Nadir Sounding of Carbon Gases using SCIAMACHY Near Infrared Channels

Abstract

The Beer InfraRed Retrieval Algorithm (BIRRA) and Column EstimatoR Vertical InfraRed Sounding Atmosphere (CERVISA) codes have been designed to retrieve vertical column densities (VCDs) of atmospheric gases in the near and thermal infrared (NIR,TIR) region of the electromagnetic spectrum by means of non-linear least squares fitting of radiances. As part of the operational SCIAMACHY level 1-2 processor, BIRRA is currently used for the specific retrieval of carbon monoxide (CO) VCDs exploiting the fitting window 4282-4301 cm-1 within the SCIAMACHY channel 8. Using appropriate fitting windows in channel 6, BIRRA also allows to gain information on greenhouse gases such as methane and carbon dioxide. Unfortunately, the increasing number of dead and bad pixels - specially in the NIR channels - reduces the available spectral information and consequently makes the VCDs retrieval more and more challenging. The proper choice of the pixel mask, fitting window, auxiliary fit parameters, as well as the filtering of the Level 2 data is crucial for obtaining a high quality atmospheric product. For validation of BIRRA the closely related CERVISA code is used to retrieve CO and CH4 from nadir infrared sounding data of AIRS, IASI, or TES. BIRRA and CERVISA share a large portion of modules, e.g., for line-by-line absorption and the nonlinear least squares solver; the essential difference is the part of the forward model devoted to radiative transfer through the atmosphere, i.e., Beer’s law for the NIR versus Schwarzschild’s equation for the TIR. CERVISA retrieval results are compared both to the operational products of the TIR sounder and to the SCIAMACHY-BIRRA product. In this work, we present recent results of carbon monoxide and methane retrievals.