Constraining Modern Climate Sensitivity and Pattern Effects with New Paleoclimate and Historical Reconstructions

Determining the modern climate’s sensitivity to greenhouse-gas forcing has been a central challenge for over 40 years. To constrain the notoriously uncertain upper bound of climate sensitivity, we must look to the natural experiments in Earth’s past. Recent advances in climate reconstruction now provide new constraints on the spatial patterns of paleoclimate and historical temperature change. These temperature patterns play a leading role in climate sensitivity due to pattern effects.We first investigate the cold Last Glacial Maximum and the warm Pliocene. By combining recent reconstructions with atmospheric general circulation models, we show why cloud feedbacks strongly amplify temperature changes in past climates and how this finding helps constrain the upper bound of modern climate sensitivity.We then turn to the recent past (1850–2023) to examine the outstanding uncertainty in radiative feedbacks over the historical record. We introduce a new coupled reconstruction, which uses data assimilation to combine observational and dynamical constraints across the atmosphere and ocean. Using the reconstruction’s ensemble members in several atmospheric general circulation models, we quantify how uncertainty in SST, sea ice, and model physics leads to time-evolving uncertainty in feedbacks over the historical record. Finally, we combine results from the paleoclimate and historical records to show that accounting for pattern effects leads to stronger constraints on modern climate sensitivity and projections of 21st-century warming.