Statistical characteristics of cloud cover in the Mediterranean-Black Sea region from satellite data

An archive of cloud amount in the Mediterranean-Black Sea region in 1984-2009 has been formed using satellite data from the Meteosat satellites. The procedures for decoding satellite images and automatical identifying of the cloud amount from single-channel satellite shots made by the Meteosat satellite in the infrared range were used. The procedure based on the analysis of the time series of radiation temperature at fixed “points” on the earth has shown an advantage over the results obtained using the simple threshold filter. To assess the uniformity and quality of the data of the generated archive, annual and monthly mean cloud amounts were compared with the satellite data archives of the ISCCP and ground-based observations of SYNOP. For a qualitative comparison of all the archives, annual mean values of the cloud amount, standard deviations and correlation coefficients between them were calculated. Correlation coefficients for monthly mean cloud numbers between the archives were 0.71–0.95 at the level of 99% confidence, and standard deviations in winter over the sea were 0.7 and over the land up to 0.9. The cloud amount error was determined by the gradations (score) and was performed for two satellite archives and the archive of ground-based observations. The errors that arise when assessing the quality of the cloud data from satellite and ground-based observations depend on the features of their detection, observation, and the processing techniques. The obtained data array showed that the zone of the maximum cloud amount over the Black Sea in winter indicates the active stage of the Black Sea depression, with its typical stationary atmospheric front. The results of the Fourier analysis showed the presence of a quasi-two-year harmonic, with the largest amplitude of oscillations occurring in winter and spring, and the smallest in summer