Impact of aerodynamic temperature on ET estimates of Mediterranean irrigated crops from an energy-water balance model

The aerodynamic temperature identifies the temperature at an effective depth within the vegetation canopy, where sensible heat flux is originated, thus regulating the surface energy balance partitioning between latent and sensible heat. As it cannot be directly measured, a common practice in thermal infrared based retrieval of evapotranspiration is to substitute it with the radiometric temperature. However, numerous works have proven the two temperatures to be quite different, with peak differences as high as & PLUSMN;10 degrees C, with direct consequences in evapotranspiration modelling over complex landscapes. In this work, the main focus is on the integration of aerodynamic temperature within a hydrological modelling scheme, in order to identify possible improvements in the latent heat (and, thus, evapotranspiration) estimation. A first part of the analysis centres on comparing the differences between indirect estimates of aerodynamic temperature and measurements of radiometric surface temperature, showing divergences sometimes exceeding 10-15 degrees C. In a second step, a sensitivity analysis of the variable with respect to some environmental parameters was conducted. Meteorological parameters were found to be influential, whereas vegetation parameters played a relatively lesser role. In order to plug the aerodynamic temperature within the selected hydrological model (FEST-EWB), a test models was developed, testing the impact of the (measured) variable within the different turbulent fluxes. As the final aim was to continuously integrate the aerodynamic temperature within the hydrological model, possible parametrizations were scruti-nised, using the meteorological dependencies already investigated in the first part of the study. The newly-developed FEST-AeroT model, with a continuous parametrization of aerodynamic temperature, produced mild improvements in sensible and latent heat estimation and virtually no gain in net radiation. The model and parametrization uncertainties may be the reason why such a high a priori temperature divergence produced so small differences in final energy flux modelling.