Spatio-temporal Standardized Precipitation Index Selection using Geospatial Technique

Agricultural drought is a slow onset disaster that occurs due to a prolonged dry period in terms of deficiency of precipitation over a long period in the agricultural region as its negative impact on employment and national GDP measures. Standardized Precipitation Index (SPI) based on Precipitation and Normalized Difference Vegetation Index (NDVI) encompass a substantial impact on the SPI-drought index due to the effectiveness of remote sensing and Geospatial techniques in tracking the spatiotemporal level of agriculture drought. The SPI plays a vital role to assess the drought condition. SPI computation accumulation can be done from 1 to 48 months. Identifying the best SPI accumulation scale is always a challenge for researchers. The objective of this research work is to determine the best SPIs using the raster precipitation index carried out at different timescales (1,3,6,9,12 months) since precipitation is directly proportional to the SPI index to determine the agriculture drought. To accomplish this research, a 31-year dataset (1991-2021) of precipitation is extracted from the agricultural drought-prone region-the Jaisalmer region of Rajasthan state from IMD Pune. Principal Component Analysis (PCA) and Linear Correlation have been applied to find a correlation between precipitation and SPI. SPI-I and SPI-3 are selected as a good correlation of 0.82 and 0.65 and R-squared 0.49 and 0.39 respectively out of different timescales (1,3,6,9,12-months).