Investigation of the Cloud Phase Distribution and Related Parameters with Datasets from a Passive Satellite Sensor

In numerical weather prediction and climate models a particular challenge is represented by the simulation of mixed-phase clouds in the temperature range between 0 and -38°C, in which hydrometeors can either consist of supercooled water or of ice. The partitioning into liquid and ice in this regime depends on many factors (e.g. cloud dynamics, aerosols as ice nuclei, ice multiplication processes...) and varies therefore depending on different cloud types, regions and seasons. We use satellite data observations to analyse the cloud phase distribution under different conditions. To get a better understanding of cloud systems, different datasets are used to investigate the liquid and ice cloud distribution. The considered datasets are CLARA-A2 and Cloud_cci. They consist of data records derived from the AVHRR passive sensor, carried by the polar orbiting meteorological satellite NOAA-19. The cloud classification scheme by Pavolonis et al., 2005 is used for the retrieval of the cloud top phase distribution and related parameters. Special attention is paid to the geographical distribution of different cloud regimes, obtained using the k-means clustering method. The regimes taken into account are ISCCP-like regimes and glaciation regimes. For the first ones, bidimensional histograms are built using the cloud optical thickness and the cloud top pressure; for the glaciation regimes, bidimensional histograms are built using the mean liquid effective radius and the ice fraction as variables. We will present analyses on the occurrence of these regimes and their connections.