Mapping Evaporation Fields With Satellite Sequential Observations of Surface Temperature

Land evaporation is a flux that links the surface and the atmosphere as well as the surface and the subsurface. Furthermore land evaporation is a flux that links the water, energy, and biogeochemical cycles. Despite its importance there are no large-scale observations of land evaporation outside of a few limited-duration field experiments and a very sparse network of tower-mounted instruments. While space-borne sensors are capable of global mapping they cannot measure evaporation directly. Indirect methods have been used to estimate evaporation from satellite data but they mostly suffer from empiricism. An alternative approach is proposed here that has several distinct advantages: 1) use of the dynamic surface energy balance with minimal requirements for auxiliary data compared to empirical approaches, 2) capability to assimilate multi-scale data at their native resolutions, 3) global applicability through the use of existing satellite data, and 4) retrieval of the key unknown controls on evaporation that are separated by time scales of variation (evaporative fraction on daily scale and forced convection parameters on multiple-days).

By assimilating the diurnal cycle of LST into a model of the dynamic surface energy balance we can estimate evaporative fraction. In a variational assimilation framework the LST measurement-model misfit is minimized while the dynamic surface energy balance is imposed as a strong constraint. Adjoint techniques allow the development of Euler-Lagrange equations for iterative optimization. The Hessian of the objective function yields a measure of the uncertainty in the evaporation estimates.

Another purpose of this proposal is to address a fundamental issue of how water and energy are couple in terrestrial systems. All models of the terrestrial water and energy balance – whether they are used in predictive mode to analyze consequences of climate variations and global change, or used in assimilation mode to develop value-added data products based on satellite measurements –either implicitly or explicity represent coupled heat and moisture diffusion in the soil-vegetation-atmosphere continuum. How various models perform is highly dependent on the nature of these couplings. Important as these links are, they remain largely unvalidated – especially at the global scale.