New study published with support from ObsSea4Clim!

The study aims to improve the representation of the Atlantic Water layer in the Arctic Ocean within #CMIP6 climate model simulations. This is crucial because Atlantic Water inflow significantly influences Arctic climate change, yet current models show large biases and uncertainty. By selecting model ensemble members that better match observed variability in the subpolar North Atlantic, this work seeks to constrain projections of Arctic Atlantification. This approach could enhance the reliability of future Arctic climate predictions.

Full article

Devilliers, M., Olsen, S. M., Langehaug, H. R., Guo, C., Mahmood, R., Tian, T., & Yang, S. (2025). Constraining CMIP6 simulations for Atlantic Water in the Arctic using an AMOC-SST index. Frontiers in Climate, 7, Article 1550772. https://doi.org/10.3389/fclim.2025.1550772

The accepted manuscript is available in Open Access on our Zenodo: https://zenodo.org/records/15322348

Key findings

● A subset of 19 out of 235 CMIP6 ensemble members was selected based on their correlation with an observed AMOC-SST index over 1960–1990.
● The selected members showed a systematically better representation of Atlantic Water core temperature (AWCT) and core depth (AWCD) in the Eurasian Arctic Ocean compared to the full ensemble.
● The selection led to reduced errors and improved correlations with reanalysis and observational datasets (ORAS5 and EN4) over 1980–2014.
● The selected members also maintained a stronger connection with observed subpolar gyre variability several decades after the selection period.
● These findings suggest that upstream Atlantic variability, particularly linked to the AMOC, plays a key role in shaping Arctic Atlantification.
● This ensemble screening method could help reduce uncertainty in future projections of Arctic Ocean changes.

Abstract

Atlantic Water plays a key role in future changes in the Arctic Ocean. It contributes to Atlantification by transporting salt and heat within the Arctic Ocean basins. Many studies also attribute the amplified warming of the Arctic Ocean to an increase in poleward ocean heat transport by warming currents or the increasing strength of ocean currents. Global models are needed to reliably predict consistent trends in heat transport, as large-scale processes are at play. However, these models are too coarse to resolve key ocean processes and to address the complex interplay between ocean dynamics and the bathymetry of the Arctic region. Here, we propose to construct a sub-ensemble of simulations based on 235 historical simulations from 12 CMIP6 models that best represent the downstream drivers of Atlantic warming. We select the model ensemble members showing the closest agreement with observed surface temperature variability over 1960–1990 in the subpolar gyre (SPG). More specifically, we use a recent index that links surface temperature in the SPG to the Atlantic Meridional Overturning Circulation (AMOC): the AMOC-SST index. The subsampled ensemble shows a better correlation with the observed AMOC-SST index over the last 35 years of the historical period (1980–2014). It also displays a reduced error and better correlation for the Atlantic Water core temperature and depth in the Eurasian Arctic Ocean when compared to reanalysis and observations. Overall, the AMOC-SST index-based selection leads to a systematic improvement in the representation of the Atlantic Water layer in the Eurasian Arctic region, suggesting a clear connection between the Arctic Ocean and surface temperature in the subpolar region, and by extension, possibly the AMOC.