LAUSR

Abstract The Scaled Drought Condition Index (SDCI) is an agricultural drought index derived from multiple remote sensing datasets including precipitation from the Tropical Rainfall Measurement Mission (TRMM) and temperature and vegetation index from the Moderate Resolution Imaging Spectroradiometer (MODIS). The SDCI can be used for large-scale drought monitoring at a high spatial resolution in dry and humid regions. However, it is not applicable to certain areas for numerous reasons, including the discontinued service and limited coverage of the TRMM, uncertainties associated with the coarse resolution of precipitation input, the fixed lag time between precipitation deficit and vegetation responses, experimental weights and an arbitrary classification scheme. Therefore, this study aims to propose an optimal SDCI (OSDCI) using an alternative Global Satellite Mapping of Precipitation (GSMaP) as precipitation input, adding a varied lag time between precipitation and vegetation response, determining the optimal weights of variables and revising the severity classification. In Central Asia, OSDCI is calibrated using the 3-month Standardized Precipitation Evapotranspiration Index (SPEI) during the 2001–2008 growing seasons as drought reference. The SPEI from 2009 to 2016, soil moisture anomalies (SMAs) derived from both the Global Land Data Assimilation System (GLDAS) and the Climate Prediction Center (CPC), and the annual crop production of wheat and barley are used for cross-validation. For comparison, three original SDCIs are also included. The results show that the OSDCI displays better performance than the three experimental SDCIs with higher correlation coefficients relative to SPEI and SMA from the CPC for different cover types, and performance is comparable to that of SMA from the GLDAS and annual crop production. Additionally, the OSDCI based on the revised classification scheme (OSDCI_rev) shows significant superiority over the OSDCI based on the original classification system (OSDCI_org) with lower drought frequency and area error. Furthermore, the OSDCI_rev can capture both the temporal evolution and the spatial patterns of droughts exemplified by the SPEI and SMAs. Overall, the results of this study show the potential use of the OSDCI with a revised classification scheme as a remote sensing drought index for agricultural drought monitoring.