Trace gas column retrieval – an error assessment study for GOME-2 Original Research Article

As a follow-on of the successful GOME instrument aboard ERS-2 the GOME-2 series is planned to operate on METOP satellites after 2005. Focussing on meteorological applications GOME-2 is designed to measure the total column and profiles of atmospheric ozone and the distribution of other key atmospheric constituents like BrO and NO2. The main differences to Global Ozone Monitoring Experiment (GOME) are the higher spatial resolution of 40×80 km2 in combination with a larger swath width of 1920 km. In this study potential error sources have been analysed most relevant for high viewing angles or small ground pixel sizes as applicable to GOME-2. Instrumental and atmospheric error sources have been identified with respect to basic signal-to-noise errors based upon an instrument model and full-spherical atmospheric radiative transfer simulations. Potential error sources such as the reflectance of the atmosphere, refraction, bi-directional surface reflection, spatial aliasing effects due to inhomogeneities of the scene, and spectral resolution (undersampling, slit width) have been investigated. Effects due to changes in instrumental design, e.g. read-out time, slit width and spectral sampling, have been taken into account. Error sources identified as most significant for trace gas column retrieval are spectral signature in the diffuser plate reflectivity, sun-glint effects, undersampling residuals, and noise. An overview of the study results and conclusions for future GOME-2 retrieval and instrumental designs are presented.