Why does the climate response to greenhouse warming and greenhouse cooling differ?

While the long-standing quest to constrain equilibrium climate sensitivity has resulted in intense scrutiny of the processes controlling idealized greenhouse warming, the processes controlling idealized greenhouse cooling have received less attention. Here, differences in the climate response to increased and decreased carbon dioxide concentrations are assessed in state-of-the-art fully coupled climate model experiments. One hundred and fifty years after an imposed instantaneous forcing change, surface global warming from a carbon dioxide doubling (abrupt-2xCO2, 2.43 K) is larger than the surface global cooling from a carbon dioxide halving (abrupt-0p5xCO2, 1.97 K). Both forcing and feedback differences explain these climate response differences. Multiple approaches show the radiative forcing for a carbon dioxide doubling is ~10% larger than for a carbon dioxide halving. In addition, radiative feedbacks are less negative in the doubling experiments than in the halving experiments. Specifically, less negative tropical shortwave cloud feedbacks and more positive subtropical cloud feedbacks lead to more greenhouse 2xCO2 warming than 0.5xCO2 greenhouse cooling. Motivated to directly isolate the influence of cloud feedbacks on these experiments, additional abrupt-2xCO2 and abrupt-0p5xCO2 experiments with disabled cloud-climate feedbacks were run. Comparison of these “cloud-locked” simulations with the original “cloud active” simulations shows cloud feedbacks help explain the nonlinear global surface temperature response to greenhouse warming and greenhouse cooling. Overall, these results demonstrate that both radiative forcing and radiative feedbacks are needed to explain differences in the surface climate response to increased and decreased carbon dioxide concentrations.