MPI-DS Seminar: From particles to planet: how individual droplets, ice, and turbulence shape clouds and climate

Clouds emerge from interactions across scales, from individual droplets and ice crystals to planetary radiation and precipitation budgets. Yet particle-scale processes and their in-teractions with turbulence occur far below the scales resolved by weather and climate models. Here, we ask how much these small-scale details matter for clouds and climate, using particle-based Lagrangian cloud models coupled with large-eddy simulations, to-gether with long-term observations and stochastic modeling. Arctic winter observations show two persistent surface radiative states, clear and cloudy, with distinct consequenc-es for the surface energy budget and sea-ice growth. Using a Langevin stochastic differ-ential equation framework applied to 27 years of high-frequency observations from Utqiaġvik, Alaska, we show that this bimodality is noise-induced. The mean dynamics (drift) support only one preferred state; the two states arise because cloud-driven variabil-ity (diffusion) peaks in the transition range between them. ERA5 captures the drift but un-derestimates this variability, consistent with its unimodal distribution. We then examine which microphysical processes organize state persistence and transitions. A large en-semble of LCM simulations evaluated against MOSAiC radar observations shows that the two states coexist only within a bounded thermal regime controlled by mixed-phase cloud glaciation, which provides a pathway from cloudy to the clear state transition. As the Arc-tic warms, this pathway becomes increasingly inefficient, shifting winter conditions to-ward a predominantly cloudy regime. Observations indicate this shift is already underway, and climate projections show it continuing, reducing opportunities for strong surface cool-ing and sea-ice growth. These results show how particle-scale microphysics can con-strain future Arctic winter climate states.