A variational land data assimilation system is used to estimate latent heat flux and surface control on evaporation. The dynamic equation for surface temperature with energy balance is used as a constraint on the estimation using the adjoint technique. Measurements of land surface temperature from satellite remote sensing are assimilated over two subregions within the Southern Great Plains 1997 hydrology field experiment. The performance of the estimation is linked to the timing of the satellite overpass. During days when the measurements close to the time of peak ground temperature are available, the estimation is adequate. The approach shows that satellite remote sensing of land temperature may be used to provide estimates of components of the surface energy balance and land surface control on evaporation. The latter parameter is related to surface soil moisture, and here they are compared with independent values derived from ground measurements. ¿ 2001 American Geophysical Union