Solving the pan evaporation process complexity using the development of multiple mode of neurocomputing models

Finding an accurate computational method for estimating pan evaporation (EPm) can be useful in the application of these methods development sustainable agricultural systems and water resources management. In present study, proposed hybrid called multiple model-support vector machine (MM-SVM) with aim showing increasing, decreasing, constant accuracy behavior this model improving results EP compared to two models ANN SVM on a monthly scale EPm four meteorological stations (Ardabil, Khalkhal, Manjil (from Iran), Grand Island USA)) located semi-arid regions, using output artificial intelligence (AI) (i.e., neural network (ANN) support (SVM)), was evaluated. The intelligent several statistical indices root mean square error (RMSE), absolute value (MAE), Kling-Gupta (KGE), coefficient determination (R2)) help case visual indicators were compared. According evaluation test phase, MM-SVM-6, ANN-5, MM-SVM-3, MM-SVM-7 RMSE = 1.088, 0.761, 0.829, 0.134 mm/day; MAE 0.79, 0.54, 0.589, 0.105 KGE 0.819, 0.903, 0.972, 0.981; R2 0.939, 0.962, 0.967, 0.996 input variables introduced as best Ardabil, Manjil, stations, respectively. able use its multi-model strategy inputs estimated by independent models, power estimate scenarios where there is high correlation between components EPm, feasible state Accept show. So that incremental, constant, decreasing modes estimation under climatic conditions studied areas quite clear. Therefore, superior study local stakeholders discussing