Determining the microwave emissivity of sea-ice at AMSU-B frequencies for use in operational numerical weather prediction assimilation schemes

摘要

Possible methods for improving the way in which the emissivity of sea ice is represented in numerical weather prediction (NWP) assimilation schemes, and hence allowing increased exploitation of satellite data in these traditionally data sparse regions, are investigated by comparing emissivities retrieved from in-situ, airborne observations in the arctic and those derived from two different models. Further information is provided by sea ice type and concentration products derived from Special Sensor Microwave Imager (SSM/I) observations.rnAnalysis of the aircraft derived emissivities shows a definite relationship between the emissivities at 157 GHz and 183 GHz irrespective of the ice type (R~2 = 0.97), and while this is not the case for the window channel at 89 GHz and the two higher frequencies (R~2 = 0.55), it will be shown that by sub setting the emissivities according to ice type results in a stronger relationship between these frequencies for first year ice (R~2 = 0.66 between 89 and 183 GHz) while the relationships for multi year ice are weaker.rnThe retrieved emissivities are also compared with two emissivity models, the fast emissivity model (FASTEM) which assigns an emissivity to a prescribed ice type and a model from the NOAA (National Oceanographic and Atmospheric Administration) which uses an empirical approach based on retrievals from Advanced Microwave Sounding Unit (AMSU) brightness temperatures. The NOAA model performed better than FASTEM, although FASTEM performance was improved by weighting the emissivity spectra by the sea ice type and concentration information given by the SSM/I derived products.