Thursday, March 13, 2008
Drought: Monitoring, estimation and prediction?
Drought is one of the most costly natural disasters, in both economic and societal terms. There have been many attempts at monitoring and predicting drought severity, mostly involving the use of the Palmer Drought Severity Index (PDSI). Although widely used, PDSI has some important disadvantages such as not taking account of cold processes, while it cannot directly characterize the different types of drought (meteorological, agricultural and hydrological). An alternative approach is the use of hydrology models to produce spatially and temporaly continuous datasets of drought-related variables such as soil moisture and runoff. This approach is demonstrated in a number of different contexts: (1) monitoring the spatial extent and severity of drought as well as forecasting drought recovery using real-time hydrologic data provided by the UW Surface Water Monitor, (2) historical reconstruction and trend evaluation of drought characteristics for the 20th century over the continental United States using single- and multi-model simulations of soil moisture and runoff, (3) identifying the most severe drought events globally for the second half of the 20th century, examining their impacts on agriculture and exploring potential teleconnections, and (4) comparing droughts between the 20th and 21st centuries in a consistent way using predictions from a Global Circulation Model.
Kostas Andreadis is a
PhD candidate in the UW Dept. of Civil and Environmental Engineering.