Evaluating eddy covariance method by large-scale weighing lysimeter in a maize field of northwest China

Accurate determination of evapotranspiration (ET) is useful to develop precise irrigation scheduling. Although eddy covariance (EC) is a direct method which is widely used to measure ET, its performance in arid region of northwest China is not clear. In this study, ET measured by EC (ET_EC) was compared with that by large-scale weighing lysimeter (ET_L) during the whole growing season of maize in 2009. Energy balance ratio was 0.84 for daytime fluxes, indicating that lack of energy balance closure occurred, so daytime ET_EC was adjusted by Bowen-ratio forced closure method. Compared to the corresponding ET_L, half-hourly daytime ET_EC was underestimated by 21.8% without the adjustment and 4.8% with the adjustment. Furthermore, nighttime ET_EC was adjusted using filtering/interpolation method. Mean error between half-hourly nighttime ET_EC and ET_L decreased from 30.2% without the adjustment to 10.3% with the adjustment. After such adjustment of day and night measurements, daily ET_EC was underestimated by 6.2% compared to ET_L. These results indicated that the adjusted ET_EC well matched with the ET_L. Moreover, the discrepancy of adjusted total ET_EC and ET_L was decreased to 3.2% after subtracting the overestimated ET by lysimeter resulting from irrigation and heavy rainfall events. Thus, after appropriate adjustments of observations, eddy covariance method is accurate in estimating maize ET in the arid region of northwest China.