This paper attempts to assess the potential and usefulness of Support Vector Machine (polynomial and radial basis function) based modelling techniques for the prediction of daily evaporation from a reservoir. The meteorological data of daily evaporation caused by temperature, solar radiation, relative humidity, and wind speed, are used in this study. The performance of the Support Vector Machine (SVM) is compared with linear regression on the basis of performance parameters (correlation coefficient and root mean squared error) having different combinations of input parameters. A comparison of the results reveals that there is better agreement when large input parameters are considered for modelling, as compared to a single parameter. The outcome of the parametric study suggests that SVM-based modelling can be applied as an alternative approach for the estimation of daily evaporation from reservoirs in hydrological analysis.
