SAN FRANCISCO BAY MODELING FOR THE ENVIRONMENTAL FATE AND TRANSPORT OF COPPER FROM VEHICLE BRAKE PAD WEAR DEBRIS WORK PLAN.

Transcription

1 SAN FRANCISCO BAY MODELING FOR THE ENVIRONMENTAL FATE AND TRANSPORT OF COPPER FROM VEHICLE BRAKE PAD WEAR DEBRIS WORK PLAN Submitted to Sustainable Conservation c/o Sarah Connick 121 Second Street, Sixth Floor San Francisco, CA by URS 1333 Broadway, Suite 800 Oakland, CA Date 8/15/2006

2 TABLE OF CONTENTS SECTION 1.0: INTRODUCTION AND OBJECTIVES Background...1 SECTION 2.0: MODELING APPROACH Short-Term Modeling Inputs to Hydrodynamic Module Inputs to Heavy Metals Module Model Calibration and Verification Sensitivity Analysis Long-Term Modeling Inputs to Box Model Validation Uncertainty Analysis Model Output...15 SECTION 3.0: REPORT, COORDINATION, AND MEETINGS...17 SECTION 4.0: SCHEDULE...18 SECTION 5.0: BUDGET...19 SECTION 6.0: REFERENCES...20 LIST OF TABLES Table 2-1 Table 2-2 Table 2-2 Table 2-3 Table 2-4 Table 2-5 Table 4-1 Table 5-1 Mass Flux Rates for Lower South San Francisco Bay Mass Flux Rates for South and Lower South San Francisco Bay Input Data for the MIKE 21 Hydrodynamics Module Input Data for the MIKE 21 Heavy Metals Module Sediment Calibration Parameters for the MIKE 21 ME Module Copper Calibration Parameters for the MIKE 21 ME Module Project Schedule Project Budget LIST OF FIGURES Figure 2-1 Conceptual model of biogeochemical processes affecting copper in the Bay Figure 2-2 Summary of Processes Modeled by MIKE 21 ME Figure 2-3 MIKE 21 Model Domain and Bathymetry of San Francisco Bay Figure 2-4 Average RMP Benthic Copper Concentrations and Data Grouping for MIKE 21 Model Figure 2-5 Initial Benthic Copper Concentrations (mg/kg) used in MIKE 21 Water Quality Modeling i

3 SECTION 1.0: INTRODUCTION AND OBJECTIVES The Brake Pad Partnership (BPP) has embarked on a study to better understand the fate and transport of copper from brake pad wear debris (BPWD) in the environment. Studies include characterization and analysis of the physical and chemical properties of BPWD, air and stormwater monitoring, and environmental modeling. Three environmental modeling studies are being performed including air deposition modeling, watershed modeling, and modeling of the San Francisco Bay (Bay). The air deposition and watershed modeling provide results that will be used as inputs to the model of the San Francisco Bay so that the effect of copper loading on the Bay can be determined. Even though recent studies have shown that current levels of copper in the Bay do not appear to be impairing its beneficial uses, there is still a need to ensure that water quality will not deteriorate in the future. Determining how current levels of copper loadings affect the concentrations of copper in the Bay sediments and water column provides a baseline for comparing against any increases or decreases in copper loadings that could occur in the future. The air deposition and watershed modeling will provide estimates of the quantity of copper entering the bay that originates from BPWD. By running model scenarios with and without the contributions of copper from the BPWD, the effects of BPWD on the dissolved and benthic copper concentrations in the Bay can be determined. 1.1 Background The Bay is a geographically and tidally complex system characterized by broad shoals and narrow channels. It can be considered as consisting of two systems. The North Bay system extends from the Golden Gate through the Central Bay, San Pablo Bay, Carquinez Strait, and Suisun Bay, ending at the Sacramento/ San Joaquin River Delta. The South Bay system extends from the Central Bay (starting about at the Bay Bridge) south to Guadalupe River and Coyote Creek near San Jose. Many factors affect flows in the Bay, but water depth is the most important factor controlling the spatial variability of both magnitude and direction of currents (Cheng et al. 1993). Ninety percent of the freshwater inflow to the Bay comes from the Delta (Cheng et al. 1993) and flows through the northern portion of the Bay resulting in a well mixed to partially mixed estuary (Walters et al. 1985; Uncles and Peterson 1995). The South Bay can be described as a tidal lagoon (Cheng et al. 1993; Uncles and Peterson 1995). Very little freshwater flows into the South Bay and salinities and flow properties are, to a great extent, controlled by the exchange of water with the Central Bay, resulting in a well mixed estuary. Salinity is uniform and mostly controlled by conditions in the Central Bay. Salinity is an important physical parameter for Bay circulation and ecology. Salinity variability is greatest in the North Bay where freshwater from the Delta enters the Bay. However, Delta outflow can also affect salinity in the South Bay (McCulloch et al. 1970). 1

4 Other factors that affect salinity in the South Bay include tidal variations, spring-neap variations, gravitational circulation, and freshwater inflow (Smith and Cheng 1987, 1991; Ford et al. 1990). Sediment originating in the Delta is the largest source of sediment load entering the Bay and comprises about 86 percent of the total fluvial sediment supply (Krone 1979). The remaining sediment is contributed from local tributaries. Most of the sediments entering the Bay in suspension are silts and clays. Silts and clays (cohesive sediment) are dominant in the South Bay and most parts of the Central and North Bays. Erosion, deposition, transport, and bed sediment processes are the four major processes controlling the dynamics of cohesive sediment in the Bay. Previous studies have shown that tidal currents, Delta outflow, and advective transport are the most important elements driving the four processes (Schoellhamer 1996; McDonald and Cheng 1994). In the North Bay, sediment transport is dominated by sediment input from the Delta and the outflux through the Golden Gate. The spring-neap tide cycle is an important factor affecting SSC and subsequent transport throughout the Bay. In the shallow sections of the Bay, wind and wind waves are also important. In the channels, tidal currents are an important factor for sediment transport (URS 2003). Regional water quality in the Bay can be characterized by a concentration gradient, with highest concentrations in the South Bay and southern sloughs, and lowest concentrations in the Central Bay due to tidal mixing and flushing. Chemicals of greatest concern in the Central, South, and Lower South Bays include dioxin and furans, mercury, dioxin-like PCBs, copper, nickel and diazinon (URS 2003). 2

5 SECTION 2.0: MODELING APPROACH A conceptual biogeochemical model of the processes affecting the fate and transport of copper in the Bay is shown in Figure 2-1. In addition to copper loads from air deposition and watershed runoff, loads from wastewater discharges, loads from commercial vessels (if available) and marinas where copper anti-fouling paints are used will be included. Figure 2-1. Conceptual model of biogeochemical processes affecting copper in the Bay In order to model the long-term effects of changing the copper loadings to the Bay, two models will be developed. The short-term model will define the Bay using a finite element grid at a 200-meter resolution and will provide hourly, 4-day average, or monthly average copper concentrations. Results from the short-term model will be used as inputs to the long-term model, which will define the Bay using a series of boxes at a more regional scale and will show decadal trends in copper concentrations. 3

6 2.1 Short-Term Modeling Bay Modeling Workplan The MIKE 21 model developed by the Danish Hydraulic Institute (DHI) will be used to simulate the hydrodynamics, sediment transport, and water quality of the Bay. MIKE 21 is a two-dimensional, free-surface flow modeling system. It is used to simulate hydraulics and hydraulics-related phenomena in estuaries, coastal waters, and seas where stratification can be neglected. It consists of a hydrodynamic module to which other modules can be added to address different phenomena. For the proposed copper modeling, the heavy metals module (MIKE 21 ME) will be added to the hydrodynamics module (MIKE 21 HD). In MIKE 21 ME, four processes control dissolved metal concentrations: (1) adsorption/desorption, (2) sedimentation/resuspension, (3) porewater diffusion, and (4) advection/dispersion. The first three processes are shown in Figure 2-2. Figure 2-2. Summary of Processes Modeled by MIKE 21 ME The first process, adsorption/desorption, is the essential link between metals in the aqueous and solid phases. Adsorption is the process by which dissolved metal ions or complexes attach themselves to the surface of particulate matter. Desorption is the detachment of the metal from the particulate surface and its return to the dissolved state. In MKE 21 ME, adsorption also includes the processes of absorption (incorporation of the metal into the solid phase) and surface precipitation. The second important process is resuspension and sedimentation of particulate matter within the Bay. Resuspension and sedimentation are influenced by the wind and currents. The third process illustrated in Figure 2-2 is diffusion of metal-enriched water between the ambient Bay and porewater, depending on the concentration gradient. Diffusion rates are dependent on the chemical environment in the pore spaces of the sediment, which may be significantly different than the overlying water. Adsorption and desorption rate constants are used to represent chemical transformation in the porewater. 4

7 The final process incorporated into the ME module is advection and dispersion of dissolved and adsorbed metals. Both dissolved and adsorbed metals are treated identically in the ME module, and parameters calculated in the simultaneous run of MIKE 21 HD govern the advection and dispersion. Attempts have been made to quantify the mass rate of copper inputs to the South Bay and the Lower South Bay (south of the Dumbarton Bridge). The Conceptual Model Report for Copper and Nickel in Lower South San Francisco Bay (Tetra Tech, 1999) provides estimates of total and dissolved copper fluxes for wet and dry seasons. A Kinetic Model of Copper Cycling in San Francisco Bay (Bessinger, et. al., 2006) provides estimates of dissolved copper fluxes for wet and dry seasons. The results from these studies are provided in Tables 2-1, and 2-2. Table 2-1 Mass Flux Rates for Lower South San Francisco Bay Total Copper Flux Rates Dissolved Copper Flux Rates Dry Season Wet Season Dry Season Wet Season kg/ dry season kg/ wet season kg/ dry season kg/ wet season Tributaries 160 3, Point Sources (POTW Inflow) Atmospheric Diffusive Bed-Particulate 7,101 5, Net Flux Past Dumbarton Bridge -7,931-6,753-1, Flux Balance / Net Change 0 3, Source: Tetra Tech, 1999 Table 2-2 Mass Flux Rates for the South and Lower South San Francisco Bay Dissolved Copper Mass Flux Rates 1997 Dry Season 1996 Wet Season Bay Segment (kg/yr) (kg/yr) South Bay Tributary Inflow ,000 POTW Inflow 3,000 8,000 Desorption 900-3,000 Diffusion 8,000 8,000 Net Central Bay Exchange -8,000-2,000 Net Flux Past Dumbarton Bridge Flux Balance / Net Change 5,000 20,000 Lower South Bay Tributary Inflow 0 4,000 POTW Inflow 1,000 1,000 Desorption Diffusion Net Flux Past Dumbarton Bridge Flux Balance / Net Change 1,000 5,000 Source: Bessinger, et. al.,

8 The values in Table 2-1 are based on copper concentrations from grab samples of the water and sediment, TSS measurements, measured and estimated external copper loads, and estimates of the volume and flushing time of the Lower South Bay. The values in Table 2-2 are based on results of a two-dimensional hydrodynamic and water quality model for specific water years and time periods. Results from both studies show that in the Lower South Bay, tributaries and publicly owned treatment works (POTWs) provide a large component of the dissolved copper load. Table 2-1 shows that the main component contributing to the total copper load in the Lower South Bay is particulate matter from the bed. Table 2-2 shows that in addition to the non-point and point sources, diffusion from the sediment porewater contributes a large portion of the dissolved copper to the South Bay north of the Dumbarton Bridge. Both studies indicate that the error in the mass balance is of approximately the same order of magnitude as the mass input components. This provides an estimate of the uncertainty in the values of the inputs, and also shows that there may be some processes for copper cycling that have not been included Inputs to Hydrodynamic Module The inputs to the hydrodynamics module include the bathymetric grid, wind, freshwater sources, and boundary conditions (tides in the Pacific Ocean and Delta inflow). Descriptions of these inputs and the source of the data are included in Table 2-3. The model domain and bathymetry are shown in Figure 2-3. Calibration parameters include the bottom roughness coefficients and the eddy viscosity. The Manning s n values range from 0.01 to The eddy viscosity is determined using the Smagorinsky formula based on velocity, with the Smagorinsky coefficient ranging from 0.05 to 0.5. Table 2-3 Input Data for the MIKE 21 Hydrodynamics Module Input Data Description Data Source Made up of a combination of grids: 30-m resolution NOAA Digital Elevation Model (DEM) for NOAA/NOS Bathymetric Grid Wind San Francisco Bay (most recent survey was from 1993) 25-m resolution grid for South San Francisco Bay below Dumbarton Bridge 100-m resolution San Francisco Bay grid grid created by URS from NOAA soundings in the Pacific Ocean Resulting grid used the UTM Zone 10 NAD 1927 projection, rotated degrees counterclockwise from North to align flow direction with the model coordinate system in the South Bay and in San Pablo Bay. The Delta was modeled as a "box" with surface area equal to 61,000 acres, an average depth of 5.1 m (16.7 ft), and volume of 1,145,000 acre-feet. Wind time series collected by the National Climatic Data Center (NCDC) at SFO at hourly intervals was applied to the entire Bay Smith and Cheng 1994 Cheng and Smith 1998 DWR 1995 NCDC wind data recorded at SFO 6

9 Input Data Description Data Source Daily averaged flows for Alameda Creek, Coyote Creek, Guadalupe River, Napa River, and Petaluma River were USGS gage data obtained from USGS gaging stations. For streams with no and watershed Freshwater Sources measured flows available, results of correlation analysis to area/ runoff available data were used to estimate runoff based on rainfall correlation and watershed characteristics. Smaller tributaries and streamflow downstream of gaging stations were omitted. Flow Boundary Conditions Tidal Boundary Conditions Delta outflow (provides 90% of freshwater to Bay) specified as the average daily flow from DWR's DAYFLOW program. Applied recorded time series of water surface elevations from the NOAA tide station at Point Reyes to ocean boundary located approx. 15 km outside of the Golden Gate Bridge. For periods without data at Point Reyes, water levels from the NOAA tide station at Monterey were substituted (without correction, even though Pt. Reyes lags Monterey by about 20 min, and the high water elevations were about 0.12 m lower). DWR's DAYFLOW NOAA (Primarily Pt. Reyes tide station with Monterey as a backup) Inputs to Heavy Metals Module The inputs to the heavy metals module include the initial benthic sediment concentrations, point source (POTW and Industrial Wastewater) inputs, tributary loads, loads from marinas, loads from commercial vessels (if available), wet and dry air deposition, and boundary conditions at the Pacific Ocean and Delta. Descriptions of these inputs and the previously used (URS 2003) and proposed source of the data for the BPP modeling are included in Table 2-4. The initial water column adsorbed copper concentrations were calculated by multiplying the initial suspended sediment concentration (SSC) field by the initial benthic sediment field determined from the San Francisco Estuary Regional Monitoring Program (RMP) annual reports (SFEI ). The RMP data are shown in Figure 2-4. The initial benthic sediment fields are shown in Figure 2-5. Initial dissolved concentrations were then calculated by assuming that the dissolved phase is in equilibrium with the adsorbed phase. The atmospheric deposition will be included in the MIKE 21 model by applying precipitation over the model domain, with values that can vary over time. It will be assumed that the copper in precipitation is all in the dissolved state. The MIKE 21 model does not have the capability to specify dry deposition of copper, so a very small quantity of precipitation will be applied for dry deposition, and the actual precipitation from the air quality modeling results will be used for wet deposition. Calibration parameters include: the settling velocity, critical velocity, and resuspension rate of the sediment; copper adsorption and desorption rate constants in the water column and porewater; and the initial active sediment layer thickness. The wet and dry season ranges of settling velocities, critical velocities, and resuspension rates are given in Table 2-5. The adsorption and desorption rate constants previously used are given in Table 2-6. The initial 7

10 C:\Projects\EBDA\data\bathy\mer200m_26c.dfs2 C:\Projects\EBDA\data\bathy\mer200m_26c.dfs (kilometer) (kilometer) Elevation (m, (m, NGVD) Above Above Below Below San Francisco Bay Modeling Work Plan MIKE 21 Model Domain and Bathymetry of San Francisco Bay FIGURE 2-3

11

12 San Francisco Bay Modeling Work Plan Initial Benthic Copper Concentrations (mg/kg) used in MIKE 21 Water Quality Modeling FIGURE 2-5

13 approximations of the adsorption and desorption rates were based on site-specific data and laboratory studies conducted on samples collected in the Bay. The initial active sediment layer previously used was 0.1 meters thick. Input Data Initial Benthic Sediment Concentrations Initial Suspended Sediment Concentration Point Source Input Description Table 2-4 Input Data for the MIKE 21 Heavy Metals Module Current Data Source Average benthic sediment concentrations from the RMP Annual Reports (SFEI , 1998, 1999) were assigned to the Delta, Suisun Bay, San Pablo Bay, Central Bay, South Bay, and Lower South Bay. The North and Central Bays were also subdivided into shoals and channels. Toxic hotspots identified by BPTCP (SFBRWBCB 1999) were overlain on existing sediment fields. Hotspot was assigned to single 200 m x 200 m grid cell of sample. A constant initial SSC of 10 mg/l was used. The effects of initial SSC dissipate quickly as local processes of erosion and deposition dominate. Following a spin-up time of several days, the model prediction will be independent of the initial conditions. 37 industrial and wastewater treatment plants (POTW sources) were included. Daily average flows and concentrations were input when electronic data were available, otherwise monthly averages were used (from 1996 and 1997 NPDES self-monitoring reports). Metals were assumed to be in the dissolved state since SSCs were generally low. SFEI , 1998, 1999 SFBRWQCB 1999 NPDES selfmonitoring reports Revised Data Source for BPP More recent data will be included, if available. Will incorporate more recent NPDES selfmonitoring reports for larger dischargers. Atmospheric Deposition Wet and dry deposition rate for copper (ug/m 2 -day) Not used Air Basin Modeling (AER future date) 11

14 Input Data Tributary Loads Ocean and Delta Boundaries Description 69 watersheds were delineated. Flows were based on USGS gage data, either measured or normalized by watershed area. Different SSC and metal concentrations were used for low and high flows. High flows were classified as greater than twice the July and August base flow for a given year. SSC and metal concentrations for base flows were obtained from BASMAA (1996) data. Storm event concentrations were obtained from a land-use summary of this data (Daum and Davis 2000). Dissolved and adsorbed metal concentrations specified at each boundary. Distribution coefficients and metal concentrations on suspended sediment for the Delta boundary were derived from the average of all measurements from the San Joaquin and Sacramento RMP monitoring stations. Distribution coefficients for the Pacific Ocean boundary were derived from the average of all measurements from the Golden Gate station. Adsorbed and dissolved concentrations for the Pacific Ocean were calculated from the relationship between the distribution coefficient and the adsorbed, dissolved, and total water-column metal concentrations measured in the northern Pacific Ocean. Current Data Source USGS gage data and watershed area/ runoff correlation, BASMAA 1996, Daum and Davis 2000 Calculated from equilibrium distribution coefficients, either adsorbed metal concentrations on the suspended sediment or total water column concentrations, and total SSCs (set at 20 mg/l at the Pacific Ocean and as a time series at the Delta). SFEI , 1998, 1999 Burton and Statham 1990 Bay Modeling Workplan Revised Data Source for BPP BASINS Model (EPA future date) will provide daily flow, SSC, and total copper concentrations for each tributary source. Optional: use leaching data to develop fraction of total copper discharged that may be solubilized 12

15 Sediment Parameter Settling Velocity Critical Velocity Resuspension Rate Table 2-5 Sediment Calibration Parameters for the MIKE 21 ME Module Bay Modeling Workplan Notes Min Value Max Value Units Spatially varying w/ lowest values in the mudflats near Dumbarton Bridge and highest values in the deepest water of the Golden Gate. Wet season values are the same as dry. Spatially varying w/ lowest values in the mudflats south of the Dumbarton Bridge and highest values near the Golden Gate. Wet season values are the same as dry except near the Golden Gate. Wet season values near the Golden Gate are approx. 10 to 15% less than dry season. Spatially varying w/ lowest values near the Golden Gate and highest values in the San Pablo, Suisun, and Honker Bay mudflats. Dry season values are generally larger than wet season values m/d m/s g/m2/d Table 2-6 Copper Calibration Parameters for the MIKE 21 ME Module Ads./Des. Material Dry Season Wet Season Units Rate Constant Value (if different from Dry) Adsorption Copper in North Bay Water /d/(mg/L) Adsorption Copper in Central Bay Water /d/(mg/L) Adsorption Copper in South Bay Water /d/(mg/L) Adsorption Copper in Lower South Bay /d/(mg/L) Water Adsorption Copper in Porewater 10 1/d/(mg/L) Desorption Copper in Bay Water /d Desorption Copper in Porewater /d Model Calibration and Verification The MIKE 21 hydrodynamic model was previously calibrated by URS (URS 2003) and will not be recalibrated as part of this study. In the previous hydrodynamic model calibration, model parameters were adjusted until the model reproduced observed current velocities and water levels. Two parameters were adjusted during calibration: Roughness coefficient used in the bottom friction formulation; Eddy viscosity parameterizes horizontal mixing of momentum. 13

16 The MIKE 21 heavy metals model was also calibrated previously (URS 2003), however there have been recent data collected by the Clean Estuary Partnership, as well as more data collected by the San Francisco Estuary Institute (SFEI) as part of their Regional Monitoring Program (RMP). Also, air deposition had not been included in the previous modeling, and the watershed inputs will also be changing slightly. The previous MIKE 21 ME calibration will be verified against the additional measured data, and will be recalibrated if necessary. If possible, calibration parameters (settling velocity, critical velocity for erosion, and resuspension rate of sediment, as well as adsorption and desorption rate constants for copper) will be chosen such that they remain constant yearround, rather than changing seasonally. However, this will depend on how well modeled concentrations match the measured data. Separate years of observed data will be used for calibration and subsequent verification Sensitivity Analysis The sensitivity analyses conducted previously (URS 2003) will be summarized. An analysis of model sensitivity to major parameters not addressed previously will be conducted, which will include the sensitivity to the initial concentration of benthic sediments. 2.2 Long-Term Modeling A compartmental water quality model will be selected for use in modeling the long-term impacts of copper from BPWD. Possible models include WASP, developed by EPA; ECOLab, developed by DHI; or the multi-box model developed by SFEI. SFEI is in the process of incorporating a sediment model developed by USGS into their multi-box model, however this updated version may not be ready in time to be used by URS for this study. URS will prepare a memo outlining the rationale behind the selection of the preferred longterm model prior to the completion of the short-term modeling. Regardless of the choice of model, the compartmental model will operate on the principle of conservation of mass. The water volume and total mass of copper would be accounted for over the duration of the simulation. The initial conditions and flows at the boundaries of each compartment would be based on output from the MIKE 21 model. The MIKE 21 model would give the flux of water, suspended sediment, and the mass of copper across each of the boundaries used for each compartment. The initial concentrations within each compartment would be based on the average concentration within that region of the MIKE 21 model Inputs to Box Model Inputs to the box model will include geometry, initial conditions, boundary conditions, time-series of external loads, physical properties of sediment, the transport scheme and kinetic parameters. The geometry will be described using a number of segments and defining the volume and type of each segment. Typical types of segments that could be included would be: 14

17 Surface water segment has an interface with the atmosphere; Sub-surface water segment water segment without atmospheric or benthic interface; Surface benthic segment benthic segment with a water interface Sub-surface benthic segment defines all benthic segments below the surface benthic segment The initial conditions within each segment will include the average initial benthic copper concentrations, initial suspended sediment concentrations, and initial total copper concentrations and the fraction dissolved. These initial concentrations will be based on the average concentrations in the MIKE 21 model within regions corresponding to each segment. The boundary conditions will be the same as the MIKE 21 model, except that they will be averaged over a longer time step, and will depend on the computational time step used in the model. The external tributary and atmospheric loads defined by the air deposition and watershed models will be applied to the corresponding segment in the compartmental model. The sediment parameters such as settling velocity, critical velocity for erosion, and resuspension rate will be based on the average values determined in the recalibration of the MIKE 21 model. The methods and parameters needed to calculate erosion and deposition may vary depending on the choice of model. The parameters describing the transport scheme may also vary depending on the choice of model. Parameters will need to define advection and dispersion, as well as diffusion in the pore water Validation The box model will be verified by comparing the results from the same period modeled with MIKE 21. Parameters used in the box model will be adjusted until results show agreement with the calibrated MIKE 21 model Uncertainty Analysis The uncertainty in the model output will be estimated by performing additional model runs using a range of values that bound the likely sediment transport and water quality parameters. 2.3 Model Output Output from the model verification runs will be provided to show how well modeled dissolved copper concentrations correspond to measured concentrations throughout the Bay. Results from the modeled scenarios will be focused on the South San Francisco Bay. 15

18 Output will include plots showing the baseline benthic and dissolved copper concentrations as well as the differences between the modeled scenarios (either increasing or decreasing copper input loads) and the baseline concentrations. Changes between the modeled scenarios and the baseline dissolved copper concentrations will be shown as differences between the mean and maximum 4-day average concentrations over different water years. Changes between the modeled scenarios and the baseline benthic copper concentrations will be shown as differences between the mean and maximum 30-day average concentrations. Output from the long-term modeling will be used to show decadal trends comparing the seasonal average concentrations of the modeled scenarios to the baseline. The long-term model will be used to estimate the rate at which the Bay reaches equilibrium with different levels of copper in the input loads. 16

19 SECTION 3.0: REPORT, COORDINATION, AND MEETINGS Coordination with the air and watershed modeling teams will be required prior to receiving the final results. This will ensure that the output from those models will provide the necessary input to the Bay model. URS will work with the project team to help develop appropriate input information for the Bay models, such as a method to extrapolate copper loads from the Castro Valley study watershed to estimate copper loads in storm water runoff from other watersheds surrounding the Bay. URS will meet with the BPP Steering Committee and Scientific Advisory Team members as needed to review progress, interpret results, and address any issues that arise. Meetings may be by teleconference or in person. Meetings will be at least quarterly but may be more frequent. URS will also give presentations on the project at BPP stakeholder meetings and provide information to support implementation of the BPP s Stakeholder Communication Plan. URS will provide brief written monthly progress reports to the BPP and stakeholders via and brief quarterly verbal updates to the BPP Steering Committee. URS will prepare a memo outlining the rationale behind the selection of the preferred long-term model within two months of starting the re-calibration of the short-term model. The memo will discuss the pros and cons of potentially viable models and approaches. Both draft and final reports summarizing the Bay modeling results and including all input data will be prepared. The draft report will be completed by August 15, 2006, provided that final report for the watershed modeling is completed on schedule (by April 14, 2006). The final report will be submitted four weeks after receiving all comments from the BPP. Deliverables shall be provided in hard copy and Adobe Acrobat and Microsoft Word electronic formats. As appropriate, numerical results and data shall be provided in Microsoft Excel. 17

20 SECTION 4.0: SCHEDULE Table 4-1 provides a schedule of tasks to complete this project and the estimated completion dates. Table 4-1 Project Schedule Milestone Weeks After Task Milestone Start date (upon receipt of final report for air deposition modeling and draft report for watershed modeling) Task 1 Re-calibrate MIKE 21 copper model and perform verification runs using results from final report for air deposition modeling (due Oct 14, 2005) and results from draft report for watershed modeling (due Dec 20, 2005) Task 2 Prepare memo outlining rationale for selection of long-term model. Task 3 Run scenarios for preliminary short-term modeling Task 4 Develop box model 11 Task 5 Run scenarios for preliminary long-term modeling 14 Receive final report for watershed modeling 0 Task 6 Perform final calibration and verification runs after receiving final report for watershed modeling (due Apr 14, 2006) 7 Task 7 Sensitivity study and analysis of uncertainty 10 Task 8 Complete final scenario runs. 13 Task 9 First draft of report 17 Receive all comments from BPP on draft Bay modeling report 0 Task 10 Final report 4 Task 11 Coordination, meetings As needed 18

21 SECTION 5.0: BUDGET Table 5-1 provides an estimate of the budgeted costs of performing the tasks listed in Section 4.0. Task Table 5-1 Project Budget Budget Task 1 Re-calibrate MIKE 21 copper model and perform verification runs using results from final report for air deposition modeling and results from draft report for watershed modeling Task 2 -- Prepare memo outlining rationale for selection of long-term model. Task 3 Run scenarios for preliminary short-term modeling $10,200 $5,000 $5,500 Task 4 Develop box model $10,000 Task 5 Run scenarios for preliminary long-term modeling Task 6 Perform final calibration and verification runs after receiving final report for watershed modeling $8,000 $14,955 Task 7 Sensitivity study and analysis of uncertainty $8,500 Task 8 Complete final scenario runs. $10,000 Task 9 First draft of report $17,772 Task 10 Final report $11,850 Task 11 Future coordination, meetings $5,000 Total: $106,777 19

22 SECTION 6.0: REFERENCES Atmospheric and Environmental Research (AER). (under development). Air Deposition Modeling of Copper from Brake Pad Wear Debris. Sustainable Conservation, San Francisco, CA. Bay Area Stormwater Management Agencies Association (BASMAA) San Francisco Bay Area Stormwater Runoff Monitoring Data Analysis Prepared by URS. Bessinger, Brad, et. al A Kinetic Model of Copper Cycling in San Francisco Bay. San Francisco Estuary and Watershed Science. Vol. 4, Issue 1. February. Burton, J.D., and P.J. Statham Trace metals in seawater. In Heavy Metals in the Marine Environment, R.W. Furness and P.S. Rainbow, eds., pp Boca Raton, FL: CRC Press. California Department of Water Resources (DWR) Delta Atlas. Cheng, R.T. and R.E. Smith A nowcast model for tides and tidal currents in San Francisco Bay, California. Ocean Community Conference 98, Proceedings, Baltimore, MD, Marine Technology Society, Nov , 1998, Cheng, R.T., V. Casulli, and J.W. Gartner Tidal residual intertidal mudflat (TRIM) model and its applications to San Francisco Bay, California. Estuarine, Coastal, and Shelf Science 369: Daum, T., and J.A. Davis Coastal Watershed Mass Loading Report. San Francisco Estuary Institute. Ford, M., J. Wang, and R.T. Cheng Predicting the vertical structure of tidal current and salinity in San Francisco Bay, California. Water Resources Res. 26: Krone, R.B Sedimentation in the San Francisco System. In San Francisco Bay: The Urbanized Estuary, pp Pacific Division of the American Association for the Advancement of Science, San Francisco, CA. McCulloch, D.S., D.H. Peterson, P.R. Carlson, and T.J. Conomos A preliminary study of the effects of water circulation in the San Francisco Bay estuary some effects of freshwater inflow on the flushing of South San Francisco Bay. U.S. Geological Survey Circulation 637-A. 20

23 McDonald, E.T. and R.T. Cheng Issues related to modeling the transport of suspended sediments in northern San Francisco Bay, California. Proceedings of the Third International Conference on Estuarine and Coastal Modeling, September, ASCE, Chicago, IL, pp National Oceanic and Atmospheric Administration/ National Ocean Service (NOAA/NOS) Bathymetric data for San Francisco Bay, 7.5 (30m) quads. ARCINFO data file downloaded from San Francisco Bay Regional Water Quality Control Board (SFBRWQCB) Bay Protection and Toxic Cleanup Program. Final Regional Toxic Hot Spot Cleanup Plan. March. San Francisco Estuary Institute (SFEI) Annual Report: San Francisco Estuary Regional Monitoring Program for Trace Substances. Richmond, CA. Schoellhamer, D.H Factors affecting suspended-solids concentrations in South San Francisco Bay, California. Journal of Geophysical Research 101: (C5): SFEI Annual Report: San Francisco Estuary Regional Monitoring Program for Trace Substances. Richmond, CA. SFEI Annual Report: San Francisco Estuary Regional Monitoring Program for Trace Substances. Richmond, CA. SFEI Annual Report: Regional Monitoring Program for Trace Substances. Richmond, CA. December. SFEI Annual Report: San Francisco Estuary Regional Monitoring Program for Trace Substances. Richmond, CA. SFEI Annual Report: San Francisco Estuary Regional Monitoring Program for Trace Substances. Richmond, CA. Smith, L.H. and R.T. Cheng Tidal and tidally averaged circulation characteristics of Suisun Bay, California. Water Resources Res. 23: Smith, R.E., and R.T. Cheng Gravitational circulation in a tidal strait. In Hydraulic Engineering 1991, Conference Proceedings, Nashville, TN, July 29-August 2, pp WW Division of the American Society of Civil Engineers. 21

24 Smith, R.E., and R.T. Cheng Defining intertidal bathymetry of the southern extreme of South San Francisco Bay. U.S. Geological Survey Report. Menlo Park, CA. Tetra Tech Conceptual Model Report for Copper and Nickel in Lower South San Francisco Bay. San Jose, CA. December. Uncles, R.J. and D.H. Peterson A computer model of long-term salinity in San Francisco Bay: sensitivity to mixing and inflows. Environment International 21 (No. 5): United States Environmental Protection Agency (EPA). (under development). Watershed Modeling for the Environmental Fate and Transport of Copper from Vehicle Brake Pad Wear Debris. Sustainable Conservation. San Francisco, CA. URS Proposed SFO Runway Reconfiguration Project, Predicted Changes in Hydrodynamics, Sediment Transport, Water Quality, and Aquatic Biotic Communities Associated with SFO Runway Reconfiguration Alternatives BX-6, A3, and BX-R, Final Technical Report. Oakland, CA. June. Walters, R.A., R.T. Cheng, and T.J. Conomos Time scales of circulation and mixing processes of San Francisco Bay waters. Hydrobiologia 129: