Evapotranspiration Partitioning and its Characteristics for Winter Wheat Based on the Concept of Underlying Water Use Efficiency in the Loess Tableland of China

Evapotranspiration (ET) partitioning is one of the most important issues in water transport process of terrestrial SPAC and water resources management. The underlying water use efficiency (uWUE) model could effectively separate plant transpiration (T) from ET and solve problems of the mismatch among soil evaporation (E), T and ET in spatial scales and the difficulty in continuous measuring of T and E in the now available experiment observations. T/ET was estimated by the ratio of an apparent uWUE (uWUE_a = GPP·VPD^0.5/ET) to a potential uWUE(uWUE_p = GPP·VPD^0.5/T). This study partitioned the ET of winter wheat on the Loess Tableland using eddy covariance data and analyzed the characteristics of T/ET at different time scales during growing seasons. In order to solve the problem of uWUE_a anomaly on a half-hour scale, the upper limit of uWUE_a was chose as data-filter threshold based on the requirement of average daily T/ET not more than 1. The planting system of Loess Tableland is one harvest a year, and winter wheat was grown continuously in the experimental years. The growing season average T/ET of winter wheat in the years of 2005-2006, 2007-2008 and 2008-2009 are 0.50, 0.67and 0.58, respectively. The dynamic change of five days average of T/ET within the growing season showed a bimodal pattern, while the ET rate was low before regreening stage and then increased rapidly. ET rate reached the maximum at booting stage, which was 5.48, 5.66 and 6.23mm/day, respectively. The first peak of T/ET appeared in the tillering stage with the average of 0.56, and the ET for the corresponding time were 0.76, 0.46 and 0.39mm/day, respectively. The second peak occurs around the booting stage and exceeded 0.80 in all three years, and the ET for the corresponding time were 3.02, 3.11 and 3.81 mm/day, respectively.