Periodic Reporting for period 4 - C2Phase (Closure of the Cloud Phase) Reporting period: 2021-10-01 to 2023-09-30 Summary of the context and overall objectives of the project Clouds strongly modulate the Earth’s radiative budget, are key elements in the hydrological cycle, and are major drivers of the uncertainty of the prediction of future climate. Clouds can consist of liquid water droplets, ice particles, or both at the same time. At temperatures below 0°C, liquid droplets remain in a metastable thermodynamic state until freezing is initiated heterogeneously, either by an ice particle with which it comes into contact or by an ice nucleating aerosol particle, or homogeneously at about -37°C. Whether and how many ice particles are present in an otherwise liquid cloud (i.e. the phase distribution) influences the dynamical development through latent heat release, the interaction with radiation, and the formation of precipitation. Despite its importance, the cloud phase distribution is poorly represented in many weather and climate models. In the ERC Starting Grant C2Phase (“Closure of the Cloud Phase”), we have set out to improve the representation of the phase distribution in models by making use of recent progress in understanding of microphysical ice formation processes and in observational capabilities with space-based passive sensors. More specifically, our objective was to test the following hypothesis: The relevant primary and secondary ice formation processes can be included in state-of-the art cloud models such that these agree in a statistical sense with space-based observations of the cloud thermodynamic phase for a wide range of conditions. The closure will be better (higher correlation between predicted and observed phase), the more physical details are included. Our research has shown that this is generally true, but that the representation of microphysical processes, in particular secondary ice formation, is still highly uncertain. Furthermore, uncertainties in the thermodynamic state of the atmosphere can introduce as large variations in the cloud phase distribution as microphysical drivers. Not all cloud-phase retrieval products for passive satellite sensors are suitable to identify fingerprints of primary and secondary ice formation. Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far We have demonstrated with idealized model studies that microphysical ice formation processes can exhibit specific fingerprints in the cloud-top phase distribution, which were expected to be observable from satellite. Furthermore, we have prepared our high resolution model for various setups which are suitable for the closure studies. Convective clouds, which transition the entire relevant temperature range, are particularly promising subjects. In a multiyear dataset obtained from the geostationary satellite instrument SEVIRI, several hundred convective clouds are tracked and their properties related to the cloud top phase transition from liquid to ice. The results pointed again at significant microphysical influence on the cloud top phase distribution, although variations of other factors complicate the analysis. However, in a model-to-satellite closure study for an observed case based on a day with a complex cloud scene (deep convective clouds over central Europe) using large-domain cloud resolving simulations, forward operators and the original retrieval schemes to compute artificial satellite products based on model output, these fingerprints were only conserved for rather strong microphysical perturbations and for a newly developed cloud phase retrieval scheme.The analysis of global datasets from polar orbiting satellites has revealed large biases, but hemispheric differences and trends with cloud optical depth (defining cloud types) are consistent independent of the retrieval scheme. A set of selected global climate and storm-resolving models are generally unable to reproduce the observed characteristic cloud phase distributions for different cloud types and regions, possibly because the investigated models do not include a representation of aerosol effects on cloud droplet and ice formation. Such effects were also found in detailed studies on observed cloud phase of low- and mid-level clouds over the Arctic and Southern Oceans. Low clouds were found to exhibit less ice when over sea then over open ocean, hinting at a marine source of ice nucleating particles. Meanwhile, mid-level clouds in the same regions were found to be impacted by long-range transported mineral dust. Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far) We have constructed a unique and large database of the cloud top properties of tracked convective clouds, which will be useful for the rest of the project and other studies beyond C2Phase. The ICON model has been updated with state-of-the-art descriptions of primary and secondary ice formation. We were able to provide guidance on an optimal parameterization of ice formation in detailed microphysics schemes, and also recommendations for the parameterization of cloud phase in global climate models. Cloud phase distribution in different sensitivity experiments in the real-case closure study.
