Spatial modelling of regional drought severity index based on multiple criteria analysis using cloud-based remote sensing data in agriculture land

A remote sensing index is widely used to monitor the three types of droughts including meteorological, agricultural, and hydrological. Combination of them rarely studied to improve the accuracy. This research aims to monitor the deficit in agricultural land using Regional Drought Severity Index (RDSI) model. RDSI tends to become the modification Drought Severity Index (DSI) in terms of data input, processing, and weighting variables for the district mapping level. A CHIRPS 1981-2020 and Landsat-8 2019-2020 are obtained and processed using Google Earth Engine (GEE) moreover Maxar High-Resolution Satellite Imagery 2019-2020 is collected from Google Earth. Meteorological, agricultural, and hydrological droughts are derived from Standardized Precipitation Index-3 (SPI-3), Vegetation Health Index (VHI), and Soil Moisture Index (SMI). Multiple Criteria Analysis (MCA) was used to combine these three types of drought. Meteorological, agricultural, and hydrological droughts are in terms of rainfall, vegetation stress, and soil and surface water content respectively. Furthermore, the RDSI model was employed to monitor drought in Kebumen regency, Indonesia in 2019 when El-Nino occurred. The model which combined these three types of deficit has an accuracy of 84.6% which higher than the SPI-3, VHI, and SMI. In Nov-Dec-Jan 2019, about 10,997 Ha is affected by drought than normal conditions in 2020. In conclucion, drought dominates coastal areas with fluvio-marine landforms in rainfed paddy fields.