The impact of the mesh size and microphysics scheme on the representation of mid-level clouds in the ICON model in hilly and complex terrain

Abstract. The rise in computational power in recent years has enabled research and national weather services to conduct high-resolution simulations down to the kilometric (Δx=𝒪(1 km)) and even to hectometric (Δx=𝒪(100 m)) scales for both weather and climate applications. We investigate with the state-of-the-art numerical weather prediction model ICON how mid-level clouds are represented on a mesh size of 1 km and 65 m, respectively, and for two bulk microphysics schemes, namely one-moment and two-moment cloud microphysics. For this analysis, we leverage the abundant observational data from two independent field campaigns in Switzerland (CLOUDLAB; hilly terrain) and Austria (CROSSINN; complex terrain). With four case studies, we show that while the temperature fields around the campaign sites are well represented in both mesh sizes, the 65 m resolution simulates a more realistic vertical velocity structure that is beneficial for cloud formation. Therefore, the largest differences for the representation of clouds lie in the two mesh sizes. The 1 km simulation in hilly terrain does not capture the observed clouds in both cloud microphysics schemes. Here, the higher resolution of the vertical velocities in the 65 m proves to be crucial for representing the investigated cloud types, and the two-moment microphysics scheme in general performs better with respect to the cloud characteristics because it considers variations in cloud droplet and ice crystal number concentrations. In complex terrain, the differences between the mesh sizes and the cloud microphysics schemes are surprisingly small, but the 65 m simulations with two-moment cloud microphysics show the most realistic cloud representation.