CO2 retrievals from the Atmospheric Infrared Sounder: Methodology and validation

[2] Although it was designed for high resolution/accurate temperature and moisture profiles, the National Aeronautics and Space Administration Earth-Observing System (NASAEOS) Atmospheric Infrared Sounder (AIRS) is capable of measuring variations in carbon trace gases such as CO2 [Chédin et al., 2003; Crevoisier et al., 2003; Engelen and Stephens, 2004; Aumann et al., 2005]. This capability coupled with the AIRS broad swath pattern, low and well characterized instrument noise, and global coverage afforded by a method termed cloud-clearing, enables derivation of the distribution of CO2 (as well as other trace gas species) in the middle-to-upper troposphere on global scales twice per day. [3] Numerous studies [Engelen et al., 2004; Crevoisier et al., 2004; Chahine et al., 2005] have shown that retrievals from AIRS show expected seasonal and latitudinal variability in the tropics as compared to JAL Matsueda flask data [Matsueda et al., 2002]. Engelen and McNally [2005] extend some of the results to higher latitudes using flask measurements from the National Oceanic and Atmospheric Administration (NOAA) Environmental Systems Research Lab/Global Monitoring Division (ESRL/GMD) formerly known as the Climate Monitoring Diagnostics Laboratory (CMDL) aircraft network. Nevertheless, attempts to use retrievals to constrain atmospheric inversions of CO2 surface fluxes [Chevallier et al., 2005] have been for the most part unsuccessful and comparison to models [Tiwari et al., 2006] have raised questions concerning the ability of models to correctly reproduce large scale circulation pathways of atmospheric CO2 and other atmospheric tracer species. Simultaneous derivation of atmospheric concentrations of CO, CH4 [Xiong et al., 2008], and O3 in the middleto-upper troposphere as well as high vertical resolution temperature and moisture profiles will enable better constraint on model transport and vertical mixing and warrant more study of the capabilities of the AIRS instrument in deriving CO2 abundances. [4] In this paper, we apply the methodology of Susskind et al. [2003] to the retrieval of CO2 from cloud-cleared radiances. In the section 2 we describe the methodology of the NOAA algorithm, and in section 3, we compare these retrievals to an extended set of NOAA ESRL/GMD aircraft measurements obtained during 2005 (C. Sweeney, private communication, 2006).