Search

Variations of rare earth element (REE) concentrations in metamorphic garnet are an important source of information of geodynamic and geochemical processes in the deeper Earth. In order to extract this information, the thermodynamic equilibrium and kinetic contributions of the REE uptake in garnet must be distinguished and quantified. Utilizing high-resolution trace element and μ-Raman mapping together with combined thermodynamic–geochemical–diffusion models, we demonstrate that the equilibrium and kinetic aspects of the REE uptake in metamorphic garnet can be discriminated by interpreting 2D trace element mapping in a single sample. The heavy (H) REE (Tb to Lu) zoning in the investigated garnet from a high-pressure blueschist comprises an inner part with an overall decrease from core to inner rim, followed by a concentric zone of HREE enrichment and a drastic HREE decrease towards the outermost rim. The central peak in the garnet core decreases in intensity with decreasing atomic number of the REE. The broad overall shape of this pattern resembles those often observed in metamorphic garnet from different rock types and tectonic settings. Superimposed on this trend is a concentric pattern of minor recurring fluctuations in the HREE concentrations with at least six regularly spaced sets of peaks and troughs along the entire garnet radius. Comparison of the observed inclusion suite, the trace element maps and thermodynamic–geochemical models show that the inner part with decreasing HREE concentrations results from fractional garnet growth in an unchanged mineral assemblage, whereas the REE enrichment zone is caused by the breakdown of titanite. We suggest that the width of the central peak is controlled by the bulk permeability of the interconnected transport matrix and the fraction of matrix minerals that the garnet equilibrates with. The superimposed REE fluctuations result from changing element transport properties of the host rock and mark recurring changes from equilibrium REE uptake to transport-limited REE uptake in garnet. Such fluctuating element transport properties can be best explained by pulse-like fluid fluxes that rhythmically change the interconnectivity of the intercrystalline transport matrix. Increasing numbers of published spatially highly resolved REE analyses show that such trace element fluctuations are common in metamorphic garnet indicating that recurring changes in rock permeabilities due to pulsed fluid fluxes are a common phenomenon during metamorphism.

Southern Ocean bathymetry constrains the path of the Antarctic Circumpolar Current (ACC), but the bathymetric influence on the coupled ocean–atmosphere system is poorly understood. Here, we investigate this impact by respectively flattening large topographic barriers around the Kerguelen Plateau, Campbell Plateau, Mid-Atlantic Ridge, and Drake Passage in four simulations in a coupled climate model. The barriers impact both the wind and buoyancy forcing of the ACC transport, which increases by between 4% and 14% when barriers are removed individually and by 56% when all barriers are removed simultaneously. The removal of Kerguelen Plateau bathymetry increases convection south of the plateau and the removal of Drake Passage bathymetry reduces convection upstream in the Ross Sea. When the barriers are removed, zonal flattening of the currents leads to sea surface temperature (SST) anomalies that strongly correlate to precipitation anomalies, with correlation coefficients ranging between r = 0.92 and r = 0.97 in the four experiments. The SST anomalies correlate to the surface winds too in some locations. However, they also generate circumpolar waves of sea level pressure (SLP) anomalies, which induce remote wind speed changes that are unconnected to the underlying SST field. The meridional variability in the wind stress curl contours over the Mid-Atlantic Ridge, the Kerguelen Plateau, and the Campbell Plateau disappears when these barriers are removed, confirming the impact of bathymetry on surface winds. However, bathymetry-induced wind changes are too small to affect the overall wave-3 asymmetry in the Southern Hemisphere westerlies. Removal of Southern Hemisphere orography is also inconsequential to the wave-3 pattern. Significance Statement Several studies in the past have pointed to the controlling effect of bathymetry on currents in the Southern Ocean circulation, but a clear idea of the importance of the major topographic barriers in the Southern Ocean is lacking. By removing these barriers systematically in a coupled climate model, we can evaluate their impact on several important components of the climate system, such as the Antarctic Circumpolar Current (ACC) pathways and strength, Antarctic Bottom Water formation, sea surface temperature, overlying winds, air–sea fluxes, and even precipitation. This helps in our understanding of what controls the pathways of the ACC and how much it matters for climate.

In climate research we often want to focus on a specific region and the most prominent processes affecting it, but how exactly do we select the borders of that region? We also often need to use long-term in situ observations to represent a larger area, but which area exactly are they representative for? In ocean sciences we usually consider basins as separate regions or even simpler, just select a rectangle of the ocean, but that does not always correspond to the real, physically relevant borders. As alternative, we use an unsupervised classification model, Gaussian Mixture Model (GMM), to separate the northwestern European seas into regions based on the sea level variability observed by altimetry satellites.After performing a principal component (PC) analysis on the 24 years of monthly sea level data, we use the stacked PC maps as input for the GMM. We used the Bayesian Information Criterion to determine into how many regions our area should be split because GMM requires the number of classes to be selected a priori. Depending on the number of PCs used, the optimal number of classes was between 12 and 18, more PCs typically allowing the separation into more regions. Due to the complexity of the data and the dependence of the results on the starting randomly chosen weights, the classification can differ to a degree with every new run of the model, even if we use the exact same data and parameters. To tackle that, instead of using one model, we use an ensemble of models and then determine which class does each grid point belong to by soft voting, i.e., each of the models provides a probability that the point belongs to a particular class and the class with the maximal sum of probabilities wins. As a result, we obtain both the classification and the likelihood of the model belonging to that class.Despite not using the coordinates of the data points in the model at all, the obtained classes are clearly location dependent, with grid points belonging to the same class always being close to each other. While many classes are defined by bathymetry changes, e.g., the continental shelf break and slope, sometimes other factors come into play, such as for the split of the Norwegian coast into two classes or for the division in the Barents Sea, which is probably based on the circulation. The North Sea is also split into three distinct regions, possibly based on sea level changes caused by dominant wind patterns.This method can be applied to almost any atmospheric or oceanic variable and used for larger or smaller areas. It is quick and practical, allowing us to delimit the area based on the information we cannot always clearly see from the data, which can facilitate better selection of the regions that need further research.

Metamorphic garnet can be the ultimate source of information about geodynamic and geochemical processes in the Earth’s crust. Compositionally zoned garnet porphyroblasts preserve records of the host rock’s reaction path, its element transport properties, the fluid-rock interaction during metamorphism as well as absolute ages and rates of these processes. Especially variations of rare earth element (REE) concentrations in garnet are reflecting many of these geodynamic and geochemical processes. In order to extract this information, the thermodynamic equilibrium and kinetic contributions of the REE uptake in garnet must be distinguished and quantified, for which high resolution major- and trace element mappings together with numerical growth simulations are an indispensable tool. Utilizing high resolution trace element and µ-Raman mapping together with combined thermodynamic-geochemical-diffusion models we discriminate the equilibrium and kinetic aspects of the heavy (H) REE uptake in a garnet porphyroblast from a high-pressure/low temperature blueschist from the Dominican Republic. Like many metamorphic garnets from different rock types and tectonic settings, the HREE zoning in the investigated garnet comprises an inner, bell-shaped part with a pronounced central peak and an overall decrease of all HREE from core to inner rim. The central peak in the garnet core decreases in intensity with decreasing atomic number of the REE. This trend is followed by a concentric zone of HREE enrichment and a drastic HREE decrease towards the outermost rim. Superimposed on this common trend is a concentric pattern of minor recurring fluctuations in the HREE concentrations with regularly spaced sets of peaks and troughs along the entire garnet radius.Comparison of the trace element mappings and thermodynamic-geochemical models show that the inner, bell-shaped part results from fractional garnet growth in an unchanged mineral assemblage. The model results further show that the width of the central peak is controlled by the bulk permeability of the interconnected transport matrix and the fraction of matrix minerals that the garnet equilibrates with. The correlation of high resolution µ-Raman mappings of the inclusion suite and the trace element mappings indicate that the REE enrichment zone is caused by HREE redistribution during the titanite-rutile transition.The superimposed REE fluctuations result from changing element transport properties of the host rock and mark recurring changes from equilibrium REE uptake to transport-limited REE uptake in garnet. Such fluctuating element transport properties can be best explained by pulse-like fluid fluxes that rhythmically change the interconnectivity of the intercrystalline transport matrix. Increasing numbers of published spatially highly resolved REE analyses show that such trace element fluctuations are common in metamorphic garnet indicating that recurring changes in rock permeabilities due to pulsed fluid fluxes are a common phenomenon during metamorphism.

The Beaufort Gyre is a major reservoir of freshwater in the Arctic Ocean and plays an important role in shaping the sea ice distribution. The variations in Beaufort Gyre strength, and its capacity to accumulate or release freshwater, are primarily determined by the stationary Beaufort High pressure system. Multiple studies have suggested a potential future weakening of the Beaufort High, thus driving a projected weakening of the Beaufort Gyre. Yet, in contrast, the ongoing reduction and thinning of the Arctic sea ice has been shown to strengthen the Gyre in recent years – a trend likely to continue into the future. We investigate these potentially competing effects of projected sea ice decline and changing atmospheric patterns on the Beaufort Gyre location, extent, and strength. To this end, we employ a novel storyline approach to identify plausible fates of the Beaufort Gyre in a warming climate, generated from the suite of available CMIP6 models. The different storylines are determined based on the strong or weak modelled response to these two local drivers - Arctic sea ice decline and Beaufort High pressure system weakening - on changes in Beaufort Gyre strength. While the CMIP6 multi-model mean response in the location, extent and strength of the Beaufort Gyre does not exhibit any distinct future changes under emission scenario SSP585, the storylines however reveal contrasting futures. For example, in a future with strong sea ice decline but only a weak decrease of Beaufort High pressure, the storyline indicates a spin-up of the Beaufort Gyre alongside a strong Arctic-wide surface salinity reduction. In contrast, a future with weak sea ice decline but a strong Beaufort High pressure decrease is characterised by a slow-down of the Beaufort Gyre and more regionally confined surface salinity changes. The storyline approach thus highlights that we urgently need to better constrain our model projections in order to reliably predict changes in upper Arctic ocean circulation and freshwater distribution, which play a crucial role for the future Arctic climate and response of both marine and terrestrial ecosystems.

Understanding the forcing of regional sea level variability is crucial as many people all over the world live along the coasts and are endangered by the sea level rise. The adding of fresh water into the oceans due to melting of the Earth’s land ice together with thermosteric changes has led to a rise of the global mean sea level (GMSL) with an accelerating rate during the twentieth century, and has now reached a mean rate of 3.7 mm per year according to IPCCs latest report. However, this change varies spatially and the dynamics behind what forces sea level variability on a regional to local scale is still less known, thus making it hard for decision makers to mitigate and adapt with appropriate strategies.Here we present a novel approach using machine learning (ML) to identify the dynamics and determine the most prominent drivers forcing coastal sea level variability. We use a recurrent neural network called Long Short-Term Memory (LSTM) with the advantage of learning data in sequences and thus capable of storing some memory from previous timesteps, which is beneficial when dealing with time series. To train the model we use hourly ERA5 10-m wind, mean sea level pressure (MSLP), sea surface temperature (SST), evaporation and precipitation data between 2009-2017 in the North Sea region. To reduce the dimensionality of the data but still preserve maximal information we conduct principal component analysis (PCA) after removing the climatology which are calculated by hourly means over the years. Depending on the explained variance of the PCs for each driver, 2-4 PCs are chosen and cross-correlated to eliminate collinearity, which could affect the model results. Before being used in the ML model the final preprocessed data are normalized by min-max scaling to optimize the learning. The target data in the model are hourly in-situ sea level observations from West-Terschelling in the Netherlands. Using in-situ observations compared to altimeter data enhances the ability of making good predictions in coastal zones as altimeter data has a tendency to degrade along the coasts. The sea level time series is preprocessed by tidal removal and de-seasoned by subtracting the hourly means. To determine which drivers are most prominent for the sea surface variability in our location, we mute one driver at a time in the training of the network and evaluate the eventual improvement or deterioration of the predictions.Our results show that the zonal wind is the most prominent forcing of sea level variability in our location, followed by meridional wind and MSLP. While the SST greatly affects the GMSL, SST seems to have little to no effect on local sea level variability compared to other drivers. This approach shows great potential and can easily be applied to any coastal zone and is thus very useful for a broad body of decision makers all over the world. Identifying the cause of local sea level variability will also enable the ability of producing better models for future predictions, which is of great importance and interest.

Polynyas are recurrent areas of open water or thin ice within the ice pack, which alter the local heat and moisture exchange and high-latitude atmosphere-ocean circulation interannual variability. They are differentiated as coastal (latent heat) or open-ocean (sensible heat) polynya according to their forming location. Especially, coastal polynyas are critical sources of dense water and the formation of Antarctic Bottom Water (AABW) following the brine enrichment of surface waters during sea-ice formation, and easily influenced by the local atmosphere conditions. However, few studies have examined the atmospheric response of open-ocean polynyas on the coastal polynyas given the fact that open-ocean polynyas have capability to re-adjust mesoscale atmosphere circulation. To better understand the surrounding impact of large open-ocean polynya events, output from CMIP6 historical experiment synoptic scale EC-Earth3 is adopted. Our results show an increasing coastal polynya frequency and extent accompanying with more active open-ocean polynya years in the Weddell Sea. The results are explained by near-surface wind speed differences in the coastal regions, which are found statistically significant between more and less active open-ocean polynya years. Furthermore, those intensifications of winds are found in days where easterly-dominated winds north-westerly to north-easterly and easterly to south-easterly cross the open-ocean polynya. Increased near-surface air temperatures as well as a deepening in sea level pressure are also observed during the years with more active open-ocean polynya events. The findings contribute to a better understanding of coastal polynya opening processes, as well as how we might expect to see the different type of polynya interact by their influence and dependence on surrounding atmospheric conditions.

The Arctic Ocean is strongly stratified by salinity in the uppermost layers. This stratification is a key attribute of the region as it acts as an effective barrier for the vertical exchanges of Atlantic Water heat, nutrients, and CO2 between intermediate depths and the surface of the Eurasian and Amerasian basins (EB and AB). Observations show that from 1970 to 2017, the stratification in the AB has strengthened, whereas, in parts of the EB, the stratification has weakened. The strengthening in the AB is linked to freshening and deepening of the halocline. In the EB, the weakened stratification is associated with salinification and shoaling of the halocline (Atlantification). Simulations from a suite of CMIP6 models project that, under a strong greenhouse-gas forcing scenario (ssp585), the overall surface freshening and warming continue in both basins, but there is a divergence in hydrographic trends in certain regions. Within the AB, there is agreement among the models that the upper layers will become more stratified. However, within the EB, models diverge regarding future stratification. This is due to different balances between trends at the surface and trends at depth, related to Fram Strait fluxes. The divergence affects projections of future state of Arctic sea ice, as models with the strongest Atlantification project the strongest decline in sea ice volume in the EB. From these simulations, one could conclude that Atlantificaton will not spread eastward into the AB; however, we need to improve models to simulate tendencies in a more delicately stratified EB correctly.

Knowing when sea ice will open is crucial, notably for scientific deployments. This was particularly obvious when the Weddell Polynya, a large opening in the winter Southern Ocean sea ice, unexpectedly re-appeared in 2016. As no precursor had been detected, observations were limited to chance autonomous sensors, and the exact cause of the opening could not be determined accurately. We investigate here whether the signature of the vertical ocean motions or that of the leads, which ultimately re-open the polynya, are detectable in spaceborne infrared temperature before the polynya opens. From the full historical sea ice concentration record, we find 30 polynyas starting from 1980. Then, using the full time series of the spaceborne infrared Advanced Very High Resolution Radiometer, we determine that these events can be detected in the 2 weeks before the polynya opens as a reduction in the variance of the data. For the three commonly used infrared brightness temperature bands, the 15 d sum and 15 d standard deviation of their area median and maximum are systematically lower than the climatology when a polynya will open. Moreover, by comparing the infrared brightness temperature to atmospheric reanalysis, hydrographic mooring data, and autonomous profilers, we find that temporal oscillations in one band and the decrease in the difference between bands may be used as proxies for upwelling of warm water and presence of leads, respectively, albeit with caution. Therefore, although infrared data are strongly limited by their horizontal resolution and sensitivity to clouds, they could be used for studying ocean or atmosphere preconditioning of polynyas in the historical record.

Arctic sea ice loss has become a symbol of ongoing climate change, yet climate models still struggle to reproduce it accurately, let alone predict it. A reason for this is the increasingly clear role of the ocean, especially that of the “Atlantic layer,” on sea ice processes. We here quantify biases in that Atlantic layer and the Arctic Ocean deeper layers in 14 representative models that participated in phase 6 of the Climate Model Intercomparison Project. Compared to observational climatologies and hydrographic profiles, the modeled Atlantic layer core is on average too cold by −0.4°C and too deep by 400 m in the Nansen Basin. The Atlantic layer is too thick, extending to the seafloor in some models. Deep and bottom waters are in contrast too warm by 1.1° and 1.2°C. Furthermore, the modeled properties hardly change throughout the Arctic. We attribute these biases to an inaccurate representation of shelf processes: only three models seem to produce dense water overflows, at too few locations, and these do not sink deep enough. No model compensates with open ocean deep convection. Therefore, the properties are set by the inaccurate volume fluxes through Fram Strait, biased low by up to 6 Sv (1 Sv ≡ 106 m3 s−1), but coupled to a too-warm Fram Strait, resulting in a somewhat accurate heat inflow. These fluxes are related to biases in the Nordic seas, themselves previously attributed to inaccurate sea ice extent and atmospheric modes of variability, thus highlighting the need for overall improvements in the different model components and their coupling. Significance Statement Coupled climate models are routinely used for climate change projection and adaptation, but they are only as good as the data used to create them. And in the deep Arctic, those data are scarce. We determine how biased 14 of the most recent models are regarding the deep Arctic Ocean and the Arctic’s only deep gateway, Fram Strait (between Greenland and Svalbard). These models are very biased: too cold where they should be warm, too warm where they should be cold, not stratified enough, not in contact with the surface as they should, moving the wrong way around the Arctic, etc. Some problems are induced by biases in regions outside of the Arctic and/or from the sea ice models.

Deep and bottom water formation are crucial components of the global ocean circulation, yet they were poorly represented in the previous generation of climate models. We here quantify biases in Antarctic Bottom Water (AABW) and North Atlantic Deep Water (NADW) formation, properties, transport, and global extent in 35 climate models that participated in the latest Climate Model Intercomparison Project (CMIP6). Several CMIP6 models are correctly forming AABW via shelf processes, but 28 models in the Southern Ocean and all 35 models in the North Atlantic form deep and bottom water via open-ocean deep convection too deeply, too often, and/or over too large an area. Models that convect the least form the most accurate AABW but the least accurate NADW. The four CESM2 models with their overflow parameterisation are among the most accurate models. In the Atlantic, the colder the AABW, the stronger the abyssal overturning at 30∘ S, and the further north the AABW layer extends. The saltier the NADW, the stronger the Atlantic Meridional Overturning Circulation (AMOC), and the further south the NADW layer extends. In the Indian and Pacific oceans in contrast, the fresher models are the ones which extend the furthest regardless of the strength of their abyssal overturning, most likely because they are also the models with the weakest fronts in the Antarctic Circumpolar Current. There are clear improvements since CMIP5: several CMIP6 models correctly represent or parameterise Antarctic shelf processes, fewer models exhibit Southern Ocean deep convection, more models convect at the right location in the Labrador Sea, bottom density biases are reduced, and abyssal overturning is more realistic. However, more improvements are required, e.g. by generalising the use of overflow parameterisations or by coupling to interactive ice sheet models, before deep and bottom water formation, and hence heat and carbon storage, are represented accurately.

Automatic and visual sea ice classification of SAR imagery is impeded by the incidence angle dependence of backscatter intensities. Knowledge of the angular dependence of different ice types is therefore necessary to account for this effect. While consistent estimates exist for HH polarization for different ice types, they are lacking HV polarization data, especially for multiyear sea ice. Here we investigate the incidence angle dependence of smooth and rough/deformed first-year and multiyear ice of different ages for wintertime dual-polarization Sentinel-1 C-band SAR imagery in the Beaufort Sea. Assuming a linear relationship, this dependence is determined using the difference in incidence angle and backscatter intensities from ascending and descending images of the same area. At cross-polarization rough/deformed first-year sea ice shows the strongest angular dependence with $-\text{0.11}$ dB/ $1^{\circ }$ followed by multiyear sea ice with $-\text{0.07}$ dB/ $\text{1}^{\circ }$ , and old multiyear ice (older than three years) with $-\text{0.04}$ dB/ $\text{1}^{\circ }$ . The noise floor is found to have a strong impact on smooth first-year ice and estimated slopes are therefore not fully reliable. At co-polarization, we obtained slope values of $-0.24$ , $-0.20$ , $-\text{0.15},$ and $-\text{0.10}$ dB/ $\text{1}^{\circ }$ for smooth first-year, rough/deformed first-year, multiyear, and old multiyear sea ice, respectively. Furthermore, we show that imperfect noise correction of the first subswath influences the obtained slopes for multiyear sea ice. We demonstrate that incidence angle normalization should not only be applied to co-polarization but should also be considered for cross-polarization images to minimize intra ice type variation in backscatter intensity throughout the entire image swath.

Climate models are our best tools to quantify ongoing changes caused by the climate crisis, but they are not perfect. The Arctic Ocean is particularly challenging to simulate: complex circulation flowing through narrow gateways and around tortuous bathymetry, dense water cascading off the steep shelf break, slow exchanges in canyons, along with known biases in sea ice and neighbouring seas.We investigate the Arctic Ocean in the historical run of 14 distinct models that participated to the latest Climate Model Intercomparison Project phase 6 (CMIP6) and find large biases in temperature, salinity, density, and depth of critical water masses, both on the shelves and in the deep basins. The biases are consistent throughout the water column and throughout the Arctic, with correlations often exceeding 0.9. However, no significant trend is observed in these biases, suggesting that the deep basins of the Arctic are not correctly ventilated already at the level of the Atlantic Water.Using the subset of models that submitted the age of water output, we confirm this absence of ventilation by dense water overflows: the overflows occur at too few locations and are diluted at shallow depths. Work is ongoing to relate these biases to the relevant processes, the upper water column, and fluxes through the various Arctic Ocean gateways.

The Arctic Ocean is strongly stratified by salinity gradients in the uppermost layers. This stratification is a key attribute of the region as it acts as an effective barrier for the vertical exchanges of Atlantic Water (AW) heat, nutrients, and CO2 between intermediate depths and the surface of the deep Eurasian and Amerasian Basins (EB and AB). Observations show that from 1970 to 2017, the stratification in the AB has strengthened, whereas, in parts of the EB, the stratification has weakened. The strengthening of the stratification in the AB is linked to a freshening and deepening of the halocline. The weakened stratification in parts of the EB is linked to a shoaling, warming, and lack of freshening of the halocline (Atlantification). Future simulations from a suite of CMIP6 models project that under a strong greenhouse-gas forcing scenario (SSP585), the AB and EB surface freshening and AW warming continues. To meaningfully compare hydrographic changes in the simulations, we present a new indicator of stratification. We find that within the AB, there is agreement among the models that the upper layers will become more stratified in the future. However, within the EB models diverge regarding future stratification. We discuss and detail some mechanisms responsible for these simulated discrepancies.

Deep water masses are the driver of the global ocean circulation, critical for transporting oxygen and nutrients throughout the water column, and a crucial mitigator of current climate change. They are also notoriously hard to observe: they form in winter in ice-infested areas, and then travel around the globe too deep for most autonomous instruments to monitor them. Therefore, although they represent at least half of the ocean volume, we still know very little about their circulation and variability.What we do know is that they are already changing, much faster than expected.From a ship in the Southern Ocean to models in the Arctic, I will share with you my obsession for these fascinating deep waters; highlight the blind spots that remain; and describe recent and upcoming deep-water-targeting projects that get me excited.

Global mean sea level is rising, however not uniformly. Regional deviations of sea surface height (SSH) are common due to local drivers, including surface winds, ocean density stratifications, vertical land- & crustal movements and more. The contribution of each background driver needs to be better understood to create reliable sea level rise projections, enable effective local policymaking and aid in urban planning decisions.In this study, we assess region-specific historic sea levels along the western Swedish coastline (Kattegat, Skagerrak & South Baltic Sea).We use monthly satellite altimetry observations spanning 26 years and daily observations spanning 6 years, as well as in situ tide gauge measurements to identify SSH covariance between sub-regions. We employed a number of manual statistical methods and found that SSH variability in the Skagerrak and Kattegat Seas behaves differently than areas south of the Danish Straits. While typically the correlation between SSH time series from different locations declines with distance, this is not seen at the entrance to the Baltic Sea due to the complexity of the region. To investigate this further and identify underlying primary forcings, we introduced re-analyzed ERA5 estimates of climatic drivers such as 10m-winds, sea surface temperature and sea level pressure, and tested these against principle components of the SSH variability signal within these regions. Zonal winds are most important for determining short-term sea level variability in areas north of the Danish Straits, while neither of these drivers successfully explain observed sea level variability south of them. As freshwater discharge from rivers and tributaries to the Baltic Sea is large, pressure- & density gradients may be more important as SSH regulators in this area.Additionally, we used neural networks to try to capture non-linear dependencies between the sea level drivers and sea level that are not apparent from statistical analyses. By predicting sea level at selected locations from different combination of drivers, we can determine which drivers have the highest influence. While feedforward neural networks did successfully predict some variability, they prove rather limited as delays between signals are present. Future tests using recurrent neural networks with a long short-term memory architecture might prove more successful.

Sea level is rising, threatening coastal population and infrastructures. While the causes for global sea level rise are relatively well determined and the projections can be made based on most likely climate scenarios, sea level changes in coastal regions are far less understood. Apart from the global temperature rise and the melting of ice sheets and glaciers, coastal areas are also affected by the spatio-temporal temperature and salinity variations, atmospheric conditions, and even vertical land motions, such as those due to glacial isostatic adjustment. It is therefore necessary to determine which factors contribute most to the sea level change in each specific region, in order to mitigate the effect sea level changes will have on local infrastructure. We here use long short-term memory (LSTM) neural networks to find the connection between the sea level variations and its many potential drivers in the Baltic, North, and Nordic Seas, at daily to decadal scales. We test all the different combinations of local atmospheric (surface pressure, wind, and precipitation from reanalysis) and oceanic (temperature and salinity, from in-situ observations), along with the remote ones (e.g. Greenland ice sheet runoff, large scale water mass circulation) to predict sea level variations at selected locations that have uninterrupted long-term (>50 years) tide gauge observations.By comparing the quality of sea level prediction from these different combinations, we find that the long-term sea level trend and low-frequency variations at most locations in our region of interest are mainly driven by the temperature rise, both local and remote, while the higher frequency variations are predominantly driven by the changes in local wind. As expected, northern locations are also affected by glacial isostatic adjustment, which counteracts the sea level rise. Precipitation, even during major storm events, seems to play an insignificant role in our region. The exception is the Baltic Sea, where wind plays less of a role, and the sea level is more affected by the influx of fresh water. While most regions are affected by sea level rise to some extent, because the causes for the local sea level changes vary, the protection from flooding and the warning techniques have to be adapted for each location.

Open-ocean and coastal polynyas, the result of high-latitude atmosphere-ocean circulation interannual variability, alter the local air-ocean heat exchange and sea ice production. Yet, the role of the ocean, especially its thermal flux, is rarely discussed. Here we examine the surface heat budget and sea ice changes during open-ocean and coastal polynya events in the Weddell Sea using satellite retrievals, in-situ observations, and the Japanese 55-year Reanalysis (JRA55). We find that the oceanic heat flux amounts to about 57.5±4 and 39±3 W/m2 within the 2016 and 2017 polynyas events, respectively; including these values in sea ice thickness. parameterizations significantly reduced their biases. Moreover, we compare sea ice mass productions within coastal and open-ocean polynyas using three methods. The results suggest that more ice production, albeit thinner ice thickness, occurs within the open-ocean than the coastal polynya. Finally, we find that wind and air temperature directly play a crucial role in controlling sea ice production in open-ocean polynyas, and undirectly via their impact on the polynya extent for coastal polynyas. The presence of wide open-ocean polynya does appear to reduce this influences on the ice production within the coastal polynya.

The retreating ice cover of the Central Arctic Ocean (CAO) fuels speculations on future fisheries. However, very little is known about the existence of harvestable fish stocks in this 3.3 million-square kilometer ecosystem around the North Pole. Crossing the Eurasian Basin, we documented an uninterrupted 3170-kilometer-long deep scattering layer (DSL) with zooplankton and small fish in the Atlantic water layer at 100- to 500-meter depth. Diel vertical migration of this central Arctic DSL was lacking most of the year when daily light variation was absent. Unexpectedly, the DSL also contained low abundances of Atlantic cod, along with lanternfish, armhook squid, and Arctic endemic ice cod. The Atlantic cod originated from Norwegian spawning grounds and had lived in Arctic water temperature for up to 6 years. The potential fish abundance was far below commercially sustainable levels and is expected to remain so because of the low productivity of the CAO.

Special Feature: The Multidisciplinary Drifting Observatory for the Study of Arctic Climate (MOSAiC).-- 42 pages, 16 figures, 3 tables, supplemental files https://doi.org/10.1525/elementa.2021.000046.-- Data accessibility statement: All data in this manuscript are publicly available from online repositories. Note that most data sets contain raw or preliminary data, while advanced versions will become available in future. The data may be found under the following references: drift track data (Figure 1, Nicolaus et al., doi:10.1594/PANGAEA.937204), observational dates (Figure 4, Nicolaus et al., doi:10.5281/zenodo.5898517), panorama photographs (Figure 5, Nicolaus et al., doi:10.1594/PANGAEA.938534), TLS data (Figure 6, Clemens-Sewall et al., doi:10.18739/A27S7HT3B), ROV radiation data (Figure 7, Nicolaus et al., doi:10.1594/PANGAEA.935688), surface albedo data on ground (Figure 8, Smith et al., broadband data under doi:10.18739/A2KK94D36 and spectral data under doi:10.18739/A2FT8DK8Z) and from the HELiX drone (Figure 8, Calmer et al., doi:10.18739/A2GH9BB0Q), on-ice RS data (Figure 10, Spreen et al., doi:10.5281/zenodo.5725870), surface images from thermal infrared and true color (Figure 11, Thielke et al, doi:10.1594/PANGAEA.934666), drift speed data from Polarstern (Figure 12, Nicolaus et al., doi:10.1594/PANGAEA.937204), deformation data from SAR (Figure 13, von Albedyll et al, doi:10.5281/zenodo.5195366), sea ice thickness and snow depth distribution (Figure 14, Hendricks et al., doi:10.5281/zenodo.5155244), sea ice physical properties (Figure 15, in Tables S2 and S3) with a sea ice core overview (Granskog et al., doi:10.5281/zenodo.4719905), snow pack properties (Figure 16, Macfarlane et al., doi: 10.1594/PANGAEA.935934), and ship radar video sequence (Jäkel et al., doi:10.5446/52953), Year-round observations of the physical snow and ice properties and processes that govern the ice pack evolution and its interaction with the atmosphere and the ocean were conducted during the Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition of the research vessel Polarstern in the Arctic Ocean from October 2019 to September 2020. This work was embedded into the interdisciplinary design of the 5 MOSAiC teams, studying the atmosphere, the sea ice, the ocean, the ecosystem, and biogeochemical processes. The overall aim of the snow and sea ice observations during MOSAiC was to characterize the physical properties of the snow and ice cover comprehensively in the central Arctic over an entire annual cycle. This objective was achieved by detailed observations of physical properties and of energy and mass balance of snow and ice. By studying snow and sea ice dynamics over nested spatial scales from centimeters to tens of kilometers, the variability across scales can be considered. On-ice observations of in situ and remote sensing properties of the different surface types over all seasons will help to improve numerical process and climate models and to establish and validate novel satellite remote sensing methods; the linkages to accompanying airborne measurements, satellite observations, and results of numerical models are discussed. We found large spatial variabilities of snow metamorphism and thermal regimes impacting sea ice growth. We conclude that the highly variable snow cover needs to be considered in more detail (in observations, remote sensing, and models) to better understand snow-related feedback processes. The ice pack revealed rapid transformations and motions along the drift in all seasons. The number of coupled ice–ocean interface processes observed in detail are expected to guide upcoming research with respect to the changing Arctic sea ice, This work was funded by the following: – the German Federal Ministry of Education and Research (BMBF) through financing the Alfred-Wegener-Institut Helmholtz-Zentrum für Polar- und Meeresforschung (AWI) and the Polarstern expedition PS122 under the grant N-2014-H-060_Dethloff, – the AWI through its projects: AWI_ROV, AWI_ICE, AWI_SNOW, AWI_ECO. The AWI buoy program and ROV work were funded by the Helmholtz strategic investment Frontiers in Arctic Marine Monitoring (FRAM), – the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) through the Transregional Collaborative Research Centre TRR-172 “ArctiC Amplification: Climate Relevant Atmospheric and SurfaCe Processes, and Feedback Mechanisms (AC)3” (grant 268020496), the International Research Training Group 1904 ArcTrain (grant 221211316), the MOSAiCmicrowaveRS project (grant 420499875), the HELiPOD grant (LA 2907/11-1), and the SCASI (NI 1096/5-1 and KA 2694/7-1) and SnowCast (AR1236/1) projects, – the BMBF through the projects Diatom-ARCTIC (03F0810A), IceSense (BMBF 03F0866A and 03F0866B), MOSAiC3-IceScan (BMBF 03F0916A), NiceLABpro (BMBF 03F0867A), SSIP (01LN1701A), and SIDFExplore (03F0868A), – the German Federal Ministry for Economic Affairs and Energy through the project ArcticSense (BMWi 50EE1917A), – the US National Science Foundation (NSF) through the project PROMIS (OPP-1724467, OPP-1724540, and OPP-1724748), the buoy work (OPP-1723400), the MiSNOW (OPP-1820927), the snow transect work (OPP-1820927), the sea ice coring work (OPP-1735862), the HELiX drone operations (OPP-1805569), surface energy fluxes (OPP-1724551), Climate Active Trace Gases (OPP-1807496), and Reactive Gas Chemistry (OPP-1914781). The last 4 of these were also supported by the NOAA Physical Sciences Laboratory, – the European Union’s Horizon 2020 research and innovation program projects ARICE (grant 730965) for berth fees associated with the participation of the DEARice team and INTAROS (grant 727890) supporting the drone and albedo measurements, – the US Department of Energy Atmospheric Radiation Measurement (ARM) and Atmospheric System Research (ASR) programs (DE-SC0019251, DE-SC0021341), – the National Aeronautics and Space Administration (NASA) project 80NSSC20K0658, – the European Space Agency (ESA) MOSAiC microwave radiometer (EMIRAD2, ELBARA, HUTRAD), (EMIRAD2, ELBARA, HUTRAD), CIMRex (contract 4000125503/18/NL/FF/gp) and GNSS-R (P.O. 5001025474, C.N. 4000128320/19/NL/FF/ab) GNSS-R (contracts P.O. 5001025474 and C.N. 4000128320/19/NL/FF/ab) projects, – the Canadian Space Agency FAST project (grant no. 19FACALB08), – EUMETSAT support for microwave scatterometer measurements, – the Research Council of Norway through the projects HAVOC (grant no. 280292), SIDRiFT (grant no. 287871), and CAATEX (grant no. 280531), – the Fram Centre (Tromsø, Norway), from its flagship program on Arctic Ocean through the PHOTA project, – the UKRI Natural Environment Research Council (NERC) and BMBF, who jointly funded the Changing Arctic Ocean program (project Diatom Arctic, NE/R012849/1 and 03F0810A), – the UK Natural Environment Research Council (project SSAASI-CLIM grant NE/S00257X/1), – the Agencia Estatal de Investigación AEI of Spain (grant no. PCI2019-111844-2, RTI2018-099008-B-C22), – the Swedish Research Council (VR, grant no. 2018-03859), – the Swedish Polar Research Secretariat for berth fees for MOSAiC, – the Swiss Polar Institute project SnowMOSAiC, – the Werner-Petersen-Foundation for the development of a remotely operated floating platform (grant no. FKZ 2019/610).

Arctic Ocean properties and processes are highly relevant to the regional and global coupled climate system, yet still scarcely observed, especially in winter. Team OCEAN conducted a full year of physical oceanography observations as part of the Multidisciplinary drifting Observatory for the Study of the Arctic Climate (MOSAiC), a drift with the Arctic sea ice from October 2019 to September 2020. An international team designed and implemented the program to characterize the Arctic Ocean system in unprecedented detail, from the seafloor to the air-sea ice-ocean interface, from sub-mesoscales to pan-Arctic. The oceanographic measurements were coordinated with the other teams to explore the ocean physics and linkages to the climate and ecosystem. This paper introduces the major components of the physical oceanography program and complements the other team overviews of the MOSAiC observational program. Team OCEAN’s sampling strategy was designed around hydrographic ship-, ice- and autonomous platform-based measurements to improve the understanding of regional circulation and mixing processes. Measurements were carried out both routinely, with a regular schedule, and in response to storms or opening leads. Here we present along-drift time series of hydrographic properties, allowing insights into the seasonal and regional evolution of the water column from winter in the Laptev Sea to early summer in Fram Strait: freshening of the surface, deepening of the mixed layer, increase in temperature and salinity of the Atlantic Water. We also highlight the presence of Canada Basin deep water intrusions and a surface meltwater layer in leads. MOSAiC most likely was the most comprehensive program ever conducted over the ice-covered Arctic Ocean. While data analysis and interpretation are ongoing, the acquired datasets will support a wide range of physical oceanography and multi-disciplinary research. They will provide a significant foundation for assessing and advancing modeling capabilities in the Arctic Ocean.

International audience; With the Arctic rapidly changing, the needs to observe, understand, and model the changes are essential. To support these needs, an annual cycle of observations of atmospheric properties, processes, and interactions were made while drifting with the sea ice across the central Arctic during the Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition from October 2019 to September 2020. An international team designed and implemented the comprehensive program to document and characterize all aspects of the Arctic atmospheric system in unprecedented detail, using a variety of approaches, and across multiple scales. These measurements were coordinated with other observational teams to explore cross-cutting and coupled interactions with the Arctic Ocean, sea ice, and ecosystem through a variety of physical and biogeochemical processes. This overview outlines the breadth and complexity of the atmospheric research program, which was organized into 4 subgroups: atmospheric state, clouds and precipitation, gases and aerosols, and energy budgets. Atmospheric variability over the annual cycle revealed important influences from a persistent large-scale winter circulation pattern, leading to some storms with pressure and winds that were outside the interquartile range of past conditions suggested by long-term reanalysis. Similarly, the MOSAiC location was warmer and wetter in summer than the reanalysis climatology, in part due to its close proximity to the sea ice edge. The comprehensiveness of the observational program for characterizing and analyzing atmospheric phenomena is demonstrated via a winter case study examining air mass transitions and a summer case study examining vertical atmospheric evolution. Overall, the MOSAiC atmospheric program successfully met its objectives and was the most comprehensive atmospheric measurement program to date conducted over the Arctic sea ice. The obtained data will support a broad range of coupled-system scientific research and provide an important foundation for advancing multiscale modeling capabilities in the Arctic.

Polynyas facilitate air–sea fluxes, impacting climate-relevant properties such as sea ice formation and deep water production. Despite their importance, polynyas have been poorly represented in past generations of climate models. Here we present a method to track the presence, frequency and spatial distribution of polynyas in the Southern Ocean in 27 models participating in the Climate Model Intercomparison Project Phase 6 (CMIP6) and two satellite-based sea ice products. Only half of the 27 models form open-water polynyas (OWPs), and most underestimate their area. As in satellite observations, three models show episodes of high OWP activity separated by decades of no OWP, while other models unrealistically create OWPs nearly every year. In contrast, the coastal polynya area is overestimated in most models, with the least accurate representations occurring in the models with the coarsest horizontal resolution. We show that the presence or absence of OWPs is linked to changes in the regional hydrography, specifically the linkages between polynya activity with deep water convection and/or the shoaling of the upper water column thermocline. Models with an accurate Antarctic Circumpolar Current transport and wind stress curl have too frequent OWPs. Biases in polynya representation continue to exist in climate models, which has an impact on the regional ocean circulation and ventilation that should be addressed. However, emerging iceberg discharge schemes, more adequate vertical grid type or overflow parameterisation are anticipated to improve polynya representations and associated climate prediction in the future.

Growing interest in understanding the relevance of marine fungi to food webs, biogeochemical cycling, and biological patterns necessitates establishing a context for interpreting future findings. To help establish this context, we summarize the diversity of cultured and observed marine planktonic fungi from across the world. While exploring this diversity, we discovered that only half of the known marine fungal species have a publicly available DNA locus, which we hypothesize will likely hinder accurate high-throughput sequencing classification in the future, as it does currently. Still, we reprocessed >600 high-throughput datasets and analyzed 4.9 × 109 sequences (4.8 × 109 shotgun metagenomic reads and 1.0 × 108 amplicon sequences) and found that every fungal phylum is represented in the global marine planktonic mycobiome; however, this mycobiome is generally predominated by three phyla: the Ascomycota, Basidiomycota, and Chytridiomycota. We hypothesize that these three clades are the most abundant due to a combination of evolutionary histories, as well as physical processes that aid in their dispersal. We found that environments with atypical salinity regimes (>5 standard deviations from the global mean: Red Sea, Baltic Sea, sea ice) hosted higher proportions of the Chytridiomycota, relative to open oceans that are dominated by Dikarya. The Baltic Sea and Mediterranean Sea had the highest fungal richness of all areas explored. An analysis of similarity identified significant differences between oceanographic regions. There were no latitudinal gradients of marine fungal richness and diversity observed. As more high-throughput sequencing data become available, expanding the collection of reference loci and genomes will be essential to understanding the ecology of marine fungi.

Previous literature has shown that Canadian Basin Deep Water (CBDW) crosses the Lomonosov Ridge into the Amundsen Basin close to the North Pole. This intrusion subsequently flows along the ridge towards Greenland and eventually all the way to the Greenland Sea, but an influence of CBDW in other parts of the Amundsen Basin has also been shown. We detect this deep CBDW intrusion, which is visible as a salinity maximum and oxygen minimum at a depth of about 2000 metres, in hydrographic measurements from MOSAiC and historical data sets. We also use measurements of CFC concentrations for increased robustness, as the high age of CBDW means the water mass is characterised by a CFC minimum. We map the recirculation of this CBDW in the Amundsen Basin and determine its spatial and temporal variability. In particular, we find that CBDW likely flows as a boundary current going eastwards along Gakkel Ridge, and even detect CBDW-like properties on the Nansen Basin side of Gakkel Ridge. As the Arctic Ocean is changing rapidly, understanding its deep circulation and its drivers is becoming increasingly urgent.

The Arctic Ocean, although remote to most of us, is linked to lower latitudes by way of climate, physics, biology and biogeochemistry. Strongly coupled to the rapidly changing Arctic atmosphere and sea-ice, the ocean is subject to amplification of change amid global trends in climate. The relatively fresh and cold upper mixed-layer in the Arctic basin exhibits a strong seasonal cycle, yet the deeper warm water of Atlantic origin largely stays isolated from the ice. Further, changes in heat, salt and momentum due to exchange with ice and atmosphere cannot penetrate to great depth due to a strong halocline. Nevertheless, we observed changes in the upper water column stratification and mixing, due to storms and freeze-induced brine release during the year-long MOSAiC experiment. This was further expressed by significant variability in (sub)mesoscale processes, including eddies and frontal adjustment. We will present results from ocean observations during the MOSAiC drift using a variety of manually-operated devices and autonomous platforms within several 10s of kilometres from the drifting icebreaker Polarstern. Preliminary analyses of our data highlight a pronounced seasonal cycle in mixed-layer depth and upper ocean stratification characteristics connected to brine release, turbulent events triggered by storms, and geographic background variability. We will further detail the observed full-depth water mass distribution and attempt to untangle temporal and spatial variability. Finally, we will give an overview of Team-OCEAN analyses and interdisciplinary projects.

When will sea ice open is a crucial information for navigation and scientific deployments. This became painfully obvious when the Weddell Polynya, a large hole in the winter Southern Ocean sea ice, unexpectedly re-opened in 2016 for the first time in forty years. With no early warning, observations were limited to chance autonomous sensors, so the much-debated cause of the opening still cannot be determined accurately. We aim here to create such an early warning system. From the full historical sea ice concentration record, we find in fact 30 polynyas since 1980. Then, using the full time series of the spaceborne infrared Advanced Very High Resolution Radiometer, we determine that these events can be detected in the two weeks before the polynya opens. Area-average median brightness temperature larger than 253 K in all three bands and area-maximum larger than 269 K along with a footprint at least larger than 4000 km2 successfully forecasts the polynyas and does not return any false positive. Or rather, it returned false positives that were in fact events that the sea ice concentration threshold had missed. Moreover, we find temporal oscillations in brightness temperature that could indicate upwelling of warm water, but also changes of sign in T45 (band 4 – band 5) which could indicate a lead. We hence combine the spaceborne infrared data with atmospheric reanalysis, hydrographic mooring data and Sentinel-1 radar imagery and find that all events, including the 2017 Weddell Polynya, are caused by both atmospheric divergence and oceanic upwelling. That is, the debate is closed: both parties are correct; the Weddell Polynya is a hybrid.

The Weddell Polynya, a large hole in the winter sea ice cover, has intrigued researchers since satellite observations began in the late 70s. There is no consensus regarding the mechanisms leading to its opening, not least because there never was an instrument deployed early enough at the right location and with the right sampling interval. But what if we could predict imminent openings, by detecting early-warning signs from space?The leading theory among oceanographers is that the polynya opens after the sea ice is melted from below by upwelled warm waters. We argue that such upwelling, or at least the increased heat flux through a thinning ice, should be visible on spaceborne thermal infrared imagery. Using microwave-based sea ice products to determine past polynya openings, we first found that there were in fact 83 Weddell / Maud Rise Polynya occurrences since winter 2000, 19 of which reaching an area larger than 1000 km2. We then created a timeseries of (cloud-filtered) daily mean brightness temperature at 3.7, 10.5 and 12 μm from Advanced Very High Resolution Radiometer datasets and found a significant warm temperature anomaly at least 10 days before the polynya opened, peaking at 4K for all bands 5-6 days before the opening. The anomaly is on average 2K stronger for the large polynyas (> 1000 km2). Moreover, the band ratios brutally change magnitude, which suggests lead formation rather than progressive melting – a hypothesis that would agree with meteorologists' theory that the polynya opens because of winds, and that we are now checking with spaceborne radar.Six days is not much, but it would be enough to re-route expeditions or autonomous sensors so that the opening can be monitored in details. And this is only the first step of our ongoing project... stay tuned to see if we can predict weeks or months ahead!

The sea ice thickness in the Weddell Sea during the austral winter normally exceeds 1 m, but in the case of a polynya, this thickness decreases to 10 cm or less. There are two theories as to why the Weddell Polynya opens: 1) comparatively warm oceanic water upwelling from its nominal depth of several hundred metres to the surface where it melts the sea ice from underneath; or 2) opening of a lead by a passing storm, lead which will then be maintained open either by the atmosphere or ocean and grow. The objective of this study is to estimate how long in advance the recent Weddell Polynya opening could have been detected by synthetic aperture radar (SAR) images due to the decrease of the sea ice thickness and/or early appearance of leads. We use high temporal and spatial resolution SAR images from the Sentinel-1 constellation (C-band) and ALOS2 (L-band) during the austral winters 2014-2018. We use an adapted version of the algorithm developed by Aldenhoff et al. (2018) to monitor changes in sea ice thickness over the polynya region. The algorithm detects the transition of the sea ice thickness through changes in small scale surface roughness and thus reduced backscatter, and allowing us to distinguish three different categories: ice, thin ice, and open water. The transition from ice to thin ice and then to open water indicates that the polynya is melted from under, whereas a direct transition from ice to open water will reveal leads. The high resolution and good coverage of the SAR imagery, and a combined effort of different satellites sensors (e.g. infrared and microwave sensors), opens the possibility of an early detection of Weddell Polynya opening.

The presence of polynyas has a large effect on air-sea fluxes and deep water production, therefore impacting climate-relevant properties such as heat and carbon exchange between the atmosphere and ocean interior. One of the key areas of deep water formation is in the Weddell Sea, where much attention has recently been placed in the reoccurance of the open ocean Maud Rise polynya. In this study, two methods are presented to track the number, area and spatial distribution of polynyas with a focus on the Weddell Sea. The analysis is applied to a set of 10 Coupled Model Intercomparison Project phase 6 (CMIP6) models and to satellite sea ice concentration data. The first approach is a sea ice threshold method applied to the CMIP6 sea ice data at the original model grid. Open water areas surrounded by sea ice are classified as polynyas. Without requiring any remapping or interpolation, this method preserves the area information of all grid cells and is well suited to compute the combined area of the polynyas in the Weddell Sea. The second approach makes use of an image analysis technique to outline areas with low sea ice concentration. This method is preferable for counting the absolute number of polynyas and obtaining statistical information about their position. Satellite sea ice concentration is used as a reference to compare the performance of the models representing polynya area and to indicate model biases in the location of polynyas. All analyzed CMIP6 models show coastal polynyas, while only about half of the models regularly form open water polynyas. The resolution (about one degree for most models) sets a limit for the number of the polynyas in the numerical models.

Mass loss from the Antarctic Ice Sheet to the ocean has increased in recent decades, largely because the thinning of its floating ice shelves has allowed the outflow of grounded ice to accelerate1,2. Enhanced basal melting of the ice shelves is thought to be the ultimate driver of change2,3, motivating a recent focus on the processes that control ocean heat transport onto and across the seabed of the Antarctic continental shelf towards the ice4–6. However, the shoreward heat flux typically far exceeds that required to match observed melt rates2,7,8, suggesting that other critical controls exist. Here we show that the depth-independent (barotropic) component of the heat flow towards an ice shelf is blocked by the marked step shape of the ice front, and that only the depth-varying (baroclinic) component, which is typically much smaller, can enter the sub-ice cavity. Our results arise from direct observations of the Getz Ice Shelf system and laboratory experiments on a rotating platform. A similar blocking of the barotropic component may occur in other areas with comparable ice–bathymetry configurations, which may explain why changes in the density structure of the water column have been found to be a better indicator of basal melt rate variability than the heat transported onto the continental shelf9. Representing the step topography of the ice front accurately in models is thus important for simulating ocean heat fluxes and induced melt rates. The front of the Getz Ice Shelf in West Antarctica creates an abrupt topographic step that deflects ocean currents, suppressing 70% of the heat delivery to the ice sheet.