A fine-grained, Multi-model system to Predict Rainfall across diverse datasets

Rainfall is crucial to agriculture and a farmer's means of subsistence. It has been very helpful to farmers, especially in the agricultural society of the Indian Subcontinent, where it is heavily used for agricultural production, ground water replenishment, and irrigation. Rainfall is therefore a vital and important event for a sizable section of the Indian population. Therefore, rainfall prediction is by far the most effective and advantageous method that has the power to drastically improve the lives of a great number of people. The process of estimating the probability of rainfall at a given place, forecasting future precipitation, and figuring out how much rain will fall in particular locations is known as rainfall prediction. Along with the probability of rainfall at that specific place, it takes into account the precipitation volume assessment, forecast accuracy, and prediction error. Consequently, an efficient method for rainfall prediction is presented in this research effort. The proposed method uses Artificial neural network (ANN) and Long short term memory (LSTM) model to estimate the rainfall prediction. This research used many datasets from the Indian subcontinent and other continent-wide datasets to estimate the rainfall. Results are measured for the Root mean square error (RMSE) parameter on a variety of datasets, demonstrating how well the system performs in comparison to some previous efforts.