Influence of components of the advection-aridity approach on evapotranspiration estimation

Several methods have been proposed to estimate areal evapotranspiration from common meteorological data. The advection-aridity approach is one such method which requires no site-calibrated parameters, and uses measurements at a single elevation which are commonly available from a meteorological station. The method is based on a complementary relationship between actual and potential evapotranspiration which postulates that a decrease in actual evapotranspiration will result in a complementary or symmetrical increase in potential evapotranspiration for a given energy input. In this paper, the advection-aridity equations were rewritten to isolate terms which account for available energy, advection or drying power of the air, and sensible heat flux. The advection-aridity method was tested on a daily basis with data for 43 days collected from a flat, semi-arid grassland over a 4 month period. The influence of each of the terms was examined to determine its relative influence in generating the complementary or symmetrical structure between actual and potential evapotranspiration. Advection and sensible heat flux were found to be very significant, whereas available energy was not. A comparison between reference values of evapotranspiration Een obtained from the energy budget with measurements of net radiation, ground heat flux, and sensible heat fluxes obtained by the eddy correlation method, and evapotranspiration estimates from the advection-aridity method, Eaa, produced a small root mean square error of 13.1 W m-2. However, significant bias was present in that Eaa overestimated Een for large values of Een, but underestimated Een for small values. Measurement error and external energy sources not accounted for in the complementary relationship or the advection-aridity approach are discussed as potential causes of this bias.