North Atlantic Ocean circulation and deep water formation under warmer climate conditions in EC-Earth3-HR

North Atlantic Ocean circulation and deep water formation under warmer climate conditions in EC-Earth3-HR | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article North Atlantic Ocean circulation and deep water formation under warmer climate conditions in EC-Earth3-HR Rene Gabriel Navarro-Labastida, Mehdi Pasha Karami, Torben Koenigk, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7542176/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract In this study, we analyze the impact of increased atmospheric CO₂ concentrations on the Atlantic Meridional Overturning Circulation (AMOC) and its dependence on North Atlantic Deep Water Formation (DWF), using the high-resolution version of the global coupled model EC-Earth3. The analyzed experiments include a pre-industrial control and two fixed CO₂ concentration scenarios, representative of stable warmer climate conditions and designed to investigate the long-term adjustments of the Earth system. Within this framework, we assess DWF using a novel method based on horizontal volume convergence in the main convective regions of the North Atlantic: the Labrador, Greenland-Iceland-Norway, and Irminger Seas. Our results suggest that under warmer climate conditions surface warming and freshening in the North Atlantic lead to disrupt deep convection, reduce DWF, and thereby weaken the AMOC. The reduction of DWF in the Labrador Sea emerges as the primary driver of AMOC weakening. In contrast, the Irminger Sea plays a central role in sustaining AMOC. The strengthened export of deep water from the Arctic Ocean also provides a stabilizing influence, though its effect is secondary to the sustained DWF in the Irminger Sea. AMOC weakening deep mixed volume deep water export Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 1 Introduction 1.1 Overview The Atlantic Meridional Overturning Circulation (AMOC) is a crucial component of the Earth's climate system, responsible for the large-scale redistribution of heat and anthropogenic carbon (Buckley and Marshall 2015; Jackson et al. 2022 ). Trans-basin observational results and projections from the latest phase of collaboration under the Coupled Model Intercomparison Project (CMIP6) both suggest a likely AMOC weakening driven by rising CO2 levels (McCarthy and Caesar 2023 ; Baker et al. 2023 , 2025 ). However, the transient response of the AMOC to climate forcing varies significantly across models, making it challenging to predict the precise future behavior of the circulation. Although most CMIP6 models assess a full AMOC collapse as unlikely within the 21st century, a recent analysis of the extended simulations of CMIP6 models reveals AMOC shutdown under the high-emission scenario. This suggests a significantly higher risk than previously assumed (Dijkhout et al. 2025). Furthermore, recent estimates using fingerprint attribution suggest that the AMOC could collapse around mid-century under the current emissions scenario (Ditlevsen and Ditlevsen 2023). Nonetheless, this approach may overlook complex interactions between the AMOC and global warming, partly because it assumes the AMOC directly drives long-term subpolar sea surface temperature changes. Despite the limitations, this has increased interest in monitoring early warning signals to anticipate the potential collapse of AMOC (Ditlevsen and Ditlevsen 2023; Rahmstorf 2023; van Westen et al. 2024 ). The potential for a nearly total AMOC collapse has been previously assessed under both hypothetical scenarios where large freshwater fluxes are introduced to the North Atlantic Ocean (Jackson et al. 2015 ; Bellomo et al. 2023 ; van Westen et al. 2024 ), and in the context of millennial-scale climate changes, using models configured for glacial conditions and long-term simulations designed to test the stability of the AMOC under past climate scenarios (Weijer et al. 2019 ). Based on such studies, it is believed that a large freshwater anomaly may reduce the Northern Atlantic ocean's salinity, halting deep-water formation in the North Atlantic and rapidly weakening the AMOC, leading to severe cooling over the European Arctic, where temperatures could drop by 10°C compared to current levels (Jackson et al. 2015 ; van Westen et al. 2024 ). Conversely, under future emission scenarios, strengthened freshwater fluxes and subsurface warming are projected to diminish deep convective activity in the North Atlantic, particularly in the Labrador Sea (Brodeau and Koenigk 2016 ; Levang and Schmitt 2020). Nevertheless, the establishment of a direct link between short-timescale convection in the Labrador Sea and deep water production and the overturning circulation remains an open challenge (Lozier et al. 2019 ; Lozier et al. 2021). This suggests that regions within the North Atlantic with weaker stratification, such as the Irminger Sea and the Greenland-Iceland-Norway Sea, may sustain deepwater production, even in the event of a total shutdown of convection in the Labrador Sea (Jackson et al. 2022 ; Dimma et al. 2022; Bretones et al. 2022; Chafik et al. 2022 ; Årthun et al. 2023, 2025 ). Therefore, comprehensive understanding of the convection process in the region is required, as it is hypothesized that these processes play a pivotal role in maintaining the AMOC's overall strength (Koenigk et al. 2021 ; Roberts et al. 2020). 1.2 Model resolution Previous multi-model analyses suggest that high-resolution models tend to simulate stronger AMOC weakening compared to lower-resolution Earth System Models (Bellomo et al. 2021; Roberts et al. 2020; Jackson et al. 2020 ). This difference is partly due to the tendency of lower-resolution models to produce an overly stable AMOC (e.g., Liu et al. 2014), whereas high-resolution models appear to capture AMOC variability more realistically (Koenigk et al. 2021 ; Jackson et al. 2020 ; Hirschi et al. 2020 ; Shan et al. 2024 ; Karami et al. 2025 ). In addition, the rate of decline still varies with model resolution due to model-dependent behavior, adding a further complication (Roberts et al. 2020; Jackson et al. 2020 ; Hirschi et al. 2020 ). Nevertheless, high-resolution coupled models offer more accurate representations of Arctic sea ice, including sea ice area, sea ice volume, the position of the ice edge (Docquier et al. 2019 ), in addition to improved representation of the poleward heat transport within the climate system (Grist et al. 2018 ; Docquier et al. 2019 ). Overall, high-resolution CMIP models enhance the representation of large-scale circulation patterns and small-scale processes, including climate extremes (Haarsma et al. 2016 ) while improving the simulation of sea surface temperature, velocity, and the depiction of boundary currents across all ocean basins (Gutjahr et al. 2019 ; Docquier et al. 2019 ). 1.3 Motivation While the aforementioned studies are valuable for understanding the transient climate response under increasing greenhouse gas concentrations, they primarily address short-term dynamics rather than the long-term adjustments of the Earth system. By isolating the ultimate impacts of long-term high CO₂ levels, we may assess more accurately the consequences of climate change under various warming scenarios, regardless of the pathways through which these levels are reached. To address this, we are analyzing fixed CO 2 concentration experiments that represent quasi-equilibrium climate states under warmer conditions. By doing this we further reduce uncertainties in future projections involving increased atmospheric CO₂ concentrations, particularly concerning North Atlantic oceanic processes such as the AMOC. The warmer climate conditions represent global temperature anomalies of 1°C and 2°C compared to the pre-industrial state. At a global temperature anomaly of 1°C, the level of Arctic warming (~ 3.7°C; Fig. S1 a in supplementary material) is comparable to conditions during the Last Interglacial period (LIG), which is often regarded as a potential analog for future high-latitude climate scenarios due to similar temperature increases (Sicard et al. 2023 ). Proxy-reconstructions and climate modeling studies suggest that during LIG, the AMOC may have been weaker than in pre-industrial times when freshwater forcing is included (e.g., Galaasen et al. 2014 ; Govin et al. 2012 ; Guarino et al. 2023 ). However, there is still some debate in this matter, as recent multimodel simulations representing different interglacial conditions show minimal changes in AMOC strength during the mid-Holocene (Jiang et al. 2023 ), whereas earlier studies suggest that a weakened AMOC likely contributed to significant large-scale climate disruptions in Europe, including shifts in temperature and precipitation patterns, as well as potentially extreme weather events (Govin et al. 2012 ). A global temperature anomaly of 2°C represents a closed analog to the global mean temperature during LIG, the climate that can be expected if the Paris Agreement is successfully implemented (Fisher et al. 2018; Rohling et al. 2019 ), and much more higher level of warming in the Arctic (~ 6.7°C; Fig. S1 a). 2 Materials and methods 2.1 Model description and experimental design In this study, we use the high-resolution version of the global coupled climate model EC-Earth3 (hereafter referred to as EC-Earth3-HR), which was used to contribute to CMIP6 in its standard-resolution (Döscher et al. 2022). Our configuration includes the IFS atmosphere model (cycle 36r4), with the HTESSEL land surface module, and the NEMO ocean model (version 3.6), coupled with the LIM3 sea ice model. The atmosphere was configured with a T511 spectral resolution (~ 40 km), while the ocean model used the ORCA025 configuration, corresponding to approximately 0.25 degrees. In the vertical domain EC-Earth3-HR employs 91 levels in the atmosphere and 75 layers in the ocean. The model underwent a tuning process, multi-centennial spin-up, followed by a 100-year spin-up with the final set of parameters. After the spin-up process, we started a 350-year pre-industrial (pi-control) control run, and a 1% per year increasing CO₂ experiment (1pctCO₂) (details in Karami et al. 2025 ). From the 1pctCO₂ experiment, two further experiments with fixed CO₂ concentrations were started from the points in time when global temperature anomalies reached around 1°C and 2°C compared to the EC-Earth3-HR pre-industrial (Fig. S1 b). In terms of the CO 2 concentrations in the 1pctCO 2 these levels of warming represent 400.9 ppm and 551.5 ppm and the years 1885 and 1917, respectively. After fixing atmospheric CO₂ concentrations, the model underwent a stabilization phase lasting approximately 33 and 40 years, before reaching quasi-equilibrium state at around 1.3°C and 2.7°C above the EC-Earth3-HR pre-industrial levels, respectively. Subsequently, both simulations were extended for an additional 100 years. Both fixed CO₂ experiments showed a rapid reduction of summer sea ice, with sea ice disappearing by the end of the simulation under the 551.5 ppm CO₂ concentration (Fig. S1 c). In this study, we focus on analyzing the two fixed CO₂ concentration experiments. Changes are assessed using the PI-control simulation as the reference state. All calculations are based on the final 100 years of each simulation. From here on, we refer to the fixed CO₂ experiment at 400.9 ppm as CO₂400 (moderate concentration) and the one at 551.5 ppm as CO₂550 (high concentration). 2.2 Deep Mixed Volume (DMV) In our analysis we use the Deep Mixed Volume (DMV) index to quantify deep convection (Brodeau and Koenigk 2016 ). The computation involves integrating the volume of mixed water masses below a critical depth, using only winter data (March). This calculation is performed for both the control and CO₂ experiments. The DMV is computed for the Labrador (LAB), Irminger (IRM), and Greenland-Iceland-Norway (GIN) regions. The calculation involves selecting relatively large areas to cover all the convective spots within each region (see Fig. 1 ). Only grid points that meet the depth criteria are considered. This refers to a critical depth of 1000 meters for LAB and IRM, and 700 meters in GIN. The selection of depth criteria is mainly based on topographical and dynamical constraints outlined in previous analyses (Brodeau and Koenigk 2016 ). Additionally, we calculate the frequency of convection events in line with DMV computation. The calculation is based on the event frequency of the winter mixed layer reaching the above-mentioned critical depths (frequency in percentage defined as the number of years with convection). This metric provides insights into the frequency and persistence of convective areas, as well as information on their spatial distribution. 2.3 Deep Water Formation (DWF) We apply a novel methodology to assess Deep Water Formation (DWF) based on calculating the horizontal volume convergence in the North Atlantic (Karami et al. 2025 ). The DWF index is computed using the same spatial division as in the DMV calculation (region boundaries in Fig. 1 ). The vertical column is divided into an upper and a lower layer in each region, divided at the same depth as in DMV computation to maintain consistency in our analysis. Following this approach, we calculate the horizontal mass convergence as the net inflow and outflow of volume across sections within each layer. Positive convergence in the upper layer represents water sinking toward the lower layer (below the critical depth), which corresponds to deep water formation. In contrast, negative convergence in the lower layer indicates the export of the downwelled upper water. The difference between upper and lower layer convergence is attributed to the net surface freshwater fluxes in the upper layer and the use of annual means in the computation of the metric. However, our results show that this contribution is small (~ 0.1 Sv) and shows minor changes between experiments. 2.4 Meridional transports at 26°N and 45°N The net meridional flow at 26°N has been vertically divided into upper and lower layers at a depth of 1000 meters. The characterization of the flow at 26°N is meant to zonally cross the Subtropical Gyre system and match our estimation of the AMOC index, defined as the time series of annual-mean maximum volume transport at this latitude (Fig. 1 ). The section is further divided horizontally into a western and eastern subsections. The boundary between subsections (80°W) is designed to avoid regions of water mass recirculation (McPhaden and Zhang 2004; Zhang and McPhaden 2006 ). The net meridional poleward transport in the upper layer is assumed to represent the flow out of the gyre toward the subpolar region, while the net equatorward transport in the lower layer is assumed to represent the return of this water as deep water. We do not explicitly isolate the wind-driven circulation contribution within the gyre; thus, transports in the upper layer include both the Ekman and the geostrophic gyre circulation. However, all Ekman transport is assumed to subduct within the subtropical gyre. The Overturning in the Subpolar North Atlantic Programme (OSNAP) is often used as an analog for AMOC circulation across the subpolar gyre (approximately at 50°N) (Fu et al. 2023; Lozier et al. 2019 ). However, by doing so, the role of the Irminger-Iceland and Labrador Seas contributions is ignored (Petit et al. 2025 ). In that regard, meridional transport at 45°N was chosen to characterize the export of deep water from the North Atlantic convective regions. At this location, the meridional flow has been vertically and horizontally divided, with horizontal outcrop at 40°W to pair the flow across the southern limits of the LAB and IRM regions (Fig. 1 ). 3 Results 3.1 North Atlantic surface waters Figure 2 shows the wintertime climatologies of sea surface salinity (SSS), sea surface temperature (SST), and potential sea surface density (SSD) within the main convective regions in the North Atlantic. In the control simulation (Fig. 2 a, d, g), warmer and saltier waters from the subtropics spread widely all over the IRM region. Following the counterclockwise circulation pattern of the subpolar gyre, this water continues spreading westward toward LAB and northward into GIN via the Iceland–Norway section. Surface density in IRM is relatively low compared to LAB and GIN (Fig. 2 g) due to the warmer waters in the IRM. In contrast, high SSD in LAB and GIN is driven by the strong surface cooling in the region, combined with the inflow of saltier waters from IRM. The relatively weak density along the eastern Greenland coastline in GIN is attributed to the influence of fresher waters spreading from the Arctic Ocean (Fig. 2 g). The spread of this water into the relatively salty GIN region promotes the formation of a poleward meridional density gradient, which is believed to be the main driver of the density-driven export of Atlantic waters toward the Arctic, primarily through the eastern boundary Nordic Seas (Årthun et al. 2023, 2025 ). Consequently, most of the dense surface water in the GIN is expected to exit the region and continue northward. In the moderate CO₂ concentration experiment (Fig. 2 b, e, h), simulated changes in SST, SSS, and SSD are explained in terms of local surface warming and freshwater flux changes, and partly by redistribution of these anomalies within the region. In detail, the general warming of the surface is due to a warmer atmosphere, except within the warming hole area, where a reduced overturning circulation leads to reduced ocean heat transport into this region (Keil et al. 2020 ). The widespread salinity reduction is likely explained in terms of the increased precipitation. The saltier and warmer waters dominate the GIN region, particularly between the Denmark Strait and the Barents Sea Opening is arguably explained in terms of the simulated sea ice retreat in the region (see blue line in Fig. 2 a, b). As a consequence, the horizontal density gradient seems to intensify, suggesting strengthened density-driven export of waters toward the Arctic. The strengthened Arctic and Atlantic waters spread along the eastern Greenland coastline are likely to contribute to further shaping the above-mentioned changes. The changes in the high CO₂ experiment (Fig. 2 c, f, i) are similar to those in the moderate concentration case but are only intensified. On average, all regions show intensification of net surface freshening and warming (Table 1 ). As in the moderate case, the strong haline change in GIN related to the sea ice edge retreat (Fig. 2 c) seems to further enhance the poleward meridional density gradient discussed above. Table 1 North Atlantic sea surface salinity, temperature, and density averaged over the entire regions of Labrador (LAB), Greenland-Iceland-Norway (GIN), and Irminger Sea (IRM). Deep Mixed Volume (DMV) is also shown. The results are for the pre-industrial control and the two fixed CO₂ experiments (CO₂400 and CO₂550). Only data from the last 100 years of simulation are used Exp. LAB GIN IRM Sea surface Salinity [psu] Control CO₂400 CO₂500 34.07 33.85 33.71 34.40 34.37 34.36 35.33 35.23 35.17 Sea surface Temperature [ o C] Control CO₂400 CO₂500 1.76 2.65 3.57 0.30 1.74 3.09 8.62 9.02 9.62 Sea surface Density [kg/m 3 ] Control CO₂400 CO₂500 27.06 26.80 26.61 27.43 27.30 27.18 27.27 27.14 26.99 DMV [x10 15 m 3 ] Control CO₂400 CO₂500 1.23 0.75 0.21 0.51 0.44 0.07 0.09 0.16 0.11 3.2 Changes in deep convection Figure 3 shows the frequency of convection events in the control simulation. Convection in LAB and GIN occurs more frequently and over larger areas compared to IRM, where it is relatively infrequent and confined to a smaller region. In the moderated CO₂ experiment (Fig. 3 b), the convection area in LAB has shrunk and shifted northward together with the northward-moving ice edge (Fig. 2 b). In IRM, both the convection area and frequency shows a moderate increase under this experiment. The intensified convection in the GIN Seas is associated with increased mixed layer depth values north of Svalbard, suggesting the emergence of new convective sites farther north. Deep convection in LAB and GIN is significantly reduced under high CO₂ concentration scenarios, with events becoming increasingly infrequent (Fig. 3 c). In IRM, convection under high CO₂ concentration shows similar values compared to those observed in the moderated case, with a higher occurrence of convection than in the control. In the control and moderated CO₂ experiments, IRM convection appears to be an extension of the LAB convection area, while under the high CO₂ experiment, it becomes a separate convection spot. In both CO₂ experiments, changes in convection area and frequency align with the weakened vertical profiles of potential density in each region (Fig. 4 ). This is especially pronounced within the upper 500 meters in LAB and GIN, suggesting enhanced vertical stratification in those regions. In contrast, vertical density changes in the IRM region are weaker, likely explaining the absence of significant changes in convection there. Consistent with the reduced densities in the LAB, GIN, and IRM seas, we find a decrease in DMV under both CO₂400 and CO₂550 (Table 1 and Fig. 4 ). Figure 5 , shows the time series for the DMV index in the LAB, GIN, and IRM regions. In the control simulation (Fig. 5 a, d, g), the strongest DMV values are observed in LAB, followed by GIN. All convective regions show strong interannual variability. In the moderate CO₂ concentration experiment (Fig. 5 b, e, h), LAB shows strong weakening. In GIN, there is also some reduction, but less pronounced compared to LAB. In the high CO₂ experiment (Fig. 5 c, f, i), DMV values are further weakened. While DMW in LAB stays still present, it has almost completely disappeared in GIN. The DMV in IRM does not show a major change between the CO₂400 and CO₂550 experiments which is consistent with the small size of the convection areas and the relatively low convection frequency in this region (Fig. 3 ). In the control simulation, DMV in the LAB tends to occur under colder and saltier surface conditions, whereas in the GIN and IRM seas it occurs under warmer and saltier conditions. (Table 2 ). In the IRM region, SST seems to be dominant. This is consistent with previous studies suggesting subpolar North Atlantic convection is influenced by large-scale climate patterns such as the Atlantic Multidecadal Oscillation (AMO) and the North Atlantic Oscillation (NAO) (Cheng and Zhang 2013; Sévellec et al. 2017 ; Dima et al. 2022 ; Lozier et al. 2010 ). This has been previously explored in terms of the interannual to multiannual variability under historical and preindustrial conditions (Petit et al. 2025 ), showing that positive NAO phases lead to intense heat loss over the western subpolar gyre. Conversely, negative NAO conditions can weaken the AMOC. More precisely, during negative NAO phases, reduced heat loss over the western subpolar gyre results in surface warming and decreased surface density, which weakens deep water formation in both the Irminger and Labrador Seas. This is later observed as a reduction in the southward outflow at 45°N. In GIN, DMV tends to occur under saltier surface conditions in both CO₂ scenarios (Table 2 ). However, SST changes arguably have a greater impact, given the larger change in SST compared to SSS. In the IRM, SST-driven DMV changes remain strong under moderate CO₂ forcing but weaken substantially under high CO₂, while the effect of SSS in DMV weakens in both CO₂ scenarios. In the GIN Seas, DMV is robustly determined by salinity changes, highlighting the dominant role of SSS in modulating DMV in this region (Table 2 ). This may suggest that as the climate warms, high SSS is no longer linked to warm SST, in contrast to the control conditions, where GIN convection occurs despite warm SST because the warm SST always occurs together with high salinity. Finally, although convective mixing affects the properties of the densest surface waters, it remains uncertain whether buoyancy-driven overturning depends strictly on localized convection (Marotzke 2000 ; Marotzke and Scott 1999). In addition, the widely held assumption that downwelling and convection occur in the same location may oversimplify the complex and often spatially decoupled processes that drive deep water formation. Table 2 Linear correlation (r) and confidence intervals at 95%. The results are for the pre-industrial control and the two fixed CO₂ experiments (CO₂400 and CO₂550). In the case of Sea Surface Temperature (SST), Sea Surface Salinity (SSS), and Deep Mixed Volume (DMV), only data from March is considered. Correlations using DMV and Deep Water Formation (DWF) were calculated within the Labrador (LAB), Greenland-Iceland-Norway (GIN), and Irminger (IRM) Seas. Correlations using the equatorward transport of deep water from the Arctic Ocean (Arc.trans) and the Atlantic Meridional Overturning Circulation (AMOC) index are shown as well. All values with no statistical significance are shown in red. Control CO₂400 CO₂550 r(SST LAB , DMV LAB ) -0.31 [-0.48, -0.12] -0.53 [-0.66, -0.38] -0.27 [-0.44, -0.07] r(SST IRM , DMV IRM ) -0.67 [-0.77, -0.55] -0.56 [-0.69, -0.41] -0.14 [-0.32, 0.06] r(SST GIN , DMV GIN ) 0.50 [0.33, 0.63] 0.04 [-0.16, 0.24] -0.14 [-0.33, 0.06] r(SSS LAB , DMV LAB ) 0.31 [0.12, 0.48] 0.26 [0.06, 0.43] 0.22 [0.02, 0.4] r(SSS IRM , DMV IRM ) -0.37 [-0.53, -0.19] -0.02 [-0.21, 0.18] 0.14 [-0.06, 0.33] r(SSS GIN , DMV GIN ) 0.64 [0.5, 0.74] 0.68 [0.56, 0.77] 0.45 [0.28, 0.6] r(Arc.trans., DWF GIN ) -0.89 [-0.93, -0.84] -0.81 [-0.87, -0.72] -0.81 [-0.87, -0.73] r(Arc.trans., AMOC) -0.01 [-0.21, 0.19] -0.14 [-0.06, 0.33] 0.33 [0.14, 0.5] 3.3 Changes in the AMOC In the control simulation, the annual mean AMOC stream function reveals a well-defined deep overturning cell reaching down to 2500 m depth (Fig. 6 a). The subduction of the subsurface branch occurs between 40°N and 60°N. A maximum circulation core of 20 Sv is observed between 30° and 45°N, roughly at 1000 m depth. An anticlockwise bottom overturning cell is also present, with stronger circulation southward at 30°N. Additionally, we have calculated the AMOC strength index (Fig. 7 ), which is defined as the time series of the annual mean maximum volume transport between 24.5°N and 27.5°N at depths of 800–1100 m. In both CO₂ concentration experiments, the AMOC weakens. The estimated weakening is 2.5 Sv (~ 13%) and 4.5 Sv (~ 22%) under the moderate and high CO₂ concentration experiments, respectively (Fig. 7 ). The increase in CO₂ concentrations causes significant changes in not only strength but also spatial distribution features, such as a shift of circulation cores toward lower latitudes, a more shallow upper cell, and a strong weakening of the bottom cell (Fig. 6 b,c). These seem to agree with the reduction of northward ocean heat transport to the subpolar North Atlantic (not shown), which leads to diminished surface cooling (Fig. 2 c, e, f) and weaker-shallower deep convection (Figs. 3 and 5 ). 3.4 Changes in deep water formation In the control simulation, the total DWF in the North Atlantic amounts to ~ 17 Sv from which the largest contribution is produced in the IRM region (~ 9 Sv), followed by LAB (~ 5 Sv) and GIN (~ 3 Sv) (Fig. 8 a, b, and Table 3 ). In the moderate CO₂ concentration experiment, total DWF decreases by 20%, with larger declines in LAB (30%) and GIN (35%), while IRM exhibits a relatively minor change (10%). Under high CO₂ concentrations, total DWF drops further by 40%, with sharp declines in LAB (70%) and GIN (60%), while IRM shows no additional change. Table 3 Volume transports (Sv ≡ 10 6 m 3 s -1 ) within Labrador (LAB), Greenland-Iceland-Norway (GIN), and Irminger (IRM) Seas (see Fig. 1 ), with positive values indicating inflows. Flows were calculated across given sections, which delimited each region (east, west, north, south). The results are for the pre-industrial control and the two fixed CO₂ experiments (CO₂400 and CO₂550). Results represent the mean values for the last 100 years of each simulation. The upper and lower flow contributions are shown. The Deep Water Formation index (DWF) is calculated as the horizontal mass convergence within each layer. Positive convergence in the upper layer represents water sinking toward the lower layer. Negative convergence in the lower layer is linked to deep water export outside the region. Total DWF is the sum of all the deep water volumes Control CO₂400 CO₂550 Upper Lower Upper Lower Upper Lower LAB East 7.57 -7.65 3.55 -6.66 1.51 -4.42 West 0.02 - 0.02 - 0.02 - North 2.27 0.07 2.22 0.05 1.92 0.01 South -4.78 2.36 -2.48 3.17 -2.12 2.95 DWF 5.08 -5.22 3.31 -3.44 1.33 -1.46 GIN East -1.87 - -2.75 - -3.05 - West - - - - - - North -0.82 1.70 -1.03 2.97 -0.90 3.57 South 5.75 -4.91 5.90 -5.18 5.31 -4.99 DWF 3.06 -3.21 2.12 -2.21 1.36 -1.42 IRM East 0.01 - 0.01 - 0.01 - West -7.57 7.65 -3.55 6.66 -1.51 4.42 North -5.75 4.91 -5.90 5.18 -5.31 4.99 South 21.81 -21.12 16.83 -19.29 14.24 -16.90 DWF 8.50 -8.56 7.39 -7.45 7.43 -7.49 Total DWF 16.64 -16.99 12.82 -13.10 10.12 -10.37 3.5 Deep water redistribution At 45°N, the North Atlantic Ocean exchanges upper subpolar waters with deep waters formed within the LAB, IRM, and GIN regions. The upper water (21.8 Sv) enters through the IRM Sea (Fig. 8 a; Table 4 ). About 40% of this volume is transformed into deep water within IRM, while the remainder is routed toward LAB (~ 35%) and GIN (~ 25%). In both regions, most of this water is further converted into deep water. The remaining portion is exported southward from LAB and northward from GIN into the Arctic Ocean and the Barents Sea. Similarly, in the lower layer deep water (21.1 Sv) exits the region through the IRM Sea (Fig. 8 b; Table 4 ), but in this case, water export reflects the combined DWF from all regions (90%), along with a contribution from the Arctic Ocean (10%). In both CO₂ experiments, DWF weakens across all regions because of the water redistribution and density changes (Fig. 4 ), with changes in upper and lower transports reflecting variations in strength rather than direction (Fig. 8 ). As a result, the overall circulation pattern remains consistent with the control simulation (arrows in Fig. 8 a, b). However, reduced DWF in LAB leads to decreased water export toward IRM. In contrast, GIN deep water export remains stable because increased deep water inflow from the Arctic compensates for reduced deep water formation in the warmer conditions (Fig. 8 b). 3.6 Meridional transports at 26°N and 45°N In the control simulation, the upper-layer western subsection at 26°N transports approximately 37 Sv poleward, with nearly half (~ 17 Sv) exported toward higher latitudes, while the remainder (~ 20 Sv) presumably recirculates within the subtropical gyre (see GC in Table 4 ). Similarly, the upper-layer eastern subsection at 45°N carries about 22 Sv poleward. Most of this volume is exported northward (~ 17 Sv), while the remainder (~ 5 Sv) presumably recirculates within the subpolar gyre. In the lower layer, the subtropical gyre circulation is 8 Sv at 26°N, and the subpolar gyre circulation is 5 Sv at 45°N (GC in Table 4 ). The stronger gyre circulation at 26°N likely reflects the section’s alignment with the core of the subtropical gyre, whereas the 45°N section lies farther south of the subpolar gyre center, accounting for the weaker flow. Under the CO₂ experiment, the simulated upper-ocean circulation changes (GC in Table 4 ) are weaker at 45°N compared to 26°N, which may be partially explained by the more favorable location of the latter. In contrast, the lower-layer circulation exhibits stronger changes at both locations (Table 4 ). However, these changes are not straightforward to link directly to the horizontal circulation of the subpolar or subtropical gyre, as the transport also includes meridional flow associated with the bottom overturning cell (Fig. 6 ). Despite the differences discussed above, the net volume transport in both sections remains comparable, suggesting a sustained latitudinal consistency of the flow with the AMOC (see net values in Table 4 ). Our simulated changes in the western subsection at 26°N align with this view, as they closely match the simulated changes in DWF. This is further backed up by CMIP6 projections under the SSP5-8.5 scenario, which suggest that the projected decline of the subtropical gyre’s western boundary southward transport is mainly driven by the thermohaline component rather than by wind-driven processes (Bryden et al. 2022; Tooth et al. 2024 ). In other words, changes in the lower branch are primarily buoyancy-driven and strongly influenced by remote surface buoyancy anomalies in the North Atlantic (Kostov et al. 2021; Asbjørnsen and Årthun 2023 ). Table 4 Meridional transports (Sv ≡ 10 6 m 3 s -1 ) in the North Atlantic at 26°N and 45°N, divided horizontally into west and east contributions, and vertically divided into upper and lower layers. Net transport is defined as the difference between transport contributions of the west and east, with net export carried within one of these subsections according to the circulation pattern. The returning flow or gyre circulation (GC) is also shown. At 26°N, we have chosen 80°W as the division between subsections. At 45°N, we have chosen 45°W, so subsections are equivalent to the flow crossing the southern limits of Labrador (LAB) and Irminger (IRM) regions in Fig. 1 Subtropical gyre (26 °N) Upper Lower West East (GC) Net West East (GC) Net Control 37.10 -20.24 16.86 -26.01 7.56 -18.45 CO₂400 33.99 -19.55 14.45 -22.56 6.47 -16.09 CO₂550 31.92 -19.32 12.60 -19.23 4.93 -14.29 Subpolar gyre (45 °N) Upper Lower West (GC) East Net West (GC) East Net Control -4.78 21.81 17.03 2.36 -21.12 -18.76 CO₂400 -2.48 16.83 14.35 3.17 -19.29 -16.13 CO₂550 -2.12 14.24 12.12 2.95 -16.90 -13.95 4 Discussion This study examines the relationship between North Atlantic deep convection, deep water formation, and the mean state of the AMOC under pre-industrial and elevated CO₂-fixed conditions. Using the DMV metric, we have quantified deep convection in the regions of LAB, GIN, and IRM. In the CO₂ experiments, deep convection across the North Atlantic shows a marked decline, consistent with the weakening of the AMOC. In that regard, to characterize deep water formation, we developed a method based on the assumption of effective subduction rates within the above-mentioned regions. Although we have not isolated the changes in DWF caused by the shoaling of overturning due to variations in convection and downwelling, our analysis shows no clear dependence on the critical depth selection. This suggests that changes in the depth structure of convergence are relatively negligible under warming conditions. Our method also enables the estimation of Arctic deep water export. Our results suggest that continued anthropogenic warming may disrupt deep convection, reduce DWF, and weaken the AMOC. In Fig. 9 , we show the time series of DWF and the AMOC index, with the sum of all DWF sources closely matching the AMOC volume in both control and CO₂ experiments. The total DWF decrease is primarily due to changes in the LAB and GIN Seas, while the IRM contribution remains strong (Figs. 8 and 9 ). This is in general agreement with previous observational data and reanalysis products that underscore the Irminger Sea's central role in regulating AMOC variability and its broader implications for climate dynamics in the North Atlantic (Petit et al. 2020 ; Chafik et al. 2022 ). The simulated strengthened Arctic deep water flow is likely driven by an increased surface density gradient in GIN, in response to the simulated sea ice retreat in the region (see blue line in Fig. 2 a, b and field in Fig. 2 h, i), which may also explain the enhanced upper-layer export from the GIN Seas toward the Arctic Ocean through the Fram Strait and Barents Sea Opening. This augmented export of dense surface waters toward the Arctic, in conjunction with the emergence of new convective sites further north as the climate warms (Bretones et al. 2022; Årthun et al. 2023, 2025 ), may be responsible for the simulated increase in returning deep water. Under both control and warming scenarios, Arctic deep water export increases while GIN-DWF decreases (see Fig. 8 b and Fig. 9 ), which is consistent with the SSP2-4.5 future scenario simulation in Karami et al. ( 2025 ). Furthermore, in agreement with previous studies (Heuzé, 2021 ; Bryden et al., 2022), our analysis reveals a statistically significant link between the increase in Arctic deep water export and the AMOC in the case of a high CO₂ concentration (Table 2 ). At this CO₂ concentration, the Arctic export doubles compared to the control simulation, while the contribution of the DWF of GIN and LAB substantially weakens. Nevertheless, deep water export from the Arctic Ocean is a secondary stabilizing factor for the future state of the AMOC rather than a dominant factor, as previously stated by ocean-only modeling studies (Sevellec et al., 2017; Heuzé, 2021 ; Dima et al., 2022 ). This suggests that the influence of Arctic outflow on the AMOC is probably indirect. The relative importance of IRM, as underscored by our analysis, stems from the ocean-atmosphere feedback encompassed by the higher-resolution global coupled model, which is clearly absent in an ocean-only model. Our method enables us to disentangle key processes by individually quantifying the most important components of the AMOC (Fig. 10 ). These components are often overlooked when relying solely on the traditional AMOC stream function representation. While the AMOC stream function is useful for representing meridional transport in both climate models and historical observations, it fails to identify the sources of deep water and the circulation processes due to its coarse representation of the meridional flow and the absence of a zonal component (Lozier et al. 2010 ). In contrast, our method explicitly includes the zonal component, setting a linkage between these processes on a long-term scale (Fig. 10 ). This enhanced view stresses how the dense water is redistributed within the North Atlantic convective regions, flows beneath the subpolar gyre, and eventually reaches the western boundary current in the subtropical gyre. Moreover, the relatively small differences between meridional poleward and equatorward transports further suggest limited compensating upwelling and interior mixing. The last reinforces the idea of strong meridional consistency in the lower branch of the AMOC on longer timescales (McCarthy and Caesar 2023 ; Frajka-Williams et al. 2023 ; Buckley and Marshall 2015). On climatic timescales, buoyancy forcing has been identified as the dominant driver of the AMOC, with a stronger effect than wind forcing (Petit et al., 2025 ; Yeager et al., 2021; Böning et al., 2023). However, recent observational estimates suggest that convection plays a limited role in driving interannual AMOC variability (Zou et al., 2020 ). Short-term fluctuations and the limited duration of observational records can obscure long-term variability, making sustained trends difficult to detect (McCarthy & Caesar, 2023 ; Frajka-Williams et al., 2023 ; Buckley & Marshall, 2015; Jackson et al., 2022 ; Lozier et al., 2010 ). Although deep water formation (DWF) is typically associated with regions of strong convection, it can also occur in areas where diffusive mixing within the water column is significant. Our method considers only mass convergence, thereby ostensibly neglecting the contribution of diffusive mixing. This may explain discrepancies observed between DWF and DMV estimates in the GIN Seas, where DMV indicates a shutdown of convection and a substantial reduction in deep water formation, while DWF does not indicate a complete shutdown (Figs. 8 b and 9 ). Moreover, interannual variability in DWF may be masked by the relatively large spatial domains used to compute the metric. The mismatch between DMV and DWF suggests that our MLD criteria may not be fully capturing the processes involved, implying that deep water sinking likely occurs via diffusive processes driven by changes in SST, SSS and SSD. Another aspect requiring further investigation is the impact of ice melt on local buoyancy and the AMOC. The weakening of deep convection and DWF in the GIN and LAB may be partially attributed to surface freshening caused by sea ice melt in these regions (Fig. 2 a–c). Therefore, future analyses should examine the dependence of these processes on water mass characteristics, explore shorter time scales, particularly across simulations with varying model resolutions. 5 Conclusions In this study, we have analyzed the impact of increased atmospheric CO₂ concentrations on the Atlantic Meridional Overturning Circulation (AMOC) and its dependence on North Atlantic Deep Water Formation (DWF) using the high-resolution version of the global coupled climate model EC-Earth3. Our high-resolution model provides enhancements in several key features of the North Atlantic, including more accurate simulations of regional circulation and improved vertical stratification, resulting from better representations of upper-ocean temperature and salinity (Karami et al. 2025 ). Analysed experiments include a pre-industrial control run and two fixed CO₂ concentration scenarios. Although the use of high-resolution limits our ability for running very long simulations, our CO2-level simulations are sufficiently long to reach a quasi-equilibrated state, and allow us to investigate long-term adjustments of the Earth systems, including changes in the DWF-AMOC relationship, to different CO2-levels. We have assessed DWF using a novel method based on horizontal volume convergence within North Atlantic regions of Labrador (LAB), Greenland-Iceland-Norway (GIN), and Irminger (IRM) Seas. The DWF weakens in both CO₂ experiments, with the LAB contribution showing the greatest reduction. In contrast, the IRM contribution remains strong, sustaining the AMOC. Similarly, the deep water export from the Arctic Ocean strengthened, which appears to act as an additional stabilizing factor for the future state of the AMOC. These changes correspond with variations in the meridional net volume transport carried at 26°N and at 45°N, suggesting strong latitudinal consistency of the meridional flow on the longer timescale. We show that the transport measured at 26 o N matches both the DWF estimations and the transport across section 45 o N. Despite this relationship not being fulfilled in terms of annual timescale variability, the match between these metrics suggests that observing DWF and export of deep water from IRM could suffice to monitor changes in AMOC strength, in both observational campaigns and under the framework of a large coupled model ensemble intercomparison. Declarations Funding This study was supported by the Horizon Europe project OptimESM “Optimal High Resolution Earth System Models for Exploring Future Climate Changes” under the European Union’s Horizon Europe research and innovation programme (grant agreement No 101081193), the FORMAS project FutureGS (Grant 2021-01374), the Swedish Research Council grant VR (2020-04791). The EC-Earth3 simulations and data handling were/was enabled by resources provided by the National Academic Infrastructure for Supercomputing in Sweden (NAISS), partially funded by the Swedish Research Council through grant agreement no. 2022-06725. Competing Interests The authors have no relevant financial or non-financial interests to disclose. Author Contributions All authors contributed to the study conception and design. Simulations were performed by Mehdi Pasha Karami and the analysis was performed by René Navarro-Labastida. The first draft of the manuscript was written by René Navarro-Labastida and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. Data availability The datasets generated and analysed during the current study are available under request. References Årthun M, Asbjørnsen H, Chafik L, Johnson H, Våge K (2013) Future strengthening of the Nordic Seas overturning circulation, Nat Commun 14, 2065. https://doi.org/10.1038/s41467-023-37846-6 Årthun M, Brakstad A, Dörr J, Johnson HL, Mans C, Semper S, Våge K (2025) Atlantification drives recent strengthening of the Arctic overturning circulation. Sci. Adv. 11. DOI:10.1126/sciadv.adu1794 Asbjørnsen H, and Årthun M (2023) Deconstructing future AMOC decline at 26.5°N, Geophysical Research Letters, 50, e2023GL103515. https://doi.org/10.1029/2023GL103515 Baker JA, Bell MJ, Jackson LC, Renshaw R, Vallis GK, Watson AJ, Wood, RA (2023) Overturning pathways control AMOC weakening in CMIP6 models, Geophysical Research Letters, 50, e2023GL103381. https://doi.org/10.1029/2023GL103381 Baker JA, Bell MJ, Jackson LC, Vallis GK, Watson AJ, Wood RA (2025) Continued Atlantic overturning circulation even under climate extremes. Nature 638, 987–994. https://doi.org/10.1038/s41586-024-08544-0 Bellomo K, Meccia VL, D’Agostino R, Fabiano F, Larson SM, vonHardenberg J, Corti S (2023) Impacts of a weakened AMOC on precipitation over the Euro-Atlantic region in the EC-Earth3 climate model, Clim Dyn 61, 3397–3416. https://doi.org/10.1007/s00382-023-06754-2 Bryden H, Beunk J, Drijfhout S, Hazeleger W, Mecking J (2024) Comparing observed and modelled components of the Atlantic Meridional Overturning Circulation at 26° N, Ocean Sci., 20, 589–599. https://doi.org/10.5194/os-20-589-2024 Brodeau L, Koenigk T (2016) Extinction of the northern oceanic deep convection in an ensemble of climate model simulations of the 20th and 21st centuries. Clim Dyn 46, 2863–2882. https://doi.org/10.1007/s00382-015-2736-5 Buckley MW, Marshall J (2016) Observations, inferences, and mechanisms of the Atlantic Meridional Overturning Circulation: A review, Rev. Geophys., 54. https://doi.org/10.1002/2015RG000493 Chafik L, Holliday NP, Bacon S, Rossby T (2022) Irminger Sea is the center of action for subpolar AMOC variability. Geophysical Research Letters, 49, e2022GL099133. https://doi.org/10.1029/2022GL099133 Cheng W, Chiang JCH, Zhang D (2013) Atlantic Meridional Overturning Circulation (AMOC) in CMIP5 Models: RCP and Historical Simulations. J. Climate, 26, 7187–7197. https://doi.org/10.1175/JCLI-D-12-00496.1 Dima M, Lohmann G, Ionita M, Knorr G, Scholz P (2022) AMOC modes linked with distinct North Atlantic deep water formation sites. Clim Dyn 59, 837–849. https://doi.org/10.1007/s00382-022-06156-w Docquier D, Grist JP, Roberts MJ, Roberts CD, Semmler T, Ponsoni L, Massonnet F, Sidorenko D, Sein DV, Iovino D, Bellucci A, Fichefet T (2019) Impact of model resolution on Arctic sea ice and North Atlantic Ocean heat transport. Clim Dyn 53, 4989–5017. https://doi.org/10.1007/s00382-019-04840-y Drijfhout S, Angevaare J, Mecking JV, van Westen RM, Rahmstorf S (2025) Shutdown of northern Atlantic overturning after 2100 following deep mixing collapse in CMIP6 projections. Environmental Research Letters. https://iopscience.iop.org/article/10.1088/1748-9326/adfa3b/meta Frajka-Williams E, Foukal N, Danabasoglu G (2023) Should AMOC observations continue: how and why?, Phil. Trans. R. Soc. A 381: 20220195. https://doi.org/10.1098/rsta.2022.0195 Fischer H, Meissner KJ, Mix AC, et al. (2018) Palaeoclimate constraints on the impact of 2 °C anthropogenic warming and beyond. Nature Geosci 11, 474–485. https://doi.org/10.1038/s41561-018-0146-0 Galaasen EV, Ninnemann US, Irvali N, Kleiven HF, Rosenthal Y, Kissel C, Hodell DA (2014) Rapid reductions in North Atlantic Deep Water during the peak of the last interglacial period, Science. DOI: 10.1126/science.1248667 Gérard J, Crucifix M (2024) Diagnosing the causes of AMOC slowdown in a coupled model: a cautionary tale, Earth Syst. Dynam., 15, 293–306. https://doi.org/10.5194/esd-15-293-2024 Govin A, Braconnot P, Capron E, Cortijo E, Duplessy JC, Jansen E, Labeyrie L, Landais A, Marti O, Michel E, Mosquet E, Risebrobakken B, Swingedouw D, Waelbroeck C (2012) Persistent influence of ice sheet melting on high northern latitude climate during the early Last Interglacial, Clim. Past, 8, 483–507. https://doi.org/10.5194/cp-8-483-2012 Grist JP, Josey SA, New AL, Roberts M, Koenigk T, Iovino D (2018) Increasing Atlantic Ocean heat transport in the latest generation coupled ocean-atmosphere models: The role of air-sea interaction. Journal of Geophysical Research: Oceans, 123, 8624–8637. https://doi.org/10.1029/2018JC014387 Guarino MV, Sime LC, Diamond R, Ridley J, Schroeder D (2023) The coupled system response to 250 years of freshwater forcing: Last Interglacial CMIP6–PMIP4 HadGEM3 simulations, Clim. Past, 19, 865–881. https://doi.org/10.5194/cp-19-865-2023 Gutjahr O, Putrasahan D, Lohmann K, Jungclaus JH, von Storch JS, Brüggemann N, Haak H, Stössel A (2019) Max Planck Institute Earth System Model (MPI-ESM1.2) for the High-Resolution Model Intercomparison Project (HighResMIP), Geosci. Model Dev., 12, 3241–3281. https://doi.org/10.5194/gmd-12-3241-2019 Haarsma RJ, Roberts MJ, Vidale PL, Senior CA, Bellucci A, Bao Q, Chang P, Corti S, Fučkar NS, Guemas V, von Hardenberg J, Hazeleger W, Kodama C, Koenigk T, Leung LR, Lu J, Luo JJ, Mao J, Mizielinski MS, Mizuta R, Nobre P, Satoh M, Scoccimarro E, Semmler T, Small J, von Storch JS (2016) High Resolution Model Intercomparison Project (HighResMIP v1.0) for CMIP6, Geosci. Model Dev., 9, 4185–4208. https://doi.org/10.5194/gmd-9-4185-2016 Hersbach H, Bell B, Berrisford P, Biavati G, Horányi A, Muñoz Sabater J, Nicolas J, Peubey C, Radu R, Rozum I, Schepers D, Simmons A, Soci C, Dee D, Thépaut JN (2023) ERA5 hourly data on single levels from 1940 to present. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). DOI: 10.24381/cds.adbb2d47 Heuzé C (2021) Antarctic Bottom Water and North Atlantic Deep Water in CMIP6 models, Ocean Sci., 17, 59–90. https://doi.org/10.5194/os-17-59-2021 Hirschi JJM, Barnier B, Böning C, Biastoch A, Blaker AT, Coward A, et al. (2020) The Atlantic meridional overturning circulation in high‐resolution models. Journal of Geophysical Research: Oceans, 125, e2019JC015522. https://doi.org/10.1029/2019JC015522 Jackson LC, Biastoch A, Buckley MW, Desbruyères DG, Frajka-Williams E, Moat B, Robson J (2022) The evolution of the North Atlantic Meridional Overturning Circulation since 1980. Nat. Rev. Earth Environ. 3, 241–254. (doi:10.1038/s43017-022-00263-2) Jackson LC, Kahana R, Graham T, Ringer MA, Woollings T, Mecking JV, Wood RA (2015) Global and European climate impacts of a slowdown of the AMOC in a high-resolution GCM. Clim Dyn 45, 3299–3316. https://doi.org/10.1007/s00382-015-2540-2 Jackson LC, Roberts MJ, Hewitt HT, Iovino D, Koenigk T, Meccia VL, Roberts CD, Ruprich-Robert Y, Wood RA (2020) Impact of ocean resolution and mean state on the rate of AMOC weakening. Clim Dyn 55, 1711–1732. https://doi.org/10.1007/s00382-020-05345-9 Jiang Z, Brierley C, Thornalley D, Sax S (2023) No changes in overall AMOC strength in interglacial PMIP4 time slices, Clim. Past, 19, 107–121. https://doi.org/10.5194/cp-19-107-2023 Karami MP, Koenigk T, Wang S, Navarro-Labastida R, Kruschke T, Carreric A, Ortega P, Wyser K, Fuentes Franco R, de Boer, AM, Sicard M, Aldama Campino A (2025) Historical Climate and Future Projection in the North Atlantic and Arctic: Insights from EC-Earth3 High-Resolution Simulations, EGUsphere [preprint]. https://doi.org/10.5194/egusphere-2025-2653 Keil P, Mauritsen T, Jungclaus J, Hedemann C, Olonscheck D, Ghosh R (2020) Multiple drivers of the North Atlantic warming hole. Nat. Clim. Chang. 10, 667–671. https://doi.org/10.1038/s41558-020-0819-8 Koenigk T, Fuentes-Franco R, Meccia VL, Gutjahr O, Jackson LC (2021) New, A.L., Ortega, P., Roberts, C.D., Roberts, M.J., Arsouze, T., Iovino, D., Moine, M.P., Sein, D.V.: Deep mixed ocean volume in the Labrador Sea in HighResMIP models. Clim Dyn 57, 1895–1918. https://doi.org/10.1007/s00382-021-05785-x Kosto Y, Johnson HL, Marshall DP, Heimbach P, Forget G, Penny-Holliday N, Lozier MS, Li F, Pillar HR, and Smith T (2021) Distinct sources of interannual subtropical and subpolar Atlantic overturning variability. Nat. Geosci. 14, 491–495. https://doi.org/10.1038/s41561-021-00759-4 Levang SJ, Schmitt RW (2021) What Causes the AMOC to Weaken in CMIP5? J. Climate, 33, 1535–1545. https://doi.org/10.1175/JCLI-D-19-0547.1 Li F, Lozier MS, Bacon S, et al. (2021) Subpolar North Atlantic western boundary density anomalies and the Meridional Overturning Circulation. Nat Commun 12, 3002. https://doi.org/10.1038/s41467-021-23350-2 Lozier M, et al. (2019) A sea change in our view of overturning in the subpolar North Atlantic.Science363,516-521. DOI:10.1126/science.aau6592 Lozier M, Roussenov V, Reed M, et al. (2010) Opposing decadal changes for the North Atlantic meridional overturning circulation. Nature Geosci 3, 728–734. https://doi.org/10.1038/ngeo947 Marotzke J (2000) Abrupt Climate Change and Thermohaline Circulation: Mechanisms and Predictability. Proceedings of the National Academy of Sciences of the United States of America, 97, 1347-1350. https://doi.org/10.1073/pnas.97.4.134 Marotzke J, Scot JR (1999) Convective Mixing and the Thermohaline Circulation. J. Phys. Oceanogr., 29, 2962–2970. https://doi.org/10.1175/1520-0485(1999)029 McCarthy GD, Caesar L (2023) Can we trust projections of AMOC weakening based on climate models that cannot reproduce the past? Phil. Trans. R. Soc. A 381: 20220193. https://doi.org/10.1098/rsta.2022.0193 McPhaden M, Zhang D (2002) Slowdown of the meridional overturning circulation in the upper Pacific Ocean. Nature 415, 603–608. doi: 10.1038/415603a Petit T, Lozier MS, Josey SA, Cunningham SA (2020) Atlantic deep water formation occurs primarily in the Iceland Basin and Irminger Sea by local buoyancy forcing. Geophysical Research Letters, 47, e2020GL091028. https://doi.org/10.1029/2020GL091028 Petit T, Robson J, Ferreira D, Yeager S, Evans DG (2025) Coherence of the AMOC over the subpolar North Atlantic on interannual to multiannual time scales. Geophysical Research Letters, 52, e2025GL115171. https://doi.org/10.1029/2025GL115171 Rahmstorf S (2024) Is the Atlantic overturning circulation approaching a tipping point? Oceanography. https://doi.org/10.5670/oceanog.2024.501 Rayner NA, Parker DE, Horton EB, Folland CK, Alexander LV, Rowell DP, Kent EC, Kaplan A (2003) Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. JGRA, 108(D14), 4407. https://doi.org/10.1029/2002JD002670 Rohling EJ, Hibbert FD, Grant KM, et al. (2019) Asynchronous Antarctic and Greenland ice-volume contributions to the last interglacial sea-level highstand. Nat Commun 10, 5040. https://doi.org/10.1038/s41467-019-12874-3 Sévellec F, Fedorov A, Liu W (2017) Arctic sea-ice decline weakens the Atlantic Meridional Overturning Circulation. Nature Clim Change 7, 604–610. https://doi.org/10.1038/nclimate3353 Shan X, Sun S, Wu L, Wu L, Spall M (2024) Role of the Labrador Current in the Atlantic Meridional Overturning Circulation response to greenhouse warming. Nat Commun 15, 7361. https://doi.org/10.1038/s41467-024-51449-9 Sicard M, de Boer AM, Coxall HK, Koenigk T, Karami MP, Jakobsson M, O’Regan M (2023) Similarities and differences in Arctic sea-ice loss during the solar-forced Last Interglacial warming (127 Kyr BP) and CO2-forced future warming. Geophysical Research Letters, 50, e2023GL104782. https://doi.org/10.1029/2023GL104782 Tooth OJ, Foukal NP, Johns WE, Johnson HL, Wilson C (2024) Lagrangian decomposition of the Atlantic Ocean heat transport at 26.5°N. Geophysical Research Letters, 51, e2023GL107399. https://doi.org/10.1029/2023GL107399 van Westen RM, Kliphuis M, Dijkstra HA (2024) Physics-based early warning signal shows that AMOC is on tipping course, Science Advances, 10, eadk1189. https://doi.org/10.1126/sciadv.adk1189 Weijer W, Cheng W, Drijfhout SS, Federov AV, Hu A, Jackson LC, Liu W, McDonagh EL, Mecking JV, Zhang J (2019) Stability of the Atlantic Meridional Overturning Circulation: A review and synthesis. Journal of Geophysical Research: Oceans, 124, 5336–5375. https://doi.org/10.1029/2019JC015083 Zhang D, McPhaden M (2006) Decadal variability of the shallow Pacific meridional overturning circulation: Relation to tropical sea surface temperatures in observations and climate change models. Ocean Model. 15, 250–273. doi: 10.1016/j.ocemod.2005.12.005 Zou S, Lozier MS, Li F, Abernathey R, Jackson L (2020) Density-compensated overturning in the Labrador Sea. Nat. Geosci. 13, 121–126. https://doi.org/10.1038/s41561-019-0517-1 Supplementary Files ARTAMOCDWFECEarth3HRSuplementaryMaterial.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7542176","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":516109320,"identity":"9ef5f583-3739-4cc9-9df1-5c31a418819b","order_by":0,"name":"Rene Gabriel Navarro-Labastida","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAtUlEQVRIiWNgGAWjYDACZgaGA0BKDsQ+8IAULcZgLQmkWJbYACKJ0mLeznvwcEHN4fT5YYcfAm2xk9NtIKBF5jBfwuEZxw7nbrydZgDUkmxsdoCAFglmHoPDPGxpuRtnJ4C0HEjcRpyWf2nphrPTP5CghbfNJkFeOocUW3j7bAw3SOcUHEgwIMYv/GeMP/N8k5CXn52++cOHCjs5glrgwACs0oBY5SAg30CK6lEwCkbBKBhRAABPTUDT/bNLXwAAAABJRU5ErkJggg==","orcid":"","institution":"SMHI: Sveriges meteorologiska och hydrologiska institut","correspondingAuthor":true,"prefix":"","firstName":"Rene","middleName":"Gabriel","lastName":"Navarro-Labastida","suffix":""},{"id":516109321,"identity":"2a0404ef-e037-43f7-811d-db4bc480db8b","order_by":1,"name":"Mehdi Pasha Karami","email":"","orcid":"","institution":"SMHI: Sveriges meteorologiska och hydrologiska institut","correspondingAuthor":false,"prefix":"","firstName":"Mehdi","middleName":"Pasha","lastName":"Karami","suffix":""},{"id":516109322,"identity":"f538e25a-e1b0-4c1f-aa74-96be38d87361","order_by":2,"name":"Torben Koenigk","email":"","orcid":"","institution":"SMHI: Sveriges meteorologiska och hydrologiska institut","correspondingAuthor":false,"prefix":"","firstName":"Torben","middleName":"","lastName":"Koenigk","suffix":""},{"id":516109323,"identity":"3459fc20-4334-4108-b3a2-f84fdd15635c","order_by":3,"name":"Agatha de Boer","email":"","orcid":"","institution":"Stockholm University: Stockholms Universitet","correspondingAuthor":false,"prefix":"","firstName":"Agatha","middleName":"","lastName":"de Boer","suffix":""},{"id":516109324,"identity":"93ea00fc-39df-4a8b-bf1e-94c9afa94113","order_by":4,"name":"Marie Sicard","email":"","orcid":"","institution":"University of Leeds School of Earth and Environment","correspondingAuthor":false,"prefix":"","firstName":"Marie","middleName":"","lastName":"Sicard","suffix":""}],"badges":[],"createdAt":"2025-09-05 08:17:26","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7542176/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7542176/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":92161152,"identity":"e67a848c-23f9-4b7e-923c-8b9eee198f37","added_by":"auto","created_at":"2025-09-25 10:07:56","extension":"xml","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9092,"visible":true,"origin":"","legend":"","description":"","filename":"cldyCLDYD2500970.xml","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/5d856ebc18a0ebb046d0ffe5.xml"},{"id":92161154,"identity":"f5ade481-88f4-42b4-b36f-93794b828968","added_by":"auto","created_at":"2025-09-25 10:07:56","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1013,"visible":true,"origin":"","legend":"","description":"","filename":"CLDYD250097021730.go.xml","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/e3b955fd9fc3f071654224c6.xml"},{"id":92162218,"identity":"fcd4fd07-4c01-4001-bf66-330269dd6e72","added_by":"auto","created_at":"2025-09-25 10:15:56","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":795,"visible":true,"origin":"","legend":"","description":"","filename":"CLDYD2500970Import.xml","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/3ca16cbc09d4c03a09acf295.xml"},{"id":92163049,"identity":"0e3274d2-3c6e-47fa-9771-658ab7441f65","added_by":"auto","created_at":"2025-09-25 10:23:56","extension":"xml","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":177021,"visible":true,"origin":"","legend":"","description":"","filename":"CLDYD25009700enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/ec6d6ccbe5ca24c7a7faef11.xml"},{"id":92161155,"identity":"c6573039-d3e3-4843-a39d-f3aab8bb23da","added_by":"auto","created_at":"2025-09-25 10:07:56","extension":"jpeg","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":51879,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/9fe8f0a84c56d7d85d8ad491.jpeg"},{"id":92161159,"identity":"14bea850-8240-451e-9ee2-5628740c0581","added_by":"auto","created_at":"2025-09-25 10:07:56","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":49381,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/6a29b0f3393661b79c140f74.png"},{"id":92161177,"identity":"4a3509f0-7e9a-4354-aa39-a4a9ecc58e64","added_by":"auto","created_at":"2025-09-25 10:07:57","extension":"jpeg","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1238984,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/f068eaacf408862b2991729f.jpeg"},{"id":92161178,"identity":"e095957a-7ed2-415f-87b6-77f3bd575a8b","added_by":"auto","created_at":"2025-09-25 10:07:57","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":453641,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/90b762bbfe8dcd1eece41f58.png"},{"id":92161168,"identity":"3e59fb62-5f9f-461b-8532-dda6a0efaf40","added_by":"auto","created_at":"2025-09-25 10:07:57","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":173614,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/7648a8d1891b7fb04adb2bbc.png"},{"id":92162226,"identity":"3468391a-feb8-4cc3-b12e-ad01ebb05f42","added_by":"auto","created_at":"2025-09-25 10:15:57","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":335291,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/8c6f465007740609c465aa11.png"},{"id":92162229,"identity":"bb1e65c0-6d26-4451-851a-8ac48dadfc9d","added_by":"auto","created_at":"2025-09-25 10:15:57","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":736909,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/dac153bc7abe804b370ef3fe.png"},{"id":92161183,"identity":"31b5574d-159c-4c16-8863-041082145d05","added_by":"auto","created_at":"2025-09-25 10:07:57","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":258825,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/045250dcd77c2e47dd0a4cde.png"},{"id":92161171,"identity":"3fec89d9-13a1-4ad3-bc20-c36b30e84e2f","added_by":"auto","created_at":"2025-09-25 10:07:57","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":83352,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/a7034442aadeea9ada7336fd.png"},{"id":92162228,"identity":"e794a97e-3bfb-401a-80c2-fdb9b5da3c5f","added_by":"auto","created_at":"2025-09-25 10:15:57","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":313948,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/9128c72bfc2cd8abccb707c6.png"},{"id":92161172,"identity":"08635d32-1b58-485d-8f0b-07665317fef8","added_by":"auto","created_at":"2025-09-25 10:07:57","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":56354,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/d1778eace5459bbaf6bce5d7.png"},{"id":92162231,"identity":"355cc2ce-287c-43d3-84ef-943326d92825","added_by":"auto","created_at":"2025-09-25 10:15:57","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":17529,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/cd7d1f8cce07c61a3f10b57e.png"},{"id":92161181,"identity":"caf80d26-05a9-4ab3-87d3-c59b3719b087","added_by":"auto","created_at":"2025-09-25 10:07:57","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":322755,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/f8fcc46dcfef901125ad10af.png"},{"id":92161174,"identity":"b6fdcb2d-3949-4395-b686-ba2187bd6a95","added_by":"auto","created_at":"2025-09-25 10:07:57","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":99611,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/e9fe638a9f1d2ff95e32cf77.png"},{"id":92163050,"identity":"2610fa40-4f87-4773-a9c7-efe192a0253a","added_by":"auto","created_at":"2025-09-25 10:23:57","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53542,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/87ed0d2d5261ecd192f8586d.png"},{"id":92161166,"identity":"ef5e87ae-5916-452f-8455-9b704933dc41","added_by":"auto","created_at":"2025-09-25 10:07:56","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":63228,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/b9929ac2bb0e5c373b89ca11.png"},{"id":92162224,"identity":"398c26c2-a121-4961-bb52-0c312116ee68","added_by":"auto","created_at":"2025-09-25 10:15:57","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":171108,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/cba45e3cc44e71193880dbbb.png"},{"id":92162230,"identity":"14f37590-5693-45b2-932b-e7e299e57c5f","added_by":"auto","created_at":"2025-09-25 10:15:57","extension":"png","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":56040,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/a77a0d9b67eb393f2c77d604.png"},{"id":92161179,"identity":"8dce460d-de1d-497b-8149-82366a2bfc0a","added_by":"auto","created_at":"2025-09-25 10:07:57","extension":"png","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":28833,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/503f808d44139910d8e0957b.png"},{"id":92161186,"identity":"4ede28bb-b969-44ee-a03f-f430308f472a","added_by":"auto","created_at":"2025-09-25 10:07:57","extension":"png","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":64704,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/4db6bbc2d38aec7708f9679f.png"},{"id":92163051,"identity":"a3d0f644-2839-49ea-b3b7-f24e104892bb","added_by":"auto","created_at":"2025-09-25 10:23:57","extension":"xml","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":175002,"visible":true,"origin":"","legend":"","description":"","filename":"CLDYD25009700structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/97b41dad77591439c758d2ea.xml"},{"id":92162227,"identity":"e20a421e-03d8-4426-aa2e-b10d28f6d4ff","added_by":"auto","created_at":"2025-09-25 10:15:57","extension":"html","order_by":27,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":183216,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/9410641e6be0056680387867.html"},{"id":92161151,"identity":"6e5f32fc-3031-43e4-96ec-fa389bab47be","added_by":"auto","created_at":"2025-09-25 10:07:56","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":495093,"visible":true,"origin":"","legend":"\u003cp\u003eBoxes indicate the North Atlantic regions of Labrador (LAB), Greenland-Iceland-Norway (GIN), and Irminger Sea (IRM) where Deep Mixed Volume (DMV) and Deep Water Formation (DWF) are calculated. At 26°N and 45°N meridional transports are calculated\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/69830e75dab6e30d9c8538e5.png"},{"id":92163047,"identity":"509d5388-8df1-4bc1-9230-27886ae76615","added_by":"auto","created_at":"2025-09-25 10:23:56","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":3365166,"visible":true,"origin":"","legend":"\u003cp\u003eAnnual mean climatologies of sea surface salinity (psu) (a-c), temperature (\u003csup\u003eo\u003c/sup\u003eC) (d-f), and potential density anomaly (Kgm\u003csup\u003e-3\u003c/sup\u003e) (g-h) using only winter-time data (January-March). Results for the climatological mean under pre-industrial control simulation (a,d,g) and the changes under the two CO₂-fixed concentration experiments: CO₂400 (b, e, h)\u0026nbsp; and CO₂550 (c, f, i), respectively. Changes (b-c, e-f, and h-i) have been calculated using the control (a, d, g) as the reference state. Calculations only consider the last 100 years of each simulation. The dashed line indicates the region of zero change. The thick blue line in Figures a-c represents the 15% sea-ice area fraction\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/85d66077181e4cb104b55011.png"},{"id":92162217,"identity":"7d37dafc-a4d7-4c88-b094-5e38e95ad514","added_by":"auto","created_at":"2025-09-25 10:15:56","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":490713,"visible":true,"origin":"","legend":"\u003cp\u003eConvection frequency (percentage of years with convection happening during March) under pre-industrial control simulation, and the two CO₂-fixed concentration experiments (CO₂400 and CO₂550). Computation considers a critical depth of 1000 meters in LAB and IRM Seas, and 700 meters depth for GIN. Only values from March during the last 100 years of each simulation are used\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/668a2a94bd7ca1a640dcbdcb.png"},{"id":92161150,"identity":"19622150-4d95-4911-a363-908a769a795f","added_by":"auto","created_at":"2025-09-25 10:07:56","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":528865,"visible":true,"origin":"","legend":"\u003cp\u003eAnnual mean, spatially averaged vertical profiles of potential density for the Labrador (LAB), Greenland-Iceland-Norway (GIN), and Irminger (IRM) regions. The results are for the pre-industrial control and the two fixed CO₂ experiments (CO₂400 and CO₂550)\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/b6d64ab8f0e5c980cf9566ee.png"},{"id":92163048,"identity":"b416c6b1-02e1-409d-95f2-09008c82341c","added_by":"auto","created_at":"2025-09-25 10:23:56","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1195702,"visible":true,"origin":"","legend":"\u003cp\u003eMarch Deep Mixed Volume (DMV) in Labrador (LAB) (a-c), Greenland-Iceland-Norway (GIN) (d-f), and Irminger Sea (IRM) (g-i) under pre-industrial control simulation and the two CO₂-fixed concentration experiments (CO₂400 and CO₂550).\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/8582882634eed1c62cfed23e.png"},{"id":92161161,"identity":"8b66997f-6648-4737-b024-675ca524ec48","added_by":"auto","created_at":"2025-09-25 10:07:56","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":1157530,"visible":true,"origin":"","legend":"\u003cp\u003eAnnual mean stream function or Atlantic Meridional Overturning Circulation (AMOC) volume transport, under pre-industrial control simulation and the two CO₂-fixed concentration experiments (CO₂400 and CO₂550)\u003c/p\u003e","description":"","filename":"floatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/40389131920aa1ed202bdeee.png"},{"id":92161164,"identity":"bce610ab-4504-4e36-a449-0b89b97c5026","added_by":"auto","created_at":"2025-09-25 10:07:56","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":845053,"visible":true,"origin":"","legend":"\u003cp\u003eAtlantic Meridional Overturning Circulation (AMOC) volume transport strength index, under pre-industrial control simulation and the two CO₂-fixed concentration experiments (CO₂400 and CO₂550)\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/49a34686478935ddafe127f2.png"},{"id":92161160,"identity":"59e3f266-f851-405f-879d-5a0cd52b6dcf","added_by":"auto","created_at":"2025-09-25 10:07:56","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":190881,"visible":true,"origin":"","legend":"\u003cp\u003eThe net horizontal mass convergences in the upper layer (a) and in the lower layer (b) within each convective region (volume transport in Sv ≡ 10\u003csup\u003e6\u003c/sup\u003e m\u003csup\u003e3\u003c/sup\u003es\u003csup\u003e-1\u003c/sup\u003e). Net convergence in the lower layer is the Deep Water Formation (DWF). Values are for the pre-industrial control (black text), and the moderate (CO₂400) and high (CO₂550) concentration CO₂ experiments (blue and red text). Mass flows across the different sections delimiting the regions of Labrador (LAB), Irminger (IRM), and Greenland-Iceland-Norway (GIN) are also shown (east, west, north, and south for each region). Section GIN-north and GIN-east are located across the Fram Strait and the Barents Sea Opening. There are no water exchanges across sections GIN-west in the upper layer and LAB-west and IRM-east in the lower layer because of topographical restrictions\u003c/p\u003e","description":"","filename":"floatimage8.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/65e81cb34a98e860fcbbff88.png"},{"id":92162221,"identity":"d7714ead-91d3-49eb-b98f-cffeea06ed21","added_by":"auto","created_at":"2025-09-25 10:15:56","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":852603,"visible":true,"origin":"","legend":"\u003cp\u003eDeep Water Formation (DWF) and AMOC index time series under pre-industrial control simulation and the two CO₂-fixed concentration experiments (CO₂400 and CO₂550). The DWF is calculated in the regions of Labrador (LAB), Irminger (IRM), and Greenland (GIN). The sum of all the individual DWF values (TOT), Arctic deep water equatorward transport (Arctic), and the net export (DWF TOT + Arctic) is also shown\u003c/p\u003e","description":"","filename":"floatimage9.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/c029556a8d99e447d430d080.png"},{"id":92162223,"identity":"4cd9863c-6323-4991-a3ca-8463d16ecf27","added_by":"auto","created_at":"2025-09-25 10:15:57","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":141750,"visible":true,"origin":"","legend":"\u003cp\u003e\u0026nbsp;Schematic showing poleward transports (gray arrows), equatorward deep transports (dark arrows), and horizontal circulation within the subtropical and subpolar gyres (white arrows) (volume transport in Sv ≡ 10\u003csup\u003e6\u003c/sup\u003e m\u003csup\u003e3\u003c/sup\u003es\u003csup\u003e-1\u003c/sup\u003e). Deep Water Formation (DWF) carried by the three convective regions of Labrador, Irminger, and Greenland-Iceland-Norway Seas (LAB-IRM-GIN) is also shown (vertical red arrow). Horizontal exchanges between LAB-IRM-GIN and the Arctic Ocean (gray area) are taking place mainly through the Arctic Gateways of Fram Strait and the Barents Sea. The DWF contribution from the Arctic Ocean has not been calculated, and deep water export toward the lower latitudes is inferred from the LAB-IRM-GIN mass balance. Non-resolved vertical recirculation and mixing because of our offline postprocessing and the use of annual means are also schematically depicted. Results are for the pre-industrial control (black text), and the moderate (CO₂400) and high (CO₂550) concentration CO₂ experiments (blue and red text)\u003c/p\u003e","description":"","filename":"floatimage10.png","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/64030b5c495ba6a01507fab9.png"},{"id":102295418,"identity":"2be8212e-6a37-45fc-876c-27ded54f4cae","added_by":"auto","created_at":"2026-02-10 10:11:08","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":10144561,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/8fa37177-0e4b-4d14-876d-07aaed142861.pdf"},{"id":92162222,"identity":"74aa813b-d94e-4bef-be66-d8bdbfa4f799","added_by":"auto","created_at":"2025-09-25 10:15:56","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":311159,"visible":true,"origin":"","legend":"","description":"","filename":"ARTAMOCDWFECEarth3HRSuplementaryMaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-7542176/v1/2228a4faa8463ff9430d69bf.docx"}],"financialInterests":"","formattedTitle":"North Atlantic Ocean circulation and deep water formation under warmer climate conditions in EC-Earth3-HR","fulltext":[{"header":"1 Introduction","content":"\u003cdiv id=\"Sec2\" class=\"Section2\"\u003e\u003ch2\u003e1.1 Overview\u003c/h2\u003e\u003cp\u003eThe Atlantic Meridional Overturning Circulation (AMOC) is a crucial component of the Earth's climate system, responsible for the large-scale redistribution of heat and anthropogenic carbon (Buckley and Marshall 2015; Jackson et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Trans-basin observational results and projections from the latest phase of collaboration under the Coupled Model Intercomparison Project (CMIP6) both suggest a likely AMOC weakening driven by rising CO2 levels (McCarthy and Caesar \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Baker et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2023\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). However, the transient response of the AMOC to climate forcing varies significantly across models, making it challenging to predict the precise future behavior of the circulation. Although most CMIP6 models assess a full AMOC collapse as unlikely within the 21st century, a recent analysis of the extended simulations of CMIP6 models reveals AMOC shutdown under the high-emission scenario. This suggests a significantly higher risk than previously assumed (Dijkhout et al. 2025). Furthermore, recent estimates using fingerprint attribution suggest that the AMOC could collapse around mid-century under the current emissions scenario (Ditlevsen and Ditlevsen 2023). Nonetheless, this approach may overlook complex interactions between the AMOC and global warming, partly because it assumes the AMOC directly drives long-term subpolar sea surface temperature changes. Despite the limitations, this has increased interest in monitoring early warning signals to anticipate the potential collapse of AMOC (Ditlevsen and Ditlevsen 2023; Rahmstorf 2023; van Westen et al. \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe potential for a nearly total AMOC collapse has been previously assessed under both hypothetical scenarios where large freshwater fluxes are introduced to the North Atlantic Ocean (Jackson et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; Bellomo et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; van Westen et al. \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and in the context of millennial-scale climate changes, using models configured for glacial conditions and long-term simulations designed to test the stability of the AMOC under past climate scenarios (Weijer et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Based on such studies, it is believed that a large freshwater anomaly may reduce the Northern Atlantic ocean's salinity, halting deep-water formation in the North Atlantic and rapidly weakening the AMOC, leading to severe cooling over the European Arctic, where temperatures could drop by 10\u0026deg;C compared to current levels (Jackson et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2015\u003c/span\u003e; van Westen et al. \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eConversely, under future emission scenarios, strengthened freshwater fluxes and subsurface warming are projected to diminish deep convective activity in the North Atlantic, particularly in the Labrador Sea (Brodeau and Koenigk \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Levang and Schmitt 2020). Nevertheless, the establishment of a direct link between short-timescale convection in the Labrador Sea and deep water production and the overturning circulation remains an open challenge (Lozier et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Lozier et al. 2021). This suggests that regions within the North Atlantic with weaker stratification, such as the Irminger Sea and the Greenland-Iceland-Norway Sea, may sustain deepwater production, even in the event of a total shutdown of convection in the Labrador Sea (Jackson et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Dimma et al. 2022; Bretones et al. 2022; Chafik et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; \u0026Aring;rthun et al. 2023, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Therefore, comprehensive understanding of the convection process in the region is required, as it is hypothesized that these processes play a pivotal role in maintaining the AMOC's overall strength (Koenigk et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Roberts et al. 2020).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e1.2 Model resolution\u003c/h2\u003e\u003cp\u003ePrevious multi-model analyses suggest that high-resolution models tend to simulate stronger AMOC weakening compared to lower-resolution Earth System Models (Bellomo et al. 2021; Roberts et al. 2020; Jackson et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). This difference is partly due to the tendency of lower-resolution models to produce an overly stable AMOC (e.g., Liu et al. 2014), whereas high-resolution models appear to capture AMOC variability more realistically (Koenigk et al. \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Jackson et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Hirschi et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Shan et al. \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Karami et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). In addition, the rate of decline still varies with model resolution due to model-dependent behavior, adding a further complication (Roberts et al. 2020; Jackson et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Hirschi et al. \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Nevertheless, high-resolution coupled models offer more accurate representations of Arctic sea ice, including sea ice area, sea ice volume, the position of the ice edge (Docquier et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), in addition to improved representation of the poleward heat transport within the climate system (Grist et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2018\u003c/span\u003e; Docquier et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Overall, high-resolution CMIP models enhance the representation of large-scale circulation patterns and small-scale processes, including climate extremes (Haarsma et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) while improving the simulation of sea surface temperature, velocity, and the depiction of boundary currents across all ocean basins (Gutjahr et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2019\u003c/span\u003e; Docquier et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e1.3 Motivation\u003c/h2\u003e\u003cp\u003eWhile the aforementioned studies are valuable for understanding the transient climate response under increasing greenhouse gas concentrations, they primarily address short-term dynamics rather than the long-term adjustments of the Earth system. By isolating the ultimate impacts of long-term high CO₂ levels, we may assess more accurately the consequences of climate change under various warming scenarios, regardless of the pathways through which these levels are reached. To address this, we are analyzing fixed CO\u003csub\u003e2\u003c/sub\u003e concentration experiments that represent quasi-equilibrium climate states under warmer conditions. By doing this we further reduce uncertainties in future projections involving increased atmospheric CO₂ concentrations, particularly concerning North Atlantic oceanic processes such as the AMOC. The warmer climate conditions represent global temperature anomalies of 1\u0026deg;C and 2\u0026deg;C compared to the pre-industrial state.\u003c/p\u003e\u003cp\u003eAt a global temperature anomaly of 1\u0026deg;C, the level of Arctic warming (~\u0026thinsp;3.7\u0026deg;C; Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ea in supplementary material) is comparable to conditions during the Last Interglacial period (LIG), which is often regarded as a potential analog for future high-latitude climate scenarios due to similar temperature increases (Sicard et al. \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Proxy-reconstructions and climate modeling studies suggest that during LIG, the AMOC may have been weaker than in pre-industrial times when freshwater forcing is included (e.g., Galaasen et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Govin et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Guarino et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). However, there is still some debate in this matter, as recent multimodel simulations representing different interglacial conditions show minimal changes in AMOC strength during the mid-Holocene (Jiang et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), whereas earlier studies suggest that a weakened AMOC likely contributed to significant large-scale climate disruptions in Europe, including shifts in temperature and precipitation patterns, as well as potentially extreme weather events (Govin et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2012\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eA global temperature anomaly of 2\u0026deg;C represents a closed analog to the global mean temperature during LIG, the climate that can be expected if the Paris Agreement is successfully implemented (Fisher et al. 2018; Rohling et al. \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), and much more higher level of warming in the Arctic (~\u0026thinsp;6.7\u0026deg;C; Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ea).\u003c/p\u003e\u003c/div\u003e"},{"header":"2 Materials and methods","content":"\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Model description and experimental design\u003c/h2\u003e\u003cp\u003eIn this study, we use the high-resolution version of the global coupled climate model EC-Earth3 (hereafter referred to as EC-Earth3-HR), which was used to contribute to CMIP6 in its standard-resolution (D\u0026ouml;scher et al. 2022). Our configuration includes the IFS atmosphere model (cycle 36r4), with the HTESSEL land surface module, and the NEMO ocean model (version 3.6), coupled with the LIM3 sea ice model. The atmosphere was configured with a T511 spectral resolution (~\u0026thinsp;40 km), while the ocean model used the ORCA025 configuration, corresponding to approximately 0.25 degrees. In the vertical domain EC-Earth3-HR employs 91 levels in the atmosphere and 75 layers in the ocean. The model underwent a tuning process, multi-centennial spin-up, followed by a 100-year spin-up with the final set of parameters. After the spin-up process, we started a 350-year pre-industrial (pi-control) control run, and a 1% per year increasing CO₂ experiment (1pctCO₂) (details in Karami et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eFrom the 1pctCO₂ experiment, two further experiments with fixed CO₂ concentrations were started from the points in time when global temperature anomalies reached around 1\u0026deg;C and 2\u0026deg;C compared to the EC-Earth3-HR pre-industrial (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003eb). In terms of the CO\u003csub\u003e2\u003c/sub\u003e concentrations in the 1pctCO\u003csub\u003e2\u003c/sub\u003e these levels of warming represent 400.9 ppm and 551.5 ppm and the years 1885 and 1917, respectively. After fixing atmospheric CO₂ concentrations, the model underwent a stabilization phase lasting approximately 33 and 40 years, before reaching quasi-equilibrium state at around 1.3\u0026deg;C and 2.7\u0026deg;C above the EC-Earth3-HR pre-industrial levels, respectively. Subsequently, both simulations were extended for an additional 100 years. Both fixed CO₂ experiments showed a rapid reduction of summer sea ice, with sea ice disappearing by the end of the simulation under the 551.5 ppm CO₂ concentration (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003ec).\u003c/p\u003e\u003cp\u003eIn this study, we focus on analyzing the two fixed CO₂ concentration experiments. Changes are assessed using the PI-control simulation as the reference state. All calculations are based on the final 100 years of each simulation. From here on, we refer to the fixed CO₂ experiment at 400.9 ppm as CO₂400 (moderate concentration) and the one at 551.5 ppm as CO₂550 (high concentration).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e2.2 Deep Mixed Volume (DMV)\u003c/h2\u003e\u003cp\u003eIn our analysis we use the Deep Mixed Volume (DMV) index to quantify deep convection (Brodeau and Koenigk \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). The computation involves integrating the volume of mixed water masses below a critical depth, using only winter data (March). This calculation is performed for both the control and CO₂ experiments. The DMV is computed for the Labrador (LAB), Irminger (IRM), and Greenland-Iceland-Norway (GIN) regions. The calculation involves selecting relatively large areas to cover all the convective spots within each region (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Only grid points that meet the depth criteria are considered. This refers to a critical depth of 1000 meters for LAB and IRM, and 700 meters in GIN. The selection of depth criteria is mainly based on topographical and dynamical constraints outlined in previous analyses (Brodeau and Koenigk \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Additionally, we calculate the frequency of convection events in line with DMV computation. The calculation is based on the event frequency of the winter mixed layer reaching the above-mentioned critical depths (frequency in percentage defined as the number of years with convection). This metric provides insights into the frequency and persistence of convective areas, as well as information on their spatial distribution.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e2.3 Deep Water Formation (DWF)\u003c/h2\u003e\u003cp\u003eWe apply a novel methodology to assess Deep Water Formation (DWF) based on calculating the horizontal volume convergence in the North Atlantic (Karami et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). The DWF index is computed using the same spatial division as in the DMV calculation (region boundaries in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The vertical column is divided into an upper and a lower layer in each region, divided at the same depth as in DMV computation to maintain consistency in our analysis.\u003c/p\u003e\u003cp\u003eFollowing this approach, we calculate the horizontal mass convergence as the net inflow and outflow of volume across sections within each layer. Positive convergence in the upper layer represents water sinking toward the lower layer (below the critical depth), which corresponds to deep water formation. In contrast, negative convergence in the lower layer indicates the export of the downwelled upper water. The difference between upper and lower layer convergence is attributed to the net surface freshwater fluxes in the upper layer and the use of annual means in the computation of the metric. However, our results show that this contribution is small (~\u0026thinsp;0.1 Sv) and shows minor changes between experiments.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e2.4 Meridional transports at 26\u0026deg;N and 45\u0026deg;N\u003c/h2\u003e\u003cp\u003eThe net meridional flow at 26\u0026deg;N has been vertically divided into upper and lower layers at a depth of 1000 meters. The characterization of the flow at 26\u0026deg;N is meant to zonally cross the Subtropical Gyre system and match our estimation of the AMOC index, defined as the time series of annual-mean maximum volume transport at this latitude (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The section is further divided horizontally into a western and eastern subsections. The boundary between subsections (80\u0026deg;W) is designed to avoid regions of water mass recirculation (McPhaden and Zhang 2004; Zhang and McPhaden \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). The net meridional poleward transport in the upper layer is assumed to represent the flow out of the gyre toward the subpolar region, while the net equatorward transport in the lower layer is assumed to represent the return of this water as deep water. We do not explicitly isolate the wind-driven circulation contribution within the gyre; thus, transports in the upper layer include both the Ekman and the geostrophic gyre circulation. However, all Ekman transport is assumed to subduct within the subtropical gyre.\u003c/p\u003e\u003cp\u003eThe Overturning in the Subpolar North Atlantic Programme (OSNAP) is often used as an analog for AMOC circulation across the subpolar gyre (approximately at 50\u0026deg;N) (Fu et al. 2023; Lozier et al. \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). However, by doing so, the role of the Irminger-Iceland and Labrador Seas contributions is ignored (Petit et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). In that regard, meridional transport at 45\u0026deg;N was chosen to characterize the export of deep water from the North Atlantic convective regions. At this location, the meridional flow has been vertically and horizontally divided, with horizontal outcrop at 40\u0026deg;W to pair the flow across the southern limits of the LAB and IRM regions (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e"},{"header":"3 Results","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e3.1 North Atlantic surface waters\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e shows the wintertime climatologies of sea surface salinity (SSS), sea surface temperature (SST), and potential sea surface density (SSD) within the main convective regions in the North Atlantic. In the control simulation (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea, d, g), warmer and saltier waters from the subtropics spread widely all over the IRM region. Following the counterclockwise circulation pattern of the subpolar gyre, this water continues spreading westward toward LAB and northward into GIN via the Iceland\u0026ndash;Norway section. Surface density in IRM is relatively low compared to LAB and GIN (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eg) due to the warmer waters in the IRM. In contrast, high SSD in LAB and GIN is driven by the strong surface cooling in the region, combined with the inflow of saltier waters from IRM. The relatively weak density along the eastern Greenland coastline in GIN is attributed to the influence of fresher waters spreading from the Arctic Ocean (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eg). The spread of this water into the relatively salty GIN region promotes the formation of a poleward meridional density gradient, which is believed to be the main driver of the density-driven export of Atlantic waters toward the Arctic, primarily through the eastern boundary Nordic Seas (\u0026Aring;rthun et al. 2023, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Consequently, most of the dense surface water in the GIN is expected to exit the region and continue northward.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eIn the moderate CO₂ concentration experiment (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb, e, h), simulated changes in SST, SSS, and SSD are explained in terms of local surface warming and freshwater flux changes, and partly by redistribution of these anomalies within the region. In detail, the general warming of the surface is due to a warmer atmosphere, except within the warming hole area, where a reduced overturning circulation leads to reduced ocean heat transport into this region (Keil et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The widespread salinity reduction is likely explained in terms of the increased precipitation.\u003c/p\u003e\u003cp\u003eThe saltier and warmer waters dominate the GIN region, particularly between the Denmark Strait and the Barents Sea Opening is arguably explained in terms of the simulated sea ice retreat in the region (see blue line in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea, b). As a consequence, the horizontal density gradient seems to intensify, suggesting strengthened density-driven export of waters toward the Arctic. The strengthened Arctic and Atlantic waters spread along the eastern Greenland coastline are likely to contribute to further shaping the above-mentioned changes.\u003c/p\u003e\u003cp\u003eThe changes in the high CO₂ experiment (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec, f, i) are similar to those in the moderate concentration case but are only intensified. On average, all regions show intensification of net surface freshening and warming (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). As in the moderate case, the strong haline change in GIN related to the sea ice edge retreat (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec) seems to further enhance the poleward meridional density gradient discussed above.\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eNorth Atlantic sea surface salinity, temperature, and density averaged over the entire regions of Labrador (LAB), Greenland-Iceland-Norway (GIN), and Irminger Sea (IRM). Deep Mixed Volume (DMV) is also shown. The results are for the pre-industrial control and the two fixed CO₂ experiments (CO₂400 and CO₂550). Only data from the last 100 years of simulation are used\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eExp.\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eLAB\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eGIN\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eIRM\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSea surface Salinity\u003c/p\u003e\u003cp\u003e[psu]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eControl\u003c/p\u003e\u003cp\u003eCO₂400\u003c/p\u003e\u003cp\u003eCO₂500\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e34.07\u003c/p\u003e\u003cp\u003e33.85\u003c/p\u003e\u003cp\u003e33.71\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e34.40\u003c/p\u003e\u003cp\u003e34.37\u003c/p\u003e\u003cp\u003e34.36\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e35.33\u003c/p\u003e\u003cp\u003e35.23\u003c/p\u003e\u003cp\u003e35.17\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSea surface Temperature\u003c/p\u003e\u003cp\u003e[\u003csup\u003eo\u003c/sup\u003eC]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eControl\u003c/p\u003e\u003cp\u003eCO₂400\u003c/p\u003e\u003cp\u003eCO₂500\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.76\u003c/p\u003e\u003cp\u003e2.65\u003c/p\u003e\u003cp\u003e3.57\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.30\u003c/p\u003e\u003cp\u003e1.74\u003c/p\u003e\u003cp\u003e3.09\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e8.62\u003c/p\u003e\u003cp\u003e9.02\u003c/p\u003e\u003cp\u003e9.62\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSea surface Density\u003c/p\u003e\u003cp\u003e[kg/m\u003csup\u003e3\u003c/sup\u003e]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eControl\u003c/p\u003e\u003cp\u003eCO₂400\u003c/p\u003e\u003cp\u003eCO₂500\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e27.06\u003c/p\u003e\u003cp\u003e26.80\u003c/p\u003e\u003cp\u003e26.61\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e27.43\u003c/p\u003e\u003cp\u003e27.30\u003c/p\u003e\u003cp\u003e27.18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e27.27\u003c/p\u003e\u003cp\u003e27.14\u003c/p\u003e\u003cp\u003e26.99\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDMV\u003c/p\u003e\u003cp\u003e[x10\u003csup\u003e15\u003c/sup\u003em\u003csup\u003e3\u003c/sup\u003e]\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eControl\u003c/p\u003e\u003cp\u003eCO₂400\u003c/p\u003e\u003cp\u003eCO₂500\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.23\u003c/p\u003e\u003cp\u003e0.75\u003c/p\u003e\u003cp\u003e0.21\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.51\u003c/p\u003e\u003cp\u003e0.44\u003c/p\u003e\u003cp\u003e0.07\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.09\u003c/p\u003e\u003cp\u003e0.16\u003c/p\u003e\u003cp\u003e0.11\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003e3.2 Changes in deep convection\u003c/h2\u003e\u003cp\u003eFigure \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e shows the frequency of convection events in the control simulation. Convection in LAB and GIN occurs more frequently and over larger areas compared to IRM, where it is relatively infrequent and confined to a smaller region. In the moderated CO₂ experiment (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eb), the convection area in LAB has shrunk and shifted northward together with the northward-moving ice edge (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). In IRM, both the convection area and frequency shows a moderate increase under this experiment. The intensified convection in the GIN Seas is associated with increased mixed layer depth values north of Svalbard, suggesting the emergence of new convective sites farther north.\u003c/p\u003e\u003cp\u003eDeep convection in LAB and GIN is significantly reduced under high CO₂ concentration scenarios, with events becoming increasingly infrequent (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). In IRM, convection under high CO₂ concentration shows similar values compared to those observed in the moderated case, with a higher occurrence of convection than in the control. In the control and moderated CO₂ experiments, IRM convection appears to be an extension of the LAB convection area, while under the high CO₂ experiment, it becomes a separate convection spot.\u003c/p\u003e\u003cp\u003eIn both CO₂ experiments, changes in convection area and frequency align with the weakened vertical profiles of potential density in each region (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). This is especially pronounced within the upper 500 meters in LAB and GIN, suggesting enhanced vertical stratification in those regions. In contrast, vertical density changes in the IRM region are weaker, likely explaining the absence of significant changes in convection there.\u003c/p\u003e\u003cp\u003eConsistent with the reduced densities in the LAB, GIN, and IRM seas, we find a decrease in DMV under both CO₂400 and CO₂550 (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e and Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Figure\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, shows the time series for the DMV index in the LAB, GIN, and IRM regions. In the control simulation (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ea, d, g), the strongest DMV values are observed in LAB, followed by GIN. All convective regions show strong interannual variability. In the moderate CO₂ concentration experiment (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eb, e, h), LAB shows strong weakening. In GIN, there is also some reduction, but less pronounced compared to LAB. In the high CO₂ experiment (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003ec, f, i), DMV values are further weakened. While DMW in LAB stays still present, it has almost completely disappeared in GIN. The DMV in IRM does not show a major change between the CO₂400 and CO₂550 experiments which is consistent with the small size of the convection areas and the relatively low convection frequency in this region (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eIn the control simulation, DMV in the LAB tends to occur under colder and saltier surface conditions, whereas in the GIN and IRM seas it occurs under warmer and saltier conditions. (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). In the IRM region, SST seems to be dominant. This is consistent with previous studies suggesting subpolar North Atlantic convection is influenced by large-scale climate patterns such as the Atlantic Multidecadal Oscillation (AMO) and the North Atlantic Oscillation (NAO) (Cheng and Zhang 2013; S\u0026eacute;vellec et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Dima et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Lozier et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). This has been previously explored in terms of the interannual to multiannual variability under historical and preindustrial conditions (Petit et al. \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), showing that positive NAO phases lead to intense heat loss over the western subpolar gyre. Conversely, negative NAO conditions can weaken the AMOC. More precisely, during negative NAO phases, reduced heat loss over the western subpolar gyre results in surface warming and decreased surface density, which weakens deep water formation in both the Irminger and Labrador Seas. This is later observed as a reduction in the southward outflow at 45\u0026deg;N.\u003c/p\u003e\u003cp\u003eIn GIN, DMV tends to occur under saltier surface conditions in both CO₂ scenarios (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). However, SST changes arguably have a greater impact, given the larger change in SST compared to SSS. In the IRM, SST-driven DMV changes remain strong under moderate CO₂ forcing but weaken substantially under high CO₂, while the effect of SSS in DMV weakens in both CO₂ scenarios. In the GIN Seas, DMV is robustly determined by salinity changes, highlighting the dominant role of SSS in modulating DMV in this region (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). This may suggest that as the climate warms, high SSS is no longer linked to warm SST, in contrast to the control conditions, where GIN convection occurs despite warm SST because the warm SST always occurs together with high salinity.\u003c/p\u003e\u003cp\u003eFinally, although convective mixing affects the properties of the densest surface waters, it remains uncertain whether buoyancy-driven overturning depends strictly on localized convection (Marotzke \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; Marotzke and Scott 1999). In addition, the widely held assumption that downwelling and convection occur in the same location may oversimplify the complex and often spatially decoupled processes that drive deep water formation.\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eTable 2\u003c/strong\u003e Linear correlation (r) and confidence intervals at 95%. The results are for the pre-industrial control and the two fixed CO₂ experiments (CO₂400 \u0026nbsp;and CO₂550). In the case of Sea Surface Temperature (SST), Sea Surface Salinity (SSS), and Deep Mixed Volume (DMV), only data from March is considered. Correlations using DMV and Deep Water Formation (DWF) were calculated within the Labrador (LAB), Greenland-Iceland-Norway (GIN), and Irminger (IRM) Seas. Correlations using the equatorward transport of deep water from the Arctic Ocean (Arc.trans) and the Atlantic Meridional Overturning Circulation (AMOC) index are shown as well. All values with no statistical significance are shown in red.\u003c/p\u003e\n\u003ctable style=\"border: none;border-collapse: collapse;width: 535px;\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 110.25pt;border: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92.95pt;border-top: 1pt solid black;border-right: 1pt solid black;border-bottom: 1pt solid black;border-image: initial;border-left: none;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003eControl\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 103.55pt;border-top: 1pt solid black;border-right: 1pt solid black;border-bottom: 1pt solid black;border-image: initial;border-left: none;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003eCO₂400\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94.45pt;border-top: 1pt solid black;border-right: 1pt solid black;border-bottom: 1pt solid black;border-image: initial;border-left: none;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003eCO₂550\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 110.25pt;border-right: 1pt solid black;border-bottom: 1pt solid black;border-left: 1pt solid black;border-image: initial;border-top: none;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003er(SST\u003csub\u003eLAB\u003c/sub\u003e, DMV\u003csub\u003eLAB\u003c/sub\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92.95pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e-0.31 [-0.48, -0.12]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 103.55pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e-0.53 [-0.66, -0.38]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94.45pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e-0.27 [-0.44, -0.07]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 110.25pt;border-right: 1pt solid black;border-bottom: 1pt solid black;border-left: 1pt solid black;border-image: initial;border-top: none;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003er(SST\u003csub\u003eIRM\u003c/sub\u003e, DMV\u003csub\u003eIRM\u003c/sub\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92.95pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e-0.67 [-0.77, -0.55]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 103.55pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e-0.56 [-0.69, -0.41]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94.45pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e\u003cspan style=\"color:red;\"\u003e-0.14 [-0.32, 0.06]\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 110.25pt;border-right: 1pt solid black;border-bottom: 1pt solid black;border-left: 1pt solid black;border-image: initial;border-top: none;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003er(SST\u003csub\u003eGIN\u003c/sub\u003e, DMV\u003csub\u003eGIN\u003c/sub\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92.95pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e0.50 [0.33, 0.63]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 103.55pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e\u003cspan style=\"color:red;\"\u003e0.04 [-0.16, 0.24]\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94.45pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e\u003cspan style=\"color:red;\"\u003e-0.14 [-0.33, 0.06]\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 110.25pt;border-right: 1pt solid black;border-bottom: 1pt solid black;border-left: 1pt solid black;border-image: initial;border-top: none;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003er(SSS\u003csub\u003eLAB\u003c/sub\u003e, DMV\u003csub\u003eLAB\u003c/sub\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92.95pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e0.31 [0.12, 0.48]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 103.55pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e0.26 [0.06, 0.43]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94.45pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e0.22 [0.02, 0.4]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 110.25pt;border-right: 1pt solid black;border-bottom: 1pt solid black;border-left: 1pt solid black;border-image: initial;border-top: none;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003er(SSS\u003csub\u003eIRM\u003c/sub\u003e, DMV\u003csub\u003eIRM\u003c/sub\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92.95pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e-0.37 [-0.53, -0.19]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 103.55pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e\u003cspan style=\"color:red;\"\u003e-0.02 [-0.21, 0.18]\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94.45pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e\u003cspan style=\"color:red;\"\u003e0.14 [-0.06, 0.33]\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 110.25pt;border-right: 1pt solid black;border-bottom: 1pt solid black;border-left: 1pt solid black;border-image: initial;border-top: none;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003er(SSS\u003csub\u003eGIN\u003c/sub\u003e, DMV\u003csub\u003eGIN\u003c/sub\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92.95pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e0.64 [0.5, 0.74]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 103.55pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e0.68 [0.56, 0.77]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94.45pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e0.45 [0.28, 0.6]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 110.25pt;border-right: 1pt solid black;border-bottom: 1pt solid black;border-left: 1pt solid black;border-image: initial;border-top: none;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003er(Arc.trans., DWF\u003csub\u003e\u0026nbsp;GIN\u003c/sub\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92.95pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e-0.89 \u0026nbsp; \u0026nbsp; [-0.93, -0.84]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 103.55pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e-0.81 \u0026nbsp; \u0026nbsp; [-0.87, -0.72]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94.45pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e-0.81 \u0026nbsp; \u0026nbsp; [-0.87, -0.73]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 110.25pt;border-right: 1pt solid black;border-bottom: 1pt solid black;border-left: 1pt solid black;border-image: initial;border-top: none;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003er(Arc.trans., AMOC)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92.95pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e\u003cspan style=\"color:red;\"\u003e-0.01 \u0026nbsp;[-0.21, 0.19]\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 103.55pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e\u003cspan style=\"color:red;\"\u003e-0.14 \u0026nbsp;[-0.06, 0.33]\u003c/span\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 94.45pt;border-top: none;border-left: none;border-bottom: 1pt solid black;border-right: 1pt solid black;padding: 2pt;height: 15.75pt;vertical-align: bottom;\"\u003e\n \u003cp style='margin:0cm;text-align:center;line-height:115%;font-size:13px;font-family:\"Times New Roman\",serif;'\u003e0.33 \u0026nbsp; \u0026nbsp; [0.14, 0.5]\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003e3.3 Changes in the AMOC\u003c/h2\u003e\u003cp\u003eIn the control simulation, the annual mean AMOC stream function reveals a well-defined deep overturning cell reaching down to 2500 m depth (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003ea). The subduction of the subsurface branch occurs between 40\u0026deg;N and 60\u0026deg;N. A maximum circulation core of 20 Sv is observed between 30\u0026deg; and 45\u0026deg;N, roughly at 1000 m depth. An anticlockwise bottom overturning cell is also present, with stronger circulation southward at 30\u0026deg;N. Additionally, we have calculated the AMOC strength index (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e), which is defined as the time series of the annual mean maximum volume transport between 24.5\u0026deg;N and 27.5\u0026deg;N at depths of 800\u0026ndash;1100 m.\u003c/p\u003e\u003cp\u003eIn both CO₂ concentration experiments, the AMOC weakens. The estimated weakening is 2.5 Sv (~\u0026thinsp;13%) and 4.5 Sv (~\u0026thinsp;22%) under the moderate and high CO₂ concentration experiments, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e). The increase in CO₂ concentrations causes significant changes in not only strength but also spatial distribution features, such as a shift of circulation cores toward lower latitudes, a more shallow upper cell, and a strong weakening of the bottom cell (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eb,c). These seem to agree with the reduction of northward ocean heat transport to the subpolar North Atlantic (not shown), which leads to diminished surface cooling (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ec, e, f) and weaker-shallower deep convection (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e and \u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003e3.4 Changes in deep water formation\u003c/h2\u003e\u003cp\u003eIn the control simulation, the total DWF in the North Atlantic amounts to ~\u0026thinsp;17 Sv from which the largest contribution is produced in the IRM region (~\u0026thinsp;9 Sv), followed by LAB (~\u0026thinsp;5 Sv) and GIN (~\u0026thinsp;3 Sv) (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003ea, b, and Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). In the moderate CO₂ concentration experiment, total DWF decreases by 20%, with larger declines in LAB (30%) and GIN (35%), while IRM exhibits a relatively minor change (10%). Under high CO₂ concentrations, total DWF drops further by 40%, with sharp declines in LAB (70%) and GIN (60%), while IRM shows no additional change.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eVolume transports (Sv\u0026thinsp;\u0026equiv;\u0026thinsp;10\u003csup\u003e6\u003c/sup\u003e m\u003csup\u003e3\u003c/sup\u003es\u003csup\u003e-1\u003c/sup\u003e) within Labrador (LAB), Greenland-Iceland-Norway (GIN), and Irminger (IRM) Seas (see Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), with positive values indicating inflows. Flows were calculated across given sections, which delimited each region (east, west, north, south). The results are for the pre-industrial control and the two fixed CO₂ experiments (CO₂400 and CO₂550). Results represent the mean values for the last 100 years of each simulation. The upper and lower flow contributions are shown. The Deep Water Formation index (DWF) is calculated as the horizontal mass convergence within each layer. Positive convergence in the upper layer represents water sinking toward the lower layer. Negative convergence in the lower layer is linked to deep water export outside the region. Total DWF is the sum of all the deep water volumes\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c4\" namest=\"c3\"\u003e\u003cp\u003eControl\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c6\" namest=\"c5\"\u003e\u003cp\u003eCO₂400\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c8\" namest=\"c7\"\u003e\u003cp\u003eCO₂550\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eUpper\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLower\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eUpper\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eLower\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eUpper\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003eLower\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"4\" rowspan=\"5\"\u003e\u003cp\u003e\u003cb\u003eLAB\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eEast\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e7.57\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-7.65\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.55\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-6.66\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e1.51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-4.42\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.02\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.02\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.02\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNorth\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.27\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.07\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.22\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e0.05\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e1.92\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e0.01\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSouth\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-4.78\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e2.36\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-2.48\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e3.17\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-2.12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e2.95\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDWF\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5.08\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-5.22\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.31\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-3.44\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e1.33\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-1.46\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"4\" rowspan=\"5\"\u003e\u003cp\u003e\u003cb\u003eGIN\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eEast\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-1.87\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-2.75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-3.05\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNorth\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-0.82\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.70\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-1.03\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e2.97\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-0.90\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e3.57\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSouth\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e5.75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-4.91\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5.90\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-5.18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e5.31\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-4.99\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDWF\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e3.06\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-3.21\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-2.21\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e1.36\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-1.42\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\" morerows=\"4\" rowspan=\"5\"\u003e\u003cp\u003e\u003cb\u003eIRM\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eEast\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.01\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.01\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e0.01\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-7.57\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e7.65\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-3.55\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e6.66\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-1.51\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e4.42\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eNorth\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-5.75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.91\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-5.90\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e5.18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-5.31\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e4.99\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eSouth\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e21.81\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-21.12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e16.83\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-19.29\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e14.24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-16.90\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDWF\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e8.50\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-8.56\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e7.39\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-7.45\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e7.43\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-7.49\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTotal DWF\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e16.64\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e-16.99\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e12.82\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-13.10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e10.12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e-10.37\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003e3.5 Deep water redistribution\u003c/h2\u003e\u003cp\u003eAt 45\u0026deg;N, the North Atlantic Ocean exchanges upper subpolar waters with deep waters formed within the LAB, IRM, and GIN regions. The upper water (21.8 Sv) enters through the IRM Sea (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003ea; Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). About 40% of this volume is transformed into deep water within IRM, while the remainder is routed toward LAB (~\u0026thinsp;35%) and GIN (~\u0026thinsp;25%). In both regions, most of this water is further converted into deep water. The remaining portion is exported southward from LAB and northward from GIN into the Arctic Ocean and the Barents Sea. Similarly, in the lower layer deep water (21.1 Sv) exits the region through the IRM Sea (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eb; Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e), but in this case, water export reflects the combined DWF from all regions (90%), along with a contribution from the Arctic Ocean (10%).\u003c/p\u003e\u003cp\u003eIn both CO₂ experiments, DWF weakens across all regions because of the water redistribution and density changes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e), with changes in upper and lower transports reflecting variations in strength rather than direction (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e). As a result, the overall circulation pattern remains consistent with the control simulation (arrows in Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003ea, b). However, reduced DWF in LAB leads to decreased water export toward IRM. In contrast, GIN deep water export remains stable because increased deep water inflow from the Arctic compensates for reduced deep water formation in the warmer conditions (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eb).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\u003ch2\u003e3.6 Meridional transports at 26\u0026deg;N and 45\u0026deg;N\u003c/h2\u003e\u003cp\u003eIn the control simulation, the upper-layer western subsection at 26\u0026deg;N transports approximately 37 Sv poleward, with nearly half (~\u0026thinsp;17 Sv) exported toward higher latitudes, while the remainder (~\u0026thinsp;20 Sv) presumably recirculates within the subtropical gyre (see GC in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Similarly, the upper-layer eastern subsection at 45\u0026deg;N carries about 22 Sv poleward. Most of this volume is exported northward (~\u0026thinsp;17 Sv), while the remainder (~\u0026thinsp;5 Sv) presumably recirculates within the subpolar gyre. In the lower layer, the subtropical gyre circulation is 8 Sv at 26\u0026deg;N, and the subpolar gyre circulation is 5 Sv at 45\u0026deg;N (GC in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The stronger gyre circulation at 26\u0026deg;N likely reflects the section\u0026rsquo;s alignment with the core of the subtropical gyre, whereas the 45\u0026deg;N section lies farther south of the subpolar gyre center, accounting for the weaker flow.\u003c/p\u003e\u003cp\u003eUnder the CO₂ experiment, the simulated upper-ocean circulation changes (GC in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) are weaker at 45\u0026deg;N compared to 26\u0026deg;N, which may be partially explained by the more favorable location of the latter. In contrast, the lower-layer circulation exhibits stronger changes at both locations (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). However, these changes are not straightforward to link directly to the horizontal circulation of the subpolar or subtropical gyre, as the transport also includes meridional flow associated with the bottom overturning cell (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eDespite the differences discussed above, the net volume transport in both sections remains comparable, suggesting a sustained latitudinal consistency of the flow with the AMOC (see net values in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Our simulated changes in the western subsection at 26\u0026deg;N align with this view, as they closely match the simulated changes in DWF. This is further backed up by CMIP6 projections under the SSP5-8.5 scenario, which suggest that the projected decline of the subtropical gyre\u0026rsquo;s western boundary southward transport is mainly driven by the thermohaline component rather than by wind-driven processes (Bryden et al. 2022; Tooth et al. \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In other words, changes in the lower branch are primarily buoyancy-driven and strongly influenced by remote surface buoyancy anomalies in the North Atlantic (Kostov et al. 2021; Asbj\u0026oslash;rnsen and \u0026Aring;rthun \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eMeridional transports (Sv\u0026thinsp;\u0026equiv;\u0026thinsp;10\u003csup\u003e6\u003c/sup\u003e m\u003csup\u003e3\u003c/sup\u003es\u003csup\u003e-1\u003c/sup\u003e) in the North Atlantic at 26\u0026deg;N and 45\u0026deg;N, divided horizontally into west and east contributions, and vertically divided into upper and lower layers. Net transport is defined as the difference between transport contributions of the west and east, with net export carried within one of these subsections according to the circulation pattern. The returning flow or gyre circulation (GC) is also shown. At 26\u0026deg;N, we have chosen 80\u0026deg;W as the division between subsections. At 45\u0026deg;N, we have chosen 45\u0026deg;W, so subsections are equivalent to the flow crossing the southern limits of Labrador (LAB) and Irminger (IRM) regions in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"7\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"7\" nameend=\"c7\" namest=\"c1\"\u003e\u003cp\u003eSubtropical gyre (26 \u0026deg;N)\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e\u003cp\u003eUpper\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c7\" namest=\"c5\"\u003e\u003cp\u003eLower\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eEast\u003c/p\u003e\u003cp\u003e(GC)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNet\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eWest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eEast\u003c/p\u003e\u003cp\u003e(GC)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNet\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eControl\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e37.10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-20.24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e16.86\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-26.01\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e7.56\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-18.45\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCO₂400\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e33.99\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-19.55\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e14.45\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-22.56\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e6.47\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-16.09\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCO₂550\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e31.92\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e-19.32\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e12.60\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e-19.23\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e4.93\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-14.29\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"7\" nameend=\"c7\" namest=\"c1\"\u003e\u003cp\u003eSubpolar gyre (45 \u0026deg;N)\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c4\" namest=\"c2\"\u003e\u003cp\u003eUpper\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c7\" namest=\"c5\"\u003e\u003cp\u003eLower\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eWest\u003c/p\u003e\u003cp\u003e(GC)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003eEast\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNet\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003eWest\u003c/p\u003e\u003cp\u003e(GC)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003eEast\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003eNet\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eControl\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-4.78\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e21.81\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e17.03\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.36\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-21.12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-18.76\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCO₂400\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-2.48\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e16.83\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e14.35\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e3.17\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-19.29\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-16.13\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCO₂550\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-2.12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e14.24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e12.12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e2.95\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e-16.90\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e-13.95\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"4 Discussion","content":"\u003cp\u003eThis study examines the relationship between North Atlantic deep convection, deep water formation, and the mean state of the AMOC under pre-industrial and elevated CO₂-fixed conditions. Using the DMV metric, we have quantified deep convection in the regions of LAB, GIN, and IRM. In the CO₂ experiments, deep convection across the North Atlantic shows a marked decline, consistent with the weakening of the AMOC. In that regard, to characterize deep water formation, we developed a method based on the assumption of effective subduction rates within the above-mentioned regions. Although we have not isolated the changes in DWF caused by the shoaling of overturning due to variations in convection and downwelling, our analysis shows no clear dependence on the critical depth selection. This suggests that changes in the depth structure of convergence are relatively negligible under warming conditions. Our method also enables the estimation of Arctic deep water export.\u003c/p\u003e\u003cp\u003eOur results suggest that continued anthropogenic warming may disrupt deep convection, reduce DWF, and weaken the AMOC. In Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e, we show the time series of DWF and the AMOC index, with the sum of all DWF sources closely matching the AMOC volume in both control and CO₂ experiments. The total DWF decrease is primarily due to changes in the LAB and GIN Seas, while the IRM contribution remains strong (Figs.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e and \u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e). This is in general agreement with previous observational data and reanalysis products that underscore the Irminger Sea's central role in regulating AMOC variability and its broader implications for climate dynamics in the North Atlantic (Petit et al. \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Chafik et al. \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe simulated strengthened Arctic deep water flow is likely driven by an increased surface density gradient in GIN, in response to the simulated sea ice retreat in the region (see blue line in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea, b and field in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eh, i), which may also explain the enhanced upper-layer export from the GIN Seas toward the Arctic Ocean through the Fram Strait and Barents Sea Opening. This augmented export of dense surface waters toward the Arctic, in conjunction with the emergence of new convective sites further north as the climate warms (Bretones et al. 2022; \u0026Aring;rthun et al. 2023, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), may be responsible for the simulated increase in returning deep water.\u003c/p\u003e\u003cp\u003eUnder both control and warming scenarios, Arctic deep water export increases while GIN-DWF decreases (see Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eb and Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e), which is consistent with the SSP2-4.5 future scenario simulation in Karami et al. (\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Furthermore, in agreement with previous studies (Heuz\u0026eacute;, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Bryden et al., 2022), our analysis reveals a statistically significant link between the increase in Arctic deep water export and the AMOC in the case of a high CO₂ concentration (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). At this CO₂ concentration, the Arctic export doubles compared to the control simulation, while the contribution of the DWF of GIN and LAB substantially weakens. Nevertheless, deep water export from the Arctic Ocean is a secondary stabilizing factor for the future state of the AMOC rather than a dominant factor, as previously stated by ocean-only modeling studies (Sevellec et al., 2017; Heuz\u0026eacute;, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Dima et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). This suggests that the influence of Arctic outflow on the AMOC is probably indirect. The relative importance of IRM, as underscored by our analysis, stems from the ocean-atmosphere feedback encompassed by the higher-resolution global coupled model, which is clearly absent in an ocean-only model.\u003c/p\u003e\u003cp\u003eOur method enables us to disentangle key processes by individually quantifying the most important components of the AMOC (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e). These components are often overlooked when relying solely on the traditional AMOC stream function representation. While the AMOC stream function is useful for representing meridional transport in both climate models and historical observations, it fails to identify the sources of deep water and the circulation processes due to its coarse representation of the meridional flow and the absence of a zonal component (Lozier et al. \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). In contrast, our method explicitly includes the zonal component, setting a linkage between these processes on a long-term scale (Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e). This enhanced view stresses how the dense water is redistributed within the North Atlantic convective regions, flows beneath the subpolar gyre, and eventually reaches the western boundary current in the subtropical gyre. Moreover, the relatively small differences between meridional poleward and equatorward transports further suggest limited compensating upwelling and interior mixing. The last reinforces the idea of strong meridional consistency in the lower branch of the AMOC on longer timescales (McCarthy and Caesar \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Frajka-Williams et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Buckley and Marshall 2015).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eOn climatic timescales, buoyancy forcing has been identified as the dominant driver of the AMOC, with a stronger effect than wind forcing (Petit et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Yeager et al., 2021; B\u0026ouml;ning et al., 2023). However, recent observational estimates suggest that convection plays a limited role in driving interannual AMOC variability (Zou et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Short-term fluctuations and the limited duration of observational records can obscure long-term variability, making sustained trends difficult to detect (McCarthy \u0026amp; Caesar, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Frajka-Williams et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Buckley \u0026amp; Marshall, 2015; Jackson et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Lozier et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2010\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAlthough deep water formation (DWF) is typically associated with regions of strong convection, it can also occur in areas where diffusive mixing within the water column is significant. Our method considers only mass convergence, thereby ostensibly neglecting the contribution of diffusive mixing. This may explain discrepancies observed between DWF and DMV estimates in the GIN Seas, where DMV indicates a shutdown of convection and a substantial reduction in deep water formation, while DWF does not indicate a complete shutdown (Figs.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eb and \u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e). Moreover, interannual variability in DWF may be masked by the relatively large spatial domains used to compute the metric. The mismatch between DMV and DWF suggests that our MLD criteria may not be fully capturing the processes involved, implying that deep water sinking likely occurs via diffusive processes driven by changes in SST, SSS and SSD.\u003c/p\u003e\u003cp\u003eAnother aspect requiring further investigation is the impact of ice melt on local buoyancy and the AMOC. The weakening of deep convection and DWF in the GIN and LAB may be partially attributed to surface freshening caused by sea ice melt in these regions (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003ea\u0026ndash;c). Therefore, future analyses should examine the dependence of these processes on water mass characteristics, explore shorter time scales, particularly across simulations with varying model resolutions.\u003c/p\u003e"},{"header":"5 Conclusions","content":"\u003cp\u003eIn this study, we have analyzed the impact of increased atmospheric CO₂ concentrations on the Atlantic Meridional Overturning Circulation (AMOC) and its dependence on North Atlantic Deep Water Formation (DWF) using the high-resolution version of the global coupled climate model EC-Earth3. Our high-resolution model provides enhancements in several key features of the North Atlantic, including more accurate simulations of regional circulation and improved vertical stratification, resulting from better representations of upper-ocean temperature and salinity (Karami et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAnalysed experiments include a pre-industrial control run and two fixed CO₂ concentration scenarios. Although the use of high-resolution limits our ability for running very long simulations, our CO2-level simulations are sufficiently long to reach a quasi-equilibrated state, and allow us to investigate long-term adjustments of the Earth systems, including changes in the DWF-AMOC relationship, to different CO2-levels.\u003c/p\u003e\u003cp\u003eWe have assessed DWF using a novel method based on horizontal volume convergence within North Atlantic regions of Labrador (LAB), Greenland-Iceland-Norway (GIN), and Irminger (IRM) Seas. The DWF weakens in both CO₂ experiments, with the LAB contribution showing the greatest reduction. In contrast, the IRM contribution remains strong, sustaining the AMOC. Similarly, the deep water export from the Arctic Ocean strengthened, which appears to act as an additional stabilizing factor for the future state of the AMOC.\u003c/p\u003e\u003cp\u003eThese changes correspond with variations in the meridional net volume transport carried at 26\u0026deg;N and at 45\u0026deg;N, suggesting strong latitudinal consistency of the meridional flow on the longer timescale. We show that the transport measured at 26\u003csup\u003eo\u003c/sup\u003eN matches both the DWF estimations and the transport across section 45\u003csup\u003eo\u003c/sup\u003eN. Despite this relationship not being fulfilled in terms of annual timescale variability, the match between these metrics suggests that observing DWF and export of deep water from IRM could suffice to monitor changes in AMOC strength, in both observational campaigns and under the framework of a large coupled model ensemble intercomparison.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cem\u003eFunding\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThis study was supported by the Horizon Europe project OptimESM \u0026ldquo;Optimal High Resolution Earth System Models for Exploring Future Climate Changes\u0026rdquo; under the European Union\u0026rsquo;s Horizon Europe research and innovation programme (grant agreement No 101081193), the FORMAS project FutureGS (Grant 2021-01374), the Swedish Research Council grant VR (2020-04791). The EC-Earth3 simulations and data handling were/was enabled by resources provided by the National Academic Infrastructure for Supercomputing in Sweden (NAISS), partially funded by the Swedish Research Council through grant agreement no. 2022-06725.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eCompeting Interests\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe authors have no relevant financial or non-financial interests to disclose.\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eAuthor Contributions\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eAll authors contributed to the study conception and design. Simulations were performed by Mehdi Pasha Karami and the analysis was performed by Ren\u0026eacute; Navarro-Labastida. The first draft of the manuscript was written by Ren\u0026eacute; Navarro-Labastida and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003eData availability\u003c/p\u003e\n\u003cp\u003eThe datasets generated and analysed during the current study are available under request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003e\u0026Aring;rthun M, Asbj\u0026oslash;rnsen H, Chafik L, Johnson H, V\u0026aring;ge K (2013) Future strengthening of the Nordic Seas overturning circulation, Nat Commun 14, 2065. https://doi.org/10.1038/s41467-023-37846-6\u003c/li\u003e\n\u003cli\u003e\u0026Aring;rthun M, Brakstad A, D\u0026ouml;rr J, Johnson HL, Mans C, Semper S, V\u0026aring;ge K (2025) Atlantification drives recent strengthening of the Arctic overturning circulation. Sci. Adv. 11. DOI:10.1126/sciadv.adu1794\u003c/li\u003e\n\u003cli\u003eAsbj\u0026oslash;rnsen H, and \u0026Aring;rthun M (2023) Deconstructing future AMOC decline at 26.5\u0026deg;N, Geophysical Research Letters, 50, e2023GL103515. https://doi.org/10.1029/2023GL103515\u003c/li\u003e\n\u003cli\u003eBaker JA, Bell MJ, Jackson LC, Renshaw R, Vallis GK, Watson AJ, Wood, RA (2023) Overturning pathways control AMOC weakening in CMIP6 models, Geophysical Research Letters, 50, e2023GL103381. https://doi.org/10.1029/2023GL103381\u003c/li\u003e\n\u003cli\u003eBaker JA, Bell MJ, Jackson LC, Vallis GK, Watson AJ, Wood RA (2025) Continued Atlantic overturning circulation even under climate extremes. Nature 638, 987\u0026ndash;994. https://doi.org/10.1038/s41586-024-08544-0\u003c/li\u003e\n\u003cli\u003eBellomo K, Meccia VL, D\u0026rsquo;Agostino R, Fabiano F, Larson SM, vonHardenberg J, Corti S (2023) Impacts of a weakened AMOC on precipitation over the Euro-Atlantic region in the EC-Earth3 climate model, Clim Dyn 61, 3397\u0026ndash;3416. https://doi.org/10.1007/s00382-023-06754-2\u003c/li\u003e\n\u003cli\u003eBryden H, Beunk J, Drijfhout S, Hazeleger W, Mecking J (2024) Comparing observed and modelled components of the Atlantic Meridional Overturning Circulation at 26\u0026deg;\u0026thinsp;N, Ocean Sci., 20, 589\u0026ndash;599. https://doi.org/10.5194/os-20-589-2024\u003c/li\u003e\n\u003cli\u003eBrodeau L, Koenigk T (2016) Extinction of the northern oceanic deep convection in an ensemble of climate model simulations of the 20th and 21st centuries. Clim Dyn 46, 2863\u0026ndash;2882. https://doi.org/10.1007/s00382-015-2736-5\u003c/li\u003e\n\u003cli\u003eBuckley MW, Marshall J (2016) Observations, inferences, and mechanisms of the Atlantic Meridional Overturning Circulation: A review, Rev. Geophys., 54. https://doi.org/10.1002/2015RG000493\u003c/li\u003e\n\u003cli\u003eChafik L, Holliday NP, Bacon S, Rossby T (2022) Irminger Sea is the center of action for subpolar AMOC variability. Geophysical Research Letters, 49, e2022GL099133. https://doi.org/10.1029/2022GL099133\u003c/li\u003e\n\u003cli\u003eCheng W, Chiang JCH, Zhang D (2013) Atlantic Meridional Overturning Circulation (AMOC) in CMIP5 Models: RCP and Historical Simulations. J. Climate, 26, 7187\u0026ndash;7197. https://doi.org/10.1175/JCLI-D-12-00496.1\u003c/li\u003e\n\u003cli\u003eDima M, Lohmann G, Ionita M, Knorr G, Scholz P (2022) AMOC modes linked with distinct North Atlantic deep water formation sites. Clim Dyn 59, 837\u0026ndash;849. https://doi.org/10.1007/s00382-022-06156-w\u003c/li\u003e\n\u003cli\u003eDocquier D, Grist JP, Roberts MJ, Roberts CD, Semmler T, Ponsoni L, Massonnet F, Sidorenko D, Sein DV, Iovino D, Bellucci A, Fichefet T (2019) Impact of model resolution on Arctic sea ice and North Atlantic Ocean heat transport. Clim Dyn 53, 4989\u0026ndash;5017. https://doi.org/10.1007/s00382-019-04840-y\u003c/li\u003e\n\u003cli\u003eDrijfhout S, Angevaare J, Mecking JV, van Westen RM, Rahmstorf S (2025) Shutdown of northern Atlantic overturning after 2100 following deep mixing collapse in CMIP6 projections. Environmental Research Letters. https://iopscience.iop.org/article/10.1088/1748-9326/adfa3b/meta\u003c/li\u003e\n\u003cli\u003eFrajka-Williams E, Foukal N, Danabasoglu G (2023) Should AMOC observations continue: how and why?, Phil. Trans. R. Soc. A 381: 20220195. https://doi.org/10.1098/rsta.2022.0195\u003c/li\u003e\n\u003cli\u003eFischer H, Meissner KJ, Mix AC, et al. (2018) Palaeoclimate constraints on the impact of 2 \u0026deg;C anthropogenic warming and beyond. Nature Geosci 11, 474\u0026ndash;485. https://doi.org/10.1038/s41561-018-0146-0\u003c/li\u003e\n\u003cli\u003eGalaasen EV, Ninnemann US, Irvali N, Kleiven HF, Rosenthal Y, Kissel C, Hodell DA (2014) Rapid reductions in North Atlantic Deep Water during the peak of the last interglacial period, Science. DOI: 10.1126/science.1248667\u003c/li\u003e\n\u003cli\u003eG\u0026eacute;rard J, Crucifix M (2024) Diagnosing the causes of AMOC slowdown in a coupled model: a cautionary tale, Earth Syst. Dynam., 15, 293\u0026ndash;306. https://doi.org/10.5194/esd-15-293-2024\u003c/li\u003e\n\u003cli\u003eGovin A, Braconnot P, Capron E, Cortijo E, Duplessy JC, Jansen E, Labeyrie L, Landais A, Marti O, Michel E, Mosquet E, Risebrobakken B, Swingedouw D, Waelbroeck C (2012) Persistent influence of ice sheet melting on high northern latitude climate during the early Last Interglacial, Clim. Past, 8, 483\u0026ndash;507. https://doi.org/10.5194/cp-8-483-2012\u003c/li\u003e\n\u003cli\u003eGrist JP, Josey SA, New AL, Roberts M, Koenigk T, Iovino D (2018) Increasing Atlantic Ocean heat transport in the latest generation coupled ocean-atmosphere models: The role of air-sea interaction. Journal of Geophysical Research: Oceans, 123, 8624\u0026ndash;8637. https://doi.org/10.1029/2018JC014387\u003c/li\u003e\n\u003cli\u003eGuarino MV, Sime LC, Diamond R, Ridley J, Schroeder D (2023) The coupled system response to 250 years of freshwater forcing: Last Interglacial CMIP6\u0026ndash;PMIP4 HadGEM3 simulations, Clim. Past, 19, 865\u0026ndash;881. https://doi.org/10.5194/cp-19-865-2023\u003c/li\u003e\n\u003cli\u003eGutjahr O, Putrasahan D, Lohmann K, Jungclaus JH, von Storch JS, Br\u0026uuml;ggemann N, Haak H, St\u0026ouml;ssel A (2019) Max Planck Institute Earth System Model (MPI-ESM1.2) for the High-Resolution Model Intercomparison Project (HighResMIP), Geosci. Model Dev., 12, 3241\u0026ndash;3281. https://doi.org/10.5194/gmd-12-3241-2019\u003c/li\u003e\n\u003cli\u003eHaarsma RJ, Roberts MJ, Vidale PL, Senior CA, Bellucci A, Bao Q, Chang P, Corti S, Fučkar NS, Guemas V, von Hardenberg J, Hazeleger W, Kodama C, Koenigk T, Leung LR, Lu J, Luo JJ, Mao J, Mizielinski MS, Mizuta R, Nobre P, Satoh M, Scoccimarro E, Semmler T, Small J, von Storch JS (2016) High Resolution Model Intercomparison Project (HighResMIP v1.0) for CMIP6, Geosci. Model Dev., 9, 4185\u0026ndash;4208. https://doi.org/10.5194/gmd-9-4185-2016\u003c/li\u003e\n\u003cli\u003eHersbach H, Bell B, Berrisford P, Biavati G, Hor\u0026aacute;nyi A, Mu\u0026ntilde;oz Sabater J, Nicolas J, Peubey C, Radu R, Rozum I, Schepers D, Simmons A, Soci C, Dee D, Th\u0026eacute;paut JN (2023) ERA5 hourly data on single levels from 1940 to present. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). DOI: 10.24381/cds.adbb2d47\u003c/li\u003e\n\u003cli\u003eHeuz\u0026eacute; C (2021) Antarctic Bottom Water and North Atlantic Deep Water in CMIP6 models, Ocean Sci., 17, 59\u0026ndash;90. https://doi.org/10.5194/os-17-59-2021\u003c/li\u003e\n\u003cli\u003eHirschi JJM, Barnier B, B\u0026ouml;ning C, Biastoch A, Blaker AT, Coward A, et al. (2020) The Atlantic meridional overturning circulation in high‐resolution models. Journal of Geophysical Research: Oceans, 125, e2019JC015522. https://doi.org/10.1029/2019JC015522\u003c/li\u003e\n\u003cli\u003eJackson LC, Biastoch A, Buckley MW, Desbruy\u0026egrave;res DG, Frajka-Williams E, Moat B, Robson J (2022) The evolution of the North Atlantic Meridional Overturning Circulation since 1980. Nat. Rev. Earth Environ. 3, 241\u0026ndash;254. (doi:10.1038/s43017-022-00263-2)\u003c/li\u003e\n\u003cli\u003eJackson LC, Kahana R, Graham T, Ringer MA, Woollings T, Mecking JV, Wood RA (2015) Global and European climate impacts of a slowdown of the AMOC in a high-resolution GCM. Clim Dyn 45, 3299\u0026ndash;3316. https://doi.org/10.1007/s00382-015-2540-2\u003c/li\u003e\n\u003cli\u003eJackson LC, Roberts MJ, Hewitt HT, Iovino D, Koenigk T, Meccia VL, Roberts CD, Ruprich-Robert Y, Wood RA (2020) Impact of ocean resolution and mean state on the rate of AMOC weakening. Clim Dyn 55, 1711\u0026ndash;1732. https://doi.org/10.1007/s00382-020-05345-9\u003c/li\u003e\n\u003cli\u003eJiang Z, Brierley C, Thornalley D, Sax S (2023) No changes in overall AMOC strength in interglacial PMIP4 time slices, Clim. Past, 19, 107\u0026ndash;121. https://doi.org/10.5194/cp-19-107-2023\u003c/li\u003e\n\u003cli\u003eKarami MP, Koenigk T, Wang S, Navarro-Labastida R, Kruschke T, Carreric A, Ortega P, Wyser K, Fuentes Franco R, de Boer, AM, Sicard M, Aldama Campino A (2025) Historical Climate and Future Projection in the North Atlantic and Arctic: Insights from EC-Earth3 High-Resolution Simulations, EGUsphere [preprint]. https://doi.org/10.5194/egusphere-2025-2653\u003c/li\u003e\n\u003cli\u003eKeil P, Mauritsen T, Jungclaus J, Hedemann C, Olonscheck D, Ghosh R (2020) Multiple drivers of the North Atlantic warming hole. Nat. Clim. Chang. 10, 667\u0026ndash;671. https://doi.org/10.1038/s41558-020-0819-8\u003c/li\u003e\n\u003cli\u003eKoenigk T, Fuentes-Franco R, Meccia VL, Gutjahr O, Jackson LC (2021) New, A.L., Ortega, P., Roberts, C.D., Roberts, M.J., Arsouze, T., Iovino, D., Moine, M.P., Sein, D.V.: Deep mixed ocean volume in the Labrador Sea in HighResMIP models. Clim Dyn 57, 1895\u0026ndash;1918. https://doi.org/10.1007/s00382-021-05785-x\u003c/li\u003e\n\u003cli\u003eKosto Y, Johnson HL, Marshall DP, Heimbach P, Forget G, Penny-Holliday N, Lozier MS, Li F, Pillar HR, and Smith T (2021) Distinct sources of interannual subtropical and subpolar Atlantic overturning variability. Nat. Geosci. 14, 491\u0026ndash;495. https://doi.org/10.1038/s41561-021-00759-4\u003c/li\u003e\n\u003cli\u003eLevang SJ, Schmitt RW (2021) What Causes the AMOC to Weaken in CMIP5? J. Climate, 33, 1535\u0026ndash;1545. https://doi.org/10.1175/JCLI-D-19-0547.1\u003c/li\u003e\n\u003cli\u003eLi F, Lozier MS, Bacon S, et al. (2021) Subpolar North Atlantic western boundary density anomalies and the Meridional Overturning Circulation. Nat Commun 12, 3002. https://doi.org/10.1038/s41467-021-23350-2\u003c/li\u003e\n\u003cli\u003eLozier M, et al. (2019) A sea change in our view of overturning in the subpolar North Atlantic.Science363,516-521. DOI:10.1126/science.aau6592\u003c/li\u003e\n\u003cli\u003eLozier M, Roussenov V, Reed M, et al. (2010) Opposing decadal changes for the North Atlantic meridional overturning circulation. Nature Geosci 3, 728\u0026ndash;734. https://doi.org/10.1038/ngeo947\u003c/li\u003e\n\u003cli\u003eMarotzke J (2000) Abrupt Climate Change and Thermohaline Circulation: Mechanisms and Predictability. Proceedings of the National Academy of Sciences of the United States of America, 97, 1347-1350. https://doi.org/10.1073/pnas.97.4.134\u003c/li\u003e\n\u003cli\u003eMarotzke J, Scot JR (1999) Convective Mixing and the Thermohaline Circulation. J. Phys. Oceanogr., 29, 2962\u0026ndash;2970. https://doi.org/10.1175/1520-0485(1999)029\u003c/li\u003e\n\u003cli\u003eMcCarthy GD, Caesar L (2023) Can we trust projections of AMOC weakening based on climate models that cannot reproduce the past? Phil. Trans. R. Soc. A 381: 20220193. https://doi.org/10.1098/rsta.2022.0193\u003c/li\u003e\n\u003cli\u003eMcPhaden M, Zhang D (2002) Slowdown of the meridional overturning circulation in the upper Pacific Ocean. Nature 415, 603\u0026ndash;608. doi: 10.1038/415603a\u003c/li\u003e\n\u003cli\u003ePetit T, Lozier MS, Josey SA, Cunningham SA (2020) Atlantic deep water formation occurs primarily in the Iceland Basin and Irminger Sea by local buoyancy forcing. Geophysical Research Letters, 47, e2020GL091028. https://doi.org/10.1029/2020GL091028\u003c/li\u003e\n\u003cli\u003ePetit T, Robson J, Ferreira D, Yeager S, Evans DG (2025) Coherence of the AMOC over the subpolar North Atlantic on interannual to multiannual time scales. Geophysical Research Letters, 52, e2025GL115171. https://doi.org/10.1029/2025GL115171\u003c/li\u003e\n\u003cli\u003eRahmstorf S (2024) Is the Atlantic overturning circulation approaching a tipping point? Oceanography. https://doi.org/10.5670/oceanog.2024.501\u003c/li\u003e\n\u003cli\u003eRayner NA, Parker DE, Horton EB, Folland CK, Alexander LV, Rowell DP, Kent EC, Kaplan A (2003) Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. JGRA, 108(D14), 4407. https://doi.org/10.1029/2002JD002670\u003c/li\u003e\n\u003cli\u003eRohling EJ, Hibbert FD, Grant KM, et al. (2019) Asynchronous Antarctic and Greenland ice-volume contributions to the last interglacial sea-level highstand. Nat Commun 10, 5040. https://doi.org/10.1038/s41467-019-12874-3\u003c/li\u003e\n\u003cli\u003eS\u0026eacute;vellec F, Fedorov A, Liu W (2017) Arctic sea-ice decline weakens the Atlantic Meridional Overturning Circulation. Nature Clim Change 7, 604\u0026ndash;610. https://doi.org/10.1038/nclimate3353\u003c/li\u003e\n\u003cli\u003eShan X, Sun S, Wu L, Wu L, Spall M (2024) Role of the Labrador Current in the Atlantic Meridional Overturning Circulation response to greenhouse warming. Nat Commun 15, 7361. https://doi.org/10.1038/s41467-024-51449-9\u003c/li\u003e\n\u003cli\u003eSicard M, de Boer AM, Coxall HK, Koenigk T, Karami MP, Jakobsson M, O\u0026rsquo;Regan M (2023) Similarities and differences in Arctic sea-ice loss during the solar-forced Last Interglacial warming (127 Kyr BP) and CO2-forced future warming. Geophysical Research Letters, 50, e2023GL104782. https://doi.org/10.1029/2023GL104782\u003c/li\u003e\n\u003cli\u003eTooth OJ, Foukal NP, Johns WE, Johnson HL, Wilson C (2024) Lagrangian decomposition of the Atlantic Ocean heat transport at 26.5\u0026deg;N. Geophysical Research Letters, 51, e2023GL107399. https://doi.org/10.1029/2023GL107399\u003c/li\u003e\n\u003cli\u003evan Westen RM, Kliphuis M, Dijkstra HA (2024) Physics-based early warning signal shows that AMOC is on tipping course, Science Advances, 10, eadk1189. https://doi.org/10.1126/sciadv.adk1189\u003c/li\u003e\n\u003cli\u003eWeijer W, Cheng W, Drijfhout SS, Federov AV, Hu A, Jackson LC, Liu W, McDonagh EL, Mecking JV, Zhang J (2019) Stability of the Atlantic Meridional Overturning Circulation: A review and synthesis. Journal of Geophysical Research: Oceans, 124, 5336\u0026ndash;5375. https://doi.org/10.1029/2019JC015083\u003c/li\u003e\n\u003cli\u003eZhang D, McPhaden M (2006) Decadal variability of the shallow Pacific meridional overturning circulation: Relation to tropical sea surface temperatures in observations and climate change models. Ocean Model. 15, 250\u0026ndash;273. doi: 10.1016/j.ocemod.2005.12.005\u003c/li\u003e\n\u003cli\u003eZou S, Lozier MS, Li F, Abernathey R, Jackson L (2020) Density-compensated overturning in the Labrador Sea. Nat. Geosci. 13, 121\u0026ndash;126. https://doi.org/10.1038/s41561-019-0517-1\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"AMOC weakening, deep mixed volume, deep water export","lastPublishedDoi":"10.21203/rs.3.rs-7542176/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7542176/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn this study, we analyze the impact of increased atmospheric CO₂ concentrations on the Atlantic Meridional Overturning Circulation (AMOC) and its dependence on North Atlantic Deep Water Formation (DWF), using the high-resolution version of the global coupled model EC-Earth3. The analyzed experiments include a pre-industrial control and two fixed CO₂ concentration scenarios, representative of stable warmer climate conditions and designed to investigate the long-term adjustments of the Earth system. Within this framework, we assess DWF using a novel method based on horizontal volume convergence in the main convective regions of the North Atlantic: the Labrador, Greenland-Iceland-Norway, and Irminger Seas. Our results suggest that under warmer climate conditions surface warming and freshening in the North Atlantic lead to disrupt deep convection, reduce DWF, and thereby weaken the AMOC. The reduction of DWF in the Labrador Sea emerges as the primary driver of AMOC weakening. In contrast, the Irminger Sea plays a central role in sustaining AMOC. The strengthened export of deep water from the Arctic Ocean also provides a stabilizing influence, though its effect is secondary to the sustained DWF in the Irminger Sea.\u003c/p\u003e","manuscriptTitle":"North Atlantic Ocean circulation and deep water formation under warmer climate conditions in EC-Earth3-HR","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-09-25 10:07:51","doi":"10.21203/rs.3.rs-7542176/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f5e7dbb6-9b99-45c2-adb4-83f22433cf7d","owner":[],"postedDate":"September 25th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-02-06T02:10:19+00:00","versionOfRecord":[],"versionCreatedAt":"2025-09-25 10:07:51","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7542176","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7542176","identity":"rs-7542176","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}