Abstract:
A rainfall forecasting model for Sri Lanka is constructed using artificial feed-forward
back-propagation neural network. The data obtained from meteorological station in
Colombo is used. The rainfall data with the humidity and atmospheric pressure
measurements taken at the ground level are used to train the network. The preliminary
results indicate that the model is successful in predicting rainy days one day ahead in
Colombo area with the accuracy of 67%. The possibilities for further improvement of the
model are discussed.