Agricultural Drought Monitoring using Satellite based Surface Soil Moisture Data

Agricultural drought refers to a period with declining Soil Moisture (SM) content and consequent crop failure from water stress. SM plays an important role in indicating water stress and thereby identifying agricultural drought. Due to the lack of large scale, fine resolution, and accurate/quality SM many agricultural drought studies are mostly based on ground-based SM observations having limited spatiotemporal variability and cannot be applied for large scale studies. Microwave remote sensing showed capability in estimating geophysical properties like SM and paved the way for a continuous agricultural drought monitoring. European Space Agency (ESA) under Climate Change Initiative (CCI) developed an active-passive multi-satellite merged ESA CCI SM dataset. In this study, ESA CCI SM’s potential in agricultural drought monitoring is explored, by deriving Empirical Standardized Soil Moisture Index (ESSMI) to identify agricultural drought in Indian state of Telangana from 2001 to 2020. Telangana is a severely drought-prone state of India heavily impacted by significant water stress and water shortages due to frequent droughts. This increases the need for accurate agricultural drought characterization in the state. Keeping in mind the necessity of drought monitoring system for Telangana and availability of large-scale satellite soil moisture data from ESA CCI, this present study employs the ESSMI using the non-parametric distribution of ESA CCI SM data, to characterize the agricultural drought in drought prone Telangana. The efficiency of ESSMI in drought monitoring is evaluated by comparing it to the Standardised Precipitation Index (SPI) and Rainfall Anomalies (RFA) calculated from India Meteorogical Department (IMD) daily gridded rainfall data. Both the indices along with the RFA identified 2009 as dry year and 2020 as wet year. Temporal evolution of monthly drought identified by ESSMI showed monthly delayed response when compared with SPI, whereas yearly ESSMI showed consistency with SPI and RFA. Different classes of drought areas identified by ESSMI are compared with SPI which showed near normal and mild dry regions for most of the study period. ESSMI is able to effectively capture near normal to moderate drought events and shows a consistent association with the SPI and RFA both in short and long term (monthly and annual) temporal scale. The study showed the overall performance of ESSMI is reliable for agricultural drought monitoring and can be used to develop effective drought warning and risk management.