Contributions of Climate Feedbacks to Changes in Atmospheric Circulation

The radiative response to CO₂ forcing and climate feedbacks is uncertain in current models, but it is not known how this uncertainty affects projections of circulation change. In this modelling study, the impact of climate feedbacks on meridional shifts of atmospheric circulation under CO₂ forcing is quantified by decomposing the circulation response into individual contributions of radiative forcing and feedbacks by surface albedo, clouds, and water vapor. On average, clouds dominate the overall poleward jet shift; while CO₂ forcing also weakly contributes to a poleward shift, the jet response to water vapor and surface albedo feedbacks ranges from neutral to an equatorward shift. The jet response can be interpreted through changes in baroclinicity: while clouds strongly enhance midlatitude baroclinicity, favoring a strengthened, poleward-shifted jet, water vapor and albedo feedbacks act to weaken low-level baroclinicity. However, large model uncertainty in radiative feedbacks causes a large spread in the baroclinicity response to CO₂ forcing. Across the CMIP5 models, this spread is dominated by differences in shortwave feedbacks by clouds and albedo. To quantify the dynamical impact of uncertainties in shortwave feedbacks, "ghost forcing" experiments are performed with prescribed shortwave heating anomalies associated with cloud and albedo feedbacks from a set of CMIP5 models. Jet shifts correlate strongly with the meridional gradient of the anomalous shortwave heating and the associated baroclinicity response. Differences in shortwave feedbacks explain about 50% of the inter-model spread in CMIP5 jet shifts for our set of models, demonstrating the importance of constraining radiative feedbacks for accurate projections of circulation changes.