The Response of Tropical Land–Ocean Precipitation Partitioning to SST and CO2 Increase in Global Storm-Resolving Simulations

Abstract The tropical land–ocean precipitation partitioning is skewed toward the land. This study assesses how carbon dioxide (CO 2 ) increase and uniform sea surface temperature increase affect this partitioning. We analyze 14-yr global simulations conducted with the icosahedral nonhydrostatic (ICON) model at 10-km horizontal grid spacing, at which convection is simulated explicitly. ICON’s precipitation partitioning shows better agreement with observations than the AMIP6 ensemble. Under 4×CO 2 , precipitation partitioning increases, favoring land precipitation, whereas it decreases upon +4-K sea surface temperatures (SSTs). We develop a diagnostic framework based on the land’s atmospheric energy and moisture budgets to decompose the response of tropical precipitation partitioning into contributions from land atmospheric heating, land circulation efficiency, land moisture cycling, and tropical radiative cooling. In ICON and the AMIP6 ensemble, the land atmospheric heating is identified as the primary controlling factor for changes in precipitation partitioning. Changes in atmospheric heating drive circulation adjustments that modulate land precipitation through changes in land moisture convergence. The response of the controlling factors is similar in ICON and in the AMIP6 ensemble, apart from two qualitative differences. First, the controlling factors are generally more stable toward the imposed forcings in ICON compared to the AMIP6 ensemble. Second, the opposing responses in precipitation partitioning upon CO 2 forcing and equivalent uniform SST increase are of virtually equal magnitude in ICON, whereas in AMIP6, precipitation partitioning responds more strongly to the uniform SST increase. These findings suggest that coarse-resolution models may underestimate the land hydrological sensitivity relative to the total tropical hydrological sensitivity. Significance Statement While there are energetic constraints for the change of precipitation with greenhouse-gas-induced warming, land and ocean precipitation changes remain poorly understood and are often treated separately. In this study, we investigate how the tropical land–sea precipitation partitioning responds to idealized forcing. We separate the effects of increased uniform sea surface temperature and carbon dioxide (CO 2 ) concentration on precipitation partitioning in atmospheric-only simulations. Using an energetic–hydrologic framework, we compare the icosahedral nonhydrostatic (ICON) model, a global storm-resolving model that explicitly simulates convection with coarser models that parameterize convection. Our work suggests that traditional climate models using convective parameterizations might underestimate the landward shift of precipitation in response to CO 2 forcing and thus overestimate the oceanward precipitation shift in response to combined CO 2 forcing and sea surface temperature (SST) increase.