Spatial and Temporal Patterns of Baseflow Recession at the Continental Scale

Baseflow is often treated as a static feature of watersheds according to a unique storage-discharge relationship. However, recent innovations in baseflow recession analysis have allowed novel discoveries regarding the regional clustering of typical baseflow characteristics, as well as the variability of both the stability of baseflow and its nonlinearity (i.e., the concavity of the hydrograph). We investigate spatial and temporal patterns in the character of baseflow recession at over 1,000 watersheds in the continental United States (US). We discover seasonal patterns in both the stability and nonlinearity of baseflow which vary systematically across large regions. Further, we relate these baseflow characteristics to their potential physical drivers, including estimates of evapotranspiration, watershed storage, the distribution of watershed storage, and precipitation. While coincident watershed storage is the best predictor of baseflow stability in many watersheds (particularly in the Appalachian Mountains), evapotranspiration from two to three months previous is the best predictor of baseflow stability in many other watersheds (particularly in the Pacific Northwest). We also make the novel discovery that baseflow nonlinearity has been increasing significantly in most watersheds across the US since 1980. Finally, we integrate these complex spatial patterns and temporal dynamics according to a simple, physically-based conceptual model.