Evaluation of seasonal evapotranspiration of winter wheat in humid region of East China using large-weighted lysimeter and three models

摘要

Water consumption demand of winter wheat remains uncertain in humid region of East China. A better prediction of water consumption can improve agricultural water management. Based on the evapotranspiration (ET) data determined by large-weighted lysimeter and the meteorological data during 2011-2018, the actual ET of winter wheat was compared with the ET simulated by Priestley-Taylor (PT), Penman-Monteith (PM) and FAO- 56 single crop coefficient (K_c) models, and the relationships between ET of winter wheat and its driving environmental factors were assessed to localize the model's parameters. The results showed that the seasonal variations of ET in winter wheat could be divided into three stages with the canopy development, i.e., Stage I: from sowing to wintering when hourly ET maintained a low level (0.03-0.35 mm/h); Stage II: from revival to dough when hourly ET increased rapidly to 0.60-1.50 mm/h during daytime; Stage III: the ripening stage with hourly ET decreasing to 0.01-1.00 mm/h. The ET of winter wheat was mainly determined by wind speed at Stage I and by net solar radiation (R_n) at Stage II and III by Besides primary determining factor net radiation at the whole season, vapor pressure deficit and air temperature was the second factor that dominated ET at Stages III and Stage III, respectively. The PM model performed excellently for hourly ET modelling at Stage II with R~2 of 0.944 during 2017-2018. All models did not perform well at Stage I and III on hourly time scale. PM model was the best among the three models for calculating daily ET at Stage II, while both PM and PT models could be used to simulate daily ET at Stage III. After local parameterization, the calibrated PT_x and K_(cx) models were validated with daily ET simulation with the coefficients of 1.52 and 1.25 at Stage I, 1.81 and 1.80 at Stage II, and 1.12 and 0.78 at Stage III, respectively. The PT_x model was suitable to calculate daily ET at Stage I and III with R~2 of 0.19 and 0.55, MAE of