Sensitivity of ICON ensemble simulations of tropical cyclone Shaheen to sea surface temperature and terrain-height changes

Controlled sensitivity experiments provide a direct way to examine how regional numerical weather prediction models respond to changes in environmental and surface boundary conditions. For tropical cyclones affecting the Arabian Sea and Oman Sea, such experiments are particularly important because the simulated storm can be strongly influenced by lower-boundary forcing and coastal terrain. Here, Tropical Cyclone Shaheen (2021), the most recent strong cyclone to directly affect northern Oman, is used as a case study. Using a regional ensemble of ICON at 6.5 km resolution, we evaluate controlled sensitivity experiments in which sea surface temperature (SST), terrain height, and sea-surface roughness length are modified. The experiments are designed to assess how these changes modify the simulated storm intensity, track, rainfall distribution, dynamical structure, and the relative contributions of different precipitation processes. Among the tested changes, SST+2 K produces the strongest model response, with enhanced intensification and increased coastal rainfall. Topography reduction mainly changes the storm track and weakens terrain-enhanced rainfall, while physically motivated changes in sea-surface roughness length have a smaller influence on storm intensity and precipitation structure. Pressure-level diagnostics show that the terrain response is associated with a vertical reorganisation of the cyclone circulation, including changes in low-level flow, moisture-flux convergence, mid-level ascent, and upper-level divergence. Across all sensitivity tests, when light precipitation is excluded by focusing on rainfall above 5 mm in a 6-h window, the 6-h bias exceeds the bias computed from the full-period accumulation. Within this rainfall category, the parametrized convective component remains small across the experiments, indicating that the dominant precipitation sensitivity arises from grid-scale processes. These results show that SST and terrain height produce the clearest ICON response during the Shaheen case, but through different pathways: SST mainly affects storm intensity and coastal rainfall magnitude, whereas terrain changes mainly affect track displacement, vertical circulation structure, and orographic rainfall distribution. The findings provide a case-based assessment of ICON sensitivity to selected lower-boundary and terrain controls and support future evaluation of grey-zone cyclone simulations.