Modeling the amount of free water surface evaporation from Mahabad Dam’s lake using artificial neural networks and comparing the model with experimental equations
Modeling the amount of free water surface evaporation from Mahabad Dam’s lake using artificial neural networks and comparing the model with experimental equations
rasul hajian1 Mohammad Ali Lotfollahi-Yaghin2
1) Iran Islamic Azad University of Maragheh Branch Email:
2) Iran University of Tabriz Email:
IOI Code
:
XYCE-ADAGE
Publication :
International Conference on Civil Engineering, Architecture and urban infrastructure(icica.ir)
Abstract :
The main purpose in the current research is to estimate the amount of free water surface evaporation from the reservoir of the Mahabad Dam using the experimental methods and also MLP and RBF neural network methods and to compare them with each other. Therefore, the daily information about the meteorological parameters from the upstream synoptic station of the dam was used for three years (2009-2011). In order to evaluate the effectiveness of each one of the mentioned methods, the criteria of the coefficient of determination (R2) and root mean square error (RMSE) are calculated and compared with each other. The results show that neural network structures have the best performance in evaluation of free water surface evaporation. Furthermore, the comparison of statistical parameters of R2 and RMSE for neural network method and experimental relations indicate that 3-RBF neural network scenario has a very good performance in evaluation of the amount of surface evaporation. It has to be mentioned that for this model, the value of R2 is 0.9041 and RMSE has been 0.0946 millimeters.
Keywords : free water surface evaporation experimental relations pan evaporation mlp network rbf network mahabad dam