Atmospheric Chemistry and Physics Discussions www.atmos-chem-phys-discuss.net 1680-7367 1680-7375 5 3 2005 10.5194/acpd-5-3313-2005 http://www.atmos-chem-phys-discuss.net/5/3313/2005/ http://www.atmos-chem-phys-discuss.net/5/3313/2005/acpd-5-3313-2005.html http://www.atmos-chem-phys-discuss.net/5/3313/2005/acpd-5-3313-2005.pdf 3313 3340 2005-05-25 Evidence of systematic errors in SCIAMACHY-observed CO₂ due to aerosols S. Houweling W. Hartmann I. Aben H. Schrijver J. Skidmore G.-J. Roelofs F.-M. Breon National Institute for Space Research (SRON), Utrecht, The Netherlands Institute for Marine and Atmospheric Research (IMAU), Utrecht, The Netherlands Laboratoire des Sciences du Climate et de l’Environnement, Gif sur Yvette, France SCIAMACHY CO₂ measurements show a large variability in total column CO₂ over the Sahara desert of up to 10% that is not anticipated from in situ measurements and cannot be explained by results of atmospheric models. Comparisons with colocated aerosol measurements by TOMS and MISR over the Sahara indicate that the seasonal variation of SCIAMACHY-observed CO₂ strongly resembles seasonal variations of windblown dust. Correlation coefficients of monthly datasets of colocated MISR aerosol optical depth and SCIAMACHY CO₂ vary between 0.6 and 0.8, indicating that about half of the CO₂ variance is explained by aerosol optical depth. Radiative transfer model calculations confirm the role of dust and can explain the size of the errors. Sensitivity tests suggest that the remaining variance may largely be explained by variations in the vertical distribution of dust. Further calculations for a few typical aerosol classes and a broad range of atmospheric conditions show that the impact of aerosols on SCIAMACHY retrieved CO₂ is by far the largest over the Sahara, but may also reach significant levels elsewhere. Inverse modelling calculations indicate that continental scale source and sink estimation on the basis of SCIAMACHY CO₂ data without aerosol correction leads to significant errors. To improve terrestrial CO₂ flux estimates by inverse modelling using SCIAMACHY measurements at 1.6μm, aerosol correction will be needed. Methods for correcting aerosol-induced errors exist, but so far mainly on the basis of theoretical considerations. As demonstrated by this study, SCIAMACHY may contribute to a verification of such methods using real data.