Evaluating the Potential Use of Remotely Sensed Soil Moisture Data for Agricultural Drought Risk Monitoring

Abstract:

The Pacific Northwest US has received historically low rainfall and snowpack during winter 2015, with drought emergence declared for both states in spring 2015. To mitigate the impacts of drought vulnerability, an operational near-real-time (NRT) drought monitoring with remote sensing technique is investigated. This study provides a comprehensive assessment of the potential of remotely sensed surface soil moisture data in monitoring agricultural drought over the Columbia River Basin (CRB), Pacific Northwest. Two satellite soil moisture datasets were evaluated, the LPRM-AMSR-E (unscaled, 2002-2011) and ESA-CCI (scaled, 1979-2013). The satellite drought monitoring skill is examined with two indices: drought area coverage (the ability of drought detection) and drought severity (according to USDM categories). The effects of satellite sensors (active, passive), multi-satellite combined, and length of climatology are also examined in this study.

In order to improve the remote sensing drought monitoring skill, statistical methods including regionalization, with the concept of “trading space for time”; and also bootstrapping are introduced.