Calendar

Name of speaker: Kira Rehfeld from the University of Tübingen

Abstract:

Stabilizing Earth’s surface temperature below 2°C above preindustrial (PI) levels reduces future climate risks. Climate models are used to explain past climate changes, to explore future climate impacts, and to investigate mitigation approaches such as carbon dioxide removal and geoengineering to address the slow pace of societal decarbonization. 

We established strong evidence that climate models simulate correct global climate variability, but underestimate regional variability on decadal to centennial timescales. This may mean that models equilibrate regional differences too quickly under stabilizing feedbacks and that regional assessments of technological climate mitigation may be biased.

A key question is which climate feedbacks could enhance decadal-to-centennial regional climate variability, while maintaining global climate stability in the model world. The instrumental era provides is too short to constrain feedback balances. I will discuss two approaches we have been exploring in order to find answers: Spectral analysis of palaeoclimate records, explicitly considering the uncertainties of natural archives, and perturbed physics ensembles for climate models.

Combining previously disparate lines of investigation provides new insights into climate feedbacks, sensitivity and variability that allow to better assess potential paths towards stabilizing Earth’s global mean temperature.