[1] We present an operational method for cloud pressure retrieval from the Earth's reflectance spectrum in the visible, using the O 2 -O 2 absorption band at 477 nm. The algorithm is simple and robust. Apart from cloud pressure, an effective cloud fraction is also retrieved. Using simulations and Global Ozone Monitoring Instrument (GOME) data the accuracy of the O 2 -O 2 retrieval method is estimated. The Ozone Monitoring Instrument (OMI), to be space-borne on board the EOS-AURA platform in 2004, will use this algorithm to produce an official cloud product. The cloud product will be used to support the cloud correction of several of the OMI trace gas retrievals.

[1] A model for the vertical variation of microphysical and optical properties of single-layer water clouds is used to design a procedure to obtain cloud droplet concentration from satellite cloud optical thickness/effective radius retrievals. The model allows for smooth vertical variations in microphysical variables including droplet concentration and liquid water content. The procedure is applied to data from the MODIS instrument aboard the EOS-TERRA satellite platform over a region near Cape Grim, Tasmania, over the Southern Ocean. At this unpolluted location, there are seasonally repeated well-described variations in cloud condensation nucleus concentration. The satellite observations show that there is a factor 2.5 difference in retrieved droplet concentration between the summer and winter seasons as measured over the 4 year period (July 2000 to July 2004). Comparison of these results with cloud condensation nuclei concentrations observed at Cape Grim showed good agreement. Furthermore, for a fixed solar zenith angle cos (45°) cloud albedo varied between 0.45 and 0.55 and exhibited a clear correlation with the retrieved droplet concentration (correlation coefficient R = 0.80). Such correlation was absent between albedo and the second retrieval product, namely the cloud depth.

We present a comparison of the Scanning Imaging Absorption Spectrometer for Atmospheric Chartography (SCIA-MACHY) and the Medium Resolution Imaging Spectrometer (MERIS) reflectance calibrations at wavelengths between 442 and 885 nm. Results show that SCIAMACHY underestimates the reflectance by 13% at 442 nm reaching up to 21% at 885 nm as compared to MERIS. We also find a small nonlinear effect for the lowest reflectance. This effect is more evident for pixels over ocean than over land.

Abstract. The calibration of SCIAMACHY was thoroughly checked since the instrument was launched on-board EN-VISAT in February 2002. While SCIAMACHY's functional performance is excellent since launch, a number of technical difficulties have appeared, that required adjustments to the calibration. The problems can be separated into three types: (1) Those caused by the instrument and/or platform environment. Among these are the high water content in the satellite structure and/or MLI layer. This results in the deposition of ice on the detectors in channels 7 and 8 which seriously affects the retrievals in the IR, mostly because of the continuous change of the slit function caused by scattering of the light through the ice layer. Additionally a light leak in channel 7 severely hampers any retrieval from this channel. (2) Problems due to errors in the on-ground calibration and/or data processing affecting for example the radiometric calibration. A new approach based on a mixture of onground and in-flight data is shortly described here. (3) Problems caused by principal limitations of the calibration concept, e.g. the possible appearance of spectral structures after the polarisation correction due to unavoidable errors in the determination of atmospheric polarisation. In this paper we give a complete overview of the calibration and problems that still have to be solved. We will also give an indication of the Correspondence to: G. Lichtenberg (guenter.lichtenberg@dlr.de) effect of calibration problems on retrievals where possible. Since the operational processing chain is currently being updated and no newly processed data are available at this point in time, for some calibration issues only a rough estimate of the effect on Level 2 products can be given. However, it is the intention of this paper to serve as a future reference for detailed studies into specific calibration issues.

No abstract