On the attribution of changing crop evapotranspiration in arid regions using four methods

Accurate evaluation of the relative contribution of climatic and vegetative drivers on land evapotranspiration (ET) is critical for interpreting the ET controlling mechanism, modeling and predicting. However, how to accurately separate and estimate the contribution of climatic and vegetative changes on ET variation remains uncertain. Our study attempted to interpret the interannual variation in ET during 2007–2013 over a maize field in northwest China, using four methods simultaneously, such as the Penman-Monteith (PM) model, the modified crop coefficient method, the Priestly-Taylor (PT) model and the regression linear model. Results indicate that compared to the ET in 2007, the ET decreased averagely by 12, 30, 29, and 73 W m⁻² in 2008, 2011, 2012 and 2013, respectively. All models yielded similar results and showed that the vegetative controls played a more important role in regulating ET relative to climatic drives, and that more than half of decrease in ET was caused by vegetative factors, while the differences in net radiation, water vapor pressure deficit and air temperature among years were lesser source of variation in ET. Furthermore, the advantage and disadvantage of the four methods were discussed. Our study confirmed the great effect of vegetative drivers in regulating crop ET, highlighted the importance of estimating canopy conductance accurately in ET modeling and prediction, and provided new approaches for separating the climatic and vegetative contribution on ET changes.