Submesoscales

Submesoscales are apparent as lateral density gradients and intense circulations throughout the upper ocean over horizontal scales of 1-10 km. They have very recently been suggested to critically affect the sensitivity of the coupled climate system by dramatically enhancing vertical fluxes of buoyancy, momentum and tracers above those due to air-sea interaction alone. By doing so, submesoscales are thus a potentially leading order process within the surface mixed layer (SML) through which the ocean communicates with the atmosphere but due to their small scale are currently unresolved in climate models. Current limitations on computing capability mean that submesoscales will remain sub-grid scale in global models for years to come, yet given their apparent importance to upper ocean processes and, for example, the transformation of water masses that occur therein, our skill in future climate predictions is clearly compromised by omitting their effects. The oceanographic community is therefore currently engaged in a wide range of research projects aimed at improving our understanding of submesoscales and how they influence the climate by modulating exchange between the ocean and atmosphere.

Projects

Mixed Layer 2004 (2004-2007)

Mixed Layer 2004 (ML04) aimed to identify the details of mixing and restratification at small scales within the surface mixed (SML) at the subtropical front in the North Pacific. The project was funded by the US National Science Foundation and led by Professors Mike Gregg and Matthew Alford at the Applied Physics Laboratory, University of Washington. Of specific focus was an investigation of the role played by horizontal density gradients at scales less than 10 km in the restratification of the SML.

Submesoscale Lateral Density Variability

In contrast to recent observations that suggested all horizontal density gradients at scales of less than 10 km are compensated, i.e. the temperature and salinity contributions to density cancel each other out, we found that, in the subtropical Front, small scale density fronts are abundant throughout the SML at scales as small as 2 km. Using measurements collected during four large scale surveys at 30 N and 28 N obtained using the Shallow Water Mapping System (SWIMS) and a ship-mounted ADCP, the SML was observed to vary in depth considerably over relatively short distances (Figure 1). Horizontal density gradients, whilst comparatively weak in places, were evident particularly at 28 N where previous observations suggest that the large-scale front is uncompensated. Our findings suggest that dynamic processes restratify the SML at scales rarely resolved by numerical models of the SML.

Hosegood, P.; Gregg, M. C.; Alford, M. H.; 2006. Submesoscale lateral density structure in the oceanic surface mixed layer. Geophysical Research Letters, 33, L22604.

Mixed Layer Restratification

Previous observations suggested that horizontal gradients were responsible for enhancing stratification above that which could be explained by insolation alone. However, measurements made a fixed location over time are unable to identify horizontal variability, despite providing high-resolution observations of the temporal evolution of the SML. During ML04, the repeated surveying of a roughly 10 x 10km area using SWIMS enabled the influence on restratification of a weak but coherent horizontal density gradient within the SML to be quantified (Figure 2). Isopycnal surfaces were observed to slump from the vertical during daytime as the surface layers warmed due to insolation In contrast to the expected geostrophic adjustment of the lateral density front, our observations demonstrated that a near-inertial wave differentially advected the density field. The wave had been generated by a storm and was trapped within the frontal region by the relative vorticity associated with the frontal jet. As a result the wave's vertical wavelength was reduced, leading to an increase in vertical shear. These results suggest that the interaction of near-inertial oscillations, which are commonplace throughout the ocean following impulsive wind events, with lateral density gradients may be an important process to upper ocean restratification but that is currently ignored in numerical models.

Hosegood, P. J., M. C. Gregg, M. H. Alford, 2008. Restratification of the Surface Mixed Layer with Submesoscale Lateral Density Gradients: Diagnosing the Importance of the Horizontal Dimension. Journal of Physical Oceanography, 38, 2438-2460

Figure 1: Buoyancy frequency squared, N2, between 0 and 150 m depth for each survey, labelled accordingly.

Figure 2: Evolution of isopycnal surfaces within the SML during a 30 hour period at the subtropical front. Red dots at the surface indicate the positions at which a SWIMS profile was completed. Horizontal positions are measured relative to a drogued drifter around which the small-scale surveys were repeated.

Wind-Forced Subduction

The role of down-front winds that blow in the direction of a frontal jet have been shown in numerical simulations to generate intense vertical circulations. Within these overturning cells, intense downdrafts occur on the cold side of submesoscale fronts whilst upwelling arises on the warm side. A defining characteristic of these fronts are large Rossby numbers within elongated filaments of elevated vorticity. Following a period of intense down-front wind stress during ML04, we observed submesoscale intrusions beneath the SML (Figure 3). The intrusions were largely density compensated but their position corresponded to a narrow filament of intense cyclonic vorticity at the surface. We showed that the intrusions were generated by the wind-stress that intensified the ageostrophic circulations at the large-scale frontogenetic front by driving a cross-front Ekman buoyancy flux. In a manner analogous to surface cooling, the advection of light water over dense triggered convective instabilities that initialized the intense vertical overturning and resulted in the subduction of surface water beneath the seasonal pycnocline.

Hosegood, P. J.; Gregg, M. C.; Alford, M. H.; 2013. Wind-driven submesoscale subduction at the north Pacific subtropical front. Journal of Geophysical Research, doi: 10.1002/jgrc.20385

Surface Mixed Layer Evolution at Submesoscales (SMILES) (2014-2017)

SMILES is a NERC-funded, 3 year project that aims to assess the role played by submesoscales in transforming subantarctic mode water (SAMW). SAMW ventilates the Southern Hemisphere thermocline and, as one of the most important pathways for transporting anthropogenic tracers to the ocean interior, contains the largest total accumulation of anthropogenic CO2 and the highest CFC column inventory in the oceans. The dynamic processes driving the transformation of SAMW within the upper ocean, assumed to be driven air-sea fluxes until now, are thus crucial in setting the properties of the ventilated thermocline and the efficacy of the Southern Ocean in sequestering anthropogenic gases and tracers.

We will use an integrated approach in SMILES that employs observations and modelling at a range of scales to quantify the effects of submesoscales on the transformation of SAMW. The range of scales measured and simulated will enable us to validate the parameterisation of submesoscales in coarse-scale models so that we can then upscale our results throughout the Southern Ocean (Figure 4).

The SMILES cruise will be conducted in the Scotia Sea during April 2015 on the RRS James Clark Ross approximately 200 nautical miles south of the Falkland Islands. It will focus on a section of the subantarctic front (SAF) to the north of which SAMW is transformed and where submesoscales are almost certain to exist due to optimal forcing conditions. This particular section of the SAF exhibits high eddy kinetic energy due to instability of the SAF as it passes through Drake Passage and flows along the North Scotia Ridge and is further subjected to persistent westerly winds that blow down the SAF and are strongly implicated in promoting submesoscales. Numerical modeling will employ a three-step modelling approach, utilizing a high resolution regional ocean model and large eddy simulations (LES) to confront these challenges with the knowledge gained contributing towards the testing of submesoscale parameterisations in a coarse scale model. John Taylor at Cambridge University will lead the modeling component of SMILES.

Figure 3: Velocity field during during a small-scale survey of the SML indicated by streamlines and cones at depths of 20 m (light blue), 70 m (green) and 120 m (red). The submesoscale subduction is apparent by the low salinity water intruding up towards the surface and the high salinity water observed immediately beneath it.

Figure 4: Schematic of our approach towards understanding submesoscale effects on SAMW transformation. Each ellipse illustrates the approximate range (which here is illustrative) spanned by the measurements, both numerical and observational. The Fox-Kemper parameterisation is included in the coarse-scale ellipse and described further below, as is the analysis.