Comparação entre métodos de estimativa da evapotranspiração de referência para o município de Petrolina, PE

Abstract

The knowledge on evapotranspiration is vital in determining the water requirements of a crop. Therefore, this paper aims to compare the performance of seven of estimation methods for the reference evapotranspiration in relation to the Penman-Monteith method, adopted as standard, for the municipality of Petrolina, state of Pernambuco, Brazil. We used daily climatic data from 2004 to 2015 to estimate the ET0 coefficient, obtained in the automatic weather station of the Test Field in Bebedouro, Embrapa in the Semi-arid climate. The statistical indicators used in the evaluation were: coefficient of determination (r2), correlation coefficient (r), agreement index (d) and performance index (c). The r2 values showed that the estimation method that best fitted to the Penman-Monteith method was Ivanov's (0.73), followed by Jensen-Haise (0.64), Makkink and Priestley-Taylor (0.63), Villa Nova (0.62), Hargreaves and Samani (0.53) and Hamon (0.45) methods. However, in relation to the index c, Hamon was classified as very poor; Makkink as poor; Hargreaves and Samani and Villa Nova as tolerable; Ivanov and Priestley-Taylor as medium; and Jensen-Haise as good. The last one was considered as the best performance rating method.