Combining SWMM and WASP modeling with high-resolution field measurements to evaluate runoff and stream water quality under different stormwater management scenarios in the Wissahickon Creek watershed, PA
Abstract
We are using models calibrated with high-resolution field data to evaluate how stormwater management strategies impact runoff and stream water quality in Wissahickon Creek, PA, an urban stream that runs through Philadelphia and its suburbs. In PCSWMM v7.2, the watershed (165 km 2 ) was subdivided into 204 subcatchments characterized with high-resolution land cover, soil, and topographic data. Four wastewater treatment plants and 69 existing stormwater basins were incorporated. Runoff flow and quality data from PCSWMM were input to 143 stream segments in Water Quality Analysis Program (WASP) to model in-stream processing and transport. Though WASP is typically driven using daily data, we use 15-minute inputs for improved accuracy. Field data for calibration include turbidity and/or total suspended solids (TSS) data at 14 sites and flow data at 5 sites for March-November 2017. Two primary calibration sites have TSS and flow data, which allowed a sediment load calibration within WASP. Field data are limited by the lack of access and/or accuracy during very high flows to represent peak TSS. 118 overland flow (OLF) samples were collected from 6 sites and 17 storms in 2018 and analyzed for TSS. TSS in OLF was variable (10-8,600 mg/L, median 470 mg/L) but consistently exceeded event mean concentrations (EMCs) for roads and residences (21-120 mg/L) from the Pennsylvania Department of Environmental Protection (PADEP) Stormwater Manual. To account for the uncertainty inherent in land-cover-based EMCs, we varied EMCs between 1X and 5X the PADEP values as a calibration parameter for stream water quality. For TSS calibration, 3X EMC performed best. Modeled management scenarios include converting existing detention basins to infiltration basins, increasing the area draining to stormwater basins from 4.3% (current conditions) to 30%, and a forested scenario to determine a theoretical upper bound on water quality improvement. Preliminary WASP results show that converting existing basins to infiltration reduced mean streamflow and TSS load at the basin outlet by 7% and 17%, respectively, while treating 30% of watershed area with infiltration basins reduced mean streamflow and TSS load by 21% and 51%. We are expanding the models to include stream temperature, DO, and nutrients, also calibrated with high-resolution field data.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.H43L2197B
- Keywords:
-
- 0414 Biogeochemical cycles;
- processes;
- and modeling;
- BIOGEOSCIENCES;
- 0432 Contaminant and organic biogeochemistry;
- BIOGEOSCIENCES;
- 1834 Human impacts;
- HYDROLOGY;
- 1880 Water management;
- HYDROLOGY