The results of the evapotranspiration obtained with the Dair_avg proceeding from 14 methods were evaluated, by using the lysimetric measures of a sward. These comparisons allowed to separate the Dar_avg calculating methods into three groups: a group always resulted into an overestimate ETo; in other one always tended to underestimating the ETo; and in the last tended to overestimating the low values (<4.5 mm day-1), while underestimating the high values of ETo. No significant statistical differences were observed (t test; p <0.05) in ETo as a function of the Dair_avg methods, when using the hourly average of the temperature and/or the air relative humidity in relation to those determined with the average of the maximum and minimum values of these elements. Among the methods proposed by paper FAO56, just the one using the average relative humidity to calculate the actual pressure (e a) showed a satisfactory estimate. The best estimates of ETo were obtained with the Dair_avg methods using the average of the air temperature in the determination of the saturation pressure (e s) and the average of the relative humidity for the actual pressure (e a). The use of saturation-vapor-pressure deficit of the air at a single time (from 9 or 10 a.m. local) as a representative of Dair_avg showed to be a satisfactory alternative in estimating the ETo under the climatic conditions of Piracicaba, SP, Brazil.
saturation vapor pressure; combination equation; sensitivity analysis
