Street, Atlanta, GA 30303, ,

Transcription

1 The Development of a Hydrodynamic and Water Quality Model to Support TMDL Determinations and Water Quality Management of a Stratified Shallow Estuary: Mobile Bay, Alabama Tim A. Wool 1, Steven R. Davie 2, Yuri M. Plis 2, John Hamrick 3 1 US EPA Region 4, Water Management Division, Atlanta Federal Center, 61 Forsyth Street, Atlanta, GA 30303, , wool.tim@epa.gov Abstract 2 Tetra Tech. Inc., 2110 Powers Ferry Road, Suite 202, Atlanta, GA, 30339, steven.davie@tetratech-ffx.com, yuri.plis@tetratech-ffx.com 3 Tetra Tech. Inc., Eaton Place, Suite 340, Fairfax, VA 22030, john.hamrick@tetratech-ffx.com Mobile Bay is located in the southwestern portion of Alabama and is currently on the State of Alabama s 303(d) list of impaired waters. Large portions of the estuary are listed for organic enrichment, low dissolved oxygen, and fecal coliform. Mobile Bay is a shallow, stratified estuary in which the water quality is controlled by freshwater inflow from the Mobile River, wind, tidal variation and anthropogenic loadings. Because of the complexities of transport and the interaction with water quality, simplistic modeling approaches could not be used to develop a TMDL or be used as management tools for the waterbody. Several components of the EPA TMDL Modeling Toolbox are being used to develop the TMDL and to be used in the future by the State of Alabama as a management tool for water quality issues in the bay. The primary components of the Toolbox that is being used are: the Water Resources Database, Environmental Fluid Dynamics Code, and the Water Quality Analysis Simulation Program. There are several other ancillary tools within the Toolbox that will aid in the application of models. Introduction Total Maximum Daily Load (TMDL) development for the Mobile Bay estuary requires the application of sophisticated modeling tools to link changes in loadings to the system through management strategies to improvement of water quality. Typically, models are

2 applied to waterbodies for the express purpose of determining the TMDL. In this application the there are two goals: (1) determine the TMDL and (2) develop a management tool that can be used by the Alabama Department of Environmental Management (ADEM) in the future to manage water quality in the bay. Mobile is located on the northern coast of the Gulf of Mexico. Geomorphologically, it is a combination of the drowned river valley and bar-built estuary. Considering its shallow depths, Mobile Bay is bathymetrically complex, with deep holes in the northeastern part and a manmade island in the northwest area. There are two openings in the lower part of the Bay: to the Gulf of Mexico and Mississippi Sound. A ship channel, 12-meters deep, cuts through the Main Pass to the Port of Mobile. Mobile Bay s water quality is highly influenced by its natural geographic location, weather patterns of the watershed, and human uses. The Mobile River system delivers 95% of the total freshwater input to the bay. The average discharge of the system is about 1,840 m 3 /s, but during winter-spring rainfalls it can exceed 14,000 m 3 /s and during summer low-flow season decrease to 240 m 3 /s. The Mobile Bay drainage system is nation s sixth largest in area and fourth largest in discharging volume. As a result, urban and agricultural development in the Bay s surrounding areas and in areas far outside the coastal region impact Mobile Bay s water quality characteristics. Hypoxic and anoxic conditions are common in Mobile Bay and are generally prevalent during the summer months. These frequently stressed water quality conditions marked by stratification with low dissolved oxygen. Specific for the bay environmental problem is the Jubilee (local name) phenomenon. It is the east shoreward movement of dense concentrations of fish and invertebrates that has been observed in Mobile Bay since at least 1867, before any significant man-made environmental impact had been registered. Presumably this phenomenon associates with the formation and shoreward movement of a low oxygen zone. This persistent pattern of hypoxia occurs when winds blowing from the mainland drive surface waters from shore, causing deeper, poorly oxygenated water to move into the shallows. Mobile Bay waters are monitored for pathogens and often are subject to closings, advisories, or warnings during situations when excess fresh waters entered Mobile Bay through Mobile River or due to sanitary sewer collection system failures. The Alabama 303(d) list of impaired waters includes parts of Mobile Bay impaired by organic enrichment/low dissolved oxygen and pathogens (Figure 1). TMDLs will be developed for all of the listed segments.

3 MOBILE COUNTY N W E BALDWIN COUNTY S Miles 303d Listed (Poly) 303d Listed (Line) Cataloging Unit Boundaries State Boundaries County Boundaries Bayou la Batre (OE/DO and Pathogens) Mobile Bay (OE/DO) Mobile Bay (Pathogens) Fish River (Pathogens) Portersville Bay (Pathogens) Mississippi Sound (Pathogens) Bon Secour Bay (Pathogens) Gulf of Mexico (Mercury) Intracoastal Waterway (OE/DO) Intracoastal Waterway (OE/DO) Figure 1 Mobile Bay 303(d) Listed Segments Water Quality Management Objective The primary purpose of the application of the EPA TMDL Toolbox (Figure 2) to Mobile Bay is to support the development of a series of TMDLs to meet a consent decree. Another objective of this application is to develop a management tool that the State of Alabama can use to water quality in Mobile Bay. The TMDL Toolbox integrates various models used to develop TMDLs as well as traditional water quality modeling. The Toolbox is developed in such manner that they can be applied in various degrees of complexity which make it ideal as no only a TMDL tool but a water quality management tool. The models will be applied in such a manner that areas of the bay can be focused to look at near and far field affects of dischargers or storm water management in the basin. This will involve the development of sub grids of the bay. These sub grids will be made up of areas of concern as Figure 2 EPA's TMDL Modeling

4 determined by the ADEM, it will enable the determination NPDES permit applications as well as control of storm water and the development of best management practices in the basin. The major tasks of the developing a modeling system for Mobile Bay were as follows: o Data Assimilation o Hydrodynamic Model Calibration o Enhanced Linkage between Hydrodynamics and Water Quality Model o Development of Sediment Diagenesis Algorithms o Water Quality Model Calibration o TMDL Development The data assimilation task is the most critical task because it determines the model calibration strategy and model calibration/confirmation periods. Ultimately, it dictates how well the model is capturing the hydrodynamic and water quality processes in the system by time series and statistical comparisons to the measured data. For Mobile Bay, all of the hydrodynamic and water quality data were gathered and put into a Water Resources Database (WRDB). This database was developed by EPA Region 4 and GAEPD and is free software that can be downloaded at and is part of the TMDL Modeling Toolbox. Data were gathered from agencies such as EPA, ADEM, USGS, Alabama Water Watch, Dauphin Island Sea Laboratory, and Mobile Bay NEP. After compiling all of the data in the bay and contributing watersheds in the vicinity of the bay, there were 1,765,497 data records from 1,076 different stations, and 77 different parameters. This is a large dataset for the Mobile Bay waterbody. WRDB tracks database entries and allows for easy organization of datasets from various agencies with different laboratory and/or sampling methods. WRDB is also functional for summarizing parameters, stations, statistics, and time periods of the data. Table 1 is an example of data summary tables from the Mobile Bay WRDB. Table 1 Summary Table of Water Quality Data from Mobile Bay WRDB Parameter Units # Obs Mean Min Max First Date Last Date BOD5 mg/l /03/ /04/2002 DO mg/l /17/ /26/2003 NH3 mg/l /03/ /26/2003 NO2-NO3 mg/l /05/ /16/2001 TP mg/l /03/ /23/2002

5 Upon significant review of the existing data in WRDB, it was determined that many of the stations in the WRDB (850 out of 1,076) were collected as part of the ALAMAP random sampling program that identifies spatial areas in the bay of noncompliance. The data collected as part of ALAMAP are a snap-shot in time of large spatial extents of the bay. While these data are beneficial to ADEM in evaluating the large waterbody, they are not useful in calibrating the complex, three-dimensional hydrodynamic and water quality models. By continuing to analyze the data and identifying calibration processes and time periods, several data gaps were determined. They were as follows: o Upstream freshwater flows into bay o Recent (5 years) water quality data in the bay o Continuous dissolved oxygen data o Chlorophyll-a measurements o Light profiles o Bottom salinity data in the channel o Longterm BOD tests in bay The largest data gap was determined to be the amount of freshwater flow coming into Mobile Bay. The most downstream USGS gages near the bay are located at Claiborne Dam ( ) and the Coffeeville Dam ( ), as shown in Figure 3. These two gages are longterm flow gaging stations and are currently used to quantify the freshwater flow delivered to the bay. A drainage area ration is used to adjust for the additional flow from areas downstream of these two gages. After several preliminary model runs, it was evident that the transport, mixing, and stratification in the bay is dominated by the amount of freshwater flow moving down the Mobile and Tensaw Rivers. Therefore, this is a critical component of calibrating the Mobile Bay hydrodynamic and water quality models.

6 Figure 3 Location of Continuous USGS Gaging Stations in the Mobile Bay Watershed.

7 There is a significant gap in water quality data in Mobile Bay for calibration of a water quality model. ADEM stopped collecting BOD, nutrient, chlorophyll-a, and dissolved oxygen data in 1993 when the ALAMAP program was initiated. EPA and ADEM coordinated a meeting with the Mobile Bay NEP in May 2003 to discuss the data gaps and coordination among multiple agencies for additional data collection efforts. A concerted effort was made to collect data during a consistent time period for April through October of The coordinated data collection efforts outlined were as follows: 3 Continuous velocity monitors (ADCP bottom-mounted or ADP side-mounted) with yearly maintenance. o Mainstem upstream of Tensaw River split o Tensaw River at Bridge o Mobile River downstream of Tensaw River split Continuous monitoring from YSI or HydroLab at 15-min intervals for DO, salinity, ph, and temp at the 8 sites below and shown in Figure 4. Servicing interval = 1 week (preferably 2-3 days depending on Dauphin Island experience). The vertical depth of the instruments should be 5-feet from the water surface (ADEM s compliance point). Secchi or PAR sensor monitoring to determine light extinction and the photic zone. Monthly Nutrient grab samples at bottom, mid-depth, and surface for the following o NH3 o TKN o NO3-NO2 o TP o Ortho-P o BOD5 o Chlorophyll-a Vertical profiles with handheld YSI or HydroLab at all 8 water quality stations listed below (Figure 4) for DO, salinity, ph, and temperature. Longterm BOD monitoring ( days) for 8 sites (Figure 4) at least once during data collection period. Mid-Depth sample preferred.

8 N W E Channel 1 MB7 %U New Channel S Stations Water Quality Stations NOAA Shoreline Fine Model Grid Hydro-lines-geo-in AL Reach File, V Miles %U MB6 Channel 2 MB9 %U MB5 MB4 MB3 MB MB2 Figure 4 Location Map of Mobile Bay Water Quality Stations for Additional Data Collection. The Environmental Fluid Dynamics Code (EFDC) was applied to the Mobile Bay Estuary. EFDC is a two- and three-dimensional, hydrodynamic model that solves the circulation and transport of material in complex surface water environments including estuaries, coastal embayments, lakes, and offshore. EFDC is an orthogonal, curvilinear grid, hydrodynamic model. EFDC provides solutions for salinity, temperature, and conservative tracers with full density feedback to handle stratified conditions. EFDC is the key hydrodynamic model in the TMD Modeling Toolbox. The model grid resolution plays an important role in defining the hydrodynamic and water quality processes that are desired in the simulations. The horizontal resolution should be fine enough to capture the dominant transport processes in the bay that have an impact on the water quality. The vertical resolution typically controls the amount of stratification occurs for hydrodynamic parameters such as salinity and temperature and water quality parameters such as dissolved oxygen. In addition to computer speed and length of simulation period, both the horizontal and vertical resolution of the model grid

9 affect the computer run times. The goal of the model development for Mobile Bay was to provide EPA and ADEM a management tool for the bay in determining NPDES permit requirements, TMDL allocations, and critical areas for additional data. There were two model grids developed for Mobile Bay. The first grid is a coarse grid containing 557 horizontal cells and 3 layers for a total number of segments of 1,671. The coarse grid is shown in Figure 5. The second one is the original fine grid that was extended to cover the western shores to Mississippi and contains 1,608 horizontal cells and 4 layers for a total number of segments of 6,432. The fine grid is shown in Figure 6. Model run times are presented in Table 2. Table 2 Comparison of Coarse and Fine Grid Computer Runs. COARSE GRID FINE GRID Number of Cells 557 x 3 layers = 1,671 1,608 x 4 layers = 6,432 Model Time Step 75 seconds 75 seconds Computer Processor 1.8 GHz 2.6 GHz 1.8 GHz 2.6 GHz Computer Hours / Simulated Year The first phase of the Mobile Bay model development was able to achieve several goals such as directing a coordinated monitoring plan for the bay, test model grids for achieving a management tool, and examining the transport patterns in the bay.

10 Figure 5 Coarse Grid for Mobile Bay Model Development.

11 Figure 6 Fine Grid for Mobile Bay Model Development.

12 The modeling will help focus future data collection in the bay by determining water quality data limited areas in the Bay. TMDL Development Supporting Tools TMDL is, by definition, the sum of the individual wasteload allocations for point and nonpoint sources and natural background with a margin of safety. The optimal solution of the wasteload allocation problem is typically requiring the application of mathematical models to estimate unknown loads, relate loads to target concentrations, and to evaluate implementation strategies to achieve water quality targets. U.S. EPA directs and supports efforts of development, testing and applications of special TMDL modeling tools and making these tools and training on these tools available to partner states and other interesting parties. The current TMDL Toolbox placed on EPA Region 4 site ( It is the collection of stand alone models for dynamic simulation of hydraulics, hydrodynamics and water quality in surface waters, including overland flow, streams, rivers, lakes, estuaries, coastal embayment and offshore. All of the models in the Toolbox have a proven record in TMDL determination. Water Quality Model Description Nutrient enrichment and eutrophication are continuing concerns in Mobile Bay. A TMDL for nutrients should be established for the bay to control nuisance phytoplankton blooms due to nutrient enrichment. High concentrations of nitrogen and phosphorus can lead to periodic phytoplankton blooms and an alteration of the natural trophic balance. These periodic blooms of phytoplankton hand in hand with strong vertical density stratification cause dissolved oxygen levels to fluctuate widely, and low dissolved oxygen concentrations in bottom waters result. The purpose of the water quality model is to predict a response in chlorophyll-a and dissolved oxygen concentrations as a function of nutrient loadings and transport throughout Mobile Bay. The model should be applied to evaluation of various loading scenarios as well as looking at use support areas within the 303(d) listed segments. The nutrient enrichment, eutrophication, and dissolved oxygen depletion processes are predicted using WASP. Several physical-chemical processes can affect the transport and interaction among the nutrients, phytoplankton, carbonaceous material, and dissolved oxygen in the aquatic environment. Figure 7 presents the principal kinetic interactions for the nutrient cycling and dissolved oxygen that were simulated in WASP.

13 Figure 7 Representation of WASP State Variables and Kinetic Interactions EPA Region 4 has taken the lead on the development, support, and training for the WASP 6.1. It is a dynamic water quality model that is routinely used throughout the United States for the development of TMDLs and other waste load allocation studies. The model contains algorithms for conducting: 1) Eutrophication/Conventional Pollutants, 2) Organic Chemicals/Simple Metals, 3) Mercury, 4) Temperature, 5) Fecal Coliform. EFDC runs independently of WASP, simulates the hydrodynamics and constituent transport and writes hydrodynamic linkage for the WASP. These widely tested and accepted modeling tools were chosen to provide an informational support to TMDL development for the Mobile Bay, Alabama. Target for TMDL The first step in determining a TMDL is to identify the appropriate endpoint that is needed for the waterbody to attain its designated use. Mobile Bay was placed on the State of Alabama s 303(d) list due to exceedances of its water quality criterion for dissolved oxygen. Dissolved oxygen along with nutrients and chlorophyll-a were chosen as TMDL endpoints. The TMDL model uses these targets, and based on the available data and information on Mobile Bay, calculates the amount of aforementioned substances that can be assimilated by the bay and meet the state standards.

14 Water Quality Model Calibration The procedure of the water quality model calibration for the shallow stratified estuary was discussed in (Wool et al., 2003). The model should be calibrated for all constituents: ammonia, nitrate, organic nitrogen, chlorophyll-a, orthophosphate, organic phosphorus, BODu, and dissolved oxygen. The calibration objectives are as follows; Parameterize the model to best represent the spatial nutrient gradients in Mobile Bay system, Account for seasonal variability in nutrient concentrations at monitoring stations, Predict the chlorophyll concentrations in both time and space. Predict the dissolved oxygen in both surface and bottom waters The judgment of fit of the calibration is both a qualitative (best professional judgment of model fit to observed data) and quantitative (statistical methods used to determine goodness of fit). The quantitative approach of calibration is beneficial in that it provides a direct means of determining how well the model is capturing the variability in the system, as well as a means to compare other modeling approaches. The qualitative approach provides insight into how well the model is representing the gradients and seasonality of water quality within Mobile Bay. On basis of calibrated water quality estimates will be develop the pollution reduction scenarios. The work on water quality model calibration and reduction scenarios establishing is in progress at present time. Conclusions The problems of an adequate understanding and description of complex processes of circulation and water quality dynamics in shallow density stratified estuaries can be effectively solved by applying the state-of-the-art three-dimensional hydrodynamic and water quality mathematical models. This approach allows taking into account, analyzing and ranking numerous interacting hydrological, meteorological and biological factors, impact of natural and anthropogenic sources of pollution. The meeting such requirements mathematical models are develop and supported by U.S. EPA in the framework of the TMDL Modeling Toolbox. In relation to water quality and hydrodynamics simulations the toolbox proposes Water Quality Analysis Simulation Program (WASP 6.1) and Environmental Fluid Dynamics Code (EFDC), respectively. Particularly, the complex of WASP EFDC is capable to provide insight into details of delicate mechanisms of water quality and hydrological regimes formations in such

15 complex water systems like shallow stratified estuaries. The main condition of successful application of WASP EFDC complex to the TMDL establishing problem is the accomplishment of versatile and accurate model calibration validation procedure. Mobile Bay case study gave us one more demonstration of successful applicability of WASP EFDC modeling tool to analysis of complex hydrological - ecological processes in the large river estuary while maintaining a reasonable computer run time for management alternatives. References Noble, M.A., Schroeder, W.W., Wiseman Jr., W.J., Ryan H.F., and Gelfenbaum G. (1996). Subtidal circulation patterns in a shallow, highly stratified estuary: Mobile Bay, Alabama. J. Geophys. Res., 101(C11), 25,689-25,703. Wool, T.A., Davie S.R., and Rodriguez H.N. (2003). Development of three-dimensional hydrodynamic and water quality models to support total maximum daily load decision process to the Neuse River Estuary, North Carolina. J. Water Resour. Plan. Manage.,129, Schroeder, W.W. (1978). Riverine influence on estuary: A case study, p In M.L. Wiley (ed.), Estuarine Interactions. Academic Press, New York