The Superiority of Data-Driven Techniques for Estimation of Daily Pan Evaporation

In the present study, estimating pan evaporation (Epan) was evaluated based on different input parameters: maximum and minimum temperatures, relative humidity, wind speed, bright sunshine hours. The techniques used for Epan were artificial neural network (ANN), wavelet-based ANN (WANN), radial function-based support vector machine (SVM-RF), linear SVM (SVM-LF), multi-linear regression (MLR) models. proposed models trained tested in three scenarios (Scenario 1, Scenario 2, 3) utilizing percentages of data points. 1 includes 60%: 40%, 2 70%: 30%, 3 80%: 20% accounting training testing dataset, respectively. various statistical tools such as Pearson’s correlation coefficient (PCC), root mean square error (RMSE), Nash–Sutcliffe efficiency (NSE), Willmott Index (WI) to evaluate performance graphical representation, a line diagram, scatter plot, Taylor also model’s performance. model results showed that SVM-RF is superior other all scenarios. most accurate values PCC, RMSE, NSE, WI found be 0.607, 1.349, 0.183, 0.749, respectively, during (60%: 40% training: testing) among This with an increase sample set training, would show less modeled result. Thus, evolved produce comparatively better outcomes foster decision-making water managers planners.