Optimization of Rainfall Prediction Using Satellite Data Through Machine Learning and Deep Learning Algorithms

In recent years, non-periodic rainfall has increased globally. The importance of accurate rainfall prediction model is increased in order to reduce the losses due to floods and drought. Rainfall prediction plays a crucial role in decision-making, flood and drought risk assessment and management. In agriculture, it guides many farmers in optimizing crop planting schedules and irrigation. It ensures better yields and reduces losses from droughts and floods. Accurate rainfall prediction benefits us in effective water resource management and energy sector planning. In India, the rainfall varies rapidly due to the vast geography. There are many factors which effects rainfall like distance from the sea, seasons, the presence of mountains, prevailing winds and precipitation. This research work mainly focuses on developing an efficient ensemble-based Machine Learning Techniques such as Extreme Gradient boosting algorithm (XG Boost) and Logistic Regression (LR) for predicting rainfall in different regions of India using satellite data and comparing results with the deep learning algorithms such as Convolutional Neural Network (CNN), Multiple Layer Perceptron (MLP), Long Short-Term Memory (LSTM), and Radial Basis Function Network (RBF). The aim of this paper is to overcome the drawbacks of existing systems and build improved prediction model with accuracy more than 90% in order to enhance water resource management, early warning systems, improved urban planning, agriculture, and climate change studies.