A SPATIOTEMPORAL DROUGHT DATABASE AND ASSOCIATED QUERIES

A new process-based data model is developed to capture drought patterns and evolution in space and time. The data model takes footprints of drought intensity and tracks these footprints over space and time to form a spatiotemporal series of drought regions that represent drought distributions and development. We test the model with data of drought intensity regions (D0 through D4) published weekly by the National Drought Mitigation Center from January 2004 to June 2006. Strengths of the process-based data model are tested by a set of spatiotemporal queries for scientific investigation. These queries are an integral part of our study. They affect not only the design of the data model, but also provide the interface between the spatiotemporal database and human researchers seeking to understand how patterns and processes of drought affect society and nature. The process-based data model and query support enable spatiotemporal summaries of droughts through innovative interrogation of historical records. These queries go beyond the simple presentation of spatiotemporal attributes from a previously constructed data table. Answers to these queries may involve complex SQL queries combined with mathematical computation and other GIS functionality. The enhanced capabilities of the process-based data model with advanced spatiotemporal query support will reveal information about drought patterns, development, and environmental and social impacts that cannot be discerned from conventional GIS or statistical approaches.