Using multiple watershed models to assess the water quality impacts of alternate land development scenarios for a small community

Article history: Received 20 April 2016 Received in revised form 25 October 2016 Accepted 6 November 2016 Available online xxxx Chesapeake Bay, the largest estuary in North America, is impaired by excess nutrient discharges, especially from urban and agricultural land.Watershed simulationmodels have provided key insights to understanding land-towater connections, but rarely are these models applied to guide local landmanagement to explore and communicate uncertainty in themodel predictions. In this study, threewatershed simulationmodels; the Soil andWater Assessment Tool (SWAT), the GeneralizedWatershed Loading Function (GWLF) model, and the Chesapeake Bay Program's Chesapeake Watershed Model (CBP-CWM) were implemented to predict water, total nitrogen, and total phosphorus discharges from small tributaries in the town of Queenstown, Maryland, USA. Based on our evaluation metrics, none of the models consistently provided better results. In general, there was a good agreement on annual average water flow between the SWAT and CBP-CWM models, and the GWLF and CBP-CWM models predicted similar TN and TP loads. Eachmodel has strengths andweaknesses inflowand nutrient predictions, and predictions differed amongmodels evenwhenmodels were initializedwith the same data. Usingmultiple models may enhance the quality of model predictions and the decision making process. However, it could also be the case that the complexity of implemented watershed models and resolution of our understanding currently are not yet suited to provide scientifically credible solutions. © 2016 Elsevier B.V. All rights reserved.