Using an artificial neural network to model monthly shoreline variations

An artificial neural network (ANN) was applied to predict monthly shoreline changes at various locations along 25 km of the Noor Bay, southern Caspian Sea. The shoreline variations in 8 stations for a period of about 11 years were studied using ANN. The model results were compared with field data. The properties of the wave (height, period, energy by different equations) and wind data were fed to a feedforward backpropagation ANN. Root mean square error (RMSE) and correlation coefficient(R) statistics are used for evaluating the accuracy of the ANN. The performance (RMSE) of ANN was 0.294, 0.124, 0.13, 0.093, 0.32, 0.255, 0.41, 0.24, 0.13, 0.15, 0.06, 0.03, 0.08, 0.08 and 0.06 m for 8 stations(Golsar-1, Golsar-2, Nilofar, University-1, University-2, Darya-362, Darya-364, Darya-341, Shahrak-321, Shahrak-322, Shahrak-324, Pars-1, Pars-2, Daryashahr-1 and Daryashahr-2, respectively) in validation. The trained ANN model results had very good agreement with the shoreline changes surveys for the validation data. Results of this study show that ANN can predict monthly shoreline changes effectively.