Water Quality Modeling for Urban Watersheds Using SWAT: Distributed Urban Best Management Practices and Onsite Septic Systems.

Sprawling urban and suburban development creates negative impacts on water quality. Understanding the causes and their environmental consequences related to the sprawl may provide insights on how water resources and quality can be managed for constructing more environmentally sustainable communities over the long term. The Soil and Water Assessment Tool (SWAT) has gained popularity among scientists in simulating water quantity and quality at the watershed scale to river basin scale, though it is optimized for rural and agricultural lands. To better represent sources and contaminant transport mechanism in urban catchments, SWAT has recently been enhanced with urban subroutines. Physically based algorithms for stormwater Best Management Practices (BMPs) were added to SWAT as a part of sub-daily simulation such as sedimentation-filtration basin, retention-irrigation basin, detention pond, and wet pond. These water quality control structures are commonly used in urban central Texas areas. Onsite Septic Systems (OWSs) are a significant source of nonpoint?source pollution to surface water and groundwater in suburban settings. Methods to quantify their effect are therefore important. The mechanics of OWSsĄŻ biogeochemical processes are implemented in SWAT for applications at the watershed scale. These recent enhancements in SWAT will be presented with case studies in which the algorithms are evaluated against measured data. Results indicate that the urban BMP module for simulating sandfilters successfully reproduce measured flow and sediment load in a small highly urban catchment in downtown Austin, Texas (NSE=0.82 and 0.83, respectively for flow and sediment). In another case study in which a total of 227 OWSs were evaluated in a suburban watershed in North Carolina, the model performed well in predicting groundwater levels (R2 = 0.82 and PBIAS = ?0.8%) and NO3?N concentration in the groundwater (R2 = 0.76, PBIAS = 2.5%) at two sites.