European Geosciences Union, Atmospheric Measurement Techniques Discussions, p. 1-42
DOI: 10.5194/amt-2016-79
Full text: Download
Surface skin temperature ( T s ) is an important parameter for characterizing the energy exchange at the ground/water-atmosphere interface. The Satellite ClOud and Radiation Property retrieval System (SatCORPS) employs a single-channel thermal-infrared- (TIR-) method to retrieve T s over clear-sky land and ocean surfaces from data taken by geostationary-Earth orbit (GEO) satellite and low-Earth orbit (LEO) satellite imagers. GEO satellites can provide somewhat continuous estimates of T s over the diurnal cycle in non-polar regions, while polar T s retrievals from LEO imagers, such as the Advanced Very High Resolution Radiometer (AVHRR) can complement the GEO measurements. The combined global coverage of remotely sensed T s , along with accompanying cloud and surface radiation parameters, produced in near-real time and from historical satellite data, should be beneficial for both weather and climate applications. For example, near-real-time hourly T s observations can be assimilated in high-temporal resolution numerical weather prediction models and historical observations can be used for validation or assimilation of climate models. Key drawbacks to the utility of TIR-derived T s , data include the limitation to clear-sky conditions, the reliance on a particular set of analyses/reanalyses necessary for atmospheric corrections, and the dependence on viewing angle. Therefore, T s validation with established references is essential, as is proper evaluation of T s sensitivity to atmospheric correction source. This article presents improvements on the NASA Langley GEO satellite and AVHRR TIR-based T s product, derived using a single-channel technique. The resulting clear-sky skin temperature values are validated with surface references and independent satellite products. Furthermore, an empirical means of correcting for the viewing-angle dependency of satellite land surface temperature (LST) is explained and validated. Application of a daytime nadir-normalization model yields improved accuracy and precision of GOES-13 LST relative to independent Moderate-resolution Imaging Spectroradiometer (MYD11_L2) LST and Atmospheric Radiation Measurement Program/NOAA ESRL Surface Radiation network ground stations. These corrections serve as a basis for a means to improve satellite-based LST accuracy, thereby leading to better monitoring and utilization of the data. The immediate availability and broad coverage of these skin temperature observations should prove valuable to modelers and climate researchers looking for improved forecasts and better understanding of the global climate model.