A methodology to compare cloud properties simulated by global climate models with those retrieved from observations by satellite-based infrared (IR) sounders has been developed. The relatively high spectral resolution in the CO2 absorption band of these instruments leads to especially reliable cirrus properties, day and night. Additionally, bulk microphysical properties can be retrieved for semi-transparent cirrus, based on the observed spectral emissivity differences between 8 and 11 mu m. The particular intention of this study is to compare macro-and microphysical properties of high cloudiness as represented by the model simulations and the satellite data. For this purpose, a method has been developed to process the model output to be comparable to the satellite measurements, as in other observational simulator packages (for example the ISCCP-simulator). This simulator method takes into account i) the differences in horizontal resolution of the model and the observations, ii) the specific observation time windows, iii) the determination of the pressure of a cloud system, identified with the pressure at the middle of the uppermost cloud, and iv) the selection of high clouds with specific cloud optical thickness ranges for the microphysical property retrieval using IR sounder data. Applying this method to simulations by the global climate model ECHAM and TOVS satellite observations has important effects. The frequency of high clouds selected from the model output by using the method is significantly smaller than the total frequency of high cloudiness in the model. Largest differences occur around the equator where the zonal mean frequency of high cloudiness is reduced by about 30 % (relative change). The selection method is essential for the comparison of modelled and observed microphysical properties of high clouds. The selection of high clouds from the ECHAM simulation according to the optical thickness range of the TOVS data results in a reduction of the mean water path of high clouds by factors of more than 3 compared to the case where also high clouds of other optical thicknesses are considered. Furthermore, the selection by optical thickness causes a significant increase in the mean effective cloud particle diameter. These changes significantly reduce the differences between the simulation and the observations. The method can also be applied for comparisons with other IR sounder climatologies such as from AIRS and IASI.
