Ocean heat transport on icy moons shapes the ice shell topography, a primary observable of these moons. Two key processes control the heat transport: baroclinic instability driven by surface buoyancy contrasts and convective instability driven by heating from the core. However, global ocean simulations cannot accurately resolve convection under realistic icy moon conditions and instead often use Earth-based convective parameterizations, which capture only vertical convective mixing and cannot represent rotation-aligned slantwise convection on icy moons. We use high-resolution convection-resolving simulations to investigate ocean heat transport by slantwise convection in a parameter regime relevant to icy moons, isolated from baroclinic instability. Total heat transport follows the Coriolis-Inertial-Archimedean scaling with an added latitude dependence. The vertical transport increases with latitude, and the meridional transport is poleward. These results indicate that slantwise convection redistributes heat toward the poles, favoring a poleward-thinning ice shell, qualitatively consistent with Enceladus's observed ice thickness distribution.

Symmetric instability has broad applications in geophysical and planetary fluid dynamics. It plays a crucial role in the formation of mesoscale rainbands at mid‐latitudes on Earth, instability in the ocean's mixed layer, and slantwise convection on gas giants and icy moon oceans. Here, we apply linear instability analysis to an arbitrary zonally symmetric Boussinesq flow on a rotating spherical planet, with applicability to icy moon oceans. We divide the instabilities into three types: (a) gravitational instability, occurring when stratification is unstable along angular momentum surfaces, (b) inertial instability, occurring when angular momentum shear is unstable along buoyancy surfaces, and (c) a mixed symmetric instability, occurring when neither of the previous conditions are fulfilled, but the potential vorticity has the opposite sign to planetary rotation. We note that where is the Brunt–Väisälä frequency—a typical criterion used to trigger convective adjustment in global ocean models—is neither necessary nor sufficient for instability. Instead, , where is the stratification along the planetary rotation axis and is the local latitude, is always sufficient for instability and also necessary in the low Rossby number limit. In this limit, relevant for deep convection in icy moon oceans, the most unstable mode is slantwise convection parallel to the planetary rotation axis. This slantwise convection differs from the parameterized convection in existing general circulation models, whose convection schemes parameterize convection in the direction of gravity. Our results suggest that convection schemes in global ocean models must be revised before being applied to icy moon oceans.

An idealized ice-ocean model is used to study the time-dependent Atlantic meridional overturning circulation (AMOC) responses to a sudden uniform surface warming and/or an amplified evaporation minus precipitation (E - P) forcing. At transient time scales, the AMOC initially weakens in response to both types of forcing as a result of buoyancy gain in the North Atlantic, but the amplified E - P response is an order of magnitude smaller when its amplitude is chosen based on the Clausius-Clapeyron scaling, consistent with its weaker initial buoyancy flux anomaly. At equilibrium, the AMOC here weakens under warming, contrasting with previous idealized modeling studies. The difference is attributed to a larger role of North Atlantic warming (acting to weaken the AMOC) and a weaker role of reduced brine rejection around Antarctica (acting to deepen and strengthen the AMOC). When E - P forcing is amplified, the AMOC strengthens, qualitatively consistent with a previously proposed passive response that predicts an enhancement of the existing salinity pattern in equilibrium, although the amplification of the salinity contrast is significantly damped by a negative salt advection feedback. For a small-amplitude change in both temperature and E - P, the AMOC response can be approximated by the linear combination of the individual responses. However, large-amplitude warming and amplified E - P forcing can lead to a positive salt advection feedback that collapses the AMOC in our simulations. To understand why the sign of the salt advection feedback varies across different simulations, its multifaceted roles are further investigated using box model theories, and their relevance to comprehensive models is discussed.

During the last glacial period, the Northern Hemisphere climate underwent dramatic swings between relatively warm periods and cold periods}the Dansgaard-Oeschger oscillations. Here, we use recent progress in our theoretical understanding of the Atlantic meridional overturning circulation to develop a simple predictive model that relates variations in the overturning circulation to rapid changes in North Atlantic sea ice and the gradual recharge and discharge of the deep ocean temperature. The robustness of the model is tested against results from idealized general circulation model simulations, and exploration of its parameter space provides insights into the mechanisms dictating the overturning circulation's response to atmospheric forcing variations. The theoretical model predicts that global atmospheric temperature and salinity fluxes control the relative length of stadial versus interstadial conditions and reproduces the evolving characteristics of the d18O isotope ice core record over the last 100 kyr when forced only by the slowly changing global mean temperature. The findings indicate that the prominent climate variability observed in the Greenland ice cores is directly influenced by the gradual evolution of global temperatures and salinity fluxes. This variability can be attributed to a relatively simple physical mechanism that involves the interplay of fast positive sea ice and salt-advection feedbacks, along with a delayed negative deep-ocean-temperature feedback.

Observational data suggest that the ice shell on Enceladus is thicker at the equator than at the pole, indicating an equator-to-pole ice flow. If the ice shell is in an equilibrium state, the mass transport of the ice flow must be balanced by the freezing and melting of the ice shell, which in turn is modulated by the ocean heat transport. Here we use a numerical ocean model to study the ice–ocean interaction and ocean circulation on Enceladus with different salinities. We find that salinity fundamentally determines the ocean stratification. A stratified layer forms in the low-salinity ocean, affecting the ocean circulation and heat transport. However, in the absence of tidal heating in the ice shell, the ocean heat transport is found to always be toward lower latitudes, resulting in freezing at the poles, which cannot maintain the ice shell geometry against the equator-to-pole ice flow. The simulation results suggest that either the ice shell on Enceladus is not in an equilibrium state or tidal dissipation in the ice shell is important in maintaining the ice shell geometry. The simulations also suggest that a positive feedback between cross-equatorial ocean heat transport and ice melting results in spontaneous symmetry breaking between the two hemispheres. This feedback may play a role in the observed interhemispheric asymmetry in the ice shell.

Due to their limited resolution, numerical ocean models need to be interpreted as representing filtered or averaged equations. How to interpret models in terms of formally averaged equations, however, is not always clear, particularly in the case of hybrid or generalized vertical coordinate models. We derive the averaged hydrostatic Boussinesq equations in generalized vertical coordinates for an arbitrary thickness weighted-average. We then consider various special cases and discuss the extent to which the averaged equations are consistent with existing model formulations. As previously discussed, the momentum equations in existing depth-coordinate models are best interpreted as representing Eulerian averages (i.e., averages taken at fixed depth), while the tracer equations can be interpreted as either Eulerian or thickness-weighted isopycnal averages. Instead we find that no averaging is fully consistent with existing formulations of the parameterizations in semi-Lagrangian discretizations of generalized vertical coordinate ocean models. Perhaps the most natural interpretation of generalized vertical coordinate models is to assume that the average follows the model’s coordinate surfaces. However, the existing model formulations are generally not consistent with coordinate-following averages, which would require “coordinate-aware” parameterizations that can account for the changing nature of the eddy terms as the coordinate changes. Alternatively, the model variables can be interpreted as representing either Eulerian or (thickness-weighted) isopycnal averages, independent of the model coordinate that is being used for the numerical discretization. Existing parameterizations in generalized vertical coordinate models, however, are usually not fully consistent with either of these interpretations. We discuss what changes are needed to achieve consistency.

We quantify the volume transport and watermass transformation rates of the global overturning circulation using the Estimating the Circulation and Climate of the Ocean version 4 release 4 (ECCOv4r4) reanalysis product. The ECCO solution shows large rates of intercell exchange between the mid-depth and abyssal cells, consistent with other recent inferences. About 10 Sv of North Atlantic deep water enters the abyssal cell in the Southern Ocean and is balanced by a similar amount of apparrent diapycnal upwelling in the Indo-Pacific. However, much of the upwelling in ECCO's deep ocean is not associated with irreversible watermass transformations, as typically assumed in theoretical models. Instead, a dominant portion of the abyssal circulation in ECCO is associated with isopycnal volume tendencies, reflecting a deep ocean in a state of change and a circulation in which transient tendencies play a leading role in the watermass budget. These volume tendencies are particularly prominent in the Indo-Pacific, where ECCO depicts a cooling and densifying deep ocean with relatively little mixing-driven upwelling, in disagreement with recent observations of deep Indo-Pacific warming trends. Although abyssal ocean observations are insufficient to exclude the trends modeled by ECCO, we note that ECCO's parameterized diapycnal mixing in the abyssal ocean is much smaller than observational studies suggest and may lead to an under-representation of Antarctic Bottom Water consumption in the abyssal ocean. Whether or not ECCO's tendencies are realistic, they are a key part of its abyssal circulation and hence need to be taken into consideration when interpreting the ECCO solution. We analyze results taken from the Estimating the Circulation and Climate of the Ocean (ECCO) state estimate in order to investigate the internal structure and watermass budget of the global ocean's large-scale circulation. The ECCO solution supports the modern view of an interconnected global ocean with substantial exchange between the overturning circulation of the Atlantic and that of the Indo-Pacific via the Southern Ocean. However, our investigation also reveals that the density structure of much of the deep ocean in the ECCO product is in a state of change, and that these changes play a key role in the watermass budget of the circulation. These results reveal disagreement between the model's representation of the deep ocean and the prevailing theoretical depictions of the ocean's large-scale circulation, which generally assume that the circulation is in a steady state. Disagreement between ECCO's deep ocean mixing rates and independent estimates indicate that the trends in ECCO may be biased, but deep ocean observations are insufficient to conclusively infer the true trends. The meridional overturning circulation (MOC) in Estimating the Circulation and Climate of the Ocean version 4 release 4 (ECCOv4r4) exhibits substantial linkage between the mid-depth and abyssal cellsTransient isopycnal volume change plays a key role in the watermass budget of the MOC in ECCOECCO's transient interior state must be taken into account when interpreting its climatological state

We investigate the role of Southern Ocean topography and wind stress in the deep and abyssal ocean overturning and water mass composition using a suite of idealized global ocean circulation models. Specifically, we address how the presence of a meridional ridge in the vicinity of Drake Passage and the formation of an associated Southern Ocean gyre influence the water mass composition of the abyssal cell. Our experiments are carried out using a numerical representation of the global ocean circulation in an idealized two-basin geometry under varying wind stress and Drake Passage ridge height. In the presence of a low Drake Passage ridge, the overall strength of the meridional overturning circulation is primarily influenced by wind stress, with a topographically induced weakening of the middepth cell and concurrent strengthening of the abyssal cell occurring only after ridge height passes 2500 m. Passive tracer experiments show that a strengthening middepth cell leads to increased abyssal ventilation by North Atlantic water masses, as more North Atlantic Deep Water (NADW) enters the Southern Ocean and then spreads into the Indo-Pacific. We repeat our tracer experiments without restoring in the high-latitude Southern Ocean in order to identify the origin of water masses that circulate through the Southern Ocean before sinking into the abyss as Antarctic Bottom Water. Our results from these "exchange" tracer experiments show that an increasing ridge height in Drake Passage and the concurrent gyre spinup lead to substantially decreased NADW-origin waters in the abyssal ocean, as more surface waters from north of the Antarctic Circumpolar Current (ACC) are transferred into the Antarctic Bottom Water formation region.

Globally ice-covered oceans have been found on multiple moons in the solar system and may also have been a feature of Earth's past. However, relatively little is understood about the dynamics of these ice-covered oceans, which affect not only the physical environment but also any potential life and its detectability. A number of studies have simulated the circulation of icy-world oceans, but have come to seemingly widely different conclusions. To better understand and narrow down these diverging results, we discuss the energetic constraints for the circulation on ice-covered oceans, focusing in particular on Snowball Earth, Europa, and Enceladus. The energy input that can drive ocean circulation on ice-covered bodies can be associated with heat and salt fluxes at the boundaries as well as ocean tides and librations. We show that heating from the solid core balanced by heat loss through the ice sheet can drive an ocean circulation, but the resulting flows would be relatively weak and strongly affected by rotation. Salt fluxes associated with freezing and melting at the ice sheet boundary are unlikely to energetically drive a circulation, although they can shape the large-scale circulation when combined with turbulent mixing. Ocean tides and librations may provide an energy source for such turbulence, but the magnitude of this energy source remains highly uncertain for the icy moons, which poses a major obstacle to predicting the ocean dynamics of icy worlds and remains an important topic for future research.

Energy exchanges between large-scale ocean currents and mesoscale eddies play an important role in setting the large-scale ocean circulation but are not fully captured in models. To better understand and quantify the ocean energy cycle, we apply along-isopycnal spatial filtering to output from an isopycnal 1/32 degrees primitive equation model with idealized Atlantic and Southern Ocean geometry and topography. We diagnose the energy cycle in two frameworks: 1) a non-thickness-weighted framework, resulting in a Lorenz-like energy cycle, and 2) a thickness-weighted framework, resulting in the Bleck energy cycle. This paper shows that framework 2 is more useful for studying energy pathways when an isopycnal average is used. Next, we investigate the Bleck cycle as a function of filter scale. Baroclinic conversion generates mesoscale eddy kinetic energy over a wide range of scales and peaks near the deformation scale at high latitudes but below the deformation scale at low latitudes. Away from topography, an inverse cascade transfers kinetic energy from the mesoscales to larger scales. The upscale energy transfer peaks near the energy-containing scale at high latitudes but below the deformation scale at low latitudes. Regions downstream of topography are characterized by a downscale kinetic energy transfer, in which mesoscale eddies are generated through barotropic instability. The scale- and flow-dependent energy pathways diagnosed in this paper provide a basis for evaluating and developing scale- and flow-aware mesoscale eddy parameterizations. Significance StatementBlowing winds provide a major energy source for the large-scale ocean circulation. A substantial fraction of this energy is converted to smaller-scale eddies, which swirl through the ocean as sea cyclones. Ocean turbulence causes these eddies to transfer part of their energy back to the large-scale ocean currents. This ocean energy cycle is not fully simulated in numerical models, but it plays an important role in transporting heat, carbon, and nutrients throughout the world's oceans. The purpose of this study is to quantify the ocean energy cycle by using fine-scale idealized numerical simulations of the Atlantic and Southern Oceans. Our results provide a basis for how to include unrepresented energy exchanges in coarse global climate models.

Understanding surface temperature is important for habitability. Recent work on Mars has found that the dependence of surface temperature on elevation (surface lapse rate) converges to zero in the limit of a thin $\mathrm{CO_2}$ atmosphere. However, the mechanisms that control the surface lapse rate are still not fully understood. It remains unclear how the surface lapse rate depends on both greenhouse effect and surface pressure. Here, we use climate models to study when and why “mountaintops are cold”. We find the tropical surface lapse rate increases with the greenhouse effect and with surface pressure. The greenhouse effect dominates the surface lapse rate transition and is robust across latitudes. The pressure effect is important at low latitudes in moderately opaque ($\tau \sim 0.1$) atmospheres. A simple model provides insights into the mechanisms of the transition. Our results suggest that topographic cold-trapping may be important for the climate of arid planets.

Abstract There are two distinct parameterizations for the restratification effect of mesoscale eddies: the Greatbatch and Lamb (1990, GL90, https://journals.ametsoc.org/view/journals/phoc/20/10/1520-0485_1990_020_1634_opvmom_2_0_co_2.xml?tab_body=abstract-display) parameterization, which mixes horizontal momentum in the vertical, and the Gent and McWilliams (1990, GM90, https://journals.ametsoc.org/view/journals/phoc/20/1/1520-0485_1990_020_0150_imiocm_2_0_co_2.xml) parameterization, which flattens isopycnals adiabatically. Even though these two parameterizations are effectively equivalent under the assumption of quasi‐geostrophy, GL90 has been used much less than GM90, and exclusively in z‐coordinate models. In this paper, we compare the GL90 and GM90 parameterizations in an idealized isopycnal coordinate model, both from a theoretical and practical perspective. From a theoretical perspective, GL90 is more attractive than GM90 for isopycnal coordinate models because GL90 provides an interpretation that is fully consistent with thickness‐weighted isopycnal averaging, while GM90 cannot be entirely reconciled with any fully isopycnal averaging framework. From a practical perspective, the GL90 and GM90 parameterizations lead to extremely similar energy levels, flow and vertical structure, even though their energetic pathways are very different. The striking resemblance between the GL90 and GM90 simulations persists from non‐eddying through eddy‐permitting resolution. We conclude that GL90 is a promising alternative to GM90 for isopycnal coordinate models, where it is more consistent with theory, computationally more efficient, easier to implement, and numerically more stable. Assessing the applicability of GL90 in realistic global ocean simulations with hybrid coordinate schemes should be a priority for future work.

We describe an idealized primitive-equation model for studying mesoscale turbulence and leverage a hierarchy of grid resolutions to make eddy-resolving calculations on the finest grids more affordable. The model has intermediate complexity, incorporating basin-scale geometry with idealized Atlantic and Southern oceans and with non-uniform ocean depth to allow for mesoscale eddy interactions with topography. The model is perfectly adiabatic and spans the Equator and thus fills a gap between quasi-geostrophic models, which cannot span two hemispheres, and idealized general circulation models, which generally include diabatic processes and buoyancy forcing. We show that the model solution is approaching convergence in mean kinetic energy for the ocean mesoscale processes of interest and has a rich range of dynamics with circulation features that emerge only due to resolving mesoscale turbulence.

Building on previous work using single-basin models, we here explore the time-dependent response of the Atlantic meridional overturning circulation (AMOC) to a sudden global temperature change in a two-basin ocean-ice model. We find that the previously identified mechanisms remain qualitatively useful to explain the transient and the long-term time-mean responses of the AMOC in our simulations. Specifically, we find an initial weakening of the AMOC in response to warming (and vice versa for cooling), controlled by the mid-depth meridional temperature contrast across the Atlantic basin. The long-term mean response instead is controlled primarily by changes in the abyssal stratification within the basin. In contrast to previous studies we find that for small-amplitude surface temperature changes, the equilibrium AMOC is almost unchanged, as the abyssal stratification remains similar due to a substantial compensation between the effects of salinity and temperature changes. The temperature-driven stratification change results from the differential warming/cooling between North Atlantic Deep Water and Antarctic Bottom Water, while the salinity change is driven by changes in Antarctic sea ice formation. Another distinct feature of our simulations is the emergence of AMOC variability in the much colder and much warmer climates. We discuss how this variability is related to variations in deep-ocean heat content, surface salinity, and sea ice in the deep convective regions, both in the North Atlantic and in the Southern Ocean, and its potential relevance to past and future climates.

The ice shell and subsurface ocean on icy worlds are strongly coupled together—heat and salinity flux from the ice shell induced by the ice thickness gradient drives circulation in the ocean, and in turn, the heat transport by ocean circulation shapes the ice shell. Since measurements in the near future are likely to remain constrained to above the ice shell, understanding this ocean−ice interaction is crucial. Using an ocean box model and a series of experiments simulating the 2D ocean circulation, we find that large icy moons with strong gravity tend to have stronger ocean heat transport under the same ice shell topography. As a result, the equilibrium ice shell geometry is expected to be flatter on moons with larger size, and vice versa. This finding is broadly consistent with the observed ice shell geometry for Enceladus and Europa.

The ice shell and subsurface ocean on icy worlds are strongly coupled together – heat and salinity flux from the ice shell induced by the ice thickness gradient drives circulation in the ocean, and in turn, the heat transport by ocean circulation shapes the ice shell. Since measurements in the near future are likely to remain constrained to above the ice shell, understanding this ocean-ice interaction is crucial. Using an ocean box model and a series of experiments simulating the 2D ocean circulation, we find that large icy moons with strong gravity tend to have stronger ocean heat transport under the same ice-shell topography. As a result, the equilibrium ice shell geometry is expected to be flatter on moons with larger size, and vice versa. This finding is broadly consistent with the observed ice shell geometry for Enceladus and Europa.

The influence of atmospheric composition on the climates of present‐day and early Earth has been studied extensively, but the role of ocean composition has received less attention. We use the ROCKE‐3D ocean‐atmosphere general circulation model to investigate the response of Earth's present‐day and Archean climate system to low versus high ocean salinity. We find that saltier oceans yield warmer climates in large part due to changes in ocean dynamics. Increasing ocean salinity from 20 to 50 g/kg results in a 71% reduction in sea ice cover in our present‐day Earth scenario. This same salinity change also halves the pCO2 threshold at which Snowball glaciation occurs in our Archean scenarios. In combination with higher levels of greenhouse gases such as CO2 and CH4, a saltier ocean may allow for a warm Archean Earth with only seasonal ice at the poles despite receiving ∼20% less energy from the Sun.

Changes in the Southern Ocean (SO) surface wind stress influence both the meridional overturning circulation (MOC) and stratification not only in the SO but in the global oceans, which can take multiple millennia to fully equilibrate. We use a hierarchy of models to investigate the time-dependent response of the MOC and low-latitude pycnocline depth (which quantifies the stratification) to SO wind stress changes: a two-layer analytical theory, a multicolumn model (PyMOC), and an idealized general circulation model (GCM). We find that in both the GCM and PyMOC, the MOC has a multidecadal adjustment time scale while the pycnocline depth has a multicentennial time scale. The two-layer theory instead predicts the MOC and pycnocline depth to adjust on the same, multidecadal time scale. We argue that this discrepancy arises because the pycnocline depth depends on the bulk stratification, while the MOC amplitude is sensitive mostly to isopycnals within the overturning cell. We can reconcile the discrepancy by interpreting the "pycnocline depth" in the theory as the depth of a specific isopycnal near the maximum of the MOC. We also find that SO stationary eddies respond very quickly to a sudden wind stress change, compensating for most of the change in the Ekman-driven MOC. This effect is missing in the theory, where the eddy-induced MOC only follows the adjustment of the pycnocline depth. Our results emphasize the importance of depth dependence in the oceans' transient response to changes in surface boundary conditions, and the distinct role played by stationary eddies in the SO. Significance StatementOur work resolves the question of why previous theories predict the ocean density structure to adjust to a change in the winds over the Southern Ocean within centuries, while climate models indicate that this adjustment takes thousands of years. The question is important because it is related to our understanding of how the ocean responds to potential climate change scenarios. Our results emphasize the importance of depth dependence in the density response (i.e., the upper ocean adjusts faster than the deep ocean), suggesting that future theoretical advancement should be made with careful considerations of the ocean's vertical structure. Our results also highlight the role of stationary meanders in the Southern Ocean's Antarctic Circumpolar Current, whose influence has not been included in the existing theories.

1University of Maryland Center for Environmental Science, Chesapeake Biological Laboratory, Solomons, MD, USA. 2Department of Geological and Atmospheric Sciences, Iowa State University, Ames, IA, USA. 3Geography Department, Durham University, Durham, UK. 4Centre for Geography and Environmental Sciences, University of Exeter, Penryn, UK. 5Climate and Global Dynamics Laboratory, National Center for Atmospheric Research, Boulder, CO, USA. 6Woods Hole Oceanographic Institution, Falmouth, MA, USA. 7Cooperative Institute for Marine and Atmospheric Studies, University of Miami, Miami, FL, USA. 8Atlantic Oceanographic and Meteorological Laboratory, National Oceanic and Atmospheric Administration, Miami, FL, USA. 9Department of the Geophysical Sciences, The University of Chicago, Chicago, IL, USA. 10Oceanography Department, Atmospheric and Environmental Research, Inc., Texas, TX, USA. 11US Geological Survey, St Petersburg Coastal and Marine Science Center, St Petersburg, FL, USA. 12Barcelona Supercomputing Center, Barcelona, Spain. 13Graduate School of Oceanography, University of Rhode Island, Kingston, RI, USA. 14University Corporation of Atmospheric Research, Boulder, CO, USA. ✉e-mail: kilbourn@umces.edu Deep oceanic overturning circulation in the Atlantic (Atlantic Meridional Overturning Circulation (AMOC)) is projected to decrease in the future in response to anthropogenic warming. Caesar et al.1 argue that an AMOC slowdown started in the nineteenth century and intensified during the mid-twentieth century. Although the argument and selected evidence proposed have some merits, we find that their conclusions might be different if a more complete array of data available in the North Atlantic region is considered. We argue that the strength of AMOC over recent centuries is still poorly constrained and the expected slowdown may not have started yet. Recently, Moffa-Sánchez et al.2 compiled a comprehensive set of palaeoclimate proxy data from the North Atlantic and Arctic regions using objective criteria to identify high-quality datasets of ocean conditions that span the past two millennia (Fig. 1). Although no direct (singular) proxy for AMOC exists, the palaeoceanographic proxy data compiled by Moffa-Sánchez et al.2 highlight the spatial and temporal complexities of the ocean state in modern times and the recent past. When all the available proxy records potentially related to AMOC variability and twentieth century observational datasets are considered, the time history of the AMOC system becomes less certain. In contrast, selecting only a subset of proxy records that share similar trends, as performed by Caesar et al.1, provides an incomplete perspective on AMOC changes through time. Increased data availability in recent decades has enabled a shift in the fields of palaeoceanography and palaeoclimatology towards more objective and transparent data selections in studies aimed at quantitatively reconstructing past variability. Such screening methods tend to minimize the impact of spurious or less reliable records on analyses, and work to enhance the common signal in proxy records. Additionally, analysing networks of suitable and carefully selected data enables robust uncertainty estimates on the resulting reconstructions, which is essential to provide confidence in the results and the ability to compare information across disciplines. Key to such work is identifying robust criteria and weighting schemes that objectively identify and utilize the most reliable data. Caesar et al.1 use a variety of proxy records in their analysis, but do not identify the reasoning or criteria for selecting those records over many others that are probably related to aspects of AMOC dynamics (see the recent review2). Objective and inclusive data selection standards are especially important when addressing AMOC, which is a system composed of many different components that can behave differently at different latitudes, depths and timescales3, and looking at any singular index of AMOC inherently oversimplifies the system. The complex signals in the available AMOC-related proxy variables over recent centuries support this notion2, although many of these studies were not considered by Caesar et al.1. In addition to the need for objective standards, we argue that most of the records compiled in the Caesar et al. article1 have substantial caveats that were not discussed. Reconstructing the strength of AMOC more than a few decades ago relies on palaeoclimate and palaeoceanographic proxies because direct measurements are unavailable. Some proxies are more directly related to components of AMOC variability than others, and some sites are better situated to record specific oceanographic and atmospheric processes than others. The limited scope of data utilized combined with the inherent uncertainties in the proxies and conflicting evidence from other sources leaves the question open as to whether the available evidence supports the conclusion that AMOC is currently undergoing an unprecedented shift and/or weakening. Key information and rationale about the records included are lacking in Caesar et al.1. For example, the Rahmstorf et al.4 AMOC reconstruction used by Caesar et al.1 is based on the subpolar North Atlantic temperature minus the Northern Hemisphere mean temperature, each constructed from tree ring and ice core records, and a scaling coefficient derived from one climate model. These data are land-based estimates influenced by atmospheric conditions, not necessarily robust indicators of marine temperatures, and the resulting index is strongly impacted by the global warming signal5. Furthermore, subpolar gyre sea surface temperatures are an unreliable indicator of AMOC variability5,6 because these temperatures can have multiple drivers and the spatial AMOC/sea surface temperature fingerprints used for such reconstructions are temporally non-stationary2,5. Variables related to marine biological processes used as evidence by Caesar et al.1 are potentially problematic as Atlantic circulation change still uncertain