Evaporation Estimation Using Support Vector Machines Technique

Abstract

Abstract—This paper is aimed at presenting a preliminary study on evaporation, as a major component of the hydrologic cycle that plays a key role in water resources development and management in arid and semi-arid regions. Debate remains as to the optimal method for estimating evaporation from free water surfaces of reservoirs. This paper investigates the abilities of Support Vector Machines (SVMs) technique to improve the accuracy of daily evaporation estimation in the Chahnimeh reservoirs of Zabol in the southeast of Iran. This paper describes SVMs technique for simulating evaporation, whose performance compared favorably based on performance criteria such as root mean square error (RMSE), mean absolute error (MAE), mean square error (MSE) and coefficient of determination (R 2) have been employed for comparison of the results obtained from this technique whit some empirical methods. In this paper, the Gamma Test (GT) has been used for the first time in modeling one of the key hydrological components: evaporation estimation modeling. Keywords—Evaporation, support vector machine, gamma test, chahnimeh reservoirs. I.