Development of Consistent Land Surface Emissivity Retrievals from a Combination of Microwave Sensors

Microwave-based precipitation retrievals rely on accurate values for surface emissivity especially over land due to their high variability and significant emission from the land surface. Using a suite of satellite remote sensing systems, including the Global Precipitation Mission Microwave Imager (GPM-GMI), the Advanced Microwave Scanning Radiometer 2 (AMSR2), AMSR-E, and the Moderate Resolution Imaging Spectroradiometer (MODIS) instrument on board Aqua and Terra satellites, surface emissivity values have been calculated for each of the AMSR-2, AMSR-E and GMI microwave bands. The objective of this study is to develop a consistent land surface emissivity for frequencies from 6.9GHz to 89.0 GHz that is free from discrepancies due to sensors configuration, footprint, and acquisition times. Among different microwave sensors, the varying overpass time of the GMI allows construction of brightness temperature (TB) diurnal cycles from consecutive days. Alternatively, the Special Sensor Microwave/Imagers (SSM/I) and AMSR2/AMSR-E instruments provide six and two overpasses, respectively, per day although at consistent times. The diurnal variation in the brightness temperature can be modeled from these sensors. The inconsistencies between these sensors can be alleviated when they were corrected using consistent GMI TB diurnal cycles, and consequently it led to a consistent blended land surface emissivity product from year 2014 to 2018. The resulting emissivity estimates can be used for a variety of hydrological applications including soil characteristics, particularly seasonal freeze and thaw (FT) rates of the land. Current FT rate products are generally derived using brightness temperature which is not independent of atmospheric effects, such as water vapor. It is shown that surface emissivity is unaffected by these atmospheric effects and is therefore a more appropriate product for the study of FT states.