Investigation of agricultural drought in paddy using vegetation indices and land surface temperature in Anuradhapura district, Sri Lanka

Abstract

Agricultural drought is defined as the decline in the productivity of crops due to irregularities in rainfall, increase in the temperature rate, etc., which causes a decrease in soil moisture. Remote sensing-based vegetation indices and land surface temperature (LST) data play a key role in monitoring and assessing agricultural drought. Therefore, the present study aimed to monitor vegetation indices and LST to assess agricultural drought stress in paddy cultivation in the Anuradhapura district using Landsat-8 satellite data. As study sites, six large paddy fields (Area: 0.1 km2 - 0.5 km2) were selected using Google Earth from drought prone Divisional Secretary’s divisions in Anuradhapura district based on recent research studies. Landsat-8 satellite images of the Anuradhapura district from 2015 to 2020 were downloaded from the Land Viewer website. The average paddy yield of Anuradhapura district in the Yala and Maha seasons during the study period was collected from the Department of Census and Statistics. Normalized Difference Vegetation Index (NDVI), LST and Vegetation Condition Index (VCI) were calculated using ArcMap’s Raster Calculator tool. NDVI, LST and VCI values of 25 locations from each map were obtained to calculate the average value of each index for each season. According to the results of VCI and yield statistics, 2016/2017-Maha, 2017-Yala, 2017/2018-Maha and 2018-Yala were identified as agricultural drought seasons affecting the paddy yield in Anuradhapura district from 2015 to 2020. There was no significant difference (p<0.05) between NDVI in Yala and NDVI in Maha season. In contrast, there was a significant difference (p<0.05) between LST in Yala and LST in Maha season, suggesting that a combination of NDVI and LST is essential to obtain accurate and reliable results. According to the regression analysis (p<0.05), the results validation of NDVI and VCI exhibited positive associations with paddy yield data, whereas LST exhibited a negative association. Therefore, the current study revealed that the vegetation indices, i.e., NDVI, VCI together with LST play a significant role in determining agricultural drought.