Development and Calibration of a High Resolution SWMM Model for Simulating the Effects of LID Retrofits on the Outflow Hydrograph of a Dense Urban Watershed

Most urban watershed models developed to comply with various regulatory programs can only be used to approximate LID implementation in their existing configuration. These models were developed in order to simulate flow through the thousands of acres that together make up each water pollution control facility (WPCF) service. Consequently, the resolution of the original models is relatively coarse. For example, tens of square blocks are aggregated into single hydrologic units called subcatchments, and only the WPCP's head works, interceptors, major trunk sewers, and large diameter sewer pipes are included in the sewer network. Because of these approximations, and the fact that LID technologies would alter flow patterns and pollutant mobilization rates within the large-scale subcatchments, LID retrofits cannot be directly simulated in many existing urban scale hydrologic and hydraulic models. This paper will present initial work on the development and calibration of a high resolution hydraulic and hydrologic model that is being used to simulate the impact of low impact development retrofits on wet weather flowrates in the combined sewer system of a dense urban watershed. The two-block, Bronx, NY study site was selected due to its location at the hydrologic edge of the sewer service area, and also because of the extent to which its soils, slopes, open space, and building stock and vintage reflected average values for New York City, as confirmed using various statistical analyses. In the model, developed using SWMM 5.0, separate subcatchments are defined for each roof, sidewalk, backyard, and street present in the study area. Realistic routing is also incorporated in the hydrologic representation. Model calibration is accomplished by systematically varying subcatchment properties and comparing time series flowrate predictions at the downstream end of the study area with actual sewer flow observations, using precipitation time series also measured on site. The subcatchment attribute combinations associated with all sets of predictions are recorded, and parameter regions associated with reasonable and unreasonable predictions are identified. Presented are some initial attempts to simulate the effects of green roof retrofits on the outflow hydrograph from the study site. A discussion of how the resolution of subcatchment discretization affects model predictions is also provided.