPDES) pennitting process. A major task of the NPDES pennitting is the monitoring of contaminants in stonnwater runoff. It is necessary to know the threshold of rainfall for measurable runoff. The source of rainfall for particular runoff events must be identified. The data provided by the vohmteers is an essential part of this monitoring. As with any endeavor that requires volunteer cooperation over an extended period, consistent participation is problematic. AMAFCA regularly sends newsletters to the volunteers infornling them of the value of their contribution. For some volunteers, the routine of regular observation is difficult to maintain; their primary participation is in reporting of severe events. Some volunteers have been exceedingly faithful for as long as 16 years. THE BASIC QUESTION A volunteer network is prone to a myriad of hunlan and technical errors, erroneous gages, misreadings, or sloppy recordkeeping. From the floodplain management perspective, a basic question arises. "Can rainfall data derived from a volunteer network be statistically valid for long-tenn meteorological assessment?" Or, "Does rainfall data derived from a volunteer network have more than anecdotal value in floodplain planning?" This paper sunmlarizes initial findings from 20 of Albuquerque's volunteer stations. A 63-year professionally recorded daily history from Los Alamos, New Mexico, is used as a benchmark of the results.

PAGE 260

Heggen, Anderson, and Hemphl'll 245 SPATIAL DISTRIBUTION The spatial distribution of convective stonns is well documented in only a few high-density instrumented experimental watersheds, such as Walnut Gulch, Arizona. Albuquerque's less-sophisticated volunteer network reveals similar distributed behavior. Consistent with general Southwestern experience, convective stonns are localized, concentrating rainfall in a few square miles. Only about five stonns per year are simultaneously noted at more than one Albuquerque volunteer station. Of 90 multi-station events, in only 17 was the second-highest measurement within 10% of the highest. In only 5 events was the third-highest observation also within that limit. The lack of depth persistence over distance clearly reveals the error in presuming the even application of a reported value over an entire watershed. As not all stations were active in all years, such a record should be seen as an illustration, not proof of storm pattern. As more stornlS are simultaneously monitored, geostatistical tools are available to quantitatively strengthen the spatial conclusions. Lack of station-to-station correlation for a given storm causes the station records to be statistically independent data samples with respect to time. The statistical implication of time-independence between gages is very much to the point of flood management. Albuquerque's storm history reveals itself to be a nunlber (yet undeternlined) of proximate, but independent, rainfall zones, having histories similar in overall rainfall, but only weakly correlated on the daily calendar. While the zones share SU0101ary statistics, given some orographic adjustment, each has its own history. If there are five zones, as an illustrative mmlber, Albuquerque overall should have five 100-year stornlS in a typical century. This understanding, demonstrated by the volunteer network, answers the frequent comment, "Why has Albuquerque had so many 1 OO-year stonns recently?" In many cases, the volunteer record docunlents storms more localized than those described by the standard NOAA area-reduction factors. Such spatial specification improves hydrologic model calibration, a successful element in the Federal Emergency Management Agency's approval of Albuquerque's runoff model AHYMO for floodplain mapping. THE HYETOGRAPH The volunteer record docunlents stonns more intense than the conventional NOAA one-hour percentages. A shortcoming in the volunteer network is the sparsity of measurements at less than a 24-hour increment. Lacking recording gages, volunteers cannot chart the hyetograph. As runoff modeling in Albuquerque is done with time steps

PAGE 261

246 Spatially Distributed Rainfall-the Use of Volunteer Gaging of five minutes or less, a corresponding time-step rainfall record would be invaluable. Likewise the volunteer network lacks ability to track storm movement across the landscape. As the runoff peak from a storm moving down a basin can be greater than the same storm moving up the basin, timing would be valuable data. Lack of recording gages, however, does not preclude storm sleuthing from the volunteer's filed notes. Comments such as "Most between 3 and 3:30 PM" assist in storm reconstruction. RETURN PERIODS The authors developed estimates for rainfall return periods for 8 volunteer stations having from 118 to 1314 days of record. The procedure looks at the distribution of the Sunl of all rainfall events equal to or greater than a given magnitude. The number of events is the ordinate on a semi-log plot. The magnitude of the event is raised to a power, typically around 2/3, to minimize the mean-squared-error between the observed magnitudes and the predicted magnitudes. Presumably this corrects for natural distribution and the orographic effects. Analysis by spreadsheet is relatively simple. Figure 1 illustrates the fit for a representative station. 3.00 CD o ,. va... .9-2.00 Q. '" 3 c 1.00 ]' 0.00 0.00 0.50 0 1>0 0 1.00 Depth (inches) An 0 0 0 1.50 Figure 1. Semi-log recurrence plot, station 9. v 2.00 The typical correlation between predicted and observed distribution exceeds 0.99. The curve fit shows intuitive graphical visual confirmation with relatively few data points, exhibiting none of the tail spin-out typical

PAGE 262

Heggen, Anderson, and Hemphill 247 of alternate models. The data below summarizes 24-hour results for Albuquerque and Los Alamos. For log(y) = m(x"n)+b, n: m: b: Standard Error: Correlation CocCficienL: Return Interval Depth, inches: 2 year: 5 year: 10 year: 100 year: 500 year: 1 ,000 year:: 10,000 year: Number of Observations> 0: Total Number of Observations: Albuquerque 0.753 -l.614 3.846 0.00085 0.9998 l.63 2.00 2.30 3.30 4.16 4.52 5.75 6,210 24,126 Los Alamos 0.753 -1.597 2.731 0.000652 0.9723 l.31 1.84 2.28 4.02 5.48 6.17 8.69 339 1,339 The combined volunteer stations show significantly more rare-event rainfall than that reported by NOAA, analysis benefiting from neighboring rainfall histories. The NOAA atlas 100-year airport depth is 2.6 inches. The airport official record evaluated in the semi-log manner above yields 2.08 inches. Using only the years of the volunteer network operation, the airport 100-year storn1 is 2.68 inches. The discrepancy appears to be a consequence of several major stonns lmevenly located over the area within the san1pling period. The lmexpectedly high correlations indicate that the data represents a statistically well-behaved natural phenomenon. The data is not corrupted by the vohmteer system. As a complement to NOAA data, a conm1unity can make statistically-defensible use of vohmteer rainfall reporting. STATISTICAL DISTRIBUTION The high degree of correlation suggests the possibility of a fundamental underlying statistical distribution for rainfall depth. A distribution commonly suggested for rare-event hydrologic data is a two-parameter Gamma flmction. A Garrm1a distribution can accOlmt for a data set containing a large proportion of zero values, a familiar aspect of Southwestern daily rainfall. Figure 2 illustrates the fit for a typical Albuquerque volunteer station.

PAGE 263

248 Spatially Distributed Rainfall-the Use of Volunteer Gaging 1.00-----.1 -.... .... 2.-.... ..... ---. _-.. c::J ... .......................................... J7 '[ 0.801-1 \/I .9-I II 0.7011Ilf----+------1-----l-----t-----1 a.. 0.60-l-----+----1-----t------'-------I --Record ...... Gamma 0.50 0.0 0.5 1.0 1.5 2.0 Depth (inches) Figure 2. Gamma function fit, station 140. The Ganuna function parameters for the eight stations are: Sta. 3 6 9 17 27 49 130 140 Lambda 1.64 1.40 1.81 1.45 1.99 1.75 1.66 2.01 Alpha 0.1001 0.0908 0.1002 0.0957 0.0989 0.1263 0.0924 0.1236 2.5 The proximity of alpha, a dimensionless shape factor, to a mean value suggests that the several stations, while exhibiting different rainfall traces, may all fall within the same population of rare event. The basis for the particular parametric values is not yet understood. Significant for this paper, volunteer-derived data reveals an underlying behavior. Volunteer-derived data again appears to be statistically reliable.

PAGE 264

Heggen, Anderson, and Hemphill 249 CONCLUSION The results illustrate insight gained through a relatively unsophisticated data acquisition program. Albuquerque's network of volunteer rainfall reporting provides a data set of remarkable statistical significance. Such a data set may be of use in establishing rainfall-prediction equations or refining the capability of existing estimates, generally based on multiple-year data strings at single NOAA stations. Volunteer rain gaging is an inexpensive, relatively quick method for it community to evaluate rainfall patterns and pursue appropriate design hydrology. With relatively little capital expense, a flood management agency can incorporate willing citizens into its data collection system and achieve analytically defensible results. REFERENCES McCuen, Richard and Willard Snyder 1986 Hydrologic Modeling, Statistical Methods and Applications. Englewood Cliffs, NJ: Prentice HaIl Miller, J.F., Frederick, R.H., and R.J. Tracey 1973 Precipitation-Frequency Atlas of the Western United States, Volume IV-New Mexico. NOAA Atlas 2. Washington, D.C.: U.S. Department of Commerce, National Oceanic and Atmospheric Administration, National Weather Service. Wright Water Engineers, Inc. 1989 Assessment of the July 9, 1988 Storm and the City of Albuquerque Flood Control System. Albuquerque, NM.

PAGE 265

New Preci pitation Frequency Studies for the United States Lesley T. Julian John L. Vogel National Weather Service INTRODUCTION Design of hydraulic structures and the management of water resources requires rainfall frequency analyses and depth-are a-duration curves to determine design storms and storm water runoff. The Hydrometeorological Branch of the National Weather Service, Office of Hydrology, will publish the Semiarid Precipitation Frequency Study (Semiarid Study) as Volume 1 of NOAA Atlas 14, Precipitation-frequency Atlas of the United States, in 1997. Each state in the semiarid region: Arizona, Nevada, New Mexico, Utah, and southeastern California, will have a separate document, numbered 1.1 to 1.5. The Semiarid Study will supersede the previous atlases, NOAA Atlas 2 (NA2) (Miller et aI., 1973) and Technical Paper 49 (TP49) (Miller, 1964) for these western states. Precipitation frequencies are provided for events as frequent as six times a year and up to 100-year return periods for durations from 5 minutes to 60 days. For durations of 24 hours and longer return frequencies up to 1000 years will be available. New depth-are a-duration curves have been developed specific to the Southwest. The Semiarid Study differs from the earlier studies in the following: 1) 230 supplemental stations, longer periods of record, and 30% more daily stations; 2) new statistical methods that permit more objective quality control, regionalization of data, and objective curve fitting techniques; and 3) direct use of partial-duration data series. An important addition to the atlas for some states is seasonality of extreme events. The seasons of extreme precipitation vary widely within the Semiarid Study area. The Semiarid Study will serve as a prototype in the process of updating frequency studies for design storms over the entire United States. The "current" atlas for the midwestern and eastern United States, Technical Paper 40 (TP40) was released 35 years ago (Hershfield, 1961). The Hydrometeorological Branch is currently preparing studies for

PAGE 266

Julian and Vogel Hawaii and Puerto Rico, and is working with the state of Alaska to determine their needs. SEMIARID STUDY 251 The Semiarid Study area is shown in Figure 1. Boundaries of 24 climatic regions, based on seasonality, topography, synoptic climatology, and other extreme rainfall characteristics are shown. Two rainfall seasons, warm and cool, were determined and the months of the seasons also are given. Warm-season rainfall is usually characterized by thunderstomls and other intense, short-duration rains, and cool-season precipitation is primarily from general storms of longer duration. DATA One of the most important aspects of any study detemlining the return frequencies of phenomena is the database. It is especially important to have high-quality data from as many long-term stations as possible. For the Semiarid Study, the primary source of rain gage records was the cooperative network of daily and hourly stations. In addition, other federal, state, and local records were sought and found throughout the region, thus providing a total of 743 daily stations with records of 19 years or more, and 207 hourly stations with record lengths of 15 years or more. An additional 230 supplementary stations with records of 10 to 15 years were obtained for remote locations, all of which are daily reporting stations. There were 122 supplementary stations from the SNOTEL (SNOw TELemetry) stations operated by the National Resources Conservation Service (formerly the Soil Conservation Service), and 108 supplementary stations from Mexico. In the Uinta Mountains in northeast Utah, the SNOTEL stations represent the only infomlation available. The period of record is not long, but these supplementary stations provide infomlation in regions where no or only limited data are otherwise available. ANALYSIS Statistics According to Hosking and Wallis (1991), the analysis procedure for defining regional frequency analysis consists of four parts: 1) perfoffil quality control on data to eliminate gross errors and inconsistencies; 2) identify homogeneous regions, so that sites within a region have approximately the same frequency distribution; 3) define a regional frequency distribution; and 4) evaluate the regional frequency distribution.

PAGE 267

Figure 1. Semiarid Study ciimatic regions. I\) (11 I\) z '"C is' a 0' -. CD .J:l c: CD .Q g? c: a. (D' CII 0' -. =r-CD C ;:::0: CD a. g? CD CII

PAGE 268

Julian and Vogel 253 These steps have been greatly simplified by the introduction of L moment statistics (Wallis, 1989; Hosking, 1990; and Hosking and Wallis, 1990). The L in L-moments stands for linear, so that L-moment statistics are linear combinations of ordered (ranked) statistics. The theoretical advantages of such statistics are the abilities to: 1) characterize a wide range of distributions; 2) produce a robust technique of handling outliers in the data sample; 3) provide a means of performing regional analysis, which is more robust than single-station analysis; and 4) maximize the utility of those stations which do not have many years of record. For the Semiarid Study, a partial-duration series and the Generalized Pareto distribution were used for L-moment analysis of the precipitation data in each near-homogeneous region. The choice of distribution was made as a result of curve-fitting tests within the L-moment software and real data comparisons with theoretical distributions. The L-moment analysis provided Regional Growth Factors (RGFs), which are used to define the return frequencies for each station. Mapping TIle mapping and analysis process is a combined hand-analysis and computer mapping technique that: 1) develops an index map, 2) determines its relation to other durations and/or return frequencies, and 3) uses the computer to do the ari tlunetic to generate other maps of interest. The 2-year, 24-hour map (index map) was hand-analyzed from exactingly quality-controlled data, and return-frequency values computed using L-moment statistical software over near-homogeneous climatic regions. The index map is multiplied by the appropriate regional growth f
PAGE 269

254 New Precipitation Frequency Studies for the United States 2-year 24-hour INDEX MAP 1 OO-yr 24-hr RGFs X------2-yr 24-hr RGFs 1-hr Ratios 24-hr 2-year 1-hour map 100-yr 1-hr RGFs x------2-yr 1-hr RGFs 1 DO-year 24-hour map 1 DO-year 1-hour map Figure 2. Flow chart of mapping procedure. of various durations, and other return frequencies up to 100 years for 1 hour and 24 hours only; and seasonal 2-year maps for 1-, 6-, and 24-hour durations). Smaller scale ratio maps will fill in intermediate durations; and all other return frequencies will be computed from the spreadsheet described above. It is planned to also put NOAA Atlas 14 on an interactive CD-ROM. Engineers, planners, water-resource managers, and others use point probabilities and depth-area-duration (DAD) curves to develop a design storm and calculate potential stom1water runoff. TP40, TP49, and NA2 supply a set of DAD curves based on data from 20 dense networks of rain gages concentrated in the East, Midwest, and along the West Coast of the United States (U.S. Weather Bureau, 1957-1960). No networks were located in the semiarid area shown in Figure 1. Thus, new DAD curves based on southwestern storms are being developed and will be given in NOAA Atlas 14. Information on the temporal distribution of precipitation within storms will also be included in the final report.

PAGE 270

Julian and Vogel 255 REFERENCES Hershfield, D.M. 1961 Rainfall Frequency Atlas of the United States for Durations from 30 Minutes to 24 Hours and Return Periods from 1 to 100 Years. Weather Bureau Technical Paper No. 40. Washington, D.C.: U.S. Weather Bureau. Hosking, J.RM. 1990 ''L-Moment: Analysis and Estimation of Distributions using Linear Combinations of Order Statistics." Journal of the Royal Statistical Society B 52 (1): 105-124. Hosking, J.RM., and J.R. Wallis 1990 Regional Flood Frequency Analysis Using L-Moments. Research Report RC 15658. Yorktown Heights, N.Y.: mM Research Division. Hosking, l.R.M. and l.R. Wallis 1991 Some Statistics Useful in Regional Frequency Analysis. Research Report, RC 17096. Yorktown Heights, N.Y.: mM Research Division. Miller, J.F. 1964 Two-to Ten-Day Precipitation for Return Periods of 2 to 100 Years in the Contiguolls United States. Technical Paper No. 49. Washington, D.C.: U.S. Weather Bureau. Miller, J.F., RH. Frederick, and R.J. Tracey 1973 Precipitation-frequency Atlas of the Western United States. NOAA Atlas 2. Silver Spring, MD: National Weather Service. u.S. Weather Bureau 1957-1960 Rainfall Intensity-Frequency Regime, Parts 1-5. Technical Paper No. 29. Washington, D.C.: U.S. Department of Commerce. Wallis, J. R. 1989 Regional Frequency Studies using L-Moments. Research Report RC 14597. Yorktown Heights, N.Y.: mM Research Division.

PAGE 271

The Big 1995 Floods in Northern California Maurice Roos California Department of Water Resources INTRODUCTION Twice during 1995 a series of winter stornlS caused severe flood damage in many areas of California. Precipitation was three times nOmlal in both January and March. In between was sandwiched a very dry February. Although flood losses were substantial, flood control projects built over the past 80 years limited the damage. The Sacramento River Flood Control Project handled the excess water quite well, although there is a need for improvements to the system, especially on the American River. Although other types of stOmlS can cause floods in California, including local floods from strong thlmderstornlS, the most feared flooding comes from big winter season stomlS covering a wide area. These stornlS are slow moving with a long westerly fetch extending toward Hawaii, the so-called "pineapple connection." Often there is a near balance between a high pressure area to the south of California and a strong low pressure area off the northern California or Oregon coast. The greater the pressure difference, the stronger the moisture-laden southwesterly winds, which dump enormous amounts of rain and snow as the air is lifted over mountain barriers such as the Sierra Nevada. The line of strongest air mass contrast, the frontal zone, can ripple back and forth several hundred kilometers but produces almost continuous rain to fairly high elevations over a broad zone in northern or central California (and less commonly in southern California). This warm southwesterly flow pattern is evident in practically all of our large general floods. The direction of orographic wind flow is important. The greatest amount of water is extracted when the wind flow is at right angles to the mountain barrier, or from the southwest for the Sierra Nevada. A southerly wind does not produce such large amounts in the Sierra, but often concentrates precipitation at the north end of the Sacramento Valley, and even the nOmlally rain-shadowed eastern slopes of the Northern Coast Range if there is a small easterly component. Of course, many stOmlS start out with a more

PAGE 272

Roos 257 southerly flow during the early phases and shift into a southwesterly and eventually westerly direction as the stonn progresses. A west or northwest direction means the flood threat is passing for two reasons: cooler air has less moisture content and cooler temperatures mean lower snow levels, which curtail the direct runoff. Many people think that snowmelt is the cause of the flooding during the big southwesterly winter storms. But melting snow is only a small portion of runoff during these events, perhaps 10 to 15%. Most of the flow is direct rain runoff from intense rain falling to high elevations. One other factor is necessary to produce large floods in northern California. That is wet ground, which requires antecedent precipitation. The most striking example is the Colunlbus Day stonn of October 1962. This stonn produced rainfall comparable to standard project flood amounts (exceeding 1-in-200-year 3-day totals), yet nmoff was less than that from a lO-year event because the rain fell on dry ground. It only produced a moderate flood, unusually early in the season, but not big enough to make the top 10 floods of the century. THE JANUARY FLOODS Water year 1995 was somewhat unusual in that we had two periods of substantial flooding and the areal extent embraced most regions of the state at one time or another. In the first large event in January, the Coast Range north of San Francisco and the upper Sacranlento Valley were hit particularly hard (Figure O. In three days, stages on the Russian River jumped from low levels to nearly as high as the record-breaking February 1986 flood (Figure 2). On the Napa and Eel rivers, also part of our flood forecasting program, water levels were not quite as high as 1986, but well over flood stage. The upper Sacramento River flood in January was generated from tillcontrolled side stream inflow from the area below Shasta Dam. Inflow to Shasta Lake neared 120,000 cubic feet per second twice during the week of stonns, but was almost completely stored. Flood levels in the upper valley exceeded 1986 at some stations but were lower than in the larger March 1983 flood. Farther downstream, peak levels were much less than the record levels of February 1986, by about four feet at Fremont Weir and Sacramento. Runoff from major Sierra rivers was not that unusual and mostly stored at the reservoirs. Peak American River inflow at Folsom Dam was about 68,000 cfs with nearly 120,000 cfs at Oroville Dam on the Feather River. Releases from the Oroville Complex to the Feather were only 5,000 cfs later in January, compared to 150,000 cfs in 1986.

PAGE 273

258 The Big 1995 Floods in Northern California en Figure 1. Location map. I{ I J 'C// / / \y \ Y 7

PAGE 274

RODS Hacienda Bridge RIVER: Napa RIVER: Scolia 259 SACRAMENTO RIVER: 50 40 3/95 1/95 1986 1983 1964 33297 33231 1986 1983 1964 Figure 2. Peak flood stages at six sites (in feet). Note that 1995 stages are preliminary, hased on telemetry. THE MARCH FLOODS As the January stonns began, major reservoir storage in northern California was quite low because 1994 had been extremely dry. Much of the flood runoff was stored in the reservoirs. Statewide storage increased nearly 8 million acre feet during January, from 75% to 104% of average. February was quite dry with a much slower storage increase, but by the time the March stonns began, many reservoirs were approaching allowable flood limits. llms, once heavy nmoff began, major releases had to be made, adding to the volume of downstream flow in the flood way system. During March, releases from Oroville and Folsom danls were boosted to around half the downstream charmel rated capacity, while later in the stonn releases from Shasta Dam for a short time reached the rated

PAGE 275

260 The Big 1995 Floods in Northern California downstream Sacramento River channel capacity in Redding of 79,000 cfs. Oroville releases peaked at 87,000 cfs and Folsom releases at 50,000 cfs. Because of less side flow from other tUlcontrolled tributaries, peak March flows in the upper Sacramento Valley were a little less than in January. In the lower system at Fremont Weir near Sacramento, stages were about one foot higher than in January due to more tributary reservoir releases, but still within design capacities. To help control Sacramento River levels in the Sacramento metropolitan area, 22 gates (of 48) in the Sacramento Weir were opened. The March floods produced a new peak of record on the Salinas River and, based on flood marks, exceeded the 1986 peak on the Napa River. They also produced water above warning stage on the lower San Joaquin River. Arroyo Pasajero flows near Coalinga, which collapsed the 1-5 bridge crossing, probably were close to a 100-year event. The real surprise was the Salinas River, which crested at Spreckels, near Monterey, about 4 feet above the previous peak of record, 26.2 feet in 1969. This was within one or two feet of the estimated stage in the legendary 1862 flood-long before the upstream Nacimiento and San Antonio dan1s were built. The Pajaro River, too, exceeded its flood stage but was not as high as its 1958 record at the Chittenden gage. You probably recall the photo graphs of the little town of Pajaro when protective river levees gave way. High water problems continued on the Sacramento River with record May floodway flows and even into SlUnmer from snowmelt on the San Joaquin River. But the big floods were in January and March. CONCLUSIONS The 1995 California winter stonns were unique with respect to the breadtll of tUlusually heavy precipitation statewide. Individual stonn series concentrated more heavily in certain regions of the state, but the ctUl1Ulative result of the January, March, and late April stonns was a seasonal precipitation total (through April 30, 1995) of 165 % of average. The North Coast was "only" 145% of average-but this region is norn1ally quite wet with about 50 inches of average annual precipitation. The major flood control work-> of the Sacramento Valley handled the rain and runoff quite well. Flows were within design. There were problems on smaller streams and on the lUlregulated or partly regulated rivers, especially in some of tlle coastal regions, i.e., the North Bay and Central Coa.<;t. Intense local convective stonns circulation did overload smal1 streams and stonn drainage facilities and produced some rare recurrence statistics. One such event was the January 1995 local rainstonn northeast of Sacramento where up to 6 inches fell in 24 hours. The 1995 floods again pointed out how vulnerable some urban areas are to flooding and raise questions about the extent of flood protection to

PAGE 276

RODS 261 build. Major urban areas, in our opinion, should have reliable protection from at least the "standard project flood," traditionally used for designing flood control systems on major rivers where the nature of failure-as well as consequences to people and property-can be catastrophic. For example, the American River, which flows through the state capitol at Sacramento, "behaved" rather well during the winter storms. In 1995, it produced much anxiety but no real threat. The maximum threeday inflow rate to Folsom reservoir from the American River was only about 55,000 cfs, which is about the one-in-five-year recurrence level. In contrast, the estimated one-in-200 year 3-day rate is about 240,000 cfs. The peak historical 3-day rate was 166,000 cfs in 1986. It has been known since the 1986 floods that the American River Flood Control System was severely undersized. The potential for disaster in Sacramento is great, because the 1,900-square-mile watershed is capable of developing a peak inflow to Folsom of about 440,000 cfs in a l-in-200-year event (a little smaIler than the standard project flood, which is about a 250-year event.) This is almost half of the 1,100,000 cfs flow rate past St. Louis during tile 1993 Midwest flood-which was carrying lhe combined flow of the Mississippi and Missouri Rivers, which drain a 700,000 square mile watershed upstream of St. Louis. If a one-in-200-year probability flood had occurred on the American River this past winter, the existing system would have been overwhelmed. TIle very thought of Folsom Dam operators being forced to release inflows of up to three times tile capacity of the Lower American River channel, putting almost 400,000 people and $35 billion in damageable property at risk, is tmly frightening. Decision time for Sacramento wiIl occur this year, seeking Congressional authorization to provide the state's capitol with an appropriately high level of reliable flood protection. REFERENCES California Department of Water Resources ]980 California Flood Management: An Evaluation of Flood Damage Prevention Programs. Sacramento, CA. U.S. Anny Corps of Engineers D87 Folsom Dam and Lake, American River, California, Water Control Manual. Sacramento District. U.S. Anny Corps of Engineers J 996 American River Watershed Project, California Part I Main Report; Part II Final Supplemental EIS/EIR. Sacramento District.

PAGE 277

This page IS intentionally blank

PAGE 278

Part 10 STORMWATER MANAGEMENT

PAGE 279

This page IS intentionally blank

PAGE 280

Urban Stormwater Regulations: A Worthy Opponent to Development Induced Flooding William J. Weaver STS Consultants, Ltd. INTRODUCTION Unbridled urbanization can cause flooding. In an attempt to stem the tide in northeast Illinois, state and local governments enacted several significant stonnwater and floodplain management regulations. The regulatory rules in force before the 1980s clearly were not containing the inexorable increase in flood problems associated with urban development. These rules contained the following flaws: Allowable storn1water release rate recipes that ignore regional hydrologic systems; Regulatory focus on floodway conveyance with a disregard for the importance of flood fringe storage; Ignorance or apathy with regard to the impact of downstream hydraulic controls or discharge capacity; and Imprudent emphasis on major flood events while ignoring lesser, more frequent events. Several major floods during the 1980s caused tremendolls damage and disruption. These events helped motivate the regulatory agencies to search for new regulatory tools to replace the outmoded regulatory recipes. The advent and proliferation of high speed and cost-efficient computers also helped to make feasible a more sophisticated regulatory approach. Regulations implemented during the 1980s provided a new problem solving approach to flood control. The State of Illinois promulgated new floodplain management regulations to plug some of the regulatory holes in the dike. These new rules primarily addressed floodway development standards. On the stonnwater management side of the coin, a concerted effort by regional

PAGE 281

266 Urban Stormwater Regulations and Development-Induced Flooding public agencies produced a comprehensive model ordinance that has been adopted by most communities and counties in northeast llIinois. Furthermore, governmental agency staff and consultants have become more familiar with sophisticated hydrologic engineering techniques. Several county agencies in northeast Illinois have recently implemented some of the most stringent floodplain and stomlwater management ordinances in the nation. Lake County regulates floodplain impacts in watersheds as small as 20 acres. DuPage County has developed dynamic flood routing models for many streanlS within its jurisdiction. Developers must use these models to evaluate downstream impacts. Although some loopholes still exist, the new regulations should, in theory, be effective. Yet some regulatory agencies still feel it is necessary to consider future land use changes and flood discharge increases when evaluating project impacts. An analysis of flooding conditions in the West Fork North Branch of the Chicago River (WFNB) at Northbrook, Illinois, sheds some light on this issue. Flooding conditions monitored before and after the new regulations were enacted indicate that flood discharge increases due to urbanization may no longer occur. The 11.6 square mile WFNB watershed (Figure 1) experienced significant urbanization between 1954 and 1987. Long-tenn streamflow records at the Dtmdee Road gage include flood hydrograph measurements for all stomlS that occurred during this period. Evaluation of stonn events that occurred both before and after sophisticated regulations were enacted provides valuable insight. HYDROLOGIC ANALYSIS Flood discharge for a given frequency in a developing watershed normally increases over time absent regulatory or stmctural controls. A comparison of actual and expected flood hydrographs generated by significant storms throughout the study period should provide insight to the effectiveness of the regulations. WFNB stream gage records are biased by urbanization; therefore, a traditional statistical analysis of the raw streanl flow database is not possible. Adjustment of the database with a technique developed by the U.S. Army Corps of Engineers (U.S. Amly Corps of Engineers, 1983) helps to eliminate this bias. The WFNB watershed rainfall-runoff process was simulated with the HEC-l computer program (U.S. Army Corps of Engineers, 1981). The model employs the Clark Unit Hydrograph method to synthesize flood discharge/frequency relationships, and to calibrate to historic flood events. The calibration process optimized the following: time of concentration (Tc), the Clark method storage value (R), and rainfall infiltration. The HEC-l model allows for an estimation of expected increases in flood discharge for selected rainfall events. Updated physiographic parameters incorporated in the model reflect conditions for

PAGE 282

Weaver TOWN LINE RO. WATERSHED BOUNDARY. HALF Ot.Y WEST FORK ---./ NORTH BRANCH CHICAGO RIVER DEERFIELD N 267 Figure 1. Vicinity map.

PAGE 283

268 Urban Stormwater Regulations and Development-Induced Flooding each projection period. Multiple regression relationships derived for each model parameter, based upon data from 16 regional stream gages, reflect regulatory and hydrologic conditions prior to 1976. The study considered several significant flood events in the 1950s, 1960s, and 1970s. HEC-1 model Tc and R relationships vary with the time frame for each storm event analyzed. A comparison of the HEC-1 flood hydrograph with actual measured flood hydrographs during these three decades produced a good correlation in every case. As such, the HEC-1 model and the Tc and R relationships should provide a reasonable representation of the impact of urbanization upon the WFNB flood discharge/frequency relationship through 1976. If flood control regulatory conditions enacted prior to 1976 remained Wlchanged, the synthetic projection of flood discharge increases caused by urbanization after 1976 should be reasonably accurate. In reality, most WFNB watershed communities and the State of Illinois implemented stringent flood control regulations during the middle to late 1970s. HEC-1 flood discharge/frequency projections would be expected to overpredict actual flood flows if the regulations have been effective. Table 1 is a sununary of flood discharge increases projected with the HEC-1 model considering regulations in place prior to 1976. Table 1. Flood discharge at Dlll1dee Road. Flood Recurrence Interval (years) 2 10 50 100 Discharge (cfs) 1950 1970 1987 250 430 860 1040 460 720 1200 1360 560 830 1400 1700 To test this hypothesis, we evaluated several significant events that occurred during the 1980s. Of particular interest was a stoml in August 1987. This event generated up to eight inches of rain in a 20 hours, and is estimated to have produced a 25-year flood event. A similar event occurred in July 1992. The HEC-l synthetic model predicted peak flood discharges that are approximately 10% higher than actual recorded flows for both of these events. Figure 2 illustrates both actual and projected hydrographs for the 1987 event.

PAGE 284

14001 2 If D '" LEGEND >f 1300 .., ILIOO S 1100 \ C 1000 OBSERVED H 900 OBSERVED A \ Go- COMPUTED R 800 G 700 E 600 500 400 C 300 F 200 S 100 0 0 5 1 0 15 20 25 30 35 40 45 50 TIME(HRS) I I\) Figure 2. West Fork North Branch at Northbrook, August 13-14, 1987 storm.

PAGE 285

270 Urban Stormwater Regulations and Development-Induced Flooding CONCLUSIONS The following conclusions are implied by the results of this study: Watershed urbanization during the 33-year study period has significantly increased flooding. Stormwater and floodplain regulations implemented since the late 1970s by WFNB watershed communities appear to have halted the adverse impacts of urbanization upon flood conditions. Flood discharges for selected events may no longer be increasing with time and urbanization. There is evidence that regulations implemented during the past 15 years in this watershed have been adequate to protect against worsening flood conditions. The proper mix of dynamic regulations and technical expertise to ensure proper application of these rules should help lessen future flooding problems. REFERENCES U.S. Army Corps of Engineers 1981 Computer Program HEC-I, Flood Hydrograph Package. Davis, CA: Corps of Engineers, Hydrologic Engineering Center. U.S. Army Corps of Engineers 1983 North Branch Chicago River Phase I General Design Memorandum, Appendix D, Hydrology and Hydraulics. Corps of Engineers, Chicago District. U.S. Army Corps of Engineers 1987 North Branch Chicago River, Phase II General Design Memorandum, Appendix A, Hydrology and Hydraulics. Corps of Engineers, Chicago District.

PAGE 286

Small Watershed Stormwater Management Programs R. W. Lindley Lindley & Sons, Inc. INTRODUCTION Defmite benefits accrue to a conmmnity with the implementation of stormwater management progranls in small (1-to 1,000-acre) watersheds. Of primary benefit is the temporary acconmlOdation of rainfall excess from high-intensity, short-duration precipitation events in socially acceptable locations. Further, planning the locations of these socially acceptable storage areas can produce aesthetically pleasing and/or useful open space facilities that would not otherwise be a part of the land use change. The case study presented examines the benefits of storm water controls installed within a 71O-acre watershed in a fully urbanized, mixed-use office/research section of northeastern Naperville, Illinois, 30 miles west of Chicago. While in the undeveloped state, several occurrences had been recorded of flood damage to downstream properties due to floodwaters from this particular uplands area. Owners of the flood-damaged property expressed concern about the potential effects of further urbanization. However, following the implementation of appropriate stonnwater management programs in the dominant area of the watershed, the flow of rainfall excess has remained within the channel limits and below the banks during several record stonn events. No further flood damage has been recorded since the urbanization of the dominant property. CASE STUDY Twenty-nine separate subcatchments where stomlwater management facilities have been constructed in the subject 71 O-acre watershed were examined by means of a detailed hydrologic analysis using the U.S. Department of Agriculture/Natural Resources Conservation Service (USDA/NRCS) TR20-87 progranl, with emphasis on the development of a 30-acre parcel and the effect of that development upon downstream

PAGE 287

272 Small Watershed Stormwater Management Programs facilities. A partial waterway analysis using the USDA/NRCS WSP-2 Lisle version was employed to establish the safe bankful open channel capacity that exists, as well as the capacity of some of the existing control structures. The lUldeveloped conditions of this watershed, as recorded on aerial photography taken in 1972, indicate that the predominant land use in the case study area was agricultural with numerous depressions existing within the area lUlder cultivation. After 20 years of urbanization, aerial photography taken in 1992 illustrates the change to predominantly urban land uses that include pennanent locations for the temporary accommodation of accumulations of excess rainfall. These temporary storage locations are interconnected by means of an lUldergrolUld convenience stonnwater drainage system. The 29 individual subcatchments in this watershed were established along ownership and development bOlmdaries. The various detention facilities constmcted were planned to regulate the stonnwater flUloff from each subcatchment area in confonnance with the conveyance available in the downstreanl drainage system. Regulation of nmoff was planned to accommodate the temporary onstream storage and safe transport downstream of the rainfall excess from a 100-year stonn. Figure 1 depicts the hydrologic flow diagram used to model the operation of both conveyance and temporary storage facilities within the subcatchments identified from the aerial photographs and augmented by topography provided by developers. These stage/discharge and stage/storage relationships were based upon plans submitted to the city for review, which were verified both by record drawings and onsite visits to ensure that actual conditions were reasonably close to those depicted. Hydrologic identifications were prepared to illustrate the manner in which the various stomlwater management facilities would flmction during stonns of various durations. The interesting feature depicted by these hydrologic models is the relationship that exists between stonn duration and the stage/storage-storage/conveyance that is provided at different locations within the watershed. In the uplands, the short duration stonn events require frequent storage accommodation for a greater nlUllber of stonns. However, as the tributary watershed increases, it is the longer duration events that tend to require more storage vohmle for the less frequent stonns. This stomlwater management system was designed and constructed within the following parameters: that 6.0 inches of rainfall with an USDA/NRCS type-2 distribution represents a 24-hour, 100-year frequency stonn. However, the TR20 model analysis was accomplished lUlder conditions of a 7.58-inch, 24-hour, Huff third-quartile, lOO-year frequency rainfall event. The major difference is that the relative ratio between

PAGE 288

014 .0Hl (94) 0.1 020 .0469(95) O.J 029 .0086 (98}O.05 LEGEND SUBCA TCHMENT 1.0. AREAISQ.MI.i IRCN) TC RESERVOIR ROUTING B AOD HYOROGRAPHS 022 .0516 (80) 0.2 027 C> z .... ,., C> ,., ;0 z '" .0047 (BJ)o.05 001 .0089 (98)0.07 022 .0067 (98)0.07 012 .OJIJ (95) 0.2 all .OJ28 (95) 0.2 OlJ .OJ59 (95) 0.2 .0025 (95)0.05 005 .0069 (91) 0.07 007 .00J8 (95)0.05 009 .0578 (98) 0.2 006 .0084 (95)0.08 017 .0252 (90) 0.' HEWLETT PACKARD 016 .1094 (82) 0.5 PROPOSD CONDITIONS CASE STUDY NAPERVILLE TR20 HYDROLOGIC flOW SCHEMA TIC TO CRESS CREEK LINDLEY ___ __ __ .... ___ ---.l __ ".-1 __ """ __ U .. aA. ... UDif .,tot, __ t"C1Il"' . .. ... -........... """""--. ........ Figure 1. The hydrologic flow diagram. r-S !a:: I\) (j

PAGE 289

274 Small Watershed Stormwater Management Programs MET UFE SITE WASHINGTON & DIEHl RD. NAPERVIlle IL. 30 PROPOSED CONDmONS -CALCS. OY UNDlEY & SONS INC. HINSDALE ILL m + + + + CROSSSEcnON 11M F I 24 o 771' o d F 10 I 7G!I' o w 12 I 761' n C F G. S 765' o ,--N=l.OW + + + + + + 10 20 30 40 Stonn Tunc In Hours MET UFE SITE WASHINGTON & DIEHL RD. NAPERVIlle IL STRUCTURE 8 + + + + + + + 50 60 110 PROPOSED CONDITIONS CALCS. DY UNDLEY & SONS INC. HINSDALE ILL nc; + L + + + CROSS SECTION 11j F I 88 0777: o d F 66. I 760' o w 44 I 764' n C F 22 S7GO' o INFlOW STRUCTIJHE12 + + + + + + + + + + 1.----"-i'EAK .. 59 CFS/218 flCS. 0.27 CFSI fIC. + + + + r STAGE + 10 20 3D 40 50 60 Storm Tunc In Hours Figure 2. Flow data for reservoir 8.

PAGE 290

Lindley 275 conveyance and watershed size increased from 0.15 cubic foot per second (cfs) per acre drained to approximately 0.30 cfs per acre drained. This resulted in an occasional surcharge of the conveyance system during a peak flow in the model. To date, no history of system surcharge or basin overflow has been reported. All of the reservoirs in this system function as on line, which means that all surface stomlwater must pass through a series of detention facilities in order to move downstream. By-pass, as such, does not exist; and each depressional storage facility is provided with, at minimum, a two-stage outlet control. The low-flow system usually is represented by the downstream conveyance stoml sewers, and the overland system is provided through a paved overflow weir. On occasion, surface flow will occur within a roadway or greenway provided for that purpose. This advantage can best be observed by exanlining Figure 2. It shows reservoir 8, located upstreanl of and dominant to reservoir 12. Note that the relative flow rate is diminished from 0.59 cfs per acre to 0.27 cfs per acre, even though the tributary watershed has increased from 32 acres to 218 acres. It is my opinion that this type of stormwater control progranl complements the manner in which nature intended to acconmlodate occasional local flooding conditions or flash flooding. The urbanization process should include planning to acconmlOdate this accumulation of rainfall excess in acceptable locations in order to prevent the occurrence of such accwllulations in a residence or other inappropriate domain. SUMMARY The conditions under which excess rainfall is temporarily stored or detained should duplicate, as closely as possible, the conditions tIDder which such rainfall naturally accunmlates, a process which has ftIDctioned since the beginning of time. Planning for a change in land use from agriculture or open space to residential, conmlercial, or industrial should not disregard this natural process. The temporary storage of surface storm water rurloff, which often did not interfere with crop production, now presents an inconvenience or an exposure to potential danlage for the urban land user. Finding acceptable locations for these rainfall accumulations is the challenge that must be met. While the volwne of surface stonnwater rurloff will always be greater from urbanized land as opposed to the rainfall excess from land under agricultural use or open space, compensation for the increased volume can be partially achieved by distributing the conveyance of surface storm water runoff over a longer time period.

PAGE 291

276 Small Watershed Stormwater Management Programs The stonnwater management concept that I have used for the past 30 years might be tenned the Honest Rational Method: It is possible to transport all of the water some of the time or some of the water all of the time. It is impossible to transport all of the water all of the time. It is possible to store all of the water some of the time or some of the water all of the time. It is impossible to store all of the water all of the time.

PAGE 292

Multi-objective Planning and Design of Stormwater Detention Facilities Ronald D. Flanagan R.D. Flanagan & Associates INTRODUCTION Multi-objective planning is the accomplishment of as many public and private policy goals as possible with a single project. Declining public budgets and workforces coupled with rising costs and demands for service are presenting the public and their elected officials with seemingly insurmountable budgetary chaIIenges. With most flmds taken up by highly visible, everyday demands of streets, crime prevention, water and sanitary sewer service, less visible needs such as stomlwater management, parks and recreation, the environment, nature and wildlife protection, hiking and biking trails, and wetlands preservation often suffer severe cuts. The future of these programs lies in the identification of multiple-use opportunities where possible in every public project. Through application of the multi-objective planning process, many diverse programs' objectives may be accomplished in spite of budget reductions. Storm water detention facilities afford an excellent opportunity for multi-objective planning. STORMWATER DETENTION FACILITIES Stormwater detention should be an important element in any community'S stonnwater management strategy. Detention facilities should be planned and located as the result of a watershed-wide basin master drainage planning study. Random location or across-the-board requirement of detention with every development may cause greater downstream flooding than no detention at all, due to improper location and timing. In addition, larger regional detention is preferred to several small sites because of savings in land and construction costs, and increased operations and maintenance efficiency. Larger facilities are also more easily planned and used for other community activities, providing ideal locations for public multiple-use areas. Almost any flood-tolerant activity is suitable for location in

PAGE 293

278 Multi-objective Planning and Design of Stormwater Facilities association with a stomlwater detention facility. Successful examples have included public park and recreation facilities, school playgrounds, parking lots, reforestation areas, wetlands, nature preserves, lakes and ponds, wildlife habitat, aquifer and groundwater recharge areas, water quality enhancement, open space buffers between incompatible urban land uses, and relief in the built environment. 1brough application of the multi objective planning process, discussed below, opportunities to identify and maximize public policy multi-use objectives may be realized. TEN-STEP MULTI-OBJECTIVE PLANNING Multi-objective planning is more complex than a straight-line single purpose planning process because of the many disciplines involved, multiple project objectives, and active citizen involvement. The multi objective planning process can be swnmarized as 10 distinct steps, shown in Figure 1. It differs from conventional planning in that it is a circular process, repeating itself in a constantly-evolving helix MONITOR! ADJUSTMENT ACTION PLAN 9 _----8 ..---1-----. CITIZEN INVOLVEMENT PLANNING PROCESS 6 ANALYSIS 3 GOALS & OBJECTIVES MANAGEMENT WORK PLAN RESOURCE INVENTORY Figure 1. The 10 steps in multi-objective planning.

PAGE 294

Flanagan 279 (1) Citizen Involvement The first step in planning a successful multi-objective storm water detention facility is to obtain active citizen involvement, especially from neighbors of the project. Properly utilized, citizens can be some of the most important members of the planning design team. Too often design professionals hold the view that "the only thing wrong with public hearings is that sooner or later, the public is bound to show up." Citizens know their neighborhoods better than anyone else, can provide valuable assistance throughout all planning phases, and can be important sources of support when project financing and implementation are considered. Planners and elected officials should lmderstand that the success of a project depends on the identification of common conmlUnity goals and the sharing of decision making with citizens, particularly project neighbors. (2) Problem Identification The second step is to identify cOllummity problems that should be addressed during the planning process. The temporary detention of a target amount of storm water is a given, but the dynanlic and exciting part of the planning process is for the citizens and design team to see how many other public/private problems can be identified and solved in the project at the same time. This is the essence of multi-objective planning. Issues such as lack of neighborhood open space, need for safe hiking and biking trails, urban bird and small manunal habitat, improvement of urban stomlwater runoff quality, and urban reforestation, would usually not be addressed in a single-purpose detention pond design study, but are integral to multi-objective planning. (3) Project Goals and Objectives The third step is to establish clear project goals and objectives, with the active participation of citizens, city staff, interest groups, and elected officials, so that all parties fully lmderstand what is to be accomplished. Clear project goals are important because little is ever accomplished unless scarce staff and financial resources are concentrated on the accomplishment of a few clearly defined priorities. Most designers tend to be cautious about setting anlbitious project goals, preferring to exceed low expectations rather than fall short of higher ones. But small plans don't generate the enthusiasm necessary for great and creative projects. A successful teclmique, developed by the Johnson Creek Planning Consortilml in Arlington, Texas, is to hold a public goal-setting meeting. At the meeting, every goal mentioned by the participants, without regard to feasibility, is posted on a wall. Everyone is given several adhesive dots to place on the goals they consider most

PAGE 295

280 Multi-objective Planning and Design of Stormwater Facilities important. Community priorities reveal themselves without any individual's or group's idea being rejected out of hand as absurd or impractical. Initial project goal setting should be limited only by the imagination of the participants. (4) Management Work Plan A management work plan is a step-by-step map to the accomplishment of the project's goals and objectives. Although it is often omitted from the planning process, it is essential to coordinating the complexities involved with multi-disciplinary design teams, active public involvement, and mUltiple public goals and objectives. A multi-disciplinary design team typically consists of hydrological engineers, planners, landscape architects, soils scientists, geologists, biologists, environmental scientists, and public relations specialists. Each project task should be clearly identified and described in the work plan, including task objective; methodology to be employed in task accomplishment; designation of team leader and task participants; participant responsibilities; level of effort in dollars, direct costs, and labor hours; time lines and key dates; and interim and final work products. A good management work plan informs each participant of his/her role, and the roles and interrelationships of others, provides project guidance, and is flexible enough to address the many unexpected contingencies that will arise in the multi-objective plan development process. (5) Resource Invento ry The inventory of resources and development of the project database is the first major planning phase, and usually consunles about 25% of the total project time and budget. In multi-objective planning, the development of the database should include a comprehensive inventory of everything that might impact, or be impacted by, the project. The inventory can include everything from citizen attitudes to park and recreation needs, other public and private plans, native vegetation, soils, underlying geology, utilities, and habitat. Thoroughness is crucial. If data is omitted from the inventory it cannot be considered in the screening of alternative plans or in the selection and refinement of the final plan. A citizens' meeting should be held at the end of the phase to report and digest the findings of the resource inventory. It is important to share information with all playersteam members, citizens, neighbors, staff, and elected officials-throughout the project. An open planning process is a major factor in ensuring project success and public support.

PAGE 296

Flanagan 281 (6) Inventory and Data Analysis The sixth step in the multi-objective pI arming process is to analyze the data gathered during the resource inventory. The analysis phase should begin about one-third of the way into the project, and usually takes about 12% of the project time and budget. Design team members should actively participate in, and fully tmderstand, the analysis of information from the various team disciplines. Unlike a single-purpose project, much give-and-take is involved in a multi-disciplinary/objective project. Often, project conflicts must be resolved by referring back to the initial project goals, or by seeking policy guidance from elected officials. Writing and preparation of the plan report document should begin or be underway during this phase of the project. (7) Alternative Plan Scenario Development Phase seven, the development of alternative plans for consideration, should begin about half way through the project, and normally requires about 15% of the time and budget. A wide range of alternatives should be evaluated, including a basic, single-purpose detention alternative for storage and cost comparisons. The alternatives should be presented at a public meeting, and citizen input solicited. The final plan will most likely be a hybrid of the most desired features of several alternatives. (8) Selected Plan Refinement Refining the final plan is often the most difficult and challenging part of ;he multi-objective plarming process, since it brings together disparate plan elements into a coherent whole and resolves all conceptual conflicts. The plan refinement phase nonnally requires about 25% of t1le project time and budget. The final selected plan should be presented in text and color graphics so as to be easily tillderstood by a non-technical public, but in sufficient detail to offer guidance to detail design engineers and landscape architects. A final project public meeting should be held to present the plan to the public and elected officials. If the plarming process was sOlmd, l1e public will be present to endorse the plan and urge its adoption by elected officials. (9) Action Plan If the plan is to be more than a paper exercise, or "shelf dOCtilllent," it must be accompanied by an action plan containing step-by-step procedures, with time lines, for implementing the plan. Action plan elements include public information, education and media relations strategies; financing alternatives and potential funding sources; and

PAGE 297

282 Multi-objective Planning and Design of Stormwater Facilities identification of support groups and allies. The action plan should be an integral part of the overall project plan, since the objective of planning is not the production of plans, but the initiation of organized, well-infonned, intelligent action. (10) Plan Monitoring and Adjustment Post-project monitoring and feedback is critical for multi-objective projects, with their complex and diverse team and discipline memberships, ambitious goals, and non-traditional plan element relationships. Usually all these elements work together more or less as planned, but sometimes the combinations produce unexpected undesirable results and must be modified. The plan must anticipate the possibility of unforeseen events following implementation, and have a process in place to deal with them. With citizen feedback, problems are identified, new goals established, and the planning process continues anew, in an upward, helix-like cycle. CONCLUSION Stonnwater detention facilities afford excellent opportunities to achieve multiple community program objectives within a single project. Detention ponds, in addition to the periodic storage of floodwaters, make excellent park, recreation, and open space areas, provide space for urban forests, wildlife habitat, wetlands, nature preserves and outdoor science classrooms, and can aid in cleaning the air and water of urban pollutants. Multi-objective planning is more complex than single-purpose planning, but pays large dividends by accomplishing many important public policy objectives with el\ch tax dollar spent. By following the 10-step multi-objective planning process, drab single-purpose stonnwater detention facilities can be transfonned into popular community assets.

PAGE 298

Innovative Approach for Peak Discharge Reduction In an Urban Environment Using a Multipurpose Detention Basin Douglas Lantz Zbigniew Osmolski Fazle Karim Pima County Department of Transportation and Flood Control District INTRODUCTION Flood flows from the Arroyo Chico and its tributaries cause severe and frequent flood damage to central Tucson, Arizona. One area subjected to flooding almost every year because of the limited capacity of the stream channel passing through it, is the residential neighborhood Colonia Solana, which is listed on the National Register of Historic Places. Urban encroachment into the floodplain has occurred over the years, severely limiting the rights-of-way needed for implementation of traditional flood control measures like channel improvement, levee, structure relocations, etc .. Another constraint is the desire of the residents and elected public representatives to preserve the historic character of the neighborhood. Consideration of various flood control alternatives indicated that using the Randolph South golf course, inmlediately upstream of the historic neighborhood, as a detention basin was the best alternative. It would satisfy the above-mentioned concerns and reduce peak flows in the downstream areas, including Colonia Solana. An innovative approach for the design of the detention basin was needed because of (1) the need to preserve the golf course flmction of Randolph Park, which provides significant economic benefit to the commtmity; and (2) the prohibitive cost of a single detention basin, which would require a high embankment and a probable maximum flood (PMF) spillway under state dam safety criteria. The innovative design consists of a series of six interconnected basins excavated within the Randolph South golf course. Individual basins were designed such that the PMF spillway will not be required, and were configured to fit between fairways and greens to preserve the golf course.

PAGE 299

284 Peak Discharge Reduction in an Urban Environment The project was designed and constructed under a cooperative agreement between the Pima COlUlty Flood Control District, City of Tucson Department of Transportation and Department of Parks and Recreation, and the U.S. Army Corps of Engineers, Los Angeles District. Construction was completed in April 1996 by Tucson Parks and Recreation, with major funding provided by the Pima County Flood Control District, to be reimbursed by the Corps under Section 104 of the Flood Control Act. HYDROLOGY The Randolph South detention basin is part of the larger Tucson Drainage Feasibility Study, underway by the U.S. Amly Corps of Engineers, which covers the 1l.35-square-mile watershed for Arroyo Chico at the Santa Cruz River (Figure l). As part of the study, a HEC-1 rainfall-runoff model was constructed and calibrated for this watershed. The basic runoff criteria for the model (S-graph, n-values, and loss rates) were determined by reconstituting six observed runoff events on High School Wash, which had rainfall and nmoff gages operated by the University of Arizona. The model included the Phoenix Valley S-graph, a lag equation in which Manning's n-value ranged from 0.035 to 0.050, and uniform loss rates ranged from 0.5 in/hr to 2.0 injhour. A 6-hour summer thunderstorm was chosen for the design stoml. This duration provides almost all of the volume produced by SUflliller thunderstorms that will be contained in the detention basin, but also contains the intense rainfall for shorter durations and is thus the critical storm in producing peak discharges as well. The 6-hour rainfall depths were developed using the National Oceanic and Atmospheric Administration (NOAA) Atlas II, Volume 8 for Arizona (NOAA, 1973). The temporal distribution was adapted from the August 1954 thunderstorm over Queen Creek, Arizona, east of Phoenix. The HEC-1 model was calibrated by adjusting loss rates and n-values to reproduce discharge frequency curves for three gages on the watershed, and volume frequency curves for two gages. The subwatershed for the Randolph South detention basin is drained by Arroyo Chico, Naylor Wash, and Paseo Grande Wash, and has a total drainage area of 3.51 square miles (Figure 1). The lOO-year inflow hydrograph produced by the calibrated model at Randolph South had a peak discharge of 3100 cfs and a nmoff volume of approximately 430 acre feet. For detailed hydraulic modeling, the 100-year inflow hydrograph was broken into six subwatershed hydro graphs, each of which entered the detention basin complex at a different point. The two main flows are from Arroyo Chico (subwatershed AC), which drains a 1.13 square miles to the east, and the combined Grande Wash (subwatershed GW) and Naylor Wash (subwatershed NW), which together drain l.9 square miles to the

PAGE 300

Lantz, Osmolski, and Karim 285 Figure 1. Randolph South detention basin watershed.

PAGE 301

286 Peak Discharge Reduction in an Urban Environment east and southeast. The remaining hydro graphs contributed runoff from the golf course area itself (subwatersheds ACRN, RNE, and ACRS) and from a highly urbanized area to the northeast (subwatershed RNELC). DESIGN OF RANDOLPH SOUTH The project area includes two existing I8-hole municipal golf courses: Randolph North and Randolph South (Figure 2). By virtue of location, relatively low user fees, and year-round weather, they are reportedly two of the busiest mlIDicipal courses in the COlIDtry. Arroyo Chico generally bisects the two courses, while Naylor Wash flows through the south course to its confluence with Arroyo Chico just upstream of Randolph Way. The basin outflow was constrained by the Arroyo Chico charmel immediately downstream of Randolph Way. The existing charmel is small, having a bankfull capacity of approximately 300 cfs (less than the 2-year flood), and is surrounded by heavy desert riparian vegetation on both sides. Since the wash and the neighborhood through which it flows are listed on the National Register of Historic Places, charmel improvements through the neighborhood were not a practical option. Preliminary design attempts for the basin looked at a single embankment along Randolph Way. This concept was rejected for two reasons. One, it would back water onto the golf course, danlaging tees and greens, during relatively frequent events. Two, it would be classified as a jurisdictional danl by the Arizona Department of Water Resources DaDl Safety Division, and require construction of a PMF spillway. Subsequent design focused on a combination of excavated basins designed to work with a new layout of the golf course (Figure 2). Because of the relatively steep 2% slope, it was possible to construct a cascade of basins through which flood flows were conveyed both in parallel and in series. As an example of parallel storage, flows from Naylor Wash are intercepted by basin 1 while flows from Arroyo Chico are intercepted by basin 3. In series flow, basin 1 drains directly to basins 2, 3, and 6, which drain through basins 4 and 5 before reaching Randolph Way. Basin 3 drains to basin 4, which in turns drains to basin 5 and to Randolph Way. Interbasin conveyance is via weirs and culverts ranging from a single 18" reinforced concrete pipe (RCP) to a 3 barrel 60" RCP. A non-jurisdictional embankment along Randolph Way collects and detains the runoff from the urban area to the northeast, and also serves as the final control point for rest of the basin. The final outflow is metered to the Arroyo Chico charmel via a single 3' x 5' concrete box culvert under Randolph Way. This overall combination of below-ground storage in six interconnected basins and an embankment at Randolph Way served the multiple objectives of the project without requiring an expensive PMF spillway. It also allowed for design of a unique and challenging golf

PAGE 302

Lantz, Osmolski, and Karim Figure 2. Randolph South detention basin. I I I I d 287

PAGE 303

288 Peak Discharge Reduction in an Urban Environment ?ASEJ CitI.';CC: W'SM r;;:l L!J B NAYLOR .. ASH ARROYO CHICO AT ALVERNON WAY G 1 1 1 1 1 1 1 STRU 1 1 1 1 1 1 1 1 l _____ T _____ J 1 I 1 1 1 1 1 : : ______ J ARROY CHICO RANDOLPH SOUTH _I' ARROYO CHICO r.;:;;J RANDOLPH NORTH [5------; ------01 L ______ [J RANDOLPH NORTHIRNEL+ U ___ ;-u __ u ELCON MALL OUTFLOW UNDER RANDOLPH WAY Figure 3. Randolph South detention basin schematic routing diagram. course, especially compared to the previous course, which was often referred to as the "pool table." MODELING AND RESERVOIR ROUTING The HEC-l rainfall-runoff model is not appropriate for modeling interconnected detention ponds because it cannot adjust the stage..

PAGE 304

Lantz, Osmolski, and Karim 289 discharge curve as tailwaters of the individual basins fluctuate. The Advanced Interconnected Channel and Pond Routing (ADICPR) Program from Streamline Technologies was written specifically to route flows through storage nodes (basins) connected by various reaches (pipes, open channels, or weirs), and was used for routing flows through the six interconnected basins. The water surface elevations at each node, and the discharge in each reach are computed for each time increment based on (1) a downstream boundary condition, (2) stage-storage relations for each node, (3) stage-discharge relations for each reach, and (4) incoming flows. Each node in the ADICPR model represents a control volume. Water enters and leaves each node by the links connected to it, and by runoff hydrographs flowing into it. Storage at each node is provided by specified stage/storage relationships (i.e. stage-volume, or stage-area). The change in storage in each node is based on the differences in inflows and outflows at each time step during a simulation, and is used to determine the water surface elevation at each node. Flows through each link (i.e., pipes, channels, or weirs) are calculated from known elevations at the ends of the link and the hydraulic properties of the link (slope, roughness, and geometry). Simultaneous solution of the elevations, flows, and storage is done by iteration. The computation time step is variable and can be reduced to fractional seconds to minimize nun1erical inaccuracies. A schematic routing diagram for the Randolph South model is shown in Figure 3. The downstream boundary condition was chosen as critical flow depth through the low flow outlet, which was about the san1e as normal depth in the downstream channel. Stage-storage relations were computed by measuring storage volumes at one-foot contour intervals from the final grading plans. Stage-discharge relations were computed internally by the ADICPR program, based on the elevations of the head water and tail water during the period of interest. Incoming hydrographs were entered at the appropriate nodes as shown in Figure 3. REFERENCES National Oceanic and Atmospheric Administration 1973. Precipitation Frequency Atlas of the Western United States, Vol. 8. NOAA Atlas 2. Silver Spring, MD: National Weather Service.

PAGE 305

Retention/Detention Basins Efficiency in the Phoenix Metro Area Maximo R. De Vera Flood Control District of Maricopa County INTRODUCTION Land development projects involving subdivision, industrial and conunercial complexes, and related developments in metropolitan Phoenix are required to retain on-site nmoff using retention/detention facilities. The Flood Control District of Maricopa County (FCDMC) also requires that for similar projects the peak discharge from post development conditions should not exceed that of pre-development. Many hydrologic modeling studies have been completed that include data on retention or detention basins, some of which may be only approximations of the actual physical configuration. Thus, some information from modeling efforts is available that can be analyzed for development of a stonnwater management strategy. This paper is an attempt to determine the efficiency of existing retention/detention basins in the Phoenix metropolitan area in reducing peak discharge that affects the design of drainage facilities and the extent of floodplain along river banks; and also the percentage utilization based on maximum storage volume requirements. The study does not differentiate between regional and onsite facilities. DRAINAGE REGULATIONS Maricopa COlmty includes the cities of Phoenix, Glendale, Tempe, and other nearby cities and towns. Drainage regulations require that stored nmoff be discharged completely from the facility within 36 hours after the stonn to minimize adverse environmental effects. All detention/retention facilities within new developments shall be designed to retain the peak The author expresses gratitude to Amir Motamedi, Flood Control District of Maricopa County for his comments; mul to Dan Sherwood, City of Glendale and Raymond Acuna, City of Phoenix, for their assistance.

PAGE 306

De Vera 291 flow and runoff volwne from the 100-year/2-hour duration event over the entire development area including right-of-ways. The City of Phoenix requires that "all developments shall make provisions to retain the runoff from a 100-year/2-hour storm falling within the boundaries of the development" (City of Phoenix, 1988). Glendale adopted the 100-year/2-hour detention basin policy in 1986 (Sherwood, 1995); before that, the regulation was based on 10-year/2-hour rainfall. Other cities and towns adopted similar regulations at about the same time. In Mesa, the maximml1 depth of retention basins as measured from natural grade to the bottom of the basin was set at 3.5 feet with the basin bottom slope at a minimml1 of 1 % and side slope not flatter than 4: 1. Other jurisdictions set a maximunl depth at 3.0 ft with the sanle side slope requirement. METHODOLOGY The study involves analysis of a hypothetical watershed using HEC-1 and hydrologic modeling results from 12 selected hydrologic studies. It also includes comparison of methods for estimating storage volmlle requirement. The selected area includes a part of the whole area studied in which model data or output is available. Efficiency as used in this study is assml1ed to be the difference between inflow and outflow divided by the inflow. The inflow may have been generated by a single sub-basin or from two or more sub-basins as extracted from the model output. The outflow depends upon the retention/detention structure configuration and imposed asslmlptions in the hydrologic model. The nmoff coefficient C is asswned to be the ratio of rainfall excess to total rainfall as extracted from the HEC-l output. In this study any basin with controlled outflow such as a low level outlet and a spillway is considered a retention basin. The basic definition stipulates that stored water is disposed of by infiltration, evaporation, dry wells, or a plmlping system. Hypothetical Urban Watershed A retention/detention basin facility in a hypothetical urban watershed was analyzed for peak discharge reduction efficiency and percent utilization. A rectangular watershed of length equal to twice its width with an area of 10 acres was assmned to have a slope of 1 %, watershed factor of 0.04, and with a 25% impervious area. Average values of the Green and Ampt loss parameters were used. The HEC-l input file is generated using MCUHP1 as developed by FCDMC (1991). The storage volume requirement is equal to the nmoff volunle estimate using V=C(P/12)A, where P =2.70" (l00-year/2-hour stonn), C as defined earlier is runoff coefficient

PAGE 307

292 Retention/Detention Basins Efficiency in Phoenix corresponding to the OO-yr storm, A is area in acres, and V is storage volume requirement in acre-ft. The lO-year, 25-year, 50-year and 100year/6 hour duration rainfall were used to generate the peak discharge and peak storage. Table 1 shows the input data for the HEC-1 model. Table l. Swnrnary of HEC-l input data for the hypothetical watershed. FRE() Rfl inch' XKSAT DTHET RTIMP TC Slllr Cod C V 11102, 10 1.99 0.06 n.15 25.11 n.175 n.157 n.70 1.5X 25 2AR (1.{I6 0.15 25.n O.15R 0.140 0.74 1.67 50 2.R3 n.n6 0.15 25.0 0.150 n.132 0.76 1.71 100 3.211 0.06 0.15 25.0 0.142 n.124 o.n 1.76 The storage volume provided is 1.76 ac-ft and it is assumed that when water surface reaches 3 feet deep, water will spill at 0.6 cfs, which should be the outflow rate to empty the reservoir in 36 hours. As a detention basin the gravity outflow pipe is asswned to be 12" in diameter, which is the minimwll size required in drainage regulations. Discharge is computed using the orifice discharge equation. The above values were included in the HEC-l data file generated by MCUHP1. Volume Estimate Comparison FCDMC (1991) compared methods used by six cities in Maricopa County to estimate retention vohmle requirements. An 83.2-acre watershed with five land use types was used. The cities used empirical overland flow equation and Marming's equation to estimate time of concentration (Table 2). Table 2. Comparative peak discharge and runoff volunle estimates. CITY METHOD Peak () V(Cily)(a V(MC) Ulc-III CilY MC c-fl) Ch:UJdler Ovcrl:UJd flow Cli. + Manning', IXR 227 13.19 11.62 Glendale Usc Tc=TI+ln. TI=L/60V In9 2'1,7 7.74 10. X X Mesa Overland flow en. wilh Tc-Ti+TI 144 231 11.'1,4 IIA9 Scollsdale Ovcrl:Uld flow Cli. Tc=Ti+TI 20X 297 In.D 12.01 Tempe Ti=KLo,,/S'" TI=L/60V 13R 231 15.XlI I 1.1 X Phoenix Overland flow eq. Tp=Tcx Ave Widlh 13X 243 7.74 12AI NOle: MC = Maricopa CoullIy mctlllld

PAGE 308

De Vera 293 WATERSHED CASE STUDIES Hydrologic models of 12 project areas with generated peak Q from 100year/6-hour storm in metropolitan Phoenix were used as case studies. In most models the existing retention/detention facilities have been included. The inflow and outflow of each retention/detention basin had been extracted from the hydrologic model output. In addition, the peak storage volmne was also extracted and compared with the design storage voltmle criteria. Analysis of Results The retention voitmle estimation by FCDMC using a simple fomlUla does not differ significantly from estimation methods used by the various cities. As shown in Table 2 the FCDMC peak Q estimates are higher than the cities while the runoff voltune estimates are about the same. The HEC-l results for the hypothetical watershed using 100-year/6hour stoml are shown in Table 3. The inflow volume in acre-feet was computed from the given rainfall excess and drainage area. Figure 1 shows a plot of efficiency and percent utilization versus return period. It can be noted that for the assumed basin configuration and flow condition the percent peak discharge reduction is generally greater for the detention basin than for a retention basin. The percent efficiency for retention basins decreases with an increase in frequency from 10 to 100 years. These 100 DB Peak Q Reduction 80 .________ _____. RB Peak Q Reduction 70 -----. C 60 DB RO Vol. Reduction ____ >-g 50 III '0 40 :c w 30 20 10 RB RO Vol. Reduction 10 25 50 100 Retum Period (yrs) Figure 1. Plot of basin percent efficiency/utilization vs. return period.

PAGE 309

294 Retention/Detention Basins Efficiency in Phoenix results indicate that with the assumed input parameters peak Q reduction appears to be affected by frequency. Table 3. Peak discharge and runoff volume reduction. FREQ(YR) NFLO(cf,) NFLO(,,) RB-OUTF DB-OUTF RB6HVOL RB24HV DBflHV()L 10 29 l.7R 1 1.0 3.0 0.291 0.774 1.537 25 3X 2.251 5.0 4.0 U.627 I.l7fl 1.79X 50 44 2.5R5 10.0 4.0 0.957 1.50l) 1.919 UK) 51 2.')45 15.0 6.0 1.314 I.X70 2.07X Table 4 shows a summary of retention/detention basins' efficiency for the 12 project areas with a total area of 205.2 square miles. It appears that about 25% of the retention/detention basins have less than 20% efficiency in reducing peak discharge and about 30% have efficiency greater than 80%. The results may indicate the distinction between detention and retention basins. The latter is expected to have greater efficiency. Figure 2 is a frequency histogram of the results for 12 project areas. Table 4. Sunmlary of retention/detention basins' efficiency. PIO! Area I.Area 0-10% 11-20%21-30% 31-40% 41% 51-60% 61-70% 71-80% 81-90% 91-100% T. No. Mean Eff 91st Ave 98 4 2 0 0 0 0 1 0 2 1 10 3620 Arrowhead Ranch 6.8 4 3 3 1 4 3 2 4 8 3 35 5360 Bethany Home 15.71 2 4 0 0 1 0 0 1 1 0 9 2900 Gilbert-Chandler( 1; 37.9 2 1 0 1 0 1 0 2 6 14 67.50 GilbertChandler(2; 34.7 4 1 1 2 1 0 0 0 1 11 292C Gilbert-Chandler(3: 5 6 2 1 1 1 1 0 6 20 45 'Xl Olive Drain 9.43 0 0 0 0 0 0 I 3 3 8 8520 Skunk Creek 23.78 2 5 1 0 0 0 1 0 0 5 14 4620 SossomanDr 4.37 4 0 0 0 0 0 0 0 0 2100 Sun City West 7.48 0 0 0 0 0 0 6 3 0 10 76 eo SC W EXpansion 298 3 1 0 0 1 0 0 2 0 12 48 ()'J Sunlakes 2.25 0 0 0 0 0 0 1 3 3 11 17 925] Total 205.2 31 19 6 5 8 5 14 16 24 37 165 3648_ Note: Losl figure In line Total Is area weighted mean of the efficiencies CONCLUSION AND RECOMMENDATIONS Results from the HEC-l model for the hypothetical watershed show that peak discharge reduction for the detention basin is nearly the same as that of a retention basin. Outflow volume reduction is greater in the retention basin than in the detention basin. It is therefore recommended that if downstream flooding is to be minimized, properly designed retention

PAGE 310

De Vera fit o 35 c 25 '2..20 10 Q: '0 5 o 0 z 0 0 0 0 0 0 0 c:i c:i c:i c:i 'i' '"'f "-ex;> o N <') lI) -0 "-co Percent Efficiency Range Figure 2. Frequency histogram of efficiency decile occurrences for 12 project areas. 295 0 8 0basins should be considered as one alternative for mitigating flooding problems. A comparison of methods to determine the design capacity of a retention/detention basin shows that the FCDMC method, which uses a simple formula, is as good as any of the more theoretical approaches used by area cities for the 100-year/2-hour rainfall event. The frequency histogram of 165 retention/detention basins in the greater Phoenix area shows that they are either inefficient with less than 20% efficiency in peak Q reduction or efficient with at least 80% efficiency. The retention basins are expected to be more efficient. Existing retention/detention facilities in the area should be assessed for their efficiency in peak discharge and runoff volume reduction. Low efficiency facilities should be improved or re-designed to improve their efficiency so that downstream flooding can be minimized or eliminated. REFERENCES City of Phoenix 1988 Storm Drainage Design Manual. Subdivision Drainage Design, Infrastructure Services. Flood Control District of Maricopa County 1991 Drainage Design Manual, Volume II-Hydrology. Sherwood, D.A. 1995 Personal communication. City of Glendale Engineering Department. Oct 16.

PAGE 311

Stormwater Management Planning for Redevelopment of Denver's Stapleton Airport John M. Pflaum McLaughlin Water Engineers, Ltd. William Wenk Wenk Associates Ben Urbonas Urban Drainage and Flood Control District INTRODUCTION In 1993, the City and COlmty of Denver and the Urban Drainage and Flood Control District retained the design teanl of McLaughlin Water Engineers, Ltd. (MWE) and Wenk Associates to prepare a Stornlwater Outfall Systems Plan for the Stapleton Airport site. With the opening of Denver International Airport, Stapleton has been closed to air service and is proposed to be redeveloped over the next 30 to 40 years. This report describes the study area and sununarizes the Stornlwater Outfall Systems Plan and its interrelationship with the Stapleton Development Plan. STUDY AREA Stapleton Airport encompasses a total of 4,723 acres (7.4 square miles) as illustrated in Figure 1. The site lies within the City and County of Denver in close proximity to downtown Denver. The site is botmded on the north by the 27-square-mile Rocky MOlmtain Arsenal, a fonner weapons and pesticide manufacturing facility that is currently being remediated and converted to a national wildlife refuge. East of Stapleton the Montbello Industrial Park is served by Havana Street, while on the west residential, conunercial, and industrial uses are served by Quebec Street. Residential neighborhoods occupy the areas to the south. The City of Conunerce City borders the site to the northwest, while the City of Aurora borders on the southeast. Interstate 70 bisects the study area.

PAGE 312

Pflaum, Wenk, and Urbonas lr, Mo."",," ........ ', L I Commerce CIt!:} i + I I I I ii}!'i. ___ Montvlew eOulewrd II I Figure 1. Existing Stapleton Airport site. + E. %th Ave. + Montbello Inc:iJ6tr 181 Park. 297

PAGE 313

298 Stormwater Planning for Redevelopment of Denver's Airport Brief History Before it was developed as an airfield, the Stapleton site was characterized by rolling sand hills prairie traversed by Sand Creek and Westerly Creek. In the late 1920s the area was identified by Denver as the site for a new mlUlicipal airport, and in 1929, the 34S-acre Denver MlUlicipal Airport was dedicated. Facilities were added during World War IT, and further expansion occurred with the growth of air traffic between 1960 and 1985. Stapleton's limitations with regard to rlUlway separation, traffic handling capacity during adverse weather, and lack of viable options for further expansion led Denver to pursue development of a new airport in 1985. Denver International Airport opened on February 28, 1995 and Stapleton International Airport was formally closed. Planning for the redevelopment and reuse of the Stapleton property began in 1989. A general concept plan for Stapleton was developed as part of a commlmity planning effort known as "Stapleton Tomorrow." In 1993, the nonprofit Stapleton Redevelopment FOlUldation (SRF) was established by cornnllUlity leaders and entered into a partrlership agreement with the City and COlUlty of Denver to assist in maximizing redevelopment opportlUlities at Stapleton. A team of planning consultants was retained by the SRF and the Stapleton Development Plan was completed over an 18-month period commencing in the fall of 1993 and concluding at the end of 1994. The MWE design team worked closely with the SRF design team to coordinate the Storm water Outfall Systems Plan with the Development Plan. Opportunities and Constraints North of Interstate 70, the north-south nmways are the dominant land feature, served by associated drainage swales and detention ponds. Due to highly permeable soils, very little surface nmoff leaves this area. Also, no outfall drainage facilities exist to convey excess nmoff from the site. South of Interstate 70 lie the existing tenninal complex and east-west rllllways. Excess rlUloff from these areas is directed to Sand Creek and Westerly Creek. Through the history of the airport these drainageways were used as single-purpose channels for conveyance of drainage flows through the airport. RlUlway and taxiway bridges were constructed and the channel banks were filled with construction spoils, broken concrete, and debris. Westerly Creek was placed in an lUldersized culvert beneath the east-west nmways. Some areas of contamination exist from fuel spills and industrial activities; however, the impacted areas comprise less than 2% of the site. With the closing of Stapleton as a single-use site, the opportunity exists for redevelopment of the area into a lUlique urban commlUlity. With

PAGE 314

Pflaum, Wenk, and Urbonas 299 over 7 square miles of publicly owned land in the heart of the city, Denver faces the largest urban redevelopment opportunity in its history. STAPLETON DEVELOPMENT PLAN The Stapleton area is now planned to be a lUlique mixed-use commlUlity capable of supporting more than 30,000 jobs and 25,000 residents. The Development Plan organized the site into eight districts, each with an identifiable center and integrated land uses of employment, housing, public transportation, and walkable scale. The open space system comprises over one-third of the site area (in excess of 1,600 acres) and serves a major role in lUlifying the eight districts and providing multi-use flUlctions of drainage, parks, greenway corridors, trails, and natural areas. Stapleton's sustainable development philosophy is characterized by the compact, mixed-use neighborhoods that are walkable and transit-oriented, with infrastructure that stresses water and energy conservation, renewable sources of energy supply, and a storm water management approach that provides opportunities for reuse of nmoff for on-site irrigation and water quality enhancement. STORMWATER MANAGEMENT PLAN The plan for management of stormwater nmoff generated by the new communities echoes the sustainable philosophy of the Development Plan. Excess nmoff will be managed by surface drainage facilities (open channels and ponds) that will be fully integrated with the commlUlities' open space, trails, and recreational uses in a system of greenway corridors. In addition, a nlUllber of best management practices (BMPs) are planned to enhance the quality of all site nmoff, including wetland channels, extended detention basins, and retention ponds with permanent pools. Figure 2 illustrates the proposed Stapleton Development Plan and the storm water management facilities planned to serve the new urban commlUlity. The drainage system is a hierarchy comprising major outfall corridors serving smaller tributary outfall channels that typically combine with or parallel transportation routes. These in tum accept nmoff from smaller, local tributary channels that serve private development parcels, where on-site BMPs are encouraged to reduce nmoff by minimizing directly connected impervious areas and using nmoff to help irrigate buffer strips and landscaping. Channel corridors will provide multiple uses, including wetland and habitat zones, water quality enhancement via infiltration, and a trail network that provides maintenance access and pedestrianfbicycle linkage with neighborhoods and parks/open spaces.

PAGE 315

300 Stormwater Planning for Redevelopment of Denver's Airport North of Interstate 70, where no outfall existed, a major park and open space corridor is planned to provide drainage conveyance to Sand Creek and incorporate a golf course, wetlands, and habitat development, a multi-use trail, and a regional water quality control pond. Tributary channels will provide similar uses and stonnwater quantity and quality detention in planned wetlands and smaller ponds. South of Interstate 70, Sand Creek and Westerly Creek will be restored and revitalized as multi use stream corridors, providing water quality enhancement features (ponds and wetlands), regional trails, and wildlife corridors. Areo Outfoll Nationol Wildlife Refuge Oullaliia land Cre,k Irandal, Oi ... "an .Ilruclure B,I\I', (for Ih, ,orin :'", .100year regional d,lentlO' :co waler qualify trealmenl Open ,pace/recreahan/wddl,f, cOllldor Sond Creek/ReglOnol Outfoll II ReceIVing Walerway Regional outfall 10 PlolI, R'm Regional wddhfe and lrail cOllldor Weslerly Creek/Soulh Areo Ouilo R,,"ving Wolerwoy Oullall 10 land Creek .Ilrudural B,I\I', Park,/lrad, inleg"l,d .1Oy,ar regional d,l"h"!.:,,, quol,ty,nhoncem,nt fribulory Outfall Channels Integral wllh porkwoys! per" Ilrudurol BfrlP, fYPlcollocol fribulory Conveyance slreel flow local,oftbottam channel' 1m all P'P" In R,O,W, ........ I..I..+--+ Privole lond On ,de BliP, Figure 2. Stapleton Development Plan, Stonnwater Outfall Systems plan.

PAGE 316

Pflaum, Wenk, and Urbonas 301 SUMMARY Redevelopment of the Stapleton site is a challenging and exciting task for the City and COlmty of Denver. The proposed surface drainage system is a marked departure from conventional drainage infrastructure, but one that offers significant savings (an estimated $20 million less than typical stonn sewer system costs) and the benefits of multiple use and shared maintenance as an integral part of the development's park, recreation, open space, and transportation systems. It is also one of the first stonnwater outfall plans to incorporate comprehensive use of BMPs for runoff water quality enhancement in accordance with the latest Denver criteria (Urban Drainage and Flood Control District, 1992). By taking advantage of the site's highly permeable soils, treatment and infiltration of excess runoff can begin at the source, with recommended on-site practices such as minimized directly-connected impervious areas, and continue with other structural BMPs along tributaries and outfalls. The Stapleton Stonnwater Outfall Systems Plan represents a comprehensive approach to urban stonnwater management. REFERENCES McLaughlin Water Engineers, Ltd. 1995 Stormwater Outfall Systems Plan-Stapleton Area. Urban Drainage and Flood Control District, City and County of Denver. Stapleton Redevelopment Foundation 1995 Stapleton Development Plan. City and County of Denver and Citizens Advisory Board. Urban Drainage and Flood Control District 1992 Urban Storm Drainage Criteria Manual Volume 3 -Best Management Practices. Denver, Colorado.

PAGE 317

Upper Peaks Branch: Flooding in an Unmapped Area Albert H. Halff Walter E. Skipwith Kevin Shunk Halff Associates, Inc. Ben Cernosek City of Dallas INTRODUCTION The Peaks Branch storm sewer system was constructed in the 1930s to drain 5.7 square miles of East Dallas. The design was based on a Master Drainage Plan and criteria developed by W.W. Homer, a noted St. Louis drainage engineer. The horseshoe and box culvert sewers were sized for the 5-year flood using the Rational Method and Manning's equation. The underground system drains to an open channel that conveys stormwater to White Rock Creek. The system was a great improvement over the network of ditches and small culverts that were responsible for annual flooding of homes and businesses in the area. As the years went by, however, development blocked the emergency overflow paths. Many homes and businesses in the drainage basin thus began to experience flooding from overloaded stonn sewers and overland accumulation in low areas. Recornnlendations for flood relief alternatives were outlined in a 1976 floodplain management report prepared for the City of Dallas. This study focused on the broad floodplain adjacent to the open channel in the lower basin. The recommended plan consisted of channel improvements plus a relief storm sewer system to handle overflows from the upper basin. Channel improvements were constructed in 1984 for $4.7 million. In 1983, design of the first stage diversion relief conduit (double box culvert) was begun; the $8.5 million structure was completed in 1989. These improvements removed almost 800 fanlilies from the regulatory floodplain in lower Peaks Branch and provided the fOlmdation for future relief of the middle and upper Peaks Branch areas.

PAGE 318

Halff, Skipwith, Shunk, and Cernosek 303 In 1989, a plan was made for relief of the middle Peaks Branch area. The recommendation for improved drainage for the Main Peaks Branch trunk stonn sewer was an extension of the existing stage 1 diversion conduit. The proposed extension would consist of a double 10 x 10 foot reinforced concrete box culvert from the end of the existing diversion to a point upstream of Fair Park, the site of the annual state fair and many museums and historical buildings. It was estimated to cost $9.8 million in 1989. The existing Peaks Branch main trunk stoml sewer was also inspected, revealing significant distress in the 60+ year old system. It was recommended that the city repair and renovate the existing system. HISTORY OF FLOODING Past Floods Since construction of the Peaks Branch Stonn Sewer System in 1933, flooding in excess of the 5-year event has occurred on numerous occasions. Complaints on record at City Hall for the Peaks Branch Stonn Sewer indicate that seven significant floods took place between 1931 and 1974. The complaints dealt primarily with street flooding south of Fair Park where the majority of overland flow from the upper basin collects. In 1991, a stoml struck the Peaks Branch Watershed, flooding homes along Alcalde Street. Gauges in Garrett Park measured 4.0 inches of rainfall in 3 hours with l. 7 inches falling in one hour on April 12, 1991. Some homes along Alcalde reported flooding to depths of 30 inches. May 5, 1995 Flood In the late evening of May 5, 1995, a severe stonn raced across north Texas, including many parts of Dallas. Nineteen deaths were reported, many occurring as a result of overloaded stonn sewer systems and flooded low-lying areas. The Red Cross reported flood danlage in excess of $l3.8 million at 317 structures (homes and businesses) throughout the ci ty. This stonn hit Peaks Branch and the Alcalde Street area especially hard. The Garrett Park gauge recorded almost 4.5 inches of rainfall in one hour. (The one-hour 100-year rainfall total for Dallas is 4.0 inches according to the National Weather Service rainfall atlas.) The rain caused dooding at the Starplex Amphitheater, Fair Park Music Hall, businesses along Exposition Boulevard, Dallas Area Rapid Transit (DARn facilities near Main and Haskell, homes in the Alcalde Street area and south of Tietze Park, and several other areas. In the Peaks Branch watershed, 128 residential and commercial structures were flooded, with danlage over $3.8 million. On Alcalde Street, the flood waters were 7.2 feet deep in the street. Eleven duplexes were flooded, some up to almost 5 feet. Other area homes and the nearby elementary school were also flooded.

PAGE 319

304 Upper Peaks Branch: Flooding in an Unmapped Area DART EMPLOYEE PARKING GARAGE In 1988, an parking garage for DART employees was built at Elm and Haskell streets in East Dallas. The garage is owned and operated by DART but was constructed from plans prepared for the Dallas Transit System (City of Dallas) in 1986. Construction was approved and permitted by the city. The DART parking garage actually is built over the existing Peaks Branch Storm Sewer. The area was once the emergency overflow path for excess stormwater. Before the DART facility was constructed, overflows of storm runoff could escape the Alcalde Street area starting at about elevation 471. Since the garage was built, stornlwater must rise to elevation 473.8 before overflowing down Elm Street. This ponds stormwater to a depth of almost 3 feet in the street before emergency overflow from the area can occur. Despite these severe flood problems, the area is not mapped as floodplain by either the City of Dallas or the National Flood Insurance Program (NFIP) Flood Insurance Rate Maps. ALTERNATIVES TO MITIGATE FLOOD DAMAGE Long-Term Alternatives The city immediately commissioned a flood mitigation study for the Alcalde Street area. The study determined that flood losses could best be reduced by constructing either relief storm sewers or detention basins in the upper Peaks Branch watershed. Typically, these types of measures provide permanent solutions to flood problems but often take 2 to 10 years to plan, design, fund, and construct. Relief and Diversion by Closed Conduits Closed conduits are underground drainage systems that convey surface runoff. Nonnally, closed conduits are used in small areas or where open drainage is not feasible due to right-of-way restrictions. This practice is widely used in Dallas. As flows become larger, however, economics usually dictate that the water be conveyed in an open channel or natural streanl bed, if possible. As previously discussed, the lack of drainage planning during the development of East Dallas in the early 1900s precluded open channel or natural drainage for Peaks Branch. Possible routes for relief of the overloaded Peaks Branch Storm Sewer were developed. This alternative generally consists of extending a double 8 x 10 foot reinforced concrete box culvert 3300 feet along one of two routes to intercept stonn sewer overflows. The cost would range from $5.3 to $5.5 million, depending on the route, and would be in addition to the $12 million estimated for building the relief system up to the Eastside/Haskell intersection as proposed in 1989.

PAGE 320

Halff, Skipwith, Shunk, and Cernosek 305 Detention Basins A flood retarding or detention basin reduces the discharge of floodwater by detaining some of the peak flow and releasing only a predetermined amount into the existing drainage system. A proposed detention basin for Peaks Branch could be located in Buckner Park. To eliminate flooding along Alcalde Street for a 100-year event, the basin must contain over 37 million gallons (114 acre feet) of stonnwater-about half as large as Dallas' recently constructed Cole Park Detention Vault. Because of its size, the basin would probably be located underground to preserve the park. Stonnwater would be diverted from the Peaks Branch system and stored until the flooding passes, then slowly returned to the system. A 20,000 gpm punlP station would be needed to empty the basin safely after the flood. A system designed for 100-year flood flows would cost $16.8 million. Even with detention, additional stonn sewer improvements would be needed downstream of Alcalde Street to protect Fair Park facilities. Short-Term Alternatives Short-ternl (within one year) alternatives to reduce flood damage along Alcalde Street include structural and nonstructural measures. Modifications to DART Employee Parking Garage TIle first-floor walls of the parking structure can be opened up to allow the passage of stonn sewer overflows. This requires eliminating at least 350 feet of wall on both north and south elevations. This modification will reduce flood levels along Alcalde Street by 2.8 feet during a 100-year flood. This alternative returns the flood levels to approximately the flood condition before construction of the garage. However, streets will continue to carry large amounts of stornlwater overflows during severe stonn events. Table 1 sununarizes flood elevations for this alternative. Table 1. Nunlber of flooded residences, Alcalde Street. Flood Existing (1995) With Modified Reduction in Magnitude Condition DART Garage Flood Elevation 5-Year 6 0 0.7 10-Year 19 7 l.8 50-Year 21 14 2.6 100-Year 21 14 2.7 May 5, 1995 Flood 21 15 2.8

PAGE 321

306 Upper Peaks Branch: Flooding in an Unmapped Area Local Storm Sewer Improvements Additional inlets and laterals in the Alcalde Street area can be constructed to provide some relief from the smaller, more frequent storms. These inlets should be located along Victor Street south of Carroll and on Alcalde Street itself. They would have little, if any, effect on large floods like the May 5, 1995 storm, but they would definitely be a required part of any relief storm sewer system extended to this area. Flood Alert Warning Systems Flood prediction and early warning systems provide time in which to prepare improvised flood defenses and evacuate flood hazard areas. The City recently installed the Dallas Area Flood Warning and Control System (DAFWC), which automates the Trinity River pwnping facilities, and provides early flood warning for drainage basins adjacent to the Dallas Floodway and ultimately throughout Dallas. Once a warning of a possible flood is received, personnel are dispatched to the area to warn the residents of the potential flood and to barricade the area. The Civil Defense is notified and goes on yellow alert. The Dallas Civil Defense functions primarily as a coordinator for the Flood Plan and helps with the evacuation warnings. If the need arises, the Civil Defense personnel contact the Red Cross for emergency aid. The Dallas police and fire departments are responsible for rescue operations. Future warning stations could be located within the Peaks Branch basin to provide more warning time. To be effective, they should be directly activated by rising flood waters in the Peaks Branch Storm Sewer System to allow maximwn time for evacuation. One disadvantage to such a system is that false alarms may occur because of the very rapid rise of flood waters in the area. Floodproofing The purpose of floodproofing is to reduce flood damage to structures and their contents, if flooding is not prevented by other means. Floodproofing could be used as a temporary measure to protect any permanent structures in the floodplain, such as park buildings. In the Alcalde Street area, floodproofing measures are of limited usefulness due to lack of warning, severity of flooding and the nature of the structures to be protected. Flood Insurance The NFIP requires that flood insurance be purchased for structures flooded by the 100-year flood, before home improvement loans or mortgage loans can be obtained from the federal government or any federally insured, regulated, or supervised lending institution. Currently, the Alcalde Street area and the other 128 flooded structures in the upper

PAGE 322

Halff, Skipwith, Shunk, and Cernosek 307 Peaks Branch basin are not identified in any floodprone area by the Federal Emergency Management Agency (FEMA). Flood insurance helps to alleviate the cost of damage after flooding has occurred, but it is ineffective in correcting or preventing floods. Floodplain Mapping Currently, the Middle and Upper Peaks Branch areas are not mapped. Some type of mapping is needed to delineate above-ground, flood prone areas for Peaks Branch. This would assist city staff with decision making about new development and redevelopment in the area and make the public aware of flooding problem areas. Designation of the area prone to flood during the lOO-year storm as "floodplain" or FP is one alternative. However, this designation has traditionally applied to open creeks and streams and is an area that is specially regarded by FEMA and the City of Dallas. In particular, the city has a special ordinance governing development in a designated floodplain. Properties would immediately be subject to all rules that apply to development in an open floodplain. For instance, any fill, excavation, or storage of materials on these newly designated floodplain properties would be illegal unless property permits were obtained, which can be a lengthy process. The technical criteria that must be met in order to obtain such a permit are currently written to apply to reclamation of tmdeveloped land. Also, property owners platting or replatting property would be required to dedicate the FP area to the city since the code does not allow private ownership of the floodplain. These restrictions could negatively impact property values in a part of Dallas that is attempting to redevelop. Another alternative is to delineate the floodprone area but designate it separately from floodplains. Such an area could be designated a "flood management zone" or FMZ. The FMZ would become the tool used by city staff to guide new development and redevelopment. The FMZ could be referenced to set minimum fill and floor elevations of proposed structures and to maintain a surface overflow path for flood waters. The FMZ could be an official designation like the FP with its own rules and requirements or could be an unofficial designation for use primarily by the city's floodplain management and building inspection staff. Either way there would be fewer restrictions than with floodplain designation. This would also provide a record for the public to be aware of flooding problems and a tool for the promotion of flood insurance. Either alternative would require a technically sound analysis of the aboveand below-ground hydraulics of the drainage system to delineate the area that would be inundated during the 100-year stornl. This would become the basis of the FP or FMZ. The FMZ would also include areas recommended for detention or for the future underground relief conduits.

PAGE 323

308 Upper Peaks Branch: Flooding in an Unmapped Area RECOMMENDATIONS Long Term The recommended long-tenn solution to reduce Alcalde Street flooding consists of extending an underground stonn sewer relief system to the area from an existing relief/diversion system. This relief system will also provide flood protection for the Music Hall and other parts of Fair Park, businesses in the Exposition Boulevard area, and DART facilities in and around the Haskell/Main Street intersection. The system is estimated to cost approximately $17.5 million. Further investigations are needed to detennine the usefulness and cost effectiveness of detention as a supplement to the underground relief stonn sewer. Short Term There were three recommended short-ternl solutions. First, the DART employee parking garage should be modified to allow stonn sewer overflows to pass. This could be accomplished by removing the back and front walls. This solution has been implemented in part by DART. In addition, a warning siren connected to the existing Dallas Flood Alert System should be placed in the Alcalde Street neighborhood to allow residents more time to move vehicles and other possessions when flood waters threaten. Lastly, the city plans local drainage improvements to fully utilize the capacity of the existing truck stonn sewer. None of these short tenn solutions will eliminate severe street flooding in the area for stonns greater then the 5-year return period. Other These recommendations should become part of a comprehensive Peaks Branch Stonn Water Management Plan whose development would include recommendations to reduce flood losses throughout the basin.

PAGE 324

An Expressway, Stormwater Management, and the Environment: A Case Study Ward S. Miller Lake County Stormwater Management Commission Richard L. Thompson T.Y. Lin International BASCOR, Inc. A 20-year-old alignment for a major expressway through a low-lying, largely undeveloped corridor of Lake COlmty, Illinois, provided a challenge to the Illinois Department of Transportation (IDOT) in the pursuit of an acceptable Environmental Impact Statement (EIS) and project. An integral part of this process was the acceptance and implementation of a higher level of drainage and environmental standards than IDOT had been accustomed to in the past. All counties are not created equal when it comes to natural resources. Lake County, the location of the northern portion of the expressway, is blessed with a plethora of wetland complexes, natural stream corridors, lakes, and depressional storage areas. Many of these areas were recently identified in an Advanced Identification (ADID) wetland inventory. Based on the plarming and design experience on the southern portion of the expressway in another county, everyone involved knew that the traditional expressway pI arming process and state design standards would not result in the project's being built. INSTITUTIONAL RESPONSE As part of their plarming process for the Year 2010 Regional Transportation Plan, the Northeastern Illinois Plarming Commission (NIPC), in conjlmction with the Chicago Area Transportation Study (CA TS), agreed to include the construction of this regional expressway in their transportation plan with the understanding that IDOT would agree to pursue the project through an intergovernmental pI arming group and that IDOT would follow the Federal Highway Administration (FHW A) EIS process. IDOT, being in total agreement with these principles, joined with NIPC to form the Corridor PI arming Council (CPC). CPC membership

PAGE 325

310 An Expressway, Stormwater Management, and the Environment consists of the chief elected official of eight of the impacted village governments and an appointed cOlmty board member representing the tmincorporated areas within the expressway corridor. The two key CPC charter provisions are: The CPC would be the focal point for the corridor land use pI arming process and the development of the EIS for the expressway. The commtmities and IDOT would develop plarming and design standards that would be applied to all development in the CPC commtmities, not just the expressway. Also, new intergovernmentally derived future land use plans for the corridor would be developed. The result of this partnership would be a lengthened pI arming process with numerous opporttmities for public input and more stringent standards that everyone had to abide by, especially in the area of natural resource protection and mitigation (Figure 1). Other State Agencies ISTHA Feder-al Agencies lOOT Other County Agencies Puulic / CltlZbri Gr-oups 8 Village Gover-nments Stor-mwuier Management Commission Figure 1. The partnership fomled for dIe plaillung process. CPC began the sometimes laborious task of developing 24 sets of standards on topics ranging from grotmdwater protection to illumination. The five sets of standards related to storrnwater management are:

PAGE 326

Miller and Thompson (1) Soil Erosion and Sedimentation Control (2) Floodplain Protection (3) Stormwater Detention and Drainage (4) Wetlands, Stream, and Lake Protection (5) Open Space. 311 Due to Lake County's history of severe flooding and its natural resources, impacts to the natural drainage system and design of the proposed drainage system becanle the focal point of expressway design. At about the same time the CPC was formed, the Lake County Stormwater Management Conunission (SMC) was created to develop and implement a unified, county-wide program. During the same period the CPC was drafting standards, the SMC was adopting the Comprehensive Stormwater Management Plan and drafting the Watershed Development Ordinance (WDO), which established the minimum cOlmty-wide development standards related to the CPC topics listed above. Upon the effective date of the adopted WDO, all public and private development, including local road building, had to abide by these new standards. At this point, the CPC had not yet finalized the related drafts of the standards. After a comparative analysis between the WDO standards, roOT design standards and the initial drafts of the CPC standards, the CPC voted to adopt the WDO standards acknowledging they afforded a higher level of natural resource and drainage system protection. Midway through the enviromnental assessment process, the Illinois State Legislature passed legislation giving the Illinois State Toll Highway Authority (ISTHA) construction responsibility for the proposed facility. With ISTHA' s more readily available flmding mechanism and record for moving quickly to construct these types of facilities, this legislation served to accelerate the consensus-building process. roOT continued to oversee the EIS, through the CPC, now in partnership with ISTHA. The CPC provided a formal mechanism by which the "will of the people" could be expressed through local elected officials. roOT wrote a letter pledging conformance with the WDO on this expressway. ISTHA executed an intergovermnental agreement with the SMC pledging conformance with the WDO on this project and other road building projects in Lake County.

PAGE 327

312 An Expressway, Storrnwater Management, and the Environment NEW PLANNING AND DESIGN STANDARDS Crucial to the success of the project was institutional responsiveness and the willingness of all parties to abide by the CPC standards. Early in the process mOT authored a comparison of their standards and the WDO standards. Some of the more significant quantitative and qualitative differences identified in the comparative analysis are shown in Table 1. The differences in these standards have significant impacts on the planning and design of a highway. A far greater number of drainage crossings are considered as floodplain, with a definition standard that is approximately six times more stringent. Compensatory storage for floodplain crossings increases dramatically as there is a requirement for compensatory storage for all fill in the floodplain, often warranting the construction of a bridge to span the entire natural floodplain. Right-of-way requirements expand with the need to provide greater detention and compensatory storage. A secondary benefit is more aesthetically pleasing open space that more adequately buffers the expressway from adjacent land uses. Right-of-way requirements are further expanded by the need to retain the first 1/2 inch of runoff and more restrictive release rates for constructed areas, resulting in an increase in storm water detention requirements and different BMP facility design techniques such as wetland and forebay features. Enlarged wetland mitigation areas were required to reflect the higher ratios necessary to replace Lake County's higher quality wetlands. Lastly, and from a qualitative point of view, mOT would normally design a drainage system that would remain within their normal linear right-of-way, be economical to construct, and present the simplest needs for maintenance. This would frequently result in piped drainage systems, in-line pipe stonnwater detention, and uniform and oftentimes lined channels. However, in the planning process for this project, mOT considered innovative watershed, floodplain, and water quality design options that accommodated the requirements of the WOO. Some of these alternatives included terraced embankments, replication of sheet flow across the right-of-way, gravel filter walls, and the provision of shallow interconnected retention and detention facilities in the median of the proposed facility. The WOO encourages use of open drainage systems, multi-purpose retention/detention/BMPs, and replication of existing drainage patterns at the sub-, sub-basin level. The different standards resulted in much different considerations during the planning process. The qualitative effects of these differences are depicted in concept in Figure 2, presenting the typical appearance by mOT standards before this project as compared to the WOO/CPC standards mOT agreed to implement.

PAGE 328

Miller and Thompson 313 Table 1. Comparative analysis of mOT and WDO criteria. TOPIC Floodplain Definition Floodplain Compensatory Storage Depressional Storage Area Compensatory Storage Water Quality Storm water Detention Release Rates Wetland Buffers Drainage System Design Focus IDOT CRITERIA Drainage area greater than one square mile (urban) 1: 1 for fill in the riverine floodway None proposed Temporary and permanent erosion control and siltation measures Rate of flow before development None Most economical and maintainable system, without creating impacts as measured by their past design standards WDO CRITERIA Drainage area greater than 1/6 square mile 1.2: 1 for fill in the riverine floodway and floodplain 1: I for fill in non-riverine floodplain ErosiOn/sediment control plus the retention of the first 1/2 inch of runoff before discharge into lakes, ponds, or wetlands or other effective BMPs 0.04 cfs/acre for 2-yr stoml 0.15 cfs/acre for 100-yr storm (Usually regarded as less than pre-development conditions) Minimum of 30 feet, utilizing native vegetation Preservation of the natural components of the drainage system, replication of existing drainage patterns

PAGE 329

1 c1 o 0 0:: I CD += nCI -CD Il..+-CD --:---1-' --1----J River Floodway I \0 age AI ea I I I-I I I I I I I Before Wl co+ ___ --t _+_0 coI 1--1 -------------------... .---Floodplain -------I -_--, --I I, I ,,-,0 I De slonal I' tlo age 0". fA ea :_-:. _<0_ _ I c o +-CD OOl C/JO CL I I I I -I ,,-,0 I I I, 0,,-0<:' I I I I -'---< -1---River .-1-----.. Floodway I I I al.1-lf'-r (1)0 a.+E:V) ,8 -C o+-.---:;:; 0.--, (1J S!' Floodplain o ----o 0:: 1 1 1 1 1_0:: ______ After Figure 2. The typical appearance by mOT standards before this project as compared to the WDOjCPC standards IDOT agreed to implement. (.0) ..... ::l -C CiJ til ':< !B o (I) .... til ::l til <0 (I) 3 (I) ::l --til ::l a. (I) m cf ::l 3 (I) ;a.

PAGE 330

Miller and Thompson 315 LESSONS LEARNED Many valuable lessons were learned from this experience that may be applied to similar expressway planning projects. New standards led to a more complex, multi-faceted planning process that resulted in institutional change. The changes reflected heightened sensitivity to environmental issues and local government input. The CPC was an effective vehicle for consensus-building. Expressway planning from a watershed perspective rather than a "within right-of-way" perspective will result in more project land acquisition but greatly reduced watershed impacts. The state road building institutions had foresight in acknowledging that typical state design standards needed to be customized to unique natural resources. The open planning process requires much longer time periods. A typical highway section would not have been appropriate throughout the length of this project due to varying environments along the corridor. On a segment-by-segment basis, the cross section had to be tailored for that segment's environment. The "micro" drainage system impact analysis and mitigation design requires much higher funding levels for planning and design. Increased mitigation and compensatory measures require much greater right-of-way needs and more attention to long-term maintenance. ,. A focal point for every alignment iteration was the trade-off between people displacement and wetland preservation. In the face of a formal, agreed-upon mechanism for local government input, the state road building institutions demonstrated great flexibility and adaptability to the lmique natural resource envirorunent. Early and frequent involvement by the myriad of review/regulatory agencies greatly improved the consensus-building process. Having already-agreed-upon county-wide stomlwater standards (the WDO) helped make drainage the focal point in planning and design. Having a cOlmty-wide interjurisdictional institution such as SMC, dedicated to a comprehensive approach to stormwater management, helped the consensus-building process. The partnership of the road building agencies, conmlunities, and the cOlmty (i.e., everyone agreeing to abide by the same elevated standards) was the essential ingredient in the success of the process.

PAGE 331

Kyle Canyon Detention Basin: Conception to Constructi on Ken Gilbreth VTN Nevada Kevin Eubanks Clark County Regional Flood Control District INTRODUCTION The Las Vegas Valley and Clark County have a long history of flooding and flood damage. The Las Vegas Valley is unique in that it is surrounded by mountain ranges with steep slopes that empty onto alluvial surfaces. Ultimately, stormwater runoff has to pass through areas that are being rapidly urbanized. The steep slopes and unpredictable flow paths on the alluvial fan surfaces compound the flooding and engineering problems facing developers and the Clark County Regional Flood Control District. The problems also include the possibility of flood waters transporting tremendous amounts of debris and sediment. In the urbanized areas of the Las Vegas Valley, development and pavement of the desert increases direct runoff and speeds its flow. It is difficult to convince newcomers that the threat of severe flooding exists in a desert region that receives only 4 inches of rain annually. Since the 1960s, the Las Vegas Valley has experienced unprecedented rapid growth. In response to severe floods and the ever-present threat of future flooding, the Clark County Regional Flood Control District was formed by the Nevada legislature in 1985 to develop a coordinated and comprehensive flood control master plan to solve flooding problems, to regulate land use in special flood hazard areas, to fund and coordinate the construction of flood control facilities, and to develop and fund a maintenance program for flood control facilities. The Clark County Regional Flood Control District administers programs that include master planning, capital improvement progranuning, Corps of Engineers cooperation, regulatory programs, flood warning, environmental mitigation, public education, and operation and maintenance. Funding for the District's programs is derived from the 1/4 of one percent sales tax.

PAGE 332

Gilbreth and Eubanks 317 The District was the first to develop a comprehensive master plan that not only takes into accOlmt existing development, but also addresses the probable effects of future development. The master plan for the Las Vegas Valley includes $900 million worth of the various forms of flood control facilities. The District covers all of Clark COWlty, with a majority of District projects located in the Las Vegas Valley. Individual master plans are developed for each of Clark COWlty'S outlying areas as well. By statute, the master plans must be updated every five years to consider the progress of the capital improvement program and private development. In our first year we received approximately $15 million in sales tax revenues. This past year we received approximately $35 million. These revenues are dedicated primarily to the capital improvement program for the construction and maintenance of flood control facilities and other District programs with less than 10% going toward District administration. In 1990, we issued $80 million in bonds so that we could accelerate construction of several needed facilities. Kyle Canyon Detention Basin was one of those projects. We have nearly completed all of the projects on our bond list and are now receiving some major flood protection benefits that didn't exist just three short years ago. To date, we have spent nearly $245 million on the projects in our master plan. The capital improvement program has been developed and is reviewed annually. The District adopts a lO-year construction program for the needed facilities. These improvements include detention basins, channels, stoml drains, and bridges. Six governnlental entities within Clark COWlty use District fWlds to implement the master plan. They include Clark Cowl1y and the cities of Las Vegas, North Las Vegas, Henderson, Boulder City, and Mesquite. Each of the entities within Clark COlmty takes the lead with respect to capital improvement programming within each hydrographic basin. According to our policies, each entity must consider 10 rating factors in assigning construction priorities when developing the 10-year construction program. The factors include population affected, assessed value of the land impacted, public perception of need, emergency access and public inconvenience, cost avoidance, availability of other fWlding sources, interrelationship to other projects, timing and implementation, envirorunental enhancement, and annual maintenance cost. THE KYLE CANYON DETENTION BASIN The Kyle Canyon Detention Basin is the largest and most expensive flood control facility the District has fWlded to date. The project lies within the Northern Las Vegas Wash hydrographic area. The City of North Las Vegas is responsible for capital improvement programming in this area

PAGE 333

318 Kyle Canyon Detention Basin and also took the lead in administration of District funds for design, right of-way, and construction of the project. Design Summary The Kyle Canyon Detention Basin is located in the northwest part of the Las Vegas Valley on land administered by the u.s. Bureau of Land Management (ELM) in Sections 14 and 23, Township 19 South, Range 59 East. The basin is designed to intercept the flow from the Harris Springs Wash and the Kyle Canyon Wash with a combined total of 57 square miles of watershed. The BLM land is located adjacent to a wildlife study area to the west of the proposed detention basin location. In addition, the BLM land east of the wildlife study area contains numerous mining claims and an application for a Native American Indian allotment. The siting of the detention basin includes locations that would bypass the mining claims or would have minimal impact on them while still providing the same level of flood control protection that was originally planned for the basin. The Harris Springs Wash Basin contains 49 square miles or 85% of the watershed and the Kyle Canyon Wash contains 8 square miles or 15% of the watershed. These subbasins correspond to the major valleys and ridges that follow the geologic formations in the Harris Springs Canyon. The distribution of soils and the configuration the drainage network is strongly affected by the prevailing geology and a more refined delineation of the basin characteristics. Hydrology Table 1 slUllllarizes the peak inflow and volume to the detention basin. Table 1. Peak inflow and vohmle to the detention basin. Storm Event Inflow (cfs) Volume (ac-ft.) 2 168 40 5 2,301 434 10 4,620 930 25 7,447 1,573 50 10,285 2,231 100 13,215 2,918

PAGE 334

Gilbreth and Eubanks 319 Sediment TIie Kyle Canyon Detention Basin receives sediment yield from two major sources, the Harris Springs Wash (49 square miles) and the Kyle Canyon Wash (8 square miles). The Kyle Canyon Detention Basin was designed with a 100-year flood sediment yield plus a 5-year expected yield, for a total of 210 acre-feet. PMF/SpiIlway The PMF calculations for the Kyle Canyon Detention Basin were determined to be approximately 123,000 cfs. The spillway for the detention basin is designed for conveyance of the PMF with 8 feet of head, plus 1 foot of freeboard. The spillway is an ogee crest made of conventional concrete that caps a stepped roller compacted concrete spillway. Low Level Outlet The low flow outlet consists of a 72-inch RCP that will convey the peak flow of 366 cfs during a 100-year storn1 (Table 2). The basin is designed to drain in approximately 7 days. Table 2. Peak flow and stage for the outlet. 2 155 3,245 5 240 3,255 10 278 3,262 25 313 3,269 50 341 3,275 100 350 3,281 Dam The Kyle Canyon Detention Basin is approximately 8,500 linear feet long with a maximum height of 55 feet. The low level outlet elevation is 3,234, with the spillway crest at 3,281 and the top of dam at 3,290 (Table 3). The upstream slope of the dam is 4:1 and the downstream at 3:1. The

PAGE 335

320 Kyle Canyon Detention Basin dam embankment contains approximately 2 million cubic yards, including a drainage blanket and toe drain system. Table 3. Stage elevation, volume, and area for the detention basin dam. 3,234 3,242 32 12 3,245 80 21 3,255 477 61 3,265 1,308 103 3,275 2,453 127 3,281 3,256 147 3,290 4,638 167 Construction The construction of the Kyle Canyon Detention Basin took less than one year (May 1994 to April 1995) and was constructed $900,000 under budget. The project team emphasized a "partnering" approach between the Clark County Regional Flood Control District, the City of North Las Vegas, VTN Nevada, and the contractor. The following are volumes of selected materials used during construction: 3 million C.Y. of dirt moved; 122,000 c.Y. of RCC; 87 million gallons of water used; and 18,000 C.Y. of soil cement.

PAGE 336

Part 11 CONSTRUCTION TECHNIQUES, BUILDING PERFORMANCE, AND DATA COLLECTION

PAGE 337

This page IS intentionally blank

PAGE 338

Manufactured Home Foundations: A Summary of Current Studies William L. Coulbourne Greenhorne & O'Mara Cecelia Rosen berg Federal Emergency Management Agency INTRODUCTION The Federal Emergency Management Agency (FEMA) and Greenhome & O'Mara (G&O) have investigated many manufactured home foundation systems that failed during severe stonns. These foundation failures occurred because the homes were not elevated and anchored to resist flotation, collapse, and lateral movement as required by the National Flood Insurance Program (NFIP) regulations. The reasons for this apparent lack of regulatory compliance include institutionalized installation practices in the industry, uncertainty at the local level regarding foundation/installation designs and teclmiques that would meet NFIP perfonnance criteria, and the difficulty of determining whether a manufactured home is compliant when portions of the fOlmdation system are buried under the home. Foundation evaluations have been requested by local officials, FEMA field personnel, members of the manufactured home industry, and manufacturers of proprietary foundation or home support systems who want to detem1ine whether specific foundations would meet NFIP criteria. In an effort to provide sound engineering guidance and regulatory interpretation to those involved in installing manufactured homes in Special Flood Hazard Areas, FEMA has requested that G&O evaluate manufactured home foundations currently in use. This evaluation will include an assessment of the recent flood damage to manufactured housing in Washington state and will lead to the development of engineering guidance for an "all hazards" approach to more prescriptive foundation installation teclmiques that manufactured home installers, homeowners, dealers, and local building, pI arming, and zoning officials can understand and follow.

PAGE 339

324 Manufactured Home Foundations: Current Studies CURRENT FOUNDATION TECHNIQUES The type of foundation on which a manufactured home is installed is largely detern1ined as much by local practice as by site conditions; soil types; wind, flood or seismic hazards; or other engineering considerations. The conventional manufactured home foundation consists of dry-stacked concrete block piers, each on a minimal footing. The blocks are stacked to a height deemed appropriate. It is frequently necessary to place wood shims between the top of the pier and the home's chassis or structural frame for the purpose of leveling the home. The frame of the home is often not anchored either to the block pier or to the ground, even in floodplains. Other foundation types that appear to be in widespread use in floodplains include dry-stacked concrete block piers accompanied by an anchoring system (either grOlmd anchors or concrete deadmen attached to the home's frame with straps) that provides resistance to overturning, concrete blocks reinforced with steel and filled with mortar, wood piles driven into the ground that form a "saddle" to hold the home, and a variety of proprietary methods, including driven piers, concrete-filled bags, and steel frames bolted to concrete footings. ENGINEERING GUIDANCE CRITERIA The primary reason for evaluating existing fotmdation designs is to develop engineering guidance, including designs that are "pre-engineered" and thus would not require significant, if any, additional site-specific engineering. With the goal of developing designs and other engineering guidance that can be used nationally, parameters have been established to focus the engineering effort. Pre-engineered designs must consider "all hazards" so that they will be suitable for an appropriate combination of loads. The designs must also meet NFIP regulatory requirements and be "economica1." The definition of economical has yet to be determined but would incorporate the concept that the cost of the pre-engineered foundation must not be significantly higher than the cost of foundation systems for non-floodplain installations. The pre-engineered designs will use current building code and design standards to provide engineering guidance while considering logistical and cost issues of importance to home owners. TECHNICAL APPROACH Exclusions In order to further focus the development of engineering guidance, it is necessary to define the conditions under which the forces expected to act on a manufactured home are so great as to preclude the use of a pre-

PAGE 340

Coulbourne and Rosenberg 325 engineered design. Consequently, a pre-engineered design is considered inappropriate for manufactured homes in the following situations: flood waters are expected to be above the floor of the home during the 100-year flood (that is, the floor of the home is below the base flood elevation), flood velocities are expected to exceed 5 feet per second (fps), the home is in a coastal V Zone, sustained wind speeds are expected to exceed 110 miles per hour (mph), the home is in a seismic zone where the snow load is greater than 20 pounds per square foot (psf), the soil bearing capacity is less than 1000 psf, or the home is on an alluvial fan. In these situations, a structural engineer should design a foundation system specifically for the conditions at the site. Inclusions The paranleters used to assess the effects of various flood depths and velocities currently include the following: the dead load of the home is assumed to be 25 psf; the impact effect of a 1000-pOlmd object striking a home or its foundation during a flood has been considered; the effects of wind speeds up to 110 mph are being studied; wind and flood forces are assumed to act simultaneously. The approach is to consider overturning moments and lateral forces that act on a manufactured home at various wind speed'S (up to 110 mph) and at various flood velocities (up to 5 fps) and flood depths (up to the top of the home's floor). A factor of safety of 1.5 is used in the calculation of moments and lateral forces. A working load capacity of 1000 pOlmds is used for helical anchors; t1lis value was taken from a study of anchor capacities done by Wiss, Jarmey, Elstner Associates in 1991. This anchor capacity is significantly lower than the 4725-pOlmd capacity required by t1le Department of Housing and Urban Development. Figure 1 shows the forces that are applied to a manufactured home during an event that creates high wind and water. The resultant moment and lateral forces will dictate the type of restraint necessary to keep the home on its fotmdation; however, there are only two engineering choices for restraint: the use of anchors and straps capable of resisting the overturning moments and lateral forces, or the use of a rigid fotmdation and a rigid foundation-to-home connection that will resist these forces. PRELIMINARY FINDINGS The following is a smnmary of the findings to date: When a manufactured home is at or above BFE, the wind forces that act on the home are much greater than the flood forces acting on it.

PAGE 341

326 WIND Flood Level \ WATER Manufactured Home Foundations: Current Studies WIND UPLIFT i LOADS l House WIND UPLIFT i LOADS l ... --. LEEWARD W1;:-1D SLIDING Grade Figure l. Forces on a manufactured home. Manufactured homes are designed for wind resistance but not for resistance to buoyancy (which reinforces the first finding). Assmning a 1000-pound capacity per anchor, the height limit for a home on an unreinforced pier is 24 inches above grade when the home is in an 80-mph wind zone. Current foundations in use seem to largely ignore soil bearing capacities and frost depths.

PAGE 342

Coulbourne and Rosenberg 327 Disaster experience has shown that manufactured home installations are not often inspected by local officials for compliance with either the manufacturer's installation instructions or with local codes and ordinances. Of the seven proprietary systems considered to date, only two nearly meet the moment and lateral force requirements established for pre engineered designs; also, many proprietary systems have a limited height range. RECENT FINDINGS FROM WASHINGTON STATE Recently, FEMA assessed flood damage sustained by manufactured homes in Washington state during flooding that occurred in the winter of 1996. Approximately 80% of the 400 buildings damaged were manufactured homes. The conclusions from the assessment are that the primary flood damage to manufactured homes was from buoyancy and from the lateral force of flood waters that pushed homes off their foundations. Sometimes these two forces acted together in such a way that the flood waters floated a home enough to reduce the weight on the foundation, allowing minimal lateral pressure to push the home off its support. Most of the damaged homes were installed before Flood Insurance Rate Maps had been issued for the affected communities, and they were installed without permits or inspections. For many of the danlaged homes, no anchors had been installed. When damaged homes clid have anchors, either the anchors were the wrong design for the soil type or they had been installed incorrectly. There were many situations where the flood flow velocity was high enough to cause scour that undermined the foundation or the anchor and ultimately caused failure. The observations and conclusions from the Washington state assessment suggest the following reconmlendations: When a manufactured home is installed, it should be elevated high enough that its floor assembly and structural frame are above the BFE. The NFIP regulations require that, at a minimunl, the top of the floor be at the BFE. If only this minimunl amount of required elevation is provided, the floor assembly and the structural frame of the home (which have a combined height of approximately 18 inches), as well as the bottoms of the home's walls, will be below the BFE and will be subject to inundation and flood flow forces during the 100-year flood. Manufactured home wall and floor assemblies are not designed to withstand either inundation or flow forces; therefore, it is recommended that manufactured homes be elevated above rather than to the BFE.

PAGE 343

328 Manufactured Home Foundations: Current Studies If ground anchors are used, they must be of the appropriate design for the type of soil at the site and must be installed correctly. Procedures must be in place for ensuring that both requirements are met. Anchors must be designed for all hazards and be of sufficient size and number to resist forces from flood, wind, and seismic events. Foundation depth must be below the level of scour expected during the design event (lOO-year flood). Foundations must be designed not to exceed the load bearing capacity of the soil. Inspections by local officials should be more rigorous. Such inspections would improve compliance with manufacturers' installation instructions and local codes and would help improve the installation techniques of local contractors. FEMA's findings to date have confirmed that damage from severe storms can be reduced if the NFIP regulations are followed. Engineering guidance for pre-engineered solutions to manufactured housing foundation needs can be developed as long as the number of variables studied is limited to those with wide applicability. Ordinary, good engineering practice (such as not exceeding the soil bearing capacity) still must be followed. The participation of local officials in the pemlitting and inspections of manufactured home installations will help ensure compliance with not only the NFIP requirements but also the appropriate state and local codes, if any. REFERENCES Federal Emergency Management Agency 1996 Installation and Anchoring of Manufactured Housing in Washingtoll State. Washington, D.C. Wiss, Janney, Elstner Associates, Inc. 1991 Testing of Soil Anchors and Strapping for HUD. WJE Report No. 901798. Washington, D.C.: Departnlent of Housing and Urban Development.

PAGE 344

Step by Step-Hand in Hand: Bringing Slab Elevation and a Technical Video to South Louisiana Patricia M. Skinner Fred E. "Gene" Baker Louisia na Cooperative Extension Service INTRODUCTION When a group of flood victims in Denham Springs, Louisiana, asked the Amite River Basin Drainage and Water Conservation District to help them bring a slab elevation contractor to Louisiana to raise their homes, the odds were against them. The only known flmding source-the Federal Emergency Management Agency's (FEMA's) Hazard Mitigation Grant Progranl (HMGP)-was limited. Their homes were 1600-4000 square feet, and the project would therefore be expensive. In addition, there was no experience at the state or local level in administering a program for retrofitting privately owned buildings. The homeowners were detennined not to be flooded again and to find government assistance to ease the cost of retrofitting their properties. TIley had investigated and rejected the alternative methods of removing the structures from their slabs and elevating the slabs by suspending them from beams placed through the interior of the house. Having discovered non-invasive slab elevation at a National Flood Insurance Program (NFlP) biennial conference, they had gone to Florida to observe the work, and had even brought the contractor to Denham Springs to give estimates. This initial work was done in 1992, with the intention of developing a pilot project for the NFIP mitigation program, a program which has since been authorized but still not implemented. Complementing the homeowner detemlination were the needs and wants of several logical partners. TIle District needed to begin work on the nonstructural component of its recently adopted Hazard Mitigation Plan. The State Hazard Mitigation Office was interested in launching a program of retrofitting individual structures; however, being committed to enacting such progranls indirectly, they needed an applicant. The city was

PAGE 345

330 Slab Elevation and a Technical Video anxious to help its residents and would derive Community Rating System (CRS) credit from the program. Although the interest among the partners was high, there was literally no experience in the state with developing or administering such a project. In fact, there was no guarantee, beyond contractors' assurances, that the elevation technique was transferable to Louisiana soil conditions and residential construction standards. EDUCATION PROVIDES A KEY BENEFIT Clearly, the general lack of specific knowledge presented opportunities for education. In the area of education, experience was not lacking. A connection between the District and the Louisiana Cooperative Extension Service (LCES) had been established when the head of the LCES engineering project served as a technical adviser to the District's Hazard Mitigation Plan. Recognizing the extraordinary educational value of local examples of any construction technique and the need to address flooding on slab-built stmctures, he helped the District's program manager (principal author) and Floodproofing and Mitigation Assistance Committee extend the scope of the project to include education. Through LCES, with its 20 years of experience in flood recovery and floodproofing education, the District was able to define the benefits of results-demonstrations and fonnal education. The final proposal, funded through HMGP, included elevation and restoration of five floodprone properties and an educational program. The goals of the project were to reduce losses on the five repetitive loss structures and, through education, to increase floodproofing by elevation in Louisiana. The education program was targeted at flood victims who might use the technology and at Louisiana housemovers who might adopt the new technique, thus making it available locally. Seminars were also conducted for Extension Service agents and for local and parish emergency and floodplain managers-people who would have the oppommity to influence future floodproofing decisions. Publications, scripted slide sets, and a video were included as deliverables to provide tools for future training and public education activities. Because the Extension Service was fonnally involved in the project, it could use its array of educational outlets to draw attention to the project, including press releases and video news releases that were aired statewide and, in one instance, nationally. RELATIONSHIPS AMONG PARTICIPANTS Arrangements among participants were fomlalized in contracts and letters of agreement. These relationships, shown in Figure 1, define the method of meeting tl1e 50% non-federal funding requirement of $277,060. The

PAGE 346

Structure of Relationships FEMA Region VI HMGP Project in the Amite Basin I I LaOEP Hazard Mitigation Office I HMGP Grantee I Amite River Basin Drainage and Water Conservation District HMGP Sub-Grantee I I I 1 I Extension Service McKee and Deville, Inc. Homeowners City of Denham Springs Education Program Foundation Designs and Inspection Control of Constmcnon Fee waivers, site surveys, legal Contract + In-kind Service Contract + In-kind Service Responsible for Cost Overruns services, debris removal Written agreements Letter of commitment I Contractors Physical work on properties Contracts with homeowner ONL Y Figure 1. Relationships among individuals and agencies in the elevation demonstration project. Monetary value of services provided by each entity under OEP contributed toward meeting the HMGP matching funds requirement. Discounted services provided by Eustis Engineering, the LSU School of Architecture, and community members are omitted from the diagram. S ::3 til ::3 Q.. .., .....

PAGE 347

332 Slab Elevation and a Technical Video District's financial conunitment to cash expenditure on the project was $50,000, which would be paid as contracts to McKee and Deville Consulting Engineers (for engineering the foundations and inspecting the elevation work) and the Extension Service (for the educational program). Also contributing to the non-federal share were $20,000 in services from the City of Denham Springs, $21,400 in architectural services from the Louisiana State University School of Architecture, and a total of $23,230 in in-kind services from the Extension Service and McKee and Deville. The most critical relationship, that between the homeowner and the District, was patterned after the Corps of Engineers Dry Creek Project (US ACE National Flood Proofing Conmlittee, 1993). In this project, the Corps used a non-standard approach that reduced administrative costs and maximized homeowner involvement and satisfaction by allowing the homeowners to control most aspects of the work done on their properties. The District adopted a similar approach with two notable differences. First, it minimized the need for construction financing by making interim progress payments to homeowners instead of lunlP sum payments on completion. Second, it facilitated negotiations with the elevation contractor. Since there was no local competition and the job was relatively small, the homeowners had to agree on one contractor for the elevation work. Beyond that, the District approved the contracts submitted by the homeowners and guaranteed payment from grant funds if work was completed as detailed by the owner. No work contracts were issued by any public body and responsibility for contractor performance, and for any and all cost overruns, rested with the homeowner. Based on their own estimates for completing their elevation projects, the homeowners agreed to spend, collectively, $162,430-the difference between the required match and the amount that could be obtained from other sources. Without that expenditure or value of service (if discounted to them), they would not qualify for the grant funding. The local conmlunity pitched in: 51mburst Bank waived fees for homeowners who needed loans, Eustis Engineering provided soil borings and load analyses at half its usual rate, and local notaries donated their service for certifying contract documents. While these offerings seldom find their way onto organizational charts and represent a small fraction of total project costs, they are great for morale. CONSTRUCTION DIFFICULTIES AND ESCALATING COSTS The realities of getting a capable contractor to leave his home market delayed implementation of the District's elevation project for 15 months. In all, by the time construction began, almost two years of inflation had made the budget figures obsolete. On top of that, the contractor worked through the middle of winter and, as it turns out, one of the wettest

PAGE 348

Skinner and Baker 333 Januarys on record. These adverse conditions, compounded by the contractor's lack of familiarity with Louisiana floodplain soils, resulted in substantial dirt-handling cost overruns. Although the state Office of Emergency Preparedness (OEP) had not agreed to any direct financial participation in the project, it did provide relief to the contractor for this situation. It was an expensive lesson, but local contractors saw the technique succeed even in the worst of conditions. Those who have since adopted the technology have done so with a clear understanding of the impact of weather and soil conditions. MEETING THE STATED GOALS The project took longer and required more effort and expense than any of the proponents had envisioned. Only personal detennination and commitment held it together. Fortunately, the shadow of despair fell on the participants in turns, and not on the whole group at once. Each participant, at some point, was ready to throw in the towel and forego personal benefits, but none was willing to deprive the partners of their benefits. In the end, the goals were achieved. Each homeowner in the project has been freed from the traunla of flooding. The project homes, each of which had flooded three times in 15 years, are not expected to draw any more flood insurance claims; FEMA will recover its investment in only two floods at each property. The City of Denham Springs can take CRS credit for having five of its repetitive loss properties retrofitted. Through this project, both the demand (educated homeowners) and the supply (educated contractors) for floodproofing of homes by elevation have been created in south Louisiana. In the first major flood in southeast Louisiana since this project began, local governments submitted proposals for elevation of homes, and Louisiana contractors bid on those slab elevation projects at costs that can meet required benefit-cost ratios. TIle goal of making this technology available in and to Louisiana has been achieved. In that achievement the Louisiana Cooperative Extension Service has served its mission of helping the people of Louisiana improve their lives through education. The project is over, but education goes on. Locally relevant educational materials are now available through the Louisiana Cooperative Extension Service. These tools include a technical slab elevation video, slide shows for homeowners and contractors and printed publications for both those audiences. With these tools in the hands of the Extension Service statewide adult education network, and with the continued support of OEP for nonstructural measures and education, many more flood victims will become aware of their personal responsibility for flood

PAGE 349

334 Slab Elevation and a Technical Video protection and will be able to make informed decisions about their floodproofing options. On the technical side, information was gained in this project that could not have been obtained by sending a film crew out of state. Now that we know the technique works on Louisiana soils, we can turn our attention to making it more cost effective. We now have a much better idea of when to recommend slab elevation. THE ULTIMATE SUCCESS-A CHANGING LANDSCAPE There have been a nllll1ber of developments during the course of this project that will further contribute to the successful adoption of slab elevation technology. The HMGP has been modified to provide more funding for mitigation and at the more favorable ratio of 75:25. The State Hazard Mitigation Office has developed a scoring mechanism to evaluate elevation proposals in competitive funding situations. Not insignificantly, individuals and local governments contemplating elevation projects now have several in-state sources of first-hand experience to whom they can turn for advice and information. When the Corps of Engineers develops a flood reduction plan for an area and that plan calls for elevation, there is now a place they can send the local officials and residents to see examples of the work. The availability of local examples also helps to make lenders more comfortable with loaning money for this procedure. In a state that has over 15,000 repetitive loss properties and in which 70% of the homes are of brick-veneer, slab-on-grade construction, elevation of slab-built homes with the slab has gone from being "unheard of' in 1993, literally and figuratively, to being "locally available" in 1996. REFERENCES U.S. Army Corps of Engineers National Flood Proofing Committee 1993 A Flood Proofing Success Story along Dry Creek at Goodlettsville, Tennessee.

PAGE 350

Performance of Flood Proofed Structures Tested by Floodwater Larry S. Buss u.s. Army Corps of Engineers INTRODUCTION A considerable amount of information on flood proofing exists. This information is generally in the fom1 of brochures, booklets, or reports describing the various flood proofing measures, where the measures should be used, and how to design a flood proofed structure. The u.s. Army Corps of Engineers' National Flood Proofing Committee (NFPC) has recognized the need for infom1ation that describes how flood proofing measures perform when they are actually tested by floodwater. The NFPC originally solicited such information from numerous federal and state agencies and other organizations. This solicitation, however, resulted in little information. As a result, the NFPC decided to seek infom1ation itself by visiting flooded areas across the United States, searching for flood proofed structures within those flooded areas, and inspecting the structures to see how well the flood proofing measures performed. Because of funding limitations, obviously not every flooded area has been visited. The NFPC will document the results of its information-gathering effort into a report. The report will present case studies of flood proofed structures and will describe how floodwater affected the structures. With each specific case described, a "lesson" will be presented; it will briefly describe what worked and what did not. This paper discusses the information-gathering project and the "lessons" learned by observing flood proofed structures. DATA COLLECTION Ten basic floods have been used thus far as the basis for data collection. Clive, Iowa-May 1986 Central Michigan-September 1986 Crystal City, Minnesota-July 1987 Montgomery County, Texas-May/June 1989

PAGE 351

336 Performance of Flood Proofed Structures Central coast, South Carolina-September 1989 Central Iowa-summer 1993 St. Louis, Missouri, vicinity-summer 1993 Southeastern Texas-October 1994 Florida panhandle-fall 1995 Eastern Pennsylvania-January 1996. Data collected to date range over a number of years and include both riverine and coastal flooding. Data prior to 1993 were taken from four flood damage assessment reports developed by URS Corporation for the Federal Emergency Management Agency (FEMA). Sites included in these reports have not been visited by a member of the NFPC. Lessons learned from data collected at these sites were developed by an engineer reviewing the data at each structure based on the effectiveness of the flood proofing measures. Subsequent to 1992, all data were collected by theNFPc. The data collection method was simply to keep informed about significant flooding events across the United States. Upon occurrence of flooding, telephone calls were placed to local Corps of Engineers offices to deternline the likelihood of flood proofed stmctures in the flooded areas. With the information, a decision was made whether or not to visit the flooded area. Not every flooded area across the United States was visited due to the lack of funding for such an effort, low likelihood of flood proofed stmctures being present in the flooded area, and the lack of need to inspect and collect data on every flood proofed structure tested by flooding. Data collection efforts initially included contacting local officials in selected communities for information on flood proofed stmctures. This procedure was eventually mostly abandoned due to the inability to gain needed information. TIle procedure evolved to locating the flooded areas, having an experienced engineer drive through the flooded areas searching for flood proofed stmctures, and visiting with residents of the flooded areas. When a flood proofed stmcture tested by floodwater was located, the engineer made a personal inspection of the site to detennine what flood proofing measures worked and what did not. DATA ANALYSIS This portion of the project was accomplished by an experienced engineer, primarily through analysis of the stmcture during the onsite inspection but also during the subsequent in-office reviews of the data collected. During the onsite field inspection, the engineer was looking for reasons why the particular measure failed if indeed it did fail or why the particular

PAGE 352

Buss 337 measure was successful. In many flood proofing applications where failure occurs usually only one or two mistakes were made that caused the flood proofing measure to fail. LESSONS LEARNED This is the most important part of the project. The intent of this project is to clearly point out to all interested parties what caused a flood proofing project to either fail or succeed. This is done by simple statements based on analytical observation rather than rigorous analytical computation. With this in mind, the following general conclusions have been made based on information coJIected to date. (1) Interior drainage systems must be included in any dry flood proofing, levee, or wall measme implemented. Soil pemleability, flood dmation, and rainfall dming the flood must be considered. (2) Flood shields must be readily accessible, must be strong enough and have adequate and flmctional seals, and must be periodically installed to ensure that installation can be done. (3) Flood proofing measmes, other than elevation, t11at have the design level exceeded allow flood damage to occur equal to t11at possible wit1lOut the flood proofing measme. lllerefore a factor of "safety" or "freeboard" needs to be considered. (4) A flood proofing measure is only as strong as its weakest point. Something as simple as improper location of the sump pump discharge line, lack of or a blocked sewer backup check valve, failure to seal arOlmd t1le electrical entrance conduit, lack of knowledge of an abandoned water line entering the structure, or failure of t1le structure's occupant to tum t1le interior drain sump on to automatic before leaving have resulted in failure of otherwise sOlmd and expensive flood proofing systems. (5) The rule of 3 feet of floodwater against a "nomlally" constructed wall as being the general upward limit on depth without failure or damage due to hydrostatic force still holds true. (6) When dry flood proofing a basement, both the ability of the walls and the floor to resist hydrostatic force must be considered.

PAGE 353

338 Performance of Flood Proofed Structures (7) Scour depth is often overlooked. Many otherwise sound flood proofmg systems have failed due to the foundation depth of the footings being less than the scour depth, causing support failure. (8) In areas subject to large amounts of scour, slabs on grade should have a perimeter footing deeper than the expected scour depth to prevent failure. (9) Enclosed areas subject to flooding in high-velocity areas should be avoided to prevent creating higher localized velocities as the floodwater flows around the enclosed area, creating conditions for even more velocity-related damage. (10) In hurricane areas, metal noncorrodible fasteners are essential to to bond the structures together to withstand the force of water and wind. (11) Levee construction should include no steeper side slopes than 1 horizontal on 3 vertical to reduce the potential for levee breaching when floodwater overtops the levee. (12) Flood wall height extension cannot be reliably accomplished without knowing the design paranleters of the flood wall footing. FUTURE WORK This project is not complete. While a considerable amount of good information has been received, more information on successes and failures of flood proofed structures is needed. Information on dry and wet flood proofing is especially needed since these types of measures are very difficult to locate when driving through a flooded area. The NFPC is requesting that any information on flood proofed structures, such as those described in this paper, be forwarded to the author for documentation. The address is U.S. Army Corps of Engineers, ATTN: CEMRO-PD-F, 215 N. 17th St., Omaha, NE 68102-4978. CONCLUSION The NFPC intends to continue this project until enough information is obtained to provide an adequate range of successes and failures of all flood proofing measures actually tested by floodwater. This is a national effort and information is requested from all entities.

PAGE 354

Buss REFERENCES URS Company, Inc. 1986 Post-Flood Disaster Assessment Report-Walnut Creek Flood, Clive, Iowa. Montvale, NJ. URS Corporation 1987 Flood Damage Assessment Report: Suburbs of Minneapolis. Paramus, NJ. URS Corporation 339 1991 Flood Damage Assessment Report: Surfside Beach to Folly Island, South Carolina, Hurricane Hugo, September 21-22, 1989, Volume 1. Paramus, NJ. URS Corporation 1987 Flood Damage Assessment Report: Central Michigan, September 10-13, 1986. Paramus, NJ.

PAGE 355

GPS Elevation Surveys-A Key to Proactive Floodplain Management David F. Maune Dewberry & Davis INTRODUCTION In 1994, after severe flooding in Georgia, Florida, Alabama, and Texas, Dewberry & Davis (D&D) surveyed nearly 8,000 flooded buildings to collect flood inventory data for the Federal Emergency Management Agency's (FEMA's) Individual Assistance Program. Concurrently, Certi fied Flood Adjusters made "windshield survey" damage estimates for those buildings; their estimates were subsequently found to be in error by 50-100%. Damage estimates can be vital for timely and correct rebuilding and buy-out decisions. These decisions often depend on whether estimated repair costs exceed 50% of the replacement value of the flooded building. In 1994, FEMA detemlined that existing computer models could more accurately estimate flood damage with three pieces of data about each buil ding: (1) the square footage of the building's footprint; (2) the building'S estimated replacement value; and (3) the depth of interior flooding (to the nearest foot). The data for (1) and (2) could be collected in advance for all floodprone buildings in a conmlunity, but FEMA needed a way to quickly obtain information on (3) for each flooded building. D&D also sought other ways to help floodplain managers to be truly proactive. Central to this was the means to better perfonn flood hazard identification and risk assessment, vital for flood mitigation initiatives. GPS "SHOOTOUT" In 1995, in cooperation with the Louisville and Jefferson County, Kentucky, Metropolitan Sewer District (MSD), FEMA sponsored a "GPS shootout" in which two global positioning system (GPS) technologies competed in vertical accuracy and cost/productivity. We call these technologies "GPS BackPack" operated by Larry N. Scartz, LTD., and "GPS TruckMAP" operated by John E. Chance & Associates. Both used Trimble 4000SSe receivers with real-time kinematic (RTK) and on-the-fly

PAGE 356

Maune 341 (OTF) reinitialization. They used alternative techniques for surveying the 3-D coordinates (latitude, longitude, and elevation) of survey target points on buildings without intruding on private property. D&D calls this "stand off surveying." For the productivity portion of the test, nearly 1,300 high-density houses were surveyed to detennine if elevation certificates could be mass produced for $30 per house, as opposed to the typical $250 per house. For the accuracy portion, 62 of the houses were selected to be independently surveyed by both methods because they presented one or more technical difficulties: (1) they were located along tree-lined streets where canopy cover would interfere with GPS signals and where D&D could test the OTF capabilities when satellite lock was lost; (2) they were on the opposite side of hills from the GPS base station, where RTK radio corrections would have difficulty reaching the GPS rover units; and/or (3) they were up to 200 feet off the road so that elevations would be "cantilevered" by significant distances. These three technical challenges were considered essential to test the true capabilities and limitations of stand-off GPS survey techniques. BackPack and TruckMAP would independently survey these 62 houses, and correct for local variations in gravity. D&D would then compare the two elevation data sets and detennine if FEMA's 6-inch vertical accuracy requirement was satisfied. If the GPS technologies perfonned well under these difficult conditions, they could be relied upon also to perfonn welltmder simpler conditions. When the two elevation data sets were laid side-by-side for the 62 houses, the results were amazing! The elevations all agreed within about one inch. The standard deviation was two-thirds of an inch, and the maximwn error was less than inches at the 95% confidence level. In high-density housing areas, both methods proved that highly accurate elevation certificates could be mass produced for less than $30 per house. Both BackPack and TruckMAP won the shootout. FEMA later sponsored GPS elevation surveys of thousands of homes in 61 counties in 8 states. With the best geodetic-grade GPS receivers and exacting procedures, D&D fOlmd that survey control points and benchmarks are typically in error by 6-12 inches, and sometimes by several feet. D&D found some new homes had been constructed at elevations that make them vulnerable to predicted floods and that about one-third of conventional elevation certificates, which establish the cost of flood insurance for post FIRM homes, were in error by over one foot when checked by more accurate survey methods. CHALLENGES PREVIOUSLY UNSOLVABLE See Table 1 for a summary of 10 common challenges that can be solved with GPS elevation surveys.

PAGE 357

342 GPS Elevation Surveys Table 1. Challenge solutions from GPS elevation surveys. CHALLENGES currently facing SOLUTIONS: With Pre-Flood GPS Elevation Surveys to Floodplain Managers Nationwide predict depth of interior flooding of floodprooe buildings: I. Benchmarks nationwide have 6-12" Use best NGS control in each counl)' for all NFIP products. errors; some are several feet in error. Strictly follow NGS "Guidelines for GPS Eevalion Surveys [5 2. Large percemage of conventional centimeter accuracy I" to survey all !loodprone buildings. Elevation Cerrificates have elevation errOrs greater than I fool. Correct GPS surveys rigorously for local gravil)' variations. 3. Cannot quantify hazards/risks from Apply FEMA/USACE computer models to reliably estimate 500-, 100-, 50-, and lO-year floods. flood damages. These models require (1) predicted flood depths, (2) square footage, and (3) replacement values. NOTE: Hazard identifications and risk assessments are key to all lIU1igation Quantify legitimate !lood risks -for individual buildings and effons. for the entire communiI)' -as basis for mitigation initiatives. 4. Difficult to justify drainage computer models detenrune expected damages from (OO-yr and improvement projects. other floods -without drainage improvements (higber BFEs) and with drainage improvemems (lower BFEs). Detennine benefits of project in terms of damages avoided. 5. Convemional Elevation Cerrificates: Produce GPS Elevalion Cerrificates: (See example on reverse) costly (typically $250), less accurate. Higbly accurate and affordable when mass produced: Elevation Accuracy: inches < $30 per building in high densil)' urban areas < $70 per building in low densil)' rural areas BFE imerpolated to O.I foot (1.2 inches) Cerrificate recommends best-buy flood insurance. 6. Pre-FIRM buildings curremly don't Community eliminates excuses for not buying flood insurance require Elevation Cerrificates to idemify by providing certificates free to Pre-FIRM and Post-FIRM actual flood risks. Subsidy is expensive; homeowners and encouraging purchase of flood insurance. Congress directed 1996 subsidy restudy. Apply for CRS credits to reduce rates and offset costs. 7. Difficult to predict candidate Use GPS elevation data to run computer models for 500-, 100-, buildings for retrofitifloodproofing. 50and lO-year floods. Idemify candidates for relocation, elevation in place, floodwalls, levees, dry/wet floodproofing. Perform benefit-<:ost analyses; take proactive steps. 8. Post-flood "windshield" damage Survey post-flood elevations of several high water marks; then, estimates have errors of 50 % to 100%. calibrate H&H models to flood evenl. 9. Over 6 momh delay for disaster Estimate damages to individual buildings and communities. inventories and "rebuildlbuy-out" decisions for substantially damaged Accelerate rebuild-buy-out decisions; expedite receipt of IFG buildings. and HMGP monies. Elevation Errors Challenges I and 2 pertain to elevation errors. Errors in survey control points, benchmarks, elevation reference marks (ERMs), etc. can undermine the accuracy and intended utility of National Flood Insurance Program (NFIP) products. Flood Insurance Studies (PISs), Flood Insurance

PAGE 358

Maune 343 Rate Maps (FIRMs) and conventional elevation certificates can all have undetected errors if they result from poor survey control. Rigorous GPS elevation surveys can resolve control point/benchmark discrepancies and identify the best control in each county for NFIP use. National Geodetic Survey (NOS) control points, regularly updated on NGS' electronic bulletin board (301-713-4181) are the most reliable. Inability to Quantify Flood Hazards and Risks Challenge 3 indicates the dilemma in being unable to accurately quantify hazards and risks from SOO-year, 100-year, 50-year, and 10-year floods. By surveying the elevation of the reference level of each building in or near a Special Flood Hazard Area (SFHA) before a flood, the comnnmity can estimate, on a house-by-house basis, the depth of interior flooding that would be caused by the standard flood events. The computer models, cited above, can then compute the estimated danlages to each building, and to the community as a whole, as a result of the standard flood events (500-year, 100-year, 50-year, and 10-year floods). Such hazard identifications and risk assessments are the key to all mitigation efforts, and the community can then be aggressive and proactive in taking mitigation initiatives to reduce future flood losses, and in promoting flood insurance to owners of at-risk homes. Difficulty in Justifying Drainage Improvement Projects Challenge 4 indicates that it is difficult to justify drainage improvement projects without detailed elevation data on individual buildings in the drainage area. For example, how does one prove whether or not it is worth $2 million to construct a drainage improvement project that will lower the base flood elevation (BFE) by two feet for an area that includes 400 floodprone homes? By knowing the elevation of the lowest floor of each home, its "footprint" square footage, and its replacement value, computer models can accurately estimate expected damages from standard flood events prior to drainage improvements, and t11en recompute t1le expected danlages with drainage improvements t1lat lower the BFEs. The drainage improvement project benefits can be detennined in terms of damages avoided. Limitations in Conventional Elevation Certificates Challenges 5 and 6 pertain to conventional elevation certificates, which are sometimes considered to be an impediment to the sale of flood insurance. All elevation certificates (conventional or GPS) are expensive when not mass-produced. Although elevation certificates are not required for pre-FIRM buildings (constructed prior to publication of FIRMs for the area), Congress has directed a 1996 study of the current subsidy for pre-

PAGE 359

344 GPS Elevation Surveys FIRM homes. Without elevation certificates, it is difficult to identify candidate buildings for retrofit/floodproofmg. Challenges 5 and 6 can be solved by producing highly accurate GPS elevation certificates, mass-produced and quality-controlled, for all buildings in or near floodplains, providing them free to pre-FIRM and post-FIRM homeowners, and encouraging them to purchase flood insurance. A sample GPS elevation certificate is shown in Figure 1. In addition to the individualized photograph of the building in question, the background map pinpoints the building's geographic location centered on the base map road network and also its position in or near the SFHA shown in blue. The BFE is interpolated to the nearest 0.1 foot, and the elevation of the "target point" surveyed on the house is also shown to the nearest 0.1 foot. Target points are most typicall y the bottom of front door (BFO) or the top of foundation (TOF). Offsets to below-ground floors are estimated, based on standard 8-foot basement foundations, or 9-foot standard offsets between floors. Corrections can be made by the insurance agent and owner if the offset distance error is significant for insurance rating purposes. The estimated depth of interior flooding from the 100year base flood is also provided on the GPS elevation certificate. GPS ELEVA nON CERTIFICATE .-.... ....... --.,.,.-,......1 T .. "' ........ ... ....... ....... /5oIM. JG PI
PAGE 360

Maune 345 GPS elevation certificates, free to all, would clearly be important in the event Congress decides to eliminate the subsidy for pre-FIRM homes. In fact, they would probably be the key to success or failure in getting pre-FIRM homeowners to purchase flood insurance at actuarial rates. Challenge 7 can be solved by using GPS elevation data to run the computer models for standard flood events to identify candidates for relocation, elevation in place, floodwalls, levees, dry or wet floodproofing. Benefit-cost analyses indicate the viability of retrofitting/floodproofing of selected buildings. LImited Response to Actual Flood Events Challenges 8 and 9 pertain to current problems in estimating actual flood damages and in expediting federal monies to assist flooded homeowners and affected communities. The solution is quite simple. By already knowing the elevation of the lowest floor of each floodprone building, its square footage, and replacement value, the community would merely need to survey the post-flood elevation of several high water marks (e.g., 14th, 12th, and 9th Street bridges) in order to calibrate the H&H models to the actual flood event. Then, floodplain managers can quickly and accurately estimate the depth of actual interior flooding, estimate the damages to individual buildings and conmlunities, accelerate rebuild/buy-out decisions, and expedite the receipt of Individual and Fanlily Grant (IFG) and Hazard Mitigation Grant Progranl (MHGP) monies. SUMMARY For all buildings in or near SFHAs, accurate elevation data collected months or years in advance of actual flooding, appears to be a key to proactive floodplain management and should be helpful in implementation of FEMA's National Mitigation Strategy. Without elevation data, floodplain managers are generally restricted to reactive measures. With accurate elevation data, floodplain managers can perfoml reliable hazard identification and risk assessments; they can take proactive measures to actually reduce flood risks; they can produce GPS elevation certificates that help homeowners recognize their true flood risk and buy best-value flood insurance to reduce their financial vulnerability; and they can help accelerate federal disaster assistance funding when flooding actually occurs. The benefits to a floodprone community appear to greatly outweigh the low, mass-produced cost to the community in obtaining the highly accurate GPS elevation surveys that make proactive floodplain management possible in the first place. For more information, contact Dr. David Maune, D&D's Director of Mapping and GPS/GIS Services, at (703) 849-0396.

PAGE 361

Streamlined Data Collection for Substantially Damaged Structures in Ohio Eric Berman Federal Emergency Management Agency, Region V Donald W. Glondys Woodward-Clyde Federal Services INTRODUCTION Under contract to the Federal Emergency Management Agency (FEMA) Mitigation Directorate and Region V, Woodward-Clyde Federal Services (WCFS) inventoried buildings with the potential for substantial damage in three Ohio conununities for this "proof-of-concept" project. The data collection methodology employed was intended to utilize user-friendly but sophisticated computer hardware and software to expedite data collection, and do so at lower cost than was previously realized in other FEMA sponsored efforts to collect similar post-flood data. Communities that participate in the National Flood Insurance Program (NFIP) have the responsibility to require and review building permits before reconstruction of flood-damaged buildings occurs. Unfortunately, after a major flood community building officials often do not have the resources to identify potentially substantially damaged buildings or to inform owners of the NFIP regulations. Additionally, building officials and owners have difficulty in understanding the NFIP substantial damage regulations, which are not always enforced by communities. Under the NFIP, if a building is more than 50% damaged, a residence is expected to be elevated if repaired at all, and commercial buildings can either be elevated or floodproofed. PROJECT The buildings inventoried were danlaged as a result of Ohio River flooding that occurred January 20 through 22, 1996. Data was collected during site visits to the three Ohio communities of Brilliant (Jefferson County), Powhatan Point (Belmont County), and Racine (Meigs County)

PAGE 362

Berman and G/ondys 347 between February 12 and 14, 1996. The locations of the buildings were derived from Region V's Preliminary Damage Assessment (PDA) reports, which contained addresses (where available) or locations of damaged buildings on a community street map. The PDAs identified 42 residential units in Brilliant, 79 residential units in Powhatan Point, and 35 residential units in Racine, for a total of 156 units that had been potentially substantially damaged. The computer hardware used for the project included a TelePad notebook computer and electronic pen, and a Logitech FotoMan Pixtura Color Digital Camera. The notebook computer is designed for field use and utilizes a pen-based system within a Windows environment. Data can be entered either by the electronic pen or the detachable notebook keyboard. The main software used for the project was GeoFirma FieldPack Mobile Professional DGPS 2.2. This software works with MapInfo (a geographical infomlation system or GIS software) to link data to real world coordinates. Additional software consisted of GeoFirma FieldPack Designer DGPS 2.2 to create the electronic data inventory forms and MapInfo to prepare the digital maps. A data collection team was composed of three members and included an engineer from WCFS, a certified flood insurance adjuster, and a FEMA Disaster Assistance Employee (DAE). The DAE provided guidance for the locations of the buildings to be inventoried and the preparation of the individual PDAs. With the exception of the certified flood insurance adjuster, it is expected that future data collection efforts will not require outside contractors. The team was tasked with the following: (1) Collect a standard set of data in conformance with the FEMA Riverine Benefit-Cost Analysis module; (2) Record a digital image (i.e. digital photograph) of each building inventoried; (3) Detemline the pre-damage value, cost of the repairs due to flood damage, and the actual cash value for each building; (4) Determine whether buildings are potentially substantially damaged; (5) Incorporate the data collected onto a digital map for each cOlIlImmity; and (6) Evaluate the project concept, methodology, and data collection process to provide FEMA with an assessment of future use of this concept. Prior to the data collection, a digital map was prepared for each commlmity. When the digital map for a commlmity is opened on the computer, a street map of the comnllmity and a program-related toolbar appear on the computer's screen. From the tool bar, a blank data entry form is opened and data entered either by the electronic pen or the computer keyboard. A digital image of the inventoried building is recorded by the digital camera for later downloading and linking to the

PAGE 363

348 Data Collection for Substantially Damaged Structures record. Only the necessary munber of digital data layers (base map and street information) were used to reduce internal storage requirements. Data was obtained from a "windshield survey" process, which was used to minimize the amount of field time spent at each building, facilitate data collection by keeping the computer equipment in the front seat of the vehicle, and eliminate the need for access permission for entry onto private property. The data collected for residential buildings was recorded in the previously prepared data inventory forms. When available, high water marks on the flood danlaged buildings from the January 1996 flood were annotated on the digital images after linking the image to the building'S record. This unique feature is available through the pen-based software used for the project and offers a data collection feature not widely available on other notebook computers. Downloading of the images from the digital camera to the notebook computer and linking the images to a specific record was very time consuming. Color digital images require 5 to 7 minutes each to download and link to a record. Conversely, grey monochrome (i.e., black and white) digital images require only 2 to 2.5 minutes for the same procedure. On day one of the site visits, full data collection plus digital image downloading averaged 13 to 15 minutes per inventoried building (for color images). As the inventory team's proficiency increased, this improved to 6.5 minutes per building by the third day (for grey monochrome images). There is no appreciable difference in image resolution between the color and grey monochrome digital images. The primary difficulty in correlating the PDA data to field conditions was attributed to the length of time between the actual flood on January 20-22 and the site visits on February 12-14, 1996. During the intervening time, additional precipitation (both snow and rain) and clean-up activities by the conununities and residents reduced the physical evidence of flooding. Additional correlation problems can be attributed to the lack of addresses on buildings or mailboxes. Future inventory efforts could be enhanced by activating the data collection team within 10 calendar days of the flood. Deployment after this 10-day period may affect the accuracy of the data collected, particularly information associated with the extent of flood damage, and therefore the cost of repairs. The project was successful in developing a portfolio of 121 residential units in the three communities in two and a half days of field work and two days of post-field processing. Of these buildings, 46 units were determined by the data collection team to be either substantially damaged or potentially substantially damaged. The data obtained during this project will be retained by FEMA due to the "proof-of-concept" status. However, it is anticipated that entire or partial results of future projects could be provided to community officials

PAGE 364

Berman and G/ondys 349 to assist their rebuilding efforts. The information would be useful to communities by encouraging full compliance with the NFIP regulations, while screening potential buildings for flood mitigation activities such as floodproofing or acquisition. EVALUATION The computer hardware and software perforn1ed well in the field. After an initial acclimation period, the notebook computer and digital can1era were considered user-friendly, requiring only a small number of operation commands for data collection. With a minimum of training, non-technical field personnel should be able to use the equipment without much difficulty. The computer software was also considered user-friendly. Preparation of the digital maps prior to field deployment facilitated overall GIS use by reducing the number of steps required to activate and use the GIS software for its intended purpose on this project. Under good conditions, it is estimated that a two-member team could collect data (exclusive of digital image downloading) for up to 100 buildings in one day. The conditions would include weather, amount of daylight present, proximity of inventoried buildings to each other (i.e., geographic area to be covered), extent of flood damage, amount of physical evidence of flooding, and availability of PDA data before field deployment of the data collection team. Data Collection Methodology The process of obtaining data via a "windshield survey" was determined to be effective for gathering the data required for this project. Future projects could require exiting the vehicle for more detailed inventories for each building, such as an examination of the building'S sides not visible from the street or the interior. Additional data requirements will have varying impacts on the rate of data collection. The PDA data provided important guidance on the location of flood damaged buildings and the extent of flooded areas and therefore increased the rate of data collection. However, because PDAs will most likely be prepared by personnel other than the data collection team, the PDA data should not be the sole source of data for the inventory. Since the technical backgrounds of the PDA and inventory teams may not be similar (particularly where appraisals are concerned), the potential for discrepancies between the quantity and location of damaged buildings will remain a possibility.

PAGE 365

350 Data Collection for Substantially Damaged Structures Data Collected The standard data inventory fonn on the computer provides a checklist for the data observed for each building and the data recorded. This project used the input data necessary to run the FEMA Riverine Benefit-Cost Analysis module as guidance for the field data obtained. The data inventory fonn used on this project can be easily modified, expanded, or reduced to accommodate any future revisions to the data requirements. Comparison to Previous Data Collection Methods Previous data collection projects of this nature involved an inventory of damaged buildings that included the elevation data necessary to complete a FEMA elevation certificate. These projects contained a number of variables which impact project cost, such as the number of contract personnel and the skills categories of the team members (i.e., appraiser, surveyor, etc.), complexity of the project, inventory data requirements, travel and per diem costs, quantity of buildings to be inventoried, need for elevation certificates, and determinations concerning substantially damaged buildings. This "proof-of-concept" project has variables similar to the previous projects, but was intent on field testing a screening process while streamlining the data collection procedures and reducing the unit cost of data collection per building. This was accomplished by reducing the size of the data collection team, reducing the level of detail for the data collected and eliminating the elevation certificates. These steps reduced the complexity of the project and therefore, the unit cost. The screening process detemlined that not every building was substantially damaged. Additional cost savings are realized because the number of buildings requiring the preparation of elevation certificates has been reduced. The cost savings are conservatively estimated to be at least 30% when compared to previous inventory efforts. If elevation certificates are required, costs can still be reduced by preparing them for only those buildings that have been determined to be substantially damaged instead of all buildings within the flooded areas of a community, as was done previously. Because of faster turnaround time between data collection and office evaluation, buildings that are not substantially damaged can be identified sooner and their owners allowed to rebuild without unnecessary delays. The equipment and data collection methodology have potential applications beyond the development of this portfolio of substantially damaged buildings. The user-friendly equipment lends itself to any type of field data collection efforts such as FEMA Community Assistance Visits

PAGE 366

Berman and G/ondys (CA Vs), evaluation of post-flood reconstruction or new construction activities, and compliance of elevated buildings, among others. RECOMMENDATIONS 351 Based on the experience of the inventory data collection team during this "proof-of-concept" project, the following recommendations are provided for FEMA's consideration: (1) Use two-or three-member teams. (2) Include a certified flood insurance adjuster on the team to determine building values and the cost of flood damage repairs. (3) Deploy the data collection team within 10 calendar days of the flood event for the highest efficiency. (4) Pre-screen damage sites before field deployment to insure that potential substantial dan1age exists. (5) Develop a standard data inventory form before field deployment of the data collection team. (6) Prepare the digital maps in advance of field deployment. (7) Obtain only grey monochrome (i.e., black and white) digital images and limit the number obtained to include only those buildings which meet a pre-determined set of criteria. (8) Provide requirements for address verification to the data collection team before field deployment. REFERENCES Woodward-Clyde Federal Services 1996 Ohio River-Portfolio of Damaged Buildings, Summary Evaluation Report (Final). Federal Emergency Management Agency, Contract EMW-95-C-4678, Task Order No. 46.

PAGE 367

This page IS intentionally blank

PAGE 368

Part 12 PLANNING FOR, USING, AND MAINTAINING STRUCTURES FOR LOSS REDUCTION

PAGE 369

This page IS intentionally blank

PAGE 370

Urban Planning for an Area Protected by Levees: The Natomas Basin in Sacramento County, California James C. Campbell San Francisco State University and Pinnacle Data Corporation INTRODUCTION Floodplain regulation has several purposes, among them reducing the potential for injury, reducing the potential for damage, preventing the unwary from buying floodprone real estate, preventing new development in floodprone areas, reducing public costs for emergency operations, reducing public costs for post-flood repairs, reducing the need for structural flood-control measures, and preserving natural floodplain values (Flood Loss Reduction Associates, 1981). Essential to the establishment of public policy for floodplains are useful flood hazard maps. As Dingman and Platt pointed out in a 1977 article, precise flood bOlmdary delineation is hydrologically impossible (Dingman and Platt, 1977). Most highly floodprone areas, however, are relatively easy for hydrologists to define-and for users to discem-on a flood map. Public planning agencies can therefore decide for themselves how broadly to define flood danger in their respective communities. Many use Flood Insurance Rate Maps (FIRMs) produced by the Federal Emergency Management Agency (FEMA) for their basic guidelines, allowing or disallowing development in 100-year floodplains depending on various local criteria. But are there places in the United States where the potential injury and damage from flooding are extreme-yet completely ignored by the FIRMs? THE BASE FLOOD The "100-year flood," or base flood, refers to a flood elevation that is likely to be equaled or exceeded at a particular site on the average of once every 100 years. However, this average is really only meaningful over a period of centuries: several 100-year floods could occur over a short period of years. Nevertheless, the floodplain defined by the 100-year flood is used by the National Flood Insurance Program (NFIP) to identify

PAGE 371

356 Urban Planning for an Area Protected by Levees areas where the risk of flooding is considered "significant." Of course, the lOO-year standard is an arbitrary distinction for two reasons: 1) the frequency of flooding during any given period of time may be greater, and 2) the level of flooding at any time may be greater. Most importantly, the base flood is a frequency threshold that takes no local variables-such as potential water depth and severity of flood damage-into account. Nevertheless, planning agencies often base their public-safety policies on this frequency criterion; those that use FIRMs as their sole regulatory flood maps are doing so by default. FIRMs, however, currently designate most areas that are structurally protected from the base flood not as Special Flood Hazard Areas but as "other flood areas" (not requiring insurance). The categories included within this designation (shaded zone X) are areas of 500-year flood, areas of 100-year flood with average depths of less than one foot, areas of 100year flood with drainage areas of less than one square mile, and areas protected by levees from the 100-year flood. ll1ese areas are a\1 considered equivalent for insurance purposes. But should they be considered equivalent for public policy purposes? LEVEES Many flood-control structures-particularly levees-are built specifically to contain the 100-year flood. But areas behind levees may be at a risk of greater flood dan1age than they would be if no levees existed at all. By definition, any flood greater than the 100-year standard is a larger flood than a 100-year structure is designed to control. "Floods exceeding the level for which levees ... are designed can cause disastrous losses of life and property" (Flood Loss Reduction Associates, 1981). As a matter of fact, most levee failures have occurred without the water elevation reaching the levee crown (EIP Associates, 1989). A levee break unleashes floodwater with very high velocity and usually floods an area to great depth because the flood-stage water level behind a levee is much higher than the level of the land that it protects. Because the depth, suddenness, and duration of a flood are primary factors in the severity of the damage caused, a levee break can cause flooding that is not only significant, but catastrophic. Although constructing a levee system to 100-year standards will reduce the frequency of flooding, the flooding (when it occurs) can be just as severe as if no levees existed-in fact, more so (Association of State Floodplain Managers, 1985). However, the NFIP does not consider severity-only frequency-in requiring insurance of homeowners. Therefore, an area protected by levees to the 100-year standard is treated as any other non floodprone area for insurance purposes-even if it is subject to

PAGE 372

Campbell 357 catastrophic flooding due to an unforeseen levee failure. This flawed philosophy-which allows but does not require residents to purchase flood insurance-combined with the public perception that property is "protected" from floods, can only induce encroachment on levee-protected floodplains. THE SACRAMENTO VALLEY In much of the Sacranlento Valley in northern California residential, agricultural, and commercial areas are protected against flooding-to some extent-by levees. Many of these structures were first built in the midto late nineteenth century; they have been extended, upgraded, or replaced during this century by federal agencies-usually the Anny Corps of Engineers (Kelley, 1989). TIle reason for these structures' existence is the extreme flood danger posed by the Sacramento River and its tributaries. During the last two centuries the Sacramento Valley experienced severe flooding in 1805, 1825, 1826, 1839, 1840, 1847, 1849, 1850, 1852, 1861, 1862,1878,1881,1890,1937,1938,1940,1943,1945, 1950,1952,1955, 1956, 1958, 1962, 1963, 1964, 1967, 1969, 1973, and 1986. In the days before major, coordinated flood control structures, the resulting "inland sea" could be 250-300 miles long and 20-60 miles wide (FEMA, 1978). At nornlal flow the Sacramento is a big river, carrying about 5,000 cubic feet of water per second, but at flood stage it has been measured at over 600,000 cubic feet per second. TIlroughout history its natural banks have never contained the wet-season flows of the river except during unusually dry years. In fact, all the streams of the Sacranlento Valley floor flow on elevated beds fonned by their own silt deposits and paralleled by natural levees created during successive flood seasons (Kelley, 1989). The usual cause of flooding in the Sacranlento Valley is a high rate of nmoff from heavy and prolonged autlUlln, winter, or spring rain, often augmented by snowmelt feeding the Sierra Nevada tributaries of the Sacramento River (FEMA, 1978). TIlis type of flooding may overtop natural as well as human-made levees. Long-tenll high water, a sudden flow increase, or a major seismic event may weaken levees to the point of breakage, at which time catastrophic flooding is likely to occur. LEVEE BREAKS The most recent major flooding in the Sacranlento Valley occurred in the winter of 1986. The weather preceding the flood consisted of a series of intense stonns that had saturated the grOlmd. In February a levee broke near the confluence of the Yuba and Feather rivers (tributaries of the Sacramento River) in Yuba County. The area immdated by this failure included several large residential subdivisions on floodplain "protected" by

PAGE 373

358 Urban Planning for an Area Protected by Levees the levee in question; about 6500 buildings were affected. Because it was caused by a structural failure, flooding was sudden and rapid, leading to considerable damage. And, because of the flat, low-lying nature of the floodplain, the depth of the flooding and its persistence were also extreme. Depths ranged up to 12 feet and some residential areas were still flooded several weeks later. The combination of high velocity, great depth, and long duration exacerbated property danlage. But because of the levee system's designation as 100-year-flood protection, neither flood insurance nor floodplain management had been required (Tobin and Montz, 1988). THE NATOMAS BASIN A disturbingly similar potential for disaster exists only some 30 to 40 miles farther south, in the Natomas Basin. About 15% (over 9,000 acres) of this low-lying area-currently consisting almost exclusively of agricultural land, primarily in rice-has been proposed for residential and other development. Most of the area currently proposed for development is incorporated within the city limits of the City of Sacramento, lying in the northwest portion of the city about three miles from the downtown section (Sacramento City Planning Commission, 1994). The Natomas Basin is at the confluence of the Sacramento and American rivers, which drain, respectively, the vast Sacramento Valley and the smaller valleys and canyons to the west of Lake Tahoe. The area's flood control structures consist of an extensive levee system as well as overflow weirs and pmnping plants. These are designed mainly to transfer excess floodwater to a system of bypass channels. This partial reduction of flood hazard has allowed considerable agricultural development since the Natomas levees were built in 1914 as well as some more recent urbanization on fomler swamp and overflow lands (FEMA, 1978). The urbanized area now consists of more than 13,000 homes, businesses, and public buildings in the soutlleastem comer of the basin. At the time of construction, the urbanized area's level of flood protection met minimum federal and local standards because it was deemed protected from the lOO year flood by the levee system. Also because of this, flood insurance was made available to residents but was not required of them (Estes, 1993). THE NORTH NATOMAS COMMUNITY PLAN A "North Natomas Commtmity Plan" (NNCP) has been promulgated by the City of Sacramento Department of Planning and Development and has been adopted by the City Council. The comnllmity plan consists of a detailed study for the physical development of the area. Further urbanization of the Natomas Basin had originally been proposed in 1985, however, the severe flooding in the Sacramento Valley in the winter of

PAGE 374

Campbell 359 1986 caused the federal agencies involved with flood control to revise their estimates of the ability of existing structures to prevent damage in the Natomas area. Before the 1986 floods, FEMA standards had indicated that North Natomas (the name given to the proposed development area) had protection from up to a 125-year storm. After that, the area was re analyzed by both the Army Corps of Engineers and by FEMA and was designated within the lOa-year floodplain (Sacramento City Planning Commission, 1994). As a result of this reduction in the officially designated amount of flood protection, the Corps of Engineers and the Sacramento Area Flood Control Agency embarked upon a series of studies and projects designed to improve the flood protection provided to the Natomas Basin, mainly through levee reconstruction and renovation (Sacramento City PI arming Commission, 1994). These included the Sacramento River Urban Levee Reconstruction Project and the Natomas Area Flood Control Improvement Project, which were supposed to provide 200-year flood protection. However, in March 1994, the Corps of Engineers revised its evaluation of the Sacramento River's east levee to state that the current levee-improvement plans were inadequate to provide even lOa-year protection to the basin (Sacramento City Planning Commission, 1994). As it stands now, by the end of 1996 the levee improvements should be in place and widespread construction will be allowed as plarmed. Limited commercial development has already begun. In the 1985 Environmental Impact Report (EIR) prepared in connection with the NNCP, the ternl "flood control" is used to define facilities reducing flood risk to the lOO-year standard. (Flood severity is not addressed.) Once this level of risk is achieved, the Sacranlento city and cOlmty governments are committed to opening up the floodplain to development, even when many areas are subject to deep, catastrophic flooding and remain, according to the group Friends of the River, within the floodplain recently mapped by the Army Corps of Engineers in the American River Watershed Investigation (Sacramento City Planning Commission, 1993). The EIR contains virtually no acknowledgment that the Natomas Basin is subject to extraordinarily dangerous flooding. Yet this is precisely the problem with the proposed cOrrllnmity: the levees that surround the basin on all sides are 15 to 20 feet higher than the inside land area; therefore, during flood events, river levels are considerably higher than the grOlmd level inside the basin. If a levee were to break, according to Friends of the River, floodwater could quickly fill most of the Natomas Basin to a depth of eight to 23 feet for one month or longer (Sacranlento City Planning Conunission, 1993). In regard to the Natomas Basin the local jurisdiction has decreed, "To develop in the [North Natomas] area, lOO-year flood protection must be achieved to avoid personal injury and property dan1age and to obtain

PAGE 375

360 Urban Planning for an Area Protected by Levees affordable insurance" (Sacramento City PI arming Commission, 1994). Is this an appropriate use of the Special Flood Hazard Area designation? And should public pi arming agencies be considering the prevention of personal injury to be on a par with the ability to obtain property insurance? REFERENCES Association of State Floodplain Managers 1985 Reducing Losses in High Risk Flood Hazard Areas: A Guidebook for Local Officials. Washington, D.C.: Federal Emergency Management Agency. Dingman, S. Lawrence, and Rutherford H. Platt 1977 "Floodplain Zoning: Implications of Hydrologic and Legal Uncertainty." Water Resources Research 13 (Jtme):519-23. EIP Associates 1989 Draft Environmental Impact Report: Land Use Planning Policy within the 100-Year Flood Plain in the City and County of Sacramento. Sacran1ento, CA: City of Sacran1ento PI arming and Development Department. Estes, G.W. 1993 "New Development in Deep Floodplains Is Bad Public Policy: The Natomas Basin E;vample." Unpublished. Federal Emergency Management Agency 1978 Flood Insurance Study: Sacramento County, Unincorporated Areas. Washington, D.C.: Federal Insurance Administration. Flood Loss Reduction Associates 1981 Floodplain Management Handbook. Washington, D.C.: U.S. Water Resources Council. Kelley, R.L. 1989 Battling the Inland Sea: American Political Culture, Public Policy, and the Sacramento Valley 1850-1986. Berkeley, CA: University of California Press.

PAGE 376

Campbell Sacramento City Planning Commission 1993 Draft Supplement to the 1986 North Natomas Community Plan EIR. Sacramento, CA: City of Sacramento Planning and Development Department. 1994 Staff report on the certification of the Supplemental EIR. Unpublished. Tobin, G.A., and B.E. Montz 1988 "Catastrophic Flooding and the Response of the Real Estate Market." Social Science 10urnaI25:167-77. 361

PAGE 377

Battelle's Levee Rehabilitation and Letter of Map Revision Daniel M. Hill Burgess & Niple, Limited INTRODUCTION Battelle is a worldwide research organization with 8,000 employees and annual revenues of $1 billion. They are headquartered in Columbus, Ohio, at their King A venue campus, which is contiguous to Ohio State University (OSU) on the Olentangy River. Battelle's campus is protected from flooding by a 1,200-foot-long levee, but had been designated in the regulatory floodplain because there were no official plans or operating procedures for the levee. Complying with floodplain building code requirements would have significantly increased construction costs for substantial improvement of existing buildings, or new buildings, needed in Battelle's continuing conunitment to provide first-class research facilities. Because the levee had withstood major flooding in the past, Battelle engaged Burgess & Niple (B&N) to seek a Letter of Map Revision (LOMR) from the Federal Emergency Management Agency (FEMA) regarding the floodplain delineation. FLOODING CONDITIONS The January 1959 flood is tile flood of record on the Olentangy River. Delaware Dam, a multipurpose (including flood control) project completed by the Corps of Engineers in 1951, controls 70% of the drainage area tributary to Battelle's location. The 1959 flood, indicated to be in excess of a 100-year flood by the Columbus Flood Insurance Study (FIS), did not overtop the levee. There was some flooding behind the levee due to a malfunctioning check valve on the interior drainage system. That valve was subsequently replaced and a formal maintenance procedure adopted. The Corps of Engineers, in their 1968 Flood Plain Information Report for Columbus, recognized the levee and did not show Battelle's campus to be in the floodplain.

PAGE 378

Hill 363 Ohio State University presently has a levee (also not recognized by FEMA) extending upstream from Battelle. That levee did not exist in 1959, but the King Avenue embankment runs 850 feet across the floodplain and prevented entry of flood waters into the Battelle campus. The King A venue embankment was considered to be a tie-back levee for LOMR purposes; it would be less costly to define its capabilities than those of the much longer OSU levee. A quick comparison of existing ground and 100-year flood elevations confirmed that the freeboard on the Olentangy River main levee was more than required, but King A venue's low point at Battelle's entrance lacked about one-half of the required 4-foot freeboard. A means to provide freeboard continuity, plus definition of the main levee and King Avenue embankment stability, were thus known to be key points in gaining FEMA's recognition of the protection works. Application for a Conditional Letter of Map Revision (CLOMR), followed by implementing necessary improvements to receive the final LOMR, was chosen as the appropriate course of action. OTHER AGENCIES The Ohio Department of Natural Resources (ODNR) has regulations and a permit system involving danl and levee safety. Their approval would be necessary for improvements to the levee system. The City of Columbus would be involved with the project because the river half of the levee is owned by the city and administered by its Department of Recreation and Parks as part of a linear parkway along the Olentangy River. The city would also be involved to satisfy FEMA's requirement that levee operation plans be under the jurisdiction of a community participating in the National Flood Insurance Program. Administration of floodplain regulations in Colunlbus resides in their Development Department, Regulations Division. Details of the levee operations plan, however, were to be coordinated with the Public Utilities Department, Sewerage and Drainage Division. INVESTIGATIONS AND DESIGN Battelle authorized B&N to begin the project on January 14, 1994. Investigations included freeboard continuity; closure devices; embankment erosion protection, stability, and settlement; and interior drainage. The CLOMR application, including a draft operation and maintenance (O&M) manual, was submitted to FEMA and ODNR in August 1994. Use of a 140-foot-long, 3-foot-high water stmctures unit across Battelle's King A venue entrance was proposed to satisfy the freeboard continuity. This product (twin water-filled geomembrane tubes contained within an outer

PAGE 379

364 Battelle's Levee Rehabilitation and LOMR geotextile tube to prevent rolling) offers the advantage of quick deployment. A 50-foot-Iong landscape mound with a maximum height of 1 foot completed the King Avenue freeboard continuity. Closure devices included totally filling an abandoned 8-inch storm drain under the main levee with grout (it already was plugged with concrete at one location) and providing temporary closures of heavy-duty plastic and sandbags for three catch basins and one manhole near the King A venue entrance. Application of the FIS HEC-2 computer model produced low flood flow velocities (maximum 2 feet per second), thus eliminating the need to add riprap erosion protection. Results of five test borings in the main levee and King A venue embankment provided acceptable stability safety factors, recognizing that the critical failure surfaces are shallow and both the levee and embankment are very wide. Settlement was not a concern because the levee had been in place over 30 years. Interior drainage calculations for concurrent river flood stage and localized stonn rainfall produced a maximunl ponding depth of about 6 inches. (See Table 1.) Table 1. Sununary of factors of safety against slope failure. Case Sudden drawdoWD (transient flow net to define undrained soils limit) Steady seepage from full flood stage (partially developed phreatic surface) Steady seepage from full flood stage with earthquake factor of Safety (FS) River King Avenue Minimum Required Embankment Factor of Safety 1 0.8 0.8 1.0 1.8 13 1.4 1.6 1.1 1.0 1 EM 1110-2-1913, COE Design and CODStruction of Levees Refinements to the O&M manual were made in September 1994 at the request of Columbus officials. These included further analyses to indi cate the number of cycles (at least 10) of repetitive flooding/drawdown needed to breach the levee crest. The O&M manual was nonetheless revised to include provisions for immediate temporary repair of any shallow slough failure. FEMA requested additional data concerning eight items on December 27, 1994, which was then prepared and submitted to FEMA and ODNR February 16, 1995. Items of significance included the following:

PAGE 380

Confirmation that the water table in all test borings was below the river bed elevation. A request for revised stability analyses to include full saturation of the levee. This prompted detailed review of flood hydrograph data, which showed the duration of flooding at the levee toe elevation had never exceeded 15 hours and the maximum duration of the lOO-year stage was only a few hours. The original partial saturation analyses was therefore still fOlmd to be appropriate. Definition of basement elevations and any related seepage problems. Data review showed river flood stage had exceeded the lowest floor elevation (there are no basements) eight times since 1959 with no seepage problems. General methodology and pressure relief parameters for the uplift analysis at the landside toe were questioned. The methodology was then independently reviewed with the Corps of Engineers and found to be appropriate. AsslUning the parking lot dry wells to function in reverse as relief wells (contrary to observed test boring data) showed the factor of safety would actually increase. Structural closure devices rather than sandbags for the manhole and three catch basins would be required. A watertight bolted lid was selected for the manhole and steel insert plates with attached drain outlet valves were proposed for the catch basins. FEMA issued the CLOMR on March 28, 1995, which then permitted resolution of ODNR comments on the proposed work. One particular issue was their standard requirement to remove all trees from the levee. The Columbus Parks Department opposed this because a treed corridor was important to their parkway plan. A compromise was reached where trees would only be removed if they became diseased or damaged, and some replacement trees could be planted on the riverbank (but not on the levee). The rehabilitation plans thus included initial removal of nine trees and planting three replacements. The other principal addition required by ODNR was an exploratory trench to locate any tmknown utilities passing under the levee plus installing a reverse filter on such pipes to control potential seepage. Further changes initiated by Battelle were incorporated: replacing the shallow landscape mound for freeboard continuity adjacent to their King A venue entrance with a solid masonry landscape wall to complement large existing planters, and replacing four pine trees adjacent to the new landscape wall. Construction plans and a design report were

PAGE 381

366 Battelle's Levee Rehabilitation and LOMR submitted to ODNR on May 15, 1995 and their approval was issued June 14, 1995. CONSTRUCTION Battelle administered the construction work, engaging a landscape contractor for tree and brush removal, new tree plantings, and landscape wall construction; plus a mechanical contractor for exploratory trench and reverse filter installation, catch basin and manhole modification, grouting the abandoned 8-inch storm drain, and procuring the water structure unit. Work was initiated in early July and was completed (other than planting new trees) by the end of August. A trial deployment of the water structures unit was made on August 26, with staff from Columbus and ODNR also invited. Total construction cost was nearly $94,000 (Table 2). Table 2. Summary of construction costs. Work Items Catch bli$ins modified with drain outlet valves (3 each) and manhole modified with watertight lid (1 each) Exploratory trench (240' long x 5' deep) plus grout fill and reverse filter on existing pipe (1 each) Brush removal (1,600 square yards) and trees removal (13 each) Trees planted (7 each) Landscape wall (95' long x 2' high, maximum) Water Structure unit (140' long x 3' high) LOMR Costs $20,250 23,500 15,000 2,300 28,300 $93,850 Reports on the construction process and "as-built" copies of the plans were submitted to ODNR September 5, 1995. The "as-builts" and request for a final LOMR were submitted to FEMA September 14. They made a request in early October for stability calculations on the landscape wall, which were promptly returned showing the safety factor to be 13.8. The

PAGE 382

Hill LOMR was issued by FEMA on December 11, 1995, some 16 months after initial submittal of the CLOMR application. O&M MANUAL 367 The O&M manual is a comprehensive document addressing both operation and maintenance activities in detail. Battelle is responsible for operation and maintenance of the entire facility, even that portion on City of Columbus property, with Columbus' approval and cooperation. The primary interior drainage pump is a permanently installed lO-inch gasoline-powered Jaeger pump rated at 3,500 gallons per minute. Three other portable pumps ranging from 6-inch to 3-inch size are also available for emergency backup. The operation plan includes testing the pumps once every 3 months, operating the gravity stonn drain check valve every 6 months, and deploying the water structure unit annually. The drain valves in the catch basins are to be cleaned out after every rain. The operation plan includes a flood warning system initially comprising the National Weather Service river stage forecasts for their upstream Delaware gage plus their local intense rainfall warnings. The warning system will ultimately involve the Cohunbus system being developed for the West Coltunbus Local Protection Project, which will utilize forecasting capabilities of the National Weather Service and the State of Ohio Rain/Snow Monitoring System (STORSMS). Two different levels of emergency condition response are provided in the operation plan, E Con 1 for monitoring/standby and E Con 2 for closure deployment and drainage pump activation. The plan includes a specific notification list, including outside personnel, for E Con 2. The maintenance plan includes provisions for mowing the embankment at least twice annually and the previously noted removal of trees as they become diseased or damaged. Observation and repair of erosion areas, seepage, or sloughs are all described. The importance of and details for grOlmdhog control are provided. Inspection twice annually by Battelle staff and every 5 years by a qualified professional engineer is included.

PAGE 383

Grade Stabilization Structures for Natural Rivers Joseph C. Hill Kenneth C. Hanson San Diego County Department of Public Works Jon Walters Nolte and Associates INTRODUCTION Many natural rivers need help. Natural rivers as defined in this paper have no major structural modifications to the river bed or banks, but may have bridges and utility crossings. Rivers with sand or gravel beds usually are subject to major changes during floods. In many cases the natural equilibrium has been upset by mining operations or other activities of civilization and it is necessary to reestablish equilibrium with flood control stabilization structures. The stabilization structures establish the upstream river bed and flood flow conditions that avoid erosion and sedimentation problems. The structures protect bridge footings, utility crossings, river banks beside houses, roads, and other infrastructure. They may also be needed to stabilize a river bed to avoid loss of vegetation. Examples of river beds needing protection include those containing golf courses, areas of riparian vegetation, landscaped parks, etc .. A key function of stabilization structures is energy dissipation. Rivers with relatively steep stream bed profiles can be effectively flattened with such structures. Each structure must effectively control the energy loss in its stilling basin to avoid adverse effects downstream. The structures are compatible with the HEC-2 process so that a structure can be easily included in a typical floodplain. Methods other than HEC-2 are required to analyze the characteristics of the hydraulic jump that dissipates energy.

PAGE 384

Hill, Hanson, and Walters 369 OVERVIEW, TYPICAL DESIGN PROCESS This section outlines the design and construction of flood control grade stabilization structures, or "drop" structures. This guidance is based on the experience of County Flood Control, Nolte and Associates, and the Corps of Engineers, who have built the stmctures in the southwest United States. The design engineer should be proficient in the design of these structures and in mapping floodplains in large rivers. The stabilization structures will stabilize the river bed and establish the upstream flood water surface, typically at the existing level. The important elements in stabilization structure design are flood flow recurrence interval relationships, river/strean1 characteristics, downstrean1 flow conditions, type and hydraulic geometry of structure (grouted rock, gabion, or reinforced concrete) and sub-structure conditions. Flood Flow Recurrence Intervals The lOO-year flow used for flood insurance purposes is usually the most important design flow. The 100-year flow for future watershed conditions should also be evaluated. Smaller (e.g., 10-year) and larger (e.g., standard project flood) flows should also be included in the design process. River/Stream Characteristics at the Site The setting for the proposed stabilization structure includes the topo graphic features, the existing and future infrastmcture, the river bed material, and the groundwater conditions. In California, the State Division of Safety of Dams should be contacted for review and conunent if it appears that their criteria apply. They detennined that the proposed Upper San Diego River structures would not impound water and t1lerefore are not considered dan1s. A preliminary report evaluating alternative designs for t11e structure relative to site conditions is essential. TIle river slope and t11e riverbed conditions upstream must be considered in establishing t1le water surface and streambed elevation upstream of the proposed structure. Crest Elevation and Width Options for various crest elevations, crest widths, and unit discharges should be considered. TIle Corps has evaluated lmit discharges in the range of 50 to 200 cfs/foot. With an established upstream water surface elevation and river bed elevation, the width of the structure will establish t11e unit discharge. A wide structure will have a relatively small lIDit discharge and a high crest elevation. A narrow structure will have a relatively large unit discharge and a low crest elevation. The crest is usually most effective when perpendicular to t11e river, but this is not always possible. Figure I shows a plan view and cross sections for a

PAGE 385

370 A L 110 -100 -110 -100 --Grade Stabilization Structures for Natural Rivers RIVERBED ELEVATION = 100 I FLOW I PLAN VIEW SECTION A-A ELEV IN FEET WIDE STRUCTURE SECTION B-B NARROW STRUCTURE SECTION B-B ELEV IN FEET --110 --100 --110 --100 Figure 1. A plan view and cross sections for a typical stabilization structure.

PAGE 386

Hill, Hanson, and Walters 371 typical stabilization structure installation. Figure 2 is a profile of a rock stabilizer. The structure should be designed with environmental considerations in mind. A Corps 404 permit is usually needed. Aesthetics should be considered to blend the structure into the existing surroundings. The structure should also be designed with consideration of construction techniques. Downstream Conditions The existing downstream river and streanl conditions are important to assure an adequate tail water elevation to provide a controlled hydraulic jump and prevent undermining the structure. The downstream variables for which future changes should be anticipated and their impact on the tail water elevation include growth of vegetation that will affect the river roughness, the stability of existing downstream structures, and the potential for erosion and sedimentation and sand mining in the river bed. Type of Structure Three basic types of stabilization stmctures are discussed here. The grouted stone structure features a spillway on a slope of about 2.5 to 1 with a hydraulic junlP type basin. Grouted stone stmctures have been constructed on the San Gabriel River and the San Diego River. The concrete structure has an impact type basin. Concrete structures with a straight drop stilling basin have been constructed on the Santa Ana River. Concrete structures may not be environmentally acceptable. Gabion structures are constructed by placing riprap in wire baskets. They are widely used to stabilize natural and hlUnan-made channels. Their relatively unobtrusive appearance is enhanced as vegetation grows naturally in the spaces between the rocks. These interstices also allow for the passage of groundwater and low flows. Gabion stmctures have been constructed in Cannel Valley in the City of San Diego. Constructio n Plans Detailed plans that clearly identify all aspects of the constmction are essential. Design details include providing protection at the ends adjacent to the crest, the toe of the stmcture, and the downstream channel banks. Detailed construction plans for the stmctures identified above are available. Sub-structure Analysis The sub-structure should be designed to insure the integrity of the structure by providing sufficient mass to resist the dynamic forces during flood flows. There should be a long enough flow path to prevent piping

PAGE 387

Riverbed DUMPED STONE GROUTED STONE FLOW ... Riverbed (low flow conditions) FILTER MATERIAL-DUMPED STONE -] PROFILE Figure 2. Rock stabilizer. c.> ....... I\) G> 8. CD 2: r::;' c,-:::J SQ c: c: CD en -o z c: @ :D <' CD Cil

PAGE 388

Hill, Hanson, and Walters 373 under and around the ends of the structure. An impervious core or cutoff walls may be necessary to accomplish this objective. CONCLUSION The pleasing visual effects of a grouted rock or gabion structure in a natural river setting are an important consideration for selecting this type of stabilization structure. There is a good basis available for designing these types of river stabilization structures. The experience gained from existing structures will be valuable for future projects. In Southern California, surplus rock is generally available from construction projects that can be stockpiled for construction of stone structures. Where environmental enhancement or aesthetics are key factors, rock structures may prove to be the most desirable option.

PAGE 389

Flood Control Planning for the American River Watershed, California Ricardo S. Pineda California Reclamation Board George T. Qualley California Department of Water Resources INTRODUCTION This paper describes the flood control planning efforts tmdertaken jointly by the U.S. AnllY Corps of Engineers, the Reclan1ation Board of the State of California, and the Sacran1ento Area Flood Control Agency (SAFCA) to significantly increase the level of flood protection in the heavily urbanized American River floodplain within the City and COtmty of Sacramento and south Sutter COtmty. These three agencies have completed comprehensive engineering and environmental studies that recommend a flood control dam at Auburn that will reliably provide a 500-year level of flood protection. The proposed "locally preferred" project is being considered in Washington for federal authorization and ftmding. While it is tmclear which of three candidate plans will be authorized by Congress and implemented, it is very clear that decisive action is needed in 1996 since the current level of flood protection in Sacramento is grossly inadequate for a commtmity of its size. A second theme is that floodplain management solutions must be responsive to the "consequence of failure" with respect to existing flood control systems protecting highly developed urban areas. Within this context, the major structural project reconunended for California's state capitol clearly is consistent with the long-term vision for floodplain management outlined in the Galloway Report, which includes special consideration for protecting critical infrastructure. Finally, some "lessons learned" by the State of California in working with the Corps of Engineers and local govenunent in tmdertaking a complex flood control investigation are described; in particular, the necessity to relate "early and often" with everyone who will either be affected by the outcome or can influence key decisions. These include

PAGE 390

Pineda and Qualley 375 communicating the risk of flooding and results of the studies to the public and to regional governmental and elected officials, building coalitions with environmental and other community organizations, and understanding the perspective of Corps headquarters and Washington-based elected officials to help guide the project through the federal authorization process. THE AMERICAN RIVER WATERSHED The American River basin drains about 2,100 square miles on the western slopes of the Sierra Nevada in Northern California. In the Sacramento area, at the confluence of the Sacramento and American rivers, the American River forms a floodplain covering roughly 110,000 acres that includes most of the City of Sacramento and the Natomas basin. The American River is a fast-moving stream with elevations ranging from 10,000 feet at the upper end to only 20 feet at its confluence with the Sacramento River. Travel time for floodwater from the upper basin to the confluence during a flood event can be less than 24 homs-which provides little lead time for flood fight activities and/or evacuation of people and property. THE FLOOD RISK In February 1986, the stonll of record in the American River watershed caused flows in the lower American River to exceed the system's design flood carrying capacity. The high flows on the American and Sacramento rivers nearly resulted in catastrophic flooding of the City of Sacramento and portions of unincorporated Sacramento and Sutter Counties. Within the floodplain, approximately 400,000 residents and over $37 billion in developed property and infrastructure are presently at risk. According to the Corps, the Sacramento area is the most developed urban area at risk from major flooding in the United States. FLOOD CONTROL STUDIES Prompted by Congressional hearings held after the 1986 flood, the Corps spearheaded a comprehensive review of Sacranlento's flood control needs. The joint studies culminated in the American River Watershed Investigation Feasibility Report (December 1991) and the American River Watershed Project Supplemental Infornlation Report (March 1996). A joint Environmental Impact Statement/Environmental Impact Report accompanied each technical report. Over $25 million has been spent over the past decade by the Corps, the state, and SAFCA to develop a long-ternl solution to Sacramento's flood threat. The Corps identified an 894,000 acre-foot flood control "dry

PAGE 391

376 Flood Control Planning for the American River Watershed dam" near Auburn as the plan that maximized the net national economic development benefits (NED plan). In late 1995, the Reclamation Board and SAFCA, representing local government, identified the Auburn flood control dam as the local sponsors' preferred plan for flood control. The recommended project is estimated to cost about $950 million, which would be spread out over a 10-year period. Two other flood damage reduction alternatives were identified in the 1996 Supplemental Information Report. The Folsom Modification Plan would provide about a 180-year level of protection and cost about $470 million. This plan would allow the maximum design release to be made from Folsom Dam much sooner than at present, and also would rely heavily on additional flood control space to be reserved in Folsom Reservoir during the flood season at the expense of water supply storage in the reservoir. The Stepped Release Plan would provide about a 235-year level of protection and would cost about $630 million. This plan would incorporate most of the elements of the Folsom Modification Plan, and would also include reinforcing levees downstream of Folsom Dam to allow release of higher flows down the lower American River-essentially "red lining" the levee system. The American River study was one of the Corps' first uses of risk based analysis to assess uncertainties in estimating and measuring design parameters, thus redefining the term "level of protection" to include such concepts as system reliability and residual risk. The study process also incorporated technically innovative flood control measures, including slurry cutoff walls on lower American River levees to ensure levee stability and control seepage; an adaptive management plan at the proposed Auburn flood control dam to minimize environmental impacts in the upstream watershed; and specific plan elements designed to mitigate hydraulic impacts. LOWER AMERICAN RIVER TASK FORCE The Lower American River Task Force was formed in 1993 by SAFCA, the Reclamation Board/Department of Water Resources, and the Corps of Engineers to address alternative ways to stabilize lower American River levees while retaining-and enhancing where possible-environmental and recreational values along the lower American River parkway. The planning process was expanded to directly include groups that had been opposed to the Auburn Flood Control Dam during the 1992 authorization process. The task force consists of 32 members representing 28 agencies and organizations. Task force activities are coordinated by professional facilitators and have been divided into four phases. Each phase has concluded with ratified proceedings that were used as a guide by the flood

PAGE 392

Pineda and Qualley 377 control project sponsors as input in the pi arming process. An important benefit of the task force was receipt and consideration of critical input from task force members early in the pi arming process, as compared to the usual process of simply "responding to connnents" after a draft report has been circulated for public and agency review. While differences of opinions among the project sponsors and the environmental and resource agency groups continue to exist, regular conummication and consensus building through task force activities has resulted in reconmlendations for substantial levee improvements that are complementary to any of the three candidate plans for long-term flood protection. Construction at the first levee improvement site will begin in 1996. THE WASHINGTON, D.C., PROCESS Because of the historic environmental and political controversy surrounding any proposal for a dam at the Auburn site, the Washington process has been difficult to predict. Working with the Corps technical staff at the district, division, and headquarters level was challenging yet rewarding. Their staff was professional, highly skilled, and responsive to our needs. However, a variety of "other influences" start coming into play at the Washington level. In the case of the American River project, balancing the federal budget is currently a critical priority-spawning proposals of new cost-sharing policies that require a larger share to be covered by nonfederal interests. Uncertainties related to both shortand long-term projections of funding availability, coupled with election-year politics, tend to foster a "nondecision atmosphere." Concurrent with the Washington-level review by budgetary and resource protection agencies, the federal political process comes into play as well. These processes constantly interact. For example, once Congressional subcommittees begin holding hearings on authorization language, technical reporting agencies such as the Corps usually are asked to testify; this testimony is often influenced by policy feedback provided to the reporting agency through the Washington-level review process. No judgment is being made on whether this is good or bad-it just is. Consequently, it becomes difficult for the reporting agencies to separate their primary mission of providing clear technical reconmlendations from the policy framework within which they exist. This overlap of "technical vs. policy" has occurred in the case of the American River project, where the proposed Chief of Engineers report (released for review in March 1996) has deferred a recommendation on a long-term flood protection plan for Sacramento lmtil completion of the 90-day state and federal agency review. The state and SAFCA have both urged the Corps to make a definitive reconunendation during June-while Congress is considering the Water Resources Development Act of 1996 authorization bill.

PAGE 393

378 Flood Control Planning for the American River Watershed All in all, the Washington process is very difficult for those who have worked for many years at the state and local level to formulate a project and obtain local support for the recommended plan. While it is not fair to say there is a lack of sensitivity for the extreme flood risk faced by a major metropolitan area like Sacramento, the fact is that members of Congress and their staffs have very short windows of time to assimilate mountains of information, and they must also deal with nunlerous other factors in their decision making. All too often, it probably comes down to whether the project proponents or opponents have crafted the most compelling "solmd-bite size" responses to key issues. A DAM DILEMMA The authors are strong advocates of a balanced approach to floodplain management, which involves careful consideration of both nonstructural and structural measures in addressing flood risk. And clearly, in the aftermath of the great flood of 1993 on the Mississippi River system, there is a move toward greater efforts for humans to coexist with major rivers and their floodplains-rather than "tame" them. The preventive concepts described in the Galloway Report-including recognition of what might be called a "chain of accountability" at the federal, state, local, and individual levels-are very appropriate as we move into the 21st century. It makes a lot of sense to relate the share of responsibility for exposure to risk to the level of decision-making regarding activities within the floodplain. On the other hand, many communities are currently "caught in the middle" of changing federal policies, and find themselves in situations of exposure to extreme flood risk that could result in loss of life and extensive property damage. TIle Galloway Report recognizes that "consequence of failure" must be addressed: Reconmlendation 4.2 specifically calls for "reducing the vulnerability of critical infrastructure to damage from the standard project flood discharge," which is defined in the report as SOO-year. The Sacramento flood situation has been studied in depth over the past 10 years, and the conclusion in 1996 is the same as in 1992: the American River is simply a much bigger river than anyone thought, and significant additional flood detention storage is needed upstream of the city to effectively handle the huge volumes of floodwater that could reasonably be expected. It would certainly be simpler if an effective long-term solution was available that did not require building a highly controversial flood control dam, but Sacramento is beyond the point where modifications to the existing system can provide the measure of security needed.

PAGE 394

Pineda and Qualley LESSONS LEARNED Working 10 years on a flood control study has led to many "lessons learned" including: 379 The nonfederal sponsor must actively participate in all phases of the technical process, assigning a study manager with adequate resources to ensure that the objectives of the nonfederal sponsor are carried out. This will help keep the study on schedule, and facilitate accountability from cost-sharing partners, consultants, and other "stakeholder" agencies and organizations. Develop community support anlong local leaders and elected officials, and encourage their participation on an executive steering committee. Establish and maintain conmlunications with traditional project opponents through a multi-organizational task force and incorporate the results of the task force into the planning process. Develop multi-objective project components that go beyond flood control, such as recreation features and environmental restoration and enhancement even if there is limited federal cost sharing for these features. Communicate with the public and conmnmity leaders through a variety of media, including public workshops and hearings; COOlllunity group meetings; conmnmity leader forun1S; project newsletters, brochures, and videos; public hearings; press conferences and press releases; meetings with newspaper editorial boards; public television specials; town hall meetings; direct mailings to community leaders and decision makers; and radio talk shows. A media consultant may be beneficial in developing a public relations and commlmications campaign, for controversial and high profile projects.

PAGE 395

Environmental Management vs. Floodplain Management at Reelfoot Creek in Western Tennessee David S. Smith WEST Consultants, Inc. Donald R. Davenport Roger A. Gaines U.S. Army Corps of Engineers, Memphis District INTRODUCTION A sediment retention basin is proposed for Reelfoot Creek in western Tennessee to help control erosion from the Reelfoot Creek watershed. Sediment retention/flood retarding structures already exist in the upper parts of the basin, but sedimentation of Reelfoot Lake remains a problem due to sediment production in the watershed. Placing a structure at the outlet of Reelfoot Creek will limit the t10w of sediment into Reelfoot Lake for fish and wildlife preservation and enhancement as well as recreational purposes, however, the structure will impact the local floodplain management. Nearby cultivated fields will be flooded more frequently and for longer durations. Transportation in the area may also be affected. The HEC-2 and HEC-6 computer models were employed to examine the changes in the basin that would be expected after 50 years of simulation. This infornlation will be used in the design of the retention basin as well as dredging and maintenance concerns at upstream bridges. BACKGROUND Reelfoot Lake, located in the northwest comer of Tennessee, is threatened with sedimentation as a result of high sediment production in the watershed. Currently, the lake has a mean depth of 5.2 feet and a normal pool volume of 80,300 acre-feet. Reelfoot Lake is fed by three major tributaries in the 240 mi2 drainage basin. The Reelfoot Creek tributary, which drains approximately 112 mi2 is the largest of the three. It has ten

PAGE 396

Smith, Davenport, and Gaines 381 sediment retention/flood retarding structures in the upper parts of the basin, which were constructed by the Soil Conservation Service between 1969 and 1995. Despite the trapping of sediment by these basins, high sediment production in the rest of the watershed continues to threaten Reelfoot Lake. To further control sedimentation of Reelfoot Lake, an in line sedimentation basin is proposed for Reelfoot Creek near the downstream outlet. The slopes of the channel and overbank areas of Reelfoot Creek near the downstream outlet are very flat, about 0.00005 lIft. Much of the land in this downstream region is cultivated by farmers who grow seasonal crops such as com, soybeans, cotton, and winter wheat. There is also a state highway in the area, Highway 22. With the sedimentation basin online, the cultivated fields will be subject to more frequent flooding, and sedimentation in the vicinity of the highway could affect local transportation if not properly maintained. APPROACH AND ANALYSIS General A sediment transport analysis is required to determine the volume of sediment expected to be deposited behind the sedimentation basin during its 50-year design life, as well as the anlount of dredging required at upstream bridges to maintain adequate conveyance. Hydrologic and hydraulic analyses are also required to determine the design flood hydrographs and corresponding water surface profiles. Two conditions were evaluated in the hydrologic and hydraulic analyses: (1) before most of the upstream sediment retention/flood retarding stmctures were on-line (1975 conditions), and (2) after construction (1995 conditions). Computer Models Hydrologic conditions before the upstream sediment retention/flood retarding stmctures were operating (1975 conditions) were compared with frequency curve data in order to calibrate the initial and uniform loss rates in HEC-l. Once these loss rates were calibrated, they were input to the 1995 conditions (with retention structures) HEC-l model. The resulting flows were transferred to the 1995 conditions HEC-2 model, and water surface profiles were computed. The HEC-2 file was then converted into a HEC-6 model, with bridge routines being replaced with a single cross section at the upstream face. The geometry of this cross section was modified for each bridge lmtil the water surface and flow velocities at all cross sections were within acceptable tolerances. Next, the inflowing sediment load, bed gradations, dredging templates, and other HEC-6 input data were entered, and the model was run using a 50-year historical

PAGE 397

382 Environmental Management vs. Floodplain Management rainfall histogram to simulate future conditions. With the HEC-6 modeling completed, the next step will be to convert the HEC-6 model back to HEC-2, and run the 1995 conditions hydrology to determine the backwater impacts of the sediment retention structure. Outlet Works A reconnaissance level investigation defined an initial configuration of the outlet works for the sediment retention structure, which is considered a low-hazard dam. The design includes three low-level spillway inlets (two 72-inch culverts each), a 175-foot primary spillway, and a 1,200-foot emergency spillway. Final design of the outlet works is contingent upon the following conditions being met. The elevation of the low-level spillway inlet should be set at the estimated elevation of the sediment pool at the end of the 50-year life of the structure. The low-level inlet should be designed for staged releases, which will provide maximum detention of the most frequent events while allowing full capacity of less frequent events. The outlet should be sized to evacuate the I-year exceedance frequency event in less than 10 days excluding capacity provided by outlets below the 50-year sediment pool. Once the submerged sediment pool is established, the I-year event should be routed using the estimated 50-year elevation of the sediment pool to set the elevation for the principal spillway. The principal spillway outlet should, as a minimum, be designed to prevent events of a magnitude less than approximately a 5to 7-year exceedance frequency from overtopping the emergency spillway. The l00-year nmoff should be routed through the basin to determine the crest elevation of the emergency spillway. This analysis should be conducted with the low level outlet and primary spillway functioning. The spillway crest and length should be determined such that the resulting 100-year peak elevation does not exceed elevation 305.0, which will limit the land acquisition costs and relocations. PRELIMINARY FINDINGS Figure 1 shows the preliminary design of a typical low-level inlet riser structure. This design allows for staged releases for low flows, so that as sediment accun1Ulates in the sedimentation basin, stop logs will be placed in the orifice openings (see upstream view, Figure 1) lmtil the sediment pool elevation is reached. For higher stages above the sediment pool elevation and below the primary spillway elevation, flows will reach the culverts by flowing into the top of the riser only. The sediment pool elevation shown is only an estimate. 111e actual elevation is found through a trial and error procedure that involves making an initial estimate, running the HEC-6 50-year simulation to verify the estimate, revising the

PAGE 398

Note: Provide stop-log slots at 24" openings. 24" openings to be permanently blocked as sediment pool fills. S'l CB 48" 24" Dio. Typical Upstream View Figure 1. Typical riser inlet structure. 10' Plan View to! I' EI. 301.0 (Sediment Pool) -._,,-,-Flow Invert EI. 290.0 Profile View li? Cii .::t III ::J 0.. S CD C/) OJ

PAGE 399

9 10pO 20100 30pO 40100 50100 60pO 70100 80100 90pO 10 900 I I I I I I I I I I I I I I I I I I I I I I I I I I I I N9tes: 1. Culvert J Invert @ g90.0 In et Structu e Constr cted to 3 1.0 2. I I Sediment Pool 301.d\ I I I I I I I I 100-yrl Pool EI. I I I I I I I of Eq I Ednergency Spillway -\ I I I I I I t I 1 divert th I Iinle Strufture I I I I I I I I I I I I I Figure 2. Profile along center of dam. (Orientation is left to right, looking downstream.) 11900 I I I I I I I I I I I I I I I (,) en m ::::J < a ::::J 3 <1> ::::J !!!.. III ::::J III <0 <1> 3 <1> a < !" JJ 0 0 Co "U Dr :i" III ::::J III co <1> -

PAGE 400

Smith, Davenport, and Gaines 385 initial estimate, and repeating. All revisions must meet the conditions discussed above, such as evacuating the I-year event in 10 days or less, and not overtopping elevation 305.0 when routing the 100-year flood through the outlet works. Figure 2 shows the preliminary design layout of the outlet works, as viewed looking downstream. The width of the primary spillway was increased from 175 feet to 600 feet to prevent the 5-year event from overtopping the emergency spillway and to prevent the 100-year event from exceeding elevation 305.0. The culvert size was reduced from two 72-inch culverts at each inlet structure to one 48-inch culvert to allow the I-year event to drain within 10 days without overtopping the primary spillway elevation. This preliminary design traps about 100% of the inflowing sand load and about 60% of the inflowing silt load. The next step will be to investigate ways to reduce the width of the primary spillway without excessively compromising sediment trapping efficiency. SUMMARY To help control sedimentation of Reelfoot Lake for recreational purposes and fish and wildlife preservation and enhancement, a sediment retention basin is proposed for Reelfoot Creek in western Tennessee. The design of the basin will depend upon the results of hydrologic, hydraulic, and sedimentation studies. As a preliminary estimate, the outlet works will be configured as shown in Figures 1 and 2. This design consists of a 600foot primary spillway, a 1200-foot emergency spillway, and three inlet riser structures, each connecting to a single 48-inch culvert. The inlet riser structures are designed to drain the I-year flood in less than 10 days, which will minimize the duration of ponding on nearby cultivated fields. For larger events, the spillways are sized such that floods up to the 100year recurrence interval will pass through the structure without exceeding elevation 305, a condition that will reduce land acquisition costs and relocations. REFERENCES U.S. Army Corps of Engineers, Memphis District 1988 Reelfoot Lake, Tennessee and Kentucky, Reconnaissance Report. u.s. Geological Survey 1985 Water Budget and Suspended-Sediment Inflow for Reelfoot Lake, Obion and Lake Counties, Northwestern Tennessee, May 1984-April 1985. Water-Resources Investigation Report 85-42.

PAGE 401

This page IS intentionally blank

PAGE 402

Part 13 CONFERENCE SUMMARY

PAGE 403

This page IS intentionally blank

PAGE 404

Coast to Coast: Twenty Years of Progress Frank H. Thomas Loudon, Tennessee The scope and quality of the presentations at this conference clearly have demonstrated the progress and accomplishments of the Association of State Floodplain Managers' first 20 years. The Association's record is an outstanding statement of leadership in and commitment to our nation's progress toward managing its floodplain resources and risks. As we draw back and look at the conference program and at the achievements of our Association, we cannot escape an obvious analogy with the concept of "confluence." In this context, confluence is not merely the coming together of natural energy in hydrologic flows. It is the integration of hmnan energy in the fonn of major issues, their associated ideas, approaches, and organizations. It is the interplay, shaping, and reshaping of thoughts. Turning first to the major issues, two pervasive ones stand out among the many faced by floodplain managers. TIle inherent policy conflicts between land use development and natural hazard loss reduction, including the preservation of natural floodplain functions, have consumed an enornlOUS amount of time and energy. We have stmggled to define and resolve an endless l1tunber of land use related problems. Similarly, we have struggled with tJle endless need to build closer working relationships within a multi-governmental, multi-hazard, and multi-disciplinary framework. To cope with tJlese issues, we have fostered development of concepts such as "multi-objective management," "unified national program," and IIm itigation." Curiously, in 1979 the tenn "mitigation" was excised from the U.S. Water Resources COlmcil's Unified National Program as being IItoo threatening." To bring life to our concepts, we have sought to assure tJlat full consideration be given risks to the natural and beneficial functions of floodplains along with risks to human life and property. We have sought to achieve equity of consideration anlOng structural and nonstmctural loss

PAGE 405

390 Coast to Coast: Twenty Years of Progress reduction measures. We have sought to integrate the application of loss reduction measures through the incentives of the National Flood Insurance Program's Community Rating System. To coordinate and support our efforts, we have built the organizational stmcture of the Association to focus thinking on specific floodplain issues. We have supported the creation of associations with similar interests-the Association of State Danl Safety Officials and the Association of State Wetland Managers. Also, working relationships have been developed with the research community through the Natural Hazards Research and Applications Information Center and with the emergency management conmlunity through the Federal Emergency Management Agency. The Association can rightly pride itself for initially focusing on a single problem-flooding-and a single resource-the floodplain-and successfully drawing together practitioners from relatively disparate disciplines and institutions into a highly effective, professional organization recognized for its expertise. As we look to the future, we must be aware that other important hazard-specific and emergency management organizations have been developing in much the same manner. One exanlple is the Central United States Earthquake ConsortilUll, another is the National Emergency Managers Association. And the insurance industry has established the Insurance Institute for Property Loss Reduction. The many streams of individual hazard management and emergency management are on the threshold of coming together. Legislation to establish a national, multi-hazard, insurance program is being discussed by the Congress. The concept of a national emergency management system is being discussed at FEMA and in the emergency management community. The future clearly points toward increased ties anlOng floodplain managers, other natural hazard managers, and emergency managers. In the next 20 years, the major issues of land use management and complex institutional frameworks will remain before us. The continued integration of concepts, approaches, and organizations will create the fabric of a national emergency management system. The prospect of such a multi-hazard, multi-risk, integrated system is exciting, if somewhat daunting. However, as we look ahead, we stand strengthened by our Association and the achievements of floodplain managers over the past two decades. We can expect to be central players in the new fabric. We need not give up our identities nor our conmlitment as floodplain managers. To continue to excel, we must further hone our flood expertise, broaden our understanding of the management of other hazards, and actively participate in the shaping of the emerging emergency management system.

PAGE 406

Thomas 391 Having viewed the progress and status of the Association with the analogy of a "confluence," which draws together and enriches human energy flows, it seems appropriate to look at the immediate future with the analogy of a delta. A delta is a distributary flow and it also means change. We leave the conference enriched by the interchange and thinking of others and with the task of distribution: the sharing of our new knowledge, viewpoints, and networks with others. And as we go forward, we must strive to keep a constant focus with one eye on the ground before our feet and one eye on the horizon. Then we can go forward into the next 20 years well grounded in our floodplain expertise and also guided in the direction of the emerging scientific and institutional framework that will allow us to serve our nation most effectively.

PAGE 407

This page IS intentionally blank

PAGE 408

,.,. Related Publications of Interest from the Natural Hazards Center All items can be ordered from the Natural Hazards Research and Applications Information Center Campus Box 482 University of Colorado Boulder, CO 80309-0482 (303) 492-6819 e-mail: hazctr@spot.colorado.edu WWW: http://addeLcolorado.edu/ hazctr/Home.html Monograph Series MG53 Coastal Erosion: Has Retreat Sounded? Rutherford H. Platt et at. 1992.210 pp. $20.00. MG54 Partnerships for Community Preparedness. David F. Gillespie. 1993. 150 pp. $20.00. Special Publications SP25 Action Agenda for Managing the Nation's Floodplains. A Review of Floodplain Management in the United States: An Assessment Report. 1992. 22 pp. $20.00. SP27 When the River Rises: Flood Control on the Boise River 1943-1985. Susan M. Stacy. 1993. 209 pp. $20.00. SP28 Guidelines for the Uniform Definition, Identification, and Measurement of Economic Damages from Natural Hauud Events. Charles W. Howe and Harold C. Cochrane. 1993. 28 pp. $20.00.

PAGE 409

SP30 NANIA: "All Together" -Comprehensive Watershed Management. Proceedings of the Eighteenth Annual Conference of the Association of State Floodplain Managers. May 8-13, 1994. Tulsa, Oklahoma. 472 pp. $20.00. SP31 From the Mountains to the Sea-Developing Local Capabilities. Proceedings of the Nineteenth Annual Conference of the Association of State Floodplain Managers. May 22-26, 1995, Portland, Maine. 1996. 490 pp. $20.00. Working Papers in Print WP81 Sullivan's Island, South Carolina-The Hurricane Hugo Experience: The First Nine Months. Jamie W. Moore and Dorothy P. Moore. 1993.64 pp. $9.00. WP82 Biological Hazards and Emergency Management. Janet K. Bradford et al. 1992. 27 pp. $9.00. WP83 Natural Hazard Trends in the United States: A Preliminary Review for the 1990s. Pamela Sands Showalter, William E. Riebsame, and Mary Fran Myers. 1993. 58 pp. $9.00. WP84 The Public Policy Response to Hurricane Hugo in South Carolina. Elliott Mittler. 1993. 72 pp. $9.00. WP85 The Evolution of Flood Hazards Programs in Asia: The Current Situation. James L. Wescoat, Jr. and Jeffrey W. Jacobs. 1993. $9.00. WP87 Insurance and Natural Disasters: An Examination of the New Zealand Earthquake and War Damage Commission. Arnold R. Parr. 1994. 27 pp. $9.00. WP88 Natural Disaster Management in Korea: An Analytic Study with Policy Implications. Wook-Joong Kim. 1994. 94 pp. $9.00. WP90 Dreading the Next Wave: Nontraditional Settlement Patterns and Typhoon Threats on Contemporary Majuro Atoll. Dirk H.R. Spennemann. 1995. 42 pp. $9.00. WP91 The Hyatt SkyWalk Disaster and Other Lessons in the Regulation of Building. William L. Waugh, Jr. and Ronald John Hy. 1995. 15 pp. $9.00. WP93 Renewing FEMA: Remaking Emergency Management. Richard Sylves. 1995. 36 pp. $9.00. Working Papers on the World Wide Web WP94 Hurricane Damage to Residential Structures: Risk and Mitigation. Jon K. Ayscue. 1996. 11/13/96. Free. (Not available in printed form.)

PAGE 410

Quick Response Reports in Print QR68 Risk Factors for Death in the 27 March 1994 Georgia and Alabama T017ll1does. Thomas W. Schmidlin and Paul S. King. 1994. 15 pp. $5.00. QR70 Children of Iniki: Effects of Evacuation and Intervention. Roger S. Hamada. 1994. 23 pp. $5.00. QR72 Immediate Emotional Response to the Southern California Firestorms. E. Allison Holman and Roxane Cohen Silver. 1994.23 pp. $5.00. QR73 Residential Loss and Displacement Among Survivors of the 1993 Altadena Fire. Norma S. Gordon et al. 1994. 15 pp. $5.00. Quick Response Reports on the World Wide Web QR76 Farmers' and Public Responses to the 1994-1995 Drought in Bangladesh: A Case Study. Bimal Kanti Paul. 1995. $5.00 for printed copy. QR77 Psychophysiological Indicators of PTSD Following Hurricane Iniki: The Multi-Sensory Interview. Kent D. Drescher and Francis R. Abeug. 1995. $5.00 for printed copy. QR78 Self Organization in Disaster Response: The Great Hanshin, Japan Earthquoke of January 17, 1995. Louise K. Comfort. 1996. $5.00 for printed copy. QR79 Transition from Response to Recovery: A Look at the Lancaster, Texas Tornado. David M. Neal. 1996. $5.00 for printed copy. QR81 Newspaper Reporting in Wake of the 1995 Spring Floods in Northern California. Ute J. Dymon and Francis P. Boscoe. 1996. $5.00 for printed copy. QR82 Early Response to Hurricane Marilyn in the U.S. Virgin Islands. Betty Hearn Morrow and A. Kathleen Ragsdale. 1996. $5.00 for printed copy. QR84 Impact of Hurricane Opal on the Florida/Alabama Coast. David M. Bush et aI. 1996. $5.00 for printed copy. QR85 The Potential Impact of Information Technology on the Structure of Inter organizational Relationships during Crisis Response: The Pennsylvania Floods of 1996. Diana Burley Gant. 1996. $5.00 for printed copy. QR86 The Politics and Administration of Presidential Disaster Declarations: The California Floods of Winter 1995, Richard Sylves. 1996. $5.00 for printed copy.

PAGE 411

QR87 Coping Self-Efficacy and Psychological Distress Following the Oklahoma City Bombing, Charles C. Benight. 1996. $5.00 for printed copy. QR88 Response to Severe Winter and Blizzard Conditions in Grundy and Buchanan County, Virginia in 1996: A Focus Group Analysis. Joseph B. Perry, Duane Dukes, and Randall Norris. 1996. $5.00 for printed copy. QR90 Tornadoes in the Districts of Jamalpur and Tangail in Bangladesh. Thomas Schmidlin and Yuichi Ono, 1996. $5.00 for printed copy. Topical Bibliographies TB16 A Bibliography of Weather and Climate Hazards. Jolana Machalek. 1992. 335 pp. $30.00. TB18 Epidemiology of Disasters: A Topical Bibliography. Eric K. Noji. 1994. 69 pp. $20.00. TB19 The Socioeconomic Aspects of Flooding in the U.S.: A Topical Bibliography. John Wiener. 1996. 49 pp. $20.00. Also available via the Internet for free at http://adder.colorado.edu/ hazctr/Home.html. The Natural Hazards Observer The Natural Hazards Center publishes a bimonthly newsletter, the Natural Hazards Observer, which covers disaster management, mitigation, and education programs; information sources; research and findings from completed projects, recent legislation; applications of research at federal, state, and local levels and by private organizations; recent publications; and future conferences. Subscriptions to the printed version of the Observer are free within the U.S., and foreign subscriptions cost $15.00 per year. Beginning with Volume XX, No.4 (March 1996), back issues of the Observer are also available at the Hazards Center's Web site: http://adder.eoiorado.edu/-hazetr/ Home.html.

PAGE 412

How to Obtain Our Publications Shipping and handling charges must be added to all orders. Based on the total number of pages in an order, they may be calculated from the chart below. For larger orders, contact the Publications Clerk, Natural Hazards Research and Applications Infonnation Center, Campus Box 482, University of Colorado, Boulder, CO 80309-0482; (303) 492-6819;fax: (303) 492-2151; e-mail:jclark@colorado.edu. All orders must be prepaid and checks should be payable to the University of Colorado. American Express, Visa, Mastercard, and Diners Club cards are accepted. To obtain a printed copy of the full 8-page list of Natural Hazards Center publications, send $3.00 to the Publications Clerk at the above address. To receive a free copy via the Internet, send an e-mail message to jclark@colorado.edu or access the Natural Hazards Center's Home Page at http://adder.colorado.edu/-hazctr/Home.html. Shipping Charges DOMESTIC # of Pages Printed Matter First Class 0-35 $3.00 $3.00 36 80 $3.50 $4.00 81 450 $4.00 $5.00 CANADA AND MEXICO # of Pages Surface Printed Air Printed 0-35 $3.00 $3.00 36 80 $3.50 $4.50 81 450 $5.00 $6.00 INTERNATIONAL # of Pages Surface Printed Air Printed 0-35 $4.00 $5.00 36 80 $5.00 $6.00 81 450 $6.00 Call for nrice