A New SEBAL Approach Modified with Backtracking Search Algorithm for Actual Evapotranspiration Mapping and On-Site Application

Abstract

Actual evapotranspiration is one of the most important component for efficient water management and planning. Until recently, evapotranspiration computations and measurements have been performed locally. But, in recent years, actual evapotranspiration computations can be calculated regionally thanks to improvements on remote sensing discipline and satellites. Surface Energy Balance Algorithm for Land (SEBAL) is one of the most commonly preferred technics for actual evapotranspiration mapping. However, this algorithm has some difficulties such as mismatch with Landsat 8 and hot–wet pixel selection which require expert knowledge. In this paper, a novel SEBAL based approach (SEBAL-BSA), which can use Landsat 8 images as data and can automatically perform hot–wet pixel selection using Backtracking Search Algorithm (BSA) with ground control points, has been proposed. SEBAL-BSA was developed and tested using January 22–2015, April 28–2015 and July 01–2015 dated Landsat 8 images in Akarsu Irrigation Water User Association command area in the Cukurova Region of Turkey and actual evapotranspiration mapping was realized without albedo measurements. Accuracy of SEBAL-BSA has been examined by comparing values of predefined ground truth points and values obtained from proposed approach. According to results of parametric matched pairs T test and nonparametric Wilcoxon sign rank test, SEBAL-BSA is highly successful. Applications also show that the SEBAL-BSA is a user-friendly approach for institutions and organizations related to water authorization and management.