Designing and Implementing Drought Monitoring Model in Hambantota District Using SPI and NDVI

Abstract

Drought is a natural hazard that threatens mainly our development of agriculture and causes environment, social and economic consequences. Hambantota district mostly faces to drought due to irregular precipitation patterns. Drought heavily affects people and arid region due to lack of efficient assessment and warning systems. In this research drought assessment in Hambantota district is done using various type of geospatial techniques. Standard Precipitation Index (SPI) and Normalized Difference Vegetation Index (NDVI) are used as the drought assessment indices. Metrological drought is monitored using Standard Precipitation Index (SPI) and finally, agricultural drought is monitored using Normalized Difference Vegetation Index (NDVI). By monitoring the ground data, remote sensing provides direct spatial information on vegetation stress occurred due to drought conditions. Near Infrared (NIR) and Red bands which contains in Landsat satellite images are used to calculate the Normalized Difference Vegetation Index (NDVI) using ArcGIS. Data has been extracted based on Grama Niladhari division (GN) points using a created model. Co-relation was built monthly based on Standard precipitations Index (SPI) and Normalized Difference Vegetation Index using the extracted data by Machine Learning (ML). Multiple Linear Regression (MLR) and Support Vector Regression (SVR) was used for predicting respectively SPI and NDVI. Predicted SPI values Root Mean Square Error (RMSE) was 0.4127 and NDVI values Root Mean Square Error was 0.1947 as results.