1. Bacterial TMDL Model for Copano Bay Research performed by Carrie Gibson at Center for Research in Water Resources Schematic processor tool developed by Tim Whiteaker at CRWR Research supported by Texas Commission for Environmental Quality

2. Presentation Outline Background Scope of Work Bacterial Loading Water Quality Model –Non-Point Source Bacterial Loading Calculations/Results –Point Source Bacterial Loading Calculations/Results –Modeling Bacteria Transport: Schematic Processor –Calibration of Model Conclusions Future Work

3. Project Location Copano Bay watershed Copano Bay

4. Background Section 303(d) of 1972 Clean Water Act (CWA) Texas Surface Water Quality Standards –Fecal coliform bacteria Oyster water use Contact Recreation Use Aransas River Mission River Copano Bay

5. Existing Monitoring Data

6. Scope of Work 1.Identify major bacterial sources in Copano Bay watershed. 2.Calculate total bacterial loadings, Total Maximum Daily Loads (TMDLs), from bacterial sources. 3.Determine amount of load reductions that is needed to meet water quality standards.

7. Potential Bacteria Sources Non-point bacteria sources Point sources –Concentrated Animal Feedlot Operations (CAFOs) –Livestock (cattle, goats, horses, sheep, hen, hogs, and chickens) –Wastewater Treatment Plants (WWTPs) –Septic Systems –Waterbirds

8. Non-Point Bacterial Loadings Basic Equation: L = Q * C L = Bacterial loadings (cfu/year) Q = Runoff (m 3 /year) C = Fecal coliform concentration (cfu/m 3 )

9. Runoff (Q) Calculations Rainfall-runoff equations derived by Ann Quenzer –Based on: land use and precipitation Precipitation Data (inches/year) Land Use Land Cover Data Runoff, Q (m 3 /year) Quenzer Equations

10. EMC (C) Calculations From Reem Jihan Zoun’s thesis, Estimation of Fecal Coliform Loadings to Galveston Bay Modified dbf table in order not to account for livestock fecal wastes twice Land Use Code CategoryFecal Colonies per 100 mL 11Open Water0 21Low Intensity Residential22,000 22High Intensity Residential22,000 23Commercial/Industrial/Transportati on 22,000 31Bare Rock/Sand/Clay0 32Quarries/Strip Mines/Gravel Pits0 41Deciduous Forest1,000 42Evergreen Forest1,000 43Mixed Forest1,000 51Shrubland2,500 61Orchards/Vineyards/Other2,500 71Grasslands/Herbaceous2,500 81Pasture/Hay2,500 82Row Crops2,500 83Small Crops2,500 85Urban/Recreational Grasses22,000 91Woody Wetlands200 92Emergent Herbaceous Wetlands200 0 0 0

11. Creation of EMC Grid Land Use Land Cover Data Join based on land use code EMC dbf table Fecal Coliform Concentration, C (cfu/m 3 )

12. Non-Point Bacterial Loading Grid C (cfu/m 3 )Q (m 3 /year)L (cfu/year) * = Annual Bacterial Loading per Grid Cell

13. Non-Point Loading per Watershed Annual Bacterial Loading per grid cell (cfu/year) Zonal Statistics Annual Bacterial Loading per Watershed (cfu/year) Bacteria Monitoring Stations USGS Gauge Stations Water Segment Endpoints Watersheds Delineated Watersheds using WRAP Hydro

14. Point Source Calculations: Livestock Cattle, goats, horses, sheep, layers, hogs, chickens Data (annual animal count per county) from: –2002 Census of Agriculture, National Agricultural Statistics Service (NASS) –2004 Texas Livestock Inventory and Production, United States Department of Agriculture (USDA), NASS, Texas Statistical Office

15. Livestock Loading Results Results Add cfu/year to non-point bacterial loading calculations Livestock Bacterial Loadings Legend Animal_cfu_year 5.37e+013 - 1.58e+015 1.59e+015 - 7.50e+015 7.51e+015 - 1.35e+016 1.36e+016 - 4.14e+016 4.15e+016 - 1.17e+017

16. Point Source Calculations: Avian Texas Colonial Waterbird Census (TCWC) Breeding Pair LocationsLocations of Applied Avian Loads

17. Avian Loading Results Results Add cfu/year to non-point bacterial loading calculations Legend Birds_cfu_year 1.48e+009 1.49e+009 - 2.22e+009 2.23e+009 - 1.22e+011 1.23e+011 - 2.75e+011 2.76e+011 - 3.96e+011

18. Bacterial Loading to Watersheds Results Legend Livestock Non-Point Avian

19. Water Quality Model Runoff (m 3 /yr) Concentration (cfu/m 3 ) Load (cfu/year) Cumulative Loading per Watershed Cumulative Runoff per Watershed Schematic Processor Created Water Quality Model using Model Builder

20. Bacterial Loading Transport using Schematic Processor Creation of Schematic Network Watershed Drainage Junction Bay Watershed to Junction Junction to Junction Junction to Bay

21. Schematic Processor Implements DLLs Dynamic linked libraries, DLLs –First-order decay Simulates decay of bacteria along stream segments load passed = load received * e -kt k = first-order decay coefficient (day -1 ) - stored as attribute in SchemaLink t = travel time along streams, t (days) - stored as attribute in SchemaLink Decay

22. Copano Bay acts as CFSTR –CFSTR Assumptions –Bay is completely mixed and acts as Continuous Flow, Stirred Tank Reactor (CFSTR) –Inflow = Outflow c = L/(Q+kV) c = concentration in bay (cfu/m 3 ) L = bacteria load entering bay (cfu/yr) Q = total flow (m 3 /yr) – stored as attribute in SchemaNode k = first-order decay coefficient (day -1 ) - stored as attribute in SchemaNode V = volume of bay (m 3 ) – stored as attribute in SchemaNode

23. Schema Links and Nodes

24. Computations along the network

25. Moving material through links and nodes

26. Processing Steps

27. DLL’s have the processes in them

28. Schematic Processor Parameters Parameters (Inputs) –SchemaLink (SrcTypes 1 and 2) Residence Time (  in days), Decay Coefficient (k in day -1 ) –SchemaNode SrcType 3 – Copano Bay –Volume (V in m 3 ), Decay Coefficient (k in day -1 ) –Cumulative Runoff (Q in m 3 /year) SrcType 1 – Watersheds –Bacterial Loading per Watershed (L in cfu/year) Determined by User Calculated from Previous Steps in Model Builder

29. Model Calibration: Aransas River Sta. 12952 USGS08189700 Sta. 17592 Sta. 12945 Sta. 14783 Sta. 12948 Calibration Locations (Four)

30. Model Calibration: Aransas River Goal: Adjust upstream k and  values of each calibration location until median concentration of existing data is achieved. Then “set” k and  parameter values and work on the next downstream calibration location (bacteria monitoring station.) Nodes/Links’ parameters that can be varied for each bacteria monitoring station calibration

31. Modeled versus Existing Data Sta. 17592 Existing F.C. Median: 260 cfu/100mL Modeled F.C. Median: 260 cfu/100 mL Sta. 12952 Existing F.C. Median: 72 cfu/100 mL Modeled F.C. Median: 72 cfu/100 mL Sta. 12948 Existing F.C. Median: 96 cfu/100 mL Modeled F.C. Median: 96 cfu/100 mL Copano Bay (Aransas R. Outlet) Existing F.C. Median: 2 cfu/100 mL Modeled F.C. Median: 2 cfu/100 mL

32. Modeled versus Existing Data

33. Conclusions Major point and non-point source bacterial loadings have been calculated. Bacterial Loadings Water Quality Model has been created. Model has been calibrated (adjusting k and  parameters) to existing median bacteria monitoring data. There is uncertainty in the calculations of bacterial loadings and in the determination of parameters.

34. Future Work Determine reasonable decay coefficient (k) values for rivers and compare to k values in calibrated model. Conduct parameter optimization and a Monte Carlo simulation on the model. Determine the current load, allowable load, and the load reductions necessary to meet water quality standards for each TCEQ segment.