Hadley Cell Extent Linked to Biases in Cloud-Circulation Coupling, with Implications for Climate Sensitivity

Understanding mid-latitude clouds and their impact on Earth’s energy balance is critical to reducing the uncertainty in climate sensitivity. We analyze model output from the CMIP5 archive and from a global climate model in idealized aquaplanet setup to examine the covariability of interannual Southern Hemisphere Hadley cell (HC) edge latitude shifts and shortwave cloud radiative effect (SWCRE). With poleward HC edge expansion, models substantially reduce the shortwave radiation reflected by clouds over the lower mid-latitude (LML) region, although no such reduction is seen in observations. These biases in HC-SWCRE covariability are linked to biases in the climatological HC edge latitude. Models with a more equatorward climatological HC edge latitude also exhibit larger increases in LML SW warming with poleward HC edge expansion. In the aquaplanet model, the cloud-radiative changes result predominantly from a subsidence-induced warming of the planetary boundary layer, which decreases low-level cloud fraction and is stronger for more equatorward HC edges. A comparison of the aquaplanet results with the CMIP5 results suggests that about half of the model uncertainty in the mid-latitude cloud-circulation coupling stems from biases in the position of the climatological circulation, and thus could be removed by improving the representation of the climatological circulation in models. This model behavior, based on interannual variability, has important implications for the CMIP5 model response to CO2-forcing. In 4xCO2-forced runs, models with excessively equatorward climatological HC edge latitudes produce stronger LML SW warming and tend to have larger climate sensitivity values than models with more realistic climatological HC extents. Thus, an improved representation of the large-scale circulation may also lead to progress in constraining estimates of the SW cloud radiative response to increasing CO2 and of climate sensitivity.