Artificial Intelligence Techniques for Rainfall Prediction in Mizoram

Rainfall prediction plays a vital role in managing hydrological events. An accurate predictive model can provide timely information to reduce the affects of extreme events such as, drought and flood. In Mizoram, agriculture is a key component of the economy. Accurate and early rainfall predictions are crucial to mitigate the adverse affects of prolonged dry periods or heavy rains on crop yield and the overall economy of the state. Artificial intelligence (AI) models can achieve precise rainfall predictions by uncovering concealed patterns within historical rainfall data. In the present study, Al models viz. extreme learning machine (ELM), artificial neural network (ANN) and support vector machine (SVM) have been used for predicting rainfall in Mizoram state of India. The meteorological data from 2002 to 2021 is used to train and test the predictive accuracies of the models. The study demonstrates reasonably accurate predictive capability of all the aforesaid AI models considered in the study. The performance of ELM was found superior, exhibiting smaller error and higher coefficient of determination than ANN and SVM.