Geospatial Analysis of Climatic and Geomorphic Interactions Influencing Stream Discharge, Appalachian Mountains, USA

Streamflow values are commonly synthesized for locations where flow measurement stations are lacking or where only intermittent measurements are available. In an Appalachian Mountains dataset comprised of 29 watersheds, the most appropriate among geomorphic, geologic, and hydrogeologic datasets were selected for use in prediction of streamflow at watershed scale. A statistical model was developed using principal components analysis (PCA) and cluster analysis (CA) for. Using CA on variables derived from the PCA, an optimum set of variables was derived for predicting streamflow. Results indicate there are two categories of watersheds in the study area. The first is strongly correlated with climatic variables (precipitation, temperature, elevation, and groundwater recharge). The second is strongly correlated with two geomorphic variables (watershed slope and percentage of forested area). The spatial distribution of cluster classifications shows that watersheds dominated by the climatic component are located along the Allegheny Front while watersheds dominated by the geomorphic component are located in the Allegheny Plateau and Valley and Ridge physiographic provinces. These variations between the Allegheny Plateau and Valley and Ridge physiographic provinces suggest that, to accurately model streamflow, modeling needs be done based on natural physiographic boundaries rather than political boundaries. In this physiographic setting, elevation seems to be a major control.