Prediction of seasonal and annual rainfall using artificial neural network

Prediction of seasonal and annual rainfall for a river basin is of utmost importance for planning and design of irrigation and drainage systems as also for command area development. Assessment of rainfall can be carried out by different approaches like deterministic, stochastic, conceptual and soft computing. This paper illustrates the use of Artificial Neural Network (ANN) for prediction of rainfall for Krishna and Godavari river basins. ANN models such as Multi-Layer Perceptron (MLPN), cascade correlation and conjugate gradient are applied to train the network data. Model performance indicators such as correlation coefficient, model efficiency and root mean square error are used to evaluate the performance of the ANN models. The study showed that the MLPN is better suited for prediction of seasonal and annual rainfall for Krishna and Godavari river basins.