Can interannual to decadal variability help increase the accuracy of climate sensitivity estimates?

Climate sensitivity refers to the amount of global surface warming that will occur in response to a doubling of atmospheric CO₂ concentrations when compared to pre-industrial levels. Understanding climate sensitivity and reducing uncertainty in the estimation of climate sensitivity are therefore critical to reducing spread in projected climate change under given scenarios. The aim of this study is to estimate real-world Equilibrium Climate Sensitivity (ECS) by exploiting relationships found between observable parameters and the magnitude of climate change. We develop an emergent constraint based on surface temperature variability, which we test using preindustrial control and historical simulations from CMIP5 and CMIP6 models. We estimate the relationship between model-to-model differences (M2MDs) in ECS and M2MDs in global, tropical and tropical Pacific temperature variability, using the various measures of variability on interannual through to multidecadal timescales. We find higher correlations between MDMDs in ECS and M2MDs in the standard deviation of temperature variability in the tropics, which peaks at the decadal timescale, with larger spread in CMIP6 models. These results are then optimally combined to constrain observed temperature decadal variability and provide a distribution of real-world ECS.