Leveraging ML to predict climate change impact on rice crop disease in Eastern India

Abstract

Rice crop disease is critical in precision agriculture due to various influencing components and unstable environments. The current study uses machine learning (ML) models to predict rice crop disease in Eastern India based on biophysical factors for current and future scenarios. The nine biophysical parameters are precipitation (Pr), maximum temperature (Tmax), minimum temperature (Tmin), soil texture (ST), available water capacity (AWC), normalized difference vegetation index (NDVI), soil-adjusted vegetation index (SAVI), normalized difference chlorophyll index (NDCI), and normalized difference moisture index (NDMI) by Random forest (RF), Gradient Boosting Machine (GBM), Extreme Gradient Boosting (XGB), Artificial Neural Net (ANN), and Support vector Machine (SVM). The multicollinearity test Boruta feature selection techniques that assessed interdependency and prioritized the factors impacting crop disease. However, climatic change scenarios were created using the most recent Climate Coupled Model Intercomparison Project Phase 6 (CMIP6) Shared Socioeconomic Pathways (SSP) 2–4.5 and SSP5-8.5 datasets. The rice crop disease validation was accomplished using 1105 field-based farmer observation recordings. According to the current findings, Purba Bardhaman district experienced a 96.72% spread of rice brown spot disease due to weather conditions. In contrast, rice blast diseases are prevalent in the north-western region of Birbhum district, affecting 72.38% of rice plants due to high temperatures, water deficits, and low soil moisture. Rice tungro disease affects 63.45% of the rice plants in Bankura district due to nitrogen and zinc deficiencies. It was discovered that the link between NDMI and NDVI is robust and positive, with values ranging from 0.8 to 1. According to SHAP analysis, Pr, Tmin, and Tmax are the top three climatic variables impacting all types of disease cases. The study’s findings could have a substantial impact on precision crop protection and meeting the United Nations Sustainable Development Goals.