Collection System Model Development Using Raster Imperviousness Data

Imperviousness has traditionally been a principal calibration parameter for collection system models. The 2007 publication of the National Land Cover Database impervious surface raster dataset, and in Massachusetts, a 1-m statewide impervious surface raster dataset, mean that imperviousness for United States urban collection systems is now readily computed. Calibration can instead focus on determining the transfer fraction from impervious to pervious surfaces. For citywide sewer and drain system models in the city of Boston, average measured imperviousness for storm drainage was 50 percent in neighborhoods with separated sewers. Model calibration yielded an average transfer fraction of 67 percent, resulting in an effective imperviousness of 17 percent. In portions of the city with combined or partially separated sewers, measured imperviousness was 72 percent. The calibrated average routing fraction was 66 percent, yielding an effective imperviousness of 25 percent. These results demonstrate that calibration to measured runoff remains critical to model development, as a high fraction of flow is routed from impervious to pervious surfaces. Methods for estimating transfer fractions exist, but further study should evaluate patterns and differences among urban areas.