Applying GRACE to drought monitoring using a data assimilation system

Since its launch in 2002, the Gravity Recovery and Climate Experiment (GRACE) has provided unprecedented observations of terrestrial water storage (TWS) anomalies, available monthly with global coverage. These observations have offered insight on regional to global scale water storage and fluxes, but hydrological applications of GRACE have been limited by the sensor's coarse resolution: monthly, > 150,000 km2, and unstratified. Recently, the authors developed a GRACE Data Assimilation System (GRACE-DAS) that skillfully downscales and disaggregates the GRACE observation by introducing GRACE data to the Catchment Land Surface Model using an iterative Ensemble Kalman Smoother. Initial studies with GRACE-DAS yielded improved simulation of groundwater and runoff variability in the Mississippi basin, and the system is currently being adapted for implementation to the US and North American Drought Monitors. Here we present the first results for GRACE-DAS applied to North America. These results demonstrate the capability of GRACE data to influence the simulation of ground water, soil moisture, and hydrological fluxes. GRACE-DAS representation of significant droughts of the past five years (2003-2007) will be presented and compared with maps produced by the North American Drought Monitor. These results provide insight on the information content available in GRACE with respect to drought monitoring, particularly as applied to early detection and physically-based vulnerability assessment.