Observational constraint on the climate sensitivity to CO2 concentrations derived from the 1971-2017 global energy budget

The estimate of the historical effective climate sensitivity (histeffCS) is revisited with updated historical observations of the global energy budget in order to derive an observational constraint on the effective sensitivity of climate to CO2 (CO2effCS). A regression method based on observations of the energy budget over 1971-2017 is used to estimate the histeffCS. Then, climate model simulations are used to evaluate the distance between the histeffCS and the CO2effCS. The observational estimate of the histeffCS and the distance between the histeffCS and the CO2effCS are combined to derive an observational constraint on CO2effCS consistent with previous studies which use the same uncertainty in aerosol forcing (4.09 [1.81;26.09] K, median and 5-95% range). The main sources of uncertainty in the CO2effCS estimate comes from the uncertainty in aerosol forcing and in the top of the atmosphere energy imbalance. The second source arises from the pattern effect correction estimated from climate models. Because climate models potentially underestimate systematically the pattern effect, we develop an observation-only approach based on a refined energy budget to account explicitly for the pattern effect. The observational approach confirms the likely underestimate of the pattern effect in climate model simulations leading to an underestimated distribution of CO2effCS. This important result suggests that observations of the global energy budget since 1971 are poorly consistent with climate sensitivity to CO2 below 2.5 K. It also suggests that climate sensitivities to CO2 above 4.5 K are likely, consistent with many CMIP6 climate model estimates.