Declaring Drought for Effective Water Management

Water managers are tasked with resolving conflicts between freshwater resource uses, which range from municipal water supply, to recreation, and to sustaining aquatic ecosystem integrity. Further complicating management, hydrologic processes experience numerous sources of periodic, quasi-periodic, and episodic variation. Water allocation trade offs are often most complex and contentious when availability is low. Drought is a “recurring extreme climatic event over land characterized by below-normal precipitation over a period of months to years” (Dai 2011). Water managers often apply indicators of climatologic and hydrologic conditions to identify when drought conditions are reached (e.g., Palmer Drought Severity Index, streamflow, respectively). These indicators inform drought declarations, with associated drought responses such as watering restrictions. Herein, we suggest techniques for predicting and declaring oncoming drought to improve the accuracy of drought declarations. We hypothesize that drought indicators in preceding months are predictive of future drought levels. Specifically, we develop predictive models using the Palmer Hydrologic Drought Index, a common drought indicator. We then demonstrate the utility of our model for drought declarations for the Middle Oconee River near Athens.