Performance comparison of complementary relationship and inverse analysis methods for evapotranspiration estimation.

Abstract

Evapotranspiration is a critical factor that plays a pivotal role in irrigation and water resource planning. It is also a major influence with regard to global warming issues. Therefore, several studies have been performed on the estimation methods for evapotranspiration; however, the most prevalent method used over a long period to estimate evapotranspiration is the complementary relationship approach. Recently, this method was modified to improve its symmetry and accuracy. However, to achieve better performance, a different evapotranspiration estimation method using an inverse analysis of the Bowen ratio was proposed by us in an earlier study. In order to present a performance comparison of these different estimation methods, the reproducibility of three types of complementary relationship methods and inverse analysis methods was assessed in this study. Our study utilized data from FLUXNET2015 and evaluated the performance of the methods using regression analysis and root mean square error (RMSE) with data from 15 test sites, mainly located in the U.S.A., for a total period of 132 years. From the results, it was observed that the inverse analysis approach demonstrated a slightly better performance than the complementary relationship methods. This study provides a valuable direction for future research works on the estimation of evapotranspiration.