The reversibility of local and regional temperature extremes in CDRMIP - Preprint

[Geoengineering News]

https://essopenarchive.org/users/890609/articles/1369111-the-reversibility-of-local-and-regional-temperature-extremes-in-cdrmip

Authors: Spencer Clark, Andrew David King, Josephine R. Brown, Liam J Cassidy, Eduardo Alastrué de Asenjo, Tilo Ziehn

12 December 2025

Abstract

Currently implemented greenhouse gas emission reduction policies are projected to result in warming of the climate system beyond the 1.5°C Paris Agreement limit. Consequently, achieving the Agreement’s objectives may only be possible following a period of ‘overshoot’, where warming temporarily exceeds 1.5°C, before later declining and stabilizing below this limit through net-negative CO2 emissions. While previous studies have illustrated that global-average warming is reversible under net-negative CO2 emissions, it remains unclear whether other human-induced climate changes will exhibit similar reversibility. In this work, we assess the reversibility and hysteresis behavior of local and regional temperature extreme frequencies in eight Earth System Models that have completed the Carbon Dioxide Removal Model Intercomparison Project Tier 1 experiment. For equivalent global warming levels reached through periods of positive and negative emissions, respectively, we observe a high degree of hysteresis and short-term irreversibility in extreme temperature frequency across many land and ocean regions, although patterns vary substantially across models. To address inter-model variation, a storyline approach is adopted to present changes in temperature extreme frequency conditioned on plausible evolutions of large-scale Earth system processes, taking the strength of the Atlantic Meridional Overturning Circulation (AMOC) as an illustrative example. We demonstrate a close linkage between AMOC evolution and local temperature extreme changes, particularly in the Northern Hemisphere, and illustrate how this accounts for much of the observed inter-model variation. This work highlights the usefulness of adopting storyline approaches for presenting regional climate information under overshoot, given limited model participation and model uncertainty.

Source: ESS Open Archive