Developing a GIS-Based Model to Track Potential Point and Non-Point Sources of Urban Stream Pollution

Abstract:

Urban streams are often characterized by diminished water quality resulting from an increase in polluted runoff from impervious surfaces. Storm activity further reduces urban stream water quality by temporarily increasing stormwater discharge from sewer overflows. This will often manifest itself in rapid declines of dissolved oxygen and peaks in specific conductivity in response to a rising biochemical oxygen demand which slowly recovers as the pollution load is washed through the stream system. This research developed a GIS-based model to track potential sources of pollution based on the dissolved oxygen and specific conductivity response of urban streams to a series of storm events, within the city of Louisville, Kentucky. Watershed outlet hydrographs were first obtained to determine the lag time of dissolved oxygen drops and specific conductivity peaks in response to set of storm events. Using a digital elevation model and the National Landcover Database, 10m resolution rasters were then created which calculated slope and flow direction/accumulation for both open channel and overland flow conditions across the watersheds. The rasters were merged and converted to flow velocities using a series of storms with different intensities. The final step utilized the Flow Length tool in ArcGIS which calculated the travel time to the watershed outlets from each pixel weighted by the open channel and overland flow conditions. Potential pollution sources could then be located by matching the dissolved oxygen and specific conductivity response lag times to the associated watershed travel times.