According to U.S. drought experts, a “flash drought” is considered to be a short-term, yet severe event, characterized by moisture deficits and abnormally high temperatures that negatively impact vegetation condition. Drought monitoring and modeling that depend on time-series satellite-derived vegetation index imagery, such as the Vegetation Drought Response Index, may fail to adequately detect rapid onset or short duration events due to the lag in vegetation reflectance response to water deficits. Changes in plant canopy temperature, on the other hand, occur from the beginning of an episode of drought-induced stress. In this study we applied a Simplified Surface Energy Balance (SSEB) model to an area in the United States Southern Great Plains to identify the occurrence, spatial distribution, and severity of flash drought. The SSEB model utilizes the Moderate Resolution Imaging Spectroradiometer (MODIS) 8-day 1-km land surface temperature (LST) data to calculate a thermal-based evapotranspiration (ET) fraction. The fraction, based on temperature differences between “hot” and “cold” pixels, was used in conjunction with global scale daily reference ET data to estimate both 8-day and seasonal water use (actual ET (ETa)) from each pixel. We concentrated our initial efforts on the 2006 growing season (May – September) which was identified by the National Drought Mitigation Center (NDMC) as a period when a flash-drought was experienced throughout Oklahoma, Kansas, and the northern Texas Panhandle. The occurrence and magnitude of flash droughts were characterized by comparing the water-use patterns (ETa) of the irrigated and non-irrigated areas with the premise that the irrigated areas are less affected by short-term hot and dry spells compared to the non-irrigated areas. Preliminary SSEB model results were mapped and compared with daily precipitation and temperature data. Results showed that the SSEB model captured both the temporal and spatial distributions of the flash drought that occurred in the southern Great Plains in 2006. Further improvements and validation of the model will include a correction for topographic-induced land surface temperature differences. We believe that the SSEB modeling approach using the MODIS LST data will have adequate sensitivity, and will improve our ability to identify quick developing drought situations (i.e., flash droughts). Operational application of this model is envisaged for monitoring and detecting landscape-scale drought occurrences in the conterminous United States.