Revisiting an Emergent Constraint on Climate Sensitivity from Global Temperature Variability using Paleoclimate Data
Abstract
Prior studies have suggested a variety of possible constraints on equilibrium climate sensitivity (ECS). One of the more notable examples relies on a measure of historical global temperature variability (Ψ), as defined in Cox et al., 2018 (doi:10.1038/nature25450), to constrain ECS. In response to this work, it has been suggested that the observational temperature record is too short to robustly measure Ψ and that the strong warming trend in recent decades influences the results. Here, we use reconstructions of global temperature over the past 1000 years along with simulations from the Paleoclimate Modeling Intercomparison Project Phases 3 and 4 (PMIP3/4) and the Community Earth System Model 1 Last Millennium Ensemble to assess if longer time series of global temperature can be exploited to constrain projections of future warming. The models under consideration are characterized by ECS values that span a wide range, from 2.1 to 4.7 K. We find that while these models exhibit a strong relationship between Ψ and ECS when Ψ is calculated over the 20th century (r=0.82), the relationship weakens substantially when Ψ is calculated over the past millennium (r=0.46). Moreover, most of the simulated estimates of Ψ over the past millennium exceed estimates of Ψ derived from temperature reconstructions. This suggests that models with lower ECS values tend to exhibit better agreement with paleoclimate reconstructions of global temperature variability. Given that Ψ over the past millennium is dominated by aerosol forcing from large volcanic eruptions while Ψ over the past century is dominated by anthropogenic warming, it is rather surprising that these two quantities are well correlated across 12 PMIP models (r=0.71). We offer possible reasons for this similarity.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFM.A45L2010C