Eleanor Frajka-Williams

The Atlantic Meridional Overturning Circulation (AMOC) is projected to weaken under anthropogenic warming, yet climate models show substantial uncertainty in the North Atlantic sea surface temperature response. Using 22 CMIP6 models, we investigate how the historical thermohaline structure of the western subpolar North Atlantic influences transient AMOC and climate evolution under SSP5-8.5 forcing. Models are classified into thermally-dominated, salinity-dominated, and balanced-compensation regimes based on subsurface temperature-salinity mean-state structure. Although all groups exhibit substantial AMOC weakening, only salinity-dominated models develop a pronounced subpolar cooling response. In contrast, thermally-dominated and balanced models exhibit a more moderate relative cooling. The cause of the large cooling in the salinity-dominated models is related to a transition from a weakly stratified, ventilated state toward a stratified regime, resulting in strong AMOC-subpolar gyre coupling. These results demonstrate that historical stratification state and overturning-gyre coupling are key to explain future North Atlantic climate change rather than AMOC alone.

In early June, 2026, a workshop was convened to address the topic of climatic changes in the Atlantic Ocean circulation – specifically, a weakening or collapse of the large-scale circulation in response to anthropogenic global warming. The gathering brought together 61 experts from across the fields of physical oceanography, climate dynamics, paleoceanography, and dynamical systems, and involved participation from 11 countries. This document reports on the workshop’s aims, structure, and themes, and provides a summary of key discussions. The information comes from notes and recordings made during the workshop itself, as well as a summary survey conducted immediately following the event. The above listed authors constituted the workshop’s organising committee. All participants were invited to contribute to and comment on this report. It should thus be considered a shared effort from that wider group. A full list of workshop participants can be found in the acknowledgements.

Ocean variability spans a wide range of scales, from centimeters to thousands of kilometers and from seconds to millennia. Turbulence and mixing occupy the smallest scales, with the Atlantic Meridional Overturning Circulation (AMOC) at the largest scales. However, small-scale processes can also affect the large-scale circulation. Past reviews have considered how mixing sets or affects the mean and structure of the AMOC. This review considers how mixing may affect the time variability of the AMOC. It first defines the AMOC, then discusses drivers of variability and how they are manifested in the observed overturning at the RAPID 26°N and Overturning in the Subpolar North Atlantic Program (OSNAP) arrays. It then considers how diabatic processes can change the large-scale overturning circulation.

The Labrador Sea is a vital component of global ocean circulation, hosting vigorous deep convection that can mix to deeper than 2,000 m. Features that drive convection are widely studied, but a lack of high‐resolution, in situ observations hinder our understanding of how finescale dynamics influence the convective process and trigger restratification in spring, particularly at the submesoscales (10 km). This study assesses contributions of temperature and salinity to both vertical and horizontal stratifications during early winter, convection, and spring restratification, using data collected by five underwater gliders deployed in the winters of 2019/20 and 2021/22. Using the Turner angle, we show that vertical stratification in the Labrador Sea shifts from salinity‐stratified in early winter to temperature‐unstable during convection, and back to salinity‐stratified in the early restratification period. Horizontally, both warm and fresh intrusions drive lateral density anomalies during restratification, with the haline influence dominating. Wavelet analysis of these anomalies illustrates increased submesoscale activity during restratification, and comparisons of thermohaline contributions show the predominance of salinity‐driven submesoscale fronts. These fronts have the potential to rapidly stratify the water column, and evidence the importance of freshwater intrusions at submesoscales in halting convection.

Ocean bottom pressure recordings are a key observation for both ocean circulation and seafloor geodesy. New self‐calibrating instruments may solve a long‐time issue of instrument drift, allowing new high precision observations. However, instruments on the seafloor may settle over days to months, potentially contaminating results. Here we present evidence for likely settling of two self‐calibrating instruments deployed along the RAPID mooring array, on the order of 8–14 cm. Settling is expected, and we model the settling based on Terzaghi's consolidation theory, which is well known in civil engineering applications. After the settling correction the difference between the two pressure records has a standard deviation of ∼1 cm water height equivalent. Settling is likely an important factor that should be considered in oceanographic and geodetic experimental design. We recommend longer term deployments of 5–10 years to allow for settling and/or selection of sites with thinner sandy sediments to minimize settling.

[1]¿p#1 newcommands The Labrador Sea is a vital component of global ocean circulation, hosting vigorous deep convection that can mix to deeper than 2000 m. Features that drive convection are widely studied, but a lack of high-resolution, in-situ observations hinder our understanding of how small scale features influence the convective process and trigger restratification in spring, particularly at the submesoscales (< 10 km). This study assesses contributions of temperature and salinity to both vertical and horizontal stratification during early winter, convection, and spring restratification, using data collected by five underwater gliders deployed during winter 2019/20 and 2021/22. Using the Turner angle, we show that vertical stratification in the Labrador Sea shifts from salinity-stratified in early winter to temperature-unstable during convection, and back to salinity-stratified in the early restratification period. Horizontally, both warm and fresh intrusions drive lateral density anomalies during restratification, with the haline influence dominating. Wavelet analysis of these anomalies illustrates increased submesoscale activity during restratification, and comparisons of thermohaline contributions show the predominance of salinity-driven submesoscale fronts. These fronts have the potential to rapidly stratify the water column, and illustrate the importance of freshwater intrusions at submesoscales in halting convection.

\textit{In situ} observations of transbasin deep ocean transports at $26^\circ$N show variability on monthly to decadal timescales (2004--2015). Satellite-based estimates of ocean bottom pressure from the Gravity Recovery and Climate Experiment (GRACE) satellites were previously used to estimate interannual variability of deep ocean transports at $26^\circ$N. Here, we use GRACE ocean bottom pressure, reanalysis winds and \textit{in situ} transport estimates at $26^\circ$N to diagnose the large-scale response of the deep ocean circulation to wind-forcing. We find that deep ocean transports -- including those associated with a reversal of the Atlantic meridional overturning circulation in 2009/10 and 2010/11 -- are part of a large-scale response to wind stress curl over the intergyre-gyre region. Wind-forcing dominates deep ocean circulation variability on monthly timescales, but interannual fluctuations in the residual \textit{in situ} transports (after removing the wind-effect) are also captured by GRACE bottom pressure measurements. On decadal timescales, uncertainty in regional trends in GRACE ocean bottom pressure preclude investigation of decadal-timescale transport trends.

The Arctic Ocean is undergoing rapid changes due to global warming and the reduction in sea ice, affecting its carbon cycle. Sparse observations of surface partial pressure of CO ₂ (pCO ₂ ) limit our understanding of these changes. Here, we use continuous fields of surface pCO ₂ derived from a gap-filling technique to assess regional and temporal variations in pCO ₂ at high latitudes. We find an increase in the annual mean pCO ₂ throughout most of the Arctic. However, the rate of change since 2005 varies strongly between regions. By performing a spatio-temporal decomposition via empirical orthogonal function (EOF) analysis, we suggest two primary drivers of seasonal variability, namely solar irradiance correlating negatively with pCO ₂ in the sub-Arctic and seasonal variations in the sea-ice cover correlating with the pCO ₂ in the high Arctic. Unraveling these dynamics is a crucial step toward better understanding the processes governing the Arctic’s carbon cycle in the future.

Turbulent vertical velocity measurements are scarce in regions prone to convection such as the Labrador Sea, which hinders our understanding of deep convection dynamics. Vertical velocity, , is retrieved from wintertime glider deployments in the convective region. From , downward convective plumes of dense waters are identified. These plumes only cover a small fraction of the convective area. Throughout the convective area, the standard deviation of agrees with scaling relations for the atmospheric surface and boundary layers. It initially depends on surface buoyancy loss in winter, and later, on wind stress after mid‐March. Both periods are characterized by positive turbulent vertical buoyancy flux. During convective periods in winter, the positive buoyancy flux is mostly forced by surface heat loss. After mid‐March, when buoyancy loss to the atmosphere is reduced, the positive buoyancy flux results from a restratifying upward freshwater flux, potentially of lateral origins and without much atmospheric influence.

The Antarctic Bottom Water (AABW) exported from the Weddell Sea has experienced warming and contraction in the past 30 yrs. Superposed on this decadal trend is substantial annual and interannual variability in the volume and properties of Weddell‐sourced AABW. Several mechanisms have been suggested to explain these variations, many of which highlight a role of wind stress, but the comparative importance and possible simultaneity of the different mechanisms remains unclear. Using data from two mooring sites within the Weddell Sea, we find a rapid intensification of the abyssal boundary current carrying AABW through Orkney Passage (OP), the most direct export pathway of AABW from the Weddell Sea, in response to periods of strong zonal wind stress and anomalous wind stress curl along the South Scotia Ridge upstream of OP. This acceleration is concomitant with a 40% reduction in northward AABW transport in late 2015. The changes in transport follow anomalous wind forcing by approximately 3 months, with the short timescale indicative of a barotropic response in the flow through OP. The bottom boundary layer over the OP's sloping topography is found to have a key role in regulating export on monthly to interannual timescales. Increased boundary current velocity leading up to the passage forms a thickened bottom boundary layer, resulting in reduced AABW thickness and density, and thus restricting northward transport of AABW through the passage. Whilst other processes are likely to dominate on longer (decadal) periods, the dynamics identified here can explain significant variability on timescales up to interannual.