Using a standardized GRACE dataset to investigate regional-scale drought severity

Accurate knowledge of surface and subsurface moisture content distribution is essential for prediction of land-atmospheric processes including the manifestation of extreme events such as floods and drought. GRACE satellites deliver observations of how masses of water are moving throughout surface and subsurface reservoirs. This paper discusses the large-scale variability of total water storage via statistical evaluation. The Upper and Lower divisions of the Solimões river basin (Western Amazon) were chosen as case study locations to develop the methodology for characterizing drought severity between 2003-2011. A quantile analysis was performed on standardized GRACE terrestrial water storage anomalies (TWSA) to define terrestrial water storage drought severity categories. Four periods of extremely low storage anomalies in years 2005, 2006, 2010 and 2011 in the Solimões were identified, which are representative of moderate-to-severe periods of drought. These GRACE-identified droughts are compared with those identified by the Standard Precipitation Index to validate the methodology and distinguish relationships between meteorological and hydrologic droughts. Peak severities in GRACE-identified drought differ spatially and temporally, lagging precipitation deficits by two-to-four months in the Solimões. This methodology is also applied to river basins in South Africa and Australia with the purpose of demonstrating that GRACE is a viable observation tool for regional-scale drought monitoring around the world.