Large-eddy simulations of subtropical cloud-topped boundary layers under idealized climate change

Subtropical marine boundary layer (MBL) clouds strongly influence the Earth's radiation budget, and their response to climate change is one of the central uncertainties in climate projections. Large-eddy simulations (LES), which explicitly resolve the turbulence and convective processes responsible for cloud formation, have been used to study this response. However, the sea surface temperature (SST) or surface fluxes are usually prescribed explicitly in these simulations, and the surface energetic constraint may be violated.

Here we present a configuration of LES experiments that satisfies energetic and large-scale dynamical constraints. The SST is allowed to evolve interactively with radiative and heat fluxes at sea surface, thus the surface energy budget is closed in the statistically steady state. The free troposphere is relaxed towards a radiative-convective equilibrium profile representing the deep tropics, and the subsidence velocity is set to be consistent with the dynamical constraints of the tropical circulations. The LES is coupled with a simple two-stream radiation scheme that includes the longwave (LW) and shortwave (SW) effects of clouds and water vapor, and a simple cloud microphysics scheme. Representative cases of cumulus and stratocumulus under current climate conditions are tested and compared with the reanalysis data, which validate the capacity of this LES configuration to realistically represent the subtropical cloud-topped boundary layer regimes.

We have investigated the steady-state subtropical MBL cloud responses over a wide range of climates with this LES configuration. The simulated cumulus cloud fractions decrease as SST warms up with increased longwave opacities. This decrease is related to the reduced buoyancy production of turbulence in the subcloud layer, due to changes in the surface fluxes. We have also run control experiments to separate the cloud responses to various surface and free-tropospheric processes, and a simple bulk model is used to interpret these LES results. The results highlight the important interactions between turbulence and convection in controlling subtropical MBL cloud changes, as well as the necessity to validate and improve the corresponding parameterizations in climate models.