Exploring the timing and impacts of the 8.2 ka event in Europe: a comparison between proxy data and model outputs

The 8.2 ka event is one of the most prominent climatic events through the Holocene. It is recognised as a short-lived (~150 years), cold event caused by an outburst of freshwater to the North Atlantic. The event has been recorded in the North Atlantic region and in Europe and even across more far-field parts of the Northern Hemisphere. Even though this is a well-studied event, the influence on Europe is still debated due to the inconsistency in timing and variability of reported impacts.

In this study we aim to improve our understanding of the regional impacts in Europe caused by the 8.2 ka event based on a comparison study between speleothem proxy data and climate model simulations. Speleothems can yield high-resolution, site-specific information about variations during short-lived climate anomalies, which can be well constrained in time with precise and accurate U-Th dating.

We have compiled all existing speleothem records covering the 8.2 ka event from the SISAL database^*1 or through communication with the authors. For our data-model comparison, the datasets were selected based on three criteria: chronology, resolution and interpretation. The included datasets cover at least the period between 7.9-8.5 ka BP, and the temporal resolution is sufficient to identify and detect the 8.2 ka event. The interpretation of the various proxies has been clearly indicated by the authors, such as reflecting temperature or precipitation variation or representing a change of moisture source. Additionally, the datasets were standardized by recalculating the U-Th ages and remodeling the age-depth relationships to ensure consistency. Finally, the paleoclimate impacts, derived from proxy data, were compared with several published climate-model simulations.

^*1Atsawawaranunt, Kamolphat, Harrison, Sandy and Comas-Bru, Laia (2019): SISAL (Speleothem Isotopes Synthesis and AnaLysis Working Group) database version 1b. University of Reading. Dataset. http://dx.doi.org/10.17864/1947.189