The study of evapotranspiration is important for agricultural planning, especially in regions characterized by high spatial and temporal variability of precipitation, as in arid and semiarid regions of northeastern Brazil. Although in Brazil there are two large meteorological observation networks (INMET and CPTEC), there are still many uncovered areas, suggesting the need for methods of calculating evapotranspiration based on few meteorological elements. Considering this need, this study aimed to determine the reference evapotranspiration (ETo) by indirect methods in various time scales in the experimental watershed of the creek Gameleira, Vitoria de Santo Antão, PE. Indirect methods have been evaluated considering the Penman-Monteith FAO-56 as the standard method. Among the methods based on air temperature, the Hargreaves-Samani (1985) presented an "optimal" performance at pentadal to monthly scales, followed by the Blaney-Criddle with performance "very good". The Campbell, Thornthwaite and the modified Thornthwaite methods showed variations of performance from "bad to very good". Among the solar radiation based methods, the best performance was the solar radiation method, followed by Makkink and Jensen-Haise. The methods based on statistical modeling did not estimate ETo with good accuracy, resulting in low determination and performance coefficients values.
Penman-Monteith; indirect methods; semi-arid
