Satellite-based remotely sensed data were used to estimate regional-scale surface energy fluxes and a water deficit index of a semi-arid heterogeneous region in southeast Arizona. Spectral reflectance and radiometric temperature of the surface, derived from the digital counts of TM bands of LANDSAT-5 satellite, were used for this purpose. These reflectance and temperature, along with conventional meteorological information of the region, were used as inputs to numerical models which estimate surface energy fluxes. Point-based meteorological data of the region were spatially extrapolated over a grid of 120 m X 120 m so that it could be used with the spatially continuous remotely sensed data. The water deficit index (WDI) was estimated using surface temperature and a spectral vegetation index, "soil adjusted vegetation index" (SAVI). The surface fluxes were net radiation flux, sensible heat flux, soil heat flux and latent heat flux. Measured values obtained from the meteorological flux measurement (METFLUX) stations in the study area were compared with the modeled fluxes. Latent heat flux (LE) was the most important one to estimate in the scope of this study. The method of spatially extrapolating the point-based meteorological information and combining with the remotely sensed data produced good estimation of LE for the region, with a mean absolute difference (MAD) of 65 W/m² over a range of 67 to 196 W/m² . Also it was found that the numerical models that were previously used to estimate daily LE values from a region using mid-day remotely sensed data (mostly from NOAAAVHRR) can also be used with the mid-morning remotely sensed data (from LANDSAT). Out of the two models tested for this purpose (`Seguin-Itier' and 'Jackson' models), one was found to need some modification so that it could use mid-morning remotely sensed data as inputs. The other was found to be useable as it is, without any modification. Outputs from both models compared well with the measured fluxes from the METFLUX stations. In an effort of estimating the water deficit of the different biomes of the region, WDI of the biomes were estimated. The main goal of this effort was to be able to monitor the surface hydrologic conditions of the region using remotely sensed vegetation and surface information, and minimum ground data. Good estimation of the water deficit condition of the area were obtained by this method. This method was found to be sensitive to a few of the ground information such as wind speed and leaf area index (LAI). It was also found that if the required ground data were correctly estimated, this method could be used as an operational procedure for monitoring the vegetation water stress of the biomes and hence for better management of the region.