Prediction of tigris River Stage in Qurna, South of Iraq, using artificial neural networks

Abstract EN

Artificial neural networks (ANNs) with back-propagation algorithm are performed for predicting the stage of Tigris River in Qurna city, Basrah, south of Iraq.

This model was adopted to investigate the applicability of ANNs as an effective tool to simulate the river stage for short term.

By using the neural network toolbox in Matlab R2007b, three models are constructed as the first experiment.

Multilayer percpetron with one hidden layer is used in the architecture of network.

The best model is selected according to the trial and error procedure based on three common statistic coefficients (coefficient of correlation, root mean square error, and coefficient of efficiency).

The best model from first experiment is used to predict the stage river for one, two, and three days ahead as the second experiment.

Results indicated the ANNs with back-propagation algorithm are a powerful technique to predict the short term stage of Tigris River.