v Bacterial Loadings Watershed Model in Copano Bay Center for Research in Water Resources On-line report 06-06 by Carrie Jo Gibson Master of Science in Engineering The University of Texas at Austin May 2006 This paper is funded in part by a grant/cooperative agreement from the National Oceanic and atmospheric administration. The views expressed herein are those of the author and do not necessarily reflect the views of NOAA or any of its subagencies. vi Abstract Bacterial Loadings Watershed Model in Copano Bay Carrie Jo Gibson, M.S.E. The University of Texas at Austin, 2006 Supervisor: David R. Maidment Copano Bay currently exceeds fecal coliform Texas Surface Water Quality Standards for oyster water use. Aransas and Mission River Tidals currently exceed enterococci water quality standards for contact recreation use. The fecal coliform Copano Bay Bacterial Loadings Model will be used to support the TCEQ Total Maximum Daily Load (TMDL) program to develop the TMDLs for the three impaired water segments. The objectives of this research are to identify the major bacterial sources in the Copano Bay watershed, to calculate the total bacterial loadings (i.e., the TMDLs) from these sources, and to estimate the load reductions needed to bring each of the impaired segments into compliance with water quality standards. vii The potential bacterial sources that were considered in the model were wastewater treatment plants (WWTPs), waterbirds, livestock, failing septic systems, and other non- point sources that originate from different types of land uses (e.g., urban, forest, etc.). This thesis presents an analysis of the existing bacterial monitoring dataset for fecal coliform, including spatial and statistical analysis of the bacterial monitoring data, an estimation of fecal coliform loadings (the input into the models), including non-point and point source calculations, and a description of bacterial transport of fecal coliform from the sources in the watersheds, rivers, and Copano Bay using the model, including explanations for how the model parameters were determined. The main assumptions used in the model were that the fecal coliform bacteria decay (first-order reaction rate) in watersheds and along streams and channels, and Copano Bay is divided up into four Continuous Flow, Stirred Tank Reactors (CFSTRs). The results of the research include the modeled median fecal coliform concentrations throughout the watershed, the impact of different bacterial sources on each of the water segments in Copano Bay watershed, and the load reductions needed (and from what sources) to meet fecal coliform water quality standards. Cattle were determined (based on model results) to be the largest fecal coliform contributor of fecal coliform in Copano Bay. viii Table of Contents CHAPTER 1: INTRODUCTION 1.1 Background........................................................................................................1 1.1.1 Purpose and Objectives..........................................................................2 1.1.2 Study Area .............................................................................................3 1.2 Texas Surface Water Quality Standards ............................................................4 1.3 Total Maximum Daily Load Program................................................................6 1.4 Potential Sources of Contamination...................................................................7 1.4.1 Point Sources .........................................................................................7 1.4.2 Non-Point Sources .................................................................................7 1.5 Outline of Thesis................................................................................................8 CHAPTER 2: LITERATURE REVIEW 2.1 Bacterial Indicators ..........................................................................................10 2.1.1 Correlation between Fecal Coliform and Enterococci.........................11 2.1.2 Correlation between Fecal Coliform and E. coli .................................17 2.2 Decay Rate of Fecal Coliform Bacteria...........................................................24 2.2.1 Factors Affecting Decay Rate of Fecal Coliform ................................24 2.2.1.1 pH.............................................................................................25 2.2.1.2 Solar Radiation Intensity..........................................................26 2.2.1.3 Temperature .............................................................................26 2.2.1.4 Salinity .....................................................................................27 2.2.2 Summary..............................................................................................28 CHAPTER 3: DATA DESCRIPTION 3.1 Bacterial Monitoring Data ...............................................................................29 3.2 Datasets Required for Loading Estimation......................................................31 3.2.1 Datasets Required for Non-Point Source Loads ..................................31 3.2.1.1 Runoff Dataset .........................................................................31 ix 3.2.1.1.1 Precipitation Data......................................................31 3.2.1.1.2 Land Use / Land Cover Data ....................................32 3.2.1.2 Event Mean Concentration (EMC) Dataset .............................32 3.2.1.3 Livestock Data .........................................................................34 3.2.1.4 Septic System Data ..................................................................34 3.2.2 Datasets Required for Point Source Loads ..........................................36 3.2.2.1 Bird Data..................................................................................36 3.2.2.2 Industrial/Municipal Wastewater Outfalls Data ......................36 3.2.2.3 Concentrated Animal Feedlot Operations (CAFOs) Data .......37 3.2.3 Water Rights Analysis Package (WRAP) Hydro Watershed Delineation Dataset..............................................................................37 3.2.3.1 DEM and Terrain Preprocessing..............................................37 3.2.3.2 National Hydrography Dataset (NHD) ....................................38 3.2.3.3 Critical Points (USGS Gauge Stations, Bacterial Monitoring Stations, Water Segment Endpoints) .......................................38 3.2.4 Datasets Required for Schematic Processor Model.............................38 3.2.4.1 USGS Stream Gauge Data.......................................................38 3.2.4.2 Bathymetry Data ......................................................................39 3.3 Map Projection and Coordinate Systems.........................................................41 CHAPTER 4: ANALYSIS OF MONITORING DATASET 4.1 Spatial Distribution of Fecal Coliform ............................................................42 4.1.1 Methodology........................................................................................42 4.1.2 Procedure of Application .....................................................................43 4.1.3 Result ...................................................................................................43 4.2 Statistical Distribution of Fecal Coliform........................................................51 4.2.1 Methodology........................................................................................51 4.2.2 Procedure of Application .....................................................................51 4.2.3 Result ...................................................................................................52 4.2.3.1 Aransas River Above Tidal......................................................52 x 4.2.3.2 Aransas River Tidal .................................................................57 4.2.3.3 Mission River Above Tidal......................................................60 4.2.3.4 Mission River Tidal .................................................................63 4.2.3.5 Copano Bay..............................................................................65 CHAPTER 5: ESTIMATION OF LOADINGS 5.1 Estimation of Non-Point Bacterial Loadings from Watersheds ......................73 5.1.1 Methodology........................................................................................73 5.1.2 Procedure of Application .....................................................................74 5.1.2.1 Watershed Delineation.............................................................74 5.1.2.2 Precipitation Data Preparation .................................................76 5.1.2.3 Land Use Land Cover Data Preparation ..................................76 5.1.2.4 Rainfall-Runoff Relationships for Different Land Uses..........78 5.1.2.5 Developing Runoff Grid ..........................................................78 5.1.2.6 Estimation of Flow from each Watershed ...............................81 5.1.2.7 Developing EMC Grid.............................................................82 5.1.2.8 Estimation of Non-Point Bacterial Loading ............................83 5.1.3 Result ...................................................................................................84 5.2 Estimation of Livestock Loading.....................................................................85 5.2.1 Methodology........................................................................................85 5.2.2 Procedure of Application .....................................................................86 5.2.2.1 Finding Livestock per County .................................................86 5.2.2.2 Calculating Density of Livestock per County..........................88 5.2.2.3 Calculating Livestock Count per Watershed ...........................90 5.2.2.4 Calculating Livestock Bacterial Loading (CFU/year) per Watershed ................................................................................91 5.2.3 Result ...................................................................................................92 5.3 Estimation of Avian Loading...........................................................................95 5.3.1 Methodology........................................................................................95 5.3.2 Procedure of Application .....................................................................96 5.3.2.1 Determining Average Count of Waterbird Species .................96 xi 5.3.2.2 Determining Excretion (g/bird) from Waterbirds....................99 5.3.2.3 Estimation of Loadings (CFU/bird).......................................100 5.3.3 Result .................................................................................................101 5.4 Estimation of Wastewater Treatment Plant (WWTP) Loadings....................103 5.4.1 Methodology......................................................................................103 5.4.2 Procedure of Application ...................................................................105 5.4.2.1 Determining the Fecal Coliform Concentration (CFU/100mL) from WWTPs.................................................105 5.4.2.2 Determination of Average Flow (m 3 /yr) from WWTPs ........107 5.4.2.3 Calculating Annual Bacterial Loading (CFU/year) from WWTPs..................................................................................108 5.4.2.4 Calculating Residence Time to Mainstream Based on Flow Length...........................................................................109 5.4.3 Result .................................................................................................112 5.5 Estimation of Loadings from Septic Systems................................................113 5.5.1 Methodology......................................................................................113 5.5.2 Procedure of Application ...................................................................115 5.5.2.1 Finding Septic Systems, Complaints, Population and Housing Units per County in 2004 ........................................115 5.5.2.2 Calculating Density of Septic Systems, Complaints Investigated, Housing Units, and Population per County in 2004........................................................................................118 5.5.2.3 Calculating Total Septic Systems, Complaints Investigated, Housing Units, and Population per Watershed......................121 5.5.2.4 Calculating Septic Systems and Complaints Investigated in Soil Groups A, B, C, and D per Watershed .......................124 5.5.2.5 Calculating Septic System Bacterial Loading (CFU/year) per Watershed ........................................................................127 5.5.3 Result .................................................................................................130 5.6 Estimation of Total Loading ..........................................................................131 xii CHAPTER 6: MODELING OF BACTERIAL TRANSPORT - SCHEMATIC PROCESSOR 6.1 Background....................................................................................................134 6.2 Methodology..................................................................................................135 6.3 Procedure of Application ...............................................................................136 6.3.1 Creation of Schematic Network.........................................................136 6.3.1.1 Segmentation of Copano Bay ................................................137 6.3.2 Use of Dynamic Linked Libraries (DLLs) ........................................139 6.3.2.1 First-order Decay: clsDecay.dll .............................................139 6.3.2.2 Continuous Flow Stirred Tank Reactor: clsCFSTR.dll .........140 6.3.3 Determination of Model Parameters..................................................141 6.3.3.1 Decay Coefficient ..................................................................141 6.3.3.2 Residence Time......................................................................147 6.3.3.2.1 3d Channel Morphology (RCMM Toolbar) ...........147 6.3.3.2.2 3d Model of Aransas River: Residence Time Determination .........................................................148 6.3.3.2.3 3d Model of Mission River: Residence Time Determination .........................................................168 6.3.3.2.4 3d Model of Copano Creek: Residence Time Determination .........................................................194 6.3.3.2.5 Determinination of Residence Times of Remaining SchemaLinks (Calibration)...................202 6.3.3.3 Volume of Copano Bay Segments.........................................210 6.3.3.4 Flow of Copano Bay Segments..............................................212 6.3.4 Implementation of Schematic Processor............................................213 6.4 Results............................................................................................................219 CHAPTER 7: MODELING OF BACTERIAL TRANSPORT - MONTE CARLO SIMULATION 7.1 Background....................................................................................................229 7.2 Methodology..................................................................................................233 7.3 Procedure of Application ...............................................................................234 xiii 7.3.1. Determination of Parameters ............................................................234 7.3.1.1 Bacterial Loading...................................................................234 7.3.1.2 Decay Coefficient ..................................................................239 7.3.1.3 Residence Time......................................................................240 7.3.1.4 Other Parameters....................................................................241 7.3.2 Calibration of Model..........................................................................242 7.3.2.1 Aransas River Above Tidal....................................................243 7.3.2.2 Aransas River Tidal ...............................................................246 7.3.2.3 Mission River Above Tidal....................................................249 7.3.2.4 Mission River Tidal ...............................................................251 7.3.2.5 Copano Bay............................................................................253 7.3.3 Calculation of Load Allocations........................................................262 7.3.3.1 Copano Bay............................................................................263 7.4 Results ............................................................................................................270 CHAPTER 8: RESULTS 8.1 Estimation of Loadings..................................................................................272 8.1.1 Schematic Processor ..........................................................................272 8.1.2 Monte Carlo Simulations ...................................................................278 8.2 Estimation of Load Allocation.......................................................................282 8.2.1 Monte Carlo Simulations ...................................................................282 8.2.2 Applied to Schematic Processor Model.............................................283 8.2.2.1 Aransas River Above Tidal....................................................284 8.2.2.2 Aransas River Tidal ...............................................................284 8.2.2.3 Mission River Above Tidal....................................................286 8.2.2.4 Mission River Tidal ...............................................................286 8.2.2.5 Copano Bay............................................................................288 CHAPTER 9: CONCLUSIONS AND RECOMMENDATIONS 9.1 Conclusions ................................................................................................291 9.2 Recommendations..........................................................................................294 xiv Appendix 4.1: Bacterial Monitoring Data (1999-2005) of Copano Bay Segment 1.....................................................................................298 Appendix 4.2: Bacterial Monitoring Data (1999-2005) of Copano Bay Segment 2.....................................................................................301 Appendix 4.3: Bacterial Monitoring Data (1999-2005) of Copano Bay Segment 3.....................................................................................305 Appendix 4.4: Bacterial Monitoring Data (1999-2005) of Copano Bay Segment 4.....................................................................................306 Appendix 5.1: Terrain Preprocessing ..................................................................312 Appendix 5.2: WRAP Hydro Process..................................................................313 Appendix 5.3: Precipitation Rasters for Land Use Classifications......................315 Appendix 5.4: Livestock Loading Calculations and Results...............................316 Appendix 5.5: Mean Flow Length for Watersheds..............................................326 Appendix 5.6: Mean Flow Length from WWTPs to Mainstreams......................328 Appendix 5.7: Septic System Loading Calculations and Results........................330 Appendix 5.8: Determination of Soil Group Areas within each Watershed and Land Use Classifications 21 and 22......................................347 Appendix 6.1: Schematic Network......................................................................349 Appendix 6.2: Travel Time Calculations.............................................................352 Appendix 6.3: Process Schematic........................................................................355 Appendix 7.1: Monte Carlo Simulation Model Worksheets ...............................359 Appendix 7.2: Load Reduction Scenario #1 Results ...........................................371 Appendix 8.1: Load Allocations for Scenario #1 ................................................397 Works Cited ......................................................................................................408 Vita ......................................................................................................413 xv List of Tables Table 2.1: Bacterial Indicators for Water Segments ..........................................11 Table 2.2: Fecal Coliform and Enterococcus Measurements at Station 12948: Aransas River Tidals.........................................................................12 Table 2.3: Fecal Coliform and Enterococcus Measurements at Station 12943: Mission River Tidal ..........................................................................15 Table 2.4: Fecal Coliform and E. coli Measurements at Station 12952: Aransas River Above Tidal.............................................................................18 Table 2.5: Fecal Coliform and E. coli Measurements at Station 12944: Mission River Above Tidal.............................................................................21 Table 2.6: Average Measured pH in Copano Bay Segments.............................25 Table 2.7: Average Measured Temperature in Copano Bay Segments .............27 Table 2.8: Average Measured Salinity in Copano Bay Segments .....................28 Table 3.1: Fecal Coliform EMC Values Based on Land Use Classifications (Zoun, 2003) .....................................................................................33 Table 3.2: Description of Source Codes for EMC Values.................................34 Table 3.3: Parameters for NAD_1983_Texas_Centric_Mapping_System_ Albers................................................................................................41 Table 4.1: Fecal Coliform Water Quality Standards for Water Segments.........42 Table 4.2: Water Segment Locations of Bacterial Monitoring Stations ............49 Table 4.3: Summary of TCEQ Fecal Coliform Monitoring Data for Water Segments...........................................................................................50 xvi Table 4.4: Summary of TCEQ Fecal Coliform Monitoring Data for Copano Bay Water Segments.........................................................................50 Table 4.5: Bacterial Monitoring Data for Station 17592 (1999-2004) ..............53 Table 4.6: Bacterial Monitoring Data for Station 12952 (1999-2004) ..............55 Table 4.7: Bacterial Monitoring Data for Station 12948 (1999-2004) ..............57 Table 4.8: Bacterial Monitoring Data for Station 12944 (1999-2004) ..............60 Table 4.9: Bacterial Monitoring Data for Station 12943 (1999-2004) ..............63 Table 4.10: Statistics of Bacterial Monitoring Data for Stations in Segment 1 (1999-2005).......................................................................................65 Table 4.11: Statistics of Bacterial Monitoring Data for Stations in Segment 2 (1999-2005).......................................................................................67 Table 4.12: Statistics of Bacterial Monitoring Data for Stations in Segment 3 (1999-2005).......................................................................................69 Table 4.13: Statistics of Bacterial Monitoring Data for Stations in Segment 4 (1999-2005).......................................................................................71 Table 5.1: Reclassified Land Use Categories ....................................................77 Table 5.2: Livestock Count per County .............................................................87 Table 5.3: Animal Density per County (Acres per Animal) ..............................89 Table 5.4: Annual Fecal Coliform Production from Livestock Animals...........92 Table 5.5: Average Waterbird Count (1973-2003)............................................97 Table 5.6: Colony Codes and Watersheds/Segments to which Loads are Applied..............................................................................................98 Table 5.7: Number of Waterbird Species Applied to each Segment/Watershed ..........................................................................99 Table 5.8: Estimated Daily Fecal Mass (g/bird) ..............................................100 xvii Table 5.9: Annual Fecal Coliform Loading per Bird.......................................101 Table 5.10: Annual Fecal Coliform Avian Loadings.........................................102 Table 5.11: Fecal Coliform Concentrations of WWTPs Applied to Model ......106 Table 5.12: Flow Rates of WWTPs ...................................................................108 Table 5.13: Annual Bacterial Loadings from WWTPs......................................109 Table 5.14: WWTP Bacterial Loading Applied to Model .................................111 Table 5.15: Septic System Data from 1990 U.S. Census and TCEQ Applications (1990-2004).....................................................................................115 Table 5.16: Annual Septic System Complaint Percentage by County...............116 Table 5.17: Housing Unit Data by County.........................................................116 Table 5.18: Septic Systems in Use, Complaints Investigated, Housing Units, and Population in 2004 ..........................................................................117 Table 5.19: Septic Systems, Complaints Investigated, Housing Units, and Population per County (Count per km 2 ) .........................................120 Table 5.20: Total Septic Systems, Complaints Investigated, Population, Housing Units per Watershed, and People/Housing Unit per Watershed.....123 Table 5.21: Number of Septic Systems and Complaints per Watershed in Soil Groups A, B, C, and D....................................................................126 Table 5.22: Annual Bacterial Loadings (TCFU/yr) from Major Bacterial Sources in Entire Copano Bay Watershed......................................133 Table 6.1: Dissolving of Copano Bay Segments (New Labeling)...................139 Table 6.2: Available Data for Segregated Model on Aransas River Segment 2........................................................................................143 Table 6.3: Calculation of Decay Coefficient for Segregated Model................146 xviii Table 6.4: Upstream Cross-Section Data Comparison (Aransas River; USGS 08189700) .......................................................................................150 Table 6.5: Downstream Cross-Section Data (Aransas River)..........................151 Table 6.6: Residence Times of Aransas River Segments for Schematic Processor Model..............................................................................168 Table 6.7: Upstream Cross-Section Data Comparison (Medio Creek; USGS 08189300) .......................................................................................172 Table 6.8: Downstream Cross-Section Data (Medio Creek) ...........................173 Table 6.9: Upstream Cross-Section Data Comparison (Mission River; USGS 08189500) .......................................................................................175 Table 6.10: Downstream Cross-Section Data (Mission River)..........................175 Table 6.11: Residence Times of Medio Creek and Mission River Segments for Schematic Processor Model............................................................194 Table 6.12: Upstream Cross-Section Data Comparison (Copano Creek; USGS 08189200) .......................................................................................197 Table 6.13: Downstream Cross-Section Data (Copano Creek) .........................197 Table 6.14: Residence Time of Copano Creek for Schematic Processor Model ..............................................................................................202 Table 6.15: Area, Depth, and Volume of Copano Bay Segments......................212 Table 6.16: Cumulative Annual Runoff to Copano Bay Segments ...................213 Table 6.17: SchemaNode Attribute Table (Calibrated to Median Fecal Coliform Concentrations)...............................................................................214 Table 6.18: SchemaLink Attribute Table (Calibrated to Median Fecal Coliform Concentrations)...............................................................................216 xix Table 6.19: Modeled versus Existing Fecal Coliform Concentrations: Schematic Processor Model..............................................................................221 Table 7.1: SchemaNode Adjusted Parameters for Calibration of Station 17592...............................................................................................243 Table 7.2: SchemaLink Adjusted Parameters for Calibration of Station 17592...............................................................................................243 Table 7.3: SchemaNode Adjusted Parameters for Calibration of Station 12952...............................................................................................245 Table 7.4: SchemaLink Adjusted Parameters for Calibration of Station 12952...............................................................................................245 Table 7.5: SchemaNode Adjusted Parameters for Calibration of Station 12948...............................................................................................247 Table 7.6: SchemaLink Adjusted Parameters for Calibration of Station 12948...............................................................................................247 Table 7.7: SchemaNode Adjusted Parameters for Calibration of Station 12944...............................................................................................249 Table 7.8: SchemaLink Adjusted Parameters for Calibration of Station 12944...............................................................................................249 Table 7.9: SchemaNode Adjusted Paramters for Calibration of Station 12943...............................................................................................251 Table 7.10: SchemaLink Adjusted Parameters for Calibration of Station 12943...............................................................................................251 Table 7.11: SchemaNode Adjusted Paramters for Calibration of Segment 1....253 Table 7.12: SchemaLink Adjusted Parameters for Calibration of Segment 1 ...253 Table 7.13: SchemaNode Adjusted Parameters for Calibration of Segment 2..255 xx Table 7.14: SchemaLink Adjusted Parameters for Calibration of Segment 2 ...256 Table 7.15: SchemaNode Adjusted Parameters for Calibration of Segment 3..258 Table 7.16: SchemaLink Adjusted Parameters for Calibration of Segment 3 ...258 Table 7.17: SchemaNode Adjusted Parameters for Calibration of Segment 4..260 Table 7.18: SchemaLink Adjusted Parameters for Calibration of Segment 4 ...260 Table 7.19: Modeled Results at SchemaNode 154 with Various WWTP/Livestock Load Reductions ...............................................265 Table 7.20: Modeled Results at SchemaNode 153 with Various Livestock Load Reductions .............................................................................268 Table 8.1: Schematic Processor Bacterial Loadings to Aransas River Above Tidal (Segment 2004)..........................................................273 Table 8.2: Schematic Processor Bacterial Loadings (from Major Sources) to Aransas River Above Tidal (Segment 2004)..............................273 Table 8.3: Schematic Processor Bacterial Loadings to Aransas River Tidal (Segment 2003) .....................................................................274 Table 8.4: Schematic Processor Bacterial Loadings (from Major Sources) to Aransas River Tidal (Segment 2003)..........................................274 Table 8.5: Schematic Processor Bacterial Loadings to Mission River Above Tidal (Segment 2002) .....................................................................275 Table 8.6: Schematic Processor Bacterial Loadings (from Major Sources) to Mission River Above Tidal (Segment 2002)..............................275 Table 8.7: Schematic Processor Bacterial Loadings to Mission River Tidal (Segment 2001) .....................................................................276 Table 8.8: Schematic Processor Bacterial Loadings (from Major Sources) to Mission River Tidal (Segment 2001)..........................................276 xxi Table 8.9: Schematic Processor Bacterial Loadings to Copano Bay (Segment 2472) ...............................................................................................277 Table 8.10: Schematic Processor Bacterial Loadings from Major Sources to Copano Bay (Segment 2472) in CFU/year .....................................277 Table 8.11: Monte Carlo Simulation Model Loadings to Aransas River Above Tidal (Segment 2004)..........................................................279 Table 8.12: Monte Carlo Simulation Model Loadings to Aransas River Tidal (Segment 2003) .....................................................................279 Table 8.13: Monte Carlo Simulation Model Loadings to Mission River Above Tidal (Segment 2002)..........................................................280 Table 8.14: Monte Carlo Simulation Model Loadings to Mission River Tidal (Segment 2001) .....................................................................280 Table 8.15: Monte Carlo Simulation Model Loadings to Copano Bay (Segment 2472)...............................................................................281 Table 8.16: Load Reduction Scenario #2 at Downstream Node of Aransas River Tidal ......................................................................................286 Table 8.17: Load Reduction Scenario #2 at Downstream Node of Mission River Tidal ......................................................................................288 Table 8.18: Load Reduction Scenario #2 at Copano Bay Aransas River Outlet, Segment 2............................................................................289 Table 8.19: Load Reduction Scenario #2 at Copano Bay Mission River Outlet, Segment 3............................................................................290 Table 8.20: Load Reduction Scenario #2 at Copano Bay ..................................290 Table 5A.1: Livestock Calculations and Results ................................................316 xxii Table 5A.2: Annual Livestock Bacterial Loading per Watershed......................325 Table 5A.3: Septic System Loading Calculations and Results...........................330 Table 5A.4: Annual Septic System Bacterial Loading per Watershed...............345 Table 7A.1: Modeled Results at SchemaNode 62 with Various WWTP Load Reductions .............................................................................373 Table 7A.2: Modeled Results at SchemaNode 68 with No Additional Load Reductions.......................................................................................375 Table 7A.3: Modeled Results at SchemaNode 75 with Various Additional WWTP Load Reductions ................................................................376 Table 7A.4: Modeled Results at SchemaNode 63 with No Additional Load Reductions.......................................................................................378 Table 7A.5: Modeled Results at SchemaNode 67 with Various Additional WWTP/Livestock/Non-point Load Reductions..............................380 Table 7A.6: Modeled Results at SchemaNode 73 with Various Additional WWTP/Livestock Load Reductions ...............................................384 Table 7A.7: Modeled Results at SchemaNode 74 with No Additional Load Reductions.......................................................................................386 Table 7A.8: Modeled Results at SchemaNode 65 with No Additional Load Reductions.......................................................................................388 Table 7A.9: Modeled Results at SchemaNode 70 with No Additional Load Reductions.......................................................................................389 Table 7A.10: Modeled Results at SchemaNode 66 with Various Additional Livestock/Non-point Load Reductions...........................................391 Table 7A.11: Modeled Results at SchemaNode 154 with No Additional Load Reductions (Aransas Tidal Load Reductions Included) ........393 xxiii Table 7A.12: Modeled Results at SchemaNode 153 with No Additional Load Reductions (Mission Tidal Load Reductions Included) .................395 Table 8A.1: Load Reduction Scenario #1 at Upstream Node of Aransas River Above Tidal...........................................................................399 Table 8A.2: Load Reduction Scenario #1 at Downstream Node of Aransas River Above Tidal...........................................................................399 Table 8A.3: Load Reduction Scenario #1 at Downstream Node of Aransas River Tidal ......................................................................................401 Table 8A.4: Load Reduction Scenario #1 at Upstream Node of Mission River Above Tidal...........................................................................402 Table 8A.5: Load Reduction Scenario #1 at Downstream Node of Mission River Above Tidal...........................................................................403 Table 8A.6: Load Reduction Scenario #1 at Downstream Node of Mission River Tidal ......................................................................................405 Table 8A.7: Load Reduction Scenario #1 at Copano Bay Aransas River Outlet, Segment 2............................................................................406 Table 8A.8: Load Reduction Scenario #1 at Copano Bay Mission River Outlet, Segment 3............................................................................407 Table 8A.9: Load Reduction Scenario #1 at Copano Bay ..................................407 xxiv List of Figures Figure 1.1: Impaired Water Segments in Copano Bay Watershed........................1 Figure 1.2: Major Highways and Counties in the Copano Bay Watershed...........3 Figure 2.1: Fecal Coliform and Enterococcus Concentrations at Station 12948: Aransas River Tidal ..........................................................................13 Figure 2.2: Relationship between Enterococcus and Fecal Coliform at Station 12948: Aransas River Tidal ..............................................................14 Figure 2.3: Fecal Coliform and Enterococcus Concentrations at Station 12943: Mission River Tidal ..........................................................................16 Figure 2.4: Relationship between Enterococcus and Fecal Coliform at Station 12943: Mission River Tidal ..............................................................17 Figure 2.5: Fecal Coliform and E. coli Concentrations at Station 12952: Aransas River Above Tidal.............................................................................19 Figure 2.6: Relationship between E. coli and Fecal Coliform at Station 12952: Aransas River Above Tidal...............................................................20 Figure 2.7: Fecal Coliform and E. coli Concentrations at Station 12944: Mission River Above Tidal.............................................................................22 Figure 2.8: Relationship between E. coli and Fecal Coliform Concentrations at Station 12944: Mission River Above Tidal ......................................23 Figure 3.1: Bathymetry Map of Copano Bay ......................................................40 Figure 4.1: Mean of Fecal Coliform Concentrations at TCEQ Bacterial Monitoring Stations (1999-2005) .....................................................43 Figure 4.2: Geometric Mean of Fecal Coliform Concentrations at TCEQ xxv Bacterial Monitoring Stations (1999-2005)......................................45 Figure 4.3: Median of Fecal Coliform Concentrations at TCEQ Bacterial Monitoring Stations (1999-2005) .....................................................46 Figure 4.4: Minimum of Fecal Coliform Concentrations at TCEQ Bacterial Monitoring Stations (1999-2005) .....................................................47 Figure 4.5: Maximum of Fecal Coliform Concentrations at Bacterial Monitoring Stations (1999-2005) .....................................................48 Figure 4.6: Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Station 17592........................................54 Figure 4.7: Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Station 12952........................................56 Figure 4.8: Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Station 12948........................................59 Figure 4.9: Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Station 12944........................................62 Figure 4.10: Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Station 12943........................................64 Figure 4.11: Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Segment 1 .............................................66 Figure 4.12: Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Segment 2 .............................................68 Figure 4.13: Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Segment 3 .............................................70 Figure 4.14 Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Segment 4 .............................................72 xxvi Figure 5.1: Watershed Delineation......................................................................75 Figure 5.2: Precipitation Data (inches/year)........................................................76 Figure 5.3: Calculation of Agricultural Runoff Grid...........................................79 Figure 5.4: Creation of Runoff Grid....................................................................80 Figure 5.5: Runoff Grid (m 3 /year).......................................................................81 Figure 5.6: Runoff per Watershed (m 3 /year).......................................................82 Figure 5.7: EMC Grid (CFU/m 3 ).........................................................................83 Figure 5.8: Creation of Bacterial Loading Grid (CFU/year)...............................83 Figure 5.9: Non-Point Bacterial Loading per Watershed (CFU/year) ................84 Figure 5.10: Determination of Area (acres) of Animals in San Patricio County ..90 Figure 5.11: Determination of Cattle Count in Watershed JunctionID 45422......91 Figure 5.12: Livestock Bacterial Loading (CFU/year) per Watershed .................93 Figure 5.13: Percent Distribution of Bacterial Loadings from Livestock Species at Watersheds....................................................................................94 Figure 5.14: Locations of Breeding Pairs and Applied Loads ..............................98 Figure 5.15: Avian Loadings on Copano Bay Water Segments and Watersheds......................................................................................102 Figure 5.16: WWTP Locations and Permit Numbers .........................................105 Figure 5.17: Annual WWTP Bacterial Loadings (CFU/year).............................112 Figure 5.18: Low and High Residential Land Use Areas and Septic System Locations around Bay .....................................................................119 Figure 5.19: Determination of Area (km 2 ) of Septic Systems and Residences in San Patricio County ........................................................................121 Figure 5.20: Determination of Septic System Count in Watershed JunctionID 45422...............................................................................................122 xxvii Figure 5.21: Hydrologic Soil Group Classifications ...........................................125 Figure 5.22: Determination of Number of Impacting Septic Systems in Watershed JunctionID 45415..........................................................129 Figure 5.23: Septic System Annual Bacterial Loading (CFU/year)....................130 Figure 5.24: Total Annual Watershed/Segment Bacterial Loadings (CFU/year) ......................................................................................131 Figure 5.25: Percent Distribution of Bacterial Loading Sources ........................132 Figure 6.1: Schematic Network and Parameters ...............................................137 Figure 6.2: Copano Bay Initial Water Segments...............................................138 Figure 6.3: Copano Bay Segmentation..............................................................139 Figure 6.4: Simulation of Decay .......................................................................140 Figure 6.5: CFSTRs...........................................................................................141 Figure 6.6: Segregated Portion of Model to Calculate Decay Coefficient Distribution .....................................................................................142 Figure 6.7: Assumptions and Derivation of k-distribution................................145 Figure 6.8: Mainstream Network for RCMM ...................................................148 Figure 6.9: Flow versus Width for USGS Station 08189700............................149 Figure 6.10: Width versus Depth (Square Cross-Section) for USGS Station 08189700.........................................................................................150 Figure 6.11: Summary of RCMM Toolbar Data Requirements (Aransas River) ...............................................................................152 Figure 6.12: Flow versus Residence Time for Aransas River Segment 1...........153 Figure 6.13: Flow versus Residence Time for Aransas River Segment 2...........154 Figure 6.14: Flow versus Residence Time for Aransas River Segment 3...........155 Figure 6.15: Flow versus Residence Time for Aransas River Segment 4...........156 xxviii Figure 6.16: Flow Cumulative Distribution Function for Aransas River Segment 1........................................................................................157 Figure 6.17: Cumulative Distribution Function, q/q 0.5 for Segment 1 ................158 Figure 6.18: Calculating Mean Flow to Aransas River Segment 2.....................160 Figure 6.19: Flow Cumulative Distribution Function for Aransas River Segment 2........................................................................................161 Figure 6.20: Flow Cumulative Distribution Function for Aransas River Segment 3........................................................................................161 Figure 6.21: Flow Cumulative Distribution Function for Aransas River Segment 4........................................................................................162 Figure 6.22: Residence Time Distribution for Aransas River Segment 1...........163 Figure 6.23: Residence Time Distribution for Aransas River Segment 2...........163 Figure 6.24: Residence Time Distribution for Aransas River Segment 3...........164 Figure 6.25: Residence Time Distribution for Aransas River Segment 4...........164 Figure 6.26: Probability Distribution of Residence Time for Aransas River Segment 1........................................................................................165 Figure 6.27: Probability Distribution of Residence Time for Aransas River Segment 2........................................................................................165 Figure 6.28: Probability Distribution of Residence Time for Aransas River Segment 3........................................................................................166 Figure 6.29: Probability Distribution of Residence Time for Aransas River Segment 4........................................................................................167 Figure 6.30: SchemaLinks of Corresponding Aransas River Segments .............168 Figure 6.31: Flow versus Width for USGS Station 08189300............................170 Figure 6.32: Flow versus Width for USGS Station 08189500............................170 xxix Figure 6.33: Width versus Depth (Square Cross-Section) for USGS Station 08189300.........................................................................................171 Figure 6.34: Width versus Depth (Square Cross-Section) for USGS Station 08189500.........................................................................................172 Figure 6.35: Summary of RCMM Toolbar Data Requirements (Medio Creek).................................................................................174 Figure 6.36: Summary of RCMM Toolbar Data Requirements (Mission River) ...............................................................................176 Figure 6.37: Flow versus Residence Time for Medio Creek Segment 1.............177 Figure 6.38: Flow versus Residence Time for Mission River Segment 2...........178 Figure 6.39: Flow versus Residence Time for Mission River Segment 3...........179 Figure 6.40: Flow versus Residence Time for Mission River Segment 4...........180 Figure 6.41: Flow versus Residence Time for Mission River Segment 5...........181 Figure 6.42: Flow Cumulative Distribution Function for Segment 1 (USGS Data; 1962-2004)................................................................182 Figure 6.43: Flow Cumulative Distribution Function for Segment 2 (USGS Data; 1939-2004)................................................................183 Figure 6.44: Cumulative Distribution Function, q/q 0.5 for Mission River Segment 2........................................................................................184 Figure 6.45: Flow Cumulative Distribution Function for Mission River Segment 3........................................................................................185 Figure 6.46: Flow Cumulative Distribution Function for Mission River Segment 4........................................................................................186 Figure 6.47: Flow Cumulative Distribution Function for Mission River Segment 5........................................................................................186 xxx Figure 6.48: Residence Time Distribution for Medio Creek Segment 1.............187 Figure 6.49: Residence Time Distribution for Mission River Segment 2...........188 Figure 6.50: Residence Time Distribution for Mission River Segment 3...........188 Figure 6.51: Residence Time Distribution for Mission River Segment 4...........189 Figure 6.52: Residence Time Distribution for Mission River Segment 5...........189 Figure 6.53: Probability Distribution of Residence Time for Medio Creek Segment 1........................................................................................190 Figure 6.54: Probability Distribution of Residence Time for Mission River Segment 2........................................................................................190 Figure 6.55: Probability Distribution of Residence Time for Mission River Segment 3........................................................................................191 Figure 6.56: Probability Distribution of Residence Time for Mission River Segment 4........................................................................................191 Figure 6.57: Probability Distribution of Residence Time for Mission River Segment 5........................................................................................192 Figure 6.58: SchemaLinks of Corresponding Medio Creek and Mission River Segments.........................................................................................193 Figure 6.59: Flow versus Width for USGS Station 08189200............................195 Figure 6.60: Width versus Depth (Square Cross-Section) for USGS Station 08189200.........................................................................................196 Figure 6.61: Summary of RCMM Toolbar Data Requirements (Copano Creek)...............................................................................198 Figure 6.62: Flow versus Residence Time for Copano Creek.............................199 Figure 6.63: Flow Cumulative Distribution Function for Copano Creek (USGS Data; 1970-2004)................................................................200 xxxi Figure 6.64: Residence Time Distribution for Copano Creek.............................201 Figure 6.65: Probability Distribution of Residence Time for Copano Creek......201 Figure 6.66: Bacterial Monitoring and USGS Stations along Aransas River .....204 Figure 6.67: Bacterial Monitoring and USGS Stations along Mission River .....205 Figure 6.68: Bacterial Monitoring and USGS Station along Copano Creek.......206 Figure 6.69: Bacterial Monitoring and USGS Stations along Copano Bay Segment 1........................................................................................207 Figure 6.70: Nodes/Links' Parameters that can be Varied for Each Station along Aransas River..................................................................................209 Figure 6.71: Residence Times (days) of SchemaLinks (k = 2 days -1 ) for Schematic Processor Model............................................................210 Figure 6.72: Bathymetry Map of Copano Bay ....................................................211 Figure 6.73: HydroIDs of SchemaNodes ............................................................215 Figure 6.74: HydroIDs of SchemaLinks .............................................................218 Figure 6.75: Fecal Coliform Concentrations (CFU/100mL) - Schematic Processor Results ............................................................................219 Figure 6.76: Bacterial Loadings (from Sources) to SchemaNodes SrcTypes 2 and 3................................................................................................223 Figure 6.77: Bacterial Loadings (from Sources) to Copano Bay (CFU/year).....225 Figure 6.78: Percent Distribution of Bacterial Loading Sources (Output)..........227 Figure 7.1: Monte Carlo Simulation Conceptual Diagram ...............................231 Figure 7.2: Bacterial Load Distribution Modeled as Lognormal Distribution .....................................................................................236 Figure 7.3: Lognormal Distribution with Median = 1 (Normalized Lognormal Distribution by μ B ).......................................................237 xxxii Figure 7.4: Normal Distribution with Mean = Median = 0 and Standard Deviation, σ ln(B/μB) ...........................................................................238 Figure 7.5: Beta Distribution.............................................................................240 Figure 7.6: Modeled versus Measured Fecal Coliform Concentrations at Station 17592...............................................................................................244 Figure 7.7: Modeled versus Measured Fecal Coliform Concentrations at Station 12952...............................................................................................246 Figure 7.8: Modeled versus Measured Fecal Coliform Concentrations at Station 12948...............................................................................................248 Figure 7.9: Modeled versus Measured Fecal Coliform Concentrations at Station 12944...............................................................................................250 Figure 7.10: Modeled versus Measured Fecal Coliform Concentrations at Station 12943...............................................................................................252 Figure 7.11: Modeled versus Measured Fecal Coliform Concentrations at Segment 1........................................................................................254 Figure 7.12: Modeled versus Measured Fecal Coliform Concentrations at Segment 2........................................................................................257 Figure 7.13: Modeled versus Measured Fecal Coliform Concentrations at Segment 3........................................................................................259 Figure 7.14: Modeled versus Measured Fecal Coliform Concentrations at Segment 4........................................................................................261 Figure 7.15: Copano Bay Segments 1, 2, 3, 4 .....................................................264 Figure 7.16: Load Reductions for SchemaNode 154: Copano Bay ....................266 Figure 7.17: Existing versus Reduced Loads in Copano Bay Segment 2 ...........267 Figure 7.18: Load Reductions for SchemaNode 153: Copano Bay ....................269 xxxiii Figure 7.19: Existing versus Reduced Loads in Copano Bay Segment 3 ...........270 Figure 7.20: Load Reductions to Satisfy Fecal Coliform Standards for Monitored Conditions.......................................................................................271 Figure 8.1: Load Reduction Scenario #2: Aransas River Tidal.........................285 Figure 8.2: Load Reduction Scenario #2: Mission River Tidal.........................287 Figure 5A.1: Watershed JunctionIDs ...................................................................325 Figure 5A.2: Flow Length Raster to Mainstream and Copano Bay.....................329 Figure 7A.1: Control Sheet of Monte Carlo Simulation Model (User Interface).........................................................................................360 Figure 7A.2: Existing versus Modeled Fecal Coliform Concentrations of Monte Carlo Simulation Model ......................................................361 Figure 7A.3: Modeled Fecal Coliform Concentrations in Table Format of Monte Carlo Simulation Model ......................................................362 Figure 7A.4: Existing Monitoring Data at SchemaNode of Interest....................363 Figure 7A.5: SchemaNode Fields for Monte Carlo Simulation Model ...............366 Figure 7A.6: SchemaLink Fields for Monte Carlo Simulation Model.................370 Figure 7A.7: Load Reductions for SchemaNode 62: Aransas River Above Tidal ................................................................................................374 Figure 7A.8: Load Reductions for SchemaNode 68: Aransas River Above Tidal ................................................................................................375 Figure 7A.9: Load Reductions for SchemaNode 75: Aransas River Above Tidal ................................................................................................377 Figure 7A.10: Load Reductions for SchemaNode 63: Aransas River Tidal........379 Figure 7A.11: Load Reductions for SchemaNode 67: Aransas River Tidal........382 xxxiv Figure 7A.12: Load Reductions for SchemaNode 73: Mission River Above Tidal ................................................................................................385 Figure 7A.13: Load Reductions for SchemaNode 74: Mission River Above Tidal ................................................................................................387 Figure 7A.14: Load Reductions for SchemaNode 65: Mission River Above Tidal ................................................................................................388 Figure 7A.15: Load Reductions for SchemaNode 70: Mission River Tidal........390 Figure 7A.16: Load Reductions for SchemaNode 66: Mission River Tidal........392 Figure 7A.17: Load Reductions for SchemaNode 154: Copano Bay (Including Aransas Tidal Reductions) ............................................394 Figure 7A.18: Load Reductions for SchemaNode 153: Copano Bay (Incluing Mission River Tidal Reducitons) ....................................395 Figure 7A.19: Load Reductions to Satisfy Fecal Coliform Standards for Modeled Conditions........................................................................396 Figure 8A.1: Load Reduciton Scenario #1: Aransas River Above Tidal.............398 Figure 8A.2: Load Reduction Scenario #1: Aransas River Tidal.........................400 Figure 8A.3: Load Reduction Scenario #1: Mission River Above Tidal.............402 Figure 8A.4: Load Reduction Scenario #1: Mission River Tidal.........................404 1 Chapter 1: Introduction 1.1 BACKGROUND Section 303(d) of the 1972 Federal Clean Water Act (CWA) requires that each State identify water bodies that do not meet the State’s water quality standards and create a priority ranking of the impaired waters based on the severity of pollution and the water body’s intended use. For the State of Texas, the Texas Commission on Environmental Quality (TCEQ) has identified three water segments that do not meet the Texas Surface Water Quality Standards. Segment 2472, Copano/Port/Mission Bay, exceeds fecal coliform bacteria water quality standards for oyster water use. Segment 2003, Aransas River Tidal, exceeds enterococci bacteria water quality standards for contact recreation use, and Segment 2001, Mission River Tidal, exceeds enterococci bacteria water quality standards for contact recreation use. The three water segments are located along Texas’s southeastern coastline, which is shown in Figure 1.1. Figure 1.1 Impaired Water Segments in Copano Bay Watershed 2003 2004 2002 2001 2472 2 1.1.1 Purpose and Objectives The Copano Bay Bacterial Loadings Model will be used to support the TCEQ Total Maximum Daily Load (TMDL) program to develop the TMDLs for the three impaired water segments. The objectives of this research are to identify the major bacterial sources in the Copano Bay watershed, to calculate the total bacterial loadings (i.e., the TMDLs) from these sources, and to estimate the load reductions needed to bring each of the impaired segments into compliance with water quality standards. The primary bacterial indicator for recreational waters was fecal coliform until recently when the Environmental Protection Agency (EPA) began recommending Escherichia coli as a better freshwater indicator and enterococci as a better marine water indicator. Thus, the bacterial indicators for the Aransas and Mission River Tidals, which are classified as marine waters, were recently changed from fecal coliform to enterococci, and the bacterial indicators for the Aransas and Mission River Above Tidals, which are classified as freshwaters, were recently changed from fecal coliform to E. coli. However, fecal coliform is still the bacterial indicator for oyster water use standards in Copano Bay. Because the transition was more recent, there is not a significant amount of enterococci or E. coli monitoring data for the Tidals and Above Tidals as compared to fecal coliform monitoring data. For this reason, and because Copano Bay is the impaired water segment which motivates this study, fecal coliform bacterial loadings were modeled for this research. Thus, the TMDLs and estimate of total load reductions for each water segment were based on fecal coliform water quality standards. However, separate models will need to be created to model the other bacterial indicators (enterococci and E. coli) for the other water segments in subsequent studies. 3 1.1.2 Study Area The geographic extent of the project includes the three previously mentioned water segments, which are all located in the Copano Bay watershed: Copano Bay (Segment 2472), Aransas River Tidal (Segment 2003), and Mission River Tidal (Segment 2001). The study area and impaired segments are shown in Figure 1.1. The Copano Bay watershed is located along the southeastern Texas coastline and has a drainage area, which all drains to Copano Bay, of 5,688 km 2 . The Copano Bay watershed covers part of Aransas, Bee, Goliad, Karnes, Refugio, and San Patricio Counties as shown in Figure 1.2. Figure 1.2 Major Highways and Counties in the Copano Bay Watershed Legend Main Streams Major Roads Watersheds Aransas County Bee County Goliad County Karnes County Refugio County San Patricio Cty. U 7 7 U 5 9 S 3 5 S I 3 7 S 2 3 9 U 7 7 A S202 U 1 8 1 S 3 5 9 S 5 1 6 S 3 6 1 S 7 3 S 2 3 9 4 1.2 TEXAS SURFACE WATER QUALITY STANDARDS The Texas Surface Water Quality Standards, which are found in the Texas Administrative Code (TAC), Title 30, Chapter 307 (TCEQ, 2005a) specify the water quality standards that must be met. Section §307.7 gives the site-specific uses and associated criteria. For this research, the fecal coliform water quality standards and criteria are given below since a fecal coliform loadings model was created. The water uses that are of concern for these impaired water segments are contact recreation use (for Aransas and Mission River Tidals and Above Tidals) and oyster water use (for Copano Bay). Contact recreation includes recreational activities that involve a significant risk of ingestion of water, including wading by children, swimming, water skiing, diving, and surfing (TCEQ, 2005a). Oyster waters (Copano Bay, Segment 2472) are waters that produce edible species of clams, oysters, or mussels (TCEQ, 2005a). The following are the criteria for fecal coliform in contact recreation waters in the state of Texas, §307.7(b)(1)(A): (i) Fecal coliform content shall not exceed 200 colonies per 100 mL as a geometric mean based on a representative sampling of not less than five samples collected over not more than 30 days. (ii) Fecal coliform content shall not equal or exceed 400 colonies per 100 mL in more than 10% of all samples, but based on at least five samples, taken during any 30-day period. If ten or fewer samples are analyzed, no more than one sample shall exceed 400 colonies per 100 mL. The following are the criteria for fecal coliform in oyster waters in the state of Texas, §307.7(b)(3)(B): (i) A 1,000 foot buffer zone, measured in the water from the shoreline at ordinary high tide, is established for all bay and gulf waters, except those contained in 5 river or coastal basins as defined in §307.2 of this title (relating to Description of Standards). Fecal coliform content in buffer zones shall not exceed 200 colonies per 100 mL as a geometric mean of not less than five samples collected over not more than 30 days or equal or exceed 400 colonies per 100 mL in more than 10% of all samples taken during a 30-day period. (ii) Median fecal coliform concentration in bay and gulf waters, exclusive of buffer zones, shall not exceed 14 colonies per 100 mL, with not more than 10% of all samples exceeding 43 colonies per 100 mL. (iii) Oyster waters should be maintained so that concentrations of toxic materials do not cause edible species of clams, oysters, and mussels to exceed accepted guidelines for the protection of public health. Guidelines are provided by U.S. Food and Drug Administration Action Levels for molluscan shellfish. 6 1.3 TOTAL MAXIMUM DAILY LOAD PROGRAM The TMDL program is a TCEQ program that is striving to improve water quality in the state of Texas. The program was created to fulfill the requirements of Section 303(d) of the Federal Clean Water Act. The pathogen TMDL is the calculated allowable bacterial loadings that a waterbody can receive without exceeding water quality standards (EPA, 2005). The TMDL can be calculated by the following equation (EPA, 2005): TMDL = LC = WLA + LA + MOS Where: LC = Loading Capacity, or the greatest loading a waterbody can receive without exceeding water quality standards; WLA = Wasteload Allocation, or the portion of the TMDL allocated to existing or future point sources; LA = Load Allocation, or the portion of the TMDL allocated to existing or future non-point sources and natural background; and MOS = Margin of Safety, or an accounting of uncertainty about the relationship between pollutant loads and receiving water quality. The margin of safety can be provided implicitly through analytical assumptions or explicitly by reserving a portion of loading capacity. Once the TMDL has been determined for each of the impaired water segments, an implementation plan can be developed to bring the segments’ water quality into compliance with the water quality standards for a specific water use. 7 1.4 POTENTIAL SOURCES OF CONTAMINATION 1.4.1 Point Sources Point sources are any sources that directly discharge pathogens into a water body (EPA, 2005). The potential point sources of pathogens in the Copano Bay watershed that are considered in the model are wastewater treatment plants (WWTPs), which may discharge fecal waste directly to the watershed due to bypass events, and waterbird colonies, which have known locations around the Bay. 1.4.2 Non-Point Sources Non-point sources are indirect sources that are far enough away “…from waterbodies to allow attenuation of the pathogens in runoff, infiltrated water, or groundwater” (EPA, 2005). The major non-point source of bacteria is the feces of warm- blooded animals. The concentration of indicator bacteria (i.e., fecal coliform, E. coli, enterococci) in the impaired water segments suggests that pathogens may be entering the water body through improperly treated sewage or failing septic systems or from the feces of livestock, pets in urban areas, aquatic birds, and mammals (TCEQ, 2005b). The potential non-point sources in the Copano Bay watershed that are considered in the model are livestock, failing septic systems, and other non-point sources that originate from different types of land uses (e.g., urban, forest, etc.). 8 1.5 OUTLINE OF THESIS Chapter 1 explains the objectives and purpose of this project, the study area, the current Texas Surface Water Quality Standards as well as the potential bacterial sources in the Copano Bay watershed and the outline of this paper. Chapter 2 describes the different types of bacterial indicators that are measured in the Copano Bay water segments (fecal coliform, E. coli, and enterococci.) This project only models fecal coliform bacteria, but correlations are made in Chapter 2 between E. coli/enterococci and fecal coliform that can be used to convert the fecal coliform bacterial loadings into E. coli/enterococci bacterial loadings. Chapter 2 also describes the factors that can affect the decay rate of fecal coliform bacteria and presents studies that have been conducted to examine the effect of environmental factors and conditions on the survival rate of fecal coliform. Chapter 3 provides descriptions of the datasets and sources that were used in this research to calculate the fecal coliform bacterial loadings, delineate watersheds, and for the Schematic Processor and Monte Carlo Simulation Models. Chapter 4 analyzes the existing bacterial monitoring dataset for fecal coliform. This includes spatial and statistical analysis of the bacterial monitoring data, indicating the location and extent of exceedances of water quality standards in the Copano Bay watershed. Chapter 5 estimates fecal coliform bacterial loadings (the input into the models), including non-point and point source calculations. Chapter 6 models the bacterial transport of fecal coliform from the sources in the watersheds, rivers, and Copano Bay using the Schematic Processor Model, including how the model parameters were determined (Section 6.3.3). The results include the modeled 9 median fecal coliform concentrations throughout the watershed and the impact of different bacterial sources on the Copano Bay watershed. The Schematic Processor Model was used to model average annual conditions, to calculate bacterial loadings in each of the water segments, and to determine the impact of the different bacterial sources on the concentrations of bacteria in the water segments. Chapter 7 explains how the bacterial transport of fecal coliform from the sources is modeled using a Monte Carlo Simulation Model. The results indicate the load reductions needed (and from what sources) to meet fecal coliform water quality standards. Chapter 8 summarizes the results from the calculations and procedures that were described in Chapters 6 and 7. The current loadings, allowable loadings, and amount the loads need to be reduced are presented for all the water segments in the Copano Bay watershed (Aransas River Above Tidal, Aransas River Tidal, Mission River Above Tidal, Mission River Tidal, and Copano Bay.) Chapter 9 discusses the conclusions and recommendations from this research. 10 Chapter 2: Literature Review 2.1 BACTERIAL INDICATORS Coliforms and fecal streptococci are measured in surface waters because they are indicators of pathogenic bacteria, viruses, and protozoans, all of which are typically found in human and animal feces (EPA, 2006). Bacterial indicators are typically not harmful themselves, but they indicate the possible presence of pathogenic microorganisms that could be harmful to human health. Testing for bacterial indicators is simpler, cheaper, and less time-consuming than testing specifically for all the different types of pathogens; thus, bacterial indicators are measured rather than the pathogenic microorganisms in surface waters (EPA, 2006). Fecal coliforms, which are a subset of total coliform bacteria, have been used as the primary bacterial indicator for recreational waters. However, as described below, the Environmental Protection Agency (EPA) recently began recommending the use of E. coli and enterococci as better indicators. E. coli are fecal coliform bacteria, and enterococci are a subgroup of the fecal streptococci and have the ability to survive in salt water (EPA, 2006). In 2001, the TCEQ collected 445 surface water samples from southeast Texas to compare the three different bacterial indicators that are measured in the Copano Bay watershed (TCEQ, 2006a): fecal coliform, E. coli, and enterococci. Based on the results of the study, EPA recommends measuring E. coli as the bacterial indicator in fresh waters and enterococci as the bacterial indicator in marine/salt waters (EPA, 2006). The study also found that the number of samples exceeding the Texas Surface Water Quality Standards was greater for E. coli and enterococci than for fecal coliform. This also occurs in the Copano Bay watershed because none of the bacterial monitoring 11 stations on the Mission and Aransas River Tidals and Above Tidals exceed fecal coliform water quality standards for contact recreation use (TCEQ, 2006b). Thus, Chapter 7 may underestimate the load reductions required to meet contact recreation use standards for E. coli and enterococci in the Tidal and Above Tidal reaches, which are recommended as the better indicators of pathogenic bacteria in fresh and salt waters, respectively. The current bacterial indicators used for each water segment are given in Table 2.1. Table 2.1 Bacterial Indicators for Water Segments Environment Water Segments Water Use Bacterial Indicator Freshwater Stream Aransas and Mission River Above Tidals Contact Recreation Use E. coli Tidal Stream Aransas and Mission River Tidals Contact Recreation Use Enterococcus Bay Copano Bay Oyster Water Use Fecal Coliform 2.1.1 Correlation between Fecal Coliform and Enterococci Because fecal coliform is the current primary bacterial indicator for Copano Bay, all the bacterial loading calculations were calculated using fecal coliform for this research. However, to determine the TMDLs for the Tidal reaches, bacterial loading calculations for enterococci would need to be performed, and the resulting concentrations would then be compared to enterococci contact recreation use standards. However, it is difficult to find studies in which a direct or consistent correlation between fecal coliform and enterococci is found (TCEQ, 2006b). For this reason, different correlations will be used for different parts of the Copano Bay watershed. The TCEQ bacterial monitoring data was analyzed for each of the Tidal reaches to compare the measurements made for fecal coliform and enterococci (when measurements of each were made at the same station and day) to determine if the concentrations of fecal coliform and enterococci are correlated in this study area. 12 For the Aransas River Tidal (Station 12948), fecal coliform and enterococci were both measured on 11 days (Table 2.2). The concentrations of fecal coliform and enterococci at Station 12948 versus time are shown in Figure 2.1, and the relationship between fecal coliform and enterococcus is shown in Figure 2.2. Table 2.2 Fecal Coliform and Enterococcus Measurements at Station 12948: Aransas River Tidal Date Fecal Coliform, (#/100mL) Enterococcus (#/100mL) 10/25/1999 12 47 1/19/2000 20 16 4/17/2000 3700 12200 7/11/2000 112 590 1/14/2002 94 1082 4/9/2002 94 1082 7/8/2002 1327 3400 10/15/2002 122 60 1/21/2003 58 29 4/22/2003 34 210 8/18/2003 28 44 13 There is a strong correlation between fecal coliform and enterococcus (shown in Figure 2.1), and the relationship between these two bacterial indicators is shown in Figure 2.2. This relationship can be used to convert the fecal coliform bacterial loadings (calculated in Chapter 5) to enterococci bacterial loadings for the Aransas River Tidal watersheds. However, this was not done for this report but is recommended for future work. 0 2000 4000 6000 8000 10000 12000 14000 2/9/1999 8/28/1999 3/15/2000 10/1/2000 4/19/2001 11/5/2001 5/24/2002 12/10/2002 6/28/2003 1/14/2004 #/ 10 0 m L FC Enterococcus Figure 2.1 Fecal Coliform and Enterococcus Concentrations at Station 12948: Aransas River Tidal 14 For the Mission River Tidal (Station 12943), fecal coliform and enterococci were both measured on 16 days (Table 2.3). The concentrations of fecal coliform and enterococci at Station 12943 versus time are shown in Figure 2.3, and the relationship between fecal coliform and enterococcus is shown in Figure 2.4. y = 3.1927x + 79.797 R 2 = 0.9832 0 2000 4000 6000 8000 10000 12000 14000 0 500 1000 1500 2000 2500 3000 3500 4000 Fecal Coliform (#/100mL) E n t e rococcus ( # / 1 00mL) Figure 2.2 Relationship between Enterococcus and Fecal Coliform at Station 12948: Aransas River Tidal 15 Table 2.3 Fecal Coliform and Enterococcus Measurements at Station 12943: Mission River Tidal Date Fecal Coliform, (#/100mL) Enterococcus (#/100mL) 10/25/1999 32 98 1/19/2000 23 23 4/17/2000 52 31 7/11/2000 41 18 10/9/2000 3 13 1/15/2001 740 700 4/10/2001 37 68 6/18/2001 42 32 10/8/2001 22 84 1/14/2002 51 150 4/9/2002 51 150 7/8/2002 130 200 10/15/2002 21 250 1/21/2003 147 39 4/22/2003 270 74 8/18/2003 55 152 16 There appears to be a reasonable correlation between fecal coliform and enterococci concentrations for the field measurements (Figure 2.4). This relationship can be used to convert the fecal coliform bacterial loadings (calculated in Chapter 5) to enterococci bacterial loadings for the Mission River Tidal watersheds. However, this was not done for this report but is recommended for future work. 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 7/24/1998 12/6/1999 4/19/2001 9/1/2002 1/14/2004 5/28/2005 #/100 mL FC Enterococcus Figure 2.3 Fecal Coliform and Enterococcus Concentrations at Station 12943: Mission River Tidal 17 Comparing the monitoring data from the Aransas River Tidal to the Mission River Tidal (Figures 2.2 and 2.4), it can be seen that the correlations between fecal coliform and enterococcus vary greatly. The discrepancy shows how it is difficult to find a direct and consistent correlation between the two bacterial indicators. Thus, it is for this reason that different correlations should be used for different areas in the Copano Bay watershed. 2.1.2 Correlation between Fecal Coliform and E. coli In order to determine the TMDLs for the Above Tidal reaches, bacterial loading calculations for E. coli would have to be performed and the results compared to E. coli contact recreation use standards because E. coli is the preferred bacterial indicator for these reaches. However, like with enterococci, it is difficult to find studies in which a y = 0.7849x + 45.892 R 2 = 0.7215 0 100 200 300 400 500 600 700 800 0 100 200 300 400 500 600 700 800 Fecal Coliform (#/100mL) E n terococc us (#/10 0 mL) Figure 2.4 Relationship between Enterococcus and Fecal Coliform at Station 12943: Mission River Tidal 18 direct or consistent correlation between fecal coliform and E. coli is found (TCEQ, 2006b). For this reason, different correlations will be used for different parts of the Copano Bay watershed. The TCEQ bacterial monitoring data were analyzed for each of the Above Tidal reaches to determine if the concentrations of fecal coliform and E. coli are correlated. For the Aransas River Above Tidal (Station 12952), fecal coliform and E. coli were both measured on 5 days (Table 2.4). The concentrations of fecal coliform and E. coli at Station 12952 versus time are shown in Figure 2.5, and the relationship between fecal coliform and E. coli is shown in Figure 2.6. Table 2.4 Fecal Coliform and E. coli Measurements at Station 12952: Aransas River Above Tidal Date Fecal Coliform, (#/100mL) E. coli (#/100mL) 7/8/2002 836 400 10/15/2002 25 3 1/21/2003 72 3 4/22/2003 130 90 8/18/2003 58 56 19 There appears to be a strong correlation between fecal coliform and E. coli concentrations (Figure 2.6). This relationship can be used to convert the fecal coliform bacterial loadings (calculated in Chapter 5) to E. coli bacterial loadings for the Aransas River Above Tidal watersheds. However, this was not done for this report but is recommended for future work. 0 100 200 300 400 500 600 700 800 900 5/24/2002 7/13/2002 9/1/2002 10/21/2002 12/10/2002 1/29/2003 3/20/2003 5/9/2003 6/28/2003 8/17/2003 10/6/2003 #/100mL FC E.coli Figure 2.5 Fecal Coliform and E. coli Concentrations at Station 12952: Aransas River Above Tidal 20 For the Mission River Above Tidal (Station 12944), fecal coliform and E. coli concentrations were both measured on 17 days (Table 2.5). The concentrations of fecal coliform and E. coli at Station 12944 versus time are shown in Figure 2.7, and the relationship between fecal coliform and E. coli is shown in Figure 2.8. y = 0.4769x + 3.4784 R 2 = 0.9765 0 50 100 150 200 250 300 350 400 450 0 100 200 300 400 500 600 700 800 900 Fecal Coliform (#/100mL) E. col i ( # / 100m L) Figure 2.6 Relationship between E. coli and Fecal Coliform at Station 12952: Aransas River Above Tidal 21 Table 2.5 Fecal Coliform and E. coli Measurements at Station 12944: Mission River Above Tidal Date Fecal Coliform, (#/100mL) E. coli (#/100mL) 10/25/1999 58 52 1/19/2000 410 470 4/17/2000 112 74 7/11/2000 94 48 10/9/2000 1382 4200 1/15/2001 410 460 4/10/2001 320 62 6/18/2001 56 30 10/8/2001 157 55 1/14/2002 54 260 4/9/2002 54 260 7/8/2002 682 627 10/15/2002 46 92 1/21/2003 142 31 4/22/2003 120 132 5/12/2003 116 42 8/18/2003 54 10 22 There appears to be a good correlation between fecal coliform and E. coli concentrations (Figure 2.8). This relationship can be used to convert the fecal coliform bacterial loadings (calculated in Chapter 5) to E. coli bacterial loadings for the Mission River Above Tidal watersheds. However, this was not done for this report but is recommended for future work. 0 500 1000 1500 2000 2500 3000 3500 4000 4500 2/9/1999 8/28/1999 3/15/2000 10/1/2000 4/19/2001 11/5/2001 5/24/2002 12/10/2002 6/28/2003 1/14/2004 #/100 mL FC E.coli Figure 2.7 Fecal Coliform and E. coli Concentrations at Station 12944: Mission River Above Tidal 23 Comparing the monitoring data from the Aransas River Above Tidal to the Mission River Above Tidal (Figures 2.6 and 2.8), it can be seen that the correlations between fecal coliform and E. coli vary greatly. The discrepancy shows how it is difficult to find a direct and consistent correlation between the two bacterial indicators. Thus, it is for this reason that different correlations should be used for different areas in the Copano Bay watershed. y = 2.6882x - 268.57 R 2 = 0.8457 -500 0 500 1000 1500 2000 2500 3000 3500 4000 4500 0 200 400 600 800 1000 1200 1400 1600 Fecal Coliform (#/100mL) E .coli (#/1 00mL) Figure 2.8 Relationship between E. coli and Fecal Coliform Concentrations at Station 12944: Mission River Above Tidal 24 2.2 DECAY RATE OF FECAL COLIFORM BACTERIA 2.2.1 Factors Affecting Decay Rate of Fecal Coliform The fecal coliform bacteria are assumed to decay by a first-order decay process (van der Steen et al., 2000), so the decay of bacteria is modeled by the decay coefficient, k. The expression for first-order decay is given as follows: c t = c o * exp -kt (2.1) Where: c t = fecal coliform concentration after time t c o = initial fecal coliform concentration at t = 0 k = first-order decay coefficient t = time Many factors affect the decay rate of fecal coliform, including solar radiation intensity, temperature, salinity, pH, dissolved oxygen (DO), turbidity, presence of toxic agents, predation and parasitism, sedimentation, and nutrient concentrations (Davies et al., 1995). Some of the most critical factors have been identified as solar radiation intensity and temperature (Brissaud et al., 2000; Burkhardt et al., 2000), and the decay constant, k, given in literature varies considerably, from 0.2 to 12 days -1 (Brissaud et al., 2000). Many empirical equations have been formulated to model the effects of the various factors on fecal coliform decay. Canale et al. (1993) considered the combined effects of irradiance and temperature as well as sedimentation effects. However, there is an insufficient amount of data to apply this expression to the Copano Bay watershed. 25 2.2.1.1 pH The fecal coliform decay rate is not significantly affected by pH when the pH is in the range of 7.2 and 9.1 (van der Steen et al., 2000). However, pH values above 9.0 increase the fecal coliform decay rate, especially under poor nutrient conditions; furthermore, increased temperatures exacerbate the pH effect (Pearson et al., 1987). Curtis et al. (1992) showed that when pH > 9.0, the fecal coliform decay rate increases. However, looking at the water quality monitoring data from 1999-2005, the measured pH at all bacteria monitoring stations is less than 9.0, so pH is assumed not to be a significant factor on the decay rate of fecal coliform in the Copano Bay watershed. The average measured pH (from 1999-2005) in the Copano Bay Segments 1 are given in Table 2.6 below. Table 2.6 Average Measured pH in Copano Bay Segments Copano Bay Segment 1 Average pH 1 8.27 2 8.00 3 No Data 4 7.98 Since all the pH measurements are within the range of 7.2 and 9.1, pH was considered an insignificant factor on the decay rate of fecal coliform bacteria. 1 Copano Bay segmentation is defined in Section 6.3.1.1 26 2.2.1.2 Solar Radiation Intensity Van der Steen et al. (2000) developed an empirical expression for the decay coefficient based on solar intensity and radiation in a solar-radiated pond environment. However, when the expression from this study was applied to the Copano Bay watershed using radiation intensities from the North American Regional Reanalysis (NARR) database, the fecal coliform decay coefficients were in the range of 35.5 days -1 to 44.5 days -1 , which are much greater than the typical range of 0.2 to 12 days -1 (Brissaud et al., 2000). Thus, the empirical equation was not applied to Copano Bay. Van der steen et al. (2000) exposed fecal coliform to solar radiation in batch reactors at 20°C and 30°C. The results showed that the temperature difference did not affect decay; the results also showed that fecal coliform decay is much more rapid under irradiated conditions than under dark conditions (van der Steen et al., 2000). Curtis et al. (1992) determined that fecal coliform decay by solar radiation is dependent on the dissolved oxygen concentration (i.e., photooxidation). Longer wavelengths (> 440 nm) could not kill fecal coliform when pH values are below 8.0. Thus, photooxidation is dependent on sunlight radiation, pH, and dissolved oxygen concentration. 2.2.1.3 Temperature The literature is divided over the effect of temperature on fecal coliform decay. Van der Steen et al. (2000) conducted experiments with buffered effluents from Upflow Anaerobic Sludge Blanket (UASB) reactors, which treat domestic wastewater. Fecal coliform decay increased as temperature increased from 10 to 30°C. Two empirical 27 expressions were developed in this study to calculate decay coefficients based on temperature. The expressions were applied to the Copano Bay watershed, and the calculated decay coefficients ranged from 0.53 to 2.93 days -1 . Auer and Niehaus (1993) measured fecal coliform in batch culture under dark conditions where the temperature ranged from 10-35°C. They concluded that there is no significant relationship between fecal coliform decay rate and temperature, and the great variation between temperature and die-off rate relationships may be due to other factors having an effect on the experiments (such as nutrients, sunlight radiation, etc.) The average annual measured water temperature (from 1999-2005) in the Copano Bay Segments 2 are given in Table 2.7 below. Table 2.7 Average Measured Temperature in Copano Bay Segments Copano Bay Segment 2 Temperature (°C) 1 19.6 2 20.0 3 17.8 4 18.9 2.2.1.4 Salinity Estuarine waters have higher salinities than freshwaters; studies show that the decay rate of fecal coliform is greater in saltwater than in freshwater (Anderson et al., 2005). The average measured salinities (from 1999-2005) in the Copano Bay Segments 2 are given in Table 2.8 below. Segments 1 and 4 have higher average salinity concentrations because they are closer to the Gulf of Mexico, which sea water typically has a salinity of approximately 35 ppt. 2 Copano Bay segmentation is defined in Section 6.3.1.1 28 Table 2.8 Average Measured Salinity in Copano Bay Segments Copano Bay Segment Salinity (ppt) 1 13.7 2 11.9 3 10.7 4 13.7 2.2.2 Summary The fecal coliform decay rate is influenced by many environmental factors. Thus, it is critical that our model uses a variable decay rate to account for day and nighttime conditions, dry and wet weather conditions, and summer versus wintertime conditions, and so forth (Kashefipour et al., 2002). Because none of the empirically-derived decay coefficient equations (based on various environmental factors) could be applied to the Copano Bay watershed due to lack of data or resulting in values that were not within the range of literature values, the decay coefficients for the watershed were calculated by directly solving for k by using available monitoring data, and the procedure is described in Section 6.3.3.1. Thus, the range of decay coefficients is 2 to 2.5 days -1 for this study. 29 Chapter 3: Data Description 3.1 BACTERIAL MONITORING DATA Bacterial monitoring data for all of the bacterial monitoring stations for the water segments in the Copano Bay watershed were used to observe and analyze the fecal coliform concentrations in the stream segments. The data were also used to calibrate the Schematic Processor Model (Chapter 6) and the Monte Carlo Simulation Model (Chapter 7) to ensure that the modeled fecal coliform concentrations agree with the existing monitoring concentrations at each bacterial monitoring station. Bacterial monitoring data were obtained for the time period of January 1999 to October 2004 from Texas Department of Health (TDH) and TCEQ Regulatory Activities and Compliance System (TRACS) database for the Copano Bay watershed. The TCEQ TRACS database stores surface water quality data from TCEQ water quality monitoring stations (TCEQ, 2006c). The monitoring data are organized by Station ID (the unique identifier for the bacteria monitoring station), the date the monitoring occurred, and the Storet code, which is a unique number that corresponds to the water quality parameter. The Storet codes associated with fecal coliform monitoring data are 79835 and 31616. Storet code 79835 describes fecal coliform concentrations in units of most probable number per 100 mL (MPN/100mL). MPN/100mL is measured using the multiple-tube fermentation technique. Approximately 84% of the fecal coliform concentrations in Copano Bay were measured using this technique. A Storet code of 31616 describes fecal coliform concentrations in units of number per 100 mL (#/100mL), which is the number of coliform bacteria per 100 mL of water and is measured using the membrane filtration method. Approximately 16% of the fecal 30 coliform concentrations in Copano Bay and 100% of the fecal coliform samples in the Aransas and Mission Rivers were measured using this technique. Both of these measurements were used interchangeably in the bacterial analysis, and the units are defined as colony forming units per 100 mL of water (CFU/100mL) subsequently in this study. Other parameter Storet codes that are used in our analysis are 31648 (E. coli, #/100mL) and 31649 (enterococcus, #/100mL). E. coli, which is a freshwater bacterial indicator, is the indicator for the Aransas and Mission Above Tidal reaches; enterococcus, which is a marine water bacterial indicator, is the indicator for the Aransas and Mission Tidal reaches. A correlation between E. coli/enterococcus and fecal coliform was found (described in Section 2.1) to create models that can determine the bacteria load reductions needed in the Above Tidal and Tidal reaches (TCEQ, 2006d). 31 3.2 DATASETS REQUIRED FOR LOADING ESTIMATION 3.2.1 Datasets Required for Non-Point Source Loads To create a model to calculate bacterial loadings, Geographic Information Systems (GIS) data layers were compiled. The basic relationship that was used to calculate non-point source bacterial loadings for the model is L = Q * C (3.1) Where: L = Bacterial Loading Q = Runoff C = Concentration GIS data layers were prepared to calculate Runoff (Q) and Concentration (C). 3.2.1.1 Runoff Dataset As explained in Section 5.1.2, several GIS data layers were used to calculate the runoff in the Copano Bay watershed. Runoff calculations were made using previously generated empirical equations (Quenzer, 1997). These runoff equations were developed by using the Microsoft Excel 5.0 Regression Tool, which was used to base the equations on a relationship among streamflow depth, precipitation depth, and percent land use in each of the nine watersheds in the Corpus Christi Bay system, which includes the Copano Bay watershed. These equations are given in Section 5.1.2.4. 3.2.1.1.1 Precipitation Data The National Resources Conservation Service (NRCS) and the Spatial Climate Analysis Service (SCAS) at Oregon State University (OSU) developed PRISM (Parameter-elevation Regressions on Independent Slopes Model), which gives the average annual precipitation from 1961-1990. These data were downloaded for the state 32 of Texas in shapefile format and were used to calculate runoff in the Copano Bay watershed. 3.2.1.1.2 Land Use / Land Cover Dataset The National Land Cover Characterization project developed a national land cover data set from Multi-Resolution Land Characterization (MRLC) data called National Land Cover Data 1992 (NLCD 92). The National Land Cover Dataset is based on 30- meter Thematic Mapper data. NLCD data also exists for 2001; however, the data do not currently exist for the geographic area of interest (i.e., the Copano Bay watershed). The 1992 dataset was used along with the average annual precipitation to calculate runoff for the Copano Bay watershed. 3.2.1.2 Event Mean Concentration (EMC) Dataset As explained in Section 5.1.2.7, several GIS data layers were used to calculate bacteria concentrations in the Copano Bay watershed. The land use/land cover dataset was obtained from the United States Geographic Survey (USGS) (Section 3.2.1.1.2). The event mean concentration (EMC) values can be approximated for each type of land use. For this research, fecal coliform EMCs for each land use code were previously determined (Zoun, 2003) and are listed in Table 3.1, and the Source Code descriptions for the EMC values in Table 3.1 are given in Table 3.2. The EMC values are average fecal coliform concentrations during an entire storm event associated with different types of land use in the Galveston Bay watershed, not the Copano Bay watershed. For this reason, we decided to find a more accurate way to account for animal fecal waste based on the numbers and types of animals in the Copano Bay watershed. Thus, the fecal coliform EMC values for land use classifications 51 (Shrubland), 71 (Grasslands/Herbaceous), and 33 81 (Pasture/Hay) were modified to zero in the non-point source calculations so that bacteria from livestock waste were not accounted for twice. Table 3.1 Fecal Coliform EMC Values Based on Land Use Classifications (Zoun, 2003) Land Use Code Land Use Category Fecal Coliform EMCs (CFU/ 100 mL) Source Code 11 Open Water 0 NPS, Judgment 21 Low Intensity Residential 22,000 NPS 22 High Intensity Residential 22,000 NPS 23 Commercial/Industrial/Transportation 22,000 Inferred from NPS 31 Bare Rock/Sand/Clay 0 Judgment 32 Quarries/Strip Mines/ Gravel Pits 0 Judgment 41 Deciduous Forest 1,000 Judgment 42 Evergreen Forest 1,000 Judgment 43 Mixed Forest 1,000 Inferred, Judgment 51 Shrubland 0 Livestock 61 Orchards/Vineyards/Other 2,500 Inferred from NPS 71 Grasslands/Herbaceous 0 Livestock 81 Pasture/Hay 0 Livestock 82 Row Crops 2,500 NPS 83 Small Grains 2,500 NPS 85 Urban/Recreational Grasses 22,000 NPS 91 Woody Wetlands 200 Judgment 92 Emergent Herbaceous Wetlands 200 Judgment 34 Table 3.2 Description of Source Codes for EMC Values Source Code Description NPS (Zoun, 2003) Galveston Bay National Estuary Program Non-point Source Characterization (NPS) study CCBNEP (Zoun, 2003) Corpus Christi Bay National Estuary Program (CCBNEP) Study Inferred (Zoun, 2003) Value inferred from observed data for similar land use category in Galveston Bay area due to lack of data for the specific land use category in Galveston Bay area Judgment (Zoun, 2003) Professional judgment by Dr. George Ward, Professor, University of Texas at Austin Livestock Land use codes where livestock animals are assumed to be present. (Note: values are assumed to be zero, so that animal feces are only accounted for once in model. Livestock fecal coliform concentrations are accounted for in Section 5.2.) 3.2.1.3 Livestock Data Livestock data (annual count per county) were obtained from the 2002 Census of Agriculture, National Agricultural Statistics Service (NASS), and the 2004 Texas Livestock Inventory and Production, United States Department of Agriculture (USDA), NASS, Texas Statistical Office. The animals that were considered in the calculations (due to census data availability) were cattle, goats, horses, sheep, hen, hogs, and chickens. 3.2.1.4 Septic System Data The number of septic systems per county was obtained from the 1990 U.S. Census of Bureau for the Copano Bay watershed 3 . However, the exact locations of the septic systems were not given, nor is information available regarding the number of 3 The 2000 Census does not include questions regarding sewage disposal, so the number of septic systems per county is unknown since 1990. 35 malfunctioning septic systems. Other data from the U.S. Census of Bureau that was used in calculating the bacterial loadings from septic systems (see Section 5.5) were the occupied housing units per county (1990 and 2002), and the population per county (2004). The Texas Department of Health (TDH) regulated septic systems prior to 1990. However, in 1991, the TCEQ was given the authority to regulate on-site sewage facilities (OSSFs), which includes authority over location, design, construction, installation, and proper functioning of OSSFs (Niemann, 2006). Thus, the number of installed septic systems from 1990 – 2004 per county was obtained from the TCEQ. The types of soil in the watershed are also important for the determination of bacterial loadings from septic systems. The State Soil Geographic (STATSGO) Database, which gives soil maps with a mapping scale of 1:250,000, was used to identify the hydrologic soil groups throughout the Copano Bay watershed (Groups A, B, C, D). For example, Group A consists of soils that have low runoff potential and high infiltration rates and typically consist of USDA soil textures of sand, loamy sand, and sandy loam. The transmission rate is typically greater than 0.76 cm/hr (Maidment, 1992). Septic systems with soils classified in Group A are more likely than Groups B, C, and D to contaminate the groundwater and surface waters. The number of malfunctioning septic systems in the other soil groups (B, C, D) in the Copano Bay watershed was estimated by looking at the Authorized Agents (AA) Monthly Reports that are submitted to the TCEQ Compliance Support Division OSSF Program. These reports are available on the TCEQ website and are called OSSF Activity Reports. The relevant information from this site is that it lists the monthly “Complaints Investigated” and “Court Cases Filed” per county for OSSFs. 36 The Comprehensive Sanitary Survey of the Shellfish Producing Waters of Copano Bay (TDH, 2000) gives approximate locations of septic systems around the Copano Bay area and reports only one malfunctioning septic system in the area around the Bay. This report was used to approximate the location of septic systems around Copano Bay. 3.2.2 Datasets Required for Point Source Loads 3.2.2.1 Bird Data Approximately 30 different types of colonial waterbird species live along the Texas coastline. The Texas Colonial Waterbird Census, using bird population data collected by volunteers from state, federal, non-profit organizations, and professional organizations, gives the number of breeding pairs of colonial waterbirds along the Texas Coast. As detailed in Section 5.3, these data were used to calculate annual waste loadings from colonial waterbirds. 3.2.2.2 Industrial/Municipal Wastewater Outfalls Data The locations of industrial/municipal wastewater outfalls were obtained from the Permitted Wastewater Outfalls shapefile provided by the TCEQ. Descriptions of the permitted facilities were obtained from Sandra Alvarado from the TCEQ TRACS database. Permit monitoring data (including fecal coliform and flow measurements) of water discharge permits (discharge monitoring reports) were obtained from the Permit Compliance System (PCS) Database from the U.S. Environmental Protection Agency (EPA). Wastewater treatment plant (WWTP) bacterial loadings are calculated in Section 5.4. 37 3.2.2.3 Concentrated Animal Feedlot Operations (CAFOs) Data A shapefile that contains CAFOs within the Copano Bay watershed was obtained from the TCEQ. However, there is only one permitted facility, and it was not recently renewed because the company is no longer operating (Alvarado, 2005). Thus, there are no CAFOs within the Copano Bay watershed at this time. 3.2.3 Water Rights Analysis Package (WRAP) Hydro Watershed Delineation Dataset Before calculating the bacterial loadings, watersheds must be delineated because the bacterial loading per watershed is needed for the Schematic Processor Model. Watershed delineation requires a Digital Elevation Model (DEM), river network, and Critical Points, which is a feature class that contains points where the fecal coliform concentration can be examined. Critical Points were determined to be USGS gauge stations, bacterial monitoring stations (so modeled values can be compared to existing monitoring data), and water segment endpoints. This process is described in Section 5.1.2.1. 3.2.3.1 DEM and Terrain Preprocessing The DEM was obtained from the National Elevation Dataset (NED) from USGS, which provides seamless coverage of the United States, providing a 1:24,000-scale DEM. The DEM, along with the National Hydrography Dataset and critical points of interest (feature class called “CriticalPoints”), provides the necessary data to conduct Terrain Preprocessing (Appendix 5.1) in Arc Hydro to determine the drainage patterns for the basin. The drainage patterns determine the pathway by which the bacteria reach the impaired water segments. 38 3.2.3.2 National Hydrography Dataset (NHD) The National Hydrography Dataset (NHD) provides digital spatial data about surface water features such as lakes, ponds, streams, rivers, springs and wells. This dataset is based on the USGS Digital Line Graph (DLG) hydrography data and on information from the EPA Reach File Version 3 (RF3). However, there were missing data in the NHD dataset (feature class called “NHDFlowline”) for the Copano Bay watershed (e.g., random gaps in the river segments at multiple locations). Using the Editor Toolbar in ArcGIS, new features were created in the NHDFlowline feature class to ensure that all the river segments were connected within the river network. 3.2.3.3 Critical Points (USGS Gauge Stations, Bacterial Monitoring Stations, Water Segment Endpoints) The locations of the USGS gauge stations were obtained from Better Assessment Science Integrating Point and Non-point Sources (BASINS). BASINS is an environmental analysis system developed by the U.S. EPA that can be used to perform watershed and water quality studies. BASINS allows a user to evaluate point and non- point source data in an easy-to-use format. To use BASINS, a user specifies a geographic area of interest, and the system downloads data from EPA, USGS, and other GIS data internet sources. The locations of the water segment endpoints and bacterial monitoring stations were received from Sandra Alvarado from the TCEQ. 3.2.4 Datasets Required for Schematic Processor Model 3.2.4.1 USGS Stream Gauge Data USGS stream gauge data were used to compare the modeled average annual runoff to existing average annual flowrates at the gauge stations as well as to find the 39 residence time distributions (Section 6.3.3.2) associated with the various river segments in the Copano Bay watershed. USGS stream gauge data was downloaded for the USGS gauge stations located in the Copano Bay watershed. The USGS station names in the watershed as well as the periods of records of the available data are given below: • USGS 08189200 Copano Ck nr Refugio, TX (1970-2004) • USGS 08189300 Medio Ck nr Beeville, TX (1962-2004) • USGS 08189500 Mission Rv at Refugio, TX (1939-2004) • USGS 08189700 Aransas Rv nr Skidmore, TX (1964-2004) The data that are used for the Schematic Processor Model are the historical available daily streamflow data and streamflow measurements, which specify measurements of the width, streamflow, and area of the channel at the USGS gauge station locations. 3.2.4.2 Bathymetry Data Bathymetry data give the water depth of water bodies. A hard copy bathymetry map, shown in Figure 3.1, (Ward, 2005) was available for Copano Bay, but not for the upstream rivers. The bathymetry map was used to calculate the volumes of each of the four Copano Bay water segments. 40 Figure 3.1 Bathymetry Map of Copano Bay 41 3.3 MAP PROJECTION AND COORDINATE SYSTEMS All the datasets used for this project (described in Section 3.2) were retrieved from sources that may have had different map projection and coordinate systems. For GIS analysis and processes, it is critical to have all the datasets in the same coordinate system; thus, all the datasets for this project were projected into the same coordinate system. The map projection that was used in the analyses of this project was Albers Conical Equal Area, and the geographic coordinate system that was used is North America Datum of 1983 (NAD 83). The projected coordinate system name is NAD_1983_Texas_Centric_Mapping_System_Albers, and the geographic coordinate system name is GCS_North_American_1983. The parameters for this projection are given in Table 3.3. Table 3.3 Parameters for NAD_1983_Texas_Centric_Mapping_System_Albers Projection Albers Conical Equal Area Datum North American Datum of 1983 (NAD83) Standard Parallel #1 (degrees) 27.5 Standard Parallel #2 (degrees) 35.0 Longitude of Central Meridian (degrees) -100.0 Latitude of Projection Origin (degrees) 18.0 False Easting (meters) 1,500,000 False Northing (meters) 6,000,000 Units of Measure Meters 42 Chapter 4: Analysis of Monitoring Dataset 4.1 SPATIAL DISTRIBUTION OF FECAL COLIFORM 4.1.1 Methodology The bacterial monitoring data for fecal coliform was analyzed for this project because the bacterial indicator for oyster water use in Copano Bay is fecal coliform, and the most data exists for this bacterial indicator. The spatial distribution of fecal coliform concentrations was analyzed in the Copano Bay watershed at the locations of the bacterial monitoring stations. The fecal coliform monitoring data came from the TCEQ TRACS database and is from the time period of January 1999 to May 2005. Fecal coliform bacteria are measured quarterly throughout the year, such that seasonal variations can be observed. The fecal coliform standards that apply to the rivers and Copano Bay are given in Table 4.1. Table 4.1 Fecal Coliform Water Quality Standards for Water Segments Water Segment Water Use Geometric Mean (CFU/100mL) Percent greater than Single Sample of 400 CFU/100mL (%) Aransas River Above Tidal Aransas River Tidal Mission River Above Tidal Mission River Tidal Contact Recreation Use < 200 < 25 Water Segment Water Use Median (CFU/100mL) 90 th -Percentile (CFU/100mL) Copano Bay Oyster Water Use < 14 < 43 43 4.1.2 Procedure of Application The minimum, maximum, geometric mean, arithmetic mean, and median of the existing fecal coliform concentrations (from 1999-2005) at each bacterial monitoring station are calculated and displayed using graduated symbols in ArcMap to convey the spatial variation of fecal coliform bacteria. 4.1.3 Result The arithmetic mean, geometric mean, median, minimum, and maximum fecal coliform data (TCEQ TRACS database from 1999-2005) are shown in Figures 4.1 - 4.5 at each of the TCEQ bacterial monitoring stations in the Copano Bay watershed. The mean of all existing monitoring data is shown in Figure 4.1. Bacterial Monitoring Stations Mean (CFU/100mL) 2 - 22 23 - 76 77 - 140 141 - 394 395 - 898 Legend Figure 4.1 Mean of Fecal Coliform Concentrations at TCEQ Bacterial Monitoring Stations (1999-2005) Poesta Creek Aransas Creek Medio Creek Blanco Creek Copano Creek Mission River Chiltipin Creek Aransas River Olmos Creek 44 The mean fecal coliform concentrations are lower at stations that are closer to Copano Bay (Figure 4.1). Thus, water quality decreases the further upstream from the Bay. This trend can be explained by the effects that various environmental factors (e.g., solar radiation intensity, temperature, and salinity) have on bacterial decay. The lower concentrations in the Bay as compared to the upstream rivers and streams are also indicative of the dilution effects and higher salinity of Copano Bay. Note that higher mean fecal coliform concentrations occur in Copano Bay at locations where rivers and streams discharge into the Bay, but the lowest mean fecal coliform concentrations in the Bay occur at locations where no rivers discharge. The highest mean fecal coliform concentrations are measured near where Olmos Creek enters Aransas Creek, where Poesta and Aransas Creeks merge to become Aransas River, and where Blanco and Medio creeks merge to become Mission River. The geometric means and median concentrations of the existing data at the monitoring stations are shown in Figures 4.2 and 4.3, respectively. Use of the geometric mean and median concentrations reduces the effects of the high fecal coliform concentrations. 45 Figure 4.2 Geometric Mean of Fecal Coliform Concentrations at TCEQ Bacterial Monitoring Stations (1999-2005) 143 - 311 52 - 142 29 - 51 10 - 28 2 - 9 Geometric Mean (CFU/100mL) Legend Bacteria Monitoring Stations 46 Ignoring the high storm events that would skew the mean fecal coliform concentration values, the highest geometric mean/median fecal coliform concentrations are also in the upstream rivers and streams (as shown in Figures 4.2 and 4.3). However, the oyster water use standards that apply to Copano Bay are more stringent than the contact recreation use standards that apply to the river segments (Table 4.1). The highest geometric mean/median FC concentrations are measured where Olmos Creek enters Aransas Creek, and where Blanco and Medio Creeks merge to become Mission River. The measured median fecal coliform concentrations in the Bay comply with oyster water use fecal coliform water quality standards (median < 14 CFU/100mL.) Legend Bacterial Monitoring Stations Median (CFU/100mL) 2 - 7 8 - 47 48 - 72 73 - 116 117 - 260 Figure 4.3 Median of Fecal Coliform Concentrations at TCEQ Bacterial Monitoring Stations (1999-2005) 47 The minimum and maximum fecal coliform concentrations of the existing data at monitoring stations are shown in Figures 4.4 and 4.5, respectively. Legend Bacterial Monitoring Stations Minimum (CFU/100mL) 1 2 - 3 4 - 12 13 - 25 26 - 54 Figure 4.4 Minimum of Fecal Coliform Concentrations at TCEQ Bacterial Monitoring Stations (1999-2005) 48 Very high fecal coliform concentrations are measured at the Aransas River and Copano Creek outlets in Copano Bay (Figure 4.5). These high fecal coliform concentrations can be attributed to storm events. It is these storm events that cause portions of Copano Bay to exceed oyster water use fecal coliform water quality standards (90 th -percentile > 43 CFU/100mL). The fecal coliform monitoring data from these bacterial stations are compared to modeled results in Chapters 6 and 7. The bacterial monitoring stations associated with each water segment in the Copano Bay watershed are identified in Table 4.2. Legend Bacterial Monitoring Stations Maximum (CFU/100mL) 11 - 170 171 - 400 401 - 836 837 - 1600 1601 - 5700 Figure 4.5 Maximum of Fecal Coliform Concentrations at Bacterial Monitoring Stations (1999-2005) 49 Table 4.2 Water Segment Locations of Bacterial Monitoring Stations Copano Bay Water Segment Segment ID Segment Name Bacterial Monitoring Stations N/A 2001 Mission River Tidal 12943 N/A 2002 Mission River Above Tidal 12944 N/A 2003 Aransas River Tidal 12948 N/A 2004 Aransas River Above Tidal 12952 1 13405, 14782, 14784, 14790 2 12945, 14783, 14787, 14788 3 14797 4 2472 Copano Bay 13404, 14779, 14780, 14785, 14792, 14793 The summary of the fecal coliform data for the water body segments (TCEQ segments) in the Copano Bay watershed can be found in Table 4.3. The summary of the fecal coliform monitoring data for the Copano Bay water segments (defined in Section 6.3.1.1) can be found in Table 4.4. 50 Table 4.3 Summary of TCEQ Fecal Coliform Monitoring Data for Water Segments TCEQ Segment ID Segment Name Number of Stations Arithmetic Mean (CFU/100mL) Median (CFU/100mL) Minimum (CFU/100mL) Maximum (CFU/100mL) Number of Data Points 2001 Mission River Tidal 1 107 47 3 740 16 2002 Mission River Tidal 1 251 116 46 1382 17 2003 Aransas River Tidal 1 394 96 12 3700 16 2004 Aransas River Above Tidal 1 224 72 25 836 5 2472 Copano Bay 15 33 2 1 1600 497 Table 4.4 Summary of TCEQ Fecal Coliform Monitoring Data for Copano Bay Water Segments Segment ID Segment Name Number of Stations Arithmetic Mean (CFU/100mL) Median (CFU/100mL) Minimum (CFU/100mL) Maximum (CFU/100mL) Number of Data Points 1 Watershed JunctionID 45405 Outlet 4 17 2 1 390 113 2 Aransas River Outlet 4 68 2 1 1600 121 3 Mission River Outlet 1 22 2 2 240 31 4 Copano Creek Outlet 6 25 2 1 1600 232 51 4.2 STATISTICAL DISTRIBUTION OF FECAL COLIFORM 4.2.1 Methodology Two criteria need to be met for fecal coliform oyster water use standards: 1) the median of measured data needs to be less than 14 CFU/100mL, and 2) 90% of the measured concentrations need to be less than 43 CFU/100mL. The water quality of the four water segments of Copano Bay needs to comply with these oyster water use standards. Two criteria need to be met for fecal coliform contact recreation use standards: 1) the geometric mean of measured data needs to be less than 200 CFU/100mL, and 2) 75% of the measured concentrations need to be less than 400 CFU/100mL. The water qualities of Aransas and Mission River Tidals and Aransas and Mission River Above Tidals need to comply with these contact recreation use standards. To compare the monitoring data to water quality standards (also summarized in Table 4.1), the fecal coliform data at each bacterial monitoring station in the upstream rivers and segments and the monitoring data in each of the Copano Bay water segments (Segments 1, 2, 3, and 4) are plotted on a log scale versus the probability of exceedance. The two criteria (contact recreation use and oyster water use) are indicated on each plot in Section 4.2.3. 4.2.2 Procedure of Application Blom’s plotting formula was used to plot the probability distributions of the existing bacterial monitoring data, and a lognormal distribution was assumed (Zoun, 2003). P = 100 % * (m – 3/8) / (n + ¼) (4.1) 52 Where: P = probability of exceedance (%) m = rank; (m = 1 for largest FC concentration) n = number of data values After calculating the probability of exceedance for each measured fecal coliform concentration using Equation 4.1., the measured fecal coliform concentrations were plotted on a log-scale versus the probability of exceedance. The measured data of all of the bacterial monitoring stations for each Copano Bay water segment (Segments 1, 2, 3, and 4) are grouped together (the monitoring station data that were applied to each segment is shown in Table 4.2) for the probability distribution plots. The measured data of the bacterial monitoring stations along the rivers are shown at each bacterial monitoring station. These probability plots are used to calibrate the Monte Carlo Simulation Model at each bacterial monitoring station location (in Chapter 7). 4.2.3 Result 4.2.3.1 Aransas River Above Tidal There is only one bacterial monitoring station on the Aransas River Above Tidal reach, but there is another bacterial monitoring station with fecal coliform monitoring data that is upstream of the Above Tidal reach and must also comply with contact recreation use standards. The latter station, Station 17592 (HydroID 61), was analyzed first because it is the most upstream station. The bacterial monitoring data from Station 17592 (from 1999-2004), the rank, probability of exceedance, and the geometric mean of the measured data are shown in Table 4.5. 53 Table 4.5 Bacterial Monitoring Data for Station 17592 (1999-2004) Date Fecal Coliform Concentration (CFU/100mL) Rank Probability of Exceedance, % 6/18/2001 5700 1 6.10 4/10/2001 1373 2 15.85 1/14/2002 500 3 25.61 4/9/2002 500 4 35.37 10/8/2001 270 5 45.12 7/11/2000 250 6 54.88 1/19/2000 147 7 64.63 1/15/2001 106 8 74.39 4/17/2000 76 9 84.15 10/25/1999 54 10 93.90 Geometric Mean (CFU/100mL) 311 > 200 Percent > 400 CFU/100mL (%) ~ 40 > 25 As shown in Table 4.5, both contact recreation use standards are exceeded at this bacterial monitoring station because the geometric mean is greater than 200 CFU/100mL, and more than 25% of the samples are greater than 400 CFU/100mL. However, this station is not along a TCEQ-defined water segment, and fecal coliform monitoring data have not been collected since April 2002. E. coli has recently been chosen as the bacterial indicator for the Aransas River Above Tidal; therefore, fecal coliform concentrations are no longer measured at this station. The probability distribution of the measured fecal coliform concentrations at Station 17592, which is a plot of the data given in Table 4.5, and the two fecal coliform contact recreation use standards are shown in Figure 4.6. 54 ") ") 125 79 61 Figure 4.6 Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Station 17592 1 10 100 1000 10000 0 10203040506070809010 Probability of Exceedance, % Fe ca l Coliform (C FU/1 00 mL) 200 CFU/100mL - Geometric Mean FC Criterion #1 Not in Compliance 400 CFU/100 mL - Single Sample Criterion #2 Not in Compliance 25% of Observed Data (allowed exceedance) 75 % of Observed Data Region not in compliance 55 Station 12952 (HydroID 68) is the bacterial monitoring station on the Aransas River Above Tidal (downstream of Station 17592.) The bacterial monitoring data that exist at Station 12952 (from 1999-2004), the rank, probability of exceedance, and the geometric mean of the measured data are shown in Table 4.6. Table 4.6 Bacterial Monitoring Data for Station 12952 (1999-2004) Date Fecal Coliform Concentration (CFU/100mL) Rank Probability of Exceedance, % 7/8/2002 836 1 11.90 4/22/2003 130 2 30.95 1/21/2003 72 3 50.00 8/18/2003 58 4 69.05 10/15/2002 25 5 88.10 Geometric Mean (CFU/100mL) 103 < 200 Percent > 400 CFU/100mL (%) ~23 < 25 Both contact recreation use standards are met at this bacterial monitoring station (as shown in Table 4.6). However, fecal coliform monitoring data have not been collected since August 2003 because E. coli is now the primary bacterial indicator for the Aransas River Above Tidal. The probability distribution of the measured fecal coliform concentrations at Station 12952, which is a plot of the data given in Table 4.6, is shown in Figure 4.7. The two fecal coliform contact recreation use standards are also shown in Figure 4.7. 56 ") ") 1 1 1 1 2 4 1 4 0 113 1 0 6 91 85 84 78 68 62 61 Figure 4.7 Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Station 12952 1 10 100 1000 0 10203040506070809010 Probability of Exceedance, % Fec a l C oli f orm (CF U/1 00 mL) 200 CFU/100mL - Geometric Mean FC Criterion #1 In Compliance 400 CFU/100 mL - Single Sample Criterion #2 In Compliance 25% of Observed Data (allowed exceedance) 75% of Observed Data 57 4.2.3.2 Aransas River Tidal Station 12948 (HydroID 75) is the only bacterial monitoring station on the Aransas River Tidal, and this segment must meet contact recreation use fecal coliform standards. The bacterial monitoring data that exists at Station 12948 (from 1999-2004), the rank, probability of exceedance, and the geometric mean of the measured data, are shown in Table 4.7. Table 4.7 Bacterial Monitoring Data for Station 12948 (1999-2004) Date Fecal Coliform Concentration (CFU/100mL) Rank Probability of Exceedance, % 4/17/2000 3700 1 3.85 7/8/2002 1327 2 10.00 1/15/2001 270 3 16.15 10/9/2000 162 4 22.31 6/18/2001 131 5 28.46 10/15/2002 122 6 34.62 7/11/2000 112 7 40.77 4/10/2001 98 8 46.92 1/14/2002 94 9 53.08 4/9/2002 94 10 59.23 1/21/2003 58 11 65.38 10/8/2001 48 12 71.54 4/22/2003 34 13 77.69 8/18/2003 28 14 83.85 1/19/2000 20 15 90.00 10/25/1999 12 16 96.15 Geometric Mean (CFU/100mL) 105 < 200 Percent > 400 CFU/100mL (%) ~15 < 25 As shown in Table 4.7, both contact recreation use standards are met at this bacterial monitoring station. However, fecal coliform monitoring data have not been collected since August 2003 because enterococcus is now the primary bacterial indicator for the 58 Aransas River Tidal. As discussed in Section 2.1, enterococcus is a better bacterial indicator in marine waters. The probability distribution of the measured fecal coliform concentrations at Station 12948, which is a plot of the data given in Table 4.7, is shown in Figure 4.8. The two fecal coliform contact recreation use standards are also shown in Figure 4.8. 59 ") 1 2 0 1 2 6 83 75 68 64 ") Figure 4.8 Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Station 12948 1 10 100 1000 10000 0 10203040506070809010 Probability of Exceedance, % Fec a l C oli f orm (CF U/1 00 mL) 200 CFU/100mL - Geometric Mean FC Criterion #1 In Compliance 400 CFU/100 mL - Single Sample Criterion #2 In Compliance 25% of Observed Data (allowed exceedance) 90 % f Ob d D t 75% of Observed Data 60 4.2.3.3 Mission River Above Tidal Station 12944 (HydroID 74) is the only bacterial monitoring station on the Mission River Above Tidal, and this segment must meet contact recreation use fecal coliform standards. The bacterial monitoring data that exists at Station 12944 (from 1999-2004), the rank, probability of exceedance, and the geometric mean of the measured data are shown in Table 4.8. Table 4.8 Bacterial Monitoring Data for Station 12944 (1999-2004) Date Fecal Coliform Concentration (CFU/100mL) Rank Probability of Exceedance, % 10/9/2000 1382 1 3.62 7/8/2002 682 2 9.42 1/19/2000 410 3 15.22 1/15/2001 410 4 21.01 4/10/2001 320 5 26.81 10/8/2001 157 6 32.61 1/21/2003 142 7 38.41 4/22/2003 120 8 44.20 5/12/2003 116 9 50.00 4/17/2000 112 10 55.80 7/11/2000 94 11 61.59 10/25/1999 58 12 67.39 6/18/2001 56 13 73.19 1/14/2002 54 14 78.99 4/9/2002 54 15 84.78 8/18/2003 54 16 90.58 10/15/2002 46 17 96.38 Geometric Mean (CFU/100mL) 142 < 200 Percent > 400 CFU/100mL (%) ~21 < 25 As shown in Table 4.8, both contact recreation use standards are met at this bacterial monitoring station. However, fecal coliform monitoring data have not been collected since August 2003 because E. coli is now the primary bacterial indicator for the Mission River Above Tidal. The probability distribution of the measured fecal coliform 61 concentrations at Station 12944, which is a plot of the data given in Table 4.8, is shown in Figure 4.9. The two fecal coliform contact recreation use standards are also shown in Figure 4.9. 62 ") ") 1 1 8 1 2 8 1 2 3 1 16 99 90 80 77 74 73 Figure 4.9 Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Station 12944 1 10 100 1000 10000 0 10203040506070809010 Probability of Exceedance, % Fe ca l Coliform (C FU/1 00 mL) 200 CFU/100mL - Geometric Mean FC Criterion #1 In Compliance 400 CFU/100 mL - Single Sample Criterion #2 In Compliance 25% of Observed Data (allowed exceedance) 75 % of Observed Data 63 4.2.3.4 Mission River Tidal Station 12943 (HydroID 70) is the only bacterial monitoring station on the Mission River Tidal, and this segment must meet contact recreation use fecal coliform standards. The bacterial monitoring data that exists at Station 12943 (from 1999-2004), the rank, probability of exceedance, and the geometric mean of the measured data are shown in Table 4.9. Table 4.9 Bacterial Monitoring Data for Station 12943 (1999-2004) Date Fecal Coliform Concentration (CFU/100mL) Rank Probability of Exceedance, % 1/15/2001 740 1 3.85 4/22/2003 270 2 10.00 1/21/2003 147 3 16.15 7/8/2002 130 4 22.31 8/18/2003 55 5 28.46 4/17/2000 52 6 34.62 1/14/2002 51 7 40.77 4/9/2002 51 8 46.92 6/18/2001 42 9 53.08 7/11/2000 41 10 59.23 4/10/2001 37 11 65.38 10/25/1999 32 12 71.54 1/19/2000 23 13 77.69 10/8/2001 22 14 83.85 10/15/2002 21 15 90.00 10/9/2000 3 16 96.15 Geometric Mean (CFU/100mL) 51 < 200 Percent > 400 CFU/100mL (%) ~10 < 25 As shown in Table 4.9, both contact recreation use standards are met at this bacterial monitoring station. However, fecal coliform monitoring data have not been collected since August 2003 because enterococcus is now the primary bacterial indicator for the Mission River Tidal. The probability distribution of the measured fecal coliform 64 concentrations at Station 12943, which is a plot of the data given in Table 4.9, is shown in Figure 4.10. The two fecal coliform contact recreation use standards are also shown in Figure 4.10. ") ") 1 15 1 3 3 1 0 9 1 3 0 97 74 70 65 100 Figure 4.10 Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Station 12943 1 10 100 1000 0 10203040506070809010 Probability of Exceedance, % F eca l C o l i f o r m ( C F U /1 00 m L ) 200 CFU/100mL - Geometric Mean FC Criterion #1 In Compliance 400 CFU/100 mL - Single Sample Criterion #2 In compliance 25% of Observed Data (allowed exceedance) 75 % of Observed Data 65 4.2.3.5 Copano Bay Bacterial monitoring data were analyzed at the four Copano Bay segments (Segments 1, 2, 3, and 4), which need to meet oyster water use fecal coliform standards. Fecal coliform remains the primary bacterial indicator in Copano Bay, so fecal coliform concentrations continue to be measured. Watershed JunctionID 45405 drains into Segment 1 (SchemaNode 155). Bacterial monitoring stations 13405, 14782, 14784, and 14790 all measure fecal coliform concentrations in Copano Bay Segment 1. The median and 90 th -percentile of the monitoring data (from 1999-2005) are shown in Table 4.10; all of the bacterial monitoring data that exists for Segment 1 (from 1999-2005) and the rank and the probability of exceedance for each measurement are given in Appendix 4.1. Table 4.10 Statistics of Bacterial Monitoring Data for Stations in Segment 1 (1999-2005) Median Fecal Coliform Concentration (CFU/100mL) 2 90 th Percentile Fecal Coliform Concentration (CFU/100mL) ~ 15 Number of Measurements 113 As shown in Table 4.10, both oyster water use standards are met in Copano Bay Segment 1. The probability distribution of the measured fecal coliform concentrations at these stations is shown in Figure 4.11. The two fecal coliform oyster water use standards are also shown in Figure 4.11. 66 Figure 4.11 Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Segment 1 ") " ) ") ") ") ") ") ") 1 5 9 1 3 5 93 88 155 1 10 100 1000 0 10203040506070809010 Probability of Exceedance, % Fec a l C oli f orm (CF U/1 00 mL) 14 CFU/100mL - Median FC Criterion #1 In Compliance 43 CFU/100 mL - 90th-Percentile Criterion #2 In Compliance 10% of Observed Data (allowed exceedance) 90 % of Observed Data 67 Aransas River and Chiltipin Creek drain into Segment 2 (SchemaNode 154). Bacterial monitoring stations 12945, 14783, 14787, and 14788 measure fecal coliform concentrations in Copano Bay Segment 2. The median and 90 th -percentile of the monitoring data (from 1999-2005) are shown in Table 4.11; all of the bacterial monitoring data that exists for Segment 2 (from 1999-2005) and the rank and the probability of exceedance for each measurement are given in Appendix 4.2. Table 4.11 Statistics of Bacterial Monitoring Data for Stations in Segment 2 (1999-2005) Median Fecal Coliform Concentration (CFU/100mL) 2 90 th Percentile Fecal Coliform Concentration (CFU/100mL) ~ 79 Number of Measurements 121 As shown in Table 4.11, Copano Bay Segment 2 complies with the median fecal coliform standard (< 14 CFU/100mL) but exceeds the 90 th -percentile fecal coliform standard of 43 CFU/100mL. The probability distribution of the measured fecal coliform concentrations at these stations is shown in Figure 4.12. The two fecal coliform oyster water use standards are also shown in Figure 4.12. 68 Figure 4.12 Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Segment 2 ") " ) ") ") ") ") ") ") 122 107 1 3 4 142 127 1 1 2 1 4 3 1 5 8 1 2 1 1 3 9 87 86 75 69 67 63 104 103 102 154 1 10 100 1000 10000 0 10203040506070809010 Probability of Exceedance, % Fe ca l Coliform (C FU/1 00 mL) 14 CFU/100mL - Median FC Criterion #1 In compliance 43 CFU/100 mL - 90th-percentile Criterion #2 Not in compliance 90% of Observed Data10% of Observed Data (allowed exceedance) Region not in compliance 69 The Mission River drains into Segment 3 (SchemaNode 153.) Bacterial monitoring station 14797 measures fecal coliform concentrations in Copano Bay Segment 3. The median and 90 th -percentile of the monitoring data (from 1999-2005) are shown in Table 4.12; the bacterial monitoring data that exists for Segment 3 (from 1999- 2005) and the rank and the probability of exceedance for each measurement are given in Appendix 4.3. Table 4.12 Statistics of Bacterial Monitoring Data for Stations in Segment 3 (1999-2005) Median Fecal Coliform Concentration (CFU/100mL) 2 90 th Percentile Fecal Coliform Concentration (CFU/100mL) > 49 Number of Measurements 31 As shown in Table 4.12, Copano Bay Segment 3 complies with the median fecal coliform standard (< 14 CFU/100mL) but exceeds the 90 th -percentile fecal coliform standard of 43 CFU/100mL). The probability distribution of the measured fecal coliform concentrations at these stations is shown in Figure 4.13. The two fecal coliform oyster water use standards are also shown in Figure 4.13. 70 Figure 4.13 Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Segment 3 ") ") 1 2 9 1 5 7 1 1 0 82 70 66 153 1 10 100 1000 0 10203040506070809010 Probability of Exceedance, % Fe ca l Coliform (C FU/1 00 mL) 14 CFU/100mL - Median FC Criterion #1 In Compliance 43 CFU/100 mL - 90th-Percentile Criterion #2 Not in Compliance 10% of Observed Data (allowed exceedance ) 90 % of Observed Data Region not in compliance 71 Copano Creek drains into Segment 4. Bacterial monitoring stations 13404, 14779, 14780, 14785, 14792, and 14793 measure fecal coliform concentrations in Copano Bay Segment 4. The median and 90 th -percentile of the monitoring data (from 1999-2005) are shown in Table 4.13; the bacterial monitoring data that exists for Segment 4 (from 1999-2005) and the rank and the probability of exceedance for each measurement are given in Appendix 4.4. Table 4.13 Statistics of Bacterial Monitoring Data for Stations in Segment 4 (1999-2005) Median Fecal Coliform Concentration (CFU/100mL) 2 90 th Percentile Fecal Coliform Concentration (CFU/100mL) ~ 13 Number of Measurements 232 As shown in Table 4.13, Copano Bay Segment 4 complies with both fecal coliform oyster water use standards. The probability distribution of the measured fecal coliform concentrations at these stations is shown in Figure 4.14. The two fecal coliform oyster water use standards are also shown in Figure 4.14. 72 Figure 4.14 Existing Fecal Coliform Concentration Measurements versus Probability of Exceedance: Segment 4 ") ") ") ") ") ") ") ") ") ") 1 3 1 1 3 8 1 3 7 1 6 1 1 6 0 1 3 6 1 4 4 98 96 95 94 89 81 72 156 105 1 10 100 1000 10000 0 10203040506070809010 Probability of Exceedance, % F eca l C o l i f o r m ( C F U /1 00 m L ) 14 CFU/100mL - Median FC Criterion #1 In Compliance 43 CFU/100 mL - 90th-Percentile Criterion #2 In Compliance 10% of Observed Data (allowed exceedance) 90 % of Observed Data 73 Chapter 5: Estimation of Loadings 5.1 ESTIMATION OF NON-POINT BACTERIAL LOADINGS FROM WATERSHEDS 5.1.1 Methodology The non-point bacterial loadings of fecal coliform flow into Copano Bay from adjacent watersheds directly into the Bay or from upstream watersheds into rivers/streams/channels that flow into Copano Bay. The Bacterial Loadings Model calculates the non-point bacterial loadings for each such watershed and models bacterial concentrations as the bacteria flow from the upstream watersheds to rivers/streams/channels to Copano Bay. Non-point bacterial loadings are calculated as the product of runoff from each of the watersheds and Event Mean Concentration (EMC) of the corresponding land use land cover classifications within each watershed. The bacteria from the non-point sources (as well as the point sources) were decayed using the Schematic Processor (described in Chapter 6 of this report) as they travel from the watershed into rivers/channels and then into Copano Bay. The Bay was assumed to be completely mixed and acts as four Continuous Flow, Stirred Tank Reactors (CFSTRs) 4 , and the inflow into each of the Bay segments equals the outflow. The following steps were used to calculate the non-point bacterial loadings for each watershed: 1. Delineate watersheds to the Critical Points (USGS gauge stations, bacterial monitoring stations, and water segment endpoints) using the Digital Elevation Model (DEM), Arc Hydro’s Terrain Preprocessing on the DEM, the National Hydrography Dataset (NHD), and Water Rights 4 See Section 6.3.1.1 for how and why Copano Bay was segmented. 74 Analysis Package (WRAP) Hydro, which is a toolbar located in Arc GIS that is used to delineate watersheds for the basin. 2. Collect mean annual precipitation data from PRISM in grid format and create a runoff grid using mathematical relationships between rainfall- runoff based on different land use characteristics. 3. Obtain the land use land cover dataset from USGS (in raster format) and convert it into an EMC grid based on the EMC associated with different land use classifications. 4. Multiply the runoff grid by the EMC grid to obtain the bacterial loading per grid cell in the watersheds. 5. Using the delineated watersheds and Spatial Analyst’s Zonal Statistics, calculate the cumulative non-point bacterial loadings per watershed. 5.1.2 Procedure of Application 5.1.2.1 Watershed Delineation Before conducting the runoff and concentration calculations and delineating the watersheds, Terrain Preprocessing (found in the Arc Hydro toolbar) was implemented on the DEM to determine the flow patterns in the basin. For this project (and in order to use WRAP Hydro), the only steps that were implemented from Terrain Preprocessing were DEM Reconditioning, Fill Sinks, and Flow Direction. The step-by-step process used to conduct Terrain Preprocessing is given in Appendix 5.1. After completing Terrain Preprocessing, WRAP Hydro was used to delineate watersheds to the Critical Points (USGS gauge stations, bacterial monitoring stations, and water segment endpoints) for the basin (procedure described in Appendix 5.2.) The delineated watersheds for the basin are shown in Figure 5.1. The Critical Points are 75 points at which the modeled fecal coliform loadings/concentrations need to be observed, analyzed, and/or compared to existing bacterial monitoring data. !( !( !( !( !( !( !( !( !( !( !( !( !( !( !( !( !(! !( !( !( !( !(!( !( !( !( !( !( !( !( !( Legend !( Bacteria Monitoring Stations !( USGS Gauge Stations !( WaterSegmentPoints NHDFlowline Watersheds Error! No text of specified style in document..1 Watershed Delineation Figure 5.1 Watershed Delineation 76 5.1.2.2 Precipitation Data Preparation The precipitation data were obtained from PRISM in polygon feature class format, which is shown in Figure 5.2. Using the “Feature to Raster” tool in Arc Toolbox, the polygon feature class was converted to a raster based on the field, “RANGE”, which is the annual precipitation in inches. The annual precipitation was then converted to millimeters by using Spatial Analyst's Raster Calculator: [Precipitation in inches/year] * (25.4 mm/inch) = [Precipitation in mm/year] = P, where [] represents a raster. Rainfall-runoff relationships exist for four different land use categories to calculate runoff (Section 5.1.2.4). Thus, the precipitation grid was divided into four rasters based on the land use categories. This procedure is further described in Section 5.1.2.3 and Appendix 5.3. 5.1.2.3 Land Use Land Cover Data Preparation The 1992 National Land Cover Dataset (NLCD), which comes in raster format, was converted to a polygon feature class using the “Raster to Polygon” tool in Arc Toolbox. There are rainfall-runoff relationships for four different land use categories to calculate runoff: “Agricultural Land”; “Rangeland, Forest, Barren, Other”; “Urban Land”, and “Open Water”; (see Section 5.1.2.4 of this report for the equations.) Because Legend AnnualPrecipitation 29 31 33 35 37 39 Figure 5.2 Precipitation Data (inches/year) 77 the land use land cover dataset for Copano Bay has 18 different land use classifications, these classifications were grouped into four redefined land use categories that were used in the rainfall-runoff equations. The land use land cover classifications were reclassified into the corresponding four land use categories for this project (shown in Table 5.1). Table 5.1 Reclassified Land Use Categories Land Use Code (Gridcode) Land Use Category Reclassified Land Use Category 61 Orchards/Vineyards/Other 81 Pasture/Hay 82 Row Crops 83 Small Grains Agricultural Land 31 Bare Rock/Sand/Clay 32 Quarries/Strip Mines/Gravel Pits 41 Deciduous Forest 42 Evergreen Forest 43 Mixed Forest 51 Shrubland 71 Grasslands/Herbaceous 91 Woody Wetlands 92 Emergent Herbaceous Wetlands Rangeland, Forest, Barren, Other 21 Low Intensity Residential 22 High Intensity Residential 23 Commercial/Industrial/Transportation 85 Urban/Recreational Grasses Urban Land 11 Open Water Open Water Thus, to calculate the runoff for each land use classification, the precipitation grid was divided into four different rasters based on these redefined land use classifications. The procedure on how to create precipitation rasters for different land use classifications is given in Appendix 5.3. 78 5.1.2.4 Rainfall-Runoff Relationships for Different Land Uses Runoff calculations were made by using empirical equations from Quenzer (1997). These equations, shown below, relate runoff to precipitation and land use. Agricultural Land: Q = 0.008312 * exp ( 0.011415 * P ) (5.1) Rangeland, Forest, Barren, Other: Q = 0.0053 * exp ( 0.010993 * P ) (5.2) Urban Land: Q = 0.24 * P (5.3) Open Water: Q = 0 (5.4) Where: Q = Runoff (mm/year) P = Precipitation (mm/year) – from PRISM These equations were used to calculate runoff in the watersheds (see Section 5.1.2.5). 5.1.2.5 Developing Runoff Grid After precipitation rasters, P, were created for each land use classification (see Section 5.1.2.3), Spatial Analyst’s Raster Calculator was used to calculate the runoff for each land use. An example calculation (Runoff raster for Agriculture) is shown in Figure 5.3: 79 Once the four Runoff rasters were created for each land use, the “Mosaic” tool in Arc Toolbox was used to combine all four rasters into a single Runoff raster, which is shown in Figure 5.4. Q Agriculture = 0.008312 * exp (0.011415 * ) = P Q Figure 5.3 Calculation of Agricultural Runoff Grid 80 Once the total Runoff raster was created (Q total in mm/year), the Raster Calculator was used to convert the runoff into m 3 /year. Because the raster contains 30m by 30m grid cells, a conversion factor of 0.9 was used. [Runoff in mm/year] * 0.9 = [Runoff in m 3 /year]. The final Runoff grid for the Copano Bay watershed is shown in Figure 5.5. Q Agriculture Q Forest Q Urban Q Water Mosaic Q Total (mm/year) Figure 5.4 Creation of Runoff Grid 81 5.1.2.6 Estimation of Flow from each Watershed Using Zonal Statistics and the delineated watersheds (Figure 5.1), the cumulative runoff (summation of runoff grid cells in each watershed) was calculated for each watershed, and the results are shown in Figure 5.6. Legend Runoff_Quenzer_m3yr 15.67449093 27.39826202 33.54451752 47.89080811 59.90475845 83.71076965 106.9795532 146.3224335 159.1055908 170.0784149 181.0511932 191.0469666 192.0240021 202.9967957 213.9695892 255.764389 341.1768799 609.2828979 Figure 5.5 Runoff Grid (m 3 /year) 82 5.1.2.7 Developing EMC Grid The EMC (fecal coliform concentration) raster was created by using the relationship between fecal coliform concentrations and land use found in Section 3.2.1.2, Table 3.1. Once the EMC table (Table 3.1) was joined to the land use land cover feature class based on the land use codes (found in both the EMC Table and the Land Use Land Cover Polygon Feature Class), a raster was created based on the EMC field (now in the land use land cover feature class) using the “Feature to Raster” tool in Arc Toolbox. Raster Calculator was then used to convert CFU/100mL to CFU/m 3 . [CFU/100mL] * Legend Runoff_m3yr_Q 3.50e+005 - 7.31e+006 7.32e+006 - 1.95e+007 1.96e+007 - 3.79e+007 3.80e+007 - 7.70e+007 7.71e+007 - 1.40e+008 Figure 5.6 Runoff per Watershed (m 3 /year) 83 10,000 = [CFU/m 3 ]. The EMC grid for the Copano Bay watershed is shown in Figure 5.7. 5.1.2.8 Estimation of Non-Point Bacterial Loading Once the Runoff raster, Q, and the EMC raster, C, were created following the procedures given in Sections 5.1.2.5 and 5.1.2.7, respectively, the bacterial load per grid cell was calculated by using Spatial Analyst’s Raster Calculator and the following equation: L = Q * C. This calculation is shown in Figure 5.8. C (CFU/m 3 ) Q (m 3 /year) L (CFU/year) *= Figure 5.8 Creation of Bacterial Loading Grid (CFU/year) Legend 0 2,000,000 10,000,000 25,000,000 220,000,000 Figure 5.7 EMC Grid (CFU/m 3 ) 84 Using Zonal Statistics and the delineated watersheds (Figure 5.1), the cumulative bacterial loadings were calculated for each watershed. The bacterial loading per watershed is shown in Figure 5.9. 5.1.3 Result After completing the procedure described in Section 5.1.2, the cumulative non- point source bacterial loadings per watershed were calculated. The bacterial loading per watershed is shown in Figure 5.9. Legend BacteriaLoadings_Q 2.89e+012 - 8.61e+013 8.62e+013 - 2.19e+014 2.20e+014 - 4.59e+014 4.60e+014 - 1.08e+015 1.09e+015 - 1.77e+015 Figure 5.9 Non-Point Bacterial Loading per Watershed (CFU/year) 85 5.2 ESTIMATION OF LIVESTOCK LOADING 5.2.1 Methodology Fecal coliform loadings (i.e., bacterial loadings) from livestock were not accounted for in the non-point bacterial loading calculations because the EMC values were determined from a Galveston Bay study and not for Copano Bay. Thus, we determined that using Census data (see Section 3.2.1.3) for livestock per county would be a more accurate way to estimate annual fecal waste from livestock animals in each watershed. The fecal waste of the following seven animal species were accounted for in the bacterial loading model: cattle, horses, hogs, sheep, hens, goats, and chickens. The annual bacterial loadings per watershed from livestock were calculated by finding the annual number of each livestock species per watershed on the following types of land: Shrubland (land use code 51), Grasslands/Herbaceous (land use code 71), and Pasture/Hay (land use code 81) 5 and multiplying the livestock counts by the amount of fecal waste produced per year per species (CFU/year-animal). The following steps were used to calculate the livestock bacterial loadings for each watershed: 1. Determine the annual livestock count of each species per county from the 2002 Census of Agriculture (NASS) and 2004 Texas Livestock Inventory and Production (USDA, NASS, Texas Statistical Office.) 2. Calculate the area (m 2 ) of the land use classifications of 51, 71, and 81 in each county in the Copano Bay watershed. 5 These are the land use classifications that have an EMC value of zero for the non-point bacterial loading calculations (Table 3.1). 86 3. Find the density of each animal per county (count/m 2 of Shrubland/Grasslands/Herbaceous/Pasture/Hay ). The following equation would be used where the number in the parentheses indicates the step in which the value was determined: (1)/(2) 4. Calculate the area of land use classifications 51, 71, and 81 of each county within each watershed (watersheds may have multiple counties.) 5. Multiply the area (m 2 ) of each county within each watershed by the animal density (count/m 2 ) to find the livestock count of each species that each county has in each watershed. (4)*(3) 6. Sum the livestock count of each type of species in each watershed to obtain the total number of each species per watershed. 7. Multiply the count of each species in each watershed by the fecal coliform typically produced each year (CFU/year-animal) that is found from the literature. 8. Sum the CFU/year for each species to get a cumulative CFU/year per watershed. 5.2.2 Procedure of Application 5.2.2.1 Finding Livestock per County Livestock data (annual count per county) were obtained from the 2002 Census of Agriculture, NASS, and the 2004 Texas Livestock Inventory and Production, United States Department of Agriculture (USDA), NASS, Texas Statistical Office. The animals that were considered in the calculations were cattle, goats, horses, sheep, hens, hogs, and chickens. The livestock data that were used for the point source calculations is given in Table 5.2. 87 Table 5.2 Livestock Count per County County Livestock 2002 Data 2004 Data Cattle 2,878 2,000 Goats 75 Unavailable Horses 46 Unavailable Sheep 0 Unavailable Hens 0 Unavailable Hogs 0 Unavailable Aransas Chickens 0 Unavailable Cattle 49,950 49,000 Goats 2344 2100 Horses 1391 Unavailable Sheep 670 Unavailable Hens 793 Unavailable Hogs 0 Unavailable Bee Chickens 0 Unavailable Cattle 63,398 66,000 Goats 795 Unavailable Horses 887 Unavailable Sheep 0 Unavailable Hens 859 Unavailable Hogs 0 Unavailable Goliad Chickens 252 Unavailable Cattle 74,623 74,000 Goats 2288 2100 Horses 973 Unavailable Sheep 327 Unavailable Hens 0 Unavailable Hogs 0 Unavailable Karnes Chickens 0 Unavailable Cattle 41,239 36,000 Goats 200 Unavailable Horses 692 Unavailable Sheep 71 Unavailable Hens 63 Unavailable Hogs 0 Unavailable Refugio Chickens 0 Unavailable Cattle 22,253 20,000 Goats 773 Unavailable Horses 662 Unavailable Sheep 0 Unavailable Hens 464 Unavailable Hogs 741 Unavailable San Patricio Chickens 0 Unavailable 88 5.2.2.2 Calculating Density of Livestock per County The density of livestock per county (acres/animal) was calculated for each animal by using the following equation: [Area in acres where the animals would be located within county] / [Total annual count of each animal]. The area where animals would be located was assumed to be from the land use land cover classifications 51 (Shrubland), 71 (Grasslands/Herbaceous), and 81 (Pasture/Hay). To find the area, the land use land cover dataset was masked by each county, and the corresponding grid cells (for land use codes 51, 71, and 81) were summed. For example, in San Patricio County (calculation shown in Figure 5.10), the total area where animals are located is 472,358,700 m 2 = 116,480 acres. Thus, the density of cattle in San Patricio county is 22,253 cattle/472,358,700 m 2 = 0.0000471 cattle/m 2 = 122 cattle/mi 2 = 5 acres per cow. The density of each livestock animal in each county is given in Table 5.3. 89 Table 5.3 Animal Density per County (Acres per Animal) County Area (acres) Livestock 2002 Density 2004 Density Cattle 18 26 Goats 640 Unavailable Horses 640 Unavailable Sheep 0 Unavailable Hen 0 Unavailable Hogs 0 Unavailable Aransas 51,200 Chickens 0 Unavailable Cattle 7 7 Goats 160 160 Horses 213 Unavailable Sheep 640 Unavailable Hen 640 Unavailable Hogs 0 Unavailable Bee 341,760 Chickens 0 Unavailable Cattle 6 6 Goats 640 Unavailable Horses 320 Unavailable Sheep 0 Unavailable Hen 320 Unavailable Hogs 0 Unavailable Goliad 361,600 Chickens 0 Unavailable Cattle 4 4 Goats 32 160 Horses 320 Unavailable Sheep 640 Unavailable Hen 0 Unavailable Hogs 0 Unavailable Karnes 318,080 Chickens 0 Unavailable Cattle 7 8 Goats 0 Unavailable Horses 320 Unavailable Sheep 0 Unavailable Hen 0 Unavailable Hogs 0 Unavailable Refugio 282,240 Chickens 0 Unavailable Cattle 5 6 Goats 160 Unavailable Horses 160 Unavailable Sheep 0 Unavailable Hen 213 Unavailable Hogs 160 Unavailable San Patricio 116,480 Chickens 0 Unavailable 90 5.2.2.3 Calculating Livestock Count per Watershed The area (mi 2 ) of each county within each delineated watershed (Figure 5.1) was determined and then multiplied by each livestock's density in each corresponding county to find each livestock count. Livestock count = Area (mi 2 ) * Density (Count/mi 2 ) All the calculations for this procedure are given in Appendix 5.4. For example, Watershed JunctionID 45422 has two counties overlapping it, so there are two different areas and cattle densities to account for in the calculation. The cattle calculation for Watershed JunctionID 45422 is shown in Figure 5.11. 149 cattle/mi 2 is the cattle density in Karnes County, and 92 cattle/mi 2 6 is the density of cattle in Bee County. Approximately 17.6 mi 2 is the area in Watershed JunctionID that is a part of Karnes 6 The densities are per square mile of land where animals would be located based on land use classifications. San Patricio County Land Use Land Cover Raster (30m)(30m) * 524,843 grid cells = 472,358,700 m 2 = 116,480 acres Figure 5.10 Determination of Area (Acres) of Animals in San Patricio County 91 County (the area of land use types 51, 71, and 81 where animals would be located). Approximately 110.6 mi 2 is the area in Watershed JunctionID that is a part of Bee County. Thus, there are approximately 12,778 cattle in Watershed JunctionID 45422. This procedure was performed for all livestock species in each watershed. Figure 5.11 Determination of Cattle Count in Watershed JunctionID 45422 7 5.2.2.4 Calculating Livestock Bacterial Loading (CFU/year) per Watershed After determining the count of each animal within each watershed (see Section 5.2.2.3), the count was multiplied by the fecal coliform produced annually (CFU/year) by each animal. The CFU/year produced by each animal considered in this model is shown in Table 5.4. Information regarding fecal coliform production by hens and goats was not found in the literature, so estimations were made from similar animals. Since goats generally have a similar body mass to sheep, the production of fecal coliform was 7 The densities and areas were rounded, so the exact total cattle count is accurate, but it may not agree exactly when carrying out the multiplication. 14,366 + Bee County Karnes County 149 cattle/mi 2 * 17.6 mi 2 = 2,625 cattle + 92 cattle/mi 2 * 110.6 mi 2 = 10,152 cattle Total = 12,778 cattle 92 assumed to be the same. The production of fecal coliform by hens was calculated by using a mass ratio based on chickens (0.66 hen:chicken mass ratio), with the assumption fecal coliform production is proportional to body mass. Table 5.4 Annual Fecal Coliform Production from Livestock Animals (EPA, 2005) Livestock CFU/year Reference Sheep 1.10 x 10 13 Metcalf and Eddy, 1991 ASAE, 1998 Goat 1.10 x 10 13 (Assumed same as sheep) Hog 3.63 x 10 12 Metcalf and Eddy, 1991 ASAE, 1998 Cattle 1.97 x 10 12 Metcalf and Eddy, 1991 Horse 1.53 x 10 11 ASAE, 1998 Chicken 1.39 x 10 11 Metcalf and Eddy, 1991 ASAE, 1998 Hen 4.61 x 10 10 Calculated from fecal coliform production of chicken (CFU/year) multiplied by hen:chicken body mass ratio For example, in Watershed JunctionID 45422: 12,778 cattle * (1.97 x 10 12 CFU/year- head of cattle) = 2.52 x 10 16 CFU/year from cattle. The CFU/year needs to be summed for all species within each watershed to find the total CFU/year excreted from livestock species, which is shown below in Figure 5.12. 5.2.3 Result After completing the procedure described in Section 5.2.2, the cumulative livestock bacterial loadings per watershed were calculated. The livestock bacterial loading per watershed is shown in Figure 5.12. 93 Comparing Figure 5.12 (livestock bacterial loading) to Figure 5.9 (non-point bacterial loading from different land use types excluding livestock), livestock bacterial loadings are orders of magnitude greater than non-point bacterial loadings. Figure 5.13 shows the percent distribution of bacterial loadings from each livestock species in the upstream watersheds. Legend Animal_cfu_year 0.00e+000 - 2.31e+014 2.32e+014 - 1.53e+015 1.54e+015 - 6.74e+015 6.75e+015 - 1.17e+016 1.18e+016 - 6.17e+016 Figure 5.12 Livestock Bacterial Loading (CFU/year) per Watershed 94 As shown in Figure 5.13, cattle are the major livestock contributor to bacterial loading based on the model assumptions and calculations. Summing the total counts of each of the livestock species for the entire Copano Bay watershed, there are 111,433 cattle followed by 2,561 horses. Thus, there are significantly more cattle and cattle bacterial loadings than from any other species. Cattle Goats Horses Sheep Layers Hogs Chickens Legend Figure 5.13 Percent Distribution of Bacterial Loadings from Livestock Species at Watersheds 95 5.3 ESTIMATION OF AVIAN LOADING 5.3.1 Methodology Fecal coliform loadings (i.e., bacterial loadings) from colonial waterbirds were determined by obtaining data from the Texas Colonial Waterbird Census (TCWC). There are approximately 30 different types of colonial waterbird species along the Texas coastline, and the TCWC gives the number of breeding pairs of different waterbird species from 1973-2003. The total fecal coliform from waterbirds was calculated and applied to the Copano Bay water segments by finding the average count of each waterbird species from 1973-2003, the annual fecal coliform production by each type of bird, and the approximate percentage of load reaching the Bay. The loading was calculated based on the following equation: Avian Loading (CFU/year) = [Number of Breeding Pairs] x [2 Birds per Breeding Pair] x [Amount of Excretion per Bird (g/bird)] x [Fecal Coliform Concentration in Excretion (CFU/g)] x [Percent of Fecal Coliform that Reaches Copano Bay] The following waterbird species were included in this model: Laughing Gull, Tricolored Heron, Black Skimmer, Neotropic Cormorant, Least Tern, Great Blue Heron, Great Egret, Snowy Egret, Roseate Spoonbill, Cattle Egret, Reddish Egret, American Oystercatcher, Fulvous Whistling Duck, Forster’s Tern, and Little Blue Heron. The following steps were used to calculate the avian bacterial loadings for each Copano Bay water segment or watershed in which waterbird colonies are present: 1. Determine the average count for each species from 1973-2003 at each location from TCWC. 96 2. Find the daily excretion (g/bird) for each species in literature. 3. Find the fecal count per excretion (CFU/g) in literature. 4. Determine the percent of bacterial loading that reaches the bay, based on the amount of time that each waterbird species spends on bay year-round. 5. Multiply CFU/bird by the number of species for each location (water segment or watershed) and sum all species’ CFU/year. This yields the total annual CFU/year contributed by colonial waterbirds to each segment/watershed. 5.3.2 Procedure of Application 5.3.2.1 Determining the Average Count of Waterbird Species Waterbird data were obtained from the TCWC. An annual count of each type of waterbird breeding pair was tabulated by volunteers from State, Federal, Non-Profit Organizations, and Professional Organizations for each year from 1973-2003. For each type of waterbird species, an average annual count was found by averaging all the counts from 1973-2003. The average waterbird count at each location is given in Table 5.5. However, these counts do not include the number of waterbirds that are not breeding pairs. 97 Table 5.5 Average Waterbird Count (1973-2003) Waterbird Species Average Breeding Pair Count Number of Locations Laughing Gull 367 1 Tricolored Heron 158 1 Cattle Egret 87 2 Neotropic Cormorant 84 3 Black Skimmer 59 1 Great Blue Heron 54 3 Least Tern 45 4 Snowy Egret 14 2 Great Egret 13 1 Roseate Spoonbill 9 1 American Oystercatcher 3 1 Fulvous Whistling Duck 2 1 Forster’s Tern 1 1 Little Blue Heron 1 1 There are eight waterbird colony locations surrounding Copano Bay. The bacterial loadings produced from these colonies are applied to either a Copano Bay segment or watershed. The locations of the breeding pairs on the Copano Bay watershed and on the portion of the model to which the bacterial loadings will be applied are shown in Figure 5.14. 98 The colony codes, which correspond to each breeding pair location, and the Copano Bay segment or watershed that the bacterial loadings will be applied to are given in Table 5.6. Table 5.6 Colony Codes and Watersheds/Segments to which Loads are Applied Copano Bay Segment/Watershed Waterbird Colony Code 609-460 609-461 1 609-480 2 614-295 3 614-296 4 609-380 614-293 JunctionID 45405 614-294 Breeding Pair Locations # # # # # # # ") ") ") ") ") ") ") Segment 2 Segment 3 Segment 4 Locations of Applied Avian Loads Segment 1 Watershed JunctionID 45405 Figure 5.14 Locations of Breeding Pairs and Applied Loads 99 Table 5.7 Number of Waterbird Species Applied to each Segment/Watershed Copano Bay Watershed JunctionID Species 1 2 3 4 45405 Laughing Gull 367 Tricolored Heron 158 Black Skimmer 59 Neotropic Cormorant 44 40 Least Tern 23 6 4 12 Great Blue Heron 14 40 22 Great Egret 13 Snowy Egret 11 3 Roseate Spoonbill 9 Cattle Egret 7 80 Reddish Egret 5 American Oystercatcher 3 Fulvous Whistling Duck 2 Forster's Tern 1 Little Blue Heron 0 1 5.3.2.2 Determining Excretion (g/bird) from Waterbirds The fecal mass produced by each type of bird was found based on the excretion of the Adult Herring Gull that was determined by Reem Zoun in her thesis, Estimation of Fecal Coliform Loadings to Galveston Bay (Zoun, 2003). She found from literature that the daily fecal mass of an Adult Herring Gull is 15 g (dry weight)/bird. The fecal mass (g/bird) for the other types of waterbirds was calculated based on the different body masses of each type of bird compared to the Adult Herring Gull. For example, the Adult Herring Gull has a mass of approximately 1225 g (Percevia, 2005a) and the Laughing Gull has a mass of approximately 325 g (USGS, 2005a); thus, assuming a constant ratio between fecal mass and bird body mass, Laughing Gull excretion (g/bird) = {15 g fecal mass/Adult Herring Gull * 325 g/Laughing Gull} / {1225 g/Adult Herring Gull} = 3.98 g fecal mass per Laughing Gull. 100 The body mass (g) and the calculated daily fecal mass (g/bird) for each type of bird is given in Table 5.8. Table 5.8 Estimated Daily Fecal Mass (g/bird) Waterbird Species Body Mass (g/bird) Fecal Mass (g/bird) Laughing Gull 325 (USGS, 2005a) 3.98 Tricolored Heron 374.5 (USGS, 2005b) 4.59 Cattle Egret 337 (Percevia, 2005b) 4.13 Neotropic Cormorant 1270 (Gil de Weir, 2005) 15.55 Black Skimmer 301.5 (USGS, 2005c) 3.69 Great Blue Heron 2400 (USGS, 2005d) 29.39 Least Tern 28 (CDEP, 2005a) 0.35 Snowy Egret 371 (USGS, 2005e) 4.54 Great Egret 1021 (CDEP, 2005b) 12.50 Roseate Spoonbill 1497 (Percevia, 2005c) 18.33 Reddish Egret 451 (Percevia, 2005d) 5.52 American Oystercatcher 602.5 (USGS, 2005f) 7.38 Fulvous Whistling Duck 670 (USGS, 2005g) 8.20 Forster's Tern 160 1.96 Little Blue Heron 366 (Percevia, 2002) 4.48 Adult Herring Gull 1225 (Percevia, 2002) 15 5.3.2.3 Estimation of Loadings (CFU/bird) The bacterial loadings from waterbirds was calculated by accounting for the number of breeding pairs at each location, the fecal mass produced per bird, the fecal coliform concentration in the fecal material, and the percent of fecal coliform loading that discharged to the Bay (based on how much time each type of bird spends on Copano Bay annually.) Based on the data of Zoun (2003), the fecal coliform concentration of avian excrement was estimated to be 10 8 CFU/g of fecal material for the fecal coliform loading calculations. Furthermore, the waterbirds spend an estimate of 50% of their time on the Bay, so 50% of the total fecal coliform loading from the waterbirds is assumed to reach 101 Copano Bay; this is a conservative estimate but would account for the fecal coliform loadings from some of the non-breeding pairs of waterbirds. The estimated fecal coliform loading for each type of waterbird is given in Table 5.9. Table 5.9 Annual Fecal Coliform Loading per Bird Bacterial Loading Reaching Bay Waterbird Species CFU/bird TCFU/bird Laughing Gull 1.99E+08 0.000199 Tricolored Heron 2.29E+08 0.000229 Cattle Egret 2.06E+08 0.000206 Neotropic Cormorant 7.78E+08 0.000778 Black Skimmer 1.85E+08 0.000185 Great Blue Heron 1.47E+09 0.000147 Least Tern 1.74E+07 0.000174 Snowy Egret 2.27E+08 0.000227 Great Egret 6.25E+08 0.000625 Roseate Spoonbill 9.17E+08 0.000917 Reddish Egret 2.76E+08 0.000276 American Oystercatcher 3.69E+08 0.000369 Fulvous Whistling Duck 4.10E+08 0.000410 Forster's Tern 9.80E+07 0.000980 Little Blue Heron 2.24E+08 0.000224 To find the CFU/year from waterbirds for each water segment and/or watershed, the values from Table 5.9 were multiplied by the number of each corresponding bird species over each water segment or watershed (given in Table 5.7). 5.3.3 Result The total CFU/year excreted by the breeding pairs of different waterbirds for each water segment and watershed is tabulated in Table 5.10 and is shown in Figure 5.15. 102 Table 5.10 Annual Fecal Coliform Avian Loadings Bacterial Loading Segment/Watershed CFU/yr TCFU/yr 1 3.96E+11 0.39600 2 2.22E+09 0.00222 3 1.48E+09 0.00148 4 1.22E+11 0.12200 JunctionID 45405 2.75E+11 0.27500 As shown in Table 5.10 and Figure 5.15, the bacterial loadings from waterbirds are significantly smaller than non-point (Figure 5.9) and livestock bacterial loadings (Figure 5.12). However, these loadings are applied directly to Copano Bay, so there is no travel time for bacterial decay. The effects of this direct loading (compared to the upstream loadings) are analyzed and evaluated in Chapter 6, where the discussion of modeling bacterial transport is discussed. Legend Segment1 Segment2 Segment3 Segment4 Watersheds 45405 Birds_cfu_year 0.00e+000 1.00e-002 - 1.48e+009 1.49e+009 - 2.22e+009 2.23e+009 - 1.22e+011 1.23e+011 - 3.96e+011 Figure 5.15 Avian Loadings (CFU/yr) on Copano Bay Water Segments and Watersheds 103 5.4 ESTIMATION OF WASTEWATER TREATMENT PLANT (WWTP) LOADINGS 5.4.1 Methodology Fecal coliform loadings (i.e., bacterial loadings) from wastewater treatment plants (WWTPs) were calculated based on discharge monitoring data obtained from the Permit Compliance System (PCS) Database from the U.S. Environmental Protection Agency (EPA). The locations of the WWTPs and their corresponding permit numbers are shown in Figure 5.16. WWTPs are required to disinfect their water (with chlorine, ozone, UV radiation, etc.) and meet Texas Surface Water Quality Standards before discharging into the receiving water bodies. However, fecal coliform bacteria are not one of the water quality characteristics that are monitored regularly because it does not require a water quality permit. Looking at the discharge monitoring reports (DMRs) of the permitted facilities in the Copano Bay watershed, some of the WWTPs have no fecal coliform monitoring data while some WWTPs only have one annual measurement (that may or may not meet water quality standards.) Thus, if fecal coliform monitoring data exist for a facility, the maximum fecal coliform concentration is used for the bacterial loading calculations (by multiplying by the average flow rate from the monitoring reports). If fecal coliform data do not exist for a facility, then fecal coliform counts from the literature were used. The following steps were used to calculate the WWTP bacterial loadings for the watershed model: 1. Calculate the average fecal coliform concentration (CFU/100mL) for each WWTP facility in the Copano Bay watershed (either from monitoring data or literature values). 104 2. Calculate the average flow rate (m 3 /year) from monitoring data for each WWTP facility in Copano Bay watershed. 3. Find the annual bacterial loading (CFU/year) for each WWTP by multiplying fecal coliform concentration (CFU/100mL) by average flow rate (m 3 /year) and 10,000 (factor for converting CFU/100mL to CFU/m 3 ). The following equation would be used where the number in the parentheses indicates the step in which the value was determined: (1) * (2) * 10,000 4. Derive relationship that calculates residence time based on mean flow length in the watershed to calculate residence time from WWTPs to mainstreams 8 . 8 Ernest To of the Center for Research in Water Resources (CRWR) derived this relationship for this project. 105 5.4.2 Procedure of Application 5.4.2.1 Determining the Fecal Coliform Concentration (CFU/100mL) from WWTPs Fecal coliform concentrations from WWTPs were determined from DMRs obtained from the PCS Database from the EPA. If fecal coliform monitoring data exist, then the maximum fecal coliform concentration was used. The fecal coliform concentrations used to calculate the average annual bacterial loading from WWTPs are given in Table 5.11. # # # # # # # # # 13892-001 10748-001 10705-001 10255-001 10237-001 10156-001 10124-004 10124-002 10055-001 Figure 5.16 WWTP Locations and Permit Numbers 106 If fecal coliform monitoring data do not exist, then the literature value of 84,000 CFU/100mL (Khan and Kamal, 2001) was used. This value is the fecal coliform count at a wastewater treatment plant discharge. Table 5.11 Fecal Coliform Concentrations of WWTPs Applied to Model 9 Permit Number Facility Description Maximum Fecal Coliform Concentration (CFU/100mL) 10124-002 City of Beeville, Moore Street WWTP 144,819 10156-001 Woodsboro WW Treatment Facility 126,388 10237-001 City of Odem WWTP 10 10255-001 Town of Refugio WW Treatment Facility 560 10705-001 City of Taft, Baird WWTP < 1 10055-001 City of Sinton Main WWTP 84,000 10124-004 Chase Field WWTP 84,000 10748-001 Pettus Municipal Utility District WWTP 84,000 13892-001 Water Reclamation Facility 84,000 9 See Chapter 9 for details of the WWTP loading overestimation issue. 107 It was recently discovered during the finishing of this report that WWTP fecal coliform concentrations of treated effluent are reported on permit renewal files, which are only available in hard copy format; however, the fecal coliform concentration of treated effluent is not reported on the DMRs. The fecal coliform concentrations that are presented in Table 5.11 are actually sludge concentrations in CFU/g of total solids, so the fecal coliform concentrations that were used in the model for WWTPs are much larger than the actual fecal coliform concentrations of treated effluent reported on the renewal permit files. However, the WWTP bacterial loading results (in subsequent chapters) are based on the concentrations that are presented in Table 5.11. See Chapter 9 for details on how to re-adjust the WWTP loading for future work. 5.4.2.2 Determination of Average Flow (m 3 /yr) from WWTPs Average flows from the WWTPs were determined from DMRs obtained from the PCS Database from the EPA. The DMRs record flow rates once a month. Flow is monitored regularly at all WWTPs within the Copano Bay watershed, so the average of all measured flows was used in the bacterial loading calculations. The flows used to calculate the average bacterial loading from WWTPs are given in Table 5.12. 108 Table 5.12 Flow Rates of WWTPs Permit Number Facility Description Average Flow Rate (m 3 /yr) Average Flow Rate (MGD) 10124-002 City of Beeville, Moore Street WWTP 3,086,069 10 2.23 10156-001 Woodsboro WW Treatment Facility 175,141 0.13 10237-001 City of Odem WWTP 207,256 0.15 10255-001 Town of Refugio WW Treatment Facility 466,003 0.34 10705-001 City of Taft, Baird WWTP 622,277 0.45 10055-001 City of Sinton Main WWTP 830,607 0.60 10124-004 Chase Field WWTP 569,976 0.41 10748-001 Pettus Municipal Utility District WWTP 88,334 0.06 13892-001 Water Reclamation Facility 11,586 0.01 5.4.2.3 Calculating Annual Bacterial Loading (CFU/year) from WWTPs The annual bacterial loading from WWTPs was calculated by multiplying the fecal coliform concentration (Section 5.4.2.1 and listed in Table 5.11) by the average flow rate (determined in Section 5.4.2.2 and listed in Table 5.12). The annual bacterial loading from each WWTP (based on measured flow rates and estimated bacteria concentrations) is given in Table 5.13. 10 Error in the permit files (missing decimal points) were discovered late in the analysis. For all calculations, a flow rate of 155,283,858 m 3 /year was used, but this is a conservative flow rate. 109 Table 5.13 Annual Bacterial Loadings from WWTPs Bacterial Loading Permit Number Facility Description CFU/yr TCFU/yr 10055-001 City of Sinton Main WWTP 6.98E+14 698 10124-002 City of Beeville, Moore Street WWTP 4.47E+15 11 4,470 10124-004 Chase Field WWTP 4.79E+14 479 10156-001 Woodsboro WW Treatment Facility 2.21E+14 221 10237-001 City of Odem WWTP 2.04E+10 0.02 10255-001 Town of Refugio WW Treatment Facility 2.61E+13 26.1 10705-001 City of Taft, Baird WWTP 1.24E+11 0.12 10748-001 Pettus Municipal Utility District WWTP 7.42E+13 74.2 13892-001 Water Reclamation Facility 9.73E+12 9.73 5.4.2.4 Calculating Residence Time to Mainstreams Based on Flow Length Because the WWTPs are located at various distances from the mainstreams that are modeled for the Copano Bay watershed, the residence times from each WWTP to the downstream main river was calculated. Ernest To of CRWR derived a relationship between residence time and mean flow length of the watersheds 12 . He isolated one portion of the model to calculate the overland flow velocity. Overland velocity can be calculated from the following equation: v = L/τ (5.5) Where: L = mean flow length τ = residence time v = overland flow velocity 11 For all subsequent calculations, a bacterial loading of 2.25E+17 was used (based on the conservative flow rate). 12 The mean flow length of each watershed is the average flow length from each watershed to the watershed drainage outlet. 110 The overland flow velocity was estimated from two Schemalink feature classes along the Aransas River (Schemalink 120 and 125) that have the most available data. The overland flow velocity was then extrapolated to the entire watershed. The methodology is described as follows: The relationship between the upstream and downstream loads is given by the equation: Lds = Lus * exp (-k*tau) (5.6) Where: Lds = downstream load Lus = upstream load k = decay rate tau = watershed residence time At both Schemalinks 120 and 125, Lds and Lus are known, therefore k*tau can be directly calculated. By assuming that k ~ 1.5 days -1 , tau can be calculated. With tau and mean_flow_length (mean flow length from watershed to the watershed drainage outlet) available for the two Schemalinks, linear regression can be applied to find the overland flow velocity for the watershed. The following equation (derived by Ernest To, CRWR) relates flow length to residence time in the Copano Bay watershed: τ = 2 x 10 -5 * L + 1.6717 (5.7) Where: τ = residence time (d) L = mean flow length (m) 111 Appendix 5.5 explains the procedure for determining the mean flow length for watersheds (which was used in deriving Equation 5.7), and Appendix 5.6 provides the procedure that was used in determining the mean flow length (based on Equation 5.7) from each WWTP to the downstream main river channels, which 3d models of the river channels were created. Table 5.14 gives the calculated flow length from each of the WWTPs to the main river channels (determined from the flow length raster that was created in Appendix 5.6) as well as the decayed bacterial loading that is applied to the model. Table 5.14 WWTP Bacterial Loading Applied to Model (See Chapter 6) (A) (B) (C) (D) (E) Permit Number Bacterial Loading (CFU/year) Mean Flow Length, L (m) Residence Time, τ (days) Decayed Bacterial Loading (CFU/year) Apply to IncVal SchemaNode HydroID 10055-001 6.98E+14 0 0 6.98E+14 92 10124-002 4.47E+15 20338.7 2.47 1 3.20E+13 62 10124-004 4.79E+14 0 0 4.79E+14 64 10156-001 2.21E+14 3572.5 2.74 2 9.17E+11 70 10237-001 2.04E+10 12258.3 1.92 4.42E+08 92 10255-001 2.61E+13 3662.87 1.74 7.96E+11 65 10705-001 6.22E+05 9844.47 1.87 1.48E+04 67 10748-001 7.42E+13 10080.6 1.87 1.75E+12 90 13892-001 9.73E+12 482.132 1.68 3.37E+11 69 1 Residence time was calculated (based on Equation 5.7) to be 2.08 days, but the residence time was adjusted to allow more decay in order to match the median fecal coliform concentration at the downstream bacterial monitoring station. 2 Residence time was calculated as 1.74 days, but one day was added since bacteria are flowing in stream before being applied to model. Column A is the bacterial loading (CFU/year) that was calculated in Section 5.4.2.3 and shown in Table 5.12. Column B is the mean flow length from WWTPs to the next downstream main channel (or Copano Bay) that was determined (as described in Appendix 5.6.) Column C is the residence time calculated using Equation 5.7 from the 112 mean flow length in Column B. Column D is the decayed bacterial loading from using the first-order decay equation and assuming a decay coefficient of 2 days -1; (see Section 6.3.3.1 for how decay coefficient was determined). (D) = (A) * exp (-2 days -1 *(C)). Column E is the SchemaNode to which the bacterial loading is applied (described in further detail in Chapter 6.) 5.4.3 Result The total annual bacterial loadings from WWTPs applied to the model are shown below in Figure 5.17. # # # # # # # # # Legend # WWTPs MainStreams Loadings (CFU/year) 0.00e+000 - 3.37e+011 3.38e+011 - 9.17e+011 9.18e+011 - 1.75e+012 1.76e+012 - 3.22e+013 3.23e+013 - 6.98e+014 Figure 5.17 Annual WWTP Bacterial Loadings (CFU/year) 113 5.5 ESTIMATION OF LOADINGS FROM SEPTIC SYSTEMS 5.5.1 Methodology Due to lack of data for septic systems, it was very difficult to quantify the fecal coliform bacterial loading that could be potentially contaminating certain areas of Copano Bay. There are many factors that affect whether or not bacteria from septic systems reach surface waters: type of soil, height of the water table, etc. Fecal coliform loadings from septic systems were estimated using data from a variety of sources (given in Section 3.2.1.4). The annual bacterial loadings per watershed from septic systems were calculated by finding the annual number of septic systems in use per watershed on the following types of land: Low and High Intensity Residential (land use codes 21 and 22, respectively.) The following steps were used to calculate the bacterial loadings from septic systems for the watershed model: 1. Determine the number of septic systems in use, number of complaints, population, and housing units per county in 2004 from U.S. Census of Bureau and TCEQ data. 2. Find the area (m 2 ) of the land use classifications of 21 and 22 in each county in the Copano Bay watershed. 3. Calculate the density of septic systems in use, complaints, population, and housing units per county (count/m 2 of High/Low Residential ) in 2004. The following equation would be used where the number in the parentheses indicates the step in which the value was determined: (1)/(2) 114 4. Find the area of land use classifications 21 and 22 of each county within each watershed, recognizing that watersheds may be part of multiple counties. 5. Find the area of each Soil Group (A, B, C, D) within each watershed in land use classifications 21 and 22. 6. Multiply the area (m 2 ) of each county within each watershed (for each soil group) by the septic system density and complaint density (count/m 2 ) to find the septic system count and complaint count per soil group for each watershed. (5) * (3) 7. Multiply the area (m 2 ) of each county within each watershed to find the population and housing count per watershed. (4) * (3) 8. Find the number of people per housing unit per watershed (from calculations made in step 7). 9. Apply the criteria given in Section 5.5.2.5 to find the number of impacting septic systems per watershed; these criteria account for the number of complaints and hydrologic soil groups in each of the watersheds. 10. Multiply the number of impacting septic systems per watershed by the number of people per housing unit (in corresponding watershed) to find the number of people that may be contaminating ground and surface waters. (9) * (8) 11. Multiply the number of humans (Step 10) by the number of fecal coliform excreted per year (CFU/year-human) to find the total CFU/year from impacting/malfunctioning septic systems per watershed. 115 5.5.2 Procedure of Application 5.5.2.1 Finding Septic Systems, Complaints, Population and Housing Units per County in 2004 Septic system data (annual count per county) were obtained from the 1990 U.S Census of Bureau, and the TCEQ provided the number of septic systems installed from 1990 - 2004. These data are given in Table 5.15. Table 5.15 Septic System Data from 1990 U.S. Census and TCEQ Applications (1990- 2004) County Sewage Disposal: Septic Tank or Cesspool (1990 Census) Applications for Septic Systems (TCEQ, 1990-2004) Aransas 6,456 2,931 Bee 3,859 616 Goliad 1,898 982 Karnes 1,765 269 Refugio 1,033 229 San Patricio 5,722 1,687 The number of complaints investigated for each county was also reported to the TCEQ. Assuming that the rate of complaint was constant, the number of complaints investigated from 1990-2004, and the complaint percentage per year was calculated. The TCEQ data (columns A-C) and the calculated data (columns D-F) are given in Table 5.16. Column D (Complaints Investigated per Year) is calculated by dividing Column B by the number of years, Column C. (D) = (B)/(C). Column E is calculated by multiplying Column D by 14 years (1990-2004). (E) = (D) * 14 years. Column F is calculated by dividing Column D (Complaints Investigated per Year) by the total number of septic system applications from TCEQ (Table 5.15) and multiplying by 100 to get the units in percent. 116 Table 5.16 Annual Septic System Complaint Percentage by County (A) (B) (C) (D) (E) (F) County Complaints Investigated Corresponding Years of Investigation Complaints Investigated per Year Complaints Investigated (1990-2004) Complaint Percentage per Year (%) Aransas 398 1993-2003 39.8 557 1.36 Bee 123 1992-2003 11.2 157 1.82 Goliad 12 1998-2003 2.4 34 0.24 Karnes 2 1998-2003 0.4 6 0.15 Refugio 18 1992-2003 1.6 23 0.71 San Patricio 1 321 1990-2004 23 321 1.36 1 No complaint investigation data was available for San Patricio County. Number of complaints was found by applying ratio of septic systems installed from Aransas County. The number of housing units per county for 2004 was determined by using the 1990 and 2002 U.S. Census data. The housing unit data available from U.S. Census are given in Table 5.17 (Columns A, B, and E.) Table 5.17 Housing Unit Data by County (A) (B) (C) (D) (E) (F) Housing Units Housing Units per Year Projected Housing Units Occupied Housing Units Occupancy Rate (%) County 1990 2002 1990-2002 2004 1990 1990/2004 Aransas 10,889 13,258 197 13,653 6,938 64 Bee 10,208 11,043 70 11,182 8,592 84 Goliad 2,835 3,480 54 3,588 2,208 78 Karnes 5,117 5,523 34 5,591 4,337 85 Refugio 3,739 3,660 -7 3,647 2,937 79 San Patricio 22,126 25,650 294 26,237 18,776 85 Column C is calculated by finding the difference between Column A and B and dividing by the number of years (1990-2002). (C) = {(B) – (A)}/(2002 – 1990). Column D (Projected Housing Units) is calculated by multiplying the housing units per year by 2 117 years (2004-2002) and adding it to the number of housing units in 2002. (D) = (C) * 2 years + (B). Column F (Occupancy Rate %) is calculated by dividing the Occupied Housing Units in 1990 (Column E) by the Housing Units in 1990 (Column A) and multiplying by 100 to get the units in percent. (F) = (E) / (A) * 100%. The number of septic systems in use (by county) can be found by assuming that the occupancy rate (%) is the same in 1990 as in 2004, and the number of complaints in 2004 is found by multiplying the number of septic systems in use by the complaint percentage per year (given in Table 5.16, Column F). The number of septic systems in use in 2004, the number of complaints in 2004, the population (from 2004 U.S. Census), and projected housing units in 2004 (from Table 5.17, Column D) are summarized in Table 5.18. When a complaint is filed for a house, it is usually a complaint that could apply to a whole neighborhood rather than just one house, so the number of complaints filed and investigated is used more as a qualitative assessment to identify areas that may have more malfunctioning septic systems. Table 5.18 Septic Systems in Use, Complaints Investigated, Housing Units, and Population in 2004 County Septic Systems in Use Complaints Investigated Housing Units Population Aransas 5,981 81 13,653 24,041 Bee 3,767 68 11,182 33,046 Goliad 2,243 5 3,588 7,104 Karnes 1,724 3 5,591 15,458 Refugio 991 7 3,647 7,640 San Patricio 6,287 85 26,237 68,187 118 5.5.2.2 Calculating Density of Septic Systems, Complaints Investigated, Housing Units, and Population per County in 2004 The density of septic systems, complaints investigated, housing units, and population per county in 2004 (count/m 2 ) were calculated by using the following equation: [Total annual count] / [Area in m 2 where the residences would be located within each county]. The area where septic systems, housing units, and populations would be located was assumed to be in the land use land cover classifications 21 and 22 (Low and High Density Residential). The low and high density residential land use areas in the Copano Bay watershed and the specified locations of septic systems around Copano Bay (TDH, 2000) are shown in Figure 5.18. Residential areas are greatly outnumbered by areas for agriculture, pasture, and shrubland (areas where livestock animals would be grazing), as shown in Figure 5.18. In order to find the residential area, the land use land cover dataset was masked by each county, and the corresponding grid cells (for land use codes 21 and 22) were summed. For example, in San Patricio County (calculation shown in Figure 5.19), the total area where septic systems, housing units, and people are located is 28,288,800 m 2 . Thus, the density of septic systems in San Patricio county is 6,287 septic systems/28,288,800 m 2 = 0.00022 septic/m 2 = 576 septic/mi 2 = 222 septic/km 2 . The density of septic systems, complaints investigated, housing units, and population per county in 2004 is given in Table 5.19. 119 # # # # # # # # # # Legend # Septic Systems around Copano Bay Low/High Residential Figure 5.18 Low and High Residential Land Use Areas and Septic System Locations around Bay 120 Table 5.19 Septic Systems, Complaints Investigated, Housing Units, and Population per County (Count per km 2 ) County Residential Area (km 2 ) Description Density (Count/km 2 ) Septic Systems 424 Complaints Investigated 6 Occupied Housing Units 967 Aransas 14.1 Population 1,700 Septic Systems 269 Complaints Investigated 5 Occupied Housing Units 799 Bee 14.0 Population 2,360 Septic Systems 952 Complaints Investigated 2 Occupied Housing Units 1,520 Goliad 2.4 Population 3,010 Septic Systems 244 Complaints Investigated 0 Occupied Housing Units 791 Karnes 7.1 Population 2,190 Septic Systems 222 Complaints Investigated 3 Occupied Housing Units 927 San Patricio 28.3 Population 2,410 121 5.5.2.3 Calculating Total Septic Systems, Complaints Investigated, Housing Units, and Population per Watershed The area of low and high residential land use classifications (m 2 ) of each county within each delineated watershed (Figure 5.1) was determined and then multiplied by each density in each corresponding county. Appendix 5.7 shows the calculations and results for this procedure. For example, Watershed JunctionID 45422 has two counties overlapping it (Bee and Karnes Counties), so there are two different areas and densities for which to account. The calculation of how many septic systems are in Watershed JunctionID 45422 is shown in Figure 5.20. 632 septic/mi 2 is the density of septic systems in Karnes County, and 697 septic/mi 2 is the density of septic systems in Bee County; these densities are per square mile of land where septic systems are assumed to be located based on land use classifications. Approximately 1.4 km 2 (0.537 mi 2 ) is the area in Watershed JunctionID 45422 that is a part of Bee County (the area of land use types 21 San Patricio County Land Use Land Cover Raster (30m)(30m) * 31,432 grid cells = 28,288,800 m 2 = 28.3 km 2 Figure 5.19 Determination of Area (km 2 ) of Septic Systems and Residences in San Patricio County 122 and 22 where septic systems and residences are assumed to be located). Approximately 0 m 2 (0 mi 2 ) is the area in Watershed JunctionID that is a part of Karnes County 13 . Thus, there are approximately 374 septic systems in Watershed JunctionID 45422. This procedure was replicated for the population, housing units, and number of complaints investigated in each watershed. Figure 5.20 Determination of Septic System Count in Watershed JunctionID 45422 13 There are no areas of low and high residential land use in Watershed JunctionID 45422 in Karnes County. 14,366 + Bee County Karnes County 632 septic/mi 2 * 0 mi 2 = 0 septics + 697 septic/mi 2 * 0.537 mi 2 = 374 septics Total = 374 septic systems 123 Table 5.20 Total Septic Systems, Complaints Investigated, Population, Housing Units per Watershed, and People/Housing Unit per Watershed Watershed JunctionID Septic Systems Complaints Population Housing Units People/Housing Unit 45422 374.36 6.80 3,284 1111.39 2.96 45408 183.25 3.19 1,668 578.96 2.88 45426 471.66 6.40 5,115 1968.30 2.60 45414 0.00 0.00 0.00 0.00 0.00 45416 1,230.96 16.72 13,350 5136.95 2.60 45405 2,256.85 30.65 11,996 5961.68 2.01 45421 0.00 0.00 0.00 0.00 0.00 45417 293.00 2.09 2,258 1077.91 2.09 45404 0.97 0.02 8.00 2.88 2.96 45409 190.81 3.46 1,674 566.48 2.96 45415 1,271.99 23.09 11,160 3776.27 2.96 45419 257.58 0.74 857 422.56 2.03 45413 1,790.42 32.50 15,708 5315.39 2.96 56830 0.00 0.00 0.00 0.00 0.00 56831 823.81 11.19 3,311 1880.51 1.76 45425 0.00 0.00 0.00 0.00 0.00 45418 224.24 1.60 1,728 824.94 2.09 45423 245.60 1.75 1,893 903.51 2.09 45406 0.00 0.00 0.00 0.00 0.00 45412 0.00 0.00 0.00 0.00 0.00 45410 82.14 0.59 633 302.18 2.09 124 As shown in Table 5.20, there are approximately 74,645 people that live in the Copano Bay drainage area. To find the number of people that are on septic systems, the number of people per housing unit was multiplied by the number of septic systems in each corresponding watershed and summed for the entire drainage area; people/housing unit and number of septic systems per watershed is given in Table 5.20. Approximately, 23,912 people (out of 74,645 people) are on septic while the remaining people are assumed to have their wastewater treated by WWTPs. 5.5.2.4 Calculating Septic Systems and Complaints Investigated in Soil Groups A, B, C, and D per Watershed After determining the density of septic systems and complaints in each county (see Section 5.5.2.2), the number of septic systems and complaints (in land use codes 21 and 22) in each Soil Group was found within each watershed. The hydrologic soil group data was retrieved from STATSGO. Soil data were downloaded for the state of Texas and then clipped to the Copano Bay watershed. A dbf table called “COMP.dbf” contains the hydrologic soil group data under the field, “Hydgrp”. “COMP.dbf” can be joined to the soil polygon feature class based on the field “MUID” that is found in each attribute table. The hydrologic soil groups in the Copano Bay watershed are shown in Figure 5.21. 125 Soil group A consists of soils that have low runoff potential and high infiltration rates and typically consist of USDA soil textures of sand, loamy sand, and sandy loam. The transmission rate is typically greater than 0.76 cm/hr (Maidment, 1992). Thus, septic systems with soils classified in Group A are more likely to allow contamination of and infiltration into the groundwater and surface waters than other soil classifications. These types of soils (as seen in Figure 5.21) are only found around the South portion of Copano Bay. Soil group B consists of soils that have moderate infiltration rates when the soil is thoroughly wetted and typically consist of USDA soil textures of silt loam and loam. The transmission rate is usually between 0.38 and 0.76 cm/h. Soil group C consists of soils that have low infiltration rates when the soil is thoroughly wetted and typically consists of USDA soil textures of sandy clay loam. The transmission rate is usually between 0.13 and 0.38 cm/h. Legend Watersheds # Septic Systems around Bay Soil Group A Soil Group B Soil Group C Soil Group D Low / High Residential # # # # # # # # # # Figure 5.21 Hydrologic Soil Group Classifications 126 Soil group D consists of soils that have high runoff potential and have very low infiltration rates when the soil is thoroughly wetted and typically consist of USDA soil textures of clay loam, silty clay loam, sandy clay, silty clay, and clay. The transmission rate is usually between 0 and 0.13 cm/h. The area (m 2 ) of each soil group within each watershed (and land use codes 21 and 22) was found and then multiplied by the corresponding densities (of each overlapping county). The procedure for determining the area of each soil group within each watershed (and land use codes 21 and 22) is described in Appendix 5.8. Table 5.21 Number of Septic Systems and Complaints per Watershed in Soil Groups A, B, C, and D Soil Group A Soil Group B JunctionID Septic Systems Complaints % Septic Systems Complaints % 45422 0 0 23 0.4 1.8 45408 0 0 0 0 45426 0 0 0 0 45414 0 0 0 0 45416 0 0 0 0 45405 1543 21 1.4 0 0 45421 0 0 0 0 45417 0 0 0 0 45404 0 0 0 0 45409 0 0 0 0 45415 0 0 1234 22.4 1.8 45419 0 0 254 0.7 0.3 45413 0 0 1787 32.4 1.8 56830 0 0 0 0 56831 419 6 1.4 0 0 45425 0 0 0 0 45418 0 0 0 0 45423 0 0 0 0 45406 0 0 0 0 45412 0 0 0 0 45410 0 0 0 0 Soil Group C Soil Group D 127 JunctionID Septic Systems Complaints % Septic Systems Complaints % 45422 300 5.5 1.8 51 0.9 1.8 45408 0 0 183 3.2 1.7 45426 0 0 472 6.4 1.4 45414 0 0 0 0 45416 0 0 1231 16.7 1.4 45405 0 0 697 9.5 1.4 45421 0 0 0 0 45417 0 0 293 2.1 0.7 45404 0 0 1 0.0 45409 0 0 191 3.5 1.8 45415 0 0 38 0.7 1.8 45419 0 0 4 0.0 45413 4 0.1 1.8 0 0 56830 0 0 0 0 56831 0 0 397 5.4 1.4 45425 0 0 0 0 45418 0 0 224 1.6 0.7 45423 0 0 246 1.8 0.7 45406 0 0 0 0 45412 0 0 0 0 45410 0 0 82 0.6 0.7 5.5.2.5 Calculating Septic System Bacterial Loading (CFU/year) per Watershed After determining the count of each septic system and complaints investigated within each soil group (see Section 5.5.2.4) and finding the population and occupied housing unit count in each watershed in low and high residential land use zones (see Section 5.5.2.3), an approximation of bacterial loadings from septic systems per watershed was found. It was assumed that all of the septic systems found in hydrologic soil group A provide little to no removal of fecal coliform bacteria before reaching groundwater and surface waters. An approximation of the bacterial loadings from septic systems in hydrologic soil groups B, C, and D was made while considering the number of 128 complaints (complaint percentage) in the corresponding watershed as well as the soil characteristics. These basic assumptions were applied to all bacterial loading calculations for each soil group: • Hydrologic soil group A: 100% of loading from septic systems flows directly into surface waters. • Hydrologic soil group B: 100% of loading from septic systems flows directly into surface waters if complaint percentage is greater than 1%. • Hydrologic soil group C: 50% of loading from septic systems flows directly into surface waters if complaint percentage is greater than 1%. • Hydrologic soil group D: 50% of the loading from septic systems flows directly into surface waters if complaint percentage is greater than 1%. The number of impacting septic systems (considering the above criteria) calculated for Watershed JunctionID 45415 is shown in Figure 5.22. The number of septic systems and complaints in Figure 5.22 are given in Table 5.21. 129 Once the number of impacting septic systems was found for each watershed, the bacterial loadings from septic systems were calculated. For example, for Watershed JunctionID 45415, the number of impacting septic systems is 1,249 septic systems (Figure 5.22), and the number of people per housing unit is 2.96 people/housing unit (Table 5.20). Assuming that the annual human production of fecal coliform is 7.3 x 10 11 CFU/year (EPA, 2005), then the total bacterial loading for Watershed JunctionID is (1,249 septic systems) * (2.96 people/housing unit) * (7.3 x 10 11 CFU/year-person) = 2.70 x 10 15 CFU/year. The same procedure was repeated for all watersheds. The total fecal coliform bacterial loadings contributed by septic systems in the Copano Bay watershed are shown in Figure 5.23. Soil Group B: 1,234 septic (complaints > 1%) Soil Group C: 0 septic systems Soil Group D: 38 septic (complaints > 1%) Legend Watersheds Soil_GroupA Soil_GroupB Soil_GroupC Soil_GroupD LandCover_Residential Bee County Impacting septic systems = 1,234 + 0.5 * 38 = 1,249 septic systems Figure 5.22 Determination of Number of Impacting Septic Systems in Watershed JunctionID 45415 130 5.5.3 Result After completing the procedure described in Section 5.5.2, the cumulative septic system bacterial loadings per watershed was calculated (following the procedure described in Section 5.5.2.5.) The septic system bacterial loading per watershed is shown in Figure 5.23. Legend Human_SS_cfu_year 0.00e+000 1.00e-002 - 4.47e+014 4.48e+014 - 1.17e+015 1.18e+015 - 2.78e+015 2.79e+015 - 3.86e+015 Figure 5.23 Septic System Annual Bacterial Loading (CFU/year) 131 5.6 ESTIMATION OF TOTAL LOADING The total bacterial loadings from the watersheds (and in the Copano Bay water segments) were calculated by summing the non-point bacterial loadings (Section 5.1), livestock loadings (Section 5.2), avian loadings (Section 5.3), WWTP loadings (Section 5.4), and septic system loadings (Section 5.5). Figure 5.24 shows the total bacterial loadings for all the watersheds and water segments in the Copano Bay watershed. Figure 5.24 Total Annual Watershed/Segment Bacterial Loadings (CFU/year) The watersheds’ bacterial loadings from cattle are significantly larger than the bacterial loadings from any other point or non-point bacteria source (as shown in Figure 5.24.) Legend 2.937742e+016 NonPoint (Urban, Forest, …) Avian Human (Septic Systems) Human (WWTPs) Cattle Horses Goats Sheep Hogs Chickens Layers 132 The percent distribution of bacterial loadings for the Copano Bay watersheds and water segments is shown in Figure 5.25, such that the relative difference of sources of bacterial loadings can be observed. Cattle are the predominant source of fecal coliform at most upstream watersheds (as shown in Figure 5.25.) At the one watershed (where septic system bacterial loadings dominate) where cattle are not the major contributor, the bacterial loadings are significantly lower than the other upstream watersheds (see Figure 5.24). It can also be seen that the livestock bacterial loadings are significantly larger than the non-point, WWTP, septic system, and avian bacterial loadings. However, avian loads are applied directly on Copano Bay, and the bacterial loadings that would affect the quality of the Bay the greatest are the watersheds directly adjacent. Legend NonPoint (Urban, Forest, …) Avian Human (Septic Systems) Human (WWTPs) Cattle Horses Goats Sheep Hogs Chickens Layers Figure 5.25 Percent Distribution of Bacterial Loading Sources 133 Table 5.22 summarizes the total bacterial loadings for the entire Copano Bay watershed from the major bacterial contributors. Table 5.22 Annual Bacterial Loadings (TCFU/yr) from Major Bacterial Sources in Entire Copano Bay Watershed Bacterial Source Number of Units Bacterial Loading (TCFU/yr) Cattle 111,433 219,635 Goats 2,299 12,589 Human (Septic Systems) 23,912 12,576 Non-Point (Urban, Forest, etc.) N/A 8,777 Sheep 659 3,607 Hogs 486 1,765 Human (WWTP) 50,733 1,213 These bacterial loadings are the input into the Schematic Processor and Monte Carlo Simulation Models. Thus, the bacteria transport (what happens to the bacteria as they flow from watersheds to rivers, along rivers, and into the Bay) was modeled to see how the bacterial loadings impact the quality of the rivers and Bay in the Copano Bay watershed. Bacteria transport is described in Chapter 6 and is modeled using the Schematic Processor. 134 Chapter 6: Modeling of Bacteria Transport – Schematic Processor 6.1 BACKGROUND Once point and non-point bacterial loadings were calculated per watershed (described in Chapter 5), the transport of bacteria from the watersheds to Copano Bay was modeled. To simulate bacterial load transport, “Process Schematic”, a script tool that was developed by Jon Goodall and Tim Whiteaker in 2003, was used to implement dynamic linked libraries (DLLs). The two processing engines (DLLs) that were used in this bacteria watershed model were clsDecay.dll, which accounts for first-order decay of bacteria along water segments, and clsCFSTR.dll, which calculates the increase in bacteria concentration in Copano Bay due to bacterial loadings from the upstream watersheds. Goodall and Whiteaker submitted a journal article describing the Schematic Processor and Schematic Network in more detail (Goodall and Whiteaker, 2006). 135 6.2 METHODOLOGY “Process Schematic”, also referred to as the Schematic Processor, can be used to model bacterial transport once the following steps have been completed: 1. Bacterial loadings have been calculated (Chapter 5). 2. Schematic Network of the Copano Bay watershed has been created. 3. The parameters of each SchemaNode and SchemaLink have been determined (through calculations and/or calibration). 136 6.3 PROCEDURE OF APPLICATION 6.3.1 Creation of Schematic Network To implement the “Process Schematic” tool, a Schematic Network of the Copano Bay watershed was created, and the procedure is described in detail in Appendix 6.1. The Schematic Network is made up of two feature classes: SchemaNode and SchemaLink. SchemaNode represents the nodes in the watershed (a watershed, drainage point, or Copano Bay.) SchemaLinks connect the SchemaNodes and are a way to model what happens to the bacteria as they travel to Copano Bay. The Schematic Network was created for the Copano Bay watershed; the Schematic Network, as well as the parameters (inputs), necessary to run the model are shown in Figure 6.1 and explained in more detail in Appendix 6.1 and Section 6.3.2. The Bay was segmented into four water segments, and the segmentation of the Bay is described in Section 6.3.1.1. 137 6.3.1.1 Copano Bay Segmentation Ward and Armstrong (1997) segmented Copano Bay into water segments based on water quality parameter trends in the bay. Their study was a trends analysis on the Corpus Christi Bay system, which includes Copano Bay, in which the spatial variation of water quality monitoring data was used to segment the Bay system. Their segmentation allows the parameters to be representative by geographical location. Legend SchemaNode 1 2 3 SchemaLink 1 2 3 Watershed Junction Copano Bay Land to Streams Streams Streams to Bay Bacterial Loading per Watershed (CFU/year), L Decay Coefficient, k (day -1 ) Volume of Copano Bay, V (m 3 ) Cumulative Runoff,, QQ (m 3 /yr)) Decay Coefficient, k (day -1 ) Travel Time, t (days) Decay Coefficient, k (day -1 ) Travel Time, t (days) Parameters Figure 6.1 Schematic Network and Parameters 138 The water segments determined by Ward and Armstrong are shown in Figure 6.2. Their 15 water segments were dissolved into four Copano Bay water segments, each supporting the drainage of the upstream watersheds. The segments were clipped to the Copano Bay watershed to calculate a surface area for each segment, which can be seen in Figure 6.3 and Table 6.1. Legend PB1 PB2 CP04 CP07 CP08 AR1 CP03 M1 M2 CP05 CP06 CP01 CP02 CP09 CP10 Figure 6.2 Copano Bay Initial Water Segments 139 Table 6.1 Dissolving of Copano Bay Segments (New Labeling) 6.3.2 Use of Dynamic Linked Libraries (DLLs) 6.3.2.1 First-order Decay: clsDecay.dll clsDecay.dll simulates the decay of bacteria along stream segments (shown in Figure 6.4) and assumes first-order decay: Initial Bay Segmentation Labels Copano Bay Segmentation Labels PB1 PB2 CP04 CP07 CP08 1 AR1 CP03 2 M1 M2 CP05 CP06 3 CP01 CP02 CP09 CP10 4 Figure 6.3 Copano Bay Segmentation 140 Decay Figure 6.4 Simulation of Decay load passed = load received * e -kτ (6.1) Where: k = first-order decay coefficient (day -1 ), which is stored as an attribute in SchemaLink. τ = residence time along streams (days), which is stored as an attribute in SchemaLink. 6.3.2.2 Continuous Flow Stirred Tank Reactor: clsCFSTR.dll clsCFSTR.dll calculates the increase in fecal coliform concentration in the Bay due to bacterial loadings. The Bay is assumed to be completely mixed and acts as four Continuous Flow, Stirred Tank Reactors (CFSTRs), which are shown in Figure 6.5; furthermore, the inflow into the Bay equals the outflow. The following equation calculates the fecal coliform concentration in the Bay: c = L / (Q + kV) (6.2) 141 Where: c = concentration in bay (CFU/m 3 ) L = bacterial load entering bay (CFU/year) Q = total flow (m 3 /year), which is stored as an attribute in SchemaNode k = first-order decay coefficient (year -1 ), which is stored as an attribute in SchemaNode V = Volume of bay (m 3 ), which is stored as an attribute in SchemaNode 6.3.3 Determination of Model Parameters The DLLs described in Section 6.3.2 have many parameters that need to be determined before implementing the Schematic Processor. Each SchemaLink and SchemaNode in Figure 6.1 has parameters associated with the feature class that need to be determined. The following sections describe how the parameters were determined for each SchemaLink and SchemaNode in the Schematic Network. 6.3.3.1 Decay Coefficient Section 2.2 describe the research that was conducted to quantify the decay coefficients that would be representative of different parts of the Copano Bay watershed. Figure 6.5 CFSTRs 142 From the literature review, the decay coefficient typically ranges between 0.5 and 3 days -1 for a typical system. However, due to lack of data on the Copano Bay watershed, these values are still inconclusive. To determine a decay coefficient distribution for the Copano Bay watershed, one portion of the model was analyzed that contains the most available data, so that the decay coefficient could be calculated. The portion of the model that was segregated is shown in Figure 6.6. The data available for this portion of the model are from one USGS gauge flow data (USGS station 08189700) and two bacterial monitoring stations (Stations 12952 and 12948.) Figure 6.6 Segregated Portion of Model to Calculate Decay Coefficient Distribution ") ") ") ") ") ")" ) ") ) ") ") ") ") ") ") ") ") ") ") ") ") l l llll l l l l l Legend l USGS Gauge Stations " ) Bacteria Monitoring Stations ") ") l Station 12952 USGS 08189700 Station 12948 143 From 1999-2005, bacterial monitoring station 12952 only has five measurements (quarterly measurements taken between 2002 and 2003), and bacterial monitoring station 12948 has sixteen measurements. Thus, five days of data (corresponding to the dates from bacterial monitoring station 12952) were used in this analysis. The data from the bacterial monitoring stations and the USGS gauge station for these five days are given in Table 6.2. The data in this table were used to calculate a decay coefficient for each corresponding day (using Equation 6.1), so a distribution of five decay coefficients could be calculated. Table 6.2 Available Data for Segregated Model on Aransas River Segment 2 Date Sta. 12952, Upstream (CFU/100mL) Sta. 12948, Downstream (CFU/100mL) USGS 08189700 Daily Flow (m 3 /s) 7/8/2002 836 1327 1.22 10/15/2002 25 122 0.19 1/21/2003 72 58 0.28 4/22/2003 130 34 0.21 8/18/2003 58 28 0.14 The bacterial concentrations were converted into bacterial loadings according to the following relationship: L = Q * c (6.3) Where: L = bacteria loading (CFU/year) Q = flow rate (m 3 /year) c = fecal coliform concentration (CFU/m 3 ) 144 The upstream bacterial loading (located at Station 12952) was calculated by multiplying the bacterial concentration at Station 12952 by the measured daily flow rate at the USGS gauge station (USGS 08189700), which is given in Table 6.2. The downstream bacterial loading (located at Station 12948) was calculated by multiplying the downstream daily flow rate by the downstream bacterial concentration (located at Station 12948). Since there is no USGS gauge station at the downstream location, the downstream flow rate was calculated according to the following equation: q ds = (q us /q mean,us ) * q mean, ds (6.4) Where: q ds = downstream flow rate at Station 12948 q us = upstream flow rate at Station 12952 q mean,us = median flow rate at upstream station (from USGS data) = 0.12 m 3 /s q mean,ds = mean downstream flow rate (from modeled flow rates, see Figure 6.18) = 2.11 m 3 /s The residence times between the upstream and downstream bacterial monitoring stations (segment referred to as Aransas River Segment 2 in Section 6.3.3.2.2) were determined by applying the equation derived in Section 6.3.3.2.2, which relates residence time and flow rate (see Figure 6.13). This equation is repeated below for convenience. τ = -0.4374*lnQ + 1.7584 (6.5) Where: τ = residence time (days) 145 Q = upstream or downstream flow rate (m 3 /s) The residence time was found for each day for both the corresponding upstream and downstream flow rates, and then the residence time was averaged. The residence time of the watershed was assumed to be 1.5 times longer than the residence time of the stream (Aransas River Segment 2). The parameters and the assumptions for calculating the decay coefficient distribution are shown in Figure 6.7. In Figure 6.7, B 1 and B 3 are the bacterial loadings calculated from the measured fecal coliform concentrations and USGS flow data (Equation 6.3). B 2 is the bacterial loading for Watershed JunctionID 45408 that was calculated in Chapter 5 (3.07 x 10 16 CFU/year). Thus, for all five days, B 2 remains constant. The decay coefficient was assumed to be constant for the Aransas River and the watershed travel due to lack of data ") ") ") ") ") ")" ) ") ) ") ") ") ") ") ") ") ") ") ") ") ") l l llll l l l l l Legend l USGS Gauge Stations " ) Bacterial Monitoring Stations ") ") l B 1 B 3 B 2 k, τ k, 1.5τ Figure 6.7 Assumptions and Derivation of k-distribution 146 availability. The relationships among the parameters/values, taking into account first- order decay, are shown below (Equation 6.6). B 1 *exp -kτ + B 2 *exp -k(1.5τ) = B 3 (6.6) For each of the five days, all the values are known except for k, so k was calculated for each day; the results are given in Table 6.3. Table 6.3 Calculation of Decay Coefficient for Segregated Model Date B 1 (CFU/year) B 2 (CFU/year) B 3 (CFU/year) τ (days) 1.5*τ (days) k (days -1 ) 7/8/2002 3.21E+14 3.07E+16 8.96E+15 1.05 1.57 0.80 10/15/2002 1.50E+12 3.07E+16 1.28E+14 1.86 2.79 1.96 1/21/2003 6.37E+12 3.07E+16 9.02E+13 1.69 2.53 2.30 4/22/2003 8.59E+12 3.07E+16 3.95E+13 1.81 2.72 2.44 8/18/2003 2.49E+12 3.07E+16 2.11E+13 2.00 3.01 2.42 The range of k values is from 0.80 to 2.42 days -1 , which are all in between the typical range of 0.5 to 3 days -1 . As can be inferred from the flow rate data in Table 6.2, calculation of the 7/8/2002 k value was based on a storm event. The other four k values are more similar to each other than to the first k value because the flow rates on 10/15/2002 – 8/18/2003 were more similar. From the calculated k-distribution, an average decay coefficient of k = 2 days -1 was used for all the SchemaNodes and SchemaLinks in the Schematic Processor Model. 147 6.3.3.2 Residence Time One of the most critical input parameters in the model is the hydraulic residence time, τ, of the bacteria for each of the water segments. Residence time is the amount of time the bacteria remain in a specific water segment; thus, the residence time corresponds to the amount of time that the bacteria decay in a specific environment (e.g., watershed, river, or bay). Residence time can be calculated according to the following relationship: τ = V/Q (6.7) Where: V = volume of the water segment Q = flow rate of the water segment Before determining the volume and flow going through each water segments, 3d representations of each of the main river channels were created using Venkatesh Merwade’s River Channel Morphology Model (RCMM) Toolbar (Merwade and Maidment, 2006), which is described in Section 6.3.3.2.1. 6.3.3.2.1 3d Channel Morphology (RCMM Toolbar) Using the National Hydrography Dataset (NHD in high resolution), a mainstream network was created (shown in Figure 6.8) to generate 3d models of the main channels in HEC-RAS. Only the main channels were modeled due to data availability. Following is a list of the criteria used to determine which river segments are main channels: • River segments that have “GNIS_Name”, which is a field in the NHD feature class. • River segments with streamflow greater than 30 cfs (based on Reach File, RF1). 148 • River segments that have a USGS gauge station measuring the streamflow. Before the toolbar can be used, specific information and shapefiles are needed. The RCMM toolbar requires the NHD centerline of the river and at least two points on the river where the width, depth, and bank elevation are known. For this project, the width, depth, and bank elevation were determined at the USGS gauge station on each river segment, and other available sources were used to determine the most downstream river segment cross-section. 6.3.3.2.2 3d Model of Aransas River: Residence Time Determination In order to find the cross-sectional area at the USGS gauge station, USGS gauge data (for USGS station 08189700) were downloaded from the USGS website from Legend USGS_point NHDFlowline_MainStreams NHDFlowline Figure 6.8 Mainsteam Network for RCMM 149 “Surface-Water Measurements”, which includes width, area, and stream flow measurements. The flow and width were then plotted for all the data available from 1987 to 2005 (Figure 6.9). The width of the channel was approximated to be 35 feet, which is the width of the channel before the width dramatically increases at higher flowrates. The channel was assumed to have a square cross-section; thus, the depth can be calculated for each USGS measurement by dividing the measured area by the measured width for each stream flow. The depth and width of the channel were plotted for all available measurements in Figure 6.10, and a best-fit line was determined. The best-fit line was then used to find the depth at the width of 35 feet. Hence, the depth at USGS gauge station 08189700 was estimated to be 1.24 feet. 0 20 40 60 80 100 120 140 160 180 200 0.01 0.1 1 10 100 1000 Flow (cfs) Wi d t h ( f t ) ~35 feet Figure 6.9 Flow versus Width for USGS Station 08189700 150 The bank elevation is given for the USGS gauge station on the USGS website in “Surface-Water Measurements”, and is 72.37 feet above sea level for Station 08189700. Because there is only one USGS gauge station along the Aransas River, other sources were used to estimate the cross-section at the most downstream point of the Aransas River. The other sources that were used were aerial photographs and the Reach File (RF1) Database. First, the sources were compared to the USGS approximation at the USGS gauge station location to see the percent difference (Table 6.4). Table 6.4 Upstream Cross-Section Data Comparison (Aransas River; USGS 08189700) Percent Error (%) Width (ft) Depth (ft) Width Depth RF1 17.77 0.44 49.22 64.46 Aerial 31.00 - 11.42 - USGS/RCMM Input 35.00 1.24 - - The aerial photograph seems to be closest to the USGS approximation for width (as shown in Table 6.4). The RF1 file is the only known data source to approximate Figure 6.10 Width versus Depth (Square Cross-Section) for USGS Station 08189700 y = 0.0798x 0.7711 R 2 = 0.5669 0 1 1 2 2 3 3 4 4 5 5 0 5 10 15 20 25 30 35 40 Width (ft) D e p th (ft) ~35 feet 1.24 ft 151 depth since the available bathymetric maps do not cover the rivers in the Copano Bay watershed. The downstream dimensions using the available sources are given in Table 6.5. Table 6.5 Downstream Cross-Section Data (Aransas River) Width (ft) Depth (ft) RF1 39.54 0.71 Aerial 205.38 - The width of the channel was approximated (taking into account the 11.42% error) to be 184.33 feet while the depth was approximated (taking into account the 64.46% error) to be 2.00 feet, which seems too low since this location is at the discharge point into Copano Bay. Thus, using the same width-to-depth ratio that exists at the USGS gauge station, the depth was approximated to be 6.52 feet. The bank elevation at the most downstream point was determined using the Digital Elevation Model (DEM). A summary of all the data that are needed for the RCMM toolbar is shown in Figure 6.11. 152 After obtaining the NHD centerline feature class, USGS gauge station feature class, and the cross-section dimensions at two points, Merwade’s RCMM toolbar was used to generate a 3d model of the Aransas River (Merwade and Maidment, 2006). However, the river was divided into four 3d segments (shown in Figure 6.11) because the residence time was needed for each of these four segments. The division of these segments was based on the watershed delineations. Each of the points represents when the river crosses a watershed; the watersheds and Schematic Network were based on Critical Points: USGS gauge stations, bacterial monitoring stations, and water segment endpoints. Using the 3d Aransas River model (HEC-RAS), a relationship was found between residence time and flow rate for each of the four segments (Figure 6.11). The relationships between residence time and flow rate are shown in Figures 6.12 - 6.15. Figure 6.11 Summary of RCMM Toolbar Data Requirements (Aransas River) Width = 35 feet Depth = 1.24 feet Bank Elevation = 72.37 feet (USGS) Width = 184 feet (Aerial) Depth = 6.52 feet (width:depth) Bank Elevation = 0.66 feet (DEM) Legend NHD_Centerline USGS_point TravelTimes_RiverBreaks Segment 3 Segment 4 Segment 1 Segment 2 153 y = -0.4738x + 0.259 R 2 = 0.9948 0.185 0.190 0.195 0.200 0.205 0.210 0.215 0.100 0.110 0.120 0.130 0.140 0.150 0.160 Flow (m 3 /s) Residence Time (days) Figure 6.12 Flow versus Residence Time for Aransas River Segment 1 154 y = -0.4374Ln(x) + 1.7584 R 2 = 0.9987 1.4 1.5 1.6 1.7 1.8 1.9 2.0 0.0 0.5 1.0 1.5 2.0 2.5 Flow (m 3 /s) Residence Time (days) Figure 6.13 Flow versus Residence Time for Aransas River Segment 2 155 y = -0.0507x + 0.8305 R 2 = 0.9959 0.62 0.63 0.64 0.65 0.66 0.67 0.68 0.69 3.0 3.2 3.4 3.6 3.8 4.0 4.2 Flow (m 3 /s) R esidence Time ( d ays) Figure 6.14 Flow versus Residence Time for Aransas River Segment 3 156 A flow cumulative distribution function (CDF) plot was created for each segment. The flow CDF was based on USGS gauge daily mean streamflow data (from 1964-2004). USGS gauge station 08189700 is on Aransas River Segment 1, so a flow CDF was found directly for Segment 1. The flow CDF for Segment 1 is shown in Figure 6.16. y = -0.0581Ln(x) + 0.2967 R 2 = 0.9974 0.170 0.180 0.190 0.200 0.210 0.220 0.230 0.240 23456789 Flow (m 3 /s) Residence Time (days) Figure 6.15 Flow versus Residence Time for Aransas River Segment 4 157 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0. 01 0. 0 4 0. 15 0 . 3 0. 45 0 . 6 0. 75 0 . 9 1. 05 1 .2 1. 3 5 1 . 5 1 0 0 Daily Flow, q (m 3 /s) Figure 6.16 Flow Cumulative Distribution Function for Aransas River Segment 1 Because there is only one USGS gauge station on the Aransas River, the flow CDF of Segment 1 was used to find the flow CDF for Segments 2, 3, and 4. The median flow for Segment 1 was determined by finding the flow when the flow CDF equals 0.5, which is approximately 0.12 m 3 /s (see Figure 6.16.) A dimensionless CDF was then found for Segment 1 by dividing the flow by the median flow (Figure 6.17). 158 The flows for Segments 2, 3, and 4 were determined according to the following relationship: q = (q/q 0.5 ) * q mean (6.8) Where: (q/q 0.5 ) = dimensionless value obtained from Segment 1 q mean = mean flow of Segments 2, 3, or 4 from water quality model Each watershed has an annual mean flow that was calculated using the runoff equations derived by Quenzer (2003), which are given in Section 5.1.2.4. The mean flow was calculated by averaging the cumulative flow at the upstream point of the segment (the sum of flow of all upstream watersheds) and the cumulative flow at the downstream point 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0 . 1 0 . 3 1 . 3 2 . 5 3 . 8 5 . 0 6. 3 7 . 5 8 .8 10 . 0 11 . 3 12 . 5 8 3 3. 3 q/q 0.5 Figure 6.17 Cumulative Distribution Function, q/q 0.5 for Segment 1 159 of the segment. The mean flow to Aransas River Segment 2 was calculated and is shown in Figure 6.18. 160 Upstream Watersheds Draining to Upstream Point of Segment 2 Segment 2 Upstream Watersheds Draining to Downstream Point of Segment 2 Figure 6.18 Calculating Mean Flow to Aransas River Segment 2 161 After finding the mean flow for each segment, the flow CDF was calculated for each segment by using the relationship in Equation 6.8. The results for Segments 2, 3, and 4 are shown in Figures 6.19 - 6.21. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0. 18 0. 70 2. 64 5. 28 7 . 91 1 0. 5 5 13 . 19 15 . 8 3 18 . 46 21 . 10 23 . 74 26 . 38 17 58 . 33 Daily Flow, q (m 3 /s) Figure 6.19 Flow Cumulative Distribution Function for Aransas River Segment 2 q mean = 2.11 m 3 /s 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0. 29 1. 14 4. 28 8. 55 12 . 83 1 7. 1 0 21 . 38 25 . 6 5 29 . 93 34 . 20 38 . 48 42 . 75 28 50 . 00 Daily Flow, q (m 3 /s) Figure 6.20 Flow Cumulative Distribution Function for Aransas River Segment 3 q mean = 3.42 m 3 /s 162 Once the flow CDF was found for the four segments, the Residence Time Distribution (RTD) was determined for each segment using the relationships between residence time and flow that are shown in Figures 6.12 - 6.15. The RTD for Segments 1, 2, 3, and 4 are shown in Figures 6.22 - 6.25. Figure 6.21 Flow Cumulative Distribution Function for Aransas River Segment 4 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0. 46 1. 84 6. 91 13 . 83 20 . 74 2 7. 6 5 34 . 56 41 . 4 8 48 . 39 55 . 30 62 . 21 69 . 13 46 08 . 33 Daily Flow, q (m 3 /s) q mean = 5.53 m 3 /s 163 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.00 0.50 1.00 1.50 2.00 2.50 Residence Time (days) Figure 6.23 Residence Time Distribution for Aransas River Segment 2 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.00 0.05 0.10 0.15 0.20 0.25 Residence Time (days) Figure 6.22 Residence Time Distribution for Aransas River Segment 1 164 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 Residence Time (days) Figure 6.24 Residence Time Distribution for Aransas River Segment 3 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Residence Time (days) Figure 6.25 Residence Time Distribution for Aransas River Segment 4 165 The probability distributions were also found for the residence times of all the segments along the Aransas River and are shown in Figures 6.26 - 6.29. 0.00 0.05 0.10 0.15 0.20 0.25 0.00 0.05 0.10 0.15 0.20 0.25 Residence Time (Days) Figure 6.26 Probability Distribution of Residence Time for Aransas River Segment 1 0.00 0.05 0.10 0.15 0.20 0.25 0.00 0.50 1.00 1.50 2.00 2.50 Residence Time (Days) Figure 6.27 Probability Distribution of Residence Time for Aransas River Segment 2 166 0.00 0.05 0.10 0.15 0.20 0.25 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 Residence Time (Days) Figure 6.28 Probability Distribution of Residence Time for Aransas River Segment 3 167 0.00 0.05 0.10 0.15 0.20 0.25 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Residence Time (Days) Figure 6.29 Probability Distribution of Residence Time for Aransas River Segment 4 The RTDs were used as approximations for the corresponding SchemaLinks’ residence times in the Schematic Processor Model. Aransas River Segments 1, 2, 3, and 4 correspond to SchemaLink HydroIDs 113, 120, 107, 112, respectively, which are shown in Figure 6.30. For this Schematic Processor Model, the most frequent residence time (from the probability and residence time distributions) is the residence time associated with the corresponding SchemaLinks, and these residence times are given in Table 6.6. 168 Table 6.6 Residence Times of Aransas River Segments for Schematic Processor Model Aransas River Segment SchemaLink HydroID Residence Time (days) 1 113 0.212 2 120 1.51 3 107 0.686 4 112 0.208 6.3.3.2.3 3d Model of Mission River: Residence Time Determination The two USGS gauge stations along the Mission River were used to create a 3d model of Mission River in HEC-RAS (USGS gauge stations 08189300 and 08189500). The same methodology that was used to find the RTD of the Aransas River segments was used. 122 1 2 0 1 2 6 10 7 142 1 1 5 1 1 1 4 0 1 1 2 1 3 3 113 1 1 6 1 3 4 1 5 0 6 1 1 1 8 122 1 2 0 1 2 8 1 2 3 1 2 6 1 3 1 1 1 1 1 2 9 107 1 3 8 1 2 4 1 3 4 125 11 5 127 1 5 9 16 1 1 1 2 1 4 3 1 3 5 1 6 0 1 5 7 1 3 3113 1 1 0 1 4 4 Figure 6.30 SchemaLinks of Corresponding Aransas River Segments 169 To find the cross-sectional dimensions at the USGS gauge station, USGS gauge data (for USGS stations 08189300 and 08189500) were downloaded from the USGS website from “Surface-Water Measurements”, which includes width, area, and stream flow measurements. The flow and width were then plotted for all available data (Figures 6.31 and 6.32). “Surface-Water Measurements” were available from 2001-2005 for USGS 08189300 and from 1971-2005 for USGS 08189500. 170 0 20 40 60 80 100 120 140 160 180 200 0.1 1 10 100 1000 10000 Flow (cfs) W i dt h ( f t ) Figure 6.32 Flow versus Width for USGS Station 08189500 ~75 feet y = 3.5815x 0.4267 R 2 = 0.8299 0 50 100 150 200 250 300 350 0.1 1 10 100 1000 10000 100000 Flow (cfs) W i dt h ( f t ) ~25 feet Figure 6.31 Flow versus Width for USGS Station 08189300 171 The width of the channel at USGS gauge station 08189300 (on Medio Creek), Figure 6.31, was approximated to be 25 feet. The width of the channel at USGS gauge station 08189500 (on Mission River), Figure 6.32, was approximated to be 75 feet. The channels were assumed to have square cross-sections; thus, the depth was calculated for each USGS measurement by dividing the measured area by the measured width for each stream flow. The depth and width of each channel were then plotted for all available measurements in Figures 6.33 and 6.34. The best-fit line (given in Figure 6.33) was used to find the depth in Medio Creek at the width of 25 feet. The approximate depth at USGS gauge station 08189300 is 1.69 feet. Similarly, using the data in Figure 6.34, the depth at USGS gauge station 08189500 at a width of 75 feet was approximated to be 8.20 feet. ~25 feet 1.69 ft y = 0.1665x 0.7193 R 2 = 0.8104 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0 5 10 15 20 25 30 35 40 Width (ft) D e pt h ( f t Figure 6.33 Width versus Depth (Square Cross-Section) for USGS Station 08189300 172 The bank elevations are given for USGS gauge stations on the USGS website (in “Surface-Water Measurements”) and are 163.00 feet above sea level for USGS 08189300 and 1.00 feet above sea level for USGS 08189500. Because there is only one USGS gauge station along Medio Creek, other sources were used to estimate the cross-section at the most downstream point of Medio Creek where it drains into Mission River. The other sources that were used were aerial photographs and the RF1 Database. First, the sources were compared to the USGS approximation at the USGS gauge station location to see the percent difference. The percentage differences are shown below in Table 6.7. Table 6.7 Upstream Cross-Section Data Comparison (Medio Creek; USGS 08189300) Percent Error (%) Width (ft) Depth (ft) Width Depth RF1 14.96 0.13 40.16 92.24 Aerial 21.03 - 15.88 - USGS/RCMM Input 25.00 1.69 - - ~75 feet 0 2 4 6 8 10 12 14 0 1020304050607080 Width (ft) D e pt h ( f t 8.20 ft Figure 6.34 Width versus Depth (Square Cross-Section) for USGS Station 08189500 173 The aerial photograph is closer to the USGS approximation for width (Table 6.7); and the RF1 file is the only known data source to approximate depth because the available bathymetric maps do not cover the rivers in the Copano Bay watershed. The downstream dimensions using the available sources are shown in Table 6.8. Table 6.8 Downstream Cross-Section Data (Medio Creek) Width (ft) Depth (ft) RF1 13.77 0.12 Aerial 22.97 - The width of the channel was approximated (taking into account the 15.88% error) to be 27.30 feet while the depth was rounded up to 2.00 feet (taking into account the 92.24% error and in order to keep the depth deeper than the upstream depth). The bank elevation at the most downstream point was determined using the DEM. A summary of all the data needed for the RCMM toolbar for Medio Creek is shown in Figure 6.35. 174 The cross-section at the most downstream point of Mission River, where it drains into Copano Bay, is determined using the same methodology as for Aransas River and Medio Creek. The other sources that were used were aerial photographs and the RF1 Database. First, the sources were compared to the USGS approximation at the USGS gauge station location to see the percentage difference. The percentage differences are shown below in Table 6.9. Width = 25.00 feet Depth = 1.69 feet Bank Elevation = 163.00 feet (USGS) Width = 27.30 feet (Aerial) Depth = 2.00 feet (RF1) Bank Elevation = 23.82 feet (DEM) Legend NHD_Centerline USGS_point TravelTime_RiverBreaks Segment 1 Figure 6.35 Summary of RCMM Toolbar Data Requirements (Medio Creek) 175 Table 6.9 Upstream Cross-Section Data Comparison (Mission River; USGS 08189500) Percent Error (%) Width (ft) Depth (ft) Width Depth RF1 35.59 0.26 52.54 96.86 Aerial 32.81 - 56.26 - USGS/RCMM Input 75.00 8.20 - - The RF1 file seems is closer to the USGS approximation for width (Table 6.9); and the RF1 file is the only known data source to approximate depth because the available bathymetric maps do not cover the rivers in the Copano Bay watershed. The downstream dimensions using the available sources are shown in Table 6.10. Table 6.10 Downstream Cross-Section Data (Mission River) Width (ft) Depth (ft) RF1 60.57 0.37 Aerial 125.66 - The width of the channel was approximated (taking into account the 52.54% error) to be 264.76 feet while the depth (taking into account the 56.26% error) was 11.83 feet. The bank elevation at the most downstream point was determined using the DEM. A summary of all the data needed for the RCMM toolbar for Mission River is shown in Figure 6.36. 176 After obtaining the NHD centerline feature class, USGS gauge station feature class, and the cross-section dimensions at two points, Merwade’s RCMM toolbar was used to generate 3d models of Medio Creek and Mission River. Medio Creek was made Segment 3 Segment 4 Width = 75.00 feet Depth = 8.20 feet Bank Elevation = 1.00 feet (USGS) Width = 265.00 feet (Aerial) Depth = 11.83 feet (RF1) Bank Elevation = 1.48 feet (DEM) Legend NHD_Centerline USGS_point TravelTime_RiverBreaks Segment 5 Segment 2 Figure 6.36 Summary of RCMM Toolbar Data Requirements (Mission River) 177 into one 3d segment (Figure 6.35), and Mission River was divided into four 3d segments (Figure 6.36) because the residence time for each of these five segments was needed. Using the 3d Medio Creek and Mission River models (HEC-RAS), a relationship was found between residence time and flow rate for each of the segments. The relationships between residence time and flow rate for Medio Creek and Mission River are shown in Figures 6.37 - 6.41. y = -0.7746Ln(x) + 1.3784 R 2 = 0.9895 1.50 2.00 2.50 3.00 3.50 4.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 Flow (m 3 /s) R esidence Time ( d ays) Figure 6.37 Flow versus Residence Time for Medio Creek Segment 1 178 y = -0.26Ln(x) + 0.8222 R 2 = 0.9856 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Flow (m 3 /s) R e sidence Time ( d ays) Figure 6.38 Flow versus Residence Time for Mission River Segment 2 179 y = -0.0859Ln(x) + 0.2433 R 2 = 0.9751 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Flow (m 3 /s) R esidence Time ( d ays) Figure 6.39 Flow versus Residence Time for Mission River Segment 3 180 y = -0.3133Ln(x) + 1.2228 R 2 = 0.9863 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 012345678910 Flow (m 3 /s) Resi dence T ime (d ays) Figure 6.40 Flow versus Residence Time for Mission River Segment 4 181 y = -0.106Ln(x) + 0.395 R 2 = 0.9833 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0123456789 Flow (m 3 /s) Resi dence T ime (d ays) Figure 6.41 Flow versus Residence Time for Mission River Segment 5 A flow CDF plot was created for each segment. The flow CDF was based on USGS gauge daily mean streamflow data. The USGS gauge station 08189300 is on Medio Creek Segment 1, so a flow CDF was found directly for Segment 1, and daily mean streamflow data is available from 1962 to 2004. The flow CDF function for Medio Creek Segment 1 is shown in Figure 6.42. 182 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 0 0. 03 0. 06 0. 09 0. 12 0. 15 0. 18 0. 21 0. 24 0. 35 0. 5 0. 65 0. 8 0. 95 10 25 2000 Daily Flow, q (m 3 /s) Figure 6.42 Flow Cumulative Distribution Function for Segment 1 (USGS Data; 1962- 2004) The flow CDF for Mission River Segment 2 is based on USGS gauge daily mean streamflow data that are available from 1939 to 2004. The USGS gauge station 08189500 is on Mission River Segment 2, so a flow CDF was found directly for Segment 2. The flow CDF for Mission River is shown in Figure 6.43. 183 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 0 0. 15 0. 3 0. 45 0. 6 0. 75 0. 9 1. 1 1. 4 1. 7 2 3. 5 5 6. 5 8 9. 5 50 200 Daily Flow, q (m 3 /s) Figure 6.43 Flow Cumulative Distribution Function for Segment 2 (USGS Data; 1939- 2004) Because there is only one USGS gauge station on the Mission River, the flow CDF of Segment 2 was used to find the flow CDF for Segments 3, 4, and 5. The median flow for Segment 2 was found by finding the flow when the CDF equals 0.5, which is approximately 0.33 m 3 /s (Figure 6.43). A dimensionless CDF was found for Segment 2 by dividing the flow by the median flow (Figure 6.44). 184 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 0 5 10 15 20 q/q 0.5 Figure 6.44 Cumulative Distribution Function, q/q 0.5 for Mission River Segment 2 The flow for Mission River Segments 3, 4, and 5 was determined by using the following relationship: q = (q/q mean ) * q mean (6.9) Where: (q/q mean ) = dimensionless value obtained from Segment 2 q mean = mean flow of either Segment 3, 4, or 5 from water quality model Each watershed has an annual mean flow that was calculated using the runoff equations derived by Quenzer (2003), which are given in Section 5.1.2.4. The mean flow was calculated by averaging the cumulative flow at the upstream point of each segment (the 185 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 0. 00 1. 88 3. 76 5. 64 7. 52 9. 40 11. 29 13. 79 17. 56 21. 32 25. 08 43. 89 62. 70 81. 51 100. 32 119. 12 626. 97 2507. 88 Daily Flow, q (m 3 /s) sum of flow of all upstream watersheds) and the cumulative flow at the downstream point of each segment. The mean flow to Aransas River Segment 2 was calculated (Figure 6.18), and the same process was used to calculate the mean flow for each of the Mission River segments. After finding the mean flow for each segment, the flow CDF was calculated for each segment with Equation 6.9. The results for Segments 3, 4, and 5 are shown in Figures 6.45 - 6.47. Figure 6.45 Flow Cumulative Distribution Function for Mission River Segment 3 q mean = 4.14 m 3 /s 186 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 0 3 6 9 12 15 18 22 27 33 39 69 98 128 157 186 981 3924 Daily Flow, q (m 3 /s) q mean = 6.47 m 3 /s Figure 6.46 Flow Cumulative Distribution Function for Mission River Segment 4 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 0. 00 3. 96 7. 92 11. 8 8 15. 8 4 19. 8 0 23. 7 5 29. 0 3 36. 9 5 44. 8 7 52. 7 9 92. 3 8 131. 97 171. 56 211. 15 250. 74 1319. 7 0 5278. 7 9 Daily Flow, q (m 3 /s) q mean = 8.71 m 3 /s Figure 6.47 Flow Cumulative Distribution Function for Mission River Segment 5 187 Once the flow CDF was determined for all five segments, the RTD was found for each segment using the relationships between residence time and flow (Figures 6.37 - 6.41). The RTDs for Segments 1, 2, 3, 4, and 5 are shown in Figures 6.48 - 6.52. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 012345 Residence Time (days) Figure 6.48 Residence Time Distribution for Medio Creek Segment 1 188 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.00 0.50 1.00 1.50 2.00 Residence Time (days) Figure 6.49 Residence Time Distribution for Mission River Segment 2 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Residence Time (days) Figure 6.50 Residence Time Distribution for Mission River Segment 3 189 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 Residence Time (days) Figure 6.51 Residence Time Distribution for Mission River Segment 4 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Residence Time (days) Figure 6.52 Residence Time Distribution for Mission River Segment 5 190 The probability distributions were also found for the residence times of all segments along the Mission River and are shown in Figures 6.53 - 6.57. 0.00 0.05 0.10 0.15 0.20 0.25 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00 Residence Time (days) Figure 6.53 Probability Distribution of Residence Time for Medio Creek Segment 1 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 Residence Time (days) Figure 6.54 Probability Distribution of Residence Time for Mission River Segment 2 191 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Residence Time (days) Figure 6.55 Probability Distribution of Residence Time for Mission River Segment 3 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.00 0.20 0.40 0.60 0.80 1.00 1.20 Residence Time (days) Figure 6.56 Probability Distribution of Residence Time for Mission River Segment 4 192 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Residence Time (day) Figure 6.57 Probability Distribution of Residence Time for Mission River Segment 5 The RTDs were used as approximations for the corresponding SchemaLinks’ residence times in the Schematic Processor Model. Medio Creek Segment 1 corresponds to SchemaLink HydroID 118 while Mission River Segments 2, 3, 4, and 5 correspond to SchemaLink HydroIDs 116, 109, 115, 110, respectively, which are shown in Figure 6.58 14 . For this Schematic Processor Model, the most frequent residence time from the probability and residence time distributions is the residence time associated with the corresponding SchemaLinks, and these residence times are given in Table 6.11. 14 SchemaLink 133 corresponds to the neighboring watershed link, not Segment 3. 193 Figure 6.58 SchemaLinks of Corresponding Medio Creek and Mission River Segments 1 1 8 1 2 8 1 2 0 1 2 9 1 2 6 1 1 5 1 3 3 113 1 1 6 1 1 0 1 5 7 1 3 0 1 1 8 122 1 2 0 1 2 8 1 2 3 1 2 6 1 3 1 1 1 1 1 2 9 107 1 3 8 1 2 4 1 3 4 1 4 0 125 11 5 127 1 5 9 16 1 1 1 2 1 4 3 1 3 5 1 6 0 1 5 7 1 3 3 1 1 0 1 44 194 Table 6.11 Residence Times of Medio Creek and Mission River Segments for Schematic Processor Model Medio Creek/Mission River Segment SchemaLink HydroID Residence Time (days) 1 118 4.95 2 116 1.42 3 109 0.22 4 115 1.01 5 110 0.29 6.3.3.2.4 3d Model of Copano Creek: Residence Time Determination A 3d model of Copano Creek was created in HEC-RAS (USGS gauge stations 08189200) using the one USGS gauge station along the creek. The same methodology that was used to find the RTDs of the Aransas and Mission River segments was applied. To find the cross-section at the USGS gauge station, USGS gauge data (for USGS station 08189200) were downloaded from the USGS website from “Surface-Water Measurements”, which includes width, area, and stream flow measurements. The relationship between flow and width (Figure 6.59) was plotted for all available data (1967-2005 at USGS 08189200). 195 0 20 40 60 80 100 120 140 160 180 200 0.01 0.1 1 10 100 1000 10000 Flow (cfs) W i dt h ( ft Figure 6.59 Flow versus Width for USGS Station 08189200 The width of the channel at USGS gauge station 08189200 on Copano Creek was approximated to be 17 feet (Figure 6.59). The channel was assumed to have a square cross-section at the USGS gauge station; thus, the depth was calculated for each USGS measurement by dividing the measured area by the measured width for each measured stream flow. In Figure 6.60, the depth and width of each channel were plotted for all available measurements, and these data were used to approximate the depth in Copano Creek at a width of 17 feet. The approximate depth at USGS gauge station 08189200 with a width of 17 feet is approximately 1.74 feet (the average depth of the seven 17-ft width measurements that were made from the available USGS data). ~17 feet Width (ft) 196 0 1 2 3 4 5 6 7 8 9 10 051015 Width (ft) D e pt h ( f t The bank elevations are given for USGS gauge stations on the USGS website in “Surface-Water Measurements”, and the bank elevation is 17.25 feet above sea level for USGS 08189200. Because there is only one USGS gauge station along Copano Creek, other sources were used to estimate the cross-section at the most downstream point of Copano Creek where it drains into Copano Bay. The other sources that were used were aerial photographs and the RF1 Database. First, the sources were compared to the USGS approximation at the USGS gauge station location to see the percentage difference. The percentage differences are shown below in Table 6.12. ~17 feet 1.74 ft Figure 6.60 Width versus Depth (Square Cross-Section) for USGS Station 08189200 D e pt h (f t) 197 Table 6.12 Upstream Cross-Section Data Comparison (Copano Creek; USGS 08189200) Percent Error (%) Width (ft) Depth (ft) Width Depth RF1 68.65 0.34 303.82 80.47 Aerial 65.60 - 285.88 - USGS/RCMM Input 17.00 1.74 - - The aerial photograph is closer to the USGS approximation for width (Table 6.12); the RF1 file is the only known data source to approximate depth. The downstream dimensions using the available sources are shown in Table 6.13. Table 6.13 Downstream Cross-Section Data (Copano Creek) Width (ft) Depth (ft) RF1 68.65 0.34 Aerial 101.68 - The width of the channel was approximated (taking into account the 285.88% error) to be 26.35 feet, which was rounded to 50.00 feet (since the downstream width should be much wider than the upstream width) while the depth was approximated (taking into account the 80.47% error) to be 1.74 feet, which is the same depth as upstream at the USGS gauge station, which the depth should be larger downstream. Thus, the depth was approximated as 5.00 feet by assuming the same width:depth ratio that exists at the USGS gauge station. The bank elevation at the most downstream point was determined using the DEM. A summary of all the data needed for the RCMM toolbar for Copano Creek is shown in Figure 6.61. 198 After obtaining the NHD centerline feature class, USGS gauge station feature class, and the cross-section dimensions at two points, Merwade’s RCMM toolbar was used to generate a 3d model of Copano Creek 15 . Copano Creek was made into one 3d segment (Figure 6.61) because the residence time was needed for this one segment. Using the 3d Copano Creek model (HEC-RAS), a relationship was found between residence time and flow rate for the one segment (Figure 6.62). 15 Merwade generated the 3d river models for this project. Width = 17.00 feet Depth = 1.74 feet Bank Elevation = 17.25 feet (USGS) Width = 50.00 feet (Aerial) Depth = 5.00 feet (width:depth) Bank Elevation = 0.07 feet (DEM) Legend NHD_Centerline USGS_point TravelTime_RiverBreaks Figure 6.61 Summary of RCMM Toolbar Data Requirements (Copano Creek) 199 y = -0.9247Ln(x) + 0.8107 R 2 = 0.9661 0.00 1.00 2.00 3.00 4.00 5.00 6.00 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 Flow (m 3 /s) Resi dence T ime (d ays) Figure 6.62 Flow versus Residence Time for Copano Creek A flow CDF plot was created for the segment. The flow CDF was based on USGS gauge daily mean streamflow data that were available from 1970 to 2004. The USGS gauge station 08189200 is on Copano Creek, so a flow CDF was found directly for this segment (Figure 6.63). 200 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0 0. 02 0. 04 0. 06 0. 08 0. 1 1 2 3 4 5 50 150 1000 Daily Flow (m 3 /s) Figure 6.63 Flow Cumulative Distribution Function for Copano Creek (USGS Data; 1970-2004) Once the flow CDF was found, the RTD was found for each segment using the relationship between residence time and flow from Figure 6.62. Figure 6.64 shows the RTD for Copano Creek. 201 0.0 0.1 0.1 0.2 0.2 0.3 0.3 0.4 0.4 0.5 012345 Residence Time (days) Figure 6.64 Residence Time Distribution for Copano Creek The probability distribution was also found for the residence times of the Copano Creek segment and is shown in Figure 6.65. 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Residence Time (days) Figure 6.65 Probability Distribution of Residence Time for Copano Creek 202 The RTD was used as an approximation for the corresponding SchemaLink’s residence time in the Schematic Processor Model. Copano Creek Segment 1 corresponds to SchemaLink HydroID 138. For this Schematic Processor Model, the most frequent residence time from the probability and residence time distributions is the residence time associated with the corresponding SchemaLink, and the residence time is given in Table 6.14. Table 6.14 Residence Time of Copano Creek for Schematic Processor Model Copano Creek River Segment SchemaLink HydroID Residence Time (days) 1 138 5.069 6.3.3.2.5 Determination of Residence Times of Remaining SchemaLinks (Calibration) The preliminary residence times of the remaining SchemaLinks (i.e., those that were not modeled as 3d main river channels) were found by using the following relationship: Travel Time = Flow Length / Velocity. The procedure for calculating the initial residence time for each of the remaining river segments in the watershed is described in Appendix 6.2. After the initial residence times of the SchemaLinks were found, the Schematic Processor Model was calibrated at each of the bacterial monitoring stations, adjusting the residence times of the SchemaLinks that were not represented by 3d HEC-RAS models. The Schematic Processor Model needs to be able to accurately model what is actually occurring in the Copano Bay watershed. The bacterial loadings and concentrations calculated from the model were compared to existing bacterial monitoring data to calibrate the model. The Schematic Processor Model is calibrated to the median fecal coliform bacteria concentrations of the bacterial monitoring data provided by TCEQ from 1999 to 2005. 203 Three bacterial monitoring stations (with fecal coliform monitoring data) exist along the Aransas River (Stations 12948, 12952, 17592) and four bacterial monitoring stations exist in Copano Bay Segment 2 (Stations 12945, 14783, 14787, 14788), which are shown in Figure 6.66. The watersheds that are shown in Figure 6.66 drain into Copano Bay Segment 2. Two bacterial monitoring stations exist along the Mission River (Stations 12943, 12944), and one bacterial monitoring station exists in Copano Bay at the Mission River outlet (Station 14797), which are shown in Figure 6.67. The watersheds that are shown in Figure 6.67 drain into Copano Bay Segment 3. Six bacterial monitoring stations are located in Copano Bay Segment 4 (Stations 14793, 14792, 14785, 14780, 14779, and 13404), which are shown in Figure 6.68. Four bacterial monitoring stations are located in Copano Bay Segment 1 (Stations 14790, 14784, 14782, 13405), which are shown in Figure 6.68. 204 ") ") ") ") ") ") ") ") ") ") kj kj kj kj kj Legend kj USGS Gauge Stations ") Bacteria Monitoring Stations Sta. 12952 USGS 08189700 Sta. 17592 Sta. 12945 Sta. 14783 Sta. 14787 Sta. 14788 Sta. 12948 Figure 6.66 Bacterial Monitoring and USGS Stations along Aransas River 205 ") ") ") l l l l Legend l USGSGageStations ") BacteriaMonitoringStations_TCEQ Sta. 14797 Sta. 12944 USGS 08189500 USGS 08189300 Sta. 12943 Figure 6.67 Bacterial Monitoring and USGS Stations along Mission River 206 ") ") ") ") ") ") ") ") l l Legend l USGSGageStations ") BacteriaMonitoringStations_TCEQ USGS 08189200 Sta. 14792 Sta. 14793 Sta. 14785 Sta. 14780 Sta. 14779 Sta. 13404 Figure 6.68 Bacterial Monitoring and USGS Station along Copano Creek 207 ") ") ") ") ") ") ) ") ") ") ") ")") l l l l l Legend l USGSGageStations ") BacteriaMonitoringStations_TCEQ Sta. 14790 Sta. 14784 Sta. 14782 Sta. 13405 Figure 6.69 Bacterial Monitoring and USGS Stations along Copano Bay Segment 1 208 To calibrate the model for the Copano Bay watershed, the residence time, τ, parameter was adjusted for the river segments that were not considered main river channels to reflect the existing median fecal coliform monitoring data at each station labeled in Figures 6.66 - 6.69. The bacterial loadings of the SchemaLinks (that transport the greatest bacterial loads) directly upstream of the bacterial monitoring station are most influential of the quality of the river water since the bacteria do not have a sufficient amount of time to decay. For this calibration, a decay coefficient of 2 days -1 was assumed. A model was created in Microsoft Excel using the Solver add-in function and the calculated bacterial loadings; using this model, we solved for the residence time of the SchemaLink that most directly affects the bacterial concentration at the bacterial monitoring station. The most upstream bacterial monitoring station locations were calibrated first; then, the next downstream bacterial monitoring station was calibrated, adjusting only those parameters that do not affect the calibration of the upstream bacterial monitoring stations. The SchemaLinks’ parameters (i.e., residence time) that can be adjusted for each bacterial monitoring station along the Aransas River are shown in Figure 6.70. Each station has a corresponding color that identifies the SchemaLinks’ parameters that can be adjusted. However, only the most influential SchemaLink (i.e., directly upstream of the bacterial monitoring station and transporting the greatest bacterial load) was adjusted at each station. After calibration was complete at each bacterial monitoring station in the Copano Bay watershed, the Schematic Processor Model was used to model the median fecal coliform concentration at each station. The residence time of each SchemaLink is shown in Figure 6.71. 209 Legend Bacterial Monitoring Stations " ) 17592 " ) 12952 " ) 12948 " ) 12945 " ) 14783 ") ") ") ") ") Figure 6.70 Nodes/Links’ Parameters that can be Varied for Each Station along Aransas River 210 6.3.3.3 Volume of Copano Bay Segments A rough approximation was made for calculating the individual volumes of the four Copano Bay segments that are shown in Figure 6.3. A bathymetry map, which shows the depth of different parts of the Bay, was used (Figure 6.72) to determine the average depth of each segment (Ward and Armstrong, 1997). The surface area of each of 3 4 5 4 . 9 5 1.5 1 . 5 1 0 . 7 2.3 0.6 86 2 . 4 0 3 6 8 2 1 . 0 1 3. 0 4 1 3 8 3 2 0 . 2 1 1 . 8 5 9 2 9 1 1 . 4 3 8 4 9 8 1 2.8 54 15 4 0 . 2 9 0 . 2 2 0.21 2 0 . 7 0 4 1 7 5 0 . 0 1 0 . 6 2 1 0 2 7 3 1 . 5 5 1 . 5 4 5 1 Figure 6.71 Residence Times (days) of SchemaLinks (k = 2 day -1 ) for Schematic Processor Model 211 the water segments was found by opening the attribute table of each polygon feature class and looking under the field “ShapeArea”, which gives the area of the segment in square meters. For each of the four water segments (defined in Section 6.3.1.1), the depth was found by weighting the depth based on area covered by each of Ward’s segments that make up the segment. The volume was then calculated by multiplying the weighted depth by the surface area of each segment. The areas, depths, and volumes of each of the segments are given in Table 6.15. Figure 6.72 Bathymetry Map of Copano Bay 212 Table 6.15 Area, Depth, and Volume of Copano Bay Segments Copano Bay Segment Original Segmentation Area (m 2 ) Depth (m) Volume (m 3 ) PB1 3,078,276 1 PB2 14,196,885 1 CP04 20,001,720 2.8 CP07 14,346,749 2 1 CP08 17,434,891 69,062,585 2 1.98 136,843,257 AR1 4,524,570 1 2 CP03 27,329,952 31854530 2 1.86 59,184,474 M1 650,022 0.5 M2 14,166,339 1 CP05 18,079,089 1.7 3 CP06 18,108,840 51,014,845 1.7 1.49 76,010,829 CP01 846,994 0.5 CP02 16,493,027 2 CP09 20,131,477 2.9 4 CP10 14,690,137 52,170,162 2.8 2.55 132,923,219 6.3.3.4 Flow of Copano Bay Segments The cumulative flow to each Copano Bay water segment was determined by finding the total annual upstream flow of the upstream watersheds to each Copano Bay segment. The cumulative annual runoff for each watershed was calculated using the runoff per watershed that was determined in Section 5.1.2.6 and runoff equations (Quenzer, 2003) that relate runoff to precipitation and land use. The cumulative runoff (m 3 /year) to each Copano Bay segment is given in Table 6.16. 213 Table 6.16 Cumulative Annual Runoff to Copano Bay Segments 6.3.4 Implementation of Schematic Processor Once the Schematic Network is created and the parameter values are input to the corresponding fields in the attribute tables of SchemaNode and SchemaLink, “Process Schematic” can be run. See Appendix 6.3 for the procedure on how to use “Process Schematic” and how to interpret the results. All of the parameters and values of each corresponding SchemaNode in the Schematic Network are given in Table 6.17. “HydroID” is the unique identifier for each SchemaNode. “FeatureID” is the unique identifier for each watershed, so the SchemaNodes that have a FeatureID are nodes that represent watersheds. “SrcType” is the type of SchemaNode that the node represents; (“1” = watershed, “2” = drainage junction, “3” = Copano Bay.) “IncVal” is the sum of all the bacterial loadings from all sources (calculated in Section 5.6) that are input into the model at the specific SchemaNode. “Flow (m 3 /yr)” is the cumulative annual runoff for each Copano Bay water segment (calculated in Section 6.3.3.4). “Volume (m 3 )” is the volume of each Copano Bay water segment (calculated in Section 6.3.3.3). “Die-off rate (days -1 )” is the decay coefficient associated with each node. “Cumulative Runoff (m 3 /yr)” is the cumulative runoff of all the upstream watersheds that drain to the node of interest. The HydroIDs of each of the SchemaNodes in the Schematic Network are shown in Figure 6.73. Copano Bay Water Segment Cumulative Annual Runoff (m 3 /year) 1 25,230,928 2 251,731,639 3 275,169,044 4 72,825,125 214 Table 6.17 SchemaNode Attribute Table (Calibrated to Median Fecal Coliform Concentrations) Hydro ID Feature ID Src Type IncVal Flow (m 3 /yr) Volume (m 3 ) Die-off Rate (d -1 ) Cumulative Runoff (m 3 /yr) 61 0 2 0.00E+00 0 0 2 1.51E+07 62 0 2 3.22E+13 0 0 2 3.30E+07 63 0 2 0.00E+00 0 0 2 1.17E+08 64 0 2 4.79E+14 0 0 2 3.47E+07 65 0 2 7.96E+11 0 0 2 1.34E+08 66 0 2 0.00E+00 0 0 2 2.75E+08 67 0 2 1.48E+04 0 0 2 2.52E+08 68 0 2 0.00E+00 0 0 2 3.47E+07 69 0 2 3.37E+11 0 0 2 1.95E+07 70 0 2 9.17E+11 0 0 2 1.35E+08 71 0 2 0.00E+00 0 0 2 1.27E+08 72 0 2 0.00E+00 0 0 2 3.63E+07 73 0 2 0.00E+00 0 0 2 1.22E+08 74 0 2 0.00E+00 0 0 2 1.27E+08 75 0 2 0.00E+00 0 0 2 9.84E+07 77 45422 1 2.95E+16 0 0 2 0.00E+00 78 45413 1 1.09E+16 0 0 2 0.00E+00 79 45404 1 1.18E+16 0 0 2 0.00E+00 80 45419 1 6.25E+16 0 0 2 0.00E+00 81 45421 1 8.60E+15 0 0 2 0.00E+00 82 45417 1 4.45E+16 0 0 2 0.00E+00 83 45408 1 3.07E+16 0 0 2 0.00E+00 84 45415 1 1.17E+16 0 0 2 0.00E+00 85 45409 1 1.58E+15 0 0 2 0.00E+00 86 45426 1 7.67E+15 0 0 2 0.00E+00 87 45416 1 1.27E+16 0 0 2 0.00E+00 88 45405 1 8.17E+15 0 0 2 0.00E+00 89 56831 1 1.16E+15 0 0 2 0.00E+00 90 0 2 1.75E+12 0 0 2 1.72E+07 91 0 2 0.00E+00 0 0 2 5.45E+06 92 0 2 6.98E+14 0 0 2 3.79E+07 93 0 2 0.00E+00 0 0 2 2.52E+07 94 0 2 0.00E+00 0 0 2 5.32E+06 95 0 2 0.00E+00 0 0 2 6.75E+07 96 56830 1 6.51E+15 0 0 2 0.00E+00 97 45412 1 6.42E+13 0 0 2 0.00E+00 215 98 0 2 0.00E+00 0 0 2 3.63E+07 99 45423 1 9.81E+14 0 0 2 0.00E+00 100 45418 1 1.67E+15 0 0 2 0.00E+00 101 0 2 0.00E+00 0 0 2 3.50E+05 102 45425 1 4.02E+13 0 0 2 0.00E+00 103 45414 1 6.89E+15 0 0 2 0.00E+00 104 45410 1 1.02E+15 0 0 2 0.00E+00 105 45406 1 7.40E+14 0 0 2 0.00E+00 153 0 3 1.48E+09 2.75E+08 7.60E+07 2 2.75E+08 154 0 3 2.22E+09 2.52E+08 5.92E+07 2 2.52E+08 155 0 3 3.96E+11 2.52E+07 1.37E+08 2 2.52E+07 156 0 3 1.22E+11 7.28E+07 1.33E+08 2 7.28E+07 99 98 97 96 95 94 93 92 91 90 89 88 87 86 85 84 83 82 81 80 79 78 77 75 74 73 72 71 70 69 68 67 66 6564 63 62 61 156 155 105 104 103 102 101 100 154 153 Figure 6.73 HydroIDs of SchemaNodes 216 All of the parameters and values of each corresponding SchemaLink in the Schematic Network are shown in Table 6.18. “HydroID” is the unique identifier for each SchemaLink. “FromNodeID” is the HydroID of the upstream SchemaNode, and “ToNodeID” is the HydroID of the downstream SchemaNode. “LinkType” is the type of SchemaLink that the link represents; (“1” = watershed travel, “2” = river, “3” = Copano Bay.) “Die-off rate (days -1 )” is the decay coefficient associated with each link. “Residence Time (days)” is the amount of time the bacteria are allowed to decay in each water segment (SchemaLink), which was determined in Section 6.3.3.2. The HydroID of each of the SchemaLinks in the Schematic Network is shown in Figure 6.74. Table 6.18 SchemaLink Attribute Table (Calibrated to Median Fecal Coliform Concentrations) HydroID FromNodeID ToNodeID LinkType Die-off Rate (days -1 ) Residence Time (days) 106 61 62 2 2 2.00 107 75 63 0.69 108 68 64 2 2 0.05 109 74 65 0.22 110 70 66 2 2 0.29 111 91 62 4.00 112 63 67 2 2 0.21 113 62 68 0.21 114 69 67 2 2 0.01 115 65 70 1.01 116 73 71 2 2 1.42 117 98 72 0.05 118 90 73 2 2 4.95 119 71 74 0.05 120 64 75 2 2 1.51 121 101 67 2 2 0.01 122 92 67 1.50 123 77 90 1 2 3.00 217 124 78 91 1 2 5.00 125 79 61 2.85 126 83 75 1 2 3.04 127 86 92 4.00 128 80 73 1 2 2.40 129 82 66 1.86 130 97 70 1 2 0.18 131 81 98 3.00 132 99 71 1 2 1.00 133 100 65 1 2 2.00 134 87 67 1.44 135 88 93 1 2 0.70 136 89 94 0.62 137 96 95 1 2 0.70 138 72 95 2 5.00 139 102 101 1 2 1.00 140 84 62 5.00 141 85 68 1 2 3.12 142 103 63 1 2 2.30 143 104 69 1 2 1.50 144 105 72 1 2 1.50 157 66 153 3 0.00 158 67 154 2 0.00 159 93 155 3 0.00 160 95 156 2 0.00 161 94 156 3 0.00 218 1 1 8 122 1 2 0 1 2 8 1 2 3 1 2 6 1 3 1 1 1 1 129 107 1 3 8 1 2 4 1 3 4 1 4 0 1 3 7 142 125 1 1 5 127 1 5 9 1 6 1 1 1 2 1 4 3 1 3 5 1 5 8 1 6 0 1 5 7 1 3 3 113 1 3 6 1 1 6 1 1 0 1 4 4 10 6 1 3 0 Figure 6.74 HydroIDs of SchemaLinks 219 6.4 RESULTS The current loadings to each of the water segments (Aransas River Tidal, Mission River Tidal, and Copano Bay) are given in Chapter 8 of this report. After one simulation of the Schematic Processor was run, the bacterial loading was converted to bacterial concentration (CFU/100mL) at each of the SchemaNodes. The modeled fecal coliform concentration at each SchemaNode is shown in Figure 6.75. Legend SchemaNode cfu_100mL 2.00 - 14.00 14.01 - 43.00 43.01 - 434.42 434.43 - 2378.70 2378.71 - 7919.26 Figure 6.75 Fecal Coliform Concentrations (CFU/100mL) – Schematic Processor Results 220 The fecal coliform concentrations of SchemaNode Src Types 2 and 3 and the modeled concentrations versus the measured median fecal coliform concentrations at the bacterial monitoring station locations are shown in Table 6.19. 221 Table 6.19 Modeled versus Existing Fecal Coliform Concentrations: Schematic Processor Model SchemaNode HydroID Bacterial Monitoring Station ID Modeled Fecal Coliform Concentration (CFU/100mL) Measured Median Fecal Coliform Concentration (CFU/100mL) 61 17592 260.00 260 62 101.27 63 79.60 64 1444.23 65 94.09 66 405.26 67 345.32 68 12952 71.98 72 69 262.88 70 12943 46.57 47 71 128.20 72 154.59 73 418.77 74 12944 116.00 116 75 12948 96.01 96 90 434.42 91 9.04 92 1847.85 93 7919.26 94 6318.24 95 2378.70 98 58.72 101 1555.86 153 14797 2.00 2 154 12945, 14783, 14787, 14788 2.00 2 155 13405, 14782, 14784, 14790 2.00 2 156 13404, 14779, 14780, 14785, 14792, 14793 2.00 2 As shown in Table 6.19, the modeled fecal coliform concentrations from the Schematic Processor Model match up very well with the existing median fecal coliform concentrations. 222 This is one simulation of the Schematic Processor model (Figure 6.75 and Table 6.19), and this simulation is supposed to be representative of average annual conditions. That is why the modeled concentrations are compared to the median fecal coliform concentrations of monitoring data from 1999 to 2005. The concentrations in all segments of the Bay are constant as modeled (at 2 CFU/100mL.) This is consistent with the monitoring data in each portion of the Bay. All segments of the Bay are currently meeting the median fecal coliform standards for oyster harvesting use (< 14 CFU/100mL). Thus, the Schematic Processor Model was not used to determine the load reductions necessary to meet water quality standards because it does not generate a probability distribution (the criterion of 90 th -percentile < 43 CFU/100mL is exceeded in two Copano Bay segments.) The Monte Carlo Simulation Model (described in Chapter 7) is used to determine load reductions. Thus, the results of the Schematic Processor model were used to explore the impact of the different bacteria sources on the four segments of Copano Bay. The bacterial loading from each source, as calculated in Chapter 5, was input to the Schematic Processor model (bacterial loading of a source = “IncVal”, Schematic Processor is run, and then the “PassedVal” and “TotVal” fields were stored as the bacterial loading impact from that particular source). The effects of bacterial decay along the SchemaLinks as the bacteria travel from the upstream watersheds (bacterial loadings not shown), along the rivers, and to Copano Bay are shown in Figure 6.76. 223 Figure 6.76 Bacterial Loadings (from Sources) to SchemaNodes SrcTypes 2 and 3 Legend 7.7e+014 Non-Point (Urban, Forest, etc.) Birds Human (Septic Systems) Human (WWTPs) Cattle Horses Goats Sheep Hogs Chickens Layers 224 The watersheds that will most influence the quality of the Bay are the watersheds directly upstream and adjacent to the Bay because the bacteria have not had as long to decay as the bacteria from watersheds farther upstream of the Bay. Looking at the watersheds directly adjacent to the Bay (Figure 6.76), it can be seen that cattle are the greatest bacteria contributors to all Copano Bay segments. The total bacterial loadings (CFU/year) to each of the Copano Bay water segments after decay and mixing in the CFSTRs are modeled, are shown in Figure 6.77. 225 Legend Non-Point (Urban, Forest, etc.) Birds Human (Septic Systems) Human (WWTPs) Cattle Horses Goats Sheep Hogs Chickens Layers 2.6E+12 cfu/year Figure 6.77 Bacterial Loading (from Sources) to Copano Bay (CFU/year) 1 2 3 4 226 As shown in Figure 6.77, Copano Bay Segments 2 and 3 have the highest bacterial loads compared to the other segments. However, due to the larger cumulative flow in these portions of the Bay and an increased number of upstream watersheds draining to these portions of the Bay, the median fecal coliform concentration is the same in each segment. Recall that Equation 6.2 is used to calculate the concentration of fecal coliform in each segment of the Bay. The percent distribution of the bacterial loadings to all four segments of Copano Bay is shown in Figure 6.78. In all four Copano Bay segments, cattle are the dominant bacterial loading contributor based on the model and bacterial loading calculations. 227 Legend Non-Point (Urban, Forest, etc.) Birds Human (Septic Systems) Human (WWTPs) Cattle Horses Goats Sheep Hogs Chickens Layers Figure 6.78 Percent Distribution of Bacterial Loading Sources (Output) 1 4 3 2 228 In Section 5.6, the total bacterial loadings input into the model are described; with this information, the effects of the bacterial loadings from different sources on the fecal coliform concentration in the Bay can be examined (since the effects of bacterial transport have been implemented). For instance, the avian loading, which is the only source load that is applied directly to the Bay, does not have as great an impact on the Bay; (see Figure 5.25). This seems reasonable since the magnitudes of the bacterial loadings from avian sources (Section 5.3.3) are so much less than the magnitudes of the bacterial loadings from cattle (Section 5.2.3). Thus, even though avian loads are applied directly to the Bay and one upstream watershed, the effects of these loadings are negligible compared to other sources (see Figure 6.78). One of the watersheds that drains into Copano Bay Segment 4 (Figure 6.76) contains fecal coliform predominantly from human sources (i.e., septic systems), as shown in Figure 5.25. However, after we account for the effects of bacterial transport and the loadings from the other upstream watersheds that drain to Segment 4, the bacterial loadings from cattle dominate. Septic system bacterial loading of one of the upstream watersheds of Segment 4 dominates (as shown in Figure 6.76), but the magnitude of the bacterial loading is significantly smaller than the other upstream watersheds that also drain into Segment 4 (where cattle is the main contributor). Thus, even though the loadings from one of the upstream watersheds is predominantly from malfunctioning septic systems, the overall bacterial load from that particular watershed is significantly lower than the bacterial loads from other contributing upstream watersheds. Malfunctioning septic systems appear to have an impact on Segments 1 and 4. However, Segments 1 and 4 are currently not exceeding Texas Surface Water Quality Standards for fecal coliform oyster harvesting use. The other results from the Schematic Processor Model are discussed in Chapter 8. 229 Chapter 7: Modeling of Bacterial Transport – Monte Carlo Simulation 7.1 BACKGROUND The 90 th -percentile fecal coliform concentration needs to be less than 43 CFU/100mL to meet oyster water use standards in Copano Bay. A second model was created to predict probability distributions of fecal coliform since the Schematic Processor Model does not have this capability. The second model (created by Ernest To, CRWR) conducts a Monte Carlo simulation analysis for the Copano Bay watershed and models bacterial transport the same as the Schematic Processor Model (applying first- order decay and treating the Bay as four CFSTRs). Monte Carlo analysis picks random numbers from a probability distribution associated with uncertain parameters to simulate random behavior based on the parameter distributions. Conducting Monte Carlo simulations generates multiple outcomes (i.e., fecal coliform concentrations) by repeatedly sampling values from probability distributions of uncertain parameters and plotting the results as a probability distribution. If the model accurately represents what is occurring in the watershed, then the output distribution (i.e., modeled fecal coliform concentrations) should match the actual distribution (i.e., measured fecal coliform concentrations) at the specific point of interest (i.e., bacterial monitoring station). A schematic diagram was created to show how the Monte Carlo simulation analysis works for this project and is shown in Figure 7.1. Shown are the parameters and inputs associated with one output location (i.e., bacterial monitoring station 17592). Because only one SchemaLink and SchemaNode exist upstream of Station 17592, there is only one k-distribution, one bacterial loading distribution, and one residence time upstream of this location. The more SchemaLinks and SchemaNodes that are upstream 230 of a bacterial monitoring station, the more inputs and parameters can affect the output distribution results. 231 B 1 Monte Carlo Simulation Model Input Parameters Bacterial Loadings (Chapter 5) k Decay Coefficient (Section 6.3.3.1) τ Residence Time (Section 6.3.3.2) Output Modeled FC Concentrations at Station 17592 ") B 1 k, τ B 2 Measured FC Concentrations at Station 17592 B 2 Figure 7.1 Monte Carlo Simulation Conceptual Diagram 232 The output distribution (shown in Figure 7.1) is then compared to the measured distribution from the bacterial monitoring station to determine if the model accurately characterizes the Copano Bay system, which it does in this situation. The schematic diagram (shown in Figure 7.1) is cited in the following sections to clarify the implementation of the model. The Monte Carlo Simulation Model was created in Microsoft Excel 2003. The original intention was to perform the Monte Carlo analysis in ArcGIS Model Builder. However, Excel was chosen instead because it has built-in procedures that can sample from different probability distributions, it has built-in graphing capabilities, and the analysis takes a shorter amount of time to run because it is faster to update spreadsheets than to access and update databases (To, 2005). This chapter discusses how the Monte Carlo Simulation Model was applied and used for the Copano Bay project. However, the procedure for creating the Monte Carlo Simulation Model and the programming code and macros behind the model will not be discussed in this report. The user interface of the Monte Carlo Simulation Model, which are all of the worksheets used in Microsoft Excel, and explanations of the important features and parameters used in running the model are shown in Appendix 7.1. For the Copano Bay project, the Monte Carlo Simulation Model was used to model existing bacterial concentration conditions at all SchemaNode locations as well as to determine the load reductions necessary to reduce the fecal coliform bacteria concentrations to meet Texas Surface Water Quality Standards. 233 7.2 METHODOLOGY To use the Monte Carlo Simulation Model, the following steps were completed: 1. Schematic Network of the Copano Bay watershed was created (Section 6.3.1). 2. Annual average bacterial loadings were calculated (Chapter 5). 3. The parameters (or parameter distributions) of each SchemaNode and SchemaLink were determined through calculations and/or calibration. The Monte Carlo Simulation Model was calibrated at each bacterial monitoring station to match the existing bacterial monitoring data (from 1999-2005). Once the model was calibrated, the model was used to determine the load reductions required from various bacterial sources to attain fecal coliform water quality standards in each of the water segments. 234 7.3 PROCEDURE OF APPLICATION 7.3.1 Determination of Parameters The following sections describe how the parameters were determined for each SchemaLink and SchemaNode in the Schematic Network for the Monte Carlo Simulation Model. 7.3.1.1 Bacterial Loading Recall that the bacterial loading was calculated by using the following equation: L = Q * C, where L = bacterial loading, Q = flow rate, and C = fecal coliform concentration. Based on this equation, it can be seen that the bacterial loading to all of the water segments is going to vary throughout the year. There are many factors (e.g., precipitation and runoff), which would affect the bacterial loadings (the input into the model) in the watershed. For example, when the precipitation and runoff in the watershed are very high, then the bacterial loadings dramatically increase and affect the quality of the receiving waters. The calculations of the average annual bacterial loadings from the point and non- point bacteria sources in the watershed are described in Chapter 5. From the data analysis in Chapter 4, the measured bacterial concentrations at the bacterial monitoring stations are very similar to a lognormal distribution; thus, the bacterial loadings for the Monte Carlo Simulation Model are assumed to be log normally distributed. At each SchemaNode in which bacteria are input into the model (e.g. SchemaNodes SrcType =1, which are watersheds), a lognormal distribution of the bacterial loadings was determined. While running a simulation of the Monte Carlo Simulation Model, the model randomly selects a bacterial loading at each SchemaNode 235 based on these lognormal distributions. For example, looking at Figure 7.1, there is only one SchemaNode that has a bacterial loading input into the model that would affect the water quality at Station 17592; thus, the Monte Carlo Simulation Model randomly selects one bacterial loading, B 1 , from the lognormal distribution and then simulates the decay of bacterial transport by randomly selecting one decay coefficient from the k-distribution (described in Section 7.3.1.2) and the given residence time for the upstream SchemaLink (described in Section 7.3.1.3) to obtain one output fecal coliform concentration by applying the equation: [B 1 *exp(-kτ)]/Q = modeled fecal coliform bacterial concentration 16 . To get a distribution of modeled fecal coliform concentrations in the output, the model is run multiple times. Two main parameters are used to create the lognormal distributions at each bacterial loading source (e.g., watershed or drainage point): the median of the bacterial loadings and a multiplication factor that is associated with the standard deviation and spread of the distribution. The median of each lognormal distribution was assumed to be the average annual bacterial loading that was calculated in Chapter 5, and the multiplication factor is described in the following 17 . Microsoft Excel cannot directly sample from a lognormal distribution with a given mean and standard deviation, but Excel can sample from a unit normal distribution (a.k.a, z-curve). As a result, the lognormal distribution has to undertake a series of transformations to create a unit normal distribution in which the mean = 0 and the standard deviation = 1. The bacterial load distributions are modeled as lognormal distributions (Figure 7.2). 16 Q is the cumulative annual runoff upstream of the point of interest, and the calculation of Q for each watershed is described in Section 5.1.2.6. 17 Ernest To, CRWR, provided the information necessary to describe the Excel process and define the multiplication factor. 236 The step-by-step procedure for the transformation of the lognormal distribution (Figure 7.2) into a unit normal distribution is described below: 1. Normalize the lognormal distribution with the median to obtain a lognormal distribution with a median of 1 (Figure 7.3). Median = μ B B = Bacterial Load f Figure 7.2 Bacterial Load Distribution Modeled as Lognormal Distribution 237 2. Take the natural-log, ln( ), of the lognormal distribution (Figure 7.3) to transform the distribution into a normal distribution with mean = median = 0 and standard deviation, σ ln(B/μB (Figure 7.4). The standard deviation of this normal distribution is also referred to as the multiplication factor. Thus, the multiplication factor = σ ln(B/μB). Median = 1 B/μ B f Figure 7.3 Lognormal Distribution with Median = 1 (Normalized Lognormal Distribution by μ B ) 238 3. Normalize the normal distribution (Figure 7.4) by the multiplication factor, σ ln(B/μB) , to obtain a unit normal distribution with standard deviation of 1 (a.k.a, the z-curve). Microsoft Excel can randomly sample from a unit normal distribution (Figure 7.4), so to obtain the fecal coliform bacterial loading from the original lognormal distribution (Figure 7.2), Excel needs to work backwards (through Steps #1-3) from the normal distribution (Figure 7.4). The formula used in Excel to find the sampled bacterial loading from the original lognormal distribution (Figure 7.2), B 1 , is given below: B 1 = μ B x EXP ( σ ln (B/μ B ) x NORMINV (ABS ( RAND( ),0,1 ) ) 0 σ ln(B/μB) ln(B/μ B ) Figure 7.4 Normal Distribution with Mean = Median = 0 and Standard Deviation, σ ln(B/μB) 239 The multiplication factor can be related to the coefficient of variation, which is the standard deviation divided by the mean. Thus, the greater the multiplication factor the greater the spread of the distribution of bacterial loadings. The multiplication factor was one of the parameters that were adjusted to try to match the modeled output fecal coliform distributions to the measured fecal coliform distributions at each bacterial monitoring station. The results of the calibration are given in Section 7.3.2. 7.3.1.2 Decay Coefficient The procedure for how a k-distribution was determined for a portion of the model is described in Section 6.3.3.1. Due to lack of data and studies in the Copano Bay watershed with regard to bacterial decay, the k-distribution given in Table 6.3 was used for all the SchemaLinks in the Schematic Network. The k-distribution varies from approximately 2 to 2.5 days -1 . A beta distribution was used to represent the k-distribution for the Monte Carlo Simulation Model as explained in Appendix 7.1. The beta distribution that was chosen for the analysis was α = 2 and β = 2, which is shown in Figure 7.5 (Wikipedia, 2006). The lower and upper limits are 2 and 2.5 days -1 , respectively. Ignoring the large storm event (see Table 6.3; k = 0.8 days -1 ), the average of the remaining four decay coefficients is approximately 2.25 days -1 . Thus, instead of skewing the distribution, it was assumed that the likelihood of the decay coefficient being 2.25 days -1 is higher than the probability of the decay coefficient being either 2 or 2.5 days -1 . Thus, the probability of the decay coefficient being 2.25 days -1 was assumed highest, and the probability of the decay coefficient being 2 or 2.5 days -1 was lowest. Each simulation the Monte Carlo Simulation Model randomly 240 selects a decay coefficient for each SchemaLink based on this probability distribution in the Copano Bay watershed. 7.3.1.3 Residence Time The initial residence times, which are shown in Figure 6.71, for the SchemaLinks in the Monte Carlo Simulation Model were the residence times that were used in the Schematic Processor Model. The reason that a distribution was not found for the residence time of each SchemaLink (particularly since precipitation and flow vary greatly throughout the year) was because the bacterial loadings have lognormal distributions associated with them. Running a Monte Carlo simulation on both parameters may counteract the intended effect. For instance, in one simulation a high bacterial loading may be randomly selected from the upstream SchemaNode (Figure 7.1), which would indicate a high flow event; however, if there is a probability distribution associated with the residence time of the SchemaLink, then the model may select a longer residence time, which would indicate a low flow event. Thus, the effects of a large storm event with high bacterial loadings would be minimal and non-realistic if the residence time was relatively Figure 7.5 Beta Distribution 2.1 2.2 2.3 2.4 2.5 2 241 large because the bacteria would have a significant amount of time to decay. In an actually large storm event, the residence time should be much smaller, allowing minimal amount of time for decay. To eliminate this ‘counteracting’ effect, each SchemaLinks’ residence time was held constant for all simulations. Some of the residence times were adjusted in an attempt to match the modeled fecal coliform distribution to the measured fecal coliform concentration distribution at each bacterial monitoring station. As in the Schematic Processor Model, only the residence times of the most influential SchemaLinks (i.e., those directly upstream of a bacterial monitoring station and transporting the highest bacterial load) were adjusted at each station. The results of the calibration are given in Section 7.3.2. It should be mentioned that residence time distributions (RTDs) for the mainstreams were determined in Section 6.3.3.2. A separate Monte Carlo analysis was conducted in which the bacterial loadings (calculated in Chapter 5) were held constant, and the RTDs of the corresponding SchemaLinks were applied to the model. In this analysis, the bacterial loadings were held constant to eliminate the ‘counteracting’ effect while distributions were applied to the residence times. However, this Monte Carlo analysis did not model the existing conditions as well as the Monte Carlo analysis in which the bacterial loadings were varied and the residence times remained constant, so this model was not used in our research. 7.3.1.4 Other Parameters In this Monte Carlo Simulation Model, the bacterial loadings and decay coefficients have probability distributions associated with them (as shown in Figure 7.1 and described in Sections 7.3.1.1 and 7.3.1.2.) The residence times of the SchemaLinks are held constant (Section 7.3.1.3.) 242 However, the remaining parameters, which are only associated with SchemaNode SrcType 3 (i.e., flow, volume, and decay coefficient) were held constant during all the analyses. Thus, for the Copano Bay segments, the parameters of flow (Section 6.3.3.4), volume (Section 6.3.3.3), and decay coefficient (2 days -1 ) are the same values that were determined in Chapter 6. 7.3.2 Calibration of Model The only two parameters of the model that were adjusted for calibration purposes (i.e., to match the modeled with the measured fecal coliform concentrations at each bacterial monitoring station) were the multiplication factor (described in Section 7.3.1.1) and the residence time of the SchemaLinks (described in Section 7.3.1.3.) This section describes how each portion of the model, based on water segment, was calibrated and shows the results of the calibration. The residence time of the SchemaLink transporting the highest bacterial load to the bacterial monitoring station greatly influences the median of the modeled concentrations. The residence time of the most influential SchemaLink was adjusted such that the modeled median matched the median of the measured data. The multiplication factor influences the shape of the curve (fecal coliform versus probability of exceedance) and 90 th -percentile modeled concentrations. The multiplication factor was adjusted such that the shape of the curve and the 90 th -percentile values matched between the model and the measured data. Thus, the combination of adjusting the residence times and multiplication factors in the model was conducted to match the modeled fecal coliform distributions to the measured fecal coliform distributions at each bacterial monitoring station. 243 7.3.2.1 Aransas River Above Tidal There is only one bacterial monitoring station along the Aransas River Above Tidal, but there is a bacterial monitoring station, Station 17592, with fecal coliform monitoring data upstream of the Above Tidal that will be analyzed first. The parameters of the SchemaNodes and SchemaLinks that were adjusted at Station 17592 are given in Tables 7.1 and 7.2, respectively. The locations of the SchemaNodes, SchemaLinks, Station 17592, and the results of the calibration are shown in Figure 7.3. Table 7.1 SchemaNode Adjusted Parameters for Calibration of Station 17592 SchemaNode (HydroID) Bacterial Loading Multiplication Factor 79 1.5 Table 7.2 SchemaLink Adjusted Parameters for Calibration of Station 17592 Residence Time (days) SchemaLink (HydroID) Initial (Schematic Processor Model) Final (Monte Carlo Simulation Model) 125 2.93 2.55 Only 9 fecal coliform concentration measurements were made at Station 17592 from 1999-2005 (Figure 7.6), and these measurements were used to calibrate the model at this location. 244 The parameters of the SchemaNodes and SchemaLinks that could have been and/or were adjusted at Station 12952, which is the next downstream bacterial monitoring station along the Aransas River Above Tidal, are given in Tables 7.3 and 7.4, ") ") 125 79 61 Cumulative Density Function (CDF) of Fecal Coliform Concentration (CFU/100mL) at Schemanode 61 0 10 20 30 40 50 60 70 80 90 100 0 500 1000 1500 2000 2500 3000 3500 4000 4500 cfu/100mL Percen t i le Monitoring data Model Model Median = 264.2 cfu/100mL 90th-percentile = 1782.2 cfu/100mL Contact recreation standard for median = 200 cfu/100mL Single sample should not exceed 400 cfu/100mL) Number of Simulations = 1000 2/3/2006 8 19 02 AM Monitoring data Median = 260 cfu/100mL 90th-percentile = 1805.7 cfu/100mL Figure 7.6 Modeled versus Measured Fecal Coliform Concentrations at Station 17592 245 respectively. The locations of the SchemaNodes, SchemaLinks, Station 12952, and the results of the calibration are shown in Figure 7.7. Table 7.3 SchemaNode Adjusted Parameters for Calibration of Station 12952 SchemaNode (HydroID) Bacteria Loading Multiplication Factor 61 1.7 62 1.7 78 1.7 84 1.7 85 1.7 91 1.7 Table 7.4 SchemaLink Adjusted Parameters for Calibration of Station 12952 Residence Time (days) SchemaLink (HydroID) Initial (Schematic Processor Model) Final (Monte Carlo Simulation Model) 106 2 2 111 4 4 113 0.212 0.212 124 5 5 140 5 141 3 3.3* * Parameters that were adjusted for calibration 246 7.3.2.2 Aransas River Tidal Station 12948 is the only bacterial monitoring station along the Aransas River Tidal. ") ") 1 1 1 1 2 4 1 4 0 113 1 0 6 91 85 84 78 68 62 61 Cumulative Density Function (CDF) of Fecal Coliform Concentration (CFU/100mL) at Schemanode 68 0 10 20 30 40 50 60 70 80 90 100 0 200 400 600 800 1000 1200 1400 cfu/100mL Percen t i le Monitoring data Model Model Median = 76.6 cfu/100mL 90th-percentile = 544 cfu/100mL Contact recreation standard for median = 200 cfu/100mL Single sample should not exceed 400 cfu/100mL) Number of Simulations = 1000 3/3/2006 1 16 34 PM Monitoring data Median = 72 cfu/100mL 90th-percentile = 553.6 cfu/100mL Figure 7.7 Modeled versus Measured Fecal Coliform Concentrations at Station 12952 247 The parameters of the SchemaNodes and SchemaLinks that could have been and/or were adjusted at Station 12948 are given in Tables 7.5 and 7.6, respectively. The locations of the SchemaNodes, SchemaLinks, Station 12948, and the results of the calibration are shown in Figure 7.8. Table 7.5 SchemaNode Adjusted Parameters for Calibration of Station 12948 SchemaNode (HydroID) Bacterial Loading Multiplication Factor 64 1.9 68 1.9 83 1.9 Table 7.6 SchemaLink Adjusted Parameters for Calibration of Station 12948 Residence Time (days) SchemaLink (HydroID) Initial (Schematic Processor Model) Final (Monte Carlo Simulation Model) 108 0.05 0.05 120 1.51 1.51 126 3.2 2.95* * Parameters that were adjusted for calibration 248 Cumulative Density Function (CDF) of Fecal Coliform Concentration (CFU/100mL) at Schemanode 75 0 10 20 30 40 50 60 70 80 90 100 0 200 400 600 800 1000 1200 1400 1600 1800 cfu/100mL Percen t i le Monitoring data Model Model Median = 94.3 cfu/100mL 90th-percentile = 727.2 cfu/100mL Contact recreation standard for median = 200 cfu/100mL Single sample should not exceed 400 cfu/100mL) Number of Simulations = 1000 3/3/2006 2 02 55 PM Monitoring data Median = 96 cfu/100mL 90th-percentile = 798.5 cfu/100mL ") 1 2 0 1 2 6 83 75 68 64 ") Figure 7.8 Modeled versus Measured Fecal Coliform Concentrations at Station 12948 249 7.3.2.3 Mission River Above Tidal Station 12944 is the only bacterial monitoring station along the Mission River Above Tidal. The parameters of the SchemaNodes and SchemaLinks that could have been and/or were adjusted at Station 12944 are given in Tables 7.7 and 7.8, respectively. The locations of the SchemaNodes, SchemaLinks, Station 12944, and the results of the calibration are shown in Figure 7.9. Table 7.7 SchemaNode Adjusted Parameters for Calibration of Station 12944 SchemaNode (HydroID) Bacterial Loading Multiplication Factor 71 1.6 73 1.6 77 1.6 80 1.6 90 1.6 99 1.6 Table 7.8 SchemaLink Adjusted Parameters for Calibration of Station 12944 Residence Time (days) SchemaLink (HydroID) Initial (Schematic Processor Model) Final (Monte Carlo Simulation Model) 116 1.42 1.36* 118 4.95 4.95 119 0.05 0.05 123 3 3 128 2.29 2.29 132 1 1 * Parameters that were adjusted for calibration 250 Cumulative Density Function (CDF) of Fecal Coliform Concentration (CFU/100mL) at Schemanode 74 0 10 20 30 40 50 60 70 80 90 100 0 200 400 600 800 1000 1200 1400 cfu/100mL Percen til e Monitoring data Model Model Median = 100.5 cfu/100mL 90th-percentile = 640.8 cfu/100mL Contact recreation standard for median = 200 cfu/100mL Single sample should not exceed 400 cfu/100mL) Number of Simulations = 1000 3/3/2006 2 31 51 PM Monitoring data Median = 116 cfu/100mL 90th-percentile = 518.8 cfu/100mL ") ") 1 1 8 1 2 8 1 2 3 1 1 6 99 90 80 77 74 73 Figure 7.9 Modeled versus Measured Fecal Coliform Concentrations at Station 12944 251 7.3.2.4 Mission River Tidal Station 12943 is the only bacterial monitoring station along the Mission River Tidal. The parameters of the SchemaNodes and SchemaLinks that could have been and/or were adjusted at Station 12943 are given in Tables 7.9 and 7.10, respectively. The locations of the SchemaNodes, SchemaLinks, Station 12943, and the results of the calibration are shown in Figure 7.10. Table 7.9 SchemaNode Adjusted Parameters for Calibration of Station 12943 SchemaNode (HydroID) Bacterial Loading Multiplication Factor 65 1.4 74 1.4 97 1.4 100 1.4 Table 7.10 SchemaLink Adjusted Parameters for Calibration of Station 12943 Residence Time (days) SchemaLink (HydroID) Initial (Schematic Processor Model) Final (Monte Carlo Simulation Model) 109 0.22 0.22 115 1.01 1.1* 130 0.19 0.19 133 2 2 * Parameters that were adjusted for calibration 252 ") ") 11 5 1 3 3 1 0 9 1 3 0 97 74 70 65 100 Cumulative Density Function (CDF) of Fecal Coliform Concentration (CFU/100mL) at Schemanode 70 0 10 20 30 40 50 60 70 80 90 100 0 50 100 150 200 250 300 350 400 450 cfu/100mL Percen til e Monitoring data Model Model Median = 49.4 cfu/100mL 90th-percentile = 209.5 cfu/100mL Contact recreation standard for median = 200 cfu/100mL Single sample should not exceed 400 cfu/100mL) Number of Simulations = 1000 3/3/2006 2 51 13 PM Monitoring data Median = 47 cfu/100mL 90th-percentile = 208.5 cfu/100mL Figure 7.10 Modeled versus Measured Fecal Coliform Concentrations at Station 12943 253 7.3.2.5 Copano Bay The four Copano Bay segments (Segments 1, 2, 3, and 4) were calibrated. Watershed JunctionID 45405 drains into Segment 1. Bacterial monitoring stations 13405, 14782, 14784, and 14790 measure fecal coliform concentrations in Copano Bay Segment 1. The parameters of the SchemaNodes and SchemaLinks that could have been and/or were adjusted at Segment 1 are given in Tables 7.11 and 7.12, respectively. The locations of the SchemaNodes, SchemaLinks, the bacterial monitoring stations, and the results of the calibration are shown in Figure 7.11. Table 7.11 SchemaNode Adjusted Parameters for Calibration of Segment 1 SchemaNode (HydroID) Bacterial Loading Multiplication Factor 88 2 93 2 155 1.6 Table 7.12 SchemaLink Adjusted Parameters for Calibration of Segment 1 Residence Time (days) SchemaLink (HydroID) Initial (Schematic Processor Model) Final (Monte Carlo Simulation Model) 135 0.487 0.6* 159 - - * Parameters that were adjusted for calibration 254 ") " ) ") ") ") ") ") ") 1 5 9 1 3 5 93 88 155 Cumulative Density Function (CDF) of Fecal Coliform Concentration (CFU/100mL) at Schemanode 155 0 10 20 30 40 50 60 70 80 90 100 102030405060708090 cfu/100mL Percen til e Monitoring data Model Model Median = 2.3 cfu/100mL 90th-percentile = 27.9 cfu/100mL Oyster water use standard for median = 14 cfu/100mL standard for 90th-percentile = 43 cfu/100mL) Number of Simulations = 1000 3/3/2006 3 33 33 PM Monitoring data Median = 2 cfu/100mL 90th-percentile = 14.6 cfu/100mL Figure 7.11 Modeled versus Measured Fecal Coliform Concentrations at Segment 1 255 Aransas River and Chilipitin Creek drain into Segment 2. Bacterial monitoring stations 12945, 14783, 14787, and 14788 measure fecal coliform concentrations in Copano Bay Segment 2. The parameters of the SchemaNodes and SchemaLinks that could have been and/or were adjusted at Segment 2 are given in Tables 7.13 and 7.14, respectively. The locations of the SchemaNodes, SchemaLinks, the bacteria monitoring stations, and the results of the calibration are shown in Figure 7.12. Table 7.13 SchemaNode Adjusted Parameters for Calibration of Segment 2 SchemaNode (HydroID) Bacterial Loading Multiplication Factor 63 1.1 67 3.5 69 1.1 75 1.1 86 1.1 87 3.5 101 1.1 102 1.1 103 1.1 104 3.5 154 3.5 256 Table 7.14 SchemaLink Adjusted Parameters for Calibration of Segment 2 Residence Time (days) SchemaLink (HydroID) Initial (Schematic Processor Model) Final (Monte Carlo Simulation Model) 107 0.686 0.686 112 0.21 0.21 114 0.01 0.01 121 0.01 0.01 122 1.5 1.5 127 4 4 134 2.31 2.6* 139 1 1 142 2.3 2.3 143 1.5 1.5 158 - - * Parameters that were adjusted for calibration 257 ") " ) ") ") ") ") ") ") 122 107 1 3 4 142 127 1 1 2 1 4 3 1 5 8 1 2 1 1 3 9 87 86 75 69 67 63 104 103 102 154 Cumulative Density Function (CDF) of Fecal Coliform Concentration (CFU/100mL) at Schemanode 154 0 10 20 30 40 50 60 70 80 90 100 0 50 100 150 200 250 300 cfu/100mL Per cent i l e Monitoring data Model Model Median = 1.7 cfu/100mL 90th-percentile = 55 cfu/100mL Oyster water use standard for median = 14 cfu/100mL standard for 90th-percentile = 43 cfu/100mL) Number of Simulations = 1000 3/3/2006 4 01 04 PM Monitoring data Median = 2 cfu/100mL 90th-percentile = 79 cfu/100mL Figure 7.12 Modeled versus Measured Fecal Coliform Concentrations at Segment 2 258 Mission River drains into Segment 3. Bacterial monitoring station 14797 measures fecal coliform concentrations in Copano Bay Segment 3. The parameters of the SchemaNodes and SchemaLinks that could have been and/or were adjusted at Segment 3 are given in Tables 7.15 and 7.16, respectively. The locations of the SchemaNodes, SchemaLinks, the bacterial monitoring stations, and the results of the calibration are shown in Figure 7.13. Table 7.15 SchemaNode Adjusted Parameters for Calibration of Segment 3 SchemaNode (HydroID) Bacterial Loading Multiplication Factor 66 2.5 70 1.4 82 2.5 153 1.4 Table 7.16 SchemaLink Adjusted Parameters for Calibration of Segment 3 Residence Time (days) SchemaLink (HydroID) Initial (Schematic Processor Model) Final (Monte Carlo Simulation Model) 110 0.29 0.29 129 2.18 1.7* 157 - - * Parameters that were adjusted for calibration 259 ") ") 1 2 9 1 5 7 1 1 0 82 70 66 153 Cumulative Density Function (CDF) of Fecal Coliform Concentration (CFU/100mL) at Schemanode 153 0 10 20 30 40 50 60 70 80 90 100 0 50 100 150 200 cfu/100mL Pe r c e n t i l e Monitoring data Model Model Median = 2 cfu/100mL 90th-percentile = 50.3 cfu/100mL Oyster water use standard for median = 14 cfu/100mL standard for 90th-percentile = 43 cfu/100mL) Number of Simulations = 1000 3/3/2006 4 14 44 PM Monitoring data Median = 2 cfu/100mL 90th-percentile = 49 cfu/100mL Figure 7.13 Modeled versus Measured Fecal Coliform Concentrations at Segment 3 260 Copano Creek drains into Segment 4. Bacterial monitoring stations 13404, 14779, 14780, 14785, 14792, and 14793 measure fecal coliform concentrations in Copano Bay Segment 4. The parameters of the SchemaNodes and SchemaLinks that were adjusted at Segment 4 are given in Tables 7.17 and 7.18, respectively. The locations of the SchemaNodes, SchemaLinks, the bacterial monitoring stations, and the results of the calibration are shown in Figure 7.14. Table 7.17 SchemaNode Adjusted Parameters for Calibration of Segment 4 SchemaNode (HydroID) Bacterial Loading Multiplication Factor 72 1.2 81 1.2 89 2 94 1.2 95 1.2 96 2 98 1.2 105 1.2 156 1.2 Table 7.18 SchemaLink Adjusted Parameters for Calibration of Segment 4 Residence Time (days) SchemaLink (HydroID) Initial (Schematic Processor Model) Final (Monte Carlo Simulation Model) 117 0.05 0.05 131 3 3 136 1.5 1.5 137 1.15 0.9* 138 5 5 144 1.5 1.5 160 - - 161 - * Parameters that were adjusted for calibration 261 ") ") ") ") ") ") ") ") ") ") 1 3 1 1 3 8 1 3 7 1 6 1 1 6 0 1 3 6 1 4 4 98 96 95 94 89 81 72 156 105 Cumulative Density Function (CDF) of Fecal Coliform Concentration (CFU/100mL) at Schemanode 156 0 10 20 30 40 50 60 70 80 90 100 0 5 10 15 20 25 30 cfu/100mL Perc en ti le Monitoring data Model Model Median = 1.3 cfu/100mL 90th-percentile = 11.9 cfu/100mL Oyster water use standard for median = 14 cfu/100mL standard for 90th-percentile = 43 cfu/100mL) Number of Simulations = 1000 3/3/2006 4 32 34 PM Monitoring data Median = 2 cfu/100mL 90th-percentile = 13 cfu/100mL Figure 7.14 Modeled versus Measured Fecal Coliform Concentrations at Segment 4 262 7.3.3 Calculation of Load Allocations Once the model was calibrated (see Section 7.3.2), the parameters of the model were not re-adjusted because the calibrated model well-represented the measured fecal coliform concentrations at each of the bacterial monitoring stations. Fecal coliform water quality standards are currently being exceeded in Copano Bay segments 2 and 3. Because this is a fecal coliform model, only the fecal coliform water quality standards were considered in the load reduction determinations. However, three indicator bacteria are used in the Copano Bay watershed (enterococci for Aransas and Mission River Tidals, E. coli for Aransas and Mission River Above Tidals, and fecal coliform for Copano Bay), so we recommend that a model be created for each indicator bacteria to determine the appropriate load reductions in the water segments using those indicators. To ensure compliance with the fecal coliform water quality standards, the model was used to investigate fecal coliform concentrations at the upstream and downstream portions of the Above Tidals and Tidals, the locations of the bacterial monitoring stations, and the Copano Bay water segments. Like the calibration of the model, load reduction determinations started at the upstream locations and proceeded toward the downstream locations because upstream load reductions affect what downstream reductions are necessary. If load reduction was necessary for a segment, the loadings at controlled point sources (e.g., WWTPs) were reduced first. If that reduction was not sufficient to meet the water quality standards, then loadings from non-point sources (e.g., livestock) were reduced until the water quality standards were met. Two scenarios of load reductions were found for each water segment. Load Reduction Scenario #1 is the load reduction necessary to meet fecal coliform water 263 quality standards for all water segments (Aransas and Mission River Tidals, Aransas and Mission River Above Tidals, and Copano Bay) at each location in the model that was analyzed. However, each portion of the model that was analyzed that did not meet fecal coliform water quality standards was not always verified by existing monitoring data; thus, the results from Load Reduction Scenario #1 are inconclusive due to lack of monitoring data and are presented in Appendix 7.2. Load Reduction Scenario #2 is the load reductions necessary to meet fecal coliform water quality standards for all water segments at each monitoring station location. Only the fecal coliform bacterial load reductions for Copano Bay are presented in this chapter because fecal coliform is the primary bacterial indicator for Copano Bay. 7.3.3.1 Copano Bay Copano Bay must meet oyster harvesting use standards for fecal coliform. The median of the samples (within a two-year period) must be less than 14 CFU/100mL, and the 90 th -percentile of the samples must be less than 43 CFU/100mL (i.e., 10% of the samples are allowed to exceed 43 CFU/100mL.) Four Copano Bay segments (Segments 1, 2, 3, and 4) were analyzed (shown in Figure 7.15). Considering all the fecal coliform monitoring data from 1999-2005, Segments 1 and 4 are currently meeting water quality standards. Segments 2 and 3 (Aransas and Mission River outlets, respectively) are currently exceeding water quality standards. Thus, only load reductions for Segments 2 and 3 must be determined. 264 Without any load reductions in the upstream watersheds of Copano Bay Segment 2, two runs of 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted. For the two separate 1000 simulation runs, the median and 90 th -percentile CFU/100mL are shown in Table 7.19. Copano Bay Segment 2 exceeds the fecal coliform water quality standard 90 th - percentile > 43 CFU/100mL based on modeled results and monitoring data (shown in Table 7.19). WWTP and livestock bacterial loadings were reduced in an attempt to meet water quality standards. The bacterial loadings were reduced at all three upstream WWTPs, and livestock bacterial loadings were reduced at the adjacent upstream watersheds to Copano Bay; these watersheds are shown in Figure 7.16. The number of runs of simulations and the modeled results at SchemaNode 154 with various load reductions of WWTP and livestock bacterial loadings are also shown in Table 7.19. Bacterial Monitoring Stations ! ! ") ") ") ") ") ") ") ") ") ") ") ") ") ")") ") 3 4 2 1 Figure 7.15 Copano Bay Segments 1, 2, 3, 4 Legend ") 265 Table 7.19 Modeled Results at SchemaNode 154 with Various WWTP/Livestock Load Reductions Run # Load Reduction (%) Bacteria Source Median (CFU/100mL) 90 th -Percentile > 43 CFU/100mL 1 0 N/A 1.8 54.89 2 0 N/A 1.69 46.29 1 50 WWTP 1.6 46.05 2 0 Livestock 1.48 46.58 1 1.45 41.62 2 50 WWTP 1.24 30.81 3 1.38 25.27 4 10 Livestock 1.38 37.31 1 50 WWTP 1.53 38.79 2 1.30 33.84 3 15 Livestock 1.40 32.88 Simulations 100 1000 Reducing the WWTP bacterial loadings 18 by 50% and livestock bacterial loadings by 15% in the adjacent upstream watersheds allows the 90 th -percentile to be approximately 35 CFU/100mL, which is less than the 43 CFU/100mL standard and results in a median less than 14 CFU/100mL (shown in Table 7.19). The reductions necessary to meet fecal coliform oyster water use standards at Copano Bay Segment 2 based on modeled results are shown in Figure 7.16. Reduction of livestock bacterial loadings would require implementations of agricultural BMPs, and reduction of WWTP bacterial loadings would require proper disinfection before discharging into surface waters. The existing monitoring data from 1999-2005 and the probability distribution when the load reductions are implemented are shown in Figure 7.17. Both criteria are met when the reductions are implemented. 18 The WWTP load reductions were based on the overestimated bacterial loadings from WWTPs (explained in Section 5.4.2). 266 Legend Livestock WWTP 50% 50% 50% 15% 15% Figure 7.16 Load Reductions for SchemaNode 154: Copano Bay 267 Without any load reductions in the upstream watersheds of Copano Bay Segment 3, two runs of 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted. For the two separate 1000 simulation runs, the median and 90 th -percentile CFU/100mL are shown in Table 7.20. Table 7.20 shows that Copano Bay Segment 3 exceeds the fecal coliform water quality standard 90 th -percentile > 43 CFU/100mL based on modeled results and monitoring data when no load reductions are applied. Livestock bacterial loadings were reduced (in the upstream watershed adjacent to the Bay, shown in Figure 7.18) in an attempt to meet water quality standards; septic systems, WWTPs, and avian loadings do not discharge directly to Copano Bay, and non- point bacterial loadings are significantly less than livestock loadings. Copano Bay (Aransas River Outlet) 0.01 0.1 1 10 100 1000 10000 100000 0 10203040506070809010 Probability of Exceedance, % Fecal Co l i f orm ( cf u /100 m L ) Existing Loads Reduced 14 cfu/100mL - Median FC standard criteria #1 43 cfu/100 mL - Standard criteria #2 90% of Observed Data 10 % of Observed Data (allowed exceedance) Region not in compliance Figure 7.17 Existing versus Reduced Loads in Copano Bay Segment 2 268 The number of runs of simulations and the modeled results at SchemaNode 153 with various load reductions of livestock bacterial loadings are also shown in Table 7.20. Table 7.20 Modeled Results at SchemaNode 153 with Various Livestock Load Reductions Run # Load Reduction (%) Bacteria Source Median (CFU/100mL) 90 th -Percentile < 43 CFU/100mL 1 0 N/A 1.9 47.2 2 0 N/A 2.2 53.4 1 10 Livestock 1.59 42.6 1 1.54 33.0 2 2.02 46.1 3 15 Livestock 1.61 42.2 1 1.49 33.7 2 20 Livestock 1.74 28.0 Simulations 100 1000 Reducing the livestock bacterial loadings by 20% in the adjacent upstream watersheds allows the 90 th -percentile to be approximately 30.85 CFU/100mL, which is less than the 43 CFU/100mL standard and results in a median less than 14 CFU/100mL (shown in Table 7.20). The load reductions necessary to meet fecal coliform oyster water use standards at Copano Bay Segment 3 based on modeled results are shown in Figure 7.18. Reduction of livestock bacterial loadings would require implementations of agricultural BMPs. The existing monitoring data from 1999-2005 and the probability distribution when the load reductions are implemented for the Mission River outlet into Copano Bay are shown in Figure 7.19. 269 Figure 7.18 Load Reductions for SchemaNode 153: Copano Bay Legend 20% Livestock WWTP 270 7.4 RESULTS The current loadings and load allocations from each source to each of the water segments (i.e., Aransas River Above Tidal, Aransas River Tidal, Mission River Above Tidal, Mission River Tidal, and Copano Bay) for the Monte Carlo Simulation Model are given in Chapter 8 of this report. Considering only fecal coliform water quality standards, the load reductions required to satisfy the standards for all portions of the model where bacterial monitoring stations indicate exceedances are shown in Figure 7.20. This load reduction scenario is referred to as Load Reduction Scenario #2. Thus, based on fecal coliform monitoring data from 1999-2005, only Copano Bay Segments 2 and 3 exceed fecal coliform water quality standards. Copano Bay (Mission River Outlet) 0.01 0.1 1 10 100 1000 10000 0 10203040506070809010 Probability of Exceedance, % F e c a l C o lif o r m (c fu /1 0 0 m L ) Existing Loads Reduced 14 cfu/100mL - Median fecal coliform standard criteria #1 43 cfu/100 mL - standard criteria #2 90% of Observed 10 % of Observed Data (allowed exceedance) Region not in compliance Figure 7.19 Existing versus Reduced Loads in Copano Bay Segment 3 271 Results are described in more detail in Chapter 8 of this report, and Chapter 9 discusses recommendations on how to reduce these bacterial loadings. Legend WWTP Livestock 20% Figure 7.20 Load Reductions to Satisfy Fecal Coliform Standards for Monitored Conditions 50% 50% 50% 15% 15% 272 Chapter 8: Results 8.1 ESTIMATION OF LOADINGS The current loadings to the water segments in the Copano Bay watershed are presented in this section using the Schematic Processor Model and Monte Carlo Simulation Model. 8.1.1 Schematic Processor The current loadings to each of the water segments (Aransas River Above Tidal, Mission River Above Tidal, Aransas River Tidal, Mission River Tidal, and Copano Bay) are presented in this section. Other results from the Schematic Processor Model are given in Section 6.4. These loadings are based on the annual bacterial loadings that were calculated in Chapter 5 and the simulation of bacterial transport (with the calibrated Schematic Processor Model) that is described in Chapter 6. Because bacterial loadings are only input into the model at locations of SchemaNodes, the upstream and downstream bacterial loadings of each of the water segments were found. These loadings are the “PassedVal” and “TotVal” in the attribute table of the SchemaNode feature class after the Schematic Processor (“Process Schematic” script) was run under calibrated conditions (described in Section 6.3.4). The bacterial loadings at the upstream and downstream nodes of the Aransas River Above Tidal (segment 2004) are shown in Table 8.1. The bacterial loadings from the major bacterial sources (identified in Section 6.4) are shown in Table 8.2. 273 Table 8.1 Schematic Processor Bacterial Loadings to Aransas River Above Tidal (Segment 2004) Location on Segment SchemaNode HydroID Load (CFU/year) Upstream 62 3.345E+13 Downstream 75 9.452E+13 Table 8.2 Schematic Processor Bacterial Loadings (from Major Sources) to Aransas River Above Tidal (Segment 2004) Bacteria Source Upstream, Node 62 (CFU/year) Downstream, Node 75 (CFU/year) Cattle 9.463E+11 5.961E+13 WWTP 3.220E+13 2.431E+13 OSSF 1.228E+11 4.611E+11 Birds 0.000E+00 0.000E+00 Non-point (Urban, Forest, etc.) 3.124E+10 2.180E+12 Total Load 3.345E+13 9.452E+13 The bacterial loadings increase from the upstream to the downstream portions of the Aransas River Above Tidal (shown in Tables 8.1 and 8.2), which can be explained by the large upstream watershed draining to the Aransas River Above Tidal. The major bacterial source at the upstream portion of the Aransas River Above Tidal is the WWTP (City of Beeville Moore Street WWTP) based on the results of the Schematic Processor Model. However, as explained in Section 5.4.2, the WWTP bacterial loading is largely overestimated for our research. The major bacterial source at the downstream of the Aransas River Above Tidal is cattle. The bacterial loadings at the upstream and downstream nodes of the Aransas River Tidal (segment 2003) are shown in Table 8.3. The bacterial loadings from the major bacterial sources (identified in Section 6.4) are shown in Table 8.4. 274 Table 8.3 Schematic Processor Bacterial Loadings to Aransas River Tidal (Segment 2003) Location on Segment SchemaNode HydroID Load (CFU/year) Upstream 75 9.452E+13 Downstream 67 8.693E+14 Table 8.4 Schematic Processor Bacterial Loadings (from Major Sources) to Aransas River Tidal (Segment 2003) Bacteria Source Upstream, Node 75 (CFU/year) Downstream, Node 67 (CFU/year) Cattle 5.961E+13 5.961E+14 WWTP 2.431E+13 3.912E+13 OSSF 4.611E+11 6.583E+13 Birds 0.000E+00 0.000E+00 Non-point (Urban, Forest, etc.) 2.180E+12 1.239E+14 Total Load 9.452E+13 8.693E+14 The bacterial loadings increase from the upstream to the downstream portions of the Aransas River Tidal (shown in Tables 8.3 and 8.4). The major bacterial source at the upstream portion of the Aransas River Tidal is cattle, which is about double the bacterial loadings from the upstream WWTPs (Water Reclamation Facility, City of Taft Baird WWTP, City of Sinton Main WWTP, City of Odem WWTP) based on the results of the Schematic Processor Model. The major bacterial source downstream of the Aransas River Tidal is cattle, followed by non-point bacterial sources (land uses of urban, forest, etc.), and then septic systems. The bacterial loadings at the upstream and downstream nodes of the Mission River Above Tidal (segment 2002) are shown in Table 8.5. The bacterial loadings from the major bacterial sources (identified in Section 6.4) are shown in Table 8.6. 275 Table 8.5 Schematic Processor Bacterial Loadings to Mission River Above Tidal (Segment 2002) Location on Segment SchemaNode HydroID Load (CFU/year) Upstream 73 5.105E+14 Downstream 65 1.262E+14 Table 8.6 Schematic Processor Bacterial Loadings (from Major Sources) to Mission River Above Tidal (Segment 2002) Bacteria Source Upstream, Node 73 (CFU/year) Downstream, Node 65 (CFU/year) Cattle 4.800E+14 1.074E+14 WWTP 2.431E+13 7.960E+11 OSSF 4.611E+11 1.815E+06 Birds 0.000E+00 0.000E+00 Non-point (Urban, Forest, etc.) 2.180E+12 1.511E+13 Total Load 5.105E+14 1.262E+14 The bacterial loading decreases from the upstream to the downstream portions of the Mission River Above Tidal (shown in Tables 8.5 and 8.6). This can be explained by the major upstream watersheds draining to the upstream portion of the Mission River Above Tidal and very small watersheds draining along the Mission River Above Tidal. The major bacterial source at the upstream portion of the Mission River Above Tidal is cattle based on the results of the Schematic Processor Model. The major bacterial source at the downstream of the Mission River Above Tidal is also cattle, followed by non-point bacterial sources (land uses of urban, forest, etc.). The bacterial loadings at the upstream and downstream nodes of the Mission River Tidal (segment 2001) are shown in Table 8.7. The bacterial loadings from the major bacterial sources (identified in Section 6.4) are shown in Table 8.8. 276 Table 8.7 Schematic Processor Bacterial Loadings to Mission River Tidal (Segment 2001) Location on Segment SchemaNode HydroID Load (CFU/year) Upstream 65 1.262E+14 Downstream 66 1.115E+15 Table 8.8 Schematic Processor Bacterial Loadings (from Major Sources) to Mission River Tidal (Segment 2001) Bacteria Source Upstream, Node 65 (CFU/year) Downstream, Node 66 (CFU/year) Cattle 1.074E+14 1.062E+15 WWTP 7.960E+11 5.726E+11 OSSF 1.815E+06 1.348E+05 Birds 0.000E+00 0.000E+00 Non-point (Urban, Forest, etc.) 1.511E+13 3.266E+13 Total Load 1.262E+14 1.115E+15 The bacterial loadings increase from the upstream to the downstream portions of the Mission River Tidal (shown in Tables 8.7 and 8.8.) The major bacterial source at the upstream portion of the Mission River Tidal is cattle based on the results of the Schematic Processor Model. The major bacterial source downstream of the Mission River Tidal is also cattle, followed by non-point bacterial sources (land uses of urban, forest, etc.). The bacterial loadings to each of the Copano Bay segments (Segments 1, 2, 3, and 4) and the total current annual bacterial loading to Copano Bay are shown in Table 8.9. The bacterial loadings were calculated by multiplying the “TotVal” by the “CumRunoff_m3_yr”, fields in the SchemaNode attribute table. Recall, that “TotVal” is the concentration of a SchemaNode SrcType 3 in CFU/m 3 , and “CumRunoff_m3_yr” is the cumulative runoff of all the upstream watersheds that drain to that particular 277 Schemanode SrcType 3. The bacterial loading was calculated for each of the four Copano Bay segments (of the four SchemaNodes SrcType 3). The bacterial loadings from the major bacterial sources for each of the Copano Bay segments (identified in Section 6.4) are shown in Table 8.10. Table 8.9 Schematic Processor Bacterial Loadings to Copano Bay (Segment 2472) Copano Bay Segment SchemaNode HydroID Load (CFU/year) Watershed 45405 (Segment 1) 155 5.062E+10 Aransas River Outlet (Segment 2) 154 5.036E+12 Mission River Outlet (Segment 3) 153 5.503E+12 Copano Creek Outlet (Segment 4) 156 1.457E+12 Total 1.205E+13 Table 8.10 Schematic Processor Bacterial Loadings from Major Sources to Copano Bay (Segment 2472) in CFU/year Bacteria Source Segment 1 (Node 155) Segment 2 (Node 154) Segment 3 (Node 153) Segment 4 (Node 156) Total Cattle 2.491E+11 3.453E+12 5.243E+12 1.202E+12 1.015E+13 WWTP 0.000E+00 2.266E+11 2.826E+09 0.000E+00 2.294E+11 OSSF 1.716E+11 3.813E+11 6.650E+02 1.718E+11 7.247E+11 Birds 1.169E+08 1.283E+07 7.287E+06 9.148E+07 2.285E+08 Non-point (Urban, Forest, etc.) 4.407E+10 7.176E+11 1.611E+11 4.637E+10 9.692E+11 Total Load 5.062E+10 5.036E+12 5.503E+12 1.457E+12 1.205E+13 Bacterial loading is greatest at the Mission River outlet (Copano Bay Segment 3) and the Aransas River outlet (Copano Bay Segment 2), which is shown in Table 8.9. Cattle are the major bacterial source based on Schematic Processor Model results (Table 8.10) for all Copano Bay segments. 278 8.1.2 Monte Carlo Simulations The current loadings to each of the water segments (Aransas River Above Tidal, Mission River Above Tidal, Aransas River Tidal, Mission River Tidal, and Copano Bay) are presented in this section using the Monte Carlo Simulation Model. Other results from the Monte Carlo Simulation Model are given in Section 7.4. The difficulty with calculating the current loadings to each water segment using the Monte Carlo Simulation Model is that 1000 simulations (user-defined) are implemented per run, so this means that each SchemaNode in the Schematic Network has 1000 bacterial concentrations/loadings associated with it (i.e., a probability distribution of concentrations/loadings). On the other hand, the Schematic Processor Model implements one simulation (that represents average annual conditions), so only one bacterial loading/concentration is associated with each SchemaNode in the Schematic Network at calibrated conditions. Thus, the median of the 1000 simulations for each SchemaNode is used to represent the ‘current loading’ to each SchemaNode (and each water segment). These values should be similar to the current loadings calculated by the Schematic Processor Model in Section 8.1.1 since the bacterial loading distributions (at input locations) were based on the assumption that the median equals the annual average bacterial loadings that were calculated in Chapter 5. However, separate runs were not implemented for each bacterial source, so only the total current loadings to each water segment were determined using the Monte Carlo Simulation Model. The Monte Carlo Simulation Model outputs fecal coliform concentrations in CFU/100mL, so the median concentration was multiplied by the cumulative flow (of the upstream watersheds that were calculated in Section 5.1.2.6) and multiplied by 10,000 to convert from CFU/100mL to CFU/m 3 to find the bacterial loading in CFU/year. 279 The bacterial loadings at the upstream and downstream nodes of the Aransas River Above Tidal (segment 2004) are shown in Table 8.11. Table 8.11 Monte Carlo Simulation Model Loadings to Aransas River Above Tidal (Segment 2004) Location on Segment SchemaNode HydroID Load (CFU/year) Upstream 62 3.712E+13 Downstream 75 9.680E+13 The bacterial loading increases from the upstream to the downstream portions of the Aransas River Above Tidal (shown in Table 8.11). The bacterial loadings at the upstream and downstream nodes of the Aransas River Tidal (segment 2003) are shown in Table 8.12. Table 8.12 Monte Carlo Simulation Model Loadings to Aransas River Tidal (Segment 2003) Location on Segment SchemaNode HydroID Load (CFU/year) Upstream 75 9.680E+13 Downstream 67 8.418E+14 The bacterial loadings increase from the upstream to the downstream portions of the Aransas River Tidal (shown in Table 8.12). The bacterial loadings at the upstream and downstream nodes of the Mission River Above Tidal (segment 2002) are shown in Table 8.13. 280 Table 8.13 Monte Carlo Simulation Model Loadings to Mission River Above Tidal (Segment 2002) Location on Segment SchemaNode HydroID Load (CFU/year) Upstream 73 3.645E+14 Downstream 65 1.397E+14 The bacterial loadings decrease from the upstream to the downstream portions of the Mission River Above Tidal (shown in Table 8.13). This can be explained by the fact that major upstream watersheds drain to the upstream portion of the Mission River Above Tidal, and very small watersheds drain along the Mission River Above Tidal. The bacterial loadings at the upstream and downstream nodes of the Mission River Tidal (segment 2001) are shown in Table 8.14. Table 8.14 Monte Carlo Simulation Model Loadings to Mission River Tidal (Segment 2001) Location on Segment SchemaNode HydroID Load (CFU/year) Upstream 65 1.397E+14 Downstream 66 1.123E+15 The bacterial loadings increase from the upstream to the downstream portions of the Mission River Tidal (shown in Table 8.14). The bacterial loadings to each of the Copano Bay segments (Segments 1, 2, 3, and 4) and the total current annual bacterial loading to Copano Bay are shown in Table 8.15. The bacterial loading was calculated for each of the four Copano Bay segments (i.e., the four SchemaNodes SrcType 3). 281 Table 8.15 Monte Carlo Simulation Model Loadings to Copano Bay (Segment 2472) Copano Bay Segment SchemaNode HydroID Load (CFU/year) Watershed 45405 (Segment 1) 155 6.504E+10 Aransas River Outlet (Segment 2) 154 4.868E+12 Mission River Outlet (Segment 3) 153 4.995E+12 Copano Creek Outlet (Segment 4) 156 8.354E+11 Total 1.076E+13 The bacterial loading is greatest at the Mission River outlet (Copano Bay Segment 3) and the Aransas River outlet (Copano Bay Segment 2), which is shown in Table 8.15. 282 8.2 ESTIMATION OF LOAD ALLOCATION The percent load reductions necessary to meet fecal coliform water quality standards were determined using the Monte Carlo Simulation Model, and the Schematic Processor Model was used to quantify the load reductions required (as well as the allowable load) for each water segment in the Copano Bay watershed. 8.2.1 Monte Carlo Simulations The percent load reductions from the point and non-point source loadings (and the locations in the watershed) were determined in Chapter 7, and the final results are given in Section 7.4. These percent load reductions, which have a margin of safety incorporated, ensure that the water segments would be in compliance with fecal coliform bacterial water quality standards. There were two scenarios of load reductions presented (in Section 7.4 and Appendix 7.2). Load Reduction Scenario #1 looked at the upstream and downstream SchemaNodes of all river segments (Aransas and Mission River Above Tidals, and Aransas and Mission River Tidals), the SchemaNodes of all the bacterial monitoring stations, as well as the SchemaNodes of the four Copano Bay segments (Segments 1, 2, 3, and 4). Load reductions for this scenario were calculated to ensure that contact recreation and oyster water use fecal coliform standards were being met at all analyzed locations (Appendix 7.2). Load Reduction Scenario #2 looked at the SchemaNodes of where bacterial monitoring stations are located in the Monte Carlo Simulation Model to ensure that contact recreation and oyster water use fecal coliform standards are being met; thus, percent reductions were determined for locations where monitoring data 283 indicated problems with complying with fecal coliform water quality standards (Section 7.4). These percent load reductions (determined using the Monte Carlo Simulation Model in Chapter 7) were used with the Schematic Processor Model (the current loadings using the calibrated conditions at the average annual conditions) to determine the load reduction requirements and the allowable load (in CFU/year) for each of the water segments. 8.2.2 Applied to Schematic Processor Model The percent load reductions from the point and non-point sources of SchemaNodes in the Copano Bay watershed were applied to the SchemaNode where the bacterial loading from the source was applied to the model. However, the load reduced at the source (where the bacterial loading is applied to the model) is not necessarily the load that would need to be reduced at the downstream water segment of interest (e.g., Aransas River Above Tidal). The bacterial loadings from sources may be further upstream from the water segment of interest. Thus, these bacterial loadings were reduced at the source and then decayed by the corresponding residence times (determined in Section 6.3.3.2) of the SchemaLinks that the bacteria travel down (either watershed travel and/or river travel) between the source and the water segment. Thus, the corresponding load reduction at the water segments was found for each of the sources in the watershed. Load Reduction Scenario #1 contains the load reductions necessary to satisfy (including a margin of safety) fecal coliform water quality standards for all portions of the model that were analyzed (upstream/downstream SchemaNodes of water segments and bacterial monitoring stations). The results for Load Reduction Scenario #1 are presented in Appendix 8.1. Load Reduction Scenario #2 are the load reductions 284 necessary to satisfy (including a margin of safety) fecal coliform water quality standards for the portions of the model that correspond to bacterial monitoring stations, where problems are proven to exist by bacterial monitoring data. The results for Load Reduction Scenario #2 are presented in the following sections. 8.2.2.1 Aransas River Above Tidal The load reduction necessary for the Aransas River Above Tidal was determined for both load reduction scenarios. Load Reduction Scenario #1 is presented in Appendix 8.1, and Load Reduction Scenario #2 is presented below. No load reductions were necessary to meet water quality standards at the bacterial monitoring stations (Load Reduction Scenario #2). Thus, the current loadings and allowable loadings (given in Table 8.1) for the upstream and downstream portions of the Aransas River Above Tidal are the allowable loads to the segment. 8.2.2.2 Aransas River Tidal The load reduction necessary for the Aransas River Tidal was determined for both load reduction scenarios. Load Reduction Scenario #1 is presented in Appendix 8.1, and Load Reduction Scenario #2 is presented below. The recommended load reductions to comply with water quality standards at the bacterial monitoring station were from upstream WWTPs and livestock bacterial sources. The percent of reductions from the corresponding sources (determined in Section 7.3.3) and the SchemaNodes and SchemaLinks of interest are shown in Figure 8.1. 285 The load reductions and allowable loads to the upstream portion of the Aransas River Tidal are the same as the load reductions and allowable loads to the downstream portion of the Aransas River Above Tidal because it is the same SchemaNode (HydroID 75) in the Schematic Network. Since there are no load reductions in the downstream portion of the Aransas River Above Tidal, there are no load reductions in the upstream portion of the Aransas River Tidal. Thus, the allowable load equals the current load of the upstream portion of the Aransas River Tidal, which is given in Table 8.3. Table 8.16 shows the load reductions at the bacterial sources for the downstream portion of the Aransas River Tidal (SchemaNode 67). ") ") ") ") ") ") ") ") ") # # # # # 122 1 3 4 1 4 3 92 87 67 104 ") ") ") ") ") ")" ) ") ") ") ") ") ") ") ") ") ")") ") ") ") ") # # # # # # # # # Legend # Outfalls_WWTPs ") BacteriaMonitoringStations_TCEQ Figure 8.1 Load Reduction Scenario #2: Aransas River Tidal 50% 50% 15% 15% 75 67 286 Table 8.16 Load Reduction Scenario #2 at Downstream Node of Aransas River Tidal Schema- Node Source Current Loading (CFU/yr) Load Reduction at Source (CFU/yr) Total Residence Time to Segment (days) Equivalent Load at Segment (CFU/yr) 67 WWTP 1.48E+04 7.40E+03 0.00 7.40E+03 69 WWTP 3.37E+11 1.69E+11 0.01 1.65E+11 87 Livestock 9.81E+15 1.47E+15 1.44 8.28E+13 92 WWTP 6.98E+14 3.49E+14 1.50 1.74E+13 104 Livestock 5.61E+14 8.42E+13 1.51 4.11E+12 Total Load Reduction (CFU/year) 1.05E+14 Current Loading (CFU/year), Table 8.3 8.69E+14 Allowable Load (CFU/year) 7.65E+14 8.2.2.3 Mission River Above Tidal The load reduction necessary for the Mission River Above Tidal was determined for both load reduction scenarios. Load Reduction Scenario #1 is presented in Appendix 8.1, and Load Reduction Scenario #2 is presented below. No load reductions were needed to meet water quality standards at the bacterial monitoring station. Thus, the current loads of the upstream and downstream portions of the Mission River Above Tidal are the allowable loads to the segment; see Table 8.5 for current upstream/downstream bacterial loadings. 8.2.2.4 Mission River Tidal The load reduction necessary for the Mission River Tidal was determined for both load reduction scenarios. Load Reduction Scenario #1 is presented in Appendix 8.1, and Load Reduction Scenario #2 is presented below. The recommended load reductions to comply with water quality standards at the 287 bacterial monitoring station were from upstream livestock bacterial sources. The percent of reductions (determined in Section 7.3.3), the corresponding sources, and the SchemaNodes and SchemaLinks of interest are shown in Figure 8.2. The load reductions and allowable loads to the upstream portion of the Mission River Tidal are the same as the load reductions and allowable loads to the downstream portion of the Mission River Above Tidal because it is the same SchemaNode (HydroID 65) in the Schematic Network. Since there are no load reductions in the downstream portion of the Mission River Above Tidal, there are no load reductions in the upstream portion of the Mission River Tidal. Thus, the allowable load equals the current load of the upstream portion of the Mission River Tidal, which is given in Table 8.7. The load reductions at the bacterial sources for the downstream portion of the Mission River Tidal (SchemaNode 66) are shown in Table 8.17. Figure 8.2 Load Reduction Scenario #2: Mission River Tidal ") ") ") ") ") ") ") # # # 1 2 911 5 1 1 0 82 66 65 20% ! ! ! ! ! ! ") ") ") ") ") ")" ) ") ") ") ") ") ") ") ") ") ")") ") ") ") ") # # # # # # # # # Legend # Outfalls_WWTPs ") BacteriaMonitoringStations_TCEQ 288 Table 8.17 Load Reduction Scenario #2 at Downstream Node of Mission River Tidal Schema- Node Source Current Loading (CFU/yr) Load Reduction at Source (CFU/yr) Total Residence Time to Segment (days) Equivalent Load at Segment (CFU/yr) 82 Livestock 4.34E+16 8.68E+15 1.86 2.11E+14 Total Load Reduction (CFU/year) 2.11E+14 Current Loading (CFU/year), Table 8.7 1.16E+15 Allowable Load (CFU/year) 9.04E+14 8.2.2.5 Copano Bay The load reduction necessary for Copano Bay was determined for both load reduction scenarios. Load Reduction Scenario #1 is presented in Appendix 8.1, and Load Reduction Scenario #2 is presented below. No load reductions were necessary for Copano Bay Segments 1 and 4 in either scenario; however, load reductions were necessary for the Aransas River outlet (Copano Bay Segment 2) and the Mission River outlet (Copano Bay Segment 3.) The load reductions and allowable loads (CFU/year) for each Copano Bay Segment were determined first, and then the total load reductions and allowable loads for Copano Bay were determined. The recommended load reductions to comply with water quality standards at the bacterial monitoring stations were from upstream WWTPs/livestock bacterial sources (shown in Figure 7.20). The load reductions that are accounted for at the Aransas River outlet (Copano Bay Segment 2) are shown in Figure 8.1. The Aransas River Tidal drains directly into Copano Bay Segment 2, and the only additional loadings to this portion of the Bay are avian. Since the avian loading cannot be reduced, the total load reduction applied at 289 Segment 2 is the same load reduction that was found for the downstream portion of the Aransas River Tidal, which is given in Table 8.16. The load reduction needed at the Copano Bay Aransas River outlet (Copano Bay Segment 2) is shown in Table 8.18. Note that these load reductions are the reductions necessary to meet fecal coliform water quality standards in Copano Bay. Table 8.18 Load Reduction Scenario #2 at Copano Bay Aransas River Outlet, Segment 2 SchemaNode Source Equivalent Load at Tidal (CFU/yr) Concentration in Bay (CFU/m 3 ) Load in Bay (CFU/yr) 67 WWTP 7.40E+03 1.70E-07 4.29E+01 69 WWTP 1.65E+11 3.80E+00 9.57E+08 87 Livestock 8.28E+13 1.91E+03 4.80E+11 92 WWTP 1.74E+13 4.00E+02 1.00E+11 104 Livestock 4.11E+12 9.45E+01 2.38E+10 Cumulative Runoff. Q (m 3 /yr), Section 6.3.3.4 2.52E+08 Volume of Copano Bay Segment, V (m 3 ), Section 6.3.3.3 5.92E+07 Decay Coefficient of Segment, k (years -1 ), Section 6.3.3.1 7.30E+02 Total Load Reduction (CFU/year) 6.05E+11 Current Loading (CFU/year), Table 8.9 5.04E+12 Allowable Load (CFU/year) 4.43E+12 The load reductions that are accounted for at the Mission River outlet (Copano Bay Segment 3) are shown in Figure 8.2. The Mission River Tidal drains directly into Copano Bay Segment 3, and the only additional loadings to this portion of the Bay are avian. Since the avian loading is not reduced, the total load reduction applied at Segment 3 is the same load reduction that was found for the downstream portion of the Mission River Tidal, which is given in Table 8.17. The load reduction needed at the Copano Bay Mission River outlet (Copano Bay Segment 3) is shown in Table 8.19. Note that these load reductions are the reductions necessary to meet fecal coliform water quality standards in Copano Bay. 290 Table 8.19 Load Reduction Scenario #2 at Copano Bay Mission River Outlet, Segment 3 SchemaNode Source Equilavent Load at Tidal (CFU/yr) Concentration in Bay (CFU/m 3 ) Load in Bay (CFU/yr) 82 Livestock 2.11E+14 3.78E+03 1.04E+12 Cumulative Runoff. Q (m 3 /yr), Section 6.3.3.4 2.75E+08 Volume of Copano Bay Segment, V (m 3 ), Section 6.3.3.3 7.60E+07 Decay Coefficient of Segment, k (years -1 ), Section 6.3.3.1 7.30E+02 Total Load Reduction (CFU/year) 1.04E+12 Current Loading (CFU/year), Table 8.9 5.50E+12 Allowable Load (CFU/year) 4.46E+12 The total load reduction and allowable loading to Copano Bay were found by summing all the load reductions and current loadings for all four Copano Bay Segments. The load reductions, current loadings, and allowable loads to meet fecal coliform standards for Load Reduction Scenario #2 are shown in Table 8.20. Table 8.20 Load Reduction Scenario #2 at Copano Bay Portion of Bay Current Load (CFU/yr) Load Reductions (CFU/yr) Allowable Load (CFU/yr) Aransas Outlet (Segment 2) 5.04E+12 6.05E+11 4.43E+12 Mission Outlet (Segment 3) 5.50E+12 1.04E+12 4.46E+12 Copano Creek Outlet (Segment 4) 1.46E+12 0.00E+00 0.00E+00 Watershed JunctionID Outlet (Segment 1) 5.06E+10 0.00E+00 0.00E+00 Total Load 1.20E+13 1.65E+12 8.89E+12 291 Chapter 9: Conclusions and Recommendations 9.1 CONCLUSIONS Conclusions are presented based on the chapter from which the conclusions were drawn. All of these conclusions are based on the modeled results and the assumptions and calculations that are presented throughout this report. In Chapter 4, the bacterial monitoring data were analyzed throughout the Copano Bay watershed. From the analyses, the highest fecal coliform concentrations are found in the upstream rivers and streams; however, the rivers and streams have less stringent standards (i.e., contact recreation use) than Copano Bay (i.e., oyster water use). Within Copano Bay, the highest fecal coliform concentrations occur at the outlets where rivers and streams discharge into the Bay. At all of the bacterial monitoring stations in Copano Bay (from 1999-2005), the median fecal coliform concentrations are less than 14 CFU/100mL (the median fecal coliform standard in the Bay). In Section 4.2.3, all bacterial monitoring stations along the upstream rivers and streams meet fecal coliform contact recreation use standards based on available data from 1999-2004, except for station 17592. Station 17592 (upstream of Aransas River Above Tidal) does not comply with contact recreation use standards based on the available data from 1999-2004 and exceeds both criteria. However, this station does not monitor the water quality of TCEQ-defined water segments. In Copano Bay, Segments 2 and 3 exceed the fecal coliform oyster water use standard for the 90 th -percentile fecal coliform concentration based on available data from 1999-2005; however, Segments 1 and 4 comply with these water quality standards. 292 In Chapter 5, the annual bacterial loading calculations were made for all the point and non-point sources included in the models, and these loadings are the ‘input’ into both models. Based on the model assumptions and calculations, cattle are the main livestock contributors and contribute the greatest bacterial loading (input) compared to all other bacterial sources considered in the models. The upstream watersheds contribute the greatest bacterial loading; however, the loadings do not directly impact Copano Bay unless directly upstream up or adjacent to the Bay. Also, it was discovered towards the end of the analyses of this report that the WWTP bacterial loadings were greatly overestimated, so the WWTP loadings are even less of a bacterial contributor than what was presented in this report. Chapter 6 discusses the Schematic Processor Model and how it was calibrated to the median fecal coliform concentrations at each bacterial monitoring station. This Model models the average, annual conditions of the bacterial loadings in the Copano Bay watershed. Section 6.4 gives the results from the calibrated Schematic Processor Model. In this section, it was shown that the bacterial loadings decay very quickly; thus, at a point of interest, the bacterial loading from the watershed directly upstream will have the greatest impact on the receiving water quality. Thus, the watersheds that will most influence the quality of Copano Bay are the watersheds directly upstream and adjacent to the Bay because the bacteria have not had sufficient time to decay due to environmental conditions. After the modeling of bacterial transport (decay and CFSTRs simulated), cattle were found to be the greatest fecal coliform bacteria contributor to all Copano Bay segments based on model assumptions and calculations (shown in Figures 6.77 and 6.78). Note that Joanna Mott’s bacteria source tracking (BST) study (Mott, 2005) concluded 293 that cattle and horses contribute to fecal contamination at many of the Copano Bay stations when there is rainfall and high river flow. Wildlife (from non-point source calculations) and gulls (avian loading calculations) contribute relatively insignificant bacterial contamination to Copano Bay, which agrees with the findings from the BST study (Mott, 2005). The greatest bacterial loadings impact Copano Bay Segments 2 and 3 (shown in Figure 6.77). WWTP bacterial loadings are insignificant compared to non-point bacterial loadings (e.g., livestock, septic systems, and urban, forest runoff). Chapter 7 explains the Monte Carlo Simulation Model and how it was calibrated to the measured bacterial probability distributions at all of the bacterial monitoring stations. This Model models the variation in bacterial loadings throughout the year, accounting for seasonal, precipitation, runoff, bacterial loading, and temperature variations. Section 7.3.3 presents the load reductions necessary to meet fecal coliform water quality standards at the bacterial monitoring stations. Because Copano Bay Segments 1 and 4 meet water quality standards, no load reductions are necessary from the watersheds that drain to these portions of the Bay. However, load reductions are necessary for Copano Bay Segments 2 and 3. To meet water quality standards in the Bay, bacterial loadings from WWTPs and livestock need to be reduced in the watersheds that drain to these portions of the Bay. The reduction of bacterial loadings from WWTPs alone is not sufficient to meet standards in the Bay. Chapter 8 presents the current and allowable loadings to each of the water segments (Aransas River Above Tidal, Aransas River Tidal, Mission River Above Tidal, Mission River Tidal, and Copano Bay). 294 9.2 RECOMMENDATIONS To conduct a TMDL study for the Aransas River Above Tidal, Aransas River Tidal, Mission River Above Tidal, and Mission River Tidal, it is critical to create bacterial models that model the chosen primary bacterial indicator for each of these water segments. Thus, E. coli and enterococci models must be created for the Copano Bay watershed. One option to create these models is to find correlations between fecal coliform and E. coli / enterococci (presented in Section 2.1). One of the assumptions in the bacterial loading calculations (Chapter 5) was that all of the loading from livestock species was assumed to reach surface waters by either pasture runoff or direct discharge into the streams. There is a stakeholder concern that this overestimates livestock bacterial loadings. To see if these bacterial loadings are an overestimate, the event mean concentrations (EMCs) associated with land use types of agriculture, pasture, rangeland (land use classifications where bacterial loadings would come from primarily livestock species) of local studies should be compared to the livestock loadings calculated in these analyses. If there is a significant difference, then more research needs to be conducted to determine the fraction of the bacterial loadings that would reach surface waters from livestock species, taking into account location, time spent in water bodies, and survival rates of bacteria. More fecal coliform (or the bacterial indicator of interest based on location in stream network) monitoring should occur at WWTPs to ensure compliance with Texas Surface Water Quality Standards. However, from the modeled results, WWTP loadings were significantly less than livestock/non-point bacterial loadings (though these loadings directly discharge into surface waters). Also, from the BST study, human and sewage are not always the primary bacterial source in Copano Bay (Mott, 2005), so reductions from 295 WWTPs alone will not eliminate the bacteria contamination in the Bay. The WWTP bacterial loadings need to be re-calculated with the monitoring data from renewal permit files. The bacterial contribution from septic systems is very uncertain because it is difficult to quantify the bacterial loading that would reach surface and ground waters. Due to lack of data, it is recommended that a sensitivity analysis be conducted to determine the impact that the percentage of failing septic systems would have on Copano Bay. Feral hogs were not included as one of the potential bacteria sources in the Copano Bay watershed, and thus were not included in the bacterial loading calculations (Chapter 5.) However, at the Stakeholder’s Meeting in Refugio County on February 6, 2006, many stakeholders mentioned that feral hogs are prominent throughout the watershed and could be a major bacterial source directly impacting the quality of rivers, streams, and Copano Bay. Thus, bacterial loadings from feral hogs should be calculated and incorporated into the Schematic Processor and Monte Carlo Simulation Models. In the BST study, the following bacterial sources were analyzed: human (sewage), cow, horse, duck, gull and wildlife (Mott, 2005). All of these bacterial sources were accounted for in the models of this report, except for the duck populations, which were not included in the avian loading calculations. Based on the BST study, there are large populations of migratory ducks that inhabit the marsh areas that surround TDH stations COP 00013 and 00014 (near Aransas River outlet) and in the Mission Bay area (Mott, 2005). Thus, bacterial loadings from ducks should be calculated and incorporated into the Schematic Processor and Monte Carlo Simulation Models since ducks were found to be the major bacterial contributors in some of the storm events and stations studied and analyzed (Mott, 2005). 296 More monitoring data should be collected along the Aransas and Mission River Tidals. In the downstream portions of these two Tidals in the Monte Carlo Simulation Model, livestock, non-point, and WWTP bacterial loadings need to be significantly reduced to meet contact recreation use standards according to fecal coliform modeled results (Appendix 7.2). However, there are no monitoring data to conclude that there is a problem with complying with standards at these two locations. Monitoring data should be collected more frequently than quarterly. It is difficult to capture the variations and peaks in bacterial loadings with one bacteria measurement every three months. The more bacterial monitoring data that can be collected, the more measured data that can be used to ensure that the model is modeling what is occurring in the watershed. Agricultural Best Management Practices (BMPs) need to be implemented to reduce livestock bacterial loadings (the major modeled bacterial contributor) in the Copano Bay watershed. The final recommendation regards the Total Maximum Daily Load (TMDL) process in general. Useful information and feedback were obtained from each stakeholder meeting. Stakeholders are much more familiar with the occurrences in their watersheds than a modeler who does not live in the watershed. However, a majority of the work for the models was completed before the first stakeholder meeting. With each meeting, more useful information and feedback were given on how to improve the accuracy of the model. Since it is the stakeholders who end up being responsible for implementing BMPs, finding ways to reduce loads, and who must approve the plan and model before implementation, it is recommended that the stakeholders be involved throughout the entire process. If the stakeholders are involved from the beginning, I believe that the process will be more time and cost efficient since calculations would not 297 need to be continually redone; stakeholders would be able to provide useful input and feedback throughout the TMDL process rather than final comments. 298 Appendix 4.1: Bacterial Monitoring Data (1999-2005) of Copano Bay Segment 1 Station Date Fecal Coliform Concentration (CFU/100mL) Rank, m Probability of Exceedance, P (%) 13405 1/16/2001 390 1 0.55 13405 7/9/2002 290 2 1.44 13405 8/19/2003 220 3 2.32 14790 11/5/2002 170 4 3.20 13405 10/10/2000 104 5 4.08 13405 1/22/2003 82 6 4.97 14782 11/5/2002 79 7 5.85 14784 2/19/2003 79 8 6.73 14782 4/8/2004 64 9 7.62 14790 4/8/2004 46 10 8.50 14784 4/8/2004 33 11 9.38 14782 2/24/1999 15 12 10.27 13405 10/26/1999 13 13 11.15 14790 3/2/2004 13 14 12.03 14790 1/20/2005 13 15 12.91 13405 1/18/2000 8 16 13.80 14790 3/22/1999 8 17 14.68 14790 12/20/2004 8 18 15.56 13405 4/23/2003 7 19 16.45 14784 5/1/2002 7 20 17.33 14790 3/28/2005 7 21 18.21 13405 10/17/2002 6 22 19.10 13405 6/19/2001 5 23 19.98 14782 2/19/2003 5 24 20.86 14784 11/5/2002 5 25 21.74 14784 1/8/2004 5 26 22.63 14790 2/24/1999 5 27 23.51 14790 10/28/2004 5 28 24.39 14790 2/15/2005 5 29 25.28 13405 1/16/2002 4 30 26.16 13405 4/10/2002 4 31 27.04 13405 10/10/2001 3 32 27.93 13405 4/18/2000 2 33 28.81 14782 3/22/1999 2 34 29.69 299 14782 4/27/1999 2 35 30.57 14782 10/11/1999 2 36 31.46 14782 11/8/1999 2 37 32.34 14782 12/29/1999 2 38 33.22 14782 1/31/2000 2 39 34.11 14782 2/8/2000 2 40 34.99 14782 3/9/2000 2 41 35.87 14782 3/20/2000 2 42 36.76 14782 3/28/2000 2 43 37.64 14782 4/26/2000 2 44 38.52 14782 12/20/2000 2 45 39.40 14782 2/1/2001 2 46 40.29 14782 5/1/2002 2 47 41.17 14782 10/21/2002 2 48 42.05 14782 12/11/2002 2 49 42.94 14782 1/6/2003 2 50 43.82 14782 2/27/2003 2 51 44.70 14782 1/8/2004 2 52 45.59 14782 2/17/2004 2 53 46.47 14782 2/26/2004 2 54 47.35 14782 3/2/2004 2 55 48.23 14782 10/28/2004 2 56 49.12 14782 11/8/2004 2 57 50.00 14782 12/20/2004 2 58 50.88 14782 1/20/2005 2 59 51.77 14782 2/15/2005 2 60 52.65 14782 3/28/2005 2 61 53.53 14784 2/24/1999 2 62 54.42 14784 3/22/1999 2 63 55.30 14784 4/27/1999 2 64 56.18 14784 10/11/1999 2 65 57.06 14784 11/8/1999 2 66 57.95 14784 12/29/1999 2 67 58.83 14784 1/31/2000 2 68 59.71 14784 2/8/2000 2 69 60.60 14784 3/9/2000 2 70 61.48 14784 3/20/2000 2 71 62.36 14784 3/28/2000 2 72 63.25 14784 4/26/2000 2 73 64.13 14784 12/20/2000 2 74 65.01 14784 2/1/2001 2 75 65.89 14784 10/21/2002 2 76 66.78 300 14784 12/11/2002 2 77 67.66 14784 1/6/2003 2 78 68.54 14784 2/27/2003 2 79 69.43 14784 2/17/2004 2 80 70.31 14784 2/26/2004 2 81 71.19 14784 3/2/2004 2 82 72.08 14784 10/28/2004 2 83 72.96 14784 11/8/2004 2 84 73.84 14784 12/20/2004 2 85 74.72 14784 1/20/2005 2 86 75.61 14784 2/15/2005 2 87 76.49 14784 3/28/2005 2 88 77.37 14790 4/27/1999 2 89 78.26 14790 10/11/1999 2 90 79.14 14790 11/8/1999 2 91 80.02 14790 12/29/1999 2 92 80.91 14790 1/31/2000 2 93 81.79 14790 2/8/2000 2 94 82.67 14790 3/9/2000 2 95 83.55 14790 3/20/2000 2 96 84.44 14790 3/28/2000 2 97 85.32 14790 4/26/2000 2 98 86.20 14790 12/20/2000 2 99 87.09 14790 2/1/2001 2 100 87.97 14790 5/1/2002 2 101 88.85 14790 10/21/2002 2 102 89.74 14790 12/11/2002 2 103 90.62 14790 1/6/2003 2 104 91.50 14790 2/19/2003 2 105 92.38 14790 2/27/2003 2 106 93.27 14790 2/17/2004 2 107 94.15 14790 1/8/2004 2 108 95.03 14790 2/26/2004 2 109 95.92 14790 11/8/2004 2 110 96.80 13405 7/12/2000 1 111 97.68 14784 1/5/2000 1 112 98.57 14784 6/22/2000 1 113 99.45 301 Appendix 4.2: Bacterial Monitoring Data (1999-2005) of Copano Bay Segment 2 Station Date Fecal Coliform Concentration (CFU/100mL) Rank, m Probability of Exceedance, P (%) 14788 4/27/1999 1600 1 0.52 14788 3/20/2000 1600 2 1.34 14783 4/8/2004 1600 3 2.16 14788 4/8/2004 540 4 2.99 12945 1/21/2003 400 5 3.81 14783 11/5/2002 350 6 4.64 14788 11/5/2002 240 7 5.46 14787 11/5/2002 220 8 6.29 12945 7/8/2002 145 9 7.11 14787 4/8/2004 130 10 7.94 12945 8/18/2003 118 11 8.76 14783 2/19/2003 110 12 9.59 14788 3/28/2000 79 13 10.41 14788 2/27/2003 79 14 11.24 12945 4/17/2000 60 15 12.06 12945 4/22/2003 58 16 12.89 14783 3/20/2000 49 17 13.71 14783 11/13/2002 48 18 14.54 14783 5/18/1999 45 19 15.36 12945 1/14/2002 39 20 16.19 12945 4/9/2002 39 21 17.01 12945 4/10/2001 37 22 17.84 14788 2/26/2004 33 23 18.66 14787 2/19/2003 33 24 19.48 14787 2/26/2004 33 25 20.31 14783 4/27/1999 33 26 21.13 14783 3/28/2005 33 27 21.96 12945 10/8/2001 29 28 22.78 14783 3/28/2000 27 29 23.61 14788 2/19/2003 23 30 24.43 14787 2/27/2003 23 31 25.26 14787 3/28/2005 23 32 26.08 14788 3/28/2005 17 33 26.91 12945 1/19/2000 16 34 27.73 12945 1/15/2001 15 35 28.56 302 12945 10/15/2002 14 36 29.38 12945 10/9/2000 14 37 30.21 14787 3/20/2000 13 38 31.03 14787 12/11/2002 11 39 31.86 12945 10/25/1999 10 40 32.68 14788 2/24/1999 9 41 33.51 14783 12/11/2002 8 42 34.33 14783 2/27/2003 8 43 35.15 14788 10/21/2002 7 44 35.98 14788 1/8/2004 7 45 36.80 14783 2/15/2005 7 46 37.63 12945 6/18/2001 6 47 38.45 14788 3/22/1999 5 48 39.28 14788 5/1/2002 5 49 40.10 14787 2/24/1999 5 50 40.93 14787 3/28/2000 5 51 41.75 14787 1/8/2004 5 52 42.58 14788 1/20/2005 4 53 43.40 14783 2/16/1999 4 54 44.23 14783 7/30/2002 4 55 45.05 14788 11/8/1999 2 56 45.88 14788 12/29/1999 2 57 46.70 14788 1/31/2000 2 58 47.53 14788 2/8/2000 2 59 48.35 14788 3/9/2000 2 60 49.18 14788 4/26/2000 2 61 50.00 14788 12/20/2000 2 62 50.82 14788 2/1/2001 2 63 51.65 14788 12/11/2002 2 64 52.47 14788 1/6/2003 2 65 53.30 14788 2/17/2004 2 66 54.12 14788 3/2/2004 2 67 54.95 14788 10/28/2004 2 68 55.77 14788 11/8/2004 2 69 56.60 14788 12/20/2004 2 70 57.42 14788 2/15/2005 2 71 58.25 14787 3/22/1999 2 72 59.07 14787 4/27/1999 2 73 59.90 14787 11/8/1999 2 74 60.72 14787 12/29/1999 2 75 61.55 14787 1/31/2000 2 76 62.37 14787 2/8/2000 2 77 63.20 303 14787 3/9/2000 2 78 64.02 14787 4/26/2000 2 79 64.85 14787 12/20/2000 2 80 65.67 14787 2/1/2001 2 81 66.49 14787 5/1/2002 2 82 67.32 14787 10/21/2002 2 83 68.14 14787 1/6/2003 2 84 68.97 14787 2/17/2004 2 85 69.79 14787 3/2/2004 2 86 70.62 14787 10/28/2004 2 87 71.44 14787 11/8/2004 2 88 72.27 14787 12/20/2004 2 89 73.09 14787 1/20/2005 2 90 73.92 14787 3/15/2005 2 91 74.74 14783 2/24/1999 2 92 75.57 14783 3/22/1999 2 93 76.39 14783 11/8/1999 2 94 77.22 14783 12/29/1999 2 95 78.04 14783 1/31/2000 2 96 78.87 14783 2/8/2000 2 97 79.69 14783 3/9/2000 2 98 80.52 14783 4/26/2000 2 99 81.34 14783 12/20/2000 2 100 82.16 14783 2/1/2001 2 101 82.99 14783 2/14/2002 2 102 83.81 14783 3/28/2002 2 103 84.64 14783 5/1/2002 2 104 85.46 14783 10/21/2002 2 105 86.29 14783 1/6/2003 2 106 87.11 14783 1/30/2003 2 107 87.94 14783 1/8/2004 2 108 88.76 14783 2/17/2004 2 109 89.59 14783 2/26/2004 2 110 90.41 14783 3/2/2004 2 111 91.24 14783 10/28/2004 2 112 92.06 14783 11/8/2004 2 113 92.89 14783 12/20/2004 2 114 93.71 14783 1/20/2005 2 115 94.54 14783 7/19/1999 1 116 95.36 14783 1/5/2000 1 117 96.19 14783 6/22/2000 1 118 97.01 14783 10/3/2000 1 119 97.84 304 14783 1/24/2001 1 120 98.66 12945 7/11/2000 1 121 99.48 305 Appendix 4.3: Bacterial Monitoring Data (1999-2005) of Copano Bay Segment 3 Station Date Fecal Coliform Concentration (CFU/100mL) Rank, m Probability of Exceedance, P (%) 14797 4/8/2004 240 1 2.00 14797 11/5/2002 220 2 5.20 14797 2/19/2003 70 3 8.40 14797 3/20/2000 49 4 11.60 14797 2/24/1999 22 5 14.80 14797 2/27/2003 17 6 18.00 14797 1/8/2004 13 7 21.20 14797 3/2/2004 5 8 24.40 14797 10/28/2004 5 9 27.60 14797 3/22/1999 2 10 30.80 14797 4/27/1999 2 11 34.00 14797 11/8/1999 2 12 37.20 14797 12/29/1999 2 13 40.40 14797 1/31/2000 2 14 43.60 14797 2/8/2000 2 15 46.80 14797 3/9/2000 2 16 50.00 14797 3/28/2000 2 17 53.20 14797 4/26/2000 2 18 56.40 14797 12/20/2000 2 19 59.60 14797 2/1/2001 2 20 62.80 14797 5/1/2002 2 21 66.00 14797 10/21/2002 2 22 69.20 14797 12/11/2002 2 23 72.40 14797 1/6/2003 2 24 75.60 14797 2/17/2004 2 25 78.80 14797 2/26/2004 2 26 82.00 14797 11/8/2004 2 27 85.20 14797 12/20/2004 2 28 88.40 14797 1/20/2005 2 29 91.60 14797 2/15/2005 2 30 94.80 14797 2/15/2005 2 31 98.00 306 Appendix 4.4: Bacterial Monitoring Data (1999-2005) of Copano Bay Segment 4 Station Date Fecal Coliform Concentration (CFU/100mL) Rank, m Probability of Exceedance, P (%) 14792 2/19/2003 1600 1 0.27 14792 2/26/2004 1600 2 0.70 14792 12/11/2002 540 3 1.13 14779 11/6/2002 350 4 1.56 14792 11/5/2002 350 5 1.99 14793 11/5/2002 130 6 2.42 14792 4/8/2004 110 7 2.85 14793 4/27/1999 79 8 3.28 13404 6/19/2001 70 9 3.71 13404 7/9/2002 57 10 4.14 14792 4/27/1999 49 11 4.57 14793 1/8/2004 49 12 5.01 14792 2/24/1999 23 13 5.44 14792 2/15/2005 22 14 5.87 14793 3/20/2000 22 15 6.30 13404 4/9/2001 21 16 6.73 13404 10/10/2001 21 17 7.16 13404 1/22/2003 14 18 7.59 13404 4/23/2003 14 19 8.02 14779 12/11/2002 14 20 8.45 14780 12/11/2002 14 21 8.88 13404 10/10/2000 13 22 9.31 14780 11/5/2002 13 23 9.74 14780 2/27/2003 13 24 10.17 14793 3/9/2000 13 25 10.60 14793 3/2/2004 13 26 11.03 14793 2/26/2004 12 27 11.46 13404 11/5/2002 11 28 11.89 13404 1/8/2004 11 29 12.33 14785 11/5/2002 11 30 12.76 14779 3/24/2005 8 31 13.19 14780 3/2/2004 8 32 13.62 14792 2/27/2003 8 33 14.05 13404 12/11/2002 7 34 14.48 14779 4/27/1999 7 35 14.91 307 14779 12/11/2002 7 36 15.34 14793 2/15/2005 7 37 15.77 13404 1/16/2001 6 38 16.20 14779 11/5/2002 5 39 16.63 14779 1/6/2003 5 40 17.06 14779 2/26/2004 5 41 17.49 14779 1/20/2005 5 42 17.92 14785 2/24/1999 5 43 18.35 14785 2/19/2003 5 44 18.78 14792 1/31/2000 5 45 19.21 14792 2/17/2004 5 46 19.64 14792 3/2/2004 5 47 20.08 14793 2/1/2001 5 48 20.51 14793 12/11/2002 5 49 20.94 14793 4/8/2004 5 50 21.37 14780 1/20/2005 4 51 21.80 14793 2/19/2003 4 52 22.23 13404 4/18/2000 3 53 22.66 13404 10/17/2002 3 54 23.09 14793 12/20/2004 3 55 23.52 13404 10/26/1999 2 56 23.95 13404 2/24/1999 2 57 24.38 13404 3/22/1999 2 58 24.81 13404 4/27/1999 2 59 25.24 13404 10/11/1999 2 60 25.67 13404 11/8/1999 2 61 26.10 13404 12/29/1999 2 62 26.53 13404 1/31/2000 2 63 26.96 13404 2/8/2000 2 64 27.40 13404 3/9/2000 2 65 27.83 13404 3/20/2000 2 66 28.26 13404 3/28/2000 2 67 28.69 13404 4/26/2000 2 68 29.12 13404 12/20/2000 2 69 29.55 13404 2/1/2001 2 70 29.98 13404 1/16/2002 2 71 30.41 13404 4/10/2002 2 72 30.84 13404 5/1/2002 2 73 31.27 13404 10/21/2002 2 74 31.70 13404 1/6/2003 2 75 32.13 13404 2/19/2003 2 76 32.56 13404 2/27/2003 2 77 32.99 308 13404 2/17/2004 2 78 33.42 13404 2/26/2004 2 79 33.85 13404 3/2/2004 2 80 34.28 13404 4/8/2004 2 81 34.71 13404 10/28/2004 2 82 35.15 13404 11/8/2004 2 83 35.58 13404 12/20/2004 2 84 36.01 13404 1/20/2005 2 85 36.44 13404 2/15/2005 2 86 36.87 14779 1/13/1999 2 87 37.30 14779 2/11/1999 2 88 37.73 14779 2/24/1999 2 89 38.16 14779 3/18/1999 2 90 38.59 14779 3/22/1999 2 91 39.02 14779 4/13/1999 2 92 39.45 14779 10/11/1999 2 93 39.88 14779 10/12/1999 2 94 40.31 14779 11/8/1999 2 95 40.74 14779 11/9/1999 2 96 41.17 14779 12/13/1999 2 97 41.60 14779 12/29/1999 2 98 42.03 14779 1/31/2000 2 99 42.47 14779 1/31/2000 2 100 42.90 14779 2/8/2000 2 101 43.33 14779 2/8/2000 2 102 43.76 14779 3/9/2000 2 103 44.19 14779 3/20/2000 2 104 44.62 14779 3/28/2000 2 105 45.05 14779 3/28/2000 2 106 45.48 14779 4/26/2000 2 107 45.91 14779 4/26/2000 2 108 46.34 14779 12/11/2000 2 109 46.77 14779 12/20/2000 2 110 47.20 14779 1/23/2001 2 111 47.63 14779 2/1/2001 2 112 48.06 14779 5/1/2002 2 113 48.49 14779 5/6/2002 2 114 48.92 14779 10/21/2002 2 115 49.35 14779 1/8/2003 2 116 49.78 14779 2/11/2003 2 117 50.22 14779 2/19/2003 2 118 50.65 14779 2/27/2003 2 119 51.08 309 14779 1/8/2004 2 120 51.51 14779 1/20/2004 2 121 51.94 14779 1/28/2004 2 122 52.37 14779 2/17/2004 2 123 52.80 14779 2/24/2004 2 124 53.23 14779 3/2/2004 2 125 53.66 14779 3/25/2004 2 126 54.09 14779 4/8/2004 2 127 54.52 14779 10/19/2004 2 128 54.95 14779 10/28/2004 2 129 55.38 14779 11/8/2004 2 130 55.81 14779 11/10/2004 2 131 56.24 14779 12/2/2004 2 132 56.67 14779 12/20/2004 2 133 57.10 14779 1/24/2005 2 134 57.53 14779 2/4/2005 2 135 57.97 14779 2/15/2005 2 136 58.40 14779 3/28/2005 2 137 58.83 14780 2/24/1999 2 138 59.26 14780 3/22/1999 2 139 59.69 14780 4/27/1999 2 140 60.12 14780 10/11/1999 2 141 60.55 14780 11/8/1999 2 142 60.98 14780 12/29/1999 2 143 61.41 14780 1/31/2000 2 144 61.84 14780 2/8/2000 2 145 62.27 14780 3/9/2000 2 146 62.70 14780 3/20/2000 2 147 63.13 14780 3/28/2000 2 148 63.56 14780 4/26/2000 2 149 63.99 14780 12/20/2000 2 150 64.42 14780 2/1/2001 2 151 64.85 14780 5/1/2002 2 152 65.29 14780 10/21/2002 2 153 65.72 14780 1/6/2003 2 154 66.15 14780 2/19/2003 2 155 66.58 14780 1/8/2004 2 156 67.01 14780 2/17/2004 2 157 67.44 14780 2/26/2004 2 158 67.87 14780 4/8/2004 2 159 68.30 14780 10/28/2004 2 160 68.73 14780 11/8/2004 2 161 69.16 310 14780 12/20/2004 2 162 69.59 14780 2/15/2005 2 163 70.02 14785 3/22/1999 2 164 70.45 14785 4/27/1999 2 165 70.88 14785 10/11/1999 2 166 71.31 14785 11/8/1999 2 167 71.74 14785 12/29/1999 2 168 72.17 14785 1/31/2000 2 169 72.60 14785 2/8/2000 2 170 73.04 14785 3/9/2000 2 171 73.47 14785 3/20/2000 2 172 73.90 14785 3/28/2000 2 173 74.33 14785 4/26/2000 2 174 74.76 14785 12/20/2000 2 175 75.19 14785 2/1/2001 2 176 75.62 14785 5/1/2002 2 177 76.05 14785 10/21/2002 2 178 76.48 14785 12/11/2002 2 179 76.91 14785 1/6/2003 2 180 77.34 14785 2/27/2003 2 181 77.77 14785 1/8/2004 2 182 78.20 14785 2/17/2004 2 183 78.63 14785 2/26/2004 2 184 79.06 14785 3/2/2004 2 185 79.49 14785 4/8/2004 2 186 79.92 14785 10/28/2004 2 187 80.36 14785 11/8/2004 2 188 80.79 14785 12/20/2004 2 189 81.22 14785 1/20/2005 2 190 81.65 14785 2/15/2005 2 191 82.08 14785 3/28/2005 2 192 82.51 14792 3/22/1999 2 193 82.94 14792 10/11/1999 2 194 83.37 14792 11/8/1999 2 195 83.80 14792 12/29/1999 2 196 84.23 14792 2/8/2000 2 197 84.66 14792 3/9/2000 2 198 85.09 14792 3/20/2000 2 199 85.52 14792 3/28/2000 2 200 85.95 14792 4/26/2000 2 201 86.38 14792 12/20/2000 2 202 86.81 14792 2/1/2001 2 203 87.24 311 14792 5/1/2002 2 204 87.67 14792 10/21/2002 2 205 88.11 14792 1/6/2003 2 206 88.54 14792 1/8/2004 2 207 88.97 14792 10/28/2004 2 208 89.40 14792 11/8/2004 2 209 89.83 14792 12/20/2004 2 210 90.26 14792 1/20/2005 2 211 90.69 14793 2/24/1999 2 212 91.12 14793 3/22/1999 2 213 91.55 14793 10/11/1999 2 214 91.98 14793 11/8/1999 2 215 92.41 14793 12/29/1999 2 216 92.84 14793 1/31/2000 2 217 93.27 14793 2/8/2000 2 218 93.70 14793 3/28/2000 2 219 94.13 14793 4/26/2000 2 220 94.56 14793 12/20/2000 2 221 94.99 14793 5/1/2002 2 222 95.43 14793 10/21/2002 2 223 95.86 14793 1/6/2003 2 224 96.29 14793 2/27/2003 2 225 96.72 14793 2/17/2004 2 226 97.15 14793 10/28/2004 2 227 97.58 14793 11/8/2004 2 228 98.01 14793 1/20/2005 2 229 98.44 13404 1/18/2000 1 230 98.87 13404 7/12/2000 1 231 99.30 13404 8/19/2003 1 232 99.73 312 Appendix 5.1: Terrain Preprocessing For this project, and to allow the use of Water Rights Analysis Package (WRAP) Hydro, the only steps that were implemented from Terrain Preprocessing (located in the Arc Hydro Toolbar) were DEM Reconditioning, Fill Sinks, and Flow Direction. Before starting the process, the DEM was clipped to the watershed basin by going to Spatial Analyst | Options and changing the “Analysis Mask” to the subbasin feature class. Then I went to Spatial Analyst | Raster Calculator and evaluated the DEM to obtain the clipped DEM. DEM Reconditioning 1. Select Terrain Preprocessing | DEM Reconditioning. 2. Select the clipped DEM as the “Raw DEM”. 3. Select the modified NHDFlowline (with all the river segments connected) as the “Agree Stream”. 4. Keep all the default settings, and the output will be “AgreeDEM”. 5. Press OK, and the “AgreeDEM” layer will be added to the map. Fill Sinks 1. Select Terrain Preprocessing | Fill Sinks. 2. Select AgreeDEM as "DEM". 3. Keep all the default settings, and the output will be "Fil". 4. Press OK, and the "Fil" layer will be added to the map. Flow Direction 1. Select Terrain Preprocessing | Flow Direction. 2. Select Fil as “Hydro DEM”. 3. Keep all the default settings, and the output will be “Fdr”. 4. Press OK, and the “Fdr” layer will be added to the map. 313 Appendix 5.2: WRAP Hydro Process Water Rights Analysis Package (WRAP) Hydro, which is a toolbar located in Arc GIS, is used to delineate watersheds. The watersheds were delineated to the Critical Points (USGS gauge stations, water segment endpoints, and bacterial monitoring stations.) Create Geometric Network 1. Using Arc Catalog, create a personal geodatabase called “WRAPHydro” within a chosen directory. 2. Create a feature dataset (called “WRAPHydro”) within the Geodatabase, and use the projection: NAD 1983 Texas Centric Mapping System Albers. This will maintain the area, which is crucial in maintaining drainage areas for non-point source calculations. 3. Import NHDFlowline (with all the river segments connected) into the feature dataset, and rename it “WRAPFlowline”. 4. Import “CriticalPoints”, which is the created feature class that contains the USGS gauge stations, bacteria monitoring stations, and water segment endpoints. (Note: before creating a geometric network, the Editor Toolbar in Arc GIS needs to be implemented to ensure that all the critical points are connected to the river network (WRAPFlowline). “Critical Points” is the target layer, and “Modify Feature” is the task. Go to Editor | Snapping… and check the box to allow the critical points to snap to the edge, WRAPFlowline. This allows one to move and snap the critical points to WRAPFlowline. 5. Right-click on the feature dataset in Arc Catalog, and go to New | Geometric Network… 6. Hit “Next”, and select “Build a geometric network from existing features.” 7. Select “WRAPFlowline” and “CriticalPoints”, name the geometric network, and hit “Next”. 8. Select “Yes”, so the complex edges will be in the network. 9. Keep all the default settings for the rest of the options, and hit “Finish”. Assign HydroIDs to the Edges 1. In the Arc Hydro Toolbar in Arc Map, go to Attribute Tools | Assign HydroID. 2. Select the WRAPFlowline and CriticalPoints layers. 3. Say “Yes” to overwrite existing HydroIDs, apply to selected features, and press “OK”. 314 Delineate Watersheds 1. Make sure the WRAP Hydro Toolbar is open in Arc Map. 2. Set spatial extent. 3. Using the Spatial Analyst Toolbar, go to Spatial Analyst | Options… 4. Select the “Extent” tab, and make sure that the Fdr or DEM grid is selected for the Analysis Extent. 5. Set flow direction. 6. Using the Arc Hydro Toolbar, go to Network Tools | Set Flow Direction… 7. Select the WRAPFlowline layer and assign based on Fdr (flow direction grid that was created in Terrain Preprocessing) attribute, and press “OK”. 8. Using the WRAPHydro Toolbar, go to Settings | Layers 9. Set “WRAPJunction” to CriticalPoints. 10. Set “HydroEdge” to WRAPFlowline. 11. Set “Flow Direction Raster” to Fdr. 12. Go to Options | Delineate Watershed. 13. Set “Source Layer” as WRAPFlowline. 14. Set “Source Attribute” as JunctionID. 15. Set the Drainage Area Units as square meters. 16. Click on “Batch Process WRAPJunctions” from the WRAP Hydro menu to delineate the watersheds. 17. Clip the watersheds, so that Copano Bay is excluded from the watershed areas. 315 Appendix 5.3: Precipitation Rasters for Land Use Classifications To calculate the runoff for each land use classification, the precipitation grid was divided into four different rasters based on land use classifications. Create Feature Classes of Different Land Uses 1. Use “Raster to Polygon” tool in Arc Toolbox to convert the land use land cover raster to a polygon feature class. 2. Right-click on land use land cover feature class (in Arc Map), go to Properties | Definition Query. 3. Select “Query Builder…” 4. Double-click on [GRIDCODE] (the field that contains the land use codes), “=” and select one of the grid code values that can be classified as either Agriculture, Forest, Urban, or Open Water. If there are multiple grid codes that could be Agriculture, Forest, Urban, or Open Water, click “AND”, and repeat step 4. (Note: what grid codes are associated with which land use classification is open to interpretation.) 5. After conducting a query for one of the land use classifications, then select all the polygons in the Arc Map view. 6. Right-click on the land use land cover polygon feature class, and Data | Export Data… and create a new feature class for that specific land use classification. 7. Repeat steps #4-6 until you have four new feature classes (Agriculture, Forest, Urban, and Open Water.) Create Precipitation Rasters for Land Use Classifications 1. Go to Spatial Analyst | Options… 2. Set the “Analysis mask” to one of the land use classification feature classes (Agriculture, Forest, Urban, Open Water). 3. Set the “Extent” and “Cell Size” to the land use land cover raster. 4. Go to Spatial Analyst | Raster Calculator… 5. Double-click on the precipitation raster, P, and “Evaluate”. 6. Right-click on the Calculation raster and Make Permanent. 7. Repeat steps #1-6 for the other three land use classifications. (You will now have the original precipitation raster divided into four precipitation rasters based on the four different land use classifications.) 316 Appendix 5.4: Livestock Loading Calculations and Results The calculations for bacterial loading due to livestock are shown in Table 5A.1, and the annual bacterial loadings due to livestock per watershed are given in Table 5A.2. Table 5A.1 includes the area of each county, the area of each county where animals were assumed to be located, and the census data for each county. The census data are from 2004 if these data existed; otherwise, the data are from 2002. The area of each watershed within each county (area where animals would be located) is also given, as well as the calculated livestock count and bacterial loading for each watershed. The locations of the Watershed JunctionIDs (used in Tables 5A.1 and 5A.2) are shown in Figure 5A.1. Table 5A.1 Livestock Loading Calculations and Results County Bee San Patricio Aransas Refugio Goliad Karnes Watershed Totals Area (m 2 ) 2344047042 1798057954 742112132 2016188169 2286808461 1965399713 Cattle 111433 Shrub./Pasture/Hay (m 2 ) 1382538600 472358700 207011700 1141603200 1462038300 1286240400 Horses 2561 Cattle/Calves 49000 20000 2000 36000 66000 74000 Goat 2299 Goats 2100 773 75 200 795 2100 Sheep 659 Horses/Ponies 1391 662 46 692 887 973 Sheep/Lambs 670 0071 0 327 Deer 0 0 0 0 Layers (20 weeks +) 793 464 0 63 859 0 Hogs and Pigs 0 741 0 0 0 0 Meat-type 0 00 0 252 0 317 Chickens Cattle/m 2 3.54E-05 4.23E-05 9.66E-06 3.15E-05 4.51E-05 5.75E-05 Goats/m 2 1.52E-06 1.64E-06 3.62E-07 1.75E-07 5.44E-07 1.63E-06 Horses/m 2 1.01E-06 1.40E-06 2.22E-07 6.06E-07 6.07E-07 7.56E-07 Sheep/m 2 4.85E-07 0.00E+00 0.00E+00 6.22E-08 0.00E+00 2.54E-07 Deer/m 2 0.00E+00 0.00E+00 0.00E+00 0.00E+00 0.00E+00 0.00E+00 Layers/m 2 5.74E-07 9.82E-07 0.00E+00 5.52E-08 5.88E-07 0.00E+00 Hogs/m 2 0.00E+00 1.57E-06 0.00E+00 0.00E+00 0.00E+00 0.00E+00 Chickens/m 2 0.00E+00 0.00E+00 0.00E+00 0.00E+00 1.72E-07 0.00E+00 WATERSHEDS (JunctionID) 45422 Shrub./Pasture/Hay (m 2 ) 286448400 45632700 Count CFU/year Cattle 10152.32 2625.34 12777.66 2.52E+16 Goats 435.10 74.50 509.60 2.79E+15 Horses 288.20 34.52 322.72 4.95E+13 Sheep 138.82 11.60 150.42 8.24E+14 Layers 164.30 0.00 164.30 7.57E+12 Hogs 0.00 0.00 0.00 0.00E+00 Chickens 0.00 0.00 0.00 0.00E+00 45408 Shrub./Pasture/Hay (m 2 ) 288088200 42202800 38923200 Count CFU/year Cattle 10210.44 1786.90 1227.43 13224.76 2.61E+16 Goats 437.59 69.06 6.82 437.59 2.40E+15 Horses 289.85 59.15 23.59 372.59 5.71E+13 Sheep 139.61 0.00 2.42 142.03 7.78E+14 318 Layers 165.24 41.46 2.15 208.85 9.62E+12 Hogs 0.00 66.20 0.00 66.20 2.40E+14 Chickens 0.00 0.00 0.00 0.00 0.00E+00 45426 Shrub./Pasture/Hay (m 2 ) 64056600 Count CFU/year Cattle 2712.20 2712.20 5.35E+15 Goats 104.83 104.83 5.74E+14 Horses 89.77 89.77 1.38E+13 Sheep 0.00 0.00 0.00E+00 Layers 62.92 62.92 2.90E+12 Hogs 100.49 100.49 3.65E+14 Chickens 0.00 0.00 0.00E+00 45414 Shrub./Pasture/Hay (m 2 ) 61632900 10715400 Count CFU/year Cattle 2609.58 337.91 2947.49 5.81E+15 Goats 100.86 1.88 102.74 5.62E+14 Horses 86.38 6.50 92.87 1.42E+13 Sheep 0.00 0.67 0.67 3.65E+12 Layers 60.54 0.59 61.13 2.82E+12 Hogs 96.68 0.00 96.68 3.51E+14 Chickens 0.00 0.00 0.00 0.00E+00 45416 Shrub./Pasture/Hay (m 2 ) 5400 106298100 2049300 4218300 Count CFU/year Cattle 0.19 4500.74 19.80 133.02 4653.75 9.17E+15 Goats 0.01 173.95 0.74 0.74 0.01 4.49E+10 Horses 0.01 148.97 0.46 2.56 151.99 2.33E+13 319 Sheep 0.00 0.00 0.00 0.26 0.26 1.45E+12 Layers 0.00 104.42 0.00 0.23 104.65 4.82E+12 Hogs 0.00 166.75 0.00 0.00 166.75 6.06E+14 Chickens 0.00 0.00 0.00 0.00 0.00 0.00E+00 45405 Shrub./Pasture/Hay (m 2 ) 35626500 55712700 Count CFU/year Cattle 1508.45 538.26 2046.71 4.03E+15 Goats 58.30 20.18 78.49 4.30E+14 Horses 49.93 12.38 62.31 9.55E+12 Sheep 0.00 0.00 0.00 0.00E+00 Layers 35.00 0.00 35.00 1.61E+12 Hogs 55.89 0.00 55.89 2.03E+14 Chickens 0.00 0.00 0.00 0.00E+00 45421 Shrub./Pasture/Hay (m 2 ) 124290000 5258700 Count CFU/year Cattle 3919.44 237.39 4156.83 8.19E+15 Goats 21.77 2.86 24.63 1.35E+14 Horses 75.34 3.19 78.53 1.20E+13 Sheep 7.73 0.00 7.73 4.23E+13 Layers 6.86 3.09 9.95 4.58E+11 Hogs 0.00 0.00 0.00 0.00E+00 Chickens 0.00 0.91 0.91 6.29E+10 45417 Shrub./Pasture/Hay (m 2 ) 51879600 409879800 152252100 Count CFU/year Cattle 1838.72 12925.40 6873.03 21637.15 4.26E+16 Goats 78.80 71.81 82.79 78.80 4.31E+14 320 Horses 52.20 248.45 92.37 393.02 6.03E+13 Sheep 25.14 25.49 0.00 50.63 2.77E+14 Layers 29.76 22.62 89.45 141.83 6.53E+12 Hogs 0.00 0.00 0.00 0.00 0.00E+00 Chickens 0.00 0.00 26.24 26.24 1.82E+12 45404 Shrub./Pasture/Hay (m 2 ) 143850600 Count CFU/year Cattle 5098.36 5098.36 1.00E+16 Goats 218.50 218.50 1.20E+15 Horses 144.73 144.73 2.22E+13 Sheep 69.71 69.71 3.82E+14 Layers 82.51 82.51 3.80E+12 Hogs 0.00 0.00 0.00E+00 Chickens 0.00 0.00 0.00E+00 45409 Shrub./Pasture/Hay (m 2 ) 16466400 Count CFU/year Cattle 583.60 583.60 1.15E+15 Goats 25.01 25.01 1.37E+14 Horses 16.57 16.57 2.54E+12 Sheep 7.98 7.98 4.37E+13 Layers 9.44 9.44 4.35E+11 Hogs 0.00 0.00 0.00E+00 Chickens 0.00 0.00 0.00E+00 45415 Shrub./Pasture/Hay (m 2 ) 105787800 Count CFU/year Cattle 3749.34 3749.34 7.39E+15 321 Goats 160.69 160.69 8.80E+14 Horses 106.44 106.44 1.63E+13 Sheep 51.27 51.27 2.81E+14 Layers 60.68 60.68 2.79E+12 Hogs 0.00 0.00 0.00E+00 Chickens 0.00 0.00 0.00E+00 45419 Shrub./Pasture/Hay (m 2 ) 254346300 66544200 406713600 5850900 Count CFU/year Cattle 9014.55 2098.44 18360.05 336.61 29809.66 5.88E+16 Goats 386.34 11.66 221.16 9.55 395.89 2.17E+15 Horses 255.90 40.34 246.75 4.43 547.41 8.39E+13 Sheep 123.26 4.14 0.00 1.49 128.89 7.06E+14 Layers 145.89 3.67 238.96 0.00 388.52 1.79E+13 Hogs 0.00 0.00 0.00 0.00 0.00 0.00E+00 Chickens 0.00 0.00 70.10 0.00 70.10 4.86E+12 45413 Shrub./Pasture/Hay (m 2 ) 81936900 Count CFU/year Cattle 2904.01 2904.01 5.72E+15 Goats 124.46 124.46 6.81E+14 Horses 82.44 82.44 1.26E+13 Sheep 39.71 39.71 2.17E+14 Layers 47.00 47.00 2.16E+12 Hogs 0.00 0.00 0.00E+00 Chickens 0.00 0.00 0.00E+00 56830 Shrub./Pasture/Hay (m 2 ) 18531000 94962600 Count CFU/year 322 Cattle 179.03 2994.61 3173.64 6.26E+15 Goats 6.71 16.64 23.35 1.28E+14 Horses 4.12 57.56 61.68 9.46E+12 Sheep 0.00 5.91 5.91 3.23E+13 Layers 0.00 5.24 5.24 2.41E+11 Hogs 0.00 0.00 0.00 0.00E+00 Chickens 0.00 0.00 0.00 0.00E+00 56831 Shrub./Pasture/Hay (m 2 ) 10971000 Count CFU/year Cattle 105.99 105.99 2.09E+14 Goats 3.97 3.97 2.18E+13 Horses 2.44 2.44 3.74E+11 Sheep 0.00 0.00 0.00E+00 Layers 0.00 0.00 0.00E+00 Hogs 0.00 0.00 0.00E+00 Chickens 0.00 0.00 0.00E+00 45425 Shrub./Pasture/Hay (m 2 ) 1773900 Count CFU/year Cattle 17.14 17.14 3.38E+13 Goats 0.64 0.64 3.52E+12 Horses 0.39 0.39 6.04E+10 Sheep 0.00 0.00 0.00E+00 Layers 0.00 0.00 0.00E+00 Hogs 0.00 0.00 0.00E+00 Chickens 0.00 0.00 0.00E+00 45418 Shrub./Pasture/Hay 24035400 Count CFU/year 323 (m 2 ) Cattle 868.25 757.95 1.49E+15 Goats 4.21 4.21 2.31E+13 Horses 14.57 14.57 2.23E+12 Sheep 1.49 1.49 8.18E+12 Layers 1.33 1.33 6.11E+10 Hogs 0.00 0.00 0.00E+00 Chickens 0.00 0.00 0.00E+00 45423 Shrub./Pasture/Hay (m 2 ) 13000500 Count CFU/year Cattle 469.63 409.97 8.08E+14 Goats 2.28 2.28 1.25E+13 Horses 7.88 7.88 1.21E+12 Sheep 0.81 0.81 4.43E+12 Layers 0.72 0.72 3.30E+10 Hogs 0.00 0.00 0.00E+00 Chickens 0.00 0.00 0.00E+00 45406 Shrub./Pasture/Hay (m 2 ) 11534400 Count CFU/year Cattle 416.67 363.73 7.17E+14 Goats 2.02 2.02 1.11E+13 Horses 6.99 6.99 1.07E+12 Sheep 0.72 0.72 3.93E+12 Layers 0.64 0.64 2.93E+10 Hogs 0.00 0.00 0.00E+00 Chickens 0.00 0.00 0.00E+00 45412 324 Shrub./Pasture/Hay (m 2 ) 793800 Count CFU/year Cattle 28.68 25.03 4.93E+13 Goats 0.14 0.14 7.61E+11 Horses 0.48 0.48 7.38E+10 Sheep 0.05 0.05 2.70E+11 Layers 0.04 0.04 2.02E+09 Hogs 0.00 0.00 0.00E+00 Chickens 0.00 0.00 0.00E+00 45410 Shrub./Pasture/Hay (m 2 ) 8829900 Count CFU/year Cattle 318.97 278.45 5.49E+14 Goats 1.55 1.55 8.47E+12 Horses 5.35 5.35 8.21E+11 Sheep 0.55 0.55 3.01E+12 Layers 0.49 0.49 2.24E+10 Hogs 0.00 0.00 0.00E+00 Chickens 0.00 0.00 0.00E+00 325 Table 5A.2 Annual Livestock Bacterial Loading per Watershed Watershed (JunctionID) CFU/year 45422 2.89E+16 45408 2.95E+16 45426 6.30E+15 45414 6.74E+15 45416 9.81E+15 45405 4.68E+15 45421 8.38E+15 45417 4.34E+16 45404 1.17E+16 45409 1.33E+15 45415 8.57E+15 45419 6.17E+16 45413 6.64E+15 56830 6.43E+15 56831 2.31E+14 45425 3.74E+13 45418 1.53E+15 45423 8.26E+14 45406 7.33E+14 45412 5.04E+13 45410 5.61E+14 45419 45417 45408 45422 45416 45426 45404 45405 56830 45415 45421 45413 45414 45410 45418 56831 45423 45409 45406 45412 45425 Figure 5A.1 Watershed JunctionIDs 326 Appendix 5.5: Mean Flow Length for Watersheds In using Equation 5.7, the mean flow lengths for watersheds are needed. The mean flow lengths for watersheds were calculated as follows: Creation of Flow Length Raster 1. Create Fdr (Flow Direction Raster that was created during Terrain Preprocessing, Appendix 5.1, with the Digital Elevation Model, DEM.) 2. In Arc Toolbox, go to the Spatial Analyst Tools, and open the tool “Flow Length”. 3. Select the Fdr raster as the “Input flow direction raster”. 4. Choose the name and directory for which the raster is to be placed. 5. Set the “Direction of measurement” to DOWNSTREAM, and press “OK”, and the flow length from each grid cell to the outlet in the Copano Bay watershed will be calculated. Determination of Mean Flow Length in Watersheds 1. Go to “Zonal Statistics as Table”. 2. Select the delineated watersheds (Figure 5.1) as the “Input raster or feature zone data”. 3. Set the “Zone field” to JunctionID (the identifier for each watershed). 4. Set the “Input value raster” to the flow length raster that has already been created. 5. Choose the name and directory for which the table is to be placed. 6. Join table (Flow Length Statistics Table) to the watershed feature class. 7. Go to CRWR Attribute Tools in Arc Toolbox and use the tool: “Copy Field to Feature Class from Table” that was created by Nate Johnson (2004). 8. Join based on JunctionID (field in Watershed feature class) and VALUE_ (field in Flow Length Statistics Table that correlates with JunctionID) and add the field, MEAN, from the Statistics Table, which will give the mean of the flow length values within each delineated watershed. (Note: this is the mean flow length from the watershed to the outlet of the Copano Bay watershed.) 9. In order to calculate the mean flow length of each watershed (from the watershed to the drainage outlet of the watershed), the flow length (from the watershed to the outlet) was calculated by: {Mean flow length of the watershed} – {flow length at the drainage junction determined from FlowLength raster}. 10. Open the attribute table of the delineated watershed feature class. 11. Go to Options… | Add Field i. Name: “FlowLength”, Type: “Double”. 327 12. Go to Editor | Start Editing…, and edit within the Personal Geodatabase that contains the delineated watersheds (Figure 5.1). 13. Open the attribute table of delineated watershed feature class, and manually input the flow lengths for each of the watersheds as explained in the statement in step #9. 14. Go to Editor | Stop Editing, and save edits. 328 Appendix 5.6: Mean Flow Length from WWTPs to Mainstreams Because the WWTPs are located at various distances from the mainstreams that are modeled for the Copano Bay watershed, the residence times from each WWTP to the downstream main river were calculated. Creation of Flow Direction Raster (Fdr) with Copano Bay and Mainstreams Omitted 1. Create Fdr (Flow Direction Raster that was created during Terrain Preprocessing, Appendix 5.1, with the Digital Elevation Model, DEM.) 2. Create a 35-meter buffer around mainstream polyline feature class. a. Go to Analysis Tools | Proximity | Buffer to access the buffer tool. b. Under “Input Features”, select the mainstream feature class. c. Create a 35-meter buffer around the mainstream polyline feature class (under “Distance [value or field]”). d. Select OK, and a mainstream polygon feature class will be created. 3. Create polygon feature class of subbasin with Copano Bay and mainstreams omitted. a. Use “Union” tool under Analysis tools to combine the feature classes: Copano Bay, subbasin, and the mainstream channels. b. Using the Editor Toolbar, delete Copano Bay and the mainstream channels from the created feature class, and save edits. 4. Go to Spatial Analyst | Options…, and set the “Analysis Mask” to the feature class that was created in step 3. 5. Set the “Extent” and “Cell Size” to the Fdr raster. 6. Go to Spatial Analyst | Raster Calculator… 7. Double-click on the Fdr raster, and “Evaluate”. 8. Right-click on the Calculation raster and Make Permanent. Creation of Flow Length Raster with Copano Bay and Mainstreams Omitted 1. In Arc Toolbox, go to the Spatial Analyst Tools, and open the tool “Flow Length”. 2. Select the Fdr raster (with Copano Bay and mainstreams omitted) as the “Input flow direction raster”. 3. Choose the name and directory for which the raster is to be placed. 4. Set the “Direction of measurement” to DOWNSTREAM, and press “OK”. 329 5. The flow length from each grid cell to either a mainstream or Copano Bay is then calculated and a flow length raster is created. 6. By using the identifier tool on the flow length raster that was just created, the distance (in meters) from the WWTP to either Copano Bay or a mainstream can be determined. # # # # # # # # # Legend # Outfalls_WWTPs NHDFlowline_MainStreams FlowLength_to_MainStreamsBay Value High : 55174 Low : 0.000000 Figure 5A.2 Flow Length Raster to Mainstream and Copano Bay 330 Appendix 5.7: Septic System Loading Calculations and Results The septic system loading calculations (how the bacterial loadings were calculated for each watershed) are shown in Table 5A.3, and the annual bacterial loadings per watershed due to septic systems are given in Table 5A.4. Table 5A.3 includes the area of each county, the area of each county classified as residential, and the area within each hydrologic soil group (within watershed and low/high residential areas), and the census data for each county. The septic system count (of each hydrologic soil group), human count, relative complaint count, number of housing units, and bacterial loadings for each watershed are also shown in Table 5A.3. The locations of the Watershed JunctionIDs (used in Tables 5A.3 and 5A.4) are shown in Figure 5A.1 in Appendix 5.4. Table 5A.3 Septic System Loading Calculations and Results County Bee San Patricio Aransas Refugio Goliad Karnes 7.3E+11 CFU/year (Humans) Area (m 2 ) 2344047042 1798057954 742112132 2016188169 2286808461 1965399713 Low/High Res. (m 2 ) 13999500 28288800 14113800 7603200 2357100 7070400 Population, 2004 Estimate 33046 68187 24041 7640 7104 15458 Projected Housing Units, 2004 11182 26237 13653 3647 3588 5591 Septic Systems in use, 2004 3767 6287 5981 991 2243 1724 Complaints, 2004 68 85 81 7 5 3 Humans/m 2 0.00236 0.00241 0.00170 0.00100 0.00301 0.00219 Housing units/m 2 0.00080 0.00093 0.00097 0.00048 0.00152 0.00079 Septic Systems/m 2 0.00027 0.00022 0.00042 0.00013 0.00095 0.00024 Complaints/m 2 0.00000 0.00000 0.00001 0.00000 0.00000 0.00000 WATERSHEDS (JunctionID) 331 45422 Low/High Res. (m 2 ) 1391400 0 Count CFU/year Soil Group A (m 2 ) 0 00 Septic Systems in Soil Group A Soil Group B (m 2 ) 85500 0 23 Septic Systems in Soil Group B Soil Group C (m 2 ) 1116000 030 Septic Systems in Soil Group C Soil Group D (m 2 ) 189900 0 51 Septic Systems in Soil Group D Humans 3284 0 3284 2.40E+15 Humans Housing Units 1111 0 1111 Housing Units Septic Systems 374 0374 Septic Systems Complaints 7 07 Complaints 45408 Low/High Res. (m 2 ) 573300 130500 0 Count CFU/year Soil Group A (m 2 ) 0 0 0 0 Septic Systems in Soil Group A Soil Group B (m 2 ) 0 0 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 0 0 Septic 332 Systems in Soil Group C Soil Group D (m 2 ) 573300 130500 0 183 Septic Systems in Soil Group D Humans 1353 315 0 1668 1.22E+15 Humans Housing Units 458 121 0 579 Housing Units Septic Systems 154 29 0 183 Septic Systems Complaints 3 0 0 3 Complaints 45426 Low/High Res. (m 2 ) 2122200 Count CFU/year Soil Group A (m 2 ) 0 0 Septic Systems in Soil Group A Soil Group B (m 2 ) 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 2122200 472 Septic Systems in Soil Group D Humans 5115 5115 3.73E+15 Humans Housing Units 1968 1968 Housing Units Septic Systems 472 472 Septic Systems 333 Complaints 6 6 Complaints 45414 Low/High Res. (m 2 ) 0 0 Count CFU/year Soil Group A (m 2 ) 0 0 0 Septic Systems in Soil Group A Soil Group B (m 2 ) 0 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 0 0 0 Septic Systems in Soil Group D Humans 0 0 0 0.00E+00 Humans Housing Units 0 0 Housing Units Septic Systems 0 0 0 Septic Systems Complaints 0 0 Complaints 45416 Low/High Res. (m 2 ) 0 5538600 0 0 Count CFU/year Soil Group A (m 2 ) 0 00 0 0 Septic Systems in Soil Group A Soil Group B (m 2 ) 0 00 0 0 Septic Systems in Soil Group B 334 Soil Group C (m 2 ) 0 00 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 0 5538600 0 0 1231 Septic Systems in Soil Group D Humans 0 13350 0 0 13350 9.75E+15 Humans Housing Units 0 5137 0 0 5137 Housing Units Septic Systems 0 1231 0 0 1231 Septic Systems Complaints 0 17 0 0 17 Complaints 45405 Low/High Res. (m 2 ) 1927800 4314600 Count CFU/year Soil Group A (m 2 ) 834300 3204000 1543 Septic Systems in Soil Group A Soil Group B (m 2 ) 00 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 00 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 1087200 1074600 697 Septic Systems in Soil Group D Humans 4647 7349 11996 8.76E+15 Humans Housing Units 1788 4174 5962 Housing Units 335 Septic Systems 428 1828 2257 Septic Systems (total) Complaints 625 31 Complaints 45421 Low/High Res. (m 2 ) 0 0 Count CFU/year Soil Group A (m 2 ) 0 0 0 Septic Systems in Soil Group A Soil Group B (m 2 ) 0 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 0 0 0 Septic Systems in Soil Group D Humans 0 0 0 0.00E+00 Humans Housing Units 0 0 0 Housing Units Septic Systems 0 0 0 Septic Systems Complaints 0 0 0 Complaints 45417 Low/High Res. (m 2 ) 0 2247300 0 Count CFU/year Soil Group A (m 2 ) 0 0 0 0 Septic Systems in Soil Group A Soil Group B (m 2 ) 0 0 0 0 Septic 336 Systems in Soil Group B Soil Group C (m 2 ) 0 0 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 0 2247300 0 293 Septic Systems in Soil Group D Humans 0 2258 0 2258 1.65E+15 Humans Housing Units 0 1078 0 1078 Housing Units Septic Systems 0 293 0 293 Septic Systems Complaints 0 2 0 2 Complaints 45404 Low/High Res. (m 2 ) 3600 Count CFU/year Soil Group A (m 2 ) 0 0 Septic Systems in Soil Group A Soil Group B (m 2 ) 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 3600 1 Septic Systems in Soil Group D Humans 8 8 6.20E+12 Humans 337 Housing Units 3 3 Housing Units Septic Systems 1 1 Septic Systems Complaints 0 0 Complaints 45409 Low/High Res. (m 2 ) 709200 Count CFU/year Soil Group A (m 2 ) 0 0 Septic Systems in Soil Group A Soil Group B (m 2 ) 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 709200 191 Septic Systems in Soil Group D Humans 1674 1674 1.22E+15 Humans Housing Units 566 566 Housing Units Septic Systems 191 191 Septic Systems Complaints 3 3 Complaints 45415 Low/High Res. (m 2 ) 4727700 Count CFU/year Soil Group A (m 2 ) 0 0 Septic Systems in Soil Group A 338 Soil Group B (m 2 ) 4588200 1234 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 139500 38 Septic Systems in Soil Group D Humans 11160 11160 8.15E+15 Humans Housing Units 3776 3776 Housing Units Septic Systems 1272 1272 Septic Systems Complaints 23 23 Complaints 45419 Low/High Res. (m 2 ) 26100 3600 262800 0 Count CFU/year Soil Group A (m 2 ) 0 0 0 00 Septic Systems in Soil Group A Soil Group B (m 2 ) 26100 0 259200 0254 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 0 00 Septic Systems in Soil Group C Soil Group D (m 2 ) 0 3600 3600 04 Septic Systems in Soil Group D 339 Humans 62 4 792 0 857 6.26E+14 Humans Housing Units 21 2 400 0423 Housing Units Septic Systems 7 0 250 0258 Septic Systems Complaints 0 0 1 01 Complaints 45413 Low/High Res. (m 2 ) 6654600 Count CFU/year Soil Group A (m 2 ) 0 0 Septic Systems in Soil Group A Soil Group B (m 2 ) 6641100 1787 Septic Systems in Soil Group B Soil Group C (m 2 ) 13500 4 Septic Systems in Soil Group C Soil Group D (m 2 ) 0 0 Septic Systems in Soil Group D Humans 15708 15708 1.15E+16 Humans Housing Units 5315 5315 Housing Units Septic Systems 1790 1790 Septic Systems Complaints 33 33 Complaints 56830 Low/High Res. (m 2 ) 0 0 Count CFU/year Soil Group A (m 2 ) 0 0 0 Septic Systems in Soil Group 340 A Soil Group B (m 2 ) 0 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 0 0 0 Septic Systems in Soil Group D Humans 0 0 0 0.00E+00 Humans Housing Units 0 0 0 Housing Units Septic Systems 0 0 0 Septic Systems Complaints 0 0 0 Complaints 56831 Low/High Res. (m 2 ) 1944000 Count CFU/year Soil Group A (m 2 ) 988200 419 Septic Systems in Soil Group A Soil Group B (m 2 ) 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 936000 397 Septic Systems in Soil Group 341 D Humans 3311 3311 2.42E+15 Humans Housing Units 1881 1881 Housing Units Septic Systems 824 824 Septic Systems Complaints 11 11 Complaints 45425 Low/High Res. (m 2 ) 0 Count CFU/year Soil Group A (m 2 ) 0 0 Septic Systems in Soil Group A Soil Group B (m 2 ) 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 0 0 Septic Systems in Soil Group D Humans 0 0 0.00E+00 Humans Housing Units 0 Housing Units Septic Systems 0 0 Septic Systems Complaints 0 Complaints 45418 Low/High Res. (m 2 ) 1719900 Count CFU/year Soil Group A (m 2 ) 0 0 Septic Systems in 342 Soil Group A Soil Group B (m 2 ) 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 1719900 224 Septic Systems in Soil Group D Humans 1728 1728 1.26E+15 Humans Housing Units 825 825 Housing Units Septic Systems 224 224 Septic Systems Complaints 2 2 Complaints 45423 Low/High Res. (m 2 ) 1883700 Count CFU/year Soil Group A (m 2 ) 0 0 Septic Systems in Soil Group A Soil Group B (m 2 ) 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 1883700 246 Septic Systems in 343 Soil Group D Humans 1893 1893 1.38E+15 Humans Housing Units 904 904 Housing Units Septic Systems 246 246 Septic Systems Complaints 2 2 Complaints 45406 Low/High Res. (m 2 ) 0 Count CFU/year Soil Group A (m 2 ) 0 0 Septic Systems in Soil Group A Soil Group B (m 2 ) 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 0 0 Septic Systems in Soil Group D Humans 0 0 0.00E+00 Humans Housing Units 0 0 Housing Units Septic Systems 0 0 Septic Systems Complaints 0 0 Complaints 45412 Low/High Res. (m 2 ) 0 Count CFU/year Soil Group A (m 2 ) 0 0 Septic 344 Systems in Soil Group A Soil Group B (m 2 ) 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 0 0 Septic Systems in Soil Group D Humans 0 0 0.00E+00 Humans Housing Units 0 0 Housing Units Septic Systems 0 0 Septic Systems Complaints 0 0 Complaints 45410 Low/High Res. (m 2 ) 630000 Count CFU/year Soil Group A (m 2 ) 0 0 Septic Systems in Soil Group A Soil Group B (m 2 ) 0 0 Septic Systems in Soil Group B Soil Group C (m 2 ) 0 0 Septic Systems in Soil Group C Soil Group D (m 2 ) 630000 82 Septic 345 Systems in Soil Group D Humans 633 633 4.62E+14 Humans Housing Units 302 302 Housing Units Septic Systems 82 82 Septic Systems Complaints 1 1 Complaints Table 5A.4 Annual Septic System Bacterial Loading per Watershed Watershed (JunctionID) CFU/year 45422 4.29E+14 45408 1.93E+14 45426 4.47E+14 45414 0.00E+00 45416 1.17E+15 45405 2.78E+15 45421 0.00E+00 45417 0.00E+00 45404 0.00E+00 45409 2.06E+14 45415 2.70E+15 45419 0.00E+00 45413 3.86E+15 56830 0.00E+00 56831 7.93E+14 45425 0.00E+00 45418 0.00E+00 346 45423 8.26E+14 45406 7.33E+14 45412 5.04E+13 45410 5.61E+14 347 Appendix 5.8: Determination of Soil Group Areas within each Watershed and Land Use Classifications 21 and 22 To calculate the number of septic systems and the number of complaints investigated for each hydrologic soil group classification within each watershed in the land use classifications 21 and 22, the soil group areas within each watershed and land use classifications 21 and 22 were determined by the procedure given below. To find the total count of septic systems and complaints investigated, the area of each soil group was multiplied by each corresponding county’s density. Join Comp.dbf to Soil Polygon Feature Class (STATSGO data) 1. Right-click on soil polygon feature class, and go to Joins and Relates | Join… 2. Base the join on the field “MUID”, which is found in both the soil polygon feature class and Comp.dbf. Create Feature Classes of Different Hydrologic Soil Groups (A, B, C, D) 1. Right-click on the soil polygon feature class (in Arc Map), go to Properties | Definition Query. 2. Select “Query Builder…” 3. Double-click on [COMP.HYDGRP] (the field that contains the hydrologic soil groups), “=” and select one of the soil groups (either A, B, C, or D.) 4. After conducting a query for one of the soil groups, select all the polygons in the Arc Map view. 5. Right-click on the soil polygon feature class, and Data | Export Data… and create a new feature class for that specific soil group. 6. Repeat steps #3-5 until there are four new feature classes (Soil Group A, Soil Group B, Soil Group C, and Soil Group D.) Create Land Use Land Cover Rasters for Each Hydrologic Soil Group 1. Go to Spatial Analyst | Options… 2. Set the “Analysis mask” to one of the soil group feature classes (A, B, C, or D). 3. Set the “Extent” and “Cell Size” to the land use land cover raster. 4. Go to Spatial Analyst | Raster Calculator… 5. Double-click on the land use land cover raster and “Evaluate”. 6. Right-click on the Calculation raster and Make Permanent. 7. Repeat steps #1-6 for the other three soil group classifications. (There will now be land use land cover rasters for each of the different soil groups.) 348 Create Land Use Land Cover Rasters for each Soil Group in each Watershed 1. Go to Spatial Analyst | Options… 2. Set the “Analysis mask” to one of the watershed feature classes (each watershed needs to be exported into its own feature class). 3. Set the “Extent” and “Cell Size” to the land use land cover raster. 4. Go to Spatial Analyst | Raster Calculator… 5. Double-click on the land use land cover raster of a soil group found within that specified watershed (in step #2) and “Evaluate”. 6. Right-click on the Calculation raster and Make Permanent. 7. Repeat steps #5-6 until all the soil groups found within the watershed (specified in step #2) have been created into land use land cover rasters within the watershed. 8. Repeat steps #2-7 until the land use land cover rasters have been divided up into all the watersheds in the Copano Bay watershed. Create Land Use Land Cover Rasters for each Soil Group in each Watershed within each County 1. Go to Spatial Analyst | Options… 2. Set the “Analysis mask” to one of the counties that is overlapping the watershed of interest (each county needs to be exported into its own feature class). 3. Set the “Extent” and “Cell Size” to the land use land cover raster. 4. Go to Spatial Analyst | Raster Calculator… 5. Double-click on the land use land cover raster that was created for a specified watershed in a specified soil group and “Evaluate”. 6. Right-click on the Calculation raster and Make Permanent. 7. Repeat steps #5-6 until all the soil groups found within the watershed (specified in step #2) have been created into land use land cover rasters within the watershed and overlapping county. 8. Repeat steps #2-7 until the land use land cover rasters (of soil groups) have been divided up into all the watersheds and counties in the Copano Bay watershed. Calculate the Area of Low/High Residential within each Soil Group within each Watershed and each County. 1. Add all the land use land cover rasters for a particular watershed (the land use land cover rasters for each soil group and county within the specified watershed) to Arc Map or preview each raster’s table in ArcCatalog. 2. Right-click on each raster, and open up the attribute table. The field, “Value”, is the land use land cover classification, and “Count” is the number of grid cells that correspond to each land use land cover classification. 3. Sum the grid cell count for the land use land cover classifications 21 and 22 (low/high residential) and multiply by 900 m 2 (the area of one grid cell) to find the total area corresponding to low/high residential for each soil group and county within each watershed. 349 Appendix 6.1: Schematic Network To run the “Process Schematic”, a Schematic Network of the Copano Bay watershed was created. The Schematic Network consists of two feature classes: SchemaNode and SchemaLink. Creation of Automated Schematic Network 1. Go to Arc Hydro Toolbar, go to Network Tools | Node/Link Schema Generation. 2. Set the Watershed Polygons as the delineated watersheds, and the Junctions as BatchPoint (the feature class that contains the critical points: USGS gauge stations, bacteria monitoring stations, and water segment endpoints). (Note: the feature classes SchemaLink and SchemaNode will be automatically created.) Modify Automated Schematic Network Because of the complex network (due to Copano Bay), the SchemaLink and SchemaNode attributes will need to be manually modified. SchemaNode Modifications 1. Go to Editor | Start Editing…, and edit within the Personal Geodatabase that contains the SchemaLink and SchemaNode feature classes. 2. Set the Task as “Create New Feature” and the Target layer as “SchemaNode”. 3. Add junctions in the middle of each of the watersheds, drainage points (BatchPoint), and four junctions in Copano Bay; (Copano Bay was divided into four segments.) 350 4. Open the attribute table of the SchemaNode feature class, and set the SrcType for each of the junctions (1 = Watershed, 2 = Junction watershed drains to, 3 = Copano Bay.) 5. Open the attribute table of the SchemaNode feature class, and set the FeatureID for each of the watershed junctions (SrcType = 1) to the JunctionID of the corresponding watershed. 6. Go to Editor | Stop Editing, and save edits. SchemaLink Modifications 1. Go to Editor | Start Editing…, and edit within the Personal Geodatabase that contains the SchemaNode and SchemaLink feature classes. 2. Set the Task as “Create New Feature” and the Target layer as “SchemaLink”. 3. Add links, so that all the SchemaNode feature classes are connected by a SchemaLink. (Go to Editor | Snapping…, and select the Vertex, Edge, and End boxes of the SchemaNode feature class, so the endpoints of SchemaLink will snap to the SchemaNodes.) 4. Open the attribute table of the SchemaLink feature class, and set the LinkType for each of the links (1 = Connects watershed to drainage junction, 2 = Connects drainage junction to drainage junction, 3 = Connects drainage junction to Copano Bay.) 5. Go to Editor | Stop Editing, and save edits. 6. Go to the Arc Hydro toolbar, and go to Attribute Tools | Assign HydroID, and assign HydroIDs for the SchemaNode and SchemaLink feature classes. 7. Start the Editor, and edit the SchemaLink feature class. 8. Open the SchemaLink attribute table, and populate the fields FromNodeID and ToNodeID with the corresponding HydroIDs of the upstream and downstream SchemaNodes. 9. Go to Editor | Stop Editing, and save edits. Schematic Network Parameters Before the Schematic Network is complete, fields need to be added to the attribute tables of the SchemaNode and SchemaLink feature classes. SchemaLink 1. Open the attribute table of the SchemaLink feature class, and go to Options | Add Field… 2. Repeat step #1 until all the following fields are added: a. Name: “DecayConst_day”, Type: “Double” b. Name: “TravelTime_day”, Type: “Double” c. Name: “TotVal”, Type: “Double” d. Name: “PassedVal”, Type: “Double” e. Name: “IncVal”, Type: “Double” 351 3. Using the Editor Toolbar, the following fields need to be populated for LinkTypes 1 and 2: “DecayConst_day” and “TravelTime_day”; see Section 6.3.3 to see how the parameters were determined for this project. SchemaNode 1. Open the attribute table of the SchemaNode feature class, and go to Options | Add Field… 2. Repeat step #1 until all the following fields are added: a. Name: “DecayCoef_day”, Type: “Double” b. Name: “FLOW_m3_yr”, Type: “Double” c. Name: “Volume”, Type: “Double” d. Name: “TotVal”, Type: “Double” e. Name: “PassedVal”, Type: “Double” f. Name: “IncVal”, Type: “Double” 3. Using the Editor Toolbar, the following fields need to be populated for SrcType 1: a. “IncVal” – The values that should populate this field are the total bacterial loadings (or the loadings that effects of transport are desired) in cfu/year per watershed. (“IncVal” = cfu/year of corresponding SchemaNode in Schematic Network.) 4. Using the Editor Toolbar, the following field needs to be populated for SrcType 2: a. “IncVal” – The values that should populate this field are the total bacterial loadings (or loadings that effects of transport are desired) in cfu/year for each node. 4. Using the Editor Toolbar, the following fields need to be populated for SrcType 3: a. “DecayCoef_day” – see Section 6.3.3.1 to see how this parameter was determined for this project. b. “Volume” – see Section 6.3.3.3 to see how this parameter was determined for this project. c. “FLOW_m3_year” – see Section 6.3.3.4 to see how this parameter was determined for this project. 352 Appendix 6.2: Travel Time Calculations The travel times for Link types 1 and 2 were determined as input to the Schematic Processor. The following equation was used to calculate the initial travel time (before calibration) for the segments that were not made into 3d river models: Travel Time = Flow Length / Velocity Flow Length Calculations 1. Create Fdr (Flow Direction Raster that was created during Terrain Preprocessing, Appendix 5.1, with the Digital Elevation Model, DEM) that does not include Copano Bay. 2. Create polygon feature class of subbasin with Copano Bay omitted. a. Use “Union” tool under Analysis tools to combine the feature classes: Copano Bay and the subbasin. b. Using the Editor Toolbar, delete Copano Bay from the created feature class, and save edits. 3. Go to Spatial Analyst | Options…, and set the “Analysis Mask” to the feature class that was created in step a. 4. Set the “Extent” and “Cell Size” to the Fdr raster. 5. Go to Spatial Analyst | Raster Calculator… 6. Double-click on the Fdr raster, and “Evaluate”. 7. Right-click on the Calculation raster and Make Permanent. 8. Create flow length raster. 9. In Arc Toolbox, go to the Spatial Analyst Tools, and open the tool “Flow Length”. 10. Select the Fdr raster (with Copano Bay omitted) as the “Input flow direction raster”. 11. Choose the name and directory for which the raster is to be placed. 12. Set the “Direction of measurement” to DOWNSTREAM, and press “OK”. 13. The flow length from each grid cell to Copano Bay is then calculated. 14. Determine mean flow length in each delineated watershed. 15. Go to “Zonal Statistics as Table”. 16. Select the delineated watersheds as the “Input raster or feature zone data”. 17. Set the “Zone field” to JunctionID (the identifier for each watershed). 18. Set the “Input value raster” to the flow length raster that was created in steps #8- 13. 19. Choose the name and directory for which the table is to be placed. 20. Join table that was created in steps #15-19 (Flow Length Statistics Table) to the watershed feature class. 21. Go to CRWR Attribute Tools in Arc Toolbox and use the tool: “Copy Field to Feature Class from Table” that was created by Nate Johnson (2004). 353 22. Join based on JunctionID (field in Watershed feature class) and VALUE_ (field in Flow Length Statistics Table that correlates with JunctionID) and add the field, MEAN, from the Statistics Table, which will give the mean of the flow length values within each delineated watershed. 23. For SchemaLink (Link type 1), the flow length (from the watershed to the stream) was calculated by: {Mean flow length of the watershed} – {flow length at the drainage junction (SchemaNode Srctype 2) determined from FlowLength raster}. 24. Open the attribute table of the delineated watershed feature class. 25. Go to Options… | Add Field i. Name: “FlowLength”, Type: “Double”. 26. Go to Editor | Start Editing…, and edit within the Personal Geodatabase that contains the delineated watersheds. 27. Open the attribute table of delineated watershed feature class, and manually input the flow lengths for each of the links (LinkType 1) as explained in the statement in step #23. 28. Go to Editor | Stop Editing, and save edits. 29. For SchemaLink (Link type 2), the flow length along the streams was calculated by: {Flow length at upstream SchemaNode} – {Flow length at downstream SchemaNode}. 30. Go to Editor | Start Editing…, and edit within the Personal Geodatabase that contains the delineated watersheds. 31. Open the attribute table of delineated watershed feature class, and manually input the flow lengths for each of the links (LinkType 2) as explained in the statement in step #29. 32. Go to Editor | Stop Editing, and save edits. Velocity Calculations For preliminary work, velocities of the streams were determined using flow – velocity relationships similar to Zoun (2003). The relationship between flow and velocity was derived from the EPA Reach File 1 (RF1) database, which documents flow and velocity for the entire United States. A regression line was fitted to the flow and velocity data for the Copano Bay watershed area (same method Zoun used to derive flow and velocity relationship in Galveston Bay area). These following equations were used to calculate the velocity for each of the stream segments (SchemaLinks). V = 63.252 * Q 0.3132 Where: V = velocity in m/day Q = flow in m 3 /year 354 The cumulative runoff at each upstream and downstream SchemaNode were averaged for each SchemaLink, and then entered into the above equation. 1. Add Velocity field to SchemaLink feature class. 2. Go to Options… | Add Field. (Name: “Velocity”, Type: “Double”). 3. Go to Editor | Start Editing…, and edit within the Personal Geodatabase that contains the Schematic Network. 4. Open the attribute table of the SchemaLink feature class, and manually input the velocities (calculated from above equation) for each of the links as previously explained. 5. Go to Editor | Stop Editing, and save edits. Travel Time Calculations 1. Open the attribute table of the SchemaLink feature class. 2. Right-click on the field: “TravelTime_day” 3. Double-click on the flow length field, push the “ / ” button, and double-click on the velocity field in order to calculate: TravelTime_day = FlowLength/Velocity. 355 Appendix 6.3: Process Schematic Once the Schematic Network was created, the input feature classes, corresponding fields, and DLLs were set in the Process Schematic as shown below. This set-up tells the script (Process Schematic) to implement decay (clsdecay.dll) on SchemaLink LinkTypes 1 and 2, which will decay bacteria as they travel from the watersheds to the streams and during their travel time along the streams), and to run a CFSTR model (clsCFSTR.dll) on SchemaNode SrcType 3 (Copano Bay waters segments). SchemaLink feature class SchemaNode feature class 356 Note: Each row (shown below) corresponds to the same row of the following fields. For example, the Processing Op in the first row, WaterQualityProcessors.ClsDecay (clsdecay.dll), simulates decay of the bacteria loadings along LinkType 1 (Op Source Type) and passes the value (Op Behavior Type: Pass). DLLs being called. “1”, “2”, or “3” SrcType (for SchemaNode) or LinkType (for SchemaLink) for which corresponding Processing Op (DLL) is being run. 357 Run Process Schematic 1. Right-click on “Process Schematic”, and Run. 2. Results are found in the attribute tables of SchemaLink and SchemaNode (under the fields: PassedVal and TotVal). a. SchemaNode i. Src Type 2 1. The populated values in the "PassedVal" and "TotVal" fields are bacterial loadings (cfu/year). 2. These values can be converted to cfu/m 3 by dividing by the cumulative upstream runoff (m 3 /year), and this concentration can be converted to cfu/100mL by dividing by 10,000. ii. Src Type 3 1. The populated values in the "TotVal" field are bacterial concentrations in cfu/m 3 . 2. This value can be converted into bacterial loading (cfu/year) by multiplying by the cumulative upstream runoff (m 3 /year) to that SchemaNode. “Link” or “Node”. In this case, decay is being run along SchemaLinks (Link), and the CFSTR model is being run on SchemaNodes SrcType 3 (Node). “Pass” or “Receive”. In this case, the bacterial loading values are going to be passed along the SchemaLinks (Pass), but the bacterial loading values are going to be received by SchemaNodes SrcType 3, Copano Bay segments (Receive) 358 359 Appendix 7.1: Monte Carlo Simulation Model Worksheets This appendix gives explanations of all the worksheets that are used in the Monte Carlo Simulation Model that was created by Ernest To. The bold headings indicate the name of the worksheet, and the descriptions below explain the important features. The key parameters and how the parameters are used in modeling fecal coliform concentrations are also discussed in the appropriate sections. Control_Sheet This worksheet (Figure 7A.1) shows the Copano Bay watershed along with all the SchemaNodes and SchemaLinks (number identifiers of the nodes and links in the Schematic Network.) SchemaNodes with data are the SchemaNodes that have bacterial monitoring stations and monitoring data at the locations of the nodes, and thus, the bacterial monitoring data (from TCEQ 1999-2005) is what is plotted as “Existing” data when the model is run. User inputs is where the user identifies the location (SchemaNode) where he/she wants to model the bacteria concentrations (CFU/100 mL) and can compare to the existing monitoring data (1999-2005) if the node is at a bacterial monitoring station. The number of simulations can be specified as well. Currently, SchemaNode 61 (Bacterial Monitoring Station 17592) is modeled with 1000 simulations. To run the model, click on the “Monte Carlo Analysis” button. 360 Figure 7A.1 Control Sheet of Monte Carlo Simulation Model (User Interface) Monte_Carlo_Graph (Figure 7A.2) Once the model is run, this worksheet (shown in Figure 7A.2) will become the active sheet. It is the graph that compares the modeled bacteria concentrations to the existing monitoring data. The yellow box shows the median and 90 th -percentile fecal coliform concentrations generated by the model (from the input parameters and bacterial loadings) and the current standards at this location (and in this case, it is contact recreation use standards for Aransas River.) The pink box shows the median and 90 th - percentile fecal coliform concentrations from the existing monitoring data (1999-2005 from bacterial monitoring station 17592.) 361 Figure 7A.2 Existing versus Modeled Fecal Coliform Concentrations of Monte Carlo Simulation Model Monte_Carlo_Table This worksheet (shown in Figure 7A.3) shows all the simulations (in this case, 1000) that were run from the model and the fecal coliform concentrations calculated at the SchemaNode of interest for each simulation. 362 Figure 7A.3 Modeled Fecal Coliform Concentrations in Table Format of Monte Carlo Simulation Model AllMonitoringData This worksheet (shown in Figure 7A.4) shows all the monitoring data that exists for the SchemaNode of interest from 1999-2005, so as you can see, for bacterial monitoring station 17592, only ten fecal coliform measurements were made in six years. 363 Figure 7A.4 Existing Monitoring Data at SchemaNode of Interest Schemanode This is the SchemaNode feature class (all the nodes in the model network that can be seen in “Control_sheet”.) The full worksheet can be seen in Figure 7A.5, which is shown in three subsequent figures. These are descriptions of all the fields on this worksheet: • “ObjectID” is the value randomly assigned to each SchemaNode feature. • “HydroID” is the number identifier for each SchemaNode (the same HydroIDs are used in this Model as in the Schematic Processor Model and Schematic Network in Chapter 6.) • “SrcType” is the type of SchemaNode. SrcType = 1 represents a watershed; SrcType = 2 represents a drainage point of a watershed; SrcType = 3 represents a Copano Bay segment. 364 • “IncVal” is the incremental loading that is added to the model at that SchemaNode location. (However, this field is inactive in this model.) • “TotVal” is the incremental loading (“IncVal”) added to the upstream loading. (However, this field is inactive in this model.) • “FLOW_m3_yr” is the flow associated with each Copano Bay segment, the flowrate of the water draining to each Copano Bay segment. (These values are the same that were calculated in Section 6.3.3.4.) • “Volume” is the volume of each Copano Bay segment (calculated from bathymetry data and surface area.) (These values are the same that were calculated in Section 6.3.3.3.) • “DecayCoef_day” is the decay coefficient associated with each SchemaNode, which is 2 days -1 . • “PassedVal” is the calculated bacterial loading that is passed onto the downstream SchemaLink. (However, this field is inactive in this model.) • “CumRunoff_m3_yr” is the cumulative runoff (from upstream watersheds) to each specified SchemaNode location. This field is used to calculate the concentration (CFU/100mL) at each location and will be described later. • “Animal_cfu_year” is the calculated annual bacterial loading (CFU/year) excreted by livestock animals. Loading calculations are described in Section 5.2. • “NonPoint_cfu_year” is the calculated annual non-point bacterial loading (CFU/year) associated with different land use types (excluding agriculture, pasture, etc.) Loading calculations are described in Section 5.1. • “Birds_cfu_year” is the calculated annual bacterial loading (CFU/year) excreted by waterbirds. Loading calculations are described in Section 5.3. • “WWTP” is the calculated annual bacterial loading (CFU/year) generated by WWTPs, and “WWTP_2” is to account for the loading if two WWTPs discharge to the same SchemaNode. Loading calculations are described in Section 5.4. • “Human_Septic” is the calculated bacterial loading (CFU/year) that is produced from septic systems. Loading calculations are described in Section 5.5. • “Total (CFU/year)” is the bacteria loading that is calculated for each node based on a lognormal distribution that is created from the median bacterial loading (Column S: “Load_param1”, which is the sum of Columns L thru Q - all the calculated bacteria loadings) and an associated multiplication factor (Column T: “Load_param2”), which is one of the parameters that is adjusted such that 365 existing monitoring data can be matched. This multiplication factor is associated with the coefficient of variance. If you go to “multi_factor_vs_cov” worksheet, you can see the associated coefficient of variance with each multiplication factor. (Note: cov = standard deviation/mean.) Thus, the higher the multiplication factor, the wider the spread and range of the bacterial loadings. The Excel formula for this column is: Column R = Column S * EXP(Column T * NORMINV(ABS(RAND()),0,1)) • “Load_param1” is the sum of all the bacterial loadings and would represent the median total bacterial loading to each SchemaNode. The Excel formula for this column is: Column S = Column L + Column M + Column N + Column O + Column P + Column Q. • “Load_param2” is a multiplication factor that is used (along with “Load_param1”) to create a bacterial loading normal distribution. This multiplication factor is associated with the coefficient of variance, and the “multi_factor_vs_cov” worksheet shows the relationship. The remaining fields are what are modified to calculate the modeled fecal coliform concentrations. • “Total_value” is the upstream decayed bacterial loading plus the incremental bacterial loading added to the model at the specified node location. • “Conc/100mL”, for SrcType 2 nodes, = “Total_value” / (10,000 * “CumRunoff_m3_yr”) 366 Figure 7A.5 SchemaNode Fields for Monte Carlo Simulation Model 367 Figure 7A.5 (Continued) 368 Figure 7A.5 (Continued) Schemalink (Figure 7A.6) This is the SchemaLink feature class (all the links in the model network that can be seen in “Control_sheet”.) These are descriptions of all the fields on this worksheet: • “FromNodeID” is the HydroID of the upstream SchemaNode to the SchemaLink. • “ToNodeID” is the HydroID of the downstream SchemaNode to the SchemaLink. • “LinkType” is the type of SchemaLink it is. LinkType = 1 represents the transport of bacteria in a watershed (applies following equation: downstream loading = upstream loading * exp (-kt)); LinkType = 2 represents transport of bacteria along a river, which applies the same first-order decay equation as LinkType = 1. LinkType = 3 represents the transport of bacteria in a bay segment (applies the following equation: upstream loading/(cumulative runoff + k*Volume).) 369 • “kd_name” is just an identifier to the decay coefficient associated with each SchemaLink. • “DecayConst_day” is the decay coefficient that is calculated for each SchemaLink (and simulation) from a beta distribution (alpha = 2 and beta = 2) and from the upper and lower boundaries of 2 (“Kd_param1”, Column J) and 2.5 days -1 (“Kd_param2”, Column K). The Excel formula for this column is: BETAINV(ABS(RAND()),2,2, Column J, Column K) • “Kd_param1” is the lower boundary of the beta distribution associated with the decay distribution, which is 2 days -1 for this model. • “Kd_param2” is the upper boundary of the beta distribution associated with the decay distribution, which is 2.5 days -1 for this model. • “Tau_name” is an identifier to the travel time that is associated with each SchemaLink. • “TravelTime_day” is the travel time (residence time) associated with each SchemaLink. This parameter is adjusted to try to match up with the median fecal coliform concentrations. (The adjacent upstream SchemaLink – to the SchemaNode of interest - is usually the most influential and sensitive to the model.) The initial residence times (calculated in Section 6.3.3.2) were used before calibration, and then the appropriate residence times were adjusted. 370 Figure 7A.6 SchemaLink Fields for Monte Carlo Simulation Model 371 Appendix 7.2: Load Reduction Scenario #1 Results This appendix gives the load reduction results of all the water segments (Aransas and Mission River Tidals, Aransas and Mission River Above Tidals, and Copano Bay) for Load Reduction Scenario #1. Load Reduction Scenario #1 is the load reduction necessary to meet fecal coliform water quality standards for all water segments at each location in the model that was analyzed. The locations where the model was analyzed were the upstream and downstream portions of the Above Tidals and Tidals, the locations of the bacterial monitoring stations, and the Copano Bay water segments. However, each portion of the model that was analyzed (that did not meet fecal coliform water quality standards) was not always verified by existing monitoring data, so the results are inconclusive based on lack of monitoring data. Thus, these load reductions are only presented to show possible problem areas. Aransas River Above Tidal The Aransas River Above Tidal (shown in Figure 7A.7) must meet contact recreation use standards for fecal coliform, but the primary bacterial indicator is E. coli for this segment. However, the results presented are based on fecal coliform water quality standards because a fecal coliform model was created. For fecal coliform, the geometric mean of the samples must be less than 200 CFU/100mL, and single samples must be less than 400 CFU/100mL, but TCEQ allows 25% of the samples to exceed 400 CFU/100mL. The upstream portion of the Aransas River Above Tidal (SchemaNode 62), which is indicated in Figure 7A.7 with an orange circle, was observed in the calibrated Monte Carlo Simulation Model first, and this node is directly upstream of Station 12952. Without any load reductions in the upstream watersheds, two runs of 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted to investigate background variation of the model. The percentile at which 400 CFU/100mL 372 was reached in the model was recorded to ensure that the percentile is greater than 75%. However, the geometric mean of 1000 samples could not be calculated using the GEOMEAN() function in Microsoft Excel, so a separate run of 100 simulations was performed to obtain the geometric mean modeled at SchemaNode 62. For the two separate 1000 and 100 simulation runs, the geometric mean and percent of samples exceeding 400 CFU/100mL are shown in Table 7A.1. As shown in Table 7A.1 (0% load reduction), the model suggests that the upstream portion of the Aransas River Above Tidal is in compliance with fecal coliform water quality standards. In Run #1, the percent of samples exceeding 400 CFU/100mL is 20.6%, which is less than 25%; however, to increase the safety factor, the effect of load reductions at the upstream WWTP was tested 19 . The location of the WWTP is shown in Figure 7A.7. The median bacterial loading from this WWTP (City of Beeville Moore Street WWTP) was calculated based on two annual fecal coliform measurements and the average flow rate (of the monthly flow rates from 1998-2005) reported on the discharge monitoring reports (DMR); see Section 5.4. The number of runs of simulations and the modeled results at SchemaNode 62 with various load reductions at the WWTP are also shown in Table 7A.1. 19 The WWTP load reductions were based on the overestimated bacterial loadings from WWTPs (explained in Section 5.4.2). 373 Table 7A.1 Modeled Results at SchemaNode 62 with Various WWTP Load Reductions Run # Load Reduction (%) Geometric Mean (CFU/100mL) Percent > 400 CFU/100mL 1 0 95.3 20.6 2 0 104.3 18.5 1 10 76.1 20.8 2 10 106.4 17.9 1 20 91.9 17.8 2 20 90.1 15.9 1 30 113.1 16.5 2 30 100.3 15.2 Simulations 100 1000 As shown in Table 7A.1, reducing the bacterial loadings from the upstream WWTP by 30% allows approximately 15% of the samples to exceed the 400 CFU/100mL standard. Modifying the disinfection process could reduce the fecal coliform load from the WWTP. With a 30% reduction in bacterial load, the geometric mean should have been less than the geometric mean with a 20% reduction in load. However, this was not observed, and, in fact, Table 7A.1 shows that the geometric mean varies greatly throughout the runs. This variation can be explained by the inherent variation of the Monte Carlo analysis; this is a plausible explanation because the WWTP loadings are significantly less than loadings from non-point sources and therefore should not have a major impact on the overall fecal coliform concentration. However, to be conservative, the 30% load reduction from the WWTP was applied for the remainder of the load reduction calculations. Recall that the WWTP loadings were largely overestimated in these calculations (Section 5.4.2). 374 The next downstream SchemaNode (indicated in Figure 7A.8 by an orange circle) along the Aransas River Above Tidal that is compared to standards in the model is SchemaNode 68, which is also the location of Station 12952. With only the 30% load reduction applied at the upstream WWTP, two runs each of 100 and 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted. For the two separate 1000 and 100 simulation runs, the geometric mean and percent of samples exceeding 400 CFU/100mL are shown in Table 7A.2. 30% Figure 7A.7 Load Reductions for SchemaNode 62: Aransas River Above Tidal 375 Table 7A.2 Modeled Results at SchemaNode 68 with No Additional Load Reductions Run # Load Reduction (%) Geometric Mean (CFU/100mL) Percent > 400 CFU/100mL 1 0 60.2 8.7 2 0 71.7 9.1 Simulations 100 1000 The upstream portion of the Aransas River Above Tidal is well in compliance with fecal coliform water quality standards based on modeled results (shown in Table 7A.2.) 30% Figure 7A.8 Load Reductions for SchemaNode 68: Aransas River Above Tidal 376 The next downstream SchemaNode (indicated in Figure 7A.9 by an orange circle) along the Aransas River Above Tidal that is compared to standards in the model is SchemaNode 75, which is also the location of Station 12948. With only the 30% load reduction applied at the upstream WWTP, two runs each of 100 and 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted. For the two separate 1000 and 100 simulation runs, the geometric mean and percent of samples exceeding 400 CFU/100mL are shown in Table 7A.3. As shown in Table 7A.3 (0% load reduction), the model suggests that the downstream portion of the Aransas River Above Tidal is in compliance with fecal coliform water quality standards. However, in Run #2, the percent of samples exceeding 400 CFU/100mL is 18.2%, which is less than 25%; however, to increase the safety factor, the effect of load reductions at the upstream WWTP was tested. The location of the WWTP is shown in Figure 7A.9. The median bacterial loading from this WWTP (Chase Field WWTP) was calculated based on literature values because no fecal coliform measurements were reported on the DMR; see Section 5.4. The number of runs of simulations and the modeled results at SchemaNode 75 with various load reductions at the WWTP are also shown in Table 7A.3. Table 7A.3 Modeled Results at SchemaNode 75 with Various Additional WWTP Load Reductions Run # Load Reduction (%) Geometric Mean (CFU/100mL) Percent > 400 CFU/100mL 1 0 81.3 17.2 2 0 97.1 18.2 1 10 93.2 17.8 2 10 122.7 15.1 1 30 97.1 15 2 30 114.0 16.5 1 45 81.3 16.2 2 45 92.3 14.1 Simulations 100 1000 377 Reducing the bacterial loadings from the upstream WWTP by 30-45% allows approximately 15% of the samples to exceed the 400 CFU/100mL standard (shown in Table 7A.3.) The variation in the runs could just be the inherent variation of the Monte Carlo analysis; this is a plausible explanation because the WWTP loadings are significantly less than loadings from non-point sources and therefore should not have a major impact on the overall fecal coliform concentration. However, to be conservative, the 45% load reduction from the WWTP was applied for the remainder of the load reduction calculations. Aransas River Tidal The Aransas River Tidal (shown in Figure 7A.10) must meet contact recreation use standards for fecal coliform, but the primary bacterial indicator is enterococci for this 30% 45% Figure 7A.9 Load Reductions for SchemaNode 75: Aransas River Above Tidal 378 segment. However, the results presented are based on fecal coliform water quality standards because a fecal coliform model was created. For fecal coliform, the geometric mean of the samples must be less than 200 CFU/100mL, and single samples must be less than 400 CFU/100mL, but TCEQ allows 25% of the samples to exceed 400 CFU/100mL. The upstream portion of the Aransas River Tidal (SchemaNode 75) was analyzed as the downstream node to Aransas River Above Tidal, so will not be analyzed again in this section. The next downstream SchemaNode (indicated in Figure 7A.10 by an orange circle) along the Aransas River Tidal that is compared to standards in the model is SchemaNode 63. With only the 30% and 45% load reductions applied at the upstream WWTPs, two runs each of 100 and 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted. For the two separate 1000 and 100 simulation runs, the geometric mean and percent of samples exceeding 400 CFU/100mL is shown in Table 7A.4. Table 7A.4 Modeled Results at SchemaNode 63 with No Additional Load Reductions Run # Load Reduction (%) Geometric Mean (CFU/100mL) Percent > 400 CFU/100mL 1 0 74.2 4.4 2 0 56.6 4.9 Simulations 100 1000 The middle portion of the Aransas River Tidal is well in compliance with fecal coliform water quality standards based on modeled results (shown in Table 7A.4.) Thus, no additional load reductions are necessary for this portion of the model. 379 The next downstream SchemaNode (indicated in Figure 7A.11 by an orange circle) along the Aransas River Tidal that is compared to standards in the model is SchemaNode 67. With only the 30% and 45% load reductions applied at the upstream WWTPs, two runs each of 100 and 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted. The geometric mean and percent of samples exceeding 400 CFU/100mL is shown in Table 7A.5. The downstream portion of the Aransas River Tidal exceeds fecal coliform water quality standards for both criteria (geometric mean > 200 CFU/100mL and more than 25% of samples > 400 CFU/100mL) based on modeled results when no additional load reductions are applied (as seen in Table 7A.5.) Figure 7A.10 Load Reductions for SchemaNode 63: Aransas River Tidal 30% 45% 380 Load reductions from the upstream WWTPs, whose locations are shown in Figure 7A.11, were considered. The median bacterial loadings from these WWTPs, which are a Water Reclamation Facility, the City of Taft Baird WWTP, the City of Sinton Main WWTP, and the City of Odem WWTP, were calculated based on literature values or DMRs 20 . If the bacterial loadings at all four WWTPs were reduced by 75%, both contact recreation use standards for fecal coliform are exceeded (shown in Table 7A.5). Thus, livestock bacterial loadings, which may be more easily controlled than other non-point sources and septic system loadings, were also reduced in an attempt to meet water quality standards. The livestock bacterial loadings were reduced at the two watersheds (Figure 7A.11) that are directly upstream of the Aransas River Tidal. Table 7A.5 gives the number of runs of simulations and the modeled results at SchemaNode 67 with various load reductions at the WWTPs and of non-point and livestock sources. Table 7A.5 Modeled Results at SchemaNode 67 with Various Additional WWTP/Livestock/Non-point Load Reductions Run # Load Reduction (%) Bacteria Source Geometric Mean (CFU/100mL) Percent > 400 CFU/100mL 1 0 N/A 454.8 44.5 1 75 WWTPs 269.9 42.2 1 50 WWTPs 423.2 38.0 2 50 Livestock 323.8 39.5 1 75 WWTPs 119.2 26.2 2 75 Livestock/Non-point 137.7 24.5 1 95 WWTPs 167.9 22.2 2 75 Livestock/Non-point 119.3 24.5 1 95 WWTPs 97.8 22.1 2 80 Livestock/Non-point 138.6 20.8 1 95 WWTPs 131.1 18.9 2 85 Livestock/Non-point 119.6 18.2 Simulations 100 1000 20 The WWTP load reductions were based on the overestimated bacterial loadings from WWTPs (explained in Section 5.4.2). 381 Reducing the bacterial loadings from the upstream WWTPs by 95% and the livestock and non-point bacterial loadings by 85% in the adjacent upstream watersheds allows approximately 18% of the samples to exceed the 400 CFU/100mL standard and results in a geometric mean less than 200 CFU/100mL (shown in Table 7A.5.) The reductions necessary to meet fecal coliform contact recreation use standards downstream of the Aransas River Tidal based on modeled results are shown in Figure 7A.11. Reduction of livestock and non-point bacterial loadings would require implementations of best management practices (BMPs), and reduction of WWTP bacterial loadings would require proper disinfection before discharging into surface waters. Significant reductions are needed to meet contact recreation use standards on the downstream portion of the Aransas River Tidal. Because there is no monitoring data available for this location, the load reductions are based on the calibration of the model at the existing monitoring stations and from modeled results. 382 Mission River Above Tidal The Mission River Above Tidal (shown in Figure 7A.12) must meet contact recreation use standards for fecal coliform, but the primary bacterial indicator is E. coli for this segment. However, the results presented are based on fecal coliform water quality standards because a fecal coliform model was created. For fecal coliform, the geometric mean of the samples must be less than 200 CFU/100mL, and single samples must be less than 400 CFU/100mL, but TCEQ allows 25% of the samples to exceed 400 CFU/100mL. The upstream portion of the Mission River Above Tidal (SchemaNode 73), which is indicated in Figure 7A.12 with an orange circle, was analyzed in the calibrated Monte Carlo Simulation Model first, and this node is upstream of Station 12944. Figure 7A.11 Load Reductions for SchemaNode 67: Aransas River Tidal 30% 45% Legend Livestock/Non-Point WWTP 95% 95% 95% 85% 85% 383 Without any load reductions in the upstream watersheds, one run of 100 and 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted. The geometric mean and percent of samples exceeding 400 CFU/100mL is shown in Table 7A.6. The upstream portion of the Mission River Above Tidal exceeds fecal coliform water quality standards for both criteria (geometric mean > 200 CFU/100mL and more than 25% of samples > 400 CFU/100mL) based on modeled results when no additional load reductions are applied (shown in Table 7A.6.) Load reductions from the WWTP upstream of SchemaNode 73, shown in Figure 7A.12, were considered first. The bacterial loadings from this WWTP (Pettus Municipal Utility District WWTP) were calculated based on literature values since no fecal coliform concentration measurements were reported on the DMRs 21 . If the bacterial loadings at the WWTP are reduced by 100%, both contact recreation use standards for fecal coliform are exceeded. It is highly probable that the difference between no load reduction and 100% load reduction at the WWTP may be due to the natural variation of the Monte Carlo Simulation Model because the WWTP bacterial loadings are several orders of magnitude less than other non-point bacterial loadings, and these loadings have a substantial amount of time to decay before reaching SchemaNode 73 in the model. Thus, livestock bacterial loadings, which are more easily controlled than other non-point sources and septic system loadings, were reduced along with WWTP loadings in an attempt to meet fecal coliform water quality standards. The livestock bacterial loadings were reduced at the two watersheds upstream of the Mission River Above Tidal (shown in Figure 7A.12.) The number of runs of simulations and the modeled results at SchemaNode 73 with various WWTP and livestock load reductions are shown in Table 7A.6. 21 The WWTP load reductions were based on the overestimated bacterial loadings from WWTPs (explained in Section 5.4.2). 384 Table 7A.6 Modeled Results at SchemaNode 73 with Various Additional WWTP/Livestock Load Reductions Run # Load Reduction (%) Bacteria Source Geometric Mean (CFU/100mL) Percent > 400 CFU/100mL 1 0 N/A 308.0 39.3 1 100 WWTP 363.0 42.7 50 WWTP 1 50 Livestock (Node 80) 225.5 26.4 50 WWTP 1 50 Livestock 121.9 24 50 WWTP 1 60 Livestock 124.9 25 1 50 WWTP 116.9 17.8 2 70 Livestock 99 19.1 1 50 WWTP 78.5 12.3 2 75 Livestock 64.5 15.1 Simulations 100 1000 Reducing the bacterial loadings from the upstream WWTPs by 50% and livestock bacterial loadings by 75% in the upstream watersheds allows approximately 14% of the samples to exceed 400 CFU/100mL and results in a geometric mean less than 200 CFU/100mL. The reductions necessary to meet fecal coliform contact recreation use standards at the upstream location of the Mission River Above Tidal based on modeled results are shown in Figure 7A.12. Reduction of livestock bacterial loadings would require implementations of agricultural best management practices (BMPs), and reduction of WWTP bacterial loadings would require proper disinfection before discharging into surface waters. Significant reductions are needed to meet contact recreation use standards at the upstream portion of the Mission River Above Tidal. However, there is no monitoring data available for this location, so these reductions are based on the calibration of the model at the existing monitoring stations and from modeled results. Without any reductions at these upstream watersheds, the bacterial monitoring station 12944 meets 385 fecal coliform contact recreation use standards in the model, which agrees with monitoring data 22 . However, the 50% load reduction from the WWTP and 75% load reduction from livestock were applied for the remainder of the load reduction calculations and scenarios (shown in Figure 7A.12). These load reductions will only significantly impact SchemaNode 73 and will not significantly affect the results of the remainder of the model due to the bacterial decay rates. 22 The model was calibrated at all bacterial monitoring stations. Figure 7A.12 Load Reductions for SchemaNode 73: Mission River Above Tidal Legend 50% 75% 75% Livestock WWTP 386 The next downstream SchemaNode (indicated in Figure 7A.13 by an orange circle) along the Mission River Above Tidal that is compared to standards in the model is SchemaNode 74, which is also the location of Station 12944. With only the 50% WWTP and 75% livestock load reductions applied at the upstream watersheds, two runs each of 100 and 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted. For the two separate 1000 and 100 simulation runs, the geometric mean and percent of samples exceeding 400 CFU/100mL are shown in Table 7A.7. Even without the 50% WWTP and 75% livestock loading reductions for SchemaNode 73, the modeled geometric mean and percent of samples greater than 400 CFU/100mL are in compliance with fecal coliform water quality standards. Table 7A.7 Modeled Results at SchemaNode 74 with No Additional Load Reductions Run # Load Reduction (%) Geometric Mean (CFU/100mL) Percent > 400 CFU/100mL 1 0 92.8 15.2 2 0 62.8 15.5 Simulations 100 1000 The middle portion of the Mission River Above Tidal is in compliance with fecal coliform water quality standards based on modeled results (shown in Table 7A.7.) Thus, no additional load reductions are necessary for this portion of the model. 387 The next downstream SchemaNode (indicated in Figure 7A.14 by an orange circle) along the Mission River Above Tidal that is compared to standards in the model is SchemaNode 65. With only the 50% WWTP and 75% livestock load reductions applied at the upstream watersheds, one run of 100 and 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted. The geometric mean and percent of samples exceeding 400 CFU/100mL are shown in Table 7A.8. Note: even without the 50% WWTP and 75% livestock loading reductions for SchemaNode 73, the modeled geometric mean and percent of samples greater than 400 CFU/100mL comply with fecal coliform water quality standards. Figure 7A.13 Load Reductions for SchemaNode 74: Mission River Above Tidal Legend 50% 75% 75% Livestock WWTP 388 Table 7A.8 Modeled Results at SchemaNode 65 with No Additional Load Reductions Run # Load Reduction (%) Geometric Mean (CFU/100mL) Percent > 400 CFU/100mL 1 0 89.6 10.5 Simulations 100 1000 The downstream portion of the Mission River Above Tidal is in compliance with fecal coliform water quality standards based on modeled results (shown in Table 7A.8.) Thus, no additional load reductions are necessary for this portion of the model. Figure 7A.14 Load Reductions for SchemaNode 65: Mission River Above Tidal Legend 50% 75% 75% Livestock WWTP 389 Mission River Tidal The Mission River Tidal (shown in Figure 7A.15) must meet contact recreation use standards for fecal coliform, but the primary bacterial indicator is enterococci for this segment. However, the results presented are based on fecal coliform water quality standards because a fecal coliform model was created. For fecal coliform, the geometric mean of the samples must be less than 200 CFU/100mL, and single samples must be less than 400 CFU/100mL, but TCEQ allows 25% of the samples to exceed 400 CFU/100mL. The upstream portion of the Mission River Tidal (SchemaNode 65) was analyzed as the downstream node to Mission River Above Tidal, so it will not be analyzed again in this section. The next downstream SchemaNode (indicated in Figure 7A.15 by an orange circle) along the Mission River Tidal that is compared to standards in the model is SchemaNode 70. With only the 50% WWTP and 75% livestock load reductions applied at the upstream watersheds, one run of 100 and 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted. The geometric mean and percent of samples exceeding 400 CFU/100mL are shown in Table 7A.9. Table 7A.9 Modeled Results at SchemaNode 70 with No Additional Load Reductions Run # Load Reduction (%) Geometric Mean (CFU/100mL) Percent > 400 CFU/100mL 1 0 59.8 2.8 Simulations 100 1000 The middle portion of the Mission River Tidal is in compliance with fecal coliform water quality standards based on modeled results (shown in Table 7A.9). Thus, no additional load reductions are necessary for this portion of the model. 390 The next downstream SchemaNode (indicated in Figure 7A.16 by an orange circle) along the Mission River Tidal that is compared to standards in the model is SchemaNode 66. With only the 50% WWTP and 75% livestock load reductions applied at the upstream watersheds, two runs of 100 and 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted. For the two separate 1000 and 100 simulation runs, the geometric mean and percent of samples exceeding 400 CFU/100mL are shown in Table 7A.10. The downstream portion of the Mission River Tidal exceeds fecal coliform water quality standards for both criteria (geometric mean > 200 CFU/100mL and more than Figure 7A.15 Load Reductions for SchemaNode 70: Mission River Tidal Legend 50% 75% 75% Livestock WWTP 391 25% of samples > 400 CFU/100mL) based on modeled results when no additional load reductions are applied (shown in Table 7A.10). There are no WWTPs or septic systems in the upstream watershed that discharges directly to the downstream portion of the Mission River Tidal. Thus, livestock and non- point bacterial loadings were reduced in an attempt to meet water quality standards. The bacterial loadings were reduced at the watershed that is directly upstream of the Mission River Tidal; this watershed is shown in Figure 7A.16. The number of runs of simulations and the modeled results at SchemaNode 66 with various load reductions of livestock and non-point bacterial loadings are also shown in Table 7A.10. Table 7A.10 Modeled Results at SchemaNode 66 with Various Additional Livestock/Non-point Load Reductions Run # Load Reduction (%) Bacteria Source Geometric Mean (CFU/100mL) Percent > 400 CFU/100mL 1 0 N/A 433.5 50.2 2 0 N/A 553.9 48.5 50 Livestock 1 0 Non-point 271.5 36.1 70 Livestock 1 0 Non-point 181.5 31.8 1 85 Livestock 77.1 22.8 2 0 Non-point 133.9 25.1 1 85 Livestock 94.3 22.4 2 50 Non-point 118.4 22.7 1 90 Livestock 118.7 18.9 2 0 Non-point 76.4 18.8 Simulations 100 1000 Reducing the livestock bacterial loadings by 90% in the adjacent upstream watershed allows approximately 19% of the samples to exceed the 400 CFU/100mL standard and results in a geometric mean less than 200 CFU/100mL (shown in Table 7A.10). The reductions necessary to meet fecal coliform contact recreation use standards 392 downstream of the Mission River Tidal based on modeled results are shown in Figure 7A.16. The reduction of non-point bacterial loadings supplemented with livestock load reductions did not affect the quality of the river. The geometric mean increased when 50% reduction of non-point bacterial loadings was applied to the model (Table 7A.10). It is highly probable that the difference between non-point load reduction and no non-point reduction may be due to the natural variation of the Monte Carlo Simulation Model because the non-point bacterial loadings are several orders of magnitude less than livestock bacterial loadings. Significant reductions are needed to meet fecal coliform contact recreation use standards on the downstream portion of the Mission River Tidal. However, no monitoring data are available for this location, so these reductions are based on the calibration of the model at the existing monitoring stations and from modeled results. Figure 7A.16 Load Reductions for SchemaNode 66: Mission River Tidal Legend 50% 75% 75% Livestock WWTP 90% 393 Copano Bay Copano Bay must meet oyster harvesting use standards for fecal coliform. The median of the samples (within a two-year period) must be less than 14 CFU/100mL, and the 90 th -percentile of the samples must be less than 43 CFU/100mL (i.e., 10% of the samples are allowed to exceed 43 CFU/100mL.) Aransas River drains into Copano Bay Segment 2 (shown in Figure 7A.17), represented by SchemaNode 154. Without any additional load reductions in the upstream watersheds than those applied in Figure 7A.11, two runs of 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted. For the two separate 1000 simulation runs, the median and 90 th -percentile CFU/100mL are shown in Table 7A.11. Table 7A.11 Modeled Results at SchemaNode 154 with No Additional Load Reductions (Aransas Tidal Load Reductions Included) Run # Load Reduction (%) Median (CFU/100mL) 90 th -percentile > 43 CFU/100mL 1 0 0.43 7.21 2 0 0.43 8.08 Simulations 100 1000 The Copano Bay Segment 2 is in compliance with fecal coliform water quality standards based on modeled results if the Aransas River Tidal load reductions are applied (shown in Table 7A.11.) Thus, no additional load reductions are necessary for this portion of the model if load reductions for the downstream Aransas River Tidal are applied. 394 However, the load reductions required for the downstream portion of the Aransas River Tidal were determined from the modeled results, and there are no monitoring data at this location to verify the results of the model. A possible reason the reductions of the loads are significantly high is because the Model models all the bacteria from the adjacent watersheds draining to the Aransas River Tidal before discharging into the Bay. However, some of this loading could be discharging directly into the Bay. Mission River drains into Copano Bay Segment 3 (shown in Figure 7A.18), represented by SchemaNode 153. Without any additional load reductions in the upstream watersheds than those in Figure 7A.18, two runs of 1000 simulations of the calibrated Monte Carlo Simulation Model were conducted. For the two separate 1000 simulation runs, the median and 90 th -percentile CFU/100mL are shown in Table 7A.12. Figure 7A.17 Load Reductions for SchemaNode 154: Copano Bay (Including Aransas Tidal Reductions) Legend Livestock/Non-Point WWTP 30% 45% 95% 95% 95% 85% 85% 395 Table 7A.12 Modeled Results at SchemaNode 153 with No Additional Load Reductions (Mission Tidal Load Reductions Included) Run # Load Reduction (%) Median (CFU/100mL) 90 th -percentile > 43 CFU/100mL 1 0 0.36 4.95 2 0 0.35 5.11 Simulations 100 1000 Copano Bay Segment 3 is in compliance with fecal coliform water quality standards based on modeled results if the Mission River Tidal load reductions are applied (shown in Table 7A.12). Thus, no additional load reductions are necessary for this portion of the model if load reductions for the downstream Mission River Tidal are applied. Figure 7A.18 Load Reductions for SchemaNode 153: Copano Bay (Including Mission River Tidal Reductions) Legend 50% 75% 75% 90% Livestock WWTP 396 However, the load reductions required for the downstream portion of the Mission River Tidal were determined from the modeled results, and there are no monitoring data at this location to verify the results of the model. Summary Considering only fecal coliform water quality standards, the load reductions required to satisfy the standards for all portions of the model that were analyzed (i.e., the upstream and downstream portion of the Above Tidals and Tidals, the locations of the bacterial monitoring stations, and the four segments in Copano Bay) are shown in Figure 7A.19. This load reduction scenario is referred to as Load Reduction Scenario #1. However, bacterial monitoring data cannot verify all the locations in the model where fecal coliform water quality standards are exceeded. Thus, these load reductions are inconclusive and are only presented to show possible problem areas. Figure 7A.19 Load Reductions to Satisfy Fecal Coliform Standards for Modeled Conditions 30% 45% Legend Livestock WWTP 95% 95% 95% 85% 85% 50% 75% 75% 90% & Non-Point & Non-Point 397 Appendix 8.2: Load Allocations for Scenario #1 This appendix gives the load allocations of all the water segments (Aransas and Mission River Tidals, Aransas and Mission River Above Tidals, and Copano Bay) for Load Reduction Scenario #1. Load Reduction Scenario #1 contains the load reduction necessary to meet fecal coliform water quality standards for all water segments at each location in the model that was analyzed. The locations where the model was analyzed were the upstream and downstream portions of the Above Tidals and Tidals, the locations of the bacterial monitoring stations, and the Copano Bay water segments. However, each portion of the model that was analyzed (that did not meet fecal coliform water quality standards) was not verified by existing monitoring data, so the results are inconclusive based on lack of monitoring data. Thus, these load reductions are only presented to show possible problem areas. Aransas River Above Tidal The recommended load reductions (for Load Reduction Scenario #1) were from upstream WWTPs. The WWTPs and the Aransas River Above Tidal, and the SchemaNodes and SchemaLinks of interest are shown in Figure 8A.1. 398 Since the current loadings were found at the upstream and downstream portions of the water segments, the corresponding load reductions were found for the upstream and downstream portions. The load reduction at the WWTP source for the upstream portion of the Aransas River Above Tidal (SchemaNode 62) is shown in Table 8A.1; only the 30% load reduction WWTP would affect the water quality at the upstream portion. Since the WWTP discharges to the upstream node of the Aransas River Above Tidal, the residence time is zero days. However, the WWTP bacterial loading was largely overestimated (explained in Section 5.4.2); thus, the load reduction from WWTP would be even less than what is indicated. ") ") ") ") # # 1 2 0 113 64 62 30% ") ") ") ") ") ")" ) ") ") ") ") ") ") ") ") ") ")") ") ") ") ") # # # # # # # # # Legend # Outfalls_WWTPs ") BacteriaMonitoringStations_TCEQ 45% Figure 8A.1 Load Reduction Scenario #1: Aransas River Above Tidal 75 399 Table 8A.1 Load Reduction Scenario #1 at Upstream Node of Aransas River Above Tidal SchemaNode Current WWTP Loading (CFU/year) Load Reduction at Source (CFU/year) Total Residence Time to Segment (days) Equivalent Load at Segment (CFU/year) 62 3.22E+13 9.66E+12 0 9.66E+12 Total Load Reduction (CFU/year) 9.66E+12 Current Loading (CFU/year), Table 8.1 3.35E+13 Allowable Load (CFU/year) 2.38E+13 The load reduction at the WWTP sources for the downstream portion of the Aransas River Above Tidal (SchemaNode 75) is shown in Table 8A.2; the 30% and 45% load reductions from WWTPs would affect the water quality at the downstream portion. The 30% WWTP loading (applied at SchemaNode 62) would decay along SchemaLinks 113 and 120 (total residence time = 1.772 days), and the 45% WWTP loading (applied at SchemaNode 64) would decay along SchemaLink 120 (residence time of 1.51 days.) Table 8A.2 Load Reduction Scenario #1 at Downstream Node of Aransas River Above Tidal SchemaNode Current WWTP Loading (CFU/year) Load Reduction at Source (CFU/year) Total Residence Time to Segment (days) Equivalent Load at Segment (CFU/year) 62 3.22E+13 9.66E+12 1.772 2.79E+11 64 4.79E+14 2.16E+14 1.510 1.05E+13 Total Load Reduction (CFU/year) 1.08E+13 Current Loading (CFU/year), Table 8.1 9.45E+13 Allowable Load (CFU/year) 8.37E+13 400 Aransas River Tidal The recommended load reductions (for Load Reduction Scenario #1) were from upstream WWTPs, livestock, and other non-point bacterial sources. The percent of reductions and the corresponding sources (determined in Section 7.3.3), and the SchemaNodes and SchemaLinks of interest are shown in Figure 8A.2. Note that the load reductions for Aransas River Above Tidal were still applied to the model (shown in Figure 8A.1). However, these load reductions were significantly less (since farther upstream) than the load reductions shown in Figure 8A.2. The load reductions and allowable loads for the upstream portion of the Aransas River Tidal were the same as the load reductions and allowable loads for the downstream portion of the Aransas River Above Tidal because it is the same SchemaNode (HydroID 75) in the Schematic Network. (See Table 8A.2 for the load reductions and allowable loads for the upstream portion of the Aransas River Tidal.) The load reductions at the bacterial sources for the downstream portion of the Aransas River Tidal (SchemaNode 67) are shown in Table 8A.3. Note: “WWTP, us” means the WWTPs that were ") ") ") ") ") ") ") ") ") # # # # # 122 1 3 4 1 4 3 92 87 67 104 ") ") ") ") ") ")" ) ") ") ") ") ") ") ") ") ") ")") ") ") ") ") # # # # # # # # # Legend # Outfalls_WWTPs ") BacteriaMonitoringStations_TCEQ Figure 8A.2 Load Reduction Scenario #1: Aransas River Tidal 95% 95% 85% 85% & Non-Point & Non-Point 75 67 401 accounted for upstream of the Aransas River Tidal, and notice that these upstream WWTPs contribute significantly less bacterial loading to the Aransas River Tidal than livestock/non-point sources. Table 8A.3 Load Reduction Scenario #1 at Downstream Node of Aransas River Tidal Schema- Node Source Current Loading (CFU/yr) Load Reduction at Source (CFU/yr) Total Residence Time to Segment (days) Equivalent Load at Segment (CFU/yr) 62 WWTP, us 3.22E+13 9.66E+12 2.67 4.65E+10 64 WWTP, us 4.79E+14 2.16E+14 2.41 1.75E+12 67 WWTP 1.48E+04 1.41E+04 0.00 1.41E+04 69 WWTP 3.37E+11 3.20E+11 0.01 3.14E+11 87 Livestock/ Non-point 1.16E+16 9.84E+15 1.44 5.54E+14 92 WWTP 6.98E+14 6.63E+14 1.50 3.30E+13 104 Livestock/ Non-point 1.02E+15 8.67E+14 1.51 4.23E+13 Total Load Reduction (CFU/year) 6.32E+14 Current Loading (CFU/year), Table 8.3 8.69E+14 Allowable Load (CFU/year) 2.38E+14 Mission River Above Tidal The recommended load reductions were from upstream WWTPs and livestock bacterial sources. The percent of reductions from the corresponding sources (determined in Section 7.3.3), and the SchemaNodes and SchemaLinks of interest are shown in Figure 8A.3. 402 The load reductions at the bacterial sources for the upstream portion of the Mission River Above Tidal (SchemaNode 73) are shown in Table 8A.4. Table 8A.4 Load Reduction Scenario #1 at Upstream Node of Mission River Above Tidal Schema- Node Source Current Loading (CFU/yr) Load Reduction at Source (CFU/yr) Total Residence Time to Segment (days) Equivalent Load at Segment (CFU/yr) 77 Livestock 2.89E+16 2.16E+16 7.95 2.69E+09 80 Livestock 6.17E+16 4.63E+16 2.40 3.78E+14 90 WWTP 1.75E+12 8.75E+11 4.95 4.39E+07 Total Load Reduction (CFU/year) 3.78E+14 Current Loading (CFU/year), Table 8.5 5.10E+14 Allowable Load (CFU/year) 1.32E+14 The load reduction at the WWTP and livestock sources for the downstream portion of the Mission River Above Tidal (SchemaNode 65) is shown in Table 8A.5; the 50% and 75% load reductions from the WWTP and livestock would affect the water ") ") ") ") ") ") ") ") ") ") ") ") ") ") ") ") ")") ") ") ") ") # # # # # # # # # ") ") ") ") # # # # # 1 1 8 1 2 8 1 2 3 90 80 77 Legend # Outfalls_WWTPs ") BacteriaMonitoringStations_TCEQ Figure 8A.3 Load Reduction Scenario #1: Mission River Above Tidal 50% 75% 75% 73 65 403 quality at all the downstream portions (just would impact less the farther downstream.) Table 8A.5 Load Reduction Scenario #1 at Downstream Node of Mission River Above Tidal Schema- Node Source Current Loading (CFU/yr) Load Reduction at Source (CFU/yr) Total Residence Time to Segment (days) Equivalent Load at Segment (CFU/yr) 77 Livestock 2.89E+16 2.16E+16 9.64 9.16E+07 80 Livestock 6.17E+16 4.63E+16 4.09 1.29E+13 90 WWTP 1.75E+12 8.75E+11 6.64 1.49E+06 Total Load Reduction (CFU/year) 1.29E+13 Current Loading (CFU/year), Table 8.5 1.26E+14 Allowable Load (CFU/year) 1.13E+14 Mission River Tidal The recommended load reductions (for Load Reduction Scenario #1) were from upstream WWTP/livestock bacterial sources (shown in Figure 8A.3) and livestock bacterial sources that are shown in Figure 8A.4. The percent of reductions and the corresponding sources (determined in Section 7.3.3), and the SchemaNodes and SchemaLinks of interest are shown in Figure 8A.4. 404 The load reductions and allowable loads for the upstream portion of the Mission River Tidal were the same as the load reductions and allowable loads for the downstream portion of the Mission River Above Tidal because it is the same SchemaNode (HydroID 65) in the Schematic Network; see Table 8A.5 for the load reductions and allowable loads for the upstream portion of the Mission River Tidal. The load reductions at the bacterial sources for the downstream portion of the Mission River Tidal (SchemaNode 66) are shown in Table 8A.6. Figure 8A.4 Load Reduction Scenario #1: Mission River Tidal ") ") ") ") ") ") ") # # # 1 2 91 1 5 1 1 0 82 66 65 90% ! ! ! ! ! ! ") ") ") ") ") ")" ) ") ") ") ") ") ") ") ") ") ")") ") ") ") ") # # # # # # # # # Legend # Outfalls_WWTPs ") BacteriaMonitoringStations_TCEQ 405 Table 8A.6 Load Reduction Scenario #1 at Downstream Node of Mission River Tidal Schema- Node Source Current Loading (CFU/yr) Load Reduction at Source (CFU/yr) Total Residence Time to Segment (days) Equivalent Load at Segment (CFU/yr) 77 Livestock, us 2.89E+16 2.16E+16 10.94 6.81E+06 80 Livestock, us 6.17E+16 4.63E+16 5.39 9.57E+11 82 Livestock 4.34E+16 3.91E+16 1.86 9.48E+14 90 WWTP, us 1.75E+12 8.75E+11 7.89 1.23E+05 Total Load Reduction (CFU/year) 9.49E+14 Current Loading (CFU/year), Table 8.7 1.12E+15 Allowable Load (CFU/year) 1.66E+14 Copano Bay The recommended load reductions (for Load Reduction Scenario #1) were from upstream WWTP/livestock/non-point bacterial sources (shown in Figure 7A.19.) No load reductions were necessary for Copano Bay segments 1 and 4. The load reductions that were accounted for at the Aransas River outlet (Copano Bay Segment 2) are shown in Figure 7A.19. The Aransas River Tidal drains directly into Copano Bay Segment 2, and the only additional loadings to this portion of the Bay were avian. Since the avian loading cannot be reduced, the total load reduction applied at Segment 2 was the same load reduction that was found for the downstream portion of the Aransas River Tidal, which is given in Table 8A.3. The load reduction that occurs at the Copano Bay Aransas River outlet (Copano Bay Segment 2) based on the upstream load reductions that were made is shown in Table 8A.7. Note that this load reduction was significantly more than what would be necessary to meet the oyster water use fecal coliform standards in Copano Bay. 406 Table 8A.7 Load Reduction Scenario #1 at Copano Bay Aransas River Outlet, Segment 2 SchemaNode Source Equivalent Load at Tidal (CFU/yr) Concentration in Bay (CFU/m 3 ) Load in Bay (CFU/yr) 62 WWTP, us 4.65E+10 1.07E+00 2.69E+08 64 WWTP, us 1.75E+12 4.03E+01 1.02E+10 67 WWTP 1.41E+04 3.24E-07 8.14E+01 69 WWTP 3.14E+11 7.22E+00 1.82E+09 87 Livestock/ Non-point 5.54E+14 1.28E+04 3.21E+12 92 WWTP 3.30E+13 7.60E+02 1.91E+11 104 Livestock/ Non-point 4.23E+13 9.74E+02 2.45E+11 Cumulative Runoff. Q (m 3 /yr), Section 6.3.3.4 2.52E+08 Volume of Copano Bay Segment, V (m 3 ), Section 6.3.3.3 5.92E+07 Decay Coefficient of Segment, k (years -1 ), Section 6.3.3.1 7.30E+02 Total Load Reduction (CFU/year) 3.66E+12 Current Loading (CFU/year), Table 8.9 5.04E+12 Allowable Load (CFU/year) 1.38E+12 “Equivalent Load at Tidal (CFU/yr)” is the reduced loading at the downstream portion of the Aransas River Tidal, which is also given in Table 8A.3. “Concentration in Bay (CFU/m 3 )” is the concentration of the bay based on each of the load reductions and is calculated using Equation 6.2, which is found in Section 6.3.2.2. The parameters for Copano Bay Segment 2 that were used to calculate this concentration are also listed in Table 8A.7. The load in the Bay was then calculated by using the following equation: “Concentration in Bay (CFU/m 3 )” * “Cumulative Runoff, Q (m 3 /yr)” to find the load equivalent in the Bay of the load reduction at each of the sources. The load reductions that were accounted for at the Mission River outlet (Copano Bay Segment 3) are also shown in Figure 7A.19. The Mission River Tidal drains directly into Copano Bay Segment 3, and the only additional loadings to this portion of the Bay were avian. Since the avian loading cannot be reduced, the total load reduction applied at Segment 3 was the same load reduction that was found for the downstream portion of the Mission River Tidal, which is given in Table 8A.6. The load reduction that occurs at the 407 Copano Bay Mission River outlet (Copano Bay Segment 3) based on the upstream load reductions that were made is shown in Table 8A.8. Note that this load reduction was significantly more than what would be necessary to meet the oyster water use fecal coliform standards in Copano Bay. Table 8A.8 Load Reduction Scenario #1 at Copano Bay Mission River Outlet, Segment 3 SchemaNode Source Equivalent Load at Tidal (CFU/yr) Concentration in Bay (CFU/m 3 ) Load in Bay (CFU/yr) 77 Livestock, us 6.81E+06 1.22E-04 3.36E+04 80 Livestock, us 9.57E+11 1.72E+01 4.72E+09 82 Livestock 9.48E+14 1.70E+04 4.68E+12 90 WWTP, us 1.23E+05 2.20E-06 6.05E+02 Cumulative Runoff. Q (m 3 /yr), Section 6.3.3.4 2.75E+08 Volume of Copano Bay Segment, V (m 3 ), Section 6.3.3.3 7.60E+07 Decay Coefficient of Segment, k (years -1 ), Section 6.3.3.1 7.30E+02 Total Load Reduction (CFU/year) 4.68E+12 Current Loading (CFU/year), Table 8.9 5.50E+12 Allowable Load (CFU/year) 8.18E+11 The total load reduction and allowable to Copano Bay was found by summing all the load reductions and current loadings from all four Copano Bay Segments. The load reductions, current loadings, and allowable loads necessary to meet fecal coliform standards for Load Reduction Scenario #1 are shown in Table 8A.9. Table 8A.9 Load Reduction Scenario #1 at Copano Bay Portion of Bay Current Load (CFU/yr) Load Reductions (CFU/yr) Allowable Load (CFU/yr) Aransas Outlet (Segment 2) 5.04E+12 3.66E+12 1.38E+12 Mission Outlet (Segment 3) 5.50E+12 4.68E+12 8.18E+11 Copano Creek Outlet (Segment 4) 1.46E+12 0.00E+00 1.46E+12 Watershed JunctionID Outlet (Segment 1) 5.06E+10 0.00E+00 5.06E+10 Total Load 1.20E+13 8.34E+12 3.70E+12 408 Works Cited Alvarado, Sandra. Personal Communication. 1 July 2005. Anderson, Kimberly L., John E. Whitlock, and Valerie J. Harwood, “Persistence and Differential Survival of Fecal Indicator Bacteria in Subtropical Waters and Sediments.” Applied and Environmental Microbiology, 2005. 71(6): p. 3041- 3048. Auer, Martin T. and Stephen L. Niehaus, “Modeling fecal coliform bacteria – I. Field and laboratory determination of loss kinetics.” Water Research, 1993. 27(4): p. 693- 701. Brissaud, F., V. Lazarova, C. Ducoup, C. Joseph, B. Levine, and M.G. Tournoud, “Hydrodynamic behaviour and faecal coliform removal in a maturation pond.”Water Science and Technology, 2000. 42(10-11): p. 119-126. Burkhardt III, William, Kevin R. Calci, William D. Watkins, Scott R. Rippey and Stuart J. Chirtel, “Inactivation of indicator microorganisms in estuarine waters.” Water Research, 2000. 34(8): p. 2207-2214. Canale, Raymond P., Martin T. Auer, Emmet M. Owens, Thomas M. Heidtke, and Steven W. Effler, “Modeling fecal coliform bacteria – II. Model development and application.” Water Research, 1993. 27(4): p. 703-714. Connecticut Department of Environmental Protection. “Wildlife in Connecticut Endangered and Threatened Species Series: Least Tern.” 5 Aug 2005 . Connecticut Department of Environmental Protection. “Wildlife in Connecticut Endangered and Threatened Species Series: Great Egret.” 5 Aug 2005 . Curtis, T.P.D.D. Mara, and S.A. Silva, “The effect of sunlight on faecal coliforms in ponds: implications for research and design.” Water Science and Technology, 1992. 26(7-8): p. 1729-1738. Davies, Cheryl M., Julian A.H. Long, Margaret Donald, and Nicholoas J. Ashbolt, “Survival of Fecal Microorganisms in Marine and Freshwater Sediments.” Applied and Environmental Microbiology, 1995. 61 (5): p. 1888-1896. 409 Environmental Protection Agency (EPA). “Monitoring and Assessing Water Quality: 5.11 Fecal Bacteria.” March 2006. 17 March 2006 . Environmental Protection Agency (EPA). “Protocol for Developing Pathogen TMDLs: First Edition.” Jan 2001. 15 May 2005 . Gil de Weir, Karine, Enrique Weir, Clark Casler, and Sara Aniyar. “Ecological Functions and Economic Value of the Neotropic Cormorant.” Wildlife and Fisheries Sciences, Texas A&M University. 5 Aug 2005 . Goodall, Jon and Timothy Whiteaker, “Integrating Arc Hydro Features with a Schematic Network.” Transactions in GIS, 2006. 10(2): p. 219. Kashefipour, S.M., B. Lin, E. Harris, R.A. Falconer, “Hydro-environmental modeling for bathing water compliance of an estuarine basin.” Water Research, 2002. 36: p. 1854-1868. Khan, S.I. and N. Kamal, “Assessment of treatment efficiency by quantitative recovery of indicator bacteria and pathogens in sewage effluents." Advances in Water and Wastewater Treatment Technology, 2001. p. 129-137. Maidment, David R. Handbook of Hydrology. New York: McGraw-Hill, Inc., 1992. Merwade, Venkatesh, and David Maidment. “River Channel Morphology Model – A tool for analyzing and extrapolating river channel bathymetry.” Center for Research in Water Resources, University of Texas at Austin. 18 March 2006 . Mott, Joanna B., and Roy L. Lehman. “Bacteria Source Tracking in Copano Bay: Phase II Final Report.” Texas A&M University – Corpus Christi, 2005. Niemann, Kerry. Personal Communication. 17 Jan 2006. Pearson, H.W., D.D. Mara, S.W. Mills, and D.J. Smallman, “Physico-Chemical Parameters Influencing Faecal Bacterial Survival in Waste Stabilization Ponds.” Water Science and Technology, 1987. 19(12): p. 145-152. Percevia. “Field Guide to Birds of North America: Herring Gull Breeding Adult.” 5 Aug 2005 . Percevia. “Field Guide to Birds of North America: Cattle Egret.” 5 Aug 2005 . 410 Percevia. “Field Guide to Birds of North America: Roseate Spoonbill.” 5 Aug 2005 . Percevia. “Field Guide to Birds of North America: Reddish Egret.” 5 Aug 2005 . Quenzer, Ann Marie. “A GIS Assessment of the Total Loads and Water Quality in the Corpus Christi Bay System.” Master’s thesis, University of Texas at Austin, 1997. Texas Commission on Environmental Quality (TCEQ). “A Comparison Between Fecal Coliform, E. coli, and Enterococci as Bacterial Indicators in Southeast Texas Surface Waters.” March 2003. 17 March 2006 . Texas Commission on Environmental Quality (TCEQ). “Draft 2004 Texas Water Quality Inventory and 303(d) List.” 23 Nov 2005. 17 March 2006 . Texas Commission on Environmental Quality (TCEQ). “2000 Texas Surface Water Quality Standards.” 26 July 2000. 5 April 2005 . Texas Commission on Environmental Quality (TCEQ). “Texas Surface Water Quality: What Is It, and How Is It Measured?” August 2005. October 2005 . Texas Commission on Environmental Quality (TCEQ). “Chapter 1: Introduction to the SWQM Portion of the TRACS Database.” 12 Feb 2006 . Texas Commission on Environmental Quality (TCEQ). “2004 Surface Water Quality Monitoring Data Management Reference Guide.” 19 Dec 2005. 12 Feb 2006 . Texas Department of Health, Seafood Safety Division. Comprehensive Sanitary Survey of the Shellfish Producing Waters of Copano Bay. Texas Department of Health, 2000. To, Ernest. “Monte Carlo analysis of the Copano Bay fecal coliform model.” 13 Dec 2005 . 411 United States Geological Survey (USGS). “Biological and Ecotoxicological Characteristics of Terrestrial Vertebrate Species Residing in Estuaries: Laughing Gull.” 5 Aug 2005 . United States Geological Survey (USGS). “Biological and Ecotoxicological Characteristics of Terrestrial Vertebrate Species Residing in Estuaries: Tricolored Heron.” 5 Aug 2005 . United States Geological Survey (USGS). “Biological and Ecotoxicological Characteristics of Terrestrial Vertebrate Species Residing in Estuaries: Black Skimmer.” 5 Aug 2005 . United States Geological Survey (USGS). “Biological and Ecotoxicological Characteristics of Terrestrial Vertebrate Species Residing in Estuaries: Great Blue Heron.” 5 Aug 2005 . United States Geological Survey (USGS). “Biological and Ecotoxicological Characteristics of Terrestrial Vertebrate Species Residing in Estuaries: Snowy Egret.” 5 Aug 2005 . United States Geological Survey (USGS). “Biological and Ecotoxicological Characteristics of Terrestrial Vertebrate Species Residing in Estuaries: American Oystercatcher.” 5 Aug 2005 . United States Geological Survey (USGS). “Biological and Ecotoxicological Characteristics of Terrestrial Vertebrate Species Residing in Estuaries: Fulvous Whistling Duck.” 5 Aug 2005 . United States Geological Survey (USGS). “Biological and Ecotoxicological Characteristics of Terrestrial Vertebrate Species Residing in Estuaries: Forster’s Tern.” 5 Aug 2005 . Van der Steen, P., A. Brenner, Y. Shabtai, and G. Oron, “The effect of environmental conditions on faecal coliform decay in post-treatment of UASB reactor effluent.” Water Science and Technology, 2000. 42(10-11): p. 111-118. Ward, George. Personal Communication. 19 May 2005. Ward, George H., and Neal E. Armstrong. “Ambient Water, Sediment, and Tissue Quality of Corpus Christi Bay Study Area: Present Status and Historical Trends.” Center for Research in Water Resources: The University of Texas at Austin, 1997. Wikipedia. “Beta distribution.” 10 Feb 2006 . 412 Zoun, Reem Jihan. “Estimation of Fecal Coliform Loadings to Galveston Bay.” Master’s thesis, The University of Texas at Austin, 2003. 413 Vita Carrie Jo Gibson was born on April 15, 1981 in Bellevue, WA to her mother and father, Deborah Lynn Mohorovich and Terrence Lynn Gibson. She graduated from Lake Washington High School in June 2000, located in Kirkland, WA, with a 4.0 GPA while competing in varsity Track and Cross Country and taking honors and Advanced Placement (AP) courses. She then attended Gonzaga University in Spokane, WA with her twin sister, Kim, where she received her Bachelor’s of Science degree in Civil Engineering May 9, 2004, maintaining a 3.91 GPA while competing in varsity Track and Cross Country. She has been pursuing her Master’s in Environmental and Water Resources Engineering (EWRE) at the University of Texas at Austin. Upon receiving her degree, she is returning to Seattle, WA and starting her career at HDR Engineering, Inc located in Bellevue, WA. Permanent address: 6210 144 th Avenue NE, Redmond, WA 98052 This dissertation was typed by Carrie Jo Gibson