Integration of Stormwater Runoff and Pollutant Model with BMP Assessment Model Using ArcView GIS

Stormwater management such as flood and erosion protection is a major concern for decision makers in watershed planning and management. Best management practices (BMPs) have been widely used as cost-effective mechanisms to solve stormwater runoff and associated water quality problems. BMPs also serve the purpose of preventing flooding, reducing soil loss, and providing water retention (Chen et al., 1995). To evaluate the effectiveness of stormwater management plans, a user-friendly stormwater modeling and BMP modeling tool is needed. Such a tool can be used to simulate existing stormwater runoff and its pollutant loads, assess water quality impacts on existing watersheds, and evaluate the performance of BMPs. This tool should be able to project future stormwater quantity and quality in drainage systems, identify key design parameters to improve BMP pollutant removal efficiencies, and evaluate stormwater management alternatives. Watershed managers and planners often need a tool that can be easily operated and can produce easily-interpreted model simulation results.

IGM SYSTEM DEVELOPMENT

The IGM system consists of six key components: (1) GIS interface; (2) pre-processor; (3) run models; (4) post-processor; (5) the SRPM model; and (6) the BMPAM model. Figure 1 presents the conceptual design of the IGM system and illustrates the linkages among the key components. As a major functional component, the GIS interface was linked with all other key components in the IGM system. Users can run SRPM or BMPAM once model inputs have been entered through the pre-processor component. The outputs from the SRPM model are fed into the BMPAM model. Simulation outputs from both models can be viewed via the post-processor component in tabular and graphic formats. Input files generated from the pre-processor, model executable files, and simulation outputs are stored in the IGM system and can be reused or modified through the GIS user interface or other UNIX/DOS operating system utilities.

The Stormwater Model menu was designed to obtain model inputs from users, save the inputs to the format required by SRPM, and run the SRPM model. The SRPM model, a simplified version of the Storm Water Management Model (SWMM) (Huber et al. 1987), is a field-scale watershed model which can be applied in urban or agricultural areas to predict stormwater runoff and associated pollutant concentrations (Xue, 1996a). The SRPM model can run continuously at an hourly time step for a storm event or for longer time periods such as five or ten years. SRPM simulates up to nine water quality constituents in the following categories: biological oxygen demand (BOD); total suspended solids (TSS); total nitrogen (TN); total phosphorus (TP); heavy metals such as zinc (Zn), lead (Pb), and copper (Cu); pesticides and other non-volatile organics; and non-reactive tracers. It generates hourly, daily, monthly, and annual runoff and the associated pollutant concentrations during simulation. SRPM has been successfully applied to two Lake Okeechobee watersheds in central Florida (Xue and Zhang, 1996; 1997).

The Stormwater Model menu has the following eight menu items: (1) Select County and Site; (2) Save the Site Selection; (3) Input Site Parameters; (4) Input Phosphorus Transport Data; (5) Input Hydrology Data; (6) Input Pollutant Parameters; (7) Save SRPM Inputs; and (8) Run SRPM (Figure 2 ). The first menu item enables users to select a county and site(s) within the county of interest from the SFWMD's land use coverages. Parameters in the land use coverage include surface area, county name, and land use type. The coverages stored in a District database and were used by other projects conducted at the District. The second menu item lets users save the land use acreage data of selected site(s) in the IGM system. One or more sites can be selected for simulation as long as the primary land use type is the same. Once the site(s) are selected, users input physical parameters of the site(s) from the third menu item, including width and slope of overland flow and infiltration parameters. Phosphorus transport data, required only for phosphorus simulation in agriculture areas, are entered from the next menu item. Monthly hydrologic data as well as water quality parameters are input from the Input Hydrology Data and Input Pollutant Parameters menu items, respectively. The next menu item saves all previous selections and entered data. Meteorological data such as precipitation and evaporation are pre-stored in the IGM system. The last menu item is Run SRPM which allows users to run the SRPM model in the system.

The second menu in the IGM system is BMP Model which was designed to obtain model inputs, save the inputs for BMPAM, and run the BMPAM model. BMPAM is a continuous simulation water quality model that routes runoff and associated pollutant loads generated by a watershed model to a stormwater control BMP. It simulates runoff and pollutant removal processes in stormwater treatment devices for both urban and agricultural areas. The mechanisms used for simulating water movement and pollutant reduction include evaporation, evapotranspiration, water storage, infiltration, overflow/outlet flow, settling of suspended solids, first order kinetics for biological decay processes, adsorption, partitioning of metals between soils and water, and plant uptake of nutrients (Xue, 1996b). As with the SRPM model, BMPAM simulates up to nine water quality constituents.

The BMP Model menu has ten menu items: (1) Select a BMP; (2) Input Hydraulic Parameters; (3) Input Infiltration Parameters; (4) Input Initial Concentrations; (5) Input Pollutant Decay Rates; (6) Input Particle Parameters; (7) Input Microphyte/Macrophyte Coefficients; (8) Input Monthly Plant Coefficients; (9) Save BMPAM Input; and (10) Run BMPAM (Figure 3 ). The first menu item enables users to select the type of BMP to be simulated such as detention basin, retention device, and vegetative control. Users input physical and hydraulic parameters for the selected BMP from the second menu item: (1) depth and surface area; (2) initial water elevation; (3) water elevation above weir; (4) weir coefficient and width; and (5) size of the upstream drainage area.

BMPAM can simulate upto three infiltration layers (i.e., top, middle and natural infiltration layers) at the bottom of a BMP to account for different soil type. Infiltration parameters for each layer within a BMP such as maximum and minimum infiltration rate, a decay parameter, infiltration regeneration ratio, area, depth, and porosity are input in the third menu item. Initial pollutant concentration and decay rates are input in the fourth and fifth menu items, respectively. For simulation of particles/solids, users can specify up to five particle classes and enter percentage, diameter, and specific gravity of each class in the Input Particle Parameters menu item. To simulate wetlands and other vegetative control practices, users need to provide microphyte and macrophyte information. In the seventh menu item, required input data are monthly coefficients for plant uptake of nutrients. The next menu item saves all entered selections and data. The last menu item will run the BMPAM model in the IGM system.

Seven menu items are contained in the Simulation Outputs menu: (1) Select One Parameter; (2) Select Two Parameters; (3) Enter Hourly Display Period; (4) Enter Daily Display Period; (5) SRPM Outputs; (6) BMPAM Outputs; and (7) Comparison for Pre-/Post- BMP (Figure 4 ). The first menu item allows users to select either runoff or a pollutant constituent to compare Pre-BMP and Post-BMP simulation results. The second menu item asks users to choose two parameters for simulation result display. The third and fourth menu items enable users to define the hourly and daily period of simulation results, respectively. The next two menu items allow users to display the hourly, daily, monthly, and annual simulation results from SRPM and BMPAM, respectively. The last menu item compares simulation results of Pre-BMP and Post-BMP conditions. Hourly and daily simulation results are plotted on line charts, whereas the monthly and annual results are plotted on bar charts. Tabular results of all displays described above also are available through this menu.

The last menu of the IGM system is Exit Project. This menu allows users to save, close and exit the current project, and open another project if they desire. Under the Exit Project menu, users can also customize the IGM system by adding and specifying new menus, menu items, and buttons (Figure 5 ).

Two sites totaling 42087 m² (10.4 acres) in an agriculture area of Okeechobee County, Florida were selected for this application of the IGM system (Figure 6 ). Hourly precipitation and evaporation data from a nearby site were pre-stored in the system. All other model inputs were entered under the Stormwater Model menu. These input data were adapted from a calibrated SRPM model applied in a nearby agricultural area (Xue and Zhang, 1997). Figures 7 and 8 show the input data for the selected site parameters and phosphorus transport data, respectively. The SRPM model was then run in the IGM system.

To evaluate nutrient removal efficiency of a BMP, a hypothetical wet detention pond was simulated through the IGM system. The surface area and the storage depth of the pond were 6070 m² (65340 ft²) and 1.2 m (4 feet), respectively. The pond had one natural infiltration layer. Nutrient uptake by plants was not considered in this application. The wet detention pond received runoff simulated from the 42087 m² agricultural area previously described. The simulated runoff and phosphorus concentration from the SRPM model were fed into the BMPAM model. Hourly meteorological data also were pre-stored in the IGM system. All other model inputs required by BMPAM were entered through the BMP Model menu. Figures 9 and 10 show the input data for the hydraulic parameters and initial pollutant concentrations used for this application, respectively.

Both SRPM and BMPAM were run continuously for a 33-month period (1989 - 1991) at hourly time steps. Simulation results from both models were viewed under the Simulation. Outputs menu in the IGM system. Figures 11 and 12 show daily total phosphorus and total nitrogen concentrations generated from SRPM and annual runoff and total phosphorus loads from BMPAM, respectively. Figure 13 shows hourly comparison of runoff for pre- and post- BMP conditions. The hourly hydrographs for pre- and post-BMP conditions demonstrate a significant reduction and delay of peak runoff after the wet detention pond was applied. A significant reduction of total phosphorus loads was observed in 1991 when a monthly comparison of pre- and post- BMP conditions was made (Figure 14 ). The last two columns of data in Figure 15 demonstrate the same comparison results in tabular format.

We gratefully appreciate the comments of Susan Gray, Weihe Guan, Guangdou Hu, and Todd Tisdale, all from South Florida Water Management District, for this paper.

Chen, Z., M. Lahlou, R. Z. Xue, and M. S. Cheng, 1995. "An Integrated Impact Assessment Model for Urban Watershed Planning." Proceedings of the 4th Biennial Stormwater Research Conference. Southwest Florida Water Management District, Brooksville, Fl., pp. 53-62.

Huber, W. C., Heaney, J. P., Nix, S. J., Dickinson, R. E., and Polmann, D. J. 1987, Storm Water Management Model - User's Manual. EPA 600/2-84-109a. U.S. Environmental Protection Agency. Cincinnati, OH.

Hutchinson, S. and L. Daniel, 1995. Inside ArcView, Santa Fe, NM:OnWord Press.

Xue, R. Z., 1996a. Stormwater Runoff and Pollutant Model - SRPM Model Documentation - Version 1.3, South Florida Water Management District, West Palm Beach, FL.

Xue, R. Z., 1996b. Best Management Practice Assessment Model - BMPAM Model Documentation - Version 1.1, South Florida Water Management District, West Palm Beach, FL.

Xue, R. Z. and J. Zhang, 1996. "Application of SRPM to Predict Stormwater Runoff and Phosphorus Loads." Proceedings of the AWRA Annual Symposium of Watershed Restoration Management: Physical, Chemical, and Biological Considerations. American Water Resources Association, Herdon, Va. TPS-96-1, pp. 461-470.

Xue, R. Z. and J. J. Zhang. 1997. "Stormwater Runoff and Pollutant Modeling in a Florida Drainage Basin". Journal of Environmental Systems. Vol. 25 (4). (in print)

Figure 1. Conceptual Design of the Integrated GIS-Modeling System

Figure 2. The Stormwater Model Menu and Its Items

Figure 3. The BMP Model Menu and Its Items

Figure 4. The Simulation Outputs Menu and Its Items

Figure 5. The Exit Project Menu and Its Items

Figure 6. Selected Two Sites in Okeechobee County, Florida

Figure 7. Input Site Data for the SRPM Model

Figure 8. Input Phosphorus Transport Data for the SRPM Model

Figure 9. Input Hydraulic Parameters for the BMPAM Model

Figure 10. Input Initial Pollutant Concentrations for the BMPAM Model

Figure 11. Simulated Daily TP and TN Concentrations from SRPM

Figure 12. Simulated Annual Runoff and TP from BMPAM

Figure 13. Simulated Hourly Runoff Before and After BMP

Figure 14. Simulated Monthly TP Loads in 1991 Before and After BMP (Graphics)

Figure 15. Simulated Monthly TP Loads in 1991 Before and After BMP (Tabular)