Cloud ice particle nucleation and atmospheric ice supersaturation in numerical weather prediction models
b'Cirrus cloud genesis is a multiscale problem. This makes the parameterization in numerical weather prediction models a challenging task. In order to improve the prediction of cirrus clouds and ice supersaturation formation in the German Weather Service (DWD) model chain, the controlling physical processes are investigated and parameterised in a new cloud ice microphysics scheme. Scale dependencies of the ice microphysical scheme were assessed by conducting simulations with an idealised and realistic regional Consortium for Small-Scale Modeling (COSMO) model setup and a global model (GME). The developed two-moment two-mode cloud ice scheme includes state-of-the-art parameterisations for the two main ice creating processes, homogeneous and heterogeneous nucleation. Homogeneous freezing of supercooled liquid aerosols is triggered in regions with high atmospheric ice supersaturations (145-160 %) and high cooling rates. Heterogeneous nucleation depends mostly on the existence of sufficient ice nuclei in the atmosphere and occurs at lower ice supersaturations. The larger heterogeneously nucleated ice crystals can deplete ice supersaturation and inhibit subsequent homogenenous freezing. In order to avoid an overestimation of heterogeneous nucleation, cloud ice sedimentation and a prognostic budget variable for activated ice nuclei are introduced. A consistent treatment of the depositional growth of the two ice particle modes and the larger snowflakes using a relaxation timescale method was applied which ensures a physical representation for depleting ice supersaturation. \\n\\nComparisons between the operational and the new cloud ice microphysics scheme in the GME revealed that the location of cirrus clouds is dominated by the model dynamics whereas the cirrus cloud structures strongly differed for the different schemes. Especially a reduction in the ice water content between 9 and 11 km was observed when using the new cloud ice scheme. This change is an improvement as demonstrated by a comparison with the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) ice water content product. Further comparisons of the GME with the Integrated Forecast System (IFS) model of the European Centre for Medium-Range Weather Forecasts (ECMWF) show a clear improvement of the ice supersaturation distribution with the new two-moment cloud ice scheme. In-cloud ice supersaturation is correctly captured, which is compliant with in-situ measurements. This is a more physical description then in the IFS model, where in-cloud ice saturation is assumed.'