Urban watersheds contain a multitude of land covers with various levels of connectivity to both one another and local surface waters. Recent advances have been made to isolate these land covers in an effort to identify attributes of interest (e.g. connected impervious area). These advancements can lead to more targeted interventions to treat the most impactful imperious areas in a given watershed (the portion that has high connectivity to the stream), as well as inform pollutant models by better attribution of build-up wash off variables based on accurate land cover estimates. A study in Knoxville, Tennessee, in the Baker Creek watershed has explored the use of microbial community finger prints to isolate runoff generated from various sources in the watershed (rooftop, roadway, greenspace, etc.). At the same time, samples were collected from the stream and stormwater outfalls at multiple locations around the watershed. A Bayesian mixing model (SourceTracker) was used to estimate the portion of each stream/outfall sample originating from the various sources in the watershed. Initial results suggest SourceTracker outputs agree well with urban hydrologic theory, where connected impervious area contributes more to flows early in the storm and during small storm events in watersheds that are highly urbanized.
