Land surface temperature (LST) from thermal remote sensing is a surface boundary condition that is strongly linked to the partitioning of the available energy between latent (evapotranspiration) and sensible heat flux. Numerous modeling approaches have been developed ranging in level of complexity from semi-empirical to numerically-based soil-vegetation-atmosphere schemes. Many of the approaches require an accurate LST because the heat fluxes are related to the surface-air temperature differences. There is also difficulty estimating appropriate exchange coefficients for heterogeneous landscapes having a mixture of soil and vegetation temperatures influencing the LST observation and associated aerodynamic temperature. For regional applications this also means requiring an accurate air temperature distribution over the area of interest. These requirements have rendered many of the modeling approaches unusable for routine applications over complex land surfaces. However a two-source energy balance (TSEB) modeling scheme using time differencing in LST observations coupled to an atmospheric boundary layer growth model has been developed to adequately address the major impediments to the application of LST in large scale evapotranspiration determination. The modeling system, Atmospheric Land EXchange Inverse (ALEXI), using geostationary LST observations and the disaggregation methodology (DisALEXI) together with data fusion techniques will be described. This modeling system is currently providing regional and continental scale evapotranspiration estimates in the U.S. and plans are to develop a global product.