Assessment of Spatiotemporal Variation of Agricultural and Meteorological Drought in Gujarat (India) Using Remote Sensing and GIS

The agricultural and meteorological droughts are prominent natural hazards, and the risk associated with them has adversely affected rainfed agriculture with a more prominent impact in arid and semiarid regions. In the present study, two drought indices, namely vegetation condition index (VCI) for spatial and temporal monitoring of agricultural droughts and standardized precipitation index (SPI) for spatial and temporal monitoring of meteorological droughts, were utilized derived considering the 30 years of dataset between 1986 and 2015. The current study aims to use geospatial techniques to classify the onset and spatial extent of the VCI used in this study to monitor agricultural drought and assess the suitability of NOAA-AVHRR, derived VCI to monitor an agricultural drought at the regional scale. The SPI drought index has gained momentum in assessing drought severity, magnitude, and its geographical extent. One of the major advantages of using the SPI is that it can identify and monitor various types of dryness conditions at various time scales. The SPI index can consider daily, weekly, and monthly variations in precipitation across multiple time scales. The present study computes the meteorological drought indices in terms of SPI using rainfall data from the year 1986 to 2015 using a 3-month time scales. SPI at 3-month time scale was interpolated to show how drought patterns are spatially distributed and how severe they are during droughts and wet years. A correlation analysis was conducted to evaluate its effectiveness in quantifying the effects of drought on crop production. In addition, the SPI 3-month time series was used to assess drought risk in Gujarat. In this study, the crop yield anomaly index was also applied for the analysis of the association between precipitation and crop yield throughout wet and dry years. In this study, a significant correlation between NDVI and rainfall (r = 0.51), NDVI and crop yield anomaly (0.42), and SPI and crop yield anomaly (0.65) was observed. The findings show that the frequency of drought events varied on a complex drought during the study period. As a result of the combined drought risk map, 21.68%, 22.66%, and 30.26% of the study area were at risk for very severe, severe, and moderate drought conditions, respectively.