Past rapid warmings as a constraint on greenhouse-gas climate feedbacks

There are large uncertainties in the estimation of greenhouse-gas climate feedback. Recent observations do not provide strong constraints because they are short and complicated by human interventions, while model-based estimates differ considerably. Rapid climate changes during the last glacial peri...

Full description

Bibliographic Details
Published in:Communications Earth & Environment
Main Authors: Liu, Mengmeng, Prentice, Iain Colin, Menviel, Laurie, Harrison, Sandy P.
Format: Article in Journal/Newspaper
Language:English
Published: 2022
Subjects:
Online Access:https://researchers.mq.edu.au/en/publications/2191fa31-6d72-4e0a-8713-c5de581ce38f
https://doi.org/10.1038/s43247-022-00536-0
https://research-management.mq.edu.au/ws/files/233440775/230697190.pdf
http://www.scopus.com/inward/record.url?scp=85137237570&partnerID=8YFLogxK
Description
Summary:There are large uncertainties in the estimation of greenhouse-gas climate feedback. Recent observations do not provide strong constraints because they are short and complicated by human interventions, while model-based estimates differ considerably. Rapid climate changes during the last glacial period (Dansgaard-Oeschger events), observed near-globally, were comparable in both rate and magnitude to current and projected 21st century climate warming and therefore provide a relevant constraint on feedback strength. Here we use these events to quantify the centennial-scale feedback strength of CO 2 , CH 4 and N 2 O by relating global mean temperature changes, simulated by an appropriately forced low-resolution climate model, to the radiative forcing of these greenhouse gases derived from their concentration changes in ice-core records. We derive feedback estimates (expressed as dimensionless gain) of 0.14 ± 0.04 for CO 2, 0.10 ± 0.02 for CH 4 , and 0.09 ± 0.03 for N 2 O. This indicates that much lower or higher estimates of gains, particularly some previously published values for CO 2 , are unrealistic.