Many recent studies have applied satellite remote sensing data to large-scale hydrologic and biospheric modeling. It is widely accepted that the thermal infrared observations from the Advanced Very High Resolution Radiometer (AVHRR) have the potential to estimate land surface conditions, such as surface temperature, near surface air temperature, and near surface water vapor. In this study, algorithms to estimate all three variables are presented and applied to an area covering the state of Oklahoma for a six day period in August, 1994. The results were validated using ground observations from the 111 station Oklahoma Mesonet. Validation of the remote sensing algorithms with Mesonet observations produced comparable results to previous validation studies. In addition, the validation process revealed inadequacies in thermal modeling that had not been detected in previous validation studies leading to the development of a new approach to estimate atmospheric water vapor.