Verification of MIKE 11-NAM Model for runoff modeling using ANN, FIS, and ARIMA methods in poorly studied basin

Abstract

Hydrological information is the basis for conducting water balance studies in any region, and surface runoff is one of the most important hydrological parameters and one of the most difficult in the process of estimation and prediction. This study aims to verification of the MIKE 11-NAM Model for runoff modeling using artificial neural network (ANN), fuzzy inference system (FIS), and autoregressive integrated moving average (ARIMA) methods at Al-Jawadiyah hydrometric station on the Orontes River in Syria. MATLAB was used to build neural and fuzzy models, where many models were built with the change in all parameters, functions, and algorithms that can be used, and the Minitab was used to build ARIMA models. Many models were prepared with the addition of seasonal effect, and the comparison results showed an advantage for artificial neural network models in terms of evaluation parameters. After that, the artificial neural network models were adopted in the process of filling the gaps in the time series of surface runoff in the study area to be used in the Mike program for modeling the runoff and through the method of trial and error with a high number of iterative cycles, model parameters were calculated and runoff values estimated. Still, the results were not good, and there were significant differences between the measured values and the values simulated by the model, and this is due to the significant lack of available data. This study recommends the use of artificial intelligence and machine learning models in the field of estimation and prediction of hydrological parameters.