Search

Antarctic sea ice extent has been persistently low since late 2016, possibly owing to changes in atmospheric and oceanic conditions. However, the relative contributions of the ocean, the atmosphere and the underlying mechanisms by which they have affected sea ice remain uncertain. To investigate possible causes for this sea‐ice decrease, we establish a seasonal timeline of sea ice changes following 2016, using remote sensing observations. Anomalies in the timing of sea ice retreat and advance are examined along with their spatial and interannual relations with various indicators of seasonal sea ice and oceanic changes. They include anomalies in winter ice thickness, spring ice removal rate due to ice melt and transport, and summer sea surface temperature. We find that the ice season has shortened at an unprecedented rate and magnitude, due to earlier retreat and later advance. We attribute this shortening to a winter ice thinning, in line with ice‐albedo feedback processes, with ice transport playing a smaller role. Reduced ice thickness has accelerated spring ice area removal as thinner sea ice requires less time to melt. The consequent earlier sea ice retreat has in turn increased ocean solar heat uptake in summer, ultimately delaying sea ice advance. We speculate that the observed winter sea ice thinning is consistent with previous evidence of subsurface warming of the Southern Ocean. Plain Language Summary: Following 2016, the Antarctic sea ice cover has been persistently low. To understand why, we retrace the seasonal timeline of sea ice and oceanic changes, using satellite observations. We find that the sea ice season has been significantly shorter following 2016, due to a later start and an earlier end. This shortening is likely caused by thinner sea ice in winter, accelerating sea ice melt in spring and causing sea ice to disappear earlier. In turn, this has caused the ice‐free ocean to absorb more solar heat in summer, and to take longer to cool down in the fall, ultimately delaying the ice season onset. Warmer waters beneath the Southern Ocean's surface, as evidenced by previous studies, could be a plausible cause of the observed thinning, by increasing the heat supply to sea ice. Key Points: The Antarctic sea ice season duration has undergone an unprecedented shortening since 2016The changes include thinner ice, faster melt, earlier retreat, larger ocean heat uptake, later advance, in line with the ice‐albedo feedbackThe near‐circumpolar ice thinning suggests a possible increase in sensible heat supply by the ocean [ABSTRACT FROM AUTHOR]

The correct representation of the Maud Rise open‐ocean polynya in the Weddell Sea remains a challenge for ocean models. Here we reproduce the most recent polynya openings in 2016–2017 using a regional configuration, and assess their dependencies on vertical convective mixing schemes and freshwater forcing, both separately and in combination. We test three vertical convective mixing schemes: the enhanced vertical diffusion (EVD), the Eddy‐Diffusivity Mass‐Flux (EDMF) parameterization, and a modified version of EDMF accounting for thermobaric effects. Using simulations for the period 2007–2017, we find that the modified EDMF reproduces the observed climatological evolution of the mixed layer depth better than the original EDMF and the EVD, but a polynya fails to open due to excessive freshwater forcing. We thus use the modified EDMF to perform sensitivity experiments with reduced precipitation during 2012–2017. The imposed freshwater forcing strongly affects the number of years with polynyas. The simulation with the best representation of the 2016–2017 polynyas is analyzed to evaluate the triggering mechanisms. The 2016 polynya was induced by the action of thermobaric instabilities on a weak ambient stratification. This opening preconditioned the water column for 2017, which produced a stronger polynya. By examining the impacts of the different convective mixing schemes, we show that the modified EDMF generates more realistic patterns of deep convection. Our results highlight the importance of surface freshwater forcing and thermobaricity in governing deep convection around Maud Rise, and the need to represent thermobaric instabilities to realistically model Maud Rise polynyas. Plain Language Summary: We investigate the impacts of representing numerical vertical mixing and surface freshwater forcing in a regional ocean model on polynyas (large openings in the pack ice) at Maud Rise, Southern Ocean. Maud Rise is prone to hosting polynyas, often associated with deep convection, which is a local vertical mixing process homogenizing the water column between surface and depths of several hundred meters. Numerical models often use simplistic strategies to represent this process, but improved parameterizations have recently become available. In this work, we test the impact of the representation of convective mixing in a particularly sensitive region. The last Maud Rise polynyas were observed in 2016 and 2017. Our regional simulation is capable of reproducing these polynyas, which has long been a challenge for ocean‐sea ice models. We show that the 2016 polynya resulted from the action of a vertical instability at depth acting on weak ambient stratification. This event preconditioned the stronger 2017 polynya and deep convection. We conclude that representing convective plumes as a sub‐grid scale process in models leads to a more realistic representation of open‐ocean polynyas and associated convection events. Key Points: The Eddy‐Diffusivity Mass‐Flux (EDMF) parameterization is tested in a regional simulation of the ocean around Maud RiseThermobaric effects on convective plumes are enabled by modifying the EDMF parameterizationSimulations of Maud Rise polynyas are highly sensitive to freshwater forcing and mixing schemes [ABSTRACT FROM AUTHOR]

The region of Maud Rise, a seamount in the Weddell Sea, is known for the occurrence of irregular polynya openings during the winter months. Hydrographic observations have shown the presence of a warmer water mass below the mixed layer along the seamount's flanks, commonly termed the warm‐water Halo, surrounding a colder region above the rise, the Taylor Cap. Here we use two observational data sets, an eddy‐permitting reanalysis product and regional high‐resolution simulations, to investigate the interannual variability of the Halo and Taylor Cap for the period 2007–2022. Observations include novel hydrographic profiles obtained in the Maud Rise area in January 2022, during the first SO‐CHIC cruise. It is demonstrated that the temperature of deep waters around Maud Rise exhibits strong interannual variability within the Halo and Taylor Cap, occasionally to such an extent that the two features become indistinguishable. A warming of deep waters by as much as 0.8°C is observed in the Taylor Cap during the years preceding the opening of a polynya in 2016 and 2017, starting in 2011. By analyzing regional simulations, we show that most of the observed variability in the Halo is forced remotely by advection of deep waters from the Weddell Gyre into the region surrounding Maud Rise. Our highest‐resolution simulation indicates that mesoscale eddies subsequently transfer the properties of the Halo's deep waters onto the Taylor Cap. The eddies responsible for such transfer originate in an abrupt retroflection along the inner flank of the Halo. Plain Language Summary: A polynya is an opening within the sea ice cover during winter. In the Weddell Sea, polynyas occur irregularly and often emerge close to a seamount, Maud Rise. This topographic obstacle disturbs the local circulation of the Weddell Gyre and leads to the formation of two local permanent features: a Taylor Cap on top of the seamount, which is mainly an isolated water mass; and surrounding this, a warm‐water Halo, located along the flanks of Maud Rise. In this study, we use observational and model analyses to show the changes in the temperature of regional water masses on a timescale of years, and to identify the drivers of these changes. The properties of the warm‐water Halo are controlled by the advection of deep waters along the south‐eastern rim of the Weddell Gyre. These waters are partly mixed into the Taylor Cap by eddies, thereby preconditioning the water column for deep convection. In the years preceding the opening of a polynya in 2016–2017, we document a warming trend in the Taylor Cap, starting in 2011. Key Points: Interannual variability of the subsurface temperature maximum at Maud Rise is documented with observationsAdvection of anomalously cold and fresh deep waters from the Weddell Gyre into the Halo leads to a near‐vanishing of the Taylor Cap in 2014Observed warming of the subsurface layer in the Taylor Cap due to eddy transport preceded the polynya opening in 2016–2017 [ABSTRACT FROM AUTHOR]

Coastlines in most ocean general circulation models are piecewise constant. Accurate representation of boundary currents along staircase‐like coastlines is a long‐standing issue in ocean modeling. Pioneering work by Adcroft and Marshall (1998, https://doi.org/10.3402/tellusa.v50i1.14514) revealed that artificial indentation of model coastlines, obtained by rotating the numerical mesh within an idealized square basin, generates a spurious form drag that slows down the circulation. Here, we revisit this problem and show how this spurious drag may be eliminated. First, we find that physical convergence to spatial resolution (i.e., the main characteristics of the flow are insensitive to the increase of the mesh resolution) allows simulations to become independent of the mesh orientation. An advection scheme with a wider stencil also reduces sensitivity to mesh orientation from coarser resolution. Second, we show that indented coastlines behave as straight and slippery shores when a true mirror boundary condition on the flow is imposed. This finding applies to both symmetric and rotational‐divergence formulations of the stress tensor, and to both flux and vector‐invariant forms of the equations. Finally, we demonstrate that the detachment of a vortex flowing past an outgoing corner of the coastline is missed with a free‐slip (zero vorticity) condition at the corner. These results provide guidance for a better numerical treatment of coastlines (and isobaths) in ocean general circulation models. Plain Language Summary: Most ocean general circulation models represent coastlines as piecewise constant, which does not accurately reflect the true boundary. This approximation is necessitated by the size and square shape of model grid cells, together with finite computational resources, making it difficult to finely represent the boundary. A long‐standing issue in ocean modeling is accurately representing boundary currents along these staircase‐like coastlines. In particular, Adcroft and Marshall (1998, https://doi.org/10.3402/tellusa.v50i1.14514) discovered that artificial indentation of the model coastlines generates a "spurious form drag" that slows down the circulation. Our study revisits this long‐standing issue and shows how this spurious drag may be eliminated. First, we demonstrate that having a sufficiently fine spatial resolution to resolve the physical processes allows the model to be insensitive to coastal indentation. We also show that indented coastlines become slippery, as if they were smooth and straight shores, when a true mirror boundary condition on the flow is imposed. Finally, we show how to faithfully simulate the retroflection of a current past a cape. In summary, these results provide guidance for a better numerical representation of marine land‐forms in numerical ocean models. Key Points: The impact of coastal indentation on circulation within an idealized square basin is studied systematically for various numerical choicesSolutions are insensitive to coastal indentation with rot‐div formulation of the diffusive tensor and sufficiently fine spatial resolutionIndented coastlines become slippery when a true mirror boundary condition is imposed on the flow for all numerical formulations [ABSTRACT FROM AUTHOR]

The intensity and spectrum of light under Arctic sea ice, key to the energy budget and primary productivity of the Arctic Ocean, are tedious to observe. Earth System Models (ESMs) are instrumental in understanding the large‐scale properties and impacts of under‐ice light. To date, however, ESM parameterizations of radiative transfer have been evaluated with a few observations only. From observational programs conducted over the past decade at four locations in the Northern Hemisphere sea ice zone, 349 observational records of under‐ice light and coincident environmental characteristics were compiled. This data set was used to evaluate seven ESM parameterizations. Snow depth, melt pond presence and, to some extent, ice thickness explain the observed variance in light intensity, in agreement with previous work. The effects of Chlorophyll‐a are also detected, with rather low intensity. The spectral distribution of under‐ice light largely differs from typical open ocean spectra but weakly varies among the 349 records except for a weak effect of snow depth on the blue light contribution. Most parameterizations considered reproduce variations in under‐ice light intensity. Large errors remain for individual records, on average by a factor of ∼3, however. Skill largely improves if more predictors are considered (snow and ponds in particular). Residual errors are attributed to missing physics in the parametrizations, inconsistencies in the model‐observation comparison protocol, and measurement errors. We provide recommendations to improve the representation of light under sea ice in the ice‐ocean model NEMO, which may also apply to other ESMs and help improve next‐generation ESMs. Plain Language Summary: The Arctic sea ice cover has rapidly decreased over the last four decades, and Earth System Models (ESMs) project a complete decay of summer Arctic sea ice in the coming decades. As a result, phytoplankton, micro‐algae growing and drifting within seawater, experience increased light supply, favorable to photosynthesis. Phytoplankton is central to the Arctic marine ecosystem, because of its role as a carbon source to the whole food chain. However, there is low confidence in the capacity of models to project Arctic phytoplankton. This is because of large difficulties in simulating nutrient and light supply mechanisms, stemming from limits in basic understanding. Here we gather and analyze a large database of observations of the sea ice environment from the Northern Hemisphere. Snow depth and melt ponds emerge as key drivers of the under‐ice light climate. We also evaluate and improve a calculation method for light intensity and color used in several ESMs, and largely improve agreement with observations as compared with reference. Our results will contribute to lower uncertainties in light climate under sea ice in future ESM projections. Key Points: Snow depth, melt ponds, ice thickness, Chlorophyll‐a in sea ice, influence under‐ice light intensity and spectral distributionSeasonal variations in light intensity under sea ice are reproduced by Earth System Model parameterizations informed by observed environmental conditionsLarge uncertainties stem from measurement errors, missing physics in the parameterizations, and inconsistent comparison protocol [ABSTRACT FROM AUTHOR]

Storfjorden, Svalbard, hosts a polynya in winter and is an important source region of Brine‐enriched Shelf Water (BSW) that, if dense enough, feeds the Arctic Ocean deep water reservoir. Changes in the BSW production may thus have far‐reaching impacts. We analyze the water mass distribution and circulation in Storfjorden and the trough south of it, Storfjordrenna, using hydrographic sections occupied in July 2016, following a winter characterized by the lowest ice coverage recorded in the Barents Sea. These observations reveal an unusual hydrographic state, characterized at the surface by the near absence of Melt Water and Storfjorden Surface Water, replaced by a saltier water mass. At depth, BSW (maximum salinity of 34.95) was found from the bottom up to 90 m, above the 120‐m deep sill at the mouth to Storfjordrenna. However, no gravity driven overflow was observed downstream of the sill: the dome of BSW remained locked over the depression in a cyclonic circulation pattern consistent with a stratified Taylor column. Observations further reveal a previously unreported intrusion of Atlantic Water (AW) far into the fjord, promoting isopycnal mixing with entrapped Arctic Water. This intrusion was possibly favored by positive wind stress curl anomalies over Svalbardbanken and Storfjordrenna. The bottom plume exiting Storfjordrenna was weak, carrying Polar Front Water rather than BSW, too light to sink underneath the AW layer at Fram Strait. Whether Storfjorden switched durably to a new hydrographic state, following the observed Atlantification of the Barents Sea after 2005, remains to be established. Plain Language Summary: Storfjorden, east of Spitsbergen, plays an important role in Arctic Ocean climate through formation of dense water as salt is added to the ocean when sea ice forms. This dense water accumulates in winter before spilling toward the deep ocean into autumn, fueling the global ocean's circulation. We analyze observations from a research cruise in July 2016, following a winter season characterized by the lowest ice coverage in the Barents Sea ever recorded. These observations reveal striking differences from previous reports, which are mostly based on data prior to the 2005 regime shift in the Barents Sea characterized by warmer temperature and reduced ice cover, an expression of its "Atlantification" reported by many authors. First, the expected overflow of the locally formed dense water was absent. The latter, less saline than usual, was instead trapped in Storfjorden suggesting an intermittent discharge regime usually observed in the fall. Another notable observation is the intrusion of Atlantic Water far inside the fjord. Such a flooding episode, increasingly frequent in the fjords of the west coast of Spitsbergen, is previously unreported in Storfjorden. These observations made in the wake of an exceptionally mild winter could prefigure more permanent changes in this important region. Key Points: Changes in the hydrography and circulation in Storfjorden a decade after a regime shift in the Barents Sea linked to its AtlantificationThe less saline Brine‐enriched Shelf Water remained entrapped in Storfjorden within a cyclonic circulation pattern with no overflowStorfjorden was flooded with Atlantic Water, promoting isopycnal mixing with Arctic Water and the local formation of East Spitsbergen Water [ABSTRACT FROM AUTHOR]

A new one‐dimensional (1‐D) parameterization of penetrative convection has been developed in order to have a better representation of the vertical mixing in ocean general circulation models. Our approach is inspired from atmospheric parameterizations of shallow convection which assumes that in the convective boundary layer, the subgrid‐scale fluxes result from two different mixing scales: small eddies, which are represented by an Eddy‐Diffusivity (ED) contribution, and large eddies associated with thermals, which are represented by a mass‐flux contribution. In the present work, the local (small eddies) and nonlocal (large eddies) contributions are unified into an Eddy‐Diffusivity‐Mass‐Flux (EDMF) parameterization which treats simultaneously the whole vertical mixing. EDMF is implemented in the community ocean model NEMO and tested in its 1‐D column version. Deepening of dense water in analytic cases, successfully reproduced in LES simulations, is more realistic with EDMF than with standard diffusion parameterizations. Also the convective events observed in the western Mediterranean at the Lion station and in the North Pacific Ocean at the PAPA station are more realistic in terms of sequencing and amplitude with EDMF. Plain Language Summary: A new representation of oceanic convection has been developed in order to have a better representation of vertical mixing in ocean models. Our proposition is to represent oceanic convection consistently to atmospheric convection. We want to represent the convective plumes in oceans in the same way that the cumulus clouds are represented in the atmosphere. In this unified approach, the oceanic vertical mixing is viewed as a combination of large eddies associated with strong nonbuoyant downdrafts and small eddies which induce local turbulence. This new paradigm of oceanic mixing leads to more realistic simulations of the hydrological properties of water masses. In the future, it is expected to obtain more reliable climate projections. Key Points: An Eddy‐Diffusivity Mass‐Flux (EDMF) parameterization is proposed to represent oceanic convection consistently to atmospheric convectionThe EDMF parameterization unifies diffusion and convection processes in ocean modelsThe EDMF parameterization represents the penetrative convection and the associated counter‐gradient heat flux in the stratified thermocline [ABSTRACT FROM AUTHOR]

The current feedback to the atmosphere (CFB) contributes to the oceanic circulation by damping eddies. In an ocean‐atmosphere coupled model, CFB can be correctly accounted for by using the wind relative to the oceanic current. However, its implementation in a forced oceanic model is less straightforward as CFB also enhances the 10‐m wind. Wind products based on observations have seen real currents that will not necessarily correspond to model currents, whereas meteorological reanalyses often neglect surface currents or use surface currents that, again, will differ from the surface currents of the forced oceanic simulation. In this study, we use a set of quasi‐global oceanic simulations, coupled or not with the atmosphere, to (i) quantify the error associated with the different existing strategies of forcing an oceanic model, (ii) test different parameterizations of the CFB, and (iii) propose the best strategy to account for CFB in forced oceanic simulation. We show that scatterometer wind or stress are not suitable to properly represent the CFB in forced oceanic simulation. We furthermore demonstrate that a parameterization of CFB based on a wind‐predicted coupling coefficient between the surface current and the stress allows us to reproduce the main characteristics of a coupled simulation. Such a parameterization can be used with any forcing set, including future coupled reanalyses, assuming that the associated oceanic surface currents are known. A further assessment of the thermal feedback of the surface wind in response to oceanic surface temperature gradients shows a weak forcing effect on oceanic currents. Key Points: The Current FeedBack to the Atmosphere (CFB) can be parameterized in a forced oceanic modelA parameterization of the CFB based on a predicted coupling coefficient is the best parameterizationScatterometers are not suitable to correctly represent the CFB in a forced oceanic model (unless coherent surface currents are known) [ABSTRACT FROM AUTHOR]

Pure ice, brine and solid minerals are the main contributors to sea ice mass. Constitutional changes with salinity and temperature exert a fundamental control on sea ice physical, chemical, and biological properties. However, current estimation methods and model representations of the sea ice phase composition suffer from two limitations—in a context of poorly quantified uncertainties. First, salt minerals are neglected. Second, formulations are inconsistent with international standards, in particular with the International Thermodynamic Equation of Seawater (TEOS‐10). To address these issues, we revisit the thermodynamics of the sea ice phase composition by confronting observations, theory, and the usual computation methods. We find remarkable agreement between observations and the Gibbs‐Pitzer theory as implemented in FREZCHEM, both for brine salinity (RMSE = 1.9 g/kg) and liquid H2O mass fraction (RMSE = 8.6 g/kg). On this basis, we propose expanded sea ice phase composition equations including minerals, expressed in terms of International Temperature Scale 1990 temperature and absolute salinity, and valid down to the eutectic temperature (−36.2 °C). These equations precisely reproduce FREZCHEM, outcompeting currently used calculation techniques. We also suggest a modification of the TEOS‐10 seawater Gibbs function giving a liquidus curve consistent with observations down to the eutectic temperature without changing TEOS‐10 inside its original validity range. Plain Language Summary: Sea ice is made of pure ice, salt water, and solid minerals. The proportions of these constituents change with temperature and salinity. The constitution of sea ice, in particular salt water encased in the ice affects how sea ice responds to warming, where and when ice algae thrive and how sea ice changes the ocean and atmosphere chemical composition. We propose revised computation means for the composition of sea ice from the sole knowledge of temperature and salinity of a sea ice sample. Our developments are based on an exhaustive collection of historical observations and theoretical arguments, stemming from developments in thermochemistry of aqueous solutions. We find very good agreement between theoretical calculations and observations. We also estimate the current uncertainty on the sea ice composition. For instance, we suggest that the degree of precision on the salinity of liquid inclusions is less than 2 g of salt per unit mass of brine (<5%) and that the amount of liquid water in brine inclusions typically is within 5–10 g/kg of sea ice (<5%). Our calculations will hopefully help other researchers to better describe sea ice composition in their observational and model studies. Key Points: Revisit sea ice phase composition from observational and theoretical sourcesQuantify uncertainties in brine salinity and mass fractionPropose sea ice phase equations with solid minerals that are compatible with international standards [ABSTRACT FROM AUTHOR]

Climate modes, such as the El Niño Southern Oscillation (ENSO), influence Tropical Cyclones (TCs) interannual activity through their effect on large-scale atmospheric environment. These climate modes also induce interannual variations of subsurface oceanic stratification, which may also influence TCs. Changes in oceanic stratification indeed modulate the amplitude of TCs-induced cooling, and hence the negative feedback of air-sea interactions on the TC intensity. Here we use a dynamical downscaling approach that couples an axisymmetric TC model to a simple ocean model to quantify this interannual oceanic control on TC activity. We perform twin experiments with contrasted oceanic stratifications representative of interannual variability in each TC-prone region. While subsurface oceanic variations do not significantly affect the number of moderate (Category 3 or less) TCs, they do induce a 30% change of Category 5 TC-days globally, and a 70% change for TCs exceeding 85 m s−1. TCs in the western Pacific and the southwestern Indian Ocean are most sensitive to oceanic interannual variability (with a ∼10 m s−1 modulation of the intensity of strongest storms at low latitude), owing to large upper ocean variations in response to ENSO. These results imply that a representation of ocean stratification variability should benefit operational forecasts of intense TCs and the understanding of their climatic variability. [ABSTRACT FROM AUTHOR]

Net Arctic Ocean primary production (PP) is expected to increase over this century, due to less perennial sea ice and more available light, but could decrease depending on changes in nitrate (NO3) supply. Here Coupled Model Intercomparison Project Phase 5 simulations performed with 11 Earth System Models are analyzed in terms of PP, surface NO3, and sea ice coverage over 1900-2100. Whereas the mean model simulates reasonably well Arctic-integrated PP (511 TgC/yr, 1998-2005) and projects a mild 58 TgC/yr increase by 2080-2099 for the strongest climate change scenario, models do not agree on the sign of future PP change. However, similar mechanisms operate in all models. The perennial ice loss-driven increase in PP is in most models NO3-limited. The Arctic surface NO3 is decreasing over the 21st century (−2.3 ± 1 mmol/m3), associated with shoaling mixed layer and with decreasing NO3 in the nearby North Atlantic and Pacific waters. However, the intermodel spread in the degree of NO3 limitation is initially high, resulting from >1000 year spin-up simulations. This initial NO3 spread, combined with the trend, causes a large variation in the timing of oligotrophy onset-which directly controls the sign of future PP change. Virtually all models agree in the open ocean zones on more spatially integrated PP and less PP per unit area. The source of model uncertainty is located in the sea ice zone, where a subtle balance between light and nutrient limitations determines the PP change. Hence, it is argued that reducing uncertainty on present Arctic NO3 in the sea ice zone would render Arctic PP projections much more consistent. [ABSTRACT FROM AUTHOR]