Utility of Radiometric–aerodynamic Temperature Relations for Heat Flux Estimation

In many land-surface models using bulk transfer (one-source) approaches, the application of radiometric surface temperature observations in energy flux computations has given mixed results. This is due in part to the non-unique relationship between the so-called aerodynamic temperature, which relates to the efficiency of heat exchange between the land surface and overlying atmosphere, and a surface temperature measurement from a thermal-infrared radiometer, which largely corresponds to a weighted soil and canopy temperature as a function of radiometer viewing angle. A number of studies over the past several years using multi-source canopy models and/or experimental data have developed simplified methods to accommodate radiometric–aerodynamic temperature differences in one-source approaches. A recent investigation related the variability in the radiometric–aerodynamic relation to solar radiation using experimental data from a variety of landscapes, while another used a multi-source canopy model combined with measurements over a wide range in vegetation density to derive a relationship based on leaf area index. In this study, simulations by a detailed multi-source soil–plant–environment model, Cupid, which considers both radiative and turbulent exchanges across the soil–canopy–air interface, are used to explore the radiometric–aerodynamic temperature relations for a semi-arid shrubland ecosystem under a range of leaf area/canopy cover, soil moisture and meteorological conditions. The simulated radiometric-aerodynamic temperatures indicate that, while solar radiation and leaf area both strongly affect the magnitude of this temperature difference, the relationships are non-unique, having significant variability depending on local conditions. These simulations also show that soil–canopy temperature differences are highly correlated with variations in the radiometric–aerodynamic temperature differences, with the slope being primarily a function of leaf area. This result suggests that two-source schemes with reliable estimates of component soil and canopy temperatures and associated resistances may be better able to accommodate variability in the radiometric–aerodynamic relation for a wider range in vegetated canopy cover conditions than is possible with one-source schemes. However, comparisons of sensible heat flux estimates with Cupid using a simplified two-source model and a one-source model accommodating variability in the radiometric-aerodynamic relation based on vegetation density gave similar scatter. On the other hand, with experimental data from the shrubland site, the two-source model generally outperformed the one-source scheme. Clearly, vegetation density/leaf area has a major effect on the radiometric–aerodynamic temperature relation and must be considered in either one-source or two-source formulations. Hence these adjusted one-source models require similar inputs as in two-source approaches, but provide as output only bulk heat fluxes; this is not as useful for monitoring vegetation conditions.