Modeling cloud phase in the Arctic and globally

Doktorsavhandling, 2026

The phase of cloud hydrometeors critically modulates the impact of clouds on the global energy balance. In the Arctic, this phase partitioning influences the timing of sea ice melt and freeze. Mixed-phase clouds (MPCs) consist of both liquid and ice phases, requiring models to accurately simulate this apportionment to represent the climate system. Yet, MPCs remain difficult to simulate. This thesis advances the modeling of microphysical processes that govern phase evolution in MPCs through a multi-scale approach using large-eddy simulations (LES) and general circulation models (GCMs).
The research introduces a stochastic ice nucleation parameterization to address the limitations of existing schemes. Applying this method to LES of an Arctic MPC produced ice mass magnitudes consistent with observations. However, the scheme was sensitive to model resolution; resolving this dependency is a prerequisite for its application to improve ice representation across both LES and GCM scales. Extending the investigation of ice nucleation, a sensitivity analysis across three GCMs revealed diverging relative importance of four microphysical processes, including ice nucleation. Even a unified secondary ice production parameterization caused varying model responses, ranging from negligible changes to substantial global impacts. The lack of consensus further illustrates the challenges of modeling cloud microphysics and questions the traditional representation of microphysics as a chain of individual processes. A sensitivity analysis of a separate Arctic MPC showed that the concentrations of droplet-forming aerosol particles and ice crystals dominated the liquid and ice mass, respectively. Ice crystal shape proved less influential for the absolute magnitudes of these phases, yet it dictated whether the cloud remained mixed-phase, glaciated, or evolved into a purely liquid cloud. These findings emphasize that models must explicitly account for ice crystal shape and that observational campaigns should prioritize measuring both concentrations and shape at cloud level rather than the surface.
Overall, this thesis advances the representation and understanding of microphysics by providing a new ice nucleation parameterization, challenging the current global modeling approach of sequencing microphysical processes, and highlighting the critical role of ice crystal shape for the cloud phase.