EVALUATION OF TWO METHODS FOR DAILY EVAPOTRANSPIRATION ESTIMATION FROM FIELD AND MODIS DATA

Daily evapotranspiration (ET) is considered more significant in many practical applications, compared to the instantaneous ET obtained from remote-sensing based models. The constant reference evaporative fraction (EFr, the ratio of actual to reference ET) method is one of the well preformed upscaling methods used to extrapolate instantaneous ET to daily timescales. The constant decoupling coefficient (Omega) method requires similar input data to the constant EFr method and can be used to calculate daily ET directly. This study evaluated the performances of the two methods underlying the estimation of daily ET. The results from field data only showed that (i) daily ET were both overestimated by two methods when compared to the uncorrected Eddy covariance (EC) measurements; (ii) the estimated daily ET had a good agreement with the measurement corrected by the Bowen Ratio (BR) method. Based on MODIS and filed data and when the ET measurements were corrected by the BR method, the results showed that (i) the constant EFr method overestimated daily ET by a bias of 5.6 W/m(2) and a root mean square error (RMSE) of 18.6 W/m(2); (ii) the constant Omega method underestimated daily ET by a smaller bias of -4.8W/m(2) and a RMSE of 22.5 W/m(2). Therefore, the constant Omega method had a similar performance with the constant EFr method, and could be applied to estimate daily ET effectively.