Atmospheric cloud-radiative heating in CMIP6 and observations and its response to surface warming

Cloud–radiation interactions are key to Earth's climate and its susceptibility to change. While their impact on Earth's energy budget has been studied in great detail, their effect on atmospheric temperatures has received little attention, despite its importance for the planetary circulation of the atmosphere and hence for regional climate and weather. Here, we present the first systematic assessment of vertically resolved cloud-radiative heating within the atmosphere in 20 CMIP6 (phase 6 of the Climate Model Intercomparison Project) models, including a comparison to satellite-based estimates. Our analysis highlights model differences in cloud-radiative heating in both the lower and upper troposphere as well as uncertainties related to cloud ice processes. It also illustrates limitations of our ability to observe cloud-radiative heating. Not surprisingly, the response of cloud-radiative heating to surface warming is also uncertain across models. Yet, in the upper troposphere, the response is very well predicted by an upward shift of the present-day heating, which we show results from the fact that cloud-radiative heating in the upper troposphere is a function of air temperature and thus decoupled from surface temperature. Our results have three important implications for upper-tropospheric cloud-radiative heating: they establish a new null hypothesis for its response to warming, offer a physics-based prediction of its response to warming based on present-day observations, and emphasize the need for improving its representation in simulations of the present-day climate, possibly by combining the benefits of upcoming km-scale models and satellite observations.