A new droplet parameterization for ICON-AES physics

Running climate simulations with ICON (Sapphire configuration) using the 1-moment microphysical scheme so far relied on very simplistic assumptions about cloud droplet number concentrations (CDNC) profiles, which do not evolve in time and space. This has direct implications for cloud radiative effects and precipitation rates. Recent developments in km-scale modeling within the WarmWorld project, however, have introduced important advances in this area.

In this presentation, I introduce a new droplet parameterization developed for ICON-AES physics. It makes use of cloud condensation nuclei (CCN) derived from Copernicus Atmosphere Monitoring Service reanalysis (CAMSRA; Block et al., ESSD, 2024). This approach provides observationally constrained input for computing CDNC fields without the need to couple ICON to a full aerosol model.

This new module is currently implemented for the 1-moment scheme, in which cloud water mass is predicted while CDNC are prescribed as time-dependent boundary conditions. CDNC are computed using a diagnostic, fitted droplet parameterization that depends on CCN and vertical velocity, adapted from Kuba and Fujiyoshi (ACP, 2006). This scheme is therefore computationally efficient and well suited for non-hydrostatic models. To fully exploit this parameterization, a CCN-supersaturation spectrum is constructed using an adaptation of Twomey’s power law when reading CCN of reduced complexity into ICON. This ensures computational efficiency and helps to correct biases recently identified in CCN evaluations.

I will discuss the scientific features of this scheme, its computational feasability, and present preliminary results.