Enhancing spatial resolution of drought monitoring through a novel random forest-based GRACE drought index: a case study in Central Yunnan

Abstract

The original coarse spatial resolution of Gravity Recovery and Climate Experiment (GRACE) data limits its application in small and medium-scale areas. The study proposes a novel random forest-based GRACE drought index (NRFGDI) to enhance the spatial resolution of drought monitoring, which is applied to central Yunnan. We first employed the random forest (RF) algorithm to improve the spatial resolution of GRACE-derived terrestrial water storage anomaly (TWSA) to 0.1°, then established the NRFGDI based on the downscaled TWSA. Subsequently, the NRFGDI was verified using traditional drought indexes, including the standardized precipitation evapotranspiration index (SPEI) and the self-calibrating Palmer drought severity index (sc-PDSI), along with the 2009/2010 Bulletin of Flood and Drought Disasters in China. The NRFGDI not only captures drought characteristics, but also offers finer spatial resolution. The study provides an effective means for medium to small-scale drought monitoring, thereby making valuable contributions to decision-making processes concerning droughts in central Yunnan.

Keywords

• Novel random forest-based GRACE drought index
• terrestrial water storage anomaly
• spatial resolution enhancement
• soil moisture
• drought monitoring