USF Libraries
USF Digital Collections

Precipitation variability of streamflow fraction in West Central Florida

MISSING IMAGE

Material Information

Title:
Precipitation variability of streamflow fraction in West Central Florida
Physical Description:
Book
Language:
English
Creator:
Scott, Michael H
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Coastal
GIS
Ungaged
Tidal effects
Estuaries
Dissertations, Academic -- Civil Engineering -- Masters -- USF
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: There is a strong interest to develop a method to estimate mean annual ungaged streamflow with varying precipitation. A method was developed utilizing GIS and other statistical analysis to estimate ungaged mean annual streamflow. This method utilizes a normalized streamflow fraction (NSF) method previously developed which relies on drainage basin area, coupled with mean annual local precipitation, to estimate the ungaged streamflow variability. This method has been applied to west central Florida.The test of the method yielded an R squared value of 0.9894, proceeded by a verification that yielded an R squared value of 0.998. This method is believed to be generally applicable to other areas and the particular results should be useful in and around west central Florida and perhaps, other coastal plain environments.
Thesis:
Thesis (M.S.C.E.)--University of South Florida, 2006.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Michael H. Scott.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 45 pages.

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 - 001916253
oclc - 181017485
usfldc doi - E14-SFE0001793
usfldc handle - e14.1793
System ID:
SFS0026111:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Precipitation Variability o f Streamf low Fraction i n West Central Florida by Michael H. Scott A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Civil Engineering Department of Civil and Environmental Engineering College of Engineering University of South Florida Major Professor: Mark Ross, Ph.D. Ahmed Said, Ph.D. Xinjian Chen, Ph.D. Date of Approval: November 17 2006 Keywords: coastal, GIS ungaged, tidal effects, estuaries Copyright 2006, Micha e l H. Scott

PAGE 2

ACKNOWLEDGEMENTS I would like to take a moment to thank all those that helped me in conquering this monumental task. To my mom and dad, thank you for supporting my efforts and aspirations to be all I could be and for the monetary support when I was broke. To my major professor, Dr. Ross, thank you for all the professional advice and for being the constant motivator to get this thesis done. To Dr. Trout, Dr. Said, and Dr. Chen, thank you for all your assistan ce in my times of crisis. To my partner in crime during this thesis, Kim, thank you for your hard work and dedication. And l ast, but certainly not least, to my beautiful wife April, thank you for your understanding, patience, and support during the past five and a half years of this endeavor, I love you so much.

PAGE 3

i TABLE OF CONTENTS LIST OF TABLES ii LIST OF FIGURES iii ABSTRACT v CHAPTER ONE INTRODUCTION 1 P urpose 2 P revious Studies 2 CHAPTER TWO D ESCRITION OF STUDY AREA 5 M ethodology 8 CH APTER THREE RESULTS 2 5 CHAPTER FOUR D ISCUSSIONS AND CONSLUSIONS 3 7 REFERENCES 40 APPENDICES 42 A ppendix A: Streamflow Estimates for West Central Florida 43

PAGE 4

ii LIST OF TABLES Table 1: Annual Precipitation Statistics for West Central Florida 1 2 Table 2: Storm Activity in West Central Florida from 1993 2003 1 8 Table 3: Critical Values for Kendalls Rank Corre lation Coeffiecient Tau with Sample Size n 24 Table 4 : Streamflow Estimates for West Central Florida 43 Table 5 : Verifi cation of Sreamflow Estimates for West Central Florida 4 5

PAGE 5

iii LIST OF FIGURES Figure 1: StreamStats Availability 3 Figure 2: West Central Florida Study Domain 7 Figure 3: Sub basin Delineation and USGS Streamflow Gages 9 Figure 4: Florida Popul ation Growth from 1960 200 0 1 1 Figure 5: Annual Precipitation for 1983 2003 12 Figure 6: Comparison of Annual Precipitation for 1983 1992 and 1993 2002 1 3 Figure 7: North and South Precipitation Zones and Corresponding Gages 1 5 Fi gure 8: Comparison of Annual Precipitation for North and South Zones (1983 2003) 16 Figure 9: Depth to water table for West Central Florida 1 7 Figure 10: Normalized Streamf low Fraction for West Central Florida 2 2 Figure 1 1 : Regression of Annual Precipitation Versus Streamflo w (1993 2003) 2 5 Figure 12: Two Year Preceding Average Precipitation Versus Streamflow ( 1993 2003) 26 Figure 13 : Mean Annual Precipitation and Streamflow (1993 2003) 27 Fi gure 14 : Modified Mean Annual Precip itation and Streamflow Period of Record (1993 2003) 28 Figure 1 5 : Three Year Preceding Average Precipitation Versus Streamflow (1993 2003) 28 Figure 1 6 : Annual Precip itation and Streamflow for North Zone (1993 2003) 29

PAGE 6

iv Figure 1 7 : Annual Precipitation Versus Streamflow for North Zone (1993 2003) 30 Figure 1 8 : Two Year Preceding Average Precipitation Versus Streamflow for North Zone (1993 2003) 30 Figure 19 : Annual Precipitation and Stream flow for South Zone (1 993 2003) 31 Figure 2 0 : Regression of Annual Pre cipitation Versus Streamflow for South Zone (1993 2003) 32 Figure 2 1 : Two Year Preceding Average Precipitation Versus Streamflow for South Zone (1993 2003) 32 Figu re 22 : NSF Versus Rainfall Sensitivity Slope, m, for We st Central Florida 33 Figure 23 : NSF Versus Rainfall Sensitivity Slope, m, fo r West Ce ntral Florida U sing Modified Two Year Preceding Average Precipitation Method 34 Figure 2 4 : Predicted Versus Observed Mean Annual Streamflow (1993 2003) 35 Figure 25 : Predicted Versus Observed Mean Annual Streamflow for Verification Data Set (1993 2003) 36

PAGE 7

v PRECIPITATION VARIABILITY OF STREAMFLOW FRACTION IN WEST CENTRAL FLO RIDA Michael H. Scott ABSTRACT There is a strong interest to develop a method to estimate mean annual ungaged streamflow with varying precipitation. A method was developed utilizing GIS and other statistical analysis to estimate ungaged mean annual streamflow. T his method u tilizes a normalized streamflow fraction (NSF) method previously developed which relies on drainage basin area, coupled with mean annual local precipitation to estimate the ungaged streamflow variability. T his method has been a pplied to w est c entral Florida The test of the method yielded an R squared value of 0.989 4, proceeded by a verification that yielded an R squared value of 0.998 This method is believed to be generally applicable to other areas and the particular result s should be useful in and around west c entral Florida and perhaps, other coastal plain environments.

PAGE 8

1 CHAPTER ONE I NTRODUCTION The re is a continued need for mean annual flow estimates using simple techniques for environmental studies and other water resource assessments. Streamflow must be considered when planning for various w ater resource projects such as estuary management, storm water impacts, and impact of development. Expensive and time consuming mathematical modeling is not always an option. Nor, in the absence of calibration data, is it necessarily more reliable. One common method for determining the streamflow for an ungaged area is to use flow measurements or records from the nearest streamflow gaging station and estimate the streamflow for the desired area scaled by the ratio of the drainage areas (USGS, 2006) Depending on the distance and hydraulic similarity between the watershed contributing to the flow station and the area to be estimated, it has been shown that considerable errors can arise (Clayback, 2006). This method usually leads to lower than actual values when a gage is installed to verify the estimates. A procedure was developed that yield s flow estimate s for ungaged areas using several key hydrologic variables This pro cedure was tested and verified on west central Florida streamflow measurements reported by USGS.

PAGE 9

2 P urpose This study presents the precipitation variability of mean annual streamflow in west central Florida. The purpose of this study is to relate the beh avior of streamflow to precipitation and to discover the time variability of this relationship and how it differs regionally within the study domain. P revious Studies There have been several studies done prior to this project regarding ungaged stream flo w; however, none have been done specifically for the preceiptation variability of streamflow in west central Florida. Additionally, of the reports that exist on this topic, all of them speak of some type of mathematical analysis on the streamflow, disrega rding any physical correlations that may exist. One method developed, estimates August median streamflow for ungaged, unregulated streams in eastern coastal Maine (Lombard, 2004). This method took into account the drainage basin area and the percent un derlain b y a sand and gravel aquifer. Lombard (2004) also related base flow measurements at partial record and short term continuous record streamflow gaging stations to concurrent daily streamflows at nearby long term continuous re cord streamflow gaging stations. A generalized least squares regression analysis was used to develop equations that were applied to estimate August median streamflow on ungaged streams. The equations that were developed resulted in an error of prediction ranging from 30 to 43 percent (Lombard, 2004). With this relatively large amount of error, it was concluded that

PAGE 10

3 improved estimates of basin characteristics could be important to the improvement of low flow estimates. The United States Geological Survey (USGS) generated a pr ogram called StreamStats that can be used to estimate streamflow statistics for ungaged sites (United States Geological Survey, 2004). This program is a web application that uses existing USGS data and ESRI ArcGIS (Ormsby, et al., 2004) to analyze surroun ding basins and their respective regression equations to predict the flows for the ungaged site. However, StreamStats assumes that the error for the ungaged sites is the same as the known sites, which could be hazardous to the accuracy of the calculated d ata. Figure 1 show s the states in which StreamStats has been implemented Figure 1: StreamStats Availability (Red = Impleme nted, Blue = Implementat ion in P rocess) (USGS, 2004) In a study done by Teemu Kokkonen (2003) rainfall runoff predictions were discussed. One aspect of this study was to look at those catchments that lacked

PAGE 11

4 streamflow records within the Coweeta Hydrologic Laboratory in North Ca rolina and to predict the runoff for these catchments using data from other catchments within the same region that had streamflow records. However, Kokkonen concluded that it would be more ideal to incorporate the observed physical catchment properties in to the model structure and parameters (Kokkonen, 2003). Horn (1988) examined annual streamflow records for 124 stream gages in and near Idaho to determine the annual flow characteristics. Two sets of equations for north central and southern Idaho provi ded the best predictive results, with the equations for north central Idaho yielding multiple correlation coefficients in excess of 0.97 (Horn, 1988). The equations coupled with the maps that were developed in this study can be used to estimate the annual flow at and ungaged location throughout Idaho. Another attempt to use spatially weighted averages to estimate ungaged streamflow was done by Altunkaynak, Ozger, and Sen (2005). A standard regional dependence function was proposed to describe the weighte d average using available data points. However, it was concluded that the discharge at any particular station was better described as a function of discharge at 3.5 closest stations. Validation of this method yielded streamflow predictions with less than 10% relative error. Kroll, Luz, Allen, and Vogel (2004) realized that regional hydrologic models of low flow processes often produce estimators with unacceptably large errors. Using the watershed boundaries from the USGS, many watershed characteristics were developed from digital grids. The inclusion of hydrogeologic indices, inparticular a new smoothed baseflow recession constant estimator, led to dramatic improvements in low flow prediction. However, no quantitative results were reported.

PAGE 12

5 CHAPTER TWO D ESCRIPTION OF STUDY AREA The study domain in west c entral Florida is approximately 10,400 square miles encompassing 16 counties with a population of 3.1 million people (Figure 2 ). The average precipitation across the domain for the eleven year p eriod used in the study was 52 inches per year, but precipitation shows substantial spatial and temporal variability. On average, the driest months of record are November and April, with the wettest being July and August (NOAA, 2006). The mean annual tem perature is 73F. The mean annual open water evaporation rate for the region is 52 inches per year (Ruskauff et al., 2003). The study area is over the surficial Floridan, and Intermediate aquifers. The surficial aquifer system is predominately sand, the I ntermediate aquifer system is interbedded siliciclastics and carbonates a nd the Floridan aquifer system consists of massive carbonates (Tihansky and Knochemas, 2001). The central part of the study domain shows carbonate units dipping and becoming overlain by the thickening Hawthorne Formation ( a distinct carbonate unit) that forms the Intermediate aquifer system south of Tampa Bay. Below the Intermediate Aquifer System is a confining unit for the Floridan Aquifer and the presence of the confining unit is the primary cause of the

PAGE 13

6 change in geologic environments between the north and south portions of w est c entral Florida. The three largest rivers in the study domain are the Withlacoochee, Hillsborough and Peace Rivers. The combined drainage area of the th ree watersheds is 5100 square miles, more than half of the study area. In the north, the Weeki Wachee, C hassahowitzka, Homosassa and Crystal Rivers all originate from coastal springs. In the south, the Alafia, Little Manatee and Manatee Rivers all termin ate into Tampa Bay along with the Hillsborough River. The Myakka and Peace Rivers terminate into Charlotte Harbor. The land surface elevations for the study domain range from just over 200 feet above sea level to sea level. Ridge systems are found in th e interior and along the eastern boundary of the study domain, with the largest being the Lake Wales Ridge to the West and the Brooksville Ridge in the no rtheastern corner of the domain.

PAGE 14

7 Figure 2: West Central Florida Study Domain

PAGE 15

8 Methodolo gy To create the study area in GIS, several shape files were used from the Southwest Florida Water Management District (SWFWMD). The files were imported into GIS and then overlaid with a basin delineation coverage obtained from a previous study ( Geurink 2000). The streamflow gaging stations used for this study were all USGS gages with an eleven year period of record from 1993 to 2003. All of the gages are read and maintained on a regular basis by the USGS Each gaging station was selected on the bas is of elevation, location, period of record and previous studies. The overall study a rea was divided into sub basins using previous USGS delineations, further interpreted to close at USGS gaging stations. Basins were closed to incorporate only areas that contributed to the gaging station; resultant basins are shown in Figure 3.

PAGE 16

9 Figure 3: Sub basin Delineation and USGS Streamflow Gages

PAGE 17

10 The pre cipitation data for the area were obtained from NO AA www.NOAA.gov Initially, twenty five years (1981 2005) o f precipitation records at thirty one unique gaging stations were downloaded from NOAA; however, most of the gages had extensive missing (or incomplete) data during part of the selected study period (1981 and 1982 rainfall records ) Therefore, all 1981 th ru 1982 records were d isgarded Also, for this particular study the available streamflow records only went to 2003, so all 2004 thru 2005 precipitation data were deleted. After these edits, a twenty one year period of record remained for rainfall with m atching streamflow data. Some of the thirty one stations had some missing records for the 1983 thru 2003 record. In order to obtain some numerical value of reliability, all of the stations were analyze d to determine how much data were missing or incomple te. Eleven of the original thirty one gages had greater than 20% missing record s, and were therefore d isgarded Nevertheless, t he twenty one years of record was considered a reasonable period of record for several reasons. Analysis showed that this pe riod exhibited the desired mean and extreme rainfall characteristic observed for the region. Also, f rom 1960 to 2000, the population in Florida had increased by 400%; however, from 1980 to 2000, the population increased by only 50% ( Figure 4 ) (Census, 200 6). Therefore, it was inferred that anthropo genic stresses might be different for the two periods.

PAGE 18

11 Figure 4: Florida Population Growth from 1960 2000 (Census, 2006) As the population increased, the state became more urbanized, therefore changing th e rainfall/ runoff characteristics of the area. The mean annual precipitation fluctuations for the twenty one year period of record are shown graphically ( Figure 5 ) with the statistical information tabulated (Table 1) Year (yrs)

PAGE 19

12 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 Calender Year Preciptation (in) Figure 5: Annual Precipitation for 1983 2003 Table 1: Annual Precipitation Statistics for West Central Florida NOAA ID Avg. (in.) Min. (in.) Max. (in.) Std. D ev 236 52.50 27.31 72.61 8.97 369 51.93 26.10 66.20 10.77 945 55.75 43.96 78.33 9.47 1046 53.14 40.62 70.98 10.47 1641 50.20 2 8.92 85.10 12.93 2288 49.74 32.87 69.16 8.48 3153 53.71 34.00 78.08 12.23 3986 53.44 33.83 67.83 9.42 4289 51.29 35.66 63.25 8.21 4707 52.74 39.90 67.27 8.66 5076 50.35 29.26 66.88 8.85 5973 48.97 29.53 64.16 8.76 6065 60.20 44.03 82.31 13.26 6414 48.63 28.58 62.92 8.21 6880 53.47 30.05 71.70 11.65 7205 53.92 39.74 71.97 8.51 7397 51.34 32.85 81.06 9.98 7851 54.65 39.27 75.89 10.41 8788 46.01 29.85 67.71 10.46 9176 50.96 29.07 75.08 10.77 Overall Avg 52.15 33.77 71.92 10.02

PAGE 20

13 For this study only eleven years of rainfall record, 1993 thru 2003, w ere used due to the limited availability of reliable streamflow data prior to 1993. It was determined that the rainfall record from 1983 thru 1992 and 1993 thru 2003 were statistically similar. The 1983 thru 1992 record had an average of 50.38 inches of precipitation with a standard deviation of 6.21 inches, while the 1993 thru 2003 record had an average precipitation of 53.48 inches with a standard deviation of 7.62 inches. Below is a graph showing the similar characteristics between the two periods ( Figure 6 ). 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 1983/1993 1984/1994 1985/1995 1986/1996 1987/1997 1988/1998 1989/1999 1990/2000 1991/2001 1992/2002 Calender Year (yr) Precipitation (in) 1983-1992 Data 1993-2002 Data Figure 6: Comparison of Annual Precipitation for 1983 1992 and 1993 2002 Note that in the graph above (Figure 6) the calendar year only goes to 2002 for the second period of record. This was done solely for graphical purposes so that two ten year periods would be plotted against each other. After defining the period of record that was to be used, the precipitation data were spatially analyzed Due mainly to hydrogeologic differences pr eviously discussed, the west central Florida area is hydrologically divided by Interstate 4 going East West from

PAGE 21

14 Tampa to Orlando ( Ross, 2005 ). North of I 4 is henceforth referred to as the northern region or north zone and to the south of I 4 is the s outhern region, or south zone. ArcGIS utilities (Ormsby, et al., 2004) were used to spatially determine which precipitation gages were in the north and which were in the south ( Figure 7).

PAGE 22

15 Figure 7: North and South Precipitation Zones and Corresponding Gages

PAGE 23

16 T he north and south zones were then statistically analyzed using Microsoft Excel. To verify that the gages should be separated in the above manner, the entire twenty one year period of record was analyzed. The north zone gages had an average mean annual rainfall of 50.57 inches while the south zone gages had an average mean annua l rainfall of 52.82 inches. Also, depicted in Figure 8 below, the south zone had a greater mean annual precipitation than the north on average 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 Calender Year (yr) Precipitation (in) North South Figure 8: Comparison of Annual Precipitation for North and South Zones (1983 2003) This characteristic of the south having a greater precipitation is likely due to slight climatological differences with greater convective storm activity towards the south Also, previously menti oned geological differences result in the south having a shallower depth to water table in certain areas ( Figure 9 ), which would allow for a greater rate of evapotransporation and therefore a more productive environment for convective rainfall than in the north.

PAGE 24

17 Figure 9: Depth to water table for West Central Florida

PAGE 25

18 When the 1993 thru 2003 period was analyzed, it was discovered that an even greater difference existed between the north and south zones. For this eleven year period, the north had an a verage mean annual precipitation of 50.25 inches, while the south had an average mean annul precipitation of 53.89 inches. This greater difference could have been due to the heavy hurricane activity in the south zone during this period of record. During t hat time, twelve named storms ( eithe r hurricanes or tropical storms) came over Florida (NOAA, 2005). All twelve of these storms impacted the south zone, while only four of them significantly affected the north zone (Table 2) Table 2: Storm Activity in W est Central Florida from 1993 2003 Storm Name Hurricane Tropical Storm North Zone South Zone Michelle x x Gabrielle x x Gordon x x Irene x x Harvey x x Floyd x x Mitch x x Earl x x Josephine x x x Opal x x x Jerry x x x Erin x x x The streamflow gages were also divided into north and south zones by I 4. Mean annual, spatially averaged s treamflow was then plotted against precipitation for the entire domain and analyzed Preliminary results suggested that their might be a significant difference between the north and south zone streamflow precipitation relationship as a result of the

PAGE 26

19 hydrologic and climologic differences However, it was determined that the difference in precipitation between the two zones was not great eno ugh to be statistically meaningful. Therefore, for the remaining procedures, the study domain was analyzed without the separation into zones. A normalized streamflow fraction (NSF) was calculated for each streamflow gaging station used in this study accor ding to the method of Clayback ( 2006). The resultant NSF values are shown in Figure 10. The mean annual regional precipitation values were plotted against the NSF for each gage to determine the correlation that existed between them. The slope was then ta ken from each of these graphs for each gage and the NSF value for a precipitation of 52 inches 0 NSF This represented the NS F for the particular gage for an average precipitation (52 inch) rainfall year. These slope and 0 NSF values were then plotted against one another In order to use the most accurate precipitation data a localized mean annual precipitation was calculated for each of the streamflow gages This was done using Theisan polygons. Polygons were creat ed f or the precipitation gages that had at least 3 years of record Then a table was made for each of the years of record from 1993 2003 in which the local precipitation value was recorded for each streamflow station used in this study. NSF and the precip itation va lues were plotted for all gages using the localized mean annual precipitation. The slope was then taken from this graph for each gage and 0 NSF These NSF values were then plotted versus the slope values

PAGE 27

20 Analysis will be done using the multi year preceding average if th e precipitation was increasing, but if the precipitation was decreasing, the current year precipitation was used (referred to as the modified two year preceding average method). This method took into account th e apparent storage delays in streamflow response to wetter cycles that were found from the regression analysis The same procedure was followed to obtain the NSF versus slope graph (Figure 23). It is the objective of this research that, with only NSF and precipi tation data, one can determine the ungaged streamflo w of a desire d site It is first proposed to formulate a non dimensional annual discharge as, 0 P A Q Q b i i = ; i Q is the annual streamflow rate, b A is the basin area and 0 P is the long term mean annual precipitation (e.g., 52 inches in west central Florida). Also, a non dimensional annual precipitation volume, i P can be defined as 0 P P P i i = where 0 P is the long term mean annual precipitation for a basin i Using a linear equation of the form b mx y + = unknown streamflow at a station can be related to mean annual precipitation, i P as, i i i i b P m Q + = ) ( (1) Where i m and i b are the stream specific precipitation sensitivity slope and intercept, respectively. Noting that the long term mean annual streamflow 0 Q should follow the same relationship the equation can be written as: i i i i b P m Q + = ) ( 0 0 (2)

PAGE 28

21 It should be noted that 0 i Q is exactly the NSF value proposed by Clayback (2006) and previously discussed. Subtracting equation 2 from equation 1 to remove the intercept yields : ) ( 0 0 i i i i P P m Q Q = (3) This form of the equation yields precipitation variability of the streamflow desired. Noting that precipitation was non di mensionalized by dividing by mean annual prec ipitation ( 0 P ), 1 0 = i P then the following simplified equation was found 0 ) 1 ( i i i Q P m Q + = (4) Adapting equation 4 to dimensional discharge in cubic feet per second (cfs), a value for streamflow in an ungaged region can be expressed as, [ ] [ ] ) )( )( ( ) 1 ( 0 0 c P A Q P m Q i b i i + = (5) where c is a coefficient of unit conversions (e.g., for Q in cfs and 0 P in inches and b A in miles squared, c = .074). Using Figure 10 for a spatial referenc e 0 i Q can be taken obtained for a particular su b basin from the method of Clayback ( 2006).

PAGE 29

22 Figure 10 : Normalized Streamflow Fraction for West Central Florida

PAGE 30

23 The above derivation demonstrated that ) ( 0 m f Q = and that ) ( m f Q i = The next step was to determine a relationship for the precipitation sensi tivity variable i m The precipitation and NSF for each gage were plotted. The slope of each line (m) and the intercept at P = 52 inche s were then plotted on one graph. The resultant equation of the best fit line is below. 1246 908 13 + = x y (6) In equation 6, 0 i Q y = and i m x = By making these substitutions and rearranging, t he variable i m can be obtained from equation ( 7 ). 908 13 1246 0 = i i Q m ( 7 ) To test the above equation for determining i Q all i m values for the gages used in this study were calculated. These values along wit h the drainage basin area and local mean annual precipitation values were used to calculate the i Q for all the gages. These predicted values were then compared to the actual annual Q mean values that were collected for this study. To verify the method, streamflow was predicted for seven streamflow gaging stations that were not used in the development of the method. Some of the data used in the study was not normally distributed. Many parametric approach es rely on the data being norma lly distributed. Nonparametri c methods can be employed on non normal data set s Nonparametric methods should be used only when the underlying distribution is unknown or cannot be transformed to make it normal (Berthouex and Brown, 1994). Some of t he r esu lts obtained in this study were suspected to be nonparametric (non normal)

PAGE 31

24 Kendalls tau, T k is a measure of correlation between the strength of the relationship between two variables, regardless of whether the relationship is increasing or decreasing. Tau measures the strength of the monotonic relationship between an ordered paired observation, X and Y A monotonic relationship shows one variable increasing w h ile the other variable always increases or always decreases. Tau is a rank based procedure a nd is therefore resistant to the effect of a small number of unusual (nonparametric) values. Tau is dependent on the ranks of the data, not the values themselves and can be used where the data is limited (Helsel and Hirsch, 2002). The T k values will gene rally be lower than values of the traditional correlation coefficient r A strong value of r is 0.9 or higher. T he tau value corresponding to the same data set is about 0.7 (Helsel and Hirsch, 2002). With this in mind, Kendalls tau was calculated in addi tion to R squared for any of the data sets that had a sample size of 20 or less Table 3 list s the Kendall values necessary for a sample size, n, to achieve a 99% confidence level (Rohlf, 1969). Table 3: Critical Values for Kendalls Rank Correlation Coe fficient Tau with Sample Size n n a = 0.01 4 5 6 1.000 7 0.905 8 0.786 9 0.722 10 0.644 11 0.600 12 0.576 13 0.564 14 0.516 15 0.505

PAGE 32

25 CHAPTER THREE R ESULTS In testing the approach, regional average precipitation was first used Comparing mean annual streamflow to regional annual precipitation only show a modest relationship (Figure 11 ). The linear regression only showed a correlation of 58%; however, a regression for the two year preceding average precipitation versus annual s treamflow showed a slightly improved correlation of 68 % (Figure 1 2 ). R 2 = 0.5793 Tau = 0.6 n = 11 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 0.E+00 1.E+05 2.E+05 3.E+05 4.E+05 5.E+05 Total Streamflow (in^3/s) Precipitation (in.) Figure 1 1 : Regression of Annual Precipitation Versus Streamflow (19 9 3 2003)

PAGE 33

26 R 2 = 0.6769 Tau = 0.64 n = 10 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 0.E+00 1.E+05 2.E+05 3.E+05 4.E+05 5.E+05 Total Streamflow (in^3/s) Precipitation (in.) Figure 1 2 : Two Year Preceding Average Precipitation Versus Streamflow (19 9 3 2003) A possible reason for the improved correlation ( 68% ), for the two year preceding average mean annual precipitation versus streamflow may be from the apparent phase lag in streamflow reaction to rainfall shown in Figure 13 From the figure, it is observed that there is a one year lag in streamflow peak, but no lag in streamflow decline. An investigation of longer term preceding average precipitation/streamflow relationships revealed no improved correlations. In fact, correlations to longer term averaging significantly degra ded ( % 37 2 = R ) (see for example the three year average relations hip shown in Figure 15 ).

PAGE 34

27 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Calender year (yr.) Precipitation (in.) 0.00E+00 5.00E+04 1.00E+05 1.50E+05 2.00E+05 2.50E+05 3.00E+05 3.50E+05 4.00E+05 4.50E+05 5.00E+05 Total Streamflow (in^3/s) precipitation streamflow Figure 1 3: Mean Annual Precipitation and Streamflow (1993 2003) The phase lag was most pronounced when there was a peak in the precipitation, upon which the s treamflow peak laged by one year. However, when there was a decline in the p recipitation, there was a direct decline in the streamflow (no lag) With this observation, a modified two year preceding average mean annual precipitation versus st re amflow was explore d ( Figure 14 ) which had a correlation further improved to 77%. The modified two year preceding average approach used the first derivative (slope) of mean annual precipitation. If the slope was positive (increasing rainfall) between two subsequent years, the two year preceding average rainfall was used. If the slope was negative (decreasing rainfall), only rainfall for that year was used in the regression. The modified approach yielded an improved correlation (77%) over the simple one year or two year rainfall relationships. Therefore, this approach was used in all further analysis.

PAGE 35

28 R 2 = 0.7754 Tau = 0.71 n = 11 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 0.E+00 1.E+05 2.E+05 3.E+05 4.E+05 5.E+05 Total Streamflow (in^3/s) Precipitation (in.) Figure 14 : Modified Mean Annual Precipitation and Streamflow (1993 2003) To verify that the modified two year preceding average precipitation wa s the be st correlation, a three year preceding average precipitation versus streamflow was plotted (Figu re 1 5 ) R 2 = 0.3674 Tau = 0.33 n = 9 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 0.E+00 1.E+05 2.E+05 3.E+05 4.E+05 5.E+05 Total Streamflow (in^3/s/yr) Precipitation (in) Figure 1 5 : Three Year Preceding Average Precipitation Versus Streamflow (1993 2003)

PAGE 36

29 The spatial variability between the precipitation and the streamf low was very prominent. As mentioned in the methodology, the precipitation difference between the north and south zones was not statistically meaningful. However, the variation in the characteristics of the precipitation is interesting to note. In the no rth zone, the phase lag was similar to that of the overall domain (Figure 1 6 ). Interestingly, there was less correlation in the north zone when a linear plot was generated of annual precipitation versus mean annual streamflow, which yielded a correlation of only 44% ( Figure 17 ). 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Calender year (yr.) Precipitation (in.) 0.00E+00 5.00E+04 1.00E+05 1.50E+05 2.00E+05 2.50E+05 3.00E+05 3.50E+05 4.00E+05 4.50E+05 5.00E+05 Total Streamflow (in^3/s) precipitation streamflow Figure 1 6 : Annual Precipitation and Streamflow for North Zone (1993 2003)

PAGE 37

30 R 2 = 0.4407 Tau = 0.53 n = 11 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 0.E+00 1.E+05 2.E+05 3.E+05 4.E+05 Total Streamflow (in^3/s) Precipitation (in.) Figure 1 7 : Annual Precipitation Versus Streamflow for North Zone (1993 2003) However, when the two year preceding average mean annual precipitation ver sus streamflow was plotted for the north zone, an improved (but not profound) correlation (69%) was calculated (Figure 1 8 ). R 2 = 0.6924 Tau = 0.64 n = 10 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 0.E+00 1.E+05 2.E+05 3.E+05 4.E+05 Total Streamflow (in^3/s) Precipitation (in.) Figure 1 8 : Two Year Preceding Average Precipitation Versus Streamflow for North Zone (1993 2003)

PAGE 38

31 Perhaps owing to the differences in hydrogeology (better confinement between surficial and confined aquifers), the south zone yielded a more in phase graph when annual precipitation and annual streamflow were plotted for the period being investigated ( Figure 19 ). 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 Calender year (yr.) Precipitation (in.) 0.00E+00 5.00E+04 1.00E+05 1.50E+05 2.00E+05 2.50E+05 3.00E+05 3.50E+05 4.00E+05 4.50E+05 5.00E+05 Total Streamflow (in^3/s) precipitation streamflow Figure 19: Annual Preci pitation and Streamflow for South Zone (1993 2003) When annual precipitation was plotted against mean annual streamflow for the south zone, a correlation of 71% was calculated ( Figure 20 ).

PAGE 39

32 R 2 = 0.7105 Tau = 0.71 n = 11 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 0.E+00 1.E+05 2.E+05 3.E+05 4.E+05 5.E+05 Total Streamflow (in^3/s) Precipitation (in.) Figure 2 0 : Regression of Annual Precipitation Versus Streamflo w for South Zone (1993 2003) The two year preceding average mean annual precipitation versus streamflow ( Figure 21 ) yielded a lower correlation (61%) in comparison to the 71% of the previous graph (Figure 20 ). R 2 = 0.6087 Tau = 0.6 n = 10 0.00 10.00 20.00 30.00 40.00 50.00 60.00 70.00 0.E+00 1.E+05 2.E+05 3.E+05 4.E+05 5.E+05 Streamflow (in^3/s) Precipitation (in.) Figure 2 1 : Two Year Preceding Average Preci pitation Versus Streamflow for South Zone (1993 2003)

PAGE 40

33 The regional approach resulted in a 47% correlation. A higher correlation was desired so that the overall accuracy of the final equation would be as high as possible. A graph of this correlation is s hown in Figure 2 2 R 2 = 0.4724 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 m i NSF Figure 2 2 : NSF Versus Rainfall Sensitivity Slope, m, for West Central Florida For the localized mean annual precipitation a correlation of 51% was achieved. This was only a slightly better result than using an overall average prec ipitation value. Lastly, incorporating the previous rainfall/streamflow findings, a modified two year preceding average analysis yielded a correlation of 48 % for NSF versus i m This correlation is shown graphically in Figure 2 3

PAGE 41

34 R 2 = 0.4845 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.01 0.02 0.03 0.04 0.05 m i NSF Fi gure 2 3 : NSF Versus Rainfall Sensitivity S lope, m, for West Central Florida U sing Modified Two Year Preceding Average Precipitation M ethod This was the graph from which the equation was taken from for determining i m previously discuss ed in the methodology chapter. The results for determining ungaged streamflow using the procedure described in the previous chapter are presented in the following figures In Figure 24 the predicted mean annual streamflow is plotted versus the observed m ean annual streamflow for the eleven year study period.

PAGE 42

35 R 2 = 0.9894 0 50 100 150 200 250 300 350 0 50 100 150 200 250 300 350 Observed Q (cfs) Predicted Q (cfs) Figure 2 4 : Predicted V ersus Observed Annual Streamflow (1993 2003) The comparison of the predicted mean annual streamflow versus the observation yielded an average relative error of 1.85 % with a standard deviation of 8.11%. The maximum relative error was 16.34 %. The absolute relative error was 6.68% with a standard deviation of 4.89%. Seven streamflow stations that were not used in the above approach were used to verify the calibration data. A Q value was predicted for each of the stations and then compared to the observed value. The results yielded and average relative error of 6.32% with a standard deviation of 3.28 %. The maximum relative error was 8.94 %. The absolute relative error was 6.32% with a standard deviation of 3.28 %. This verification is shown graphically in Figure 25

PAGE 43

36 R 2 = 0.998 Tau = 1 n = 7 0 20 40 60 80 100 120 140 160 180 0 50 100 150 200 Observed Q (cfs) Predicted Q (cfs) Figure 25 : Predicted Versus Observed Mean Annual Streamflow for Verification Data Set (1993 2003)

PAGE 44

37 CHAPTER FOUR D ISCUSSIONS AND CONCLUSIONS The resu lts show that streamflow sensitivity to rainfall varies in west central Florida One interesting finding, is the apparent phase lag occurring in the streamflow with precipitation Depending on the area, streamflow may vary directly with precipitation or they may be a two year lag for wetting conditions. Where the two year lag occurs, it is theorized that aquifer and/or storages may need to be filled for longer than a year before the full response of streamflow is seen. For the north and south zone analy sis, the results were significantly different for each zone. The north zone had a correlation of only 44% when precipitation was plotted against streamflow, while the south had a correlation of 71%. However, for the two year preceding average precipitati on, the north had a correlation of 69%, while the south had a correlation of 61%, which was less than the 71% correlation mentioned previously. This suggests that the north has a stronger phase lag than the south. Also, it could be concluded that since t he north two year preceding average precipitation had a steeper slope than the south, that the north is more sensitive to a rainfall deficit. Hydrogeologically, the south has a shallow depth to water table in certain areas Therefore it has less stora ge capacity in the vadose zone than the north. On the other hand, the north has a deeper depth to water table and concurrently more water storage

PAGE 45

38 capacity The north is also more karst, with lower confinement (leaky) and thus may also be responding to po tentiometric head changes occurring over two years. Both areas have about the same coverage of wetlands (25%), so this does not seem to be responsible for the difference. The eleven year period of record used for this investigation did seem suffici ent, b ut this is uncertain. The precipitation data were characteristically similar to that of the prior 10 year period. There does seem to be some spatial variability in the rainfall for the north and south zones. However, statistically, there is not a signif icant enough difference to be able to draw any definitive conclusions regarding this anomaly. In regards to the normalized streamflow fraction (NSF), the average precipitations used for the north and south zones had a difference of approximately 3.5 inc hes. As mentioned above, this difference does not seem to be statistically meaningful The equation that was developed in this study for the ungaged steamflow, does seem to be relatively predictive, even though it was developed with parameters derived wi th relatively modest correlations. The result was a very good (0.99) correlation to observed streamflow. This equation would allow one in the field to determine the ungaged stream flow for a ny given area in the study domain This equation also demonstrat es the time variability of the streamflow fraction and the differences between hydrogeologic settings. To further explore the validity of the model, a validation data set (same region) was analyzed and demonstrated the same degree of high correlation (0 .99) in results. This further supports the preliminary findings that the method may be viable for other areas and other times that were not included in the study area.

PAGE 46

39 It is unclear whether the method will work in other regions or other times with the sa me success. The methodology maybe applicable outside of the west central Florida area; however, further testing remains.

PAGE 47

40 REFERENCES Altunkaynak, A., Ozger, M., Sen, Z., 2005. Regional Streamflow Estimation by Standard Regional Dependence Function Approach Journal of Hydraulic Engineering, vol. 131, no. 11, p. 1001 1006. Census Scope, 2006. www.censusscope.org. Clayback, Kim, 2006. Investigation of Normalized Streamflow in West Central Florida and Extrapolation to Ung ages Coastal Fringe Tribut aries, Masters Thesis, Department of Civil and Environmental Engineering, University of South Florida, Tampa, Florida. Geurink, J.S., Nachabe, M., Ross, M.A., Tara, P., 2000. Development of Interfacial Boundary Conditions for the Southern District Ground Water Model of the Southwest Florida Water Management District, Center for Modeling Hydrologic and Aquatic Systems, Report # CMHAS.SWFWMD.00.03. Helsel, D.R., Hirsch, R.M., 2002. Statistical Methods in Water Resources, Techniques of Water Resource Invest igations of the U.S.G.S. http://water.usgs.gov/pubs/twri/twri4a3. Horn, D., 1988. Annual Flow Statistics for Ungaged Streams in Idaho, Journal of Irrigation and Drainage Engineering, vol. 114, no. 3, p. 463 475. Kokkonen, T.S., Jakeman, A.J., Young, P.C and Koivusalo, H.J., 2003. Predicting Daily Flows in Ungauged Catchments: Model Regionalization from Catchment Descriptors at the Coweeta Hydrologic Laboratory, North Carolina, Hydrological Processes vol. 17, no. 11, p. 2219 2238. Kroll, C., Luz, J., A llen, B., Vogel, R., 2004, Journal of Hydrologic Engineering, vol. 9, no. 2, p. 116 125. Lombard, Pamela J., 2004. August Median Streamflow on Ungaged Streams in Eastern Coastal Maine. NOAA 2005. available at: www.ncdc.noaa.gov Ormsby, T., Napoleon, E., Burke, R., Groessl, C., Feaster, L., 2004. Getting to Know ArcGIS Desktop. ESRI Press, Redlands, California.

PAGE 48

41 Rohlf, J. F., Sokal, R. R., 1969. Statistical Tables, Second ed, Freeman and Company, New York. Ross, Mark, 2005. Interview with Michael Scott, September 9, 2005. Ruskauff, G., Aly, A., Ewing, J., Jobes, T., Donigan, A., Tara, P., Trout, K., Ross, M., 2003. The Integrated Northern Tampa Bay Hydrologic Model (INTB) Volume 3, Tampa Bay Water, Tampa FL. SWFWMD 2006. available at: www.swfwmd.state.fl.us Tihansky, A.B., Knochemus, L.A., 2001. Karst Features and Hydrogeology in West Central Florida A Field Perspective, U.S. Geological Survey Report, Tampa, FL. United States Geological Survey, 2004. StreamStats available at: www.usgs.gov

PAGE 49

42 APPENDICES

PAGE 50

43 A ppendix A : Streamflow Estimates for West Central Florida Table 4 : Streamflow E stimates for W e st C entral Florida Station Area (sq m ile) NSF Predicted Q (cfs) Observed Q (cfs) Error Relative Error ABS Error ABS Relative Error 2256500 308.59 0.2302 286.6 285.55 1.01 0.35% 1.01 0.35% 2267000 46.05 0.1987 33.8 36.78 2.97 8.07% 2.97 8.07% 2268390 53.06 0.1794 35.2 38.26 3.09 8.07% 3 .09 8.07% 2269520 118.32 0.1290 61.3 61.36 0.09 0.15% 0.09 0.15% 2270000 38.96 0.1286 20.1 20.14 0.03 0.15% 0.03 0.15% 2270500 231.25 0.3300 295.6 306.69 11.05 3.60% 11.05 3.60% 2271500 113.22 0.1434 62.9 65.25 2.35 3.60% 2.35 3.60% 2293987 17 0.74 0.1434 96.6 98.44 1.86 1.89% 1.86 1.89% 2294217 59.54 0.1689 40.9 40.43 0.46 1.14% 0.46 1.14% 2294491 85.82 0.2545 85.7 87.78 2.06 2.35% 2.06 2.35% 2294650 88.57 0.3746 130.2 133.37 3.13 2.35% 3.13 2.35% 2294898 74.91 0.1558 45.7 46.91 1.21 2.58% 1.21 2.58% 2295013 46.29 0.1601 29.0 29.79 0.77 2.58% 0.77 2.58% 2295420 125.2 0.2758 135.2 138.77 3.58 2.58% 3.58 2.58% 2295637 188.02 0.2387 175.7 180.37 4.65 2.58% 4.65 2.58% 2296500 326.47 0.2266 289.6 297.29 7.67 2.58% 7.67 2.58% 2296750 207.39 0.2268 220.0 189.07 30.88 16.33% 30.88 16.33% 2297100 120.94 0.3204 181.2 155.76 25.44 16.33% 25.44 16.33% 2297155 40.93 0.1964 37.1 32.31 4.76 14.74% 4.76 14.74% 2297310 176.4 0.2828 233.3 200.50 32.75 16.33% 32.75 16.33% 2298123 223.02 0.2809 249.0 251.83 2.86 1.13% 2.86 1.13% 2298202 145.83 0.2860 165.8 167.66 1.90 1.13% 1.90 1.13% 2298608 124.06 0.3958 229.6 197.39 32.24 16.33% 32.24 16.33% 2298830 101.47 0.2904 121.1 118.45 2.63 2.22% 2.63 2.22% 2299410 35.83 0.3275 54.9 47.1 7 7.71 16.34% 7.71 16.34% 2299450 50 0.5481 128.1 110.15 18.00 16.34% 18.00 16.34% 2299861 6 0.2792 7.5 6.73 0.79 11.76% 0.79 11.76% 2299950 66.54 0.3501 103.8 93.63 10.21 10.90% 10.21 10.90% 2300018 60.36 0.2980 79.3 72.30 6.99 9.67% 6.99 9.67% 23000 32 25.21 0.3776 42.8 38.26 4.49 11.73% 4.49 11.73% 2300042 33.29 0.3943 58.9 52.75 6.19 11.73% 6.19 11.73% 2300100 30.9 0.3017 41.6 37.48 4.09 10.91% 4.09 10.91% 2300500 120.84 0.3991 212.6 193.86 18.73 9.66% 18.73 9.66% 2300700 28.55 0.4342 54.6 49.83 4.82 9.67% 4.82 9.67% 2301000 136.03 0.2717 146.6 148.54 1.93 1.30% 1.93 1.30% 2301300 112.25 0.2361 111.3 106.55 4.75 4.46% 4.75 4.46% 2301500 91.13 0.1746 66.8 63.94 2.85 4.46% 2.85 4.46%

PAGE 51

44 Appendix A: (Continued) Table 4 : (C ontinued ) Station Area (sq mile) NSF Predicted Q (cfs) Observ ed Q (cfs) Error Relative Error ABS Error ABS Relative Error 2301738 2.9 0.2366 2.4 2.76 0.31 11.18% 0.31 11.18% 2301740 6.09 0.1604 3.5 3.93 0.44 11.24% 0.44 11.24% 2301745 2 0.3443 2.5 2.77 0.31 11.12% 0.31 11.12% 2301750 14.21 0.1772 9.0 10.12 1.14 11.24% 1.14 11.24% 2301900 9.28 0.3375 12.7 11.75 0.91 7.75% 0.91 7.75% 2301990 76.47 0.2561 77.6 73.45 4.12 5.62% 4.12 5.62% 2302500 98.61 0.2413 94.2 89.24 5.01 5.62% 5.01 5.62% 2303000 42.98 0.3894 66.3 62.77 3.53 5.62% 3.53 5.62% 2303205 2 1.63 0.2426 20.8 19.68 1.11 5.62% 1.11 5.62% 2303350 17.25 0.3639 24.9 23.54 1.32 5.63% 1.32 5.63% 2306000 16.24 0.5185 30.0 31.58 1.54 4.86% 1.54 4.86% 2306647 14.38 0.4271 21.9 23.04 1.12 4.86% 1.12 4.86% 2307000 46.64 0.2272 37.8 39.75 1.94 4. 88% 1.94 4.88% 2307200 5.22 0.1797 3.3 3.52 0.17 4.86% 0.17 4.86% 2307359 27.83 0.1123 11.2 11.73 0.57 4.89% 0.57 4.89% 2309848 13.18 0.1062 5.0 5.25 0.26 4.89% 0.26 4.89% 2310000 56.42 0.2332 46.9 49.35 2.41 4.88% 2.41 4.88% 2310147 21.8 0.06 82 5.3 5.58 0.27 4.91% 0.27 4.91% 2310280 148.84 0.0092 4.9 5.16 0.25 4.93% 0.25 4.93% 2310300 32.56 0.1325 15.8 16.18 0.36 2.25% 0.36 2.25% 2310525 10.44 3.9279 150.4 153.80 3.41 2.21% 3.41 2.21% Average: 80.98 0.3224 86.5 8 3.56 2.93 1.85% 5.20 6.68% Max: 326.47 3.9279 295.6 306.69 32.75 16.34% 32.75 16.34% Min: 2.00 0.0092 2.4 2.76 11.05 11.24% 0.03 0.15% Std Dev: 74.99 0.4937 82.7 80.30 9.00 8.11% 7.90 4.89% Median: 54.74 0.2486 56.9 51.29 0.21 1.13% 2.52 4.90%

PAGE 52

45 Appendix A: (Continued) Table 5 : Verification of Streamflow E stimates for W est C entral Florida S tation Area (sq mile) NSF Predicted Q (cfs) Observed Q (cfs) Error Relative Error ABS Error ABS Relative Error 2310947 352.83 0.1180 170.1 156.14 13.96 8.94% 13.96 8.94% 2311500 66.83 0.1917 52.4 48.06 4.30 8.94% 4.30 8.94% 2312000 152.96 0.1960 122 .5 112.45 10.06 8.94% 10.06 8.94% 2312180 87.97 0.1533 55.1 50.58 4.52 8.94% 4.52 8.94% 2312200 50.61 0.4390 85.6 83.32 2.27 2.72% 2.27 2.72% 2312500 116.92 0.1617 72.8 70.90 1.92 2.71% 1.92 2.71% 2236350 41.83 0.1111 18.0 17.43 0.52 3.01% 0.52 3.01% Average: 124.28 0.20 82.35 76.98 5.36 6.32% 5.36 6.32% Max: 352.83 0.44 170.10 156.14 13.96 8.94% 13.96 8.94% Min: 41.83 0.11 17.96 17.43 0.52 2.71% 0.52 2.71% Std Dev: 107.98 0.11 50.34 45.97 4.88 3.28% 4.88 3.28% Median: 87.97 0.16 72.83 70.90 4.30 8.94% 4.30 8.94%


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 Ka
controlfield tag 001 001916253
003 fts
005 20071109130214.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 071109s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001793
040
FHM
c FHM
035
(OCoLC)181017485
049
FHMM
090
TA145 (ONLINE)
1 100
Scott, Michael H.
0 245
Precipitation variability of streamflow fraction in West Central Florida
h [electronic resource] /
by Michael H. Scott.
260
[Tampa, Fla] :
b University of South Florida,
2006.
3 520
ABSTRACT: There is a strong interest to develop a method to estimate mean annual ungaged streamflow with varying precipitation. A method was developed utilizing GIS and other statistical analysis to estimate ungaged mean annual streamflow. This method utilizes a normalized streamflow fraction (NSF) method previously developed which relies on drainage basin area, coupled with mean annual local precipitation, to estimate the ungaged streamflow variability. This method has been applied to west central Florida.The test of the method yielded an R squared value of 0.9894, proceeded by a verification that yielded an R squared value of 0.998. This method is believed to be generally applicable to other areas and the particular results should be useful in and around west central Florida and perhaps, other coastal plain environments.
502
Thesis (M.S.C.E.)--University of South Florida, 2006.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 45 pages.
590
Adviser: Mark Ross, Ph.D.
653
Coastal.
GIS
Ungaged.
Tidal effects.
Estuaries.
690
Dissertations, Academic
z USF
x Civil Engineering
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1793