EOF regression analytical model with applications to the retrieval of atmospheric temperature and gas constituents concentration from high spectral resolution infrared observations

The study describes and demonstrates a methodology for the statistical retrieval of temperature and gas species concentration that uses the spectral radiance measured by new generation of high-resolution satellite-borne infrared sensors. These include, e.g., AIRS (Atmospheric InfraRed Sounder) on AQUA satellite and IASI (Infrared Atmospheric Sounding Interferometer) on the European Meteorological Operational satellite. These spectrometers are characterized by a wide band spectral coverage (645 to 2700 cm-1 or 3.7 to 15.5 μm) and a spectral sampling rate in the range 0.25 to 2 cm-1. The performance of the retrieval scheme has been assessed on the basis of numerical exercises. Furthermore, examples of retrievals based on real spectra measured over sea surface are given to demonstrate the ability of the scheme to obtain accurate estimation of geophysical parameters. The problem of how many principal components to retain within the regression scheme has been addressed at a length and an original procedure is presented and discussed. Furthermore, the problem of statistical interdependency of retrieval and its vertical spatial resolution has been analyzed and a new index has been designed, which is capable to quantitatively deal with such an issue. The whole methodology has been derived for a generic signal-noise model and can therefore be used to design and implement retrieval algorithms also outside the specific area of high spectral resolution infrared observations. © 2009 by Nova Science Publishers, Inc. All rights reserved.