| description abstract | The cloud mask is an essential product derived from satellite data. Whereas cloud analysis applications typically make use of information from cloudy pixels, many other applications require cloud-free conditions. For this reason many organizations have their own cloud masks tuned to serve their particular needs. Being a fundamental product, continuous quality monitoring and validation of these cloud masks are vital. This study evaluated the performance of the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT) Meteorological Products Extraction Facility cloud mask (MPEF), together with the Nowcasting Satellite Application Facility (SAFNWC) cloud masks provided by Météo-France (SAFNWC/MSG) and the Swedish Meteorological and Hydrological Institute (SAFNWC/PPS), in the high-latitude area of greater Helsinki in Finland. The first two used the Spinning Enhanced Visible and Infrared Imager (SEVIRI) instrument from the geostationary Meteosat-8 satellite, whereas the last used the Advanced Very High Resolution Radiometer (AVHRR) instrument on board the polar-orbiting NOAA satellite series. Ceilometer data from the Helsinki Testbed, an extensive observation network covering the greater Helsinki area in Finland, were used as reference data in the cloud mask comparison. A computational method, called bootstrapping, is introduced to account for the strong temporal and spatial correlation of the ceilometer observations. The method also allows the calculation of the confidence intervals (CI) for the results. This study comprised data from February and August 2006. In general, the SAFNWC/MSG algorithm performed better than MPEF. Differences were found especially in the early morning low cloud detection. The SAFNWC/PPS cloud mask performed very well in August, better than geostationary-based masks, but had problems in February when its performance was worse. The use of the CIs gave the results more depth, and their use should be encouraged. | |
