Land surface temperature retrieval at high spatial and temporal resolutions over the southwestern United States

Land surface temperature ( LST) and its diurnal variation are important when evaluating climate change, land-atmosphere energy exchange, and the global hydrological cycle. These characteristics are observable from satellites using thermal infrared measurements, but doing so at both high spatial and temporal resolutions has been difficult. Accurate temporal and spatial knowledge of LST is critical in global-scale hydrological assimilation to improve estimates of soil moisture and evapotranspiration. Historically, satellite retrieval of global LST at high spatial resolutions ( 1 km) has relied on NOAA polar-orbiting satellites recently augmented by Moderate Resolution Imaging Spectroradiometer ( MODIS) data on board the Earth Observing System ( EOS). Each satellite instrument in a polar orbit typically provides one to two observations per day. High temporal sampling of LST is achievable with geostationary satellites but at spatial resolutions is too coarse to distinguish different land surface types ( 4-5 km) and with lower accuracy. We describe an approach which employs MODIS data as a calibration source for Geostationary Environmental Satellite ( GOES) data, then uses both data sets to yield half-hourly LST values, at 1 km spatial resolution, and returns LST with an accuracy better than 2 degrees C. The approach requires good cloud clearing, atmospheric correction, and an underlying LST model to propagate values between observations. Retrieved LST against ground truth data indicate the approach is accurate to about 2 degrees C.