Estimation Of Actual Evapotranspiration Of Maize Field Using Shuttleworth-Wallace Model

Maize is one of the major crops in semiarid areas in northwestern China. Estimating the evapotranspiration of the maize correctly and understanding the water consumption characteristics comprehensiely will play a key-role in precision irrigation. Based on field data of meteorological factors, soil water potential and crop growth state, we simulated evapotranspiration and transpiration of maize with Shuttleworth-Wallace model. Evapotranspiration simulated by Shuttleworth-Wallace model and Penman-Monteith model were tested with the measured value of eddy covariance system. Then we also verified the effect of Shuttleworth-Wallace model on simulating transpiration by compared with the value obtained by packaged sap flow gauges. Moreover, sensitivity of five resistance parameters had been analyzed. Results were as follows:(1) Based on the data measured by eddy covariance, we found that the energy closure of the study site was high, and the growing season was higher than the non-growing season. Besides, the footprint source area calculated with FSAM model suggested that our flux data adequately revealed the characteristics of surface flux in agro-ecosystem. Thus, it can be considered that eddy correlation system data of the experiment station is reliable.(2) The daily evapotranspiration simulated by Shuttleworth-Wallace model and Penman-Monteith model meet the reliability criteria, and the simulation results of dual-source model are similar to that of P-M model, however, the S-W results are slightly better than that of the PM model. The simulated values of the two models are both higher than the measured values of the eddy covariance method, and showed a more reasonable result when daily evapotranspiration is greater than 2.5mm/d in sunny or cloud day. And the hourly-scaled evapotranspiration estimated by models could be well simulated in sunny and cloudy days, while in the overcast and rainy days it presented relatively poor accuracy.(3) Transpiration simulated by dual-source model also reached a reliable simulation accuracy. In general, the transpiration of maize estimated by model is smaller than the value measured by sap flow gauges. And the hourly-scaled transpiration estimated by model could be well simulated in overcast and rainy days, while in the typical sunny days, there is a deviation between the simulated and measured values of maize transpiration rate. Otherwise, the dual-source model can simulated transpiration more satisfactorily in daily-scaled than in hourly-scaled.(4) The sensitivity analysis of resistance parameters showed that the evapotranspiration, the transpiration and evaporation had different sensitivity to resistance parameters, respectively. Evapotranspiration estimated by S-W model was most sensitive to the resistance parameters rac and raa; while the transpiration part was most sensitive to rac, and the soil evaporation part was most sensitive to rac, rss. Therefore, it is much more important to determine the resistance parameters rac, rss, and raa when using the empirical resistance formula.