Revista mexicana de ciencias agrícolas
versión impresa ISSN 2007-0934
VASQUEZ-MENDEZ, Rebeca; VENTURA-RAMOS, Eusebio Jr. y ACOSTA-GALLEGOS, Jorge A.. Suitability of evapotranspiration estimation methods for semiarid central Mexico. Rev. Mex. Cienc. Agríc [online]. 2011, vol.2, n.3, pp.399-415. ISSN 2007-0934.
In arid and semiarid areas, the evapotranspiration process constitutes a major loss of water to the atmosphere, prompting the need for a better understanding and quantification of this phenomenon, especially in semiarid environments where water resources are scarce. There are many available methods to estimate evapotranspiration; however, their performance in specific environments must be evaluated before making a selection. A weather station was used to collect data from a semiarid zone of Central Mexico with the purpose to evaluate eight different methods for daily, weekly and monthly periods of observation. The estimated results were compared with measured data from an ET gage company, as this device has been calibrated and proven to be closely associated to reference evapotranspiration data. The mean absolute error, root mean squared error, regression line intercept and slope, as well as determination coefficient R2, were obtained to statistically evaluate the performance of evaluated methods. A poor performance for all time scales was found for the Oudin, McGuinness, Jensen and Haise, and Romanenko methods, with high values of mean absolute error and root mean squared error. Low values of the intercept, slope and R2 parameters were also obtained for such methods, making them not suitable and consequently not suited for the study area. On the other hand, the Penman-Monteith and Penman methods, showed the best performance in different time scales. Similar results have been fund by other authors, supporting Penman-Monteith as a method of reference. A multiple regression equation to predict reference from climatic variables was also obtained with R2= 0.8, 0.82 and 0.91 for daily, weekly and monthly scales, respectively, indicating the possibility of using regression models for semiarid conditions.
Palabras llave : climatic variables; evapotranspiration; semiarid environment.