Publications

How a pair of pollutant parcels deviates from each other with an initial separation distance r₀, known as relative dispersion or Richardson dispersion, is relevant in many circumstances. This study examines the spatiotemporal similarity of the Richardson relative dispersion in the Gulf of Mexico by reanalyzing the Lagrangian trajectory of the surface drifter provided by two famous field experiments, that is, the Grand Lagrangian Deployment and the Lagrangian Submesoscale Experiment. The experimental dispersion curve indicates a critical separation time. When above this critical time, the dispersion shows an asymptotic power law growth independent of the initial separation distance r₀. Below it, the dispersion curve shows a strong spatiotemporal dependence with two spatiotemporal similarity regimes that can be identified for both experiments by looking at the isoline of the normalized dispersion curve. A new similarity variable is introduced to successfully collapse measured dispersion curves. However, the observed spatiotemporal similarity cannot be reproduced by the submesoscale preserved model. Thus, our results suggest that small-scale fluctuations play a crucial role in the relative dispersion of oceanic flows.

Abstract. This study investigates the use of a masked autoencoder (MAE) to address the challenge of filling gaps in high-resolution (1 km) sea surface temperature (SST) fields caused by cloud cover, which often result in gaps in the SST data and/or blurry imagery in blended SST products. Our study demonstrates that MAE, a deep learning model, can efficiently learn the anisotropic nature of small-scale ocean fronts from numerical simulations and reconstruct the artificially masked SST images. The MAE model is trained and evaluated on synthetic SST fields and tested on real satellite SST data from the Visible Infrared Imaging Radiometer Suite (VIIRS) sensor on the Suomi NPP satellite. We demonstrate that the MAE model trained on numerical simulations can provide a computationally efficient alternative for filling gaps in satellite SST. MAE can reconstruct randomly occluded images with a root mean square error (RMSE) of under 0.2 °C for masking ratios of up to 80 %. A trained MAE model in inference mode is exceptionally efficient, requiring 3 orders of magnitude (approximately 5000×) less time compared to the conventional approaches of cubic radial basis interpolation and Kriging tested on a single CPU. The ability to reconstruct high-resolution SST fields under cloud cover has important implications for understanding and predicting global and regional climates and detecting small-scale SST fronts that play a crucial role in the exchange of heat, carbon, and nutrients between the ocean surface and deeper layers. Our findings highlight the potential of deep learning models such as MAE to improve the accuracy and resolution of SST data at kilometer scales. This presents a promising avenue for future research in the field of small-scale ocean remote sensing analyses.

Energetic mesoscale eddies are often accompanied by strong submesoscale variability, which plays a significant role in connecting mesoscale and turbulent motions in the ocean and leads to strong vertical motions. The product of a high-resolution (1/48°) oceanic numerical model, the LLC4320, is employed to investigate the seasonal variations of vertical heat transport induced by submesoscale processes within multiple mesoscale eddies in the Kuroshio Extension (KE) region. In different seasons, the submesoscale vertical heat transport exhibits a consistent upward pattern, with notably higher magnitudes observed during winter. In winter, the maxima value of submesoscale vertical heat flux (SVHF) can account for approximately 60% of the total vertical heat flux (VHF). This is equivalent to the average net sea surface heat flux in a single eddy region. In summer and autumn, the maxima absolute value of submesoscale vertical heat flux can account for approximately 30% of the total VHF. Energy analysis reveals that baroclinic instability associated with vertical buoyancy flux has a crucial effect on generating submesoscale processes within the eddy region. The submesoscale motions are influenced by the mixed layer instability, strain-induced frontogenesis, turbulent thermal wind and turbulent thermal wind-induced frontogenesis within the upper mixed layer, while they are largely associated with the strain-induced frontogenesis in the ocean interior. Furthermore, the upward low-frequency submesoscale vertical heat transport is generated by submesoscale secondary circulation at eddy peripheries.

Fronts in the Bay of Bengal (BoB) are active and can potentially impact the regional dynamics such as temperature variability, salinity distribution and oceanic circulation. Based on the high resolution model output (LLC4320), this study investigates the characteristics of submesoscale fronts in the northern BoB and associated compensation/reinforcement effects. At sea surface, horizontal gradients of salinity and density are remarkable in the northern BoB, and they are nearly 3 times larger than temperature gradients. As the depth deepens, temperature gradients increase and become comparable to salinity gradients, while density gradients decrease a lot due to the increasing effects of compensation at subsurface. Statistical results show the dominance of salinity-controlled fronts over temperature-controlled fronts, and compensated fronts over reinforced fronts. The surface cooling/heating results in significant temporal variation of compensation at surface, but this variation is limited at subsurface by the blocking of the mixed layer base. The submesoscale-selective feature of compensation is much more pronounced at subsurface layer than surface layer. From statistical analysis and idealized numerical model, we found the slump of salinity-controlled compensated fronts are important in generating temperature inversion and maintaining barrier layer. This study validates the compensation theories originating from observations, and further illustrates the importance of subsurface compensated fronts using spatially continuous, regionally extended and longer-term model output. The subsurface-intensified submesoscale-selective compensation is proved for the first time in this study.

We investigate the influence of tides on the exchange of water between the Arabian Gulf and the Sea of Oman through the Strait of Hormuz using a high-resolution numerical model. Two numerical simulations are contrasted, one with and one without tidal forcing. We find that tides suppress exchange through the Strait, by ∼20% in the annual mean, being largest in the summer (∼30%) and diminishing in the winter (∼13%). Tides enhance the parameterised shear-driven vertical mixing inside the Gulf and Strait, mixing warm, relatively fresh surface waters downward thus reducing the density of bottom waters flowing outwards. This reduces the lateral difference of density between Gulf and Sea of Oman and hence the exchange through the Strait. Maximum reductions occur in summer when both the vertical stratification and mixing is the largest.

We used NASA's high-resolution (1/48° or 2.3 km, hourly) Estimating the Circulation and Climate of the Ocean (ECCO) estimates of salinity at a 1 m depth from November 2011 to October 2012 to detect semi-diurnal and diurnal internal tides (ITs) in the Andaman Sea and determine their characteristics in three 2° × 2° boxes off the Myanmar coast (box A), central Andaman Sea (box B), and off the Thailand coast (box C). We also used observed salinity and temperature data for the above period at the BD12-moored buoy in the central Andaman Sea. ECCO salinity data were bandpass-filtered with 11-14 h and 22-26 h periods. Large variations in filtered ECCO salinity (~0.1 psu) in the boxes corresponded with near-surface imprints of propagating ITs. Observed data from the box B domain reveals strong salinity stratification (halocline) in the upper 40 m. Our analyses reveal that the shallow halocline affects the signatures of propagating semi-diurnal ITs reaching the surface, but diurnal ITs propagating in the halocline reach up to the surface and bring variability in ECCO salinity. In box A, the semi-diurnal IT characteristics are higher speeds (0.96 m/s) with larger wavelengths (45 km), that are closer to theoretical mode 2 estimates, but the diurnal ITs propagating in the box A domain, with a possible source over the shelf of Gulf of Martaban, attain lower values (0.45 m/s, 38 km). In box B, the propagation speed is lower (higher) for semi-diurnal (diurnal) ITs. Estimates for box C are closer to those for box A.

Motivated by the importance of mixing arising from dissipating internal waves (IWs), vertical profiles of internal-wave dissipation from a high-resolution regional ocean model are compared with finestructure estimates made from observations. A horizontal viscosity scheme restricted to only act on horizontally rotational modes (such as eddies) is introduced and tested. At lower resolutions with horizontal grid spacings of 2 km, the modeled IW dissipation from numerical model agrees reasonably well with observations in some cases when the restricted form of horizontal viscosity is used. This suggests the possibility that if restricted forms of horizontal viscosity are adopted by global models with similar resolutions, they could be used to diagnose and map IW dissipation distributions. At higher resolutions with horizontal grid spacings of ∼250 m, the dissipation from vertical shear and horizontal viscosity act much more strongly resulting in dissipation overestimates; however, the vertical-shear dissipation itself is found to agree well with observations.

Mesoscale currents account for 80% of the ocean's kinetic energy, whereas submesoscale currents capture 50% of the vertical velocity variance. SWOT's first sea surface height (SSH) observations have a spatial resolution an order of magnitude greater than traditional nadir-looking altimeters and capture mesoscale and submesoscale features. This enables the derivation of submesoscale vertical velocities, crucial for the vertical transport of heat, carbon and nutrients between the ocean interior and the surface. This work focuses on a mesoscale energetic region south of Tasmania using a coupled ocean-atmosphere simulation at km-scale resolution and preliminary SWOT SSH observations. Vertical velocities (w), temperature anomalies and vertical heat fluxes (VHF) from the surface down to 1,000 m are reconstructed using effective surface Quasi-Geostrophic (sQG) theory. An independent method for reconstructing temperature anomalies, mimicking an operational gridded product, is also developed. Results show that sQG reconstructs 90% of the modeled w and VHF rms at scales down to 30 km just below the mixed layer and 50%-70% of the rms for scales larger than 70 km at greater depth, with a spatial correlation of ~0.6. The reconstruction is spectrally coherent (>0.65) for scales larger than 30-40 km at the surface, slightly degrading (~0.55) at depth. Two temperature anomaly data sets yield similar results, indicating the dominance of w on VHF. The RMS of sQG and VHF derived from SWOT are twice as large as those derived from conventional altimetry, highlighting the potential of SWOT for reconstructing energetic meso and submesoscale dynamics in the ocean interior.

In this study, we conducted a novel massive Lagrangian simulation experiment based on a global 1/48° tide-resolving numerical simulation of the ocean circulation. This first-time twin experiment enables a comparison between Eulerian (fixed-point) and Lagrangian (along-flow) estimates of kinetic energy (KE) across the global ocean, and the quantification of systematic differences between both types of estimations. This comparison represents an important step forward for the mapping of upper ocean high-frequency variability from Lagrangian observations of the Global Drifter Program. Eulerian KE rotary frequency spectra and band-integrated energy levels (e.g., tidal and near-inertial) serve as references and are compared to Lagrangian estimates. Our analysis reveals that, except for the near-inertial band, Lagrangian velocity spectra are systematically smoother, for example, with wider and lower spectral peaks compared to Eulerian counterparts. On average, Lagrangian KE levels derived from spectral band integrations tend to underestimate Eulerian KE levels at low-frequency and tidal bands, especially in regions with strong low-frequency KE. Better agreement between Lagrangian and Eulerian low-frequency and tidal KE levels is generally found in regions with weak low-frequency KE and/or convergent surface circulation, where Lagrangian particles tend to accumulate. Conversely, Lagrangian and Eulerian near-inertial spectra and energy levels are comparable. Our results demonstrate that Lagrangian estimates may provide a distorted view of low-frequency and tidal variance. To accurately map near-surface velocity climatology at these frequencies from drifter database, conversion methods accounting for the Lagrangian bias need to be developed.

While much attention has been paid to the climatic controls of species' range limits, other factors such as dispersal limitation are also important. Temperature is an important control of the distribution of coastal mangrove forests, and mangrove expansion at multiple poleward range limits has been linked to increasing temperatures. However, mangrove abundances at other poleward range limits have been surprisingly insensitive to climate change, indicating other drivers of range limitation. For example, along the west coast of North America, the poleward mangrove range limits are found on the Baja California and mainland coasts of Mexico, between 26°48' and 30°18'N. Non-climatic factors may play an important role in setting these range limits as 1) the abundance of range limit populations has been relatively insensitive to climate variability and 2) an introduced population of mangroves has persisted hundreds of kilometers north of the natural range limits. We combined a species distribution model with a high-resolution oceanographic transport model to identify the roles of climate and dispersal limitation in controlling mangrove distributions. We identified estuarine habitat that is likely climatically suitable for mangroves north of the current range limits. However, propagules from current mangrove populations are unlikely to reach these suitable locations due to prevailing ocean currents and geomorphic factors that create a patchy distribution of estuarine habitat with large between-patch distances. Thus, although climate change is driving range shifts of mangroves in multiple regions around the world, dispersal is currently limiting poleward mangrove expansion at several range limits, including the west coast of North America.

The Kuroshio Loop Current (KLC) is an important form of Kuroshio intrusion into the northeastern South China Sea (NESCS), which has significant influences on dynamical and biogeochemical processes in the NESCS. Recent studies suggested that the KLC is a hot spot of submesoscale processes (submesoscales) with spatiotemporal scales of O(1-10) km and O(1-10) days, but submesoscales' roles in energy cascade and salt and heat transports remain obscure. Here, we investigate this issue through analyzing outputs from a 1/48° simulation. The kinetic energy exchange rate between submesoscale and larger-scale processes (KER) is overall positive in the KLC region, which suggests the dominance of forward cascade. The magnitude of KER is comparable with the temporal change rate of larger-scale kinetic energy in the upper 200 m. We also find that magnitude and direction of KER are closely associated with strain rate and horizontal divergence of background flows, respectively. In addition, the KLC region shows elevated submesoscale salinity and heat diffusivities with magnitudes reaching O(10²) m²  s⁻¹ . During the KLC period, horizontal mixing by submesoscales can transport 0.90 × 10¹³  kg salt and 0.71 × 10²⁰  J heat westward into the NESCS interior, which are an order of magnitude larger than those caused by the mesoscale eddy shedding from the KLC. These results suggest that submesoscales play important roles not only in energy cascade but also in salt and heat transports in the KLC region. Therefore, the roles of submesoscales should be taken into account when studying energy, salt, and heat budgets in the NESCS.

Clouds and other data artefacts frequently limit the retrieval of key variables from remotely sensed Earth observations. We train a natural language processing (NLP)-inspired algorithm with high-fidelity ocean simulations to accurately reconstruct masked or missing data in sea surface temperature (SST) fields-one of 54 essential climate variables identified by the Global Climate Observing System. We demonstrate that the resulting model, referred to as Enki, repeatedly outperforms previously adopted inpainting techniques by up to an order of magnitude in reconstruction error, while displaying exceptional performance even in circumstances where the majority of pixels are masked. Furthermore, experiments on real infrared sensor data with masked percentages of at least 40% show reconstruction errors of less than the known uncertainty of this sensor (root mean square error (RMSE) ≲0.1 K). We attribute Enki's success to the attentive nature of NLP combined with realistic SST model outputs-an approach that could be extended to other remotely sensed variables. This study demonstrates that systems built upon Enki-or other advanced systems like it-may therefore yield the optimal solution to mitigating masked pixels in in climate-critical ocean datasets sampling a rapidly changing Earth.

Turbulent mixing in the ocean exerts an important control on the rate and structure of the overturning circulation. However, the balance of processes underpinning this mixing is subject to significant uncertainties, limiting our understanding of the overturning's deep upwelling limb. Here, we investigate the hitherto primarily neglected role of tens of thousands of seamounts in sustaining deep-ocean upwelling. Dynamical theory indicates that seamounts may stir and mix deep waters by generating lee waves and topographic wake vortices. At low latitudes, stirring and mixing are predicted to be enhanced by a layered vortex regime in the wakes. Using three realistic regional simulations spanning equatorial to middle latitudes, we show that layered wake vortices and elevated mixing are widespread around seamounts. We identify scalings that relate mixing rate within seamount wakes to topographic and hydrographic parameters. We then apply such scalings to a global seamount dataset and an ocean climatology to show that seamount-generated mixing makes an important contribution to the upwelling of deep waters. Our work thus brings seamounts to the fore of the deep-ocean mixing problem and urges observational, theoretical, and modeling efforts toward incorporating the seamounts' mixing effects in conceptual and numerical ocean circulation models.

The Pine Island and Thwaites Ice Shelves (PIIS/TIS) in the Amundsen Sea are melting rapidly and impacting global sea levels. The thermocline depth (TD) variability, the interface between cold Winter Water and warm modified Circumpolar Deep Water (mCDW), at the PIIS/TIS front strongly correlates with basal melt rates, but the drivers of its interannual variability remain uncertain. Here, using an ocean model, we propose that the strength of the eastern Amundsen Sea on-shelf circulation primarily controls TD variability and consequent PIIS/TIS melt rates. The TD variability occurs because the on-shelf circulation meanders following the submarine glacial trough, creating vertical velocity through bottom Ekman dynamics. We suggest that a strong or weak ocean circulation, possibly linked to remote winds in the Bellingshausen Sea, generates corresponding changes in bottom Ekman convergence, which modulates mCDW upwelling and TD variability. We show that interannual variability of off-shelf zonal winds has a minor effect on ocean heat intrusion into PIIS/TIS cavities, contrary to the widely accepted concept.

The South China Sea (SCS) is the largest marginal sea in the northwestern Pacific, and it is known for its complex multiscale dynamic processes, including basin-scale circulations, mesoscale eddies, submesoscale processes (submesoscales), and small-scale internal gravity waves. Compared with dynamic processes of other scales, submesoscales are a relatively new dynamic concept; they have gained rapidly increasing attention in recent decades due to their uniquely important roles in oceanic dynamics and biogeochemistry. Considerable progress on submesoscales has been achieved by the SCS regional oceanography community due to improvements in observation and simulation capabilities in the past decade. This paper comprehensively reviews recent research advances on the dynamic aspects of submesoscales in the SCS, including submesoscale resolving/permitting observations and simulations; the general characteristics, spatiotemporal variations, and generation mechanisms of submesoscales; and the roles of submesoscales in energy cascade and vertical tracer transport and the associated parameterizations. The most important advances are as follows: (a) Novel submesoscale observations have been made in the SCS, such as through submesoscale and mesoscale nested mooring arrays. (b) Findings have shown that the spatiotemporal characteristics and generation mechanisms of submesoscales in the SCS are regionally dependent. (c) A generation mechanism called mixed transitional layer instability (MTI) was proposed, and its strength is significantly modulated by strain-induced frontogenesis. (d) A new parameterization of submesoscale vertical buoyancy flux was developed based on the mechanism of MTI modulated by frontogenesis. In addition to reviewing recent advances in this field, this paper presents research prospects on SCS submesoscales.

Extracting balanced geostrophic motions (BM) from sea surface height (SSH) observations obtained by wide-swath altimetry holds great significance in enhancing our understanding of oceanic dynamic processes at submesoscale wavelength. However, SSH observations derived from wide-swath altimetry are characterized by high spatial resolution while relatively low temporal resolution, thereby posing challenges to extract the BM from a single SSH snapshot. To address this issue, this paper proposes a deep learning model called the BM-UBM Network, which takes an instantaneous SSH snapshot as input and outputs the projection corresponding to the BM. Training experiments are conducted both in the Gulf Stream and South China Sea, and three metrics are considered to diagnose model's outputs. The favorable results highlight the potential capability of the BM-UBM Network to process SSH measurements obtained by wide-swath altimetry.

The ice shelves in the Bellingshausen Sea are melting and thinning rapidly due to modified Circumpolar Deep Water (mCDW) intrusions carrying heat toward ice-shelf cavities. Observations are, however, sparse in time and space, and extensive model-data comparisons have never been possible. Here, using a circulation model of the region and ship-based observations, we show that the simulated water mass distributions in several troughs traversing mCDW inflows are in good agreement with observations, implying that our model has the skills to simulate hydrographic structures as well as on-shelf ocean circulations. It takes 7.9 and 11.7 months for mCDW to travel to the George VI Ice Shelf cavities through the Belgica and Marguerite troughs, respectively. Ice-shelf melting is mainly caused by mCDW intrusions along the Belgica and Marguerite troughs, with the heat transport through the former being ∼2.8 times larger than that through the latter. The mCDW intrusions toward the George VI Ice Shelf show little seasonal variability, while those toward the Venable Ice Shelf show seasonal variability, with higher velocities in summer likely caused by coastal trapped waves. We also conduct particle experiments tracking glacial meltwater. After 2 years of model integration, ∼33% of the released particles are located in the Amundsen Sea, supporting a linkage between Bellingshausen Sea ice-shelf meltwater and Amundsen Sea upper ocean hydrography.

We investigate the interaction between an anticyclonic eddy (AE) and semidiurnal internal tide (SIT) on the continental slope of the northeastern South China Sea (SCS), using a high spatiotemporal resolution numerical model. Two key findings are as follows: first, the AE promotes energy conversion from low-mode to higher-mode SIT. Additionally, production terms indicate that energy is also transferred from the SIT field to the eddy field at an average rate of 3.0 mW m⁻² (accounting for 7 % of the incoming energy flux of SIT when integrated over the eddy diameter). Second, the AE can modify the spatial distribution of tidal-induced dissipation by refracting, scattering, and reflecting low-mode SIT. The phase and group velocities of the SIT are significantly influenced by the eddy field, resulting in a northward or southward shift in the internal tidal rays. These findings deepen our understanding of the complex interactions between AE and SIT, as well as their impacts on energy conversion, wave propagation, and coastal processes.

The Southern Ocean plays a major role in controlling the evolution of Antarctic glaciers and in turn their impact on sea level rise. We present the Southern Ocean high-resolution (SOhi) simulation of the MITgcm ocean model to reproduce ice-ocean interaction at 1/24° around Antarctica, including all ice shelf cavities and oceanic tides. We evaluate the model accuracy on the continental shelf using Marine Mammals Exploring the Oceans Pole to Pole data and compare the results with three other MITgcm ocean models (ECCO4, SOSE, and LLC4320) and the ISMIP6 temperature reconstruction. Below 400 m, all the models exhibit a warm bias on the continental shelf, but the bias is reduced in the high-resolution simulations. We hypothesize some of the bias is due to an overestimation of sea ice cover, which reduces heat loss to the atmosphere. Both high-resolution and accurate bathymetry are required to improve model accuracy around Antarctica.

Although eddy parameterization schemes are often based on the local assumption, previous studies indicate that the nonlocality of total eddy mixing is prevalent at the Kuroshio Extension (KE). For eddy-permitting climate models, only mixing induced by eddies smaller than the resolvable scale of climate models (L^*) needs to be parameterized. Therefore, here we aim to estimate and predict the nonlocality of scale-dependent eddy mixing at the KE region. We consider the separation scale L^* ranging from 0.2° - 2.5°, which is comparable to the typical resolution of the ocean component of climate models. Using a submesoscale-permitting model solution (MITgcm llc4320) and Lagrangian particles, we estimate the scale-dependent mixing (SDM) nonlocality ellipses and then diagnose the square root of the ellipse area (L_{n, particle}). L_{n, particle} is a metric to quantify the degree of SDM nonlocality. We found that, for all the available L^* values we consider, the SDM nonlocality is prevalent in the KE region, and mostly elevated values of L_{n, particle} occur within the KE jet. As L^* decreases from 2.5° to 0.2° , the ratio L_{n, particle}/L^* increases from 0.8 to 8.9. This result indicates that the SDM nonlocality is more non-negligible for smaller L^*, which corresponds to climate models with relatively high resolution. As to the SDM nonlocality prediction, we found that compared to the conventional scaling and the curve-fitting methods, the random forest approach can better represent L_{n, particle} , especially in the coastal regions and within the intense KE jet. The area of the Eulerian momentum ellipses well capture the spatial pattern, but not the magnitude, of L_{n, particle} . Our efforts suggest that eddy parameterization schemes for eddy-permitting models may be improved by taking into account mixing nonlocality.

For eddy-permitting climate models, only eddies smaller than the smallest resolvable scale need to be parameterized. Therefore, it is important to study the diffusivities induced by eddies smaller than a specific separation scale L^*, that is, the scale-dependent eddy diffusivities. Using a submesoscale-permitting model solution (MITgcm llc4320), we estimate the scale-dependent eddy diffusivity in the Kuroshio Extension. We find that, as the separation scale L^* increases, the diffusivity increases, and the spatial structure approaches that of the total eddy diffusivity. We quantify this scale dependence through fitting the diffusivity to L^*n. Our derivation shows that n is approximately (a + 1)/2, where a is the eddy kinetic energy spectral slope. For domain-averaged diffusivity, n is 1.33. We then extend four existing mixing theories by including scale dependence. Our results show that both of the theories designed for intense-jet regions, the suppressed mixing length theory and the multiwavenumber theory, closely match the magnitude of the scale-dependent diffusivity but fail to capture well the diffusivity's spatial structure. However, the other two theories based on eddy size and Rhines scale can reasonably represent the spatial structure. Based on this finding, we propose an empirical formula for scale-dependent eddy diffusivity that well represents both the magnitude and the spatial structure of the eddy diffusivity. Our work demonstrates that climate models should use scale-dependent diffusivity, and designing appropriate empirical formulas may be a reasonable approach to represent these scale-dependent diffusivities. Also, our diagnostic framework and theories for scale-dependent eddy diffusivity may be applicable to the global ocean.

Oceanic eddy-induced meridional heat transport (EHT) is an important process in the Southern Ocean heat budget, the variability of which significantly modulates global meridional overturning circulation (MOC) and Antarctic sea-ice extent. Although it is recognized that mesoscale eddies with scales of ~40-300 km greatly contribute to the EHT, the role of submesoscale eddies with scales of ~1-40 km remains unclear. Here, using two state-of-the-art high-resolution simulations (resolutions of 1/48° and 1/24°), we find that submesoscale eddies significantly enhance the total poleward EHT in the Southern Ocean with an enhancement percentage reaching 19-48% in the Antarctic Circumpolar Current band. By comparing the eddy energy budgets between the two simulations, we detect that the primary role of submesoscale eddies is to strengthen mesoscale eddies (and thus their heat transport capability) through inverse energy cascade rather than directly through submesoscale heat fluxes. Due to the submesoscale-mediated enhancement of mesoscale eddies in the 1/48° simulation, the clockwise upper cell and anti-clockwise lower cell of the residual-mean MOC in the Southern Ocean are weakened and strengthened, respectively. This finding identifies a potential route to improve the mesoscale parameterization in climate models for more accurate simulations of the MOC and sea ice variability in the Southern Ocean.

We have assembled 2 851 702 nearly cloud-free cutout images (sized 144 km x 144 km) of sea surface temperature (SST) data from the entire 2012-2020 Level-2 Visible Infrared Imaging Radiometer Suite (VIIRS) dataset to perform a quantitative comparison to the ocean model output from the MIT General Circulation Model (MITgcm). Specifically, we evaluate outputs from the LLC4320 (LLC, latitude-longitude-polar cap) 1/48^o global-ocean simulation for a 1-year period starting on 17 November 2011 but otherwise matched in geography and the day of the year to the VIIRS observations. In lieu of simple (e.g., mean, standard deviation) or complex (e.g., power spectrum) statistics, we analyze the cutouts of SST anomalies with an unsupervised probabilistic autoencoder (PAE) trained to learn the distribution of structures in SST anomaly (SSTa) on ~ 10-80 km scales (i.e., submesoscale to mesoscale). A principal finding is that the LLC4320 simulation reproduces, over a large fraction of the ocean, the observed distribution of SSTa patterns well, both globally and regionally. Globally, the medians of the structure distributions match to within 2σ for 65 % of the ocean, despite a modest, latitude-dependent offset. Regionally, the model outputs reproduce mesoscale variations in SSTa patterns revealed by the PAE in the VIIRS data, including subtle features imprinted by variations in bathymetry. We also identify significant differences in the distribution of SSTa patterns in several regions: (1) in an equatorial band equatorward of 15^o; (2) in the Antarctic Circumpolar Current (ACC), especially in the eastern half of the Indian Ocean; and (3) in the vicinity of the point at which western boundary currents separate from the continental margin. It is clear that region 3 is a result of premature separation in the simulated western boundary currents. The model output in region 2, the southern Indian Ocean, tends to predict more structure than observed, perhaps arising from a misrepresentation of the mixed layer or of energy dissipation and stirring in the simulation. The differences in region 1, the equatorial band, are also likely due to model errors, perhaps arising from the shortness of the simulation or from the lack of high-frequency and/or wavenumber atmospheric forcing. Although we do not yet know the exact causes for these model-data SSTa differences, we expect that this type of comparison will help guide future developments of high-resolution global-ocean simulations.

Melt rates of West Antarctic ice shelves in the Amundsen Sea track large decadal variations in the volume of warm water at their outlets. This variability is generally attributed to wind-driven variations in warm water transport toward ice shelves. Inspired by conceptual representations of the global overturning circulation, we introduce a simple model for the evolution of the thermocline, which caps the warm water layer at the ice-shelf front. This model demonstrates that interannual variations in coastal polynya buoyancy forcing can generate large decadal-scale thermocline depth variations, even when the supply of warm water from the shelf-break is fixed. The modeled variability involves transitions between bistable high and low melt regimes, enabled by feedbacks between basal melt rates and ice front stratification strength. Our simple model captures observed variations in near-coast thermocline depth and stratification strength, and poses an alternative mechanism for warm water volume changes to wind-driven theories.

Ice shelf basal melting is the primary mechanism driving mass loss from the Antarctic Ice Sheet, yet it is unknown how the localized melt enhancement from subglacial discharge will affect future Antarctic glacial retreat. We develop a parameterization of ice shelf basal melt that accounts for both ocean and subglacial discharge forcing and apply it in future projections of Denman and Scott Glaciers, East Antarctica, through 2300. In forward simulations, subglacial discharge accelerates the onset of retreat of these systems into the deepest continental trench on Earth by 25 years. During this retreat, Denman Glacier alone contributes 0.33 millimeters per year to global sea level rise, comparable to half of the contemporary sea level contribution of the entire Antarctic Ice Sheet. Our results stress the importance of resolving complex interactions between the ice, ocean, and subglacial environments in future Antarctic Ice Sheet projections.

The Surface Water and Ocean Topography (SWOT) satellite is expected to observe sea surface height (SSH) down to scales approaching ∼15 km, revealing submesoscale patterns that have never before been observed on global scales. Features at these soon-to-be-observed scales, however, are expected to be significantly influenced by internal gravity waves, fronts, and other ageostrophic processes, presenting a serious challenge for estimating surface velocities from SWOT observations. Here we show that a data-driven approach can be used to estimate the surface flow, particularly the kinematic signatures of smaller scale flows, from SSH observations, and that it performs significantly better than using the geostrophic relationship. We use a Convolutional Neural Network (CNN) trained on submesoscale-permitting high-resolution simulations to test the possibility of reconstructing surface vorticity, strain, and divergence from snapshots of SSH. By evaluating success using pointwise accuracy and vorticity-strain-divergence joint distributions, we show that the CNN works well when inertial gravity wave amplitudes are relatively weak. When the wave amplitudes are strong, reconstructions of vorticity and strain are less accurate; however, we find that the CNN naturally filters the wave-divergence, making divergence a surprisingly reliable field to reconstruct. We also show that when applied to realistic simulations, a CNN model pretrained with simpler simulation data performs well, indicating a possible path forward for estimating real flow statistics with limited observations.

The internal-wave (IW) continuum of a regional ocean model is studied in terms of the vertical spectral kinetic-energy (KE) fluxes and transfers at high vertical wavenumbers. Previous work has shown that this model permits a partial representation of the IW cascade. In this work, vertical spectral KE flux is decomposed into catalyst, source, and destination vertical modes and frequency bands of nonlinear scattering, a framework that allows for the discernment of different types of nonlinear interactions involving both waves and eddies. Energy transfer within the supertidal IW continuum is found to be strongly dependent on resolution. Specifically, at a horizontal grid spacing of 1/48°, most KE in the supertidal continuum arrives there from lower frequency modes through a single nonlinear interaction, while at 1/384° and with sufficient vertical resolution KE transfers within the supertidal IW continuum are comparable in size to KE transfer from lower-frequency modes. Additionally, comparisons are made with existing theoretical and observational work on energy pathways in the IW continuum. Induced diffusion (ID) is found to be associated with a weak forward frequency transfer within the supertidal IW continuum. ID is also limited to the highest vertical wavenumbers and is more sensitive to resolution relative to spectrally local interactions (LI). At the same time, ID-like processes involving high vertical-wavenumber near-inertial and tidal waves as well as low-vertical-wavenumber eddy fields are substantial, suggesting that the processes giving rise to a Garrett-Munk-like spectra in the present numerical simulation and perhaps the real ocean may be more varied than in idealized or wave-only frameworks.

Abstract. We investigate ocean dynamics at different scales in the Agulhas Current system, a region of important interocean exchange of heat and energy. While ocean observations and some of the most advanced climate models capture the larger mesoscale dynamics (> 100 km), the smaller-scale fronts and eddies are underrepresented. The recently launched NASA-CNES Surface Water and Ocean Topography (SWOT) wide-swath altimeter mission observes the smaller ocean geostrophic scales down to 15 km in wavelength globally. Here we will analyse different eddy diagnostics in the Agulhas Current region and quantify the contributions from the larger mesoscales observable today and the smaller scales to be observed with SWOT. Surface geostrophic diagnostics of eddy kinetic energy, strain, and energy cascades are estimated from modelled sea surface height (SSH) fields of the Massachusetts Institute of Technology general circulation model (MITgcm) latitude-longitude polar cap (LLC4320) simulation subsampled at 1/10?. In this region, the smaller scales (<150 km) have a strong signature on the horizontal geostrophic strain rate and for all eddy diagnostics in the Western Boundary Current and along the meandering Agulhas Extension. We investigate the horizontal cascade of energy using a coarse-graining technique, and we observe that the wavelength range where the inverse cascade occurs is biased towards larger mesoscale wavelengths with today's altimetric sampling. We also calculate the projected sampling of the eddy diagnostics under the SWOT swaths built with the NASA-CNES simulator to include the satellite position and realistic noise. For the swaths, a neural network noise mitigation method is implemented to reduce the residual SWOT random error before calculating eddy diagnostics. In terms of SSH, observable wavelengths of 15 to 20 km are retrieved after neural network noise mitigation, as opposed to wavelengths larger than 40 km before the noise reduction.

Surface ocean temperature and velocity anomalies at meso- and sub-meso-scales induce wind stress anomalies. These wind-front interactions, referred to as thermal (TFB) and current (CFB) feedbacks, respectively, have been studied in isolation at mesoscale, yet they have rarely been considered in tandem. Here, we assess the combined influence of TFB and CFB and their relative impact on surface wind stress derivatives. Analyses are based on output from two regions of the Southern Ocean in a coupled simulation with local ocean resolution of 2 km. Considering both TFB and CFB shows regimes of interference, which remain mostly linear down to the simulation resolution. The jointly-generated wind stress curl anomalies approach 10⁻⁵  Nm⁻³ , ∼20 times stronger than at mesoscale. The synergy of both feedbacks improves the ability to reconstruct wind stress curl magnitude and structure from both surface vorticity and SST gradients by 12%-37% on average, compared with using either feedback alone.

Near Inertial Oscillations (NIOs) are ocean oscillations forced by intermittent winds. They are most energetic at mid-latitudes, particularly in regions with atmospheric storm tracks. Wind-driven, large-scale NIOs are quickly scattered by ocean mesoscale eddies (with sizes ranging from 100 to 400 km), causing a significant portion of the NIO energy to propagate into the subsurface ocean interior. This kinetic energy pathway illustrates that the wind energy input to NIO is critical for maintaining deep ocean stratification and thus closing the total energy budget, as emphasised by numerous modelling studies. However, this wind energy input to NIO remains poorly observed on a global scale. A remote sensing approach that observes winds and ocean currents co-located in time and space with high resolution is necessary to capture the intermittent air-sea coupling. The current satellite observations do not meet these requirements. This study assesses the potential of a new satellite mission concept, Ocean DYnamics and Surface Exchange with the Atmosphere (OSYSEA), to recover wind-forced NIOs from co-located winds and currents. To do this, we use an Observation System Simulation Experiment (OSSE) based on hourly observations of ocean surface currents and surface winds from five surface moorings covering latitudes from 15° to 50°. ODYSEA wind and current observations are expected to have a spatial resolution of 10 km with about a 12 h sampling frequency in mid-latitudes. Results show that NIOs can be recovered with high accuracy using the ODYSEA spatial and temporal resolution, but only if observations are made over a wide area of 1800 km. A narrower swath (1000 km) may lead to significant aliasing.

Recent simulations suggest that submesoscale motions with scales smaller than 30 km and frequencies greater than 1 day⁻¹ drive upward vertical heat transport. These simulations have prompted us to revisit the mechanisms that explain high-frequency (HF) vertical heat fluxes (VHFs) within the surface mixed layer (ML). Here, an idealized numerical simulation of a re-entrant channel flow with an unbalanced submesoscale thermal front is used to analyze the impact of surface cooling on HF VHFs. Two types of simulations are analyzed: forced and unforced. The VHFs cospectrum analysis shows that surface diurnal cooling increases VHFs, reaching frequencies larger than 1 day⁻¹ . However, the fastest-growing length scale of ML instabilities limits the extension of positive VHFs toward fine scales. Symmetric and gravitational instabilities are the main conduits producing ageostrophic HF and small-scale structures, which in turn enhance upward VHFs across the diurnal frequency. A comparison between forced-idealized simulations with the K-profile parameterization scheme and a realistic regional simulation in the frequency-wavenumber space, reveals that the two simulation types reproduce similar VHFs near the diurnal frequency. However, the realistic simulation displays higher VHFs than the forced-idealized simulation. This study emphasizes that surface diurnal cooling significantly impacts HF VHFs. However, this impact is not sufficient to reach the HF VHFs estimated in realistic submesoscale-permitting and tidal-resolving simulations.

Internal waves (IWs) are generated in all the oceans, and their amplitudes are large, especially in regions that receive a large amount of freshwater from nearby rivers, which promote highly stratified waters. When barotropic tides encounter regions of shallow bottom-topography, internal tides (known as IWs of the tidal period) are generated and propagated along the pycnocline due to halocline or thermocline. In the North Indian Ocean, the Bay of Bengal (BoB) and the Andaman Sea receive a large volume of freshwater from major rivers and net precipitation during the summer monsoon. This study addresses the characteristics of internal tides in the BoB and Andaman Sea using NASA's Estimating the Circulation and Climate of the Ocean (ECCO) project's high-resolution (1/48° and hourly) salinity estimates at 1 m depth (hereafter written as ECCO salinity) during September 2011-October 2012, time series of temperature, and salinity profiles from moored buoys. A comparison is made between ECCO salinity and NASA's Soil Moisture Active Passive (SMAP) salinity and Aquarius salinity. The time series of ECCO salinity and observed salinity are subjected to bandpass filtering with an 11-14 h period and 22-26 h period to detect and estimate the characteristics of semi-diurnal and diurnal period internal tides. Our analysis reveals that the ECCO salinity captured well the surface imprints of diurnal period internal tide propagating through shallow pycnocline (~50 m depth) due to halocline, and the latter suppresses the impact of semi-diurnal period internal tide propagating at thermocline (~100 m depth) reaching the sea surface. The semi-diurnal (diurnal) period internal tides have their wavelengths and phase speeds increased (decreased) from the central Andaman Sea to the Sri Lanka coast. Propagation of diurnal period internal tide is dominant in the northern BoB and northern Andaman Sea.

Meltwater content and pathways determine the impact of Antarctica's melting ice shelves on ocean circulation and climate. Using ocean glider observations, we quantify meltwater distribution and transport within the Bellingshausen Sea's Belgica Trough. Meltwater is present at different densities and with different turbidities: both are indicative of a layer's ice shelf of origin. To investigate how ice-shelf origin separates meltwater into different export pathways, we compare these observations with high-resolution tracer-release model simulations. Meltwater filaments branch off the Antarctic Coastal Current into the southwestern trough. Meltwater also enters the Belgica Trough in the northwest via an extended western pathway, hence the greater observed southward (0.50 mSv) than northward (0.17 mSv) meltwater transport. Together, the observations and simulations reveal meltwater retention within a cyclonic in-trough gyre, which has the potential to promote climactically important feedbacks on circulation and future melting.

The kinetic energy transfer between the atmosphere and oceans, called wind work, affects ocean dynamics, including near-inertial oscillations and internal gravity waves, mesoscale eddies, and large-scale zonal jets. For the most part, the recent numerical estimates of global wind work amplitude are almost five times larger than those reported 10 years ago. This large increase is explained by the impact of the broad range of spatial and temporal scales covered by winds and currents, the smallest of which has only recently been uncovered by increasingly high-resolution modeling efforts. However, existing satellite observations do not fully sample this broad range of scales. The present study assesses the capabilities of ODYSEA, a conceptual satellite mission to estimate the amplitude of wind work in the global ocean. To this end, we use an ODYSEA measurement simulator fed by the outputs of a km scale coupled ocean-atmosphere model to estimate wind work globally. The results indicate that compared with numerical truth estimates, the ODYSEA instrument performs well globally, except for latitudes north of 40^oN during summer due to unresolved storm evolution. This performance is explained by the wide-swath properties of ODYSEA (a 1700 km wide swath with 5 km posting for winds and surface currents), its twice-a-day (daily) coverage at mid-latitudes (low latitudes), and the insensitivity of the wind work to uncorrelated errors in the estimated wind and current.

This study examines the process of water-column stratification breakdown in Antarctic coastal polynyas adjacent to an ice shelf with a cavity underneath. This first part of a two-part sequence seeks to quantify the influence of offshore katabatic winds, alongshore winds, air temperature, and initial ambient stratification on the timescales of polynya destratification through combining process-oriented numerical simulations and analytical scaling. In particular, the often-neglected influence of wind-driven circulation on the lateral transport of the water formed at the polynya surface - which we call Polynya Source Water (PSW) - is systematically examined here. First, an ice shelf-sea ice-ocean coupled numerical model is adapted to simulate the process of PSW formation in polynyas of various configurations. The simulations highlight that i) before reaching the bottom, majority of the PSW is actually carried away from the polynya by katabatic wind-induced offshore outflow, diminishing water-column mixing in the polynya and intrusion of the PSW into the neighboring ice shelf cavity, and ii) alongshore coastal easterly winds, through inducing onshore Ekman transport, reduce offshore loss of the PSW and enhance polynya mixing and PSW intrusion into the cavity. Second, an analytical scaling of the destratification timescale is derived based on fundamental physical principles to quantitatively synthesize the influence of the physical factors, which is then verified by independent numerical sensitivity simulations. This work provides insights into the mechanisms that drive temporal and cross-polynya variations in stratification and PSW formation in Antarctic coastal polynyas, and establishes a framework for studying differences among the polynyas in the ocean.

Submesoscale oceanic structures (<10-20 km) such as eddies and fronts are often difficult to describe given the influence of the mesoscale. In order to characterize the surface signatures of submesoscale structures, we utilize a custom spatial filtering function to separate the meso- and large-scale sea surface height (SSH) signal from the small scale SSH signal of 1/48° high resolution estimates provided by NASA's Estimating the Circulation and Climate of the Oceans (ECCO) project. In this study, we use ECCO estimates from a 14-month global simulation between September 2011 and November 2012 with a 2 km horizontal grid spacing in the Gulf of Mexico. We then use an eddy detection and tracking algorithm to identify persistent circular features on both scales, giving rise to an atlas of submesoscale eddy-like variabilities (SEVs). We briefly investigate the geographic and temporal variability of SEVs as a whole before collocating SEVs inside mesoscale eddies, allowing us to evaluate the characteristics of internal SEVs and the impact of SEVs on mesoscale eddies. We find that SEVs, both anticyclonic and cyclonic, are ubiquitous inside mesoscale eddies with lifetimes longer than a week, accounting for an average of 10-20% of the spatial area and eddy kinetic energy of mesoscale eddies. We also show that internal SEVs are persistently associated with temperature and salinity anomalies in both eddy centers and edges of up to 0.1 °C and 0.05 psu, with anticyclonic internal SEVs being warmer and fresher while cyclonic internal SEVs are colder and saltier. Finally, we examine the life cycle of an anticyclonic Loop Current eddy, demonstrating that the number and intensity of internal SEVs within increases as the eddy approaches separation from the Loop Current until a maximum is obtained just after separation. In light of forthcoming submesoscale SSH observations from NASA's Surface Water and Ocean Topography (SWOT) mission, our results showcase the variability of submesoscale eddy structures and their possible implications for biogeochemical cycling, the inverse energy cascade, and Loop Current prediction techniques.

Oceanic submesoscale flows are considered to be a crucial conduit for the downscale transfer of oceanic mesoscale kinetic energy and upper-ocean material exchange, both laterally and vertically, but defining observations revealing submesoscale dynamics and/or transport properties remain sparse. Here, we report on an elaborate observation of a warm and fresh filament intruding into a cyclonic mesoscale eddy. By integrating cruise measurements, satellite observations, particle-tracking simulations, and the trajectory of a surface drifter, we show that the filament originated from an anticyclonic eddy immediately to the west of the cyclonic eddy, and the evolution of the filament was mainly due to the geostrophic flows associated with the eddy pair. Our observations reveal the mass exchange of the eddy pair and suggest that submesoscale flows can degrade the coherence of mesoscale eddies, providing important implications for the transport properties of mesoscale eddies. Vigorous submesoscale turbulence was found within the eddy core region, due to filamentous intrusion and frontogenesis. Our findings have thus offered novel insights into the dynamics and transport properties of oceanic submesoscale flows, which should be taken into account in their simulation and parameterization in ocean and climate models.

Results of coarse-resolution climate models are sensitive to the specification of ocean eddy mixing coefficients. Therefore, it is important to estimate, rationalize and predict eddy diffusivities. Here, we estimate the seasonal variability of surface eddy diffusivities in the Kuroshio Extension region using numerical particles advected by a submesoscale-permitting model solution. We find that both the spatial structure and the domain-averaged value of the particle-based eddy diffusivities have a significant seasonal cycle. We also assess the predictability of cross-stream mixing lengths in this region using the methods of machine learning, suppressed mixing length theory (SMLT), and multiple linear regression (LR). The predictors we choose are all variables from SMLT that represent eddy- and mean-flow properties, and these predictors correlate well with the particle-based cross-stream mixing lengths. We demonstrate that, compared to SMLT and LR, machine learning methods, in particular the random forest (RF) and convolutional neural network (CNN), can better represent both the spatial structure and the domain-averaged value of cross-stream mixing lengths. The skill in predicting the mixing lengths with CNN has much less seasonal variability than that with RF. Our results indicate that the machine learning approach may be useful in future development of eddy parameterization schemes.

The submesoscale energetics of the eastern Gulf of Mexico (GoM) are diagnosed using outputs from a 1/48° MITgcm simulation. Employed is a recently developed, localized multiscale energetics formalism with three temporal-scale ranges (or scale windows), namely, a background flow window, a mesoscale window, and a submesoscale window. It is found that the energy cascades are highly inhomogeneous in space. Over the eastern continental slope of the Campeche Bank, the submesoscale eddies are generated via barotropic instability, with forward cascades of kinetic energy (KE) following a weak seasonal variation. In the deep basin of the eastern GoM, the submesoscale KE exhibits a seasonal cycle, peaking in winter, maintained via baroclinic instability, with forward available potential energy (APE) cascades in the mixed layer, followed by a strong buoyancy conversion. A spatially coherent pool of inverse KE cascade is found to extract energy from the submesoscale KE reservoir in this region to replenish the background flow. The northern GoM features the strongest submesoscale signals with a similar seasonality as seen in the deep basin. The dominant source for the submesoscale KE during winter is from buoyancy conversion and also from the forward KE cascades from mesoscale processes. To maintain the balance, the excess submesoscale KE must be dissipated by smaller-scale processes via a forward cascade, implying a direct route to finescale dissipation. Our results highlight that the role of submesoscale turbulence in the ocean energy cycle is region and time dependent.

Symmetric instability (SI) extracts kinetic energy from fronts in the surface mixed layer (SML), potentially affecting the SML structure and dynamics. Here, a global submesoscale-permitting ocean model named MITgcm LLC4320 simulation is used to examine the Stone linear prediction of the maximum SI scale to estimate grid spacings needed to begin resolving SI. Furthermore, potential effects of SI on the usable wind work are estimated roughly: this estimate of SI "activity" is useful for assessing if these modes should be resolved or parameterized. The maximum SI scale varies by latitude with median values from 568 to 23 m. Strong seasonality is observed in the SI scale and activity. The median scale in winter is 188 m globally, 2.5 times of that of summer (75 m). SI is more active in winter: 15% of the time compared with 6% in summer. The strongest SI activity is found in the western Pacific, western Atlantic, and Southern Oceans. The required grid spacings for a global model to begin resolving SI eddies in the SML are 24 m (50% of regions resolved) and 7.9 m (90%) in winter, decreasing to 9.4 m (50%) and 3.6 m (90%) in summer. It is also estimated that SI may reduce usable wind work by an upper bound of 0.83 mW m⁻² globally, or 5% of the global magnitude. The sensitivity of these estimates to empirical thresholds is provided in the text.

The spatial scale of submesoscales is an important parameter for studies of submesoscale dynamics and multiscale interactions. The horizontal spatial scales of baroclinic, geostrophic-branch mixed layer instabilities (MLI) are investigated globally (without the equatorial or Arctic oceans) based on observations and simulations in the surface and bottom mixed layers away from significant topography. Three high-vertical-resolution boundary layer schemes driven with profiles from a MITgcm global submesoscale-permitting model improve robustness. The fastest-growing MLI wavelength decreases toward the poles. The zonal median surface MLI wavelength is 51-2.9 km when estimated from the observations and from 32, 25, and 27 km to 2.5, 1.2, and 1.1 km under the K -profile parameterization (KPP), Mellor-Yamada (MY), and κ-ε schemes, respectively. The surface MLI wavelength has a strong seasonality with a median value 1.6 times smaller in summer (10 km) than winter (16 km) globally from the observations. The median bottom MLI wavelengths estimated from simulations are 2.1, 1.4, and 0.41 km globally under the KPP, MY, and κ-ε schemes, respectively, with little seasonality. The estimated required ocean model grid spacings to resolve wintertime surface mixed layer eddies are 1.9 km (50% of regions resolved) and 0.92 km (90%) globally. To resolve summertime eddies or MLI seasonality requires grids finer than 1.3 km (50%) and 0.55 km (90%). To resolve bottom mixed layer eddies, grids finer than 257, 178, and 51 m (50%) and 107, 87, and 17 m (90%) are estimated under the KPP, MY, and κ-ε schemes.

Submesoscale processes in the upper ocean vary seasonally, in tight correspondence with mixed layer thickness variability. Based on a global high-resolution MITgcm simulation, seasonal evaluation of strong vorticity and spectral analysis of the kinetic energy in the Kuroshio Extension System show the strongest submesoscales occur in March, implying a lag of about a month behind mixed layer thickness maximum in February. An analysis of spectral energy sources and transfers indicates that the seasonality of the submesoscale energy content is a result of the competition between the conversion of available potential energy into submesoscale kinetic energy via a buoyancy production/vertical buoyancy flux associated with mixed layer instability and nonlinear energy transfers to other scales associated with an energy cascade. The buoyancy production is seasonally in phase with the mixed layer depth, but the transfers of energy across scales makes energizing the reservoir of submesoscale kinetic energy lag behind by a month.