Predicting the effects of land cover and stormwater management on stream health in Baltimore County, Maryland

Abstract

Maintaining stream health in urban and agricultural riverscapes is a challenge for environmental managers, and requires a better understanding of stream ecology and biogeography. This study develops a spatial generalized least squares model for predicting the effects of land cover and stormwater management (SWM) on Piedmont stream health in Baltimore County, Maryland. Such models are needed to better integrate stream health management with local land use planning and SWM programs. The model predicts stream health impairment with >80% accuracy. Model performance is suitable for supporting a precautionary policy for preventing water quality impairment. Results show how SWM reshapes catchments, and interacts with stream health through coupled human-environment systems. Riparian buffers had little impact on stream health, while large effects were found for impervious surfaces and forest cover. Continuing development of SWM data resources and refinements to the model promise new knowledge and more nuanced and precise predictions.