Rainfall Prediction in Semi-Arid Regions in Jordan Using Back Propagation Neural Networks

In Jordan, agriculture and irrigation depend highly on rainfalls. Rainfall prediction is a challenging area of investigation for scientists. In this paper, a precipitation prediction model using artificial neural networks (ANNs) is proposed. The seasonal amount of rainfall in several areas in Jordan is predicted using rainfall rate time-series data, these rainfall rate data has been recorded from 26 stations located in different areas in Jordan. A feed forward ANN based on backpropagation (BP) is designed and trained to predict the future rainfalls in Jordan. Results are encouraging and accurate for rainfall prediction in Jordan.