Evaluation of Empirical Remote Sensing-Based Equations for Estimating Soil Heat Flux

Abstract

Accurate estimation of evapotranspiration from remotely sensed data typically requires accurate estimation of net radiation, sensible heatflux, and soil heatflux (G). The focus of this paper is on evaluating the accuracy of commonly used empirical soil-heat-flux-estimating equations. The equations were applied to estimate the ratio (Γ) of G ton et radiation using 250-m resolution Moderate Resolution Imaging Spectroradiometer (or MODIS) vegetation index data from the Terra and Aqua satellites and ground measurements at three sites in China that represent humid, semi-arid, and arid climates. The results revealed the following: equations that determine Γ on the basis of the vegetation index alone give highly inaccurate results; information on soil moisture condition is critical to obtain more reliable Γ estimates; soil moisture condition of bare soil is related to its albedo in a log-linear way. A new equation has been developed that considers the effect of soil moisture condition through the albedo of bare soil; this equation has been shown to produce highly accurate results over a wide range of climate and land cover types. It is expected that the proposed G-estimating equation could be embedded in remote sensing evapotranspiration algorithms in future studies.