Evaluation of applicability of winter wheat evapotranspiration estimation model in Northwest China
ZHANG Fujuan1, CUI Ningbo1, 2, 3*, ZHAO Lu1, CAI Huanjie2, HU Xiaotao2, ZHANG Nian1, XIAO Lu1, HU Haibo4, YANG Dewen4
1. State Key Laboratory of Hydraulics and Mountain River Engineering, Sichuan University, Chengdu, Sichuan 610065, China; 2. Institute of Water-saving Agriculture in Arid Areas of China, Northwest A&F University, Yangling, Shaanxi 712100, China; 3. Provincial Key Laboratory of Water-Saving Agriculture in Hill Area of Southern China, Chengdu, Sichuan 610066, China; 4. Sichuan Panzhihua Academy of Agricultural and Forestry Sciences, Panzhihua, Sichuan 617061, China
Abstract：In order to accurately estimate the evapotranspiration(ET)of Winter Wheat in Northwest China, the double crop coefficient model, single crop coefficient model and Priestley Taylor(PT)model are applied to simulate ET, and the accuracy is compared with the standard value of ET measured by large lysimETer. The results show that mETeorological factors are the main factors of ET from sowing to green(stage I)and heading to milk maturity(stage III), and crop factors are the dominant factors of ET from milk maturity to harvest(stage IV). The effects of the two factors on ET from turning green to heading(stage II)and the whole growth period are similar. The R2 of two crop coefficient model, single crop coefficient model and PT model are 0.511 8, 0.239 3 and 0.374 2 respectively in stage Ⅰ. The RMSE ranges from 0.284 6 mm/d to 0.366 3 mm/d, and the overall evaluation index GPI rankings are 1, 3 and 2. The R2 of the three models in stage II are all above 0.7, and the RMSE ranges from 0.540 9 to 0.844 0 mm/d. The dual crop coefficient model has the best si-mulation effect. The R2 of each model in stage Ⅲ is higher than 0.6, and the RMSE ranges from 0.8288 to 1.258 7 mm/d. The GPI of the dual crop coefficient model ranks first. The R2 of the three models in stage IV are 0.799 1, 0.671 6, 0.270 8, and the RMSE ranges from 0.968 1 to 1.946 2 mm/d. The simulation accuracy of the crop coefficient model is significantly higher than that of the PT model. The RMSE of each model ranges from 0.551 5 to 0.893 6 mm/d in the whole growth period, and the R2 of double crop coefficient model is 0.902 2.
