Potential for the use of reconstructed IASI radiances in the detection of atmospheric trace gases
journal contributionposted on 2013-10-17, 11:07 authored by N.C. Atkinson, F.I. Hilton, Samuel Michael Illingworth, J.R. Eyre, T. Hultberg
Principal component (PC) analysis has received considerable attention as a technique for the extraction of meteorological signals from hyperspectral infra-red sounders such as the Infrared Atmospheric Sounding Interferometer (IASI) and the Atmospheric Infrared Sounder (AIRS). In addition to achieving substantial bit-volume reductions for dissemination purposes, the technique can also be used to generate reconstructed radiances in which random instrument noise has been reduced. Studies on PC analysis of hyperspectral infrared sounder data have been undertaken in the context of numerical weather prediction, instrument monitoring and geophysical variable retrieval, as well as data compression. This study examines the potential of PC analysis for chemistry applications. A major concern in the use of PC analysis for chemistry is that the spectral features associated with trace gases may not be well represented in the reconstructed spectra, either due to deficiencies in the training set or due to the limited number of PC scores used in the radiance reconstruction. In this paper we show examples of reconstructed IASI radiances for several trace gases: ammonia, sulphur dioxide, methane and carbon monoxide. It is shown that care must be taken in the selection of spectra for the initial training set: an iterative technique, in which outlier spectra are added to a base training set, gives the best results. For the four trace gases examined, key features of the chemical signatures are retained in the reconstructed radiances, whilst achieving a substantial reduction in instrument noise. A new regional re-transmission service for IASI is scheduled to start in 2010, as part of the EUMETSAT Advanced Retransmission Service (EARS). For this EARS-IASI service it is intended to include PC scores as part of the data stream. The paper describes the generation of the reference eigenvectors for this new service.
CitationAtmospheric Measurement Techniques, 2010, 3 (4), 991-1003
- VoR (Version of Record)