Comparison of Different Complementary Relationship Models for Regional Evapotranspiration Estimation
基于互补相关原理的区域蒸散量估算模型比较

Based on meteorological and hydrological data in the Yellow River Basin covering 1981-2000, Advection-Aridity model, CRAE model and Granger model were validated at different temporal scales and in different climatic conditions by means of remote sensing and digital elevation model. Different climatic factors' effects on performance of models and variation characteristics of empirical parameters were analysed. Results show that with the exception of several extreme arid years, annual errors of Advection-Aridity model, CRAE model and Granger model were less than 10%. Monthly evapotranspirations from Advection- Aridity model were rational. However, CRAE model and Granger model overestimates monthly evapotranspirations during the winter. Spatial distribution of water balance closure errors for Advection-Aridity model is similar to that of Granger model. The closure errors for Advection-Aridity model and Granger model are less than that of CRAE model. In arid and humid climates or under more and less available energy conditions, Advection-Aridity model, CRAE model and Granger model did not perform very well.