Evaluation of the surface renewal method to estimate wheat evapotranspiration

Abstract

The performance of the surface renewal method to estimate latent heat fluxes (LE) over a wheat crop was evaluated by comparison against values of LE measured independently using a weighing lysimeter. High-frequency temperature readings were taken at 1.5 m above ground from 29 April to 7 June 2000 over a 0.7–0.8 m high wheat crop. Surface renewal analysis was applied for two time lags r (0.75 and 0.25 s) to estimate half-hour sensible heat flux (H) and, subsequently, LE by solving the energy balance equation, using concurrent measurements of net radiation and soil heat flux. When H was estimated using sensor measurement height (z) in the computations, indices of agreement (IA) between lysimeter and surface renewal LE were above 0.94 and relative errors varied between 8.5 and 14.9% for time lag r=0.75 s for all analyzed days but 7 June. Results were slightly poorer for time lag r=0.25 s. When z−hc or z−d (hc being the crop height and d being the zero plane displacement) were used instead of z to compute H, surface renewal LE estimates slightly improved, particularly for the z−d case. The improvement was particularly noticeable for 7 June. The use of z−hc or z−d was thus more appropriate for these measurements, with the result that it was not necessary to calibrate the weighing factor α, as required by the standard surface renewal method. Unfortunately, although of similar magnitude than those reported for other micrometeorological methods, surface renewal errors found in this paper were biased and LE was underestimated. Further research and testing of the surface renewal method is therefore required to remove biases from the estimates of LE.