Linking hectometer simulations and remote sensing observations to derive statistical cloud properties

Cloud (process) parameterizations need to describe statistical properties and distributions which are unresolved by the applied resolution. Ideally, these parameterizations are representative for different regions and synoptic conditions. In this study, hectometer simulations with the ICON model and a two-moment microphysics scheme are performed in a semi-operational setup at two very different locations: Jülich (Germany, since fall 2021) and Ny-Ålesund (Svalbard, since fall 2020). The setup contains a 100 km domain with 600 m resolution and a daily start at midnight. As forcing data the operational global weather forecasts from the Deutscher Wetterdienst (german weather service) are used. These longterm simulation database is challenged and benchmarked by dedicated remote sensing observations of clouds. To enable the comparison between observations and simulations, retrieval methods are applied to the observational data as well as an instrument simulator (PAMTRA) on the model output.

We will show results from the comparison of this growing longterm dataset with a special focus on the representation of cloud properties, the capturing of the general situation, as well as changes in resolved vs unresolved scales per region and season.