Analytical framework based on thermodynamics to estimate spatially distributed surface energy fluxes from remotely sensed radiations

Abstract

The Limited surface observations of turbulent heat fluxes result in incomplete knowledge about the surface energy balance that drives the climate system. The highly parameterized surface energy balance models suffer from significant uncertainties. Remote sensing information of incoming and outgoing radiation fluxes are important input variables for turbulent heat flux models, though analytical solutions of surface energy budget from these variables are yet to be derived. Here, we developed a novel, purely physics-based analytical method grounded on the thermodynamic principle of maximum power. The approach derives the total turbulent heat flux only from the four inputs of incoming and outgoing longwave and shortwave radiations at the land surface, which are available from remotely sensed satellite data. The proposed approach does not use any parameterization, unlike the existing surface energy balance models, and hence does not suffer from uncertainty due to the same. We validated our methodology with 102 eddy covariance observation stations around the globe with different land use land covers from FLUXNET2015 and available urban datasets. Based on monthly averages of the total turbulent flux estimates for the eddy covariance sites, we observed root-mean-square error (RMSE) of 23.2 ± 10.9 Wm−2, a mean bias error (MBE) of 11.9 ± 13.1 Wm−2 and R2 value of 0.86 ± 0.15. Using the satellite observations of radiation fluxes from CERES at a spatial resolution of 10, we obtained the global flux field of total turbulent flux (QJ) and land surface heat storage (ΔQs) fluxes. On validation of QJ with FLUXNET sites for six grids for different land use land cover, we found RMSE of 20.1 ± 7 Wm−2, MBE of 14.4 ± 9 Wm−2 and R2 value of 0.96 ± 0.02. Further, from the evaporative stress factor of GLEAM, which is based on microwave remotely sensed vegetation optical depth and root zone soil moisture, we have obtained spatially distributed global estimates of Sensible(H) and latent heat (LE). In addition, our analytical estimates address the distribution of residual energy associated with the surface energy balance closure problem driven by the land use land cover. The theoretical estimates of all surface energy balance components from remote sensing based observations will improve our understanding of surface warming for different land use land covers across the globe.