METEOSAT 8 SEVIRI and NOAA AVHRR Cloud Products. A Climate Monitoring SAF

Typ: Rapport
Serie: Meteorologi 127
Författare: Sheldon Johnston / Karl-Göran Karlsson


The goals of this study are to compare the MSG SEVIRI and PPS AVHRR monthly mean cloud products of the CM-SAF. The study was done in two parts: first comparing the cloud mask products and then comparing the cloud top temperature and height products. This was done over a region from Greenland to eastern Russia and as far south as the Sahara. The study covered four seasonally-representative months. For the cloud mask using PPS version 1.0, the results showed large problems over the Sahara and parts of Spain during the summer months. This was primarily due to the high reflectances in channel 3a and most prominent with NOAA 17. Much larger differences were found over water than over land surfaces, with the exception of Scandinavia where the differences were comparable to those found over water. The cloud-contaminated values were removed in one plot and this revealed that PPS had a larger number of cloud-contaminated pixels than MSG. This agrees with the concept that MSG reports increased cloudiness at higher viewing angles. This also explains why the differences over Scandinavia were so large and positive in value. The NOAA images at high latitudes have better spatial resolution and reports fewer cloudy and cloud-contaminated pixels than MSG. Sub-pixel and thin clouds greatly affected how well the two products converged. An attempt to use a weighted factor to adjust the effect of cloud-contaminated pixels on the total cloud cover failed to improve the convergence between the two cloud masks. The effect of the MSG viewing angle and the subsequent effects of reporting more cloudy pixels (or cloud-contaminated pixel – to include thin clouds) could be seen throughout all four months in the form of larger positive differences at latitudes approaching 80 degrees. Significant changes were seen with results from the PPS version 1.1. A significant decrease in the difference over the Sahara was the most discernable change. On the other hand, for NOAA 17, the agreement with MSG during twilight conditions was reduced by almost one half. The comparison of the cloud top temperature and height products revealed that MSG reported more low clouds during the summer months than PPS. This was mostly like due to the presence of convective clouds and the angle at which they are viewed (small cumulus clouds when viewed from nadir has a smaller diameter than when viewed slantwise).