Identification of Drought and developing a crop yield model using Meteorological and Geospatial data

Due to lack of precipitation,drought has been increasing day by day which results in a severe effect on the productivity of rain-fed crops and indirect effect on the economy of the country. Drought is a climatic phenomenon which cannot be eliminated but it can be mitigated by utilizing Spatio-temporal information, which is related to available crop yield data. In this study ,drought year was identified by using 1-month Standard Precipitation Index (SPI). This index was in good relation with Normalized Difference Vegetation Index (NDVI)because both are dependent on Precipitation data. Based on this NDVI-SPI,regression models were developed. The obtained NDVI is correlated with different types of cropyield and crop yield models were developed. The highest correlation was found in Anantapur district for Sugarcane crop, Jowar in Chittoor, Horse Gram in Kadapa and Sesamum in Kurnool. The RMSE of predicted yield in a drought year was found to be 5.05% for sugarcane which was about 4.92 t/ha of average yield, 13.3% for Jowar which was about 0.006 t/ha of average yield, 9.04% for Horse gram which was about 0.02 t/ha of average yield and 6% for Sesamum which was about 0.01 t/ha of average yield