Global-coupled storm-resolving model: a tool to better understand tropical precipitation


In this study, we use the one-year simulation of the global-coupled ICON with the Sapphire (ICON-S) configuration in a mesh grid of 5 km to analyze the impact of resolving meso-beta scale processes on representing tropical precipitation. The zonal and meridional seasonal migration and the expansion/shrinking of the terrestrial rainbelt are represented in ICON-S similarly to observations. This result contrasts with the flawed diurnal precipitation cycle, with a more intense peak occurring two hours earlier than in observations. The rainbelt seasonality in the Eastern Pacific and the Atlantic is also well represented in ICON-S. This is not the case in the Indo-Pacific region, where ICON-S struggles to simulate the sea surface temperature pattern distorting the representation of the tropical rainbelt in this region.

With this analysis, we show that the improvements in the representation of precipitation shown on regional domains using storm-resolving models are, indeed, because of resolving meso-beta scale processes and not the influence of the boundary conditions. Resolving meso-beta scale processes also leads to a better representation of the rainbelt and the SST pattern in the E. Pacific and the Atlantic, but it is not enough in the Indo-Pacific region. The extraordinary performance of ICON-S over land in reproducing the seasonal precipitation cycle, even with ocean biases, suggests a mechanism not present in the ocean that makes precipitation over land so robust. The small interannual variability of the summer precipitation area over land in observations also points to the existence of this mechanism.