The LOIS Shelf Edge Study (SES) has the following objectives:
SES involves POL as host laboratory, DML, PML and six university departments; it focuses on the outer continental shelf and slope near 56°N, west of Scotland.
Physical processes control the large-scale movement and irreversible small-scale mixing of water and its constituents. At the shelf edge, steep bathymetry may inhibit ocean-shelf exchange but, in combination with stratification, gives rise to special processes and modelling challenges.
An assessment has been made of coastal-trapped waves; along-slope currents, instability and meanders; eddies; upwelling, fronts and filaments; downwelling, cascading; tides, surges; internal tides and waves: as potentially influential processes in ocean-shelf exchange, water- mass structure and general circulation, according to their scales and context.
In particular, some further consideration has been given to cascading, the result of dense (eg. winter-cooled) water finding a route off the shelf below the less-dense adjacent slope water. Initially, cascading may be caused by instability of the flow in geostrophic balance with the cross-slope gradient of density, or it may be in the bottom Ekman layer under this flow. Any depression in the shelf edge is liable to facilitate and concentrate the process; hydraulic control may then operate. It has also been found that conditions leading to cascading, viz. denser water on the shelf (an up-slope increase in density) result in a thicker bottom boundary layer. Indeed, the boundary-layer solution is not valid (ie. breaks down) if the slope and density gradient are jointly large enough. If the latter is caused by winter cooling on the shelf, then the critical topographic scale is O(10 km) at mid- to high latitudes. Such a scale is common at the shelf edge; this boundary layer thickening may also be an influence on cascading.
In typical contexts, several processes give currents O(0.1 m/s) and cross-slope exchange O(1m^2/s); therefore, some atypicality of the context is likely to determine one or more dominant processes. Some external forcing agent (eg. winds, oceanic pressure fields) appears to be important if cross-slope transports are to be sustained. Internal tides and internal/inertial waves are important for interior mixing.
The 1995-96 experimental programme began with a cruise led by POL scientists in March 1995. Although the cruise was severely constrained by weather, a high-resolution bathymetric survey was largely completed for the area from 56° to 57°N, in depths 150 m to 1000 m. The survey used the swath bathymetry system fitted to RRS Charles Darwin, producing charts at 1 : 50 000 resolution, excellent for the development of fine-resolution models of the area. Some 85 hours of side-scan-sonar data were also obtained using the Towed Ocean Bottom Instrument (TOBI). The records show slumps, slides, channels, levees, ridges, wavy bedforms and changing geological structures. These features will help to guide the coring and geochemical sampling in later SES cruises.
The above sea-bed survey had immediate benefit by influencing the choice of positions for moorings laid on the second leg of the cruise (to be maintained throughout the SES experiment; POL leads the mooring team). The initial array comprised current meters, bottom pressure recorders, thermistors, transmissometers, a sediment trap, nutrient analysers and a meteorological buoy. Conductivity-Temperature-Depth (CTD) and other profiling and water sampling were carried out at the locations of moorings with analyses for salinity, nutrients and preservation for later analysis of dissolved organic nitrogen, particulates and plankton. Coring and sea-bed photography were carried out near to the mooring locations; the bottom proved to be covered with boulders on the shelf and smaller pebbles down to 300 m, so that coring had limited success in these shallower depths. The ship-borne ADCP operated throughout; northward currents exceeded 0.5 m/s when the tidal and along-slope currents combined.
To maintain the moorings and carry out intensive spatial surveys over a full annual cycle, six seasonal cruises will follow in 1995 and 1996 at approximately 3-month intervals.
Collaboration with the University of Wales, Bangor (UWB) School of Ocean Sciences, in developing a 3D semi-analytic wave model, has proceeded with the calculation of two analytic test cases (steep slope; near-inertial frequency) where super-inertial wave energy loss is small. However, the 3D semi-analytic results have proved difficult to interpret in relation to physically realisable motion; a 3D numerical approach is to be adopted.
A POL 3D hydrodynamic model of the shelf-edge region west of Scotland was used to study spatial variability of the tides and wind-driven circulation. Its grid resolution 1/12° x 1/12° can resolve the barotropic but not the baroclinic flow. In the vertical, a co-ordinate transformation follows near-surface and near-bed boundary layers with enhanced resolution. Vertical mixing is modelled using various turbulence energy closure methods, including prognostic equations for turbulence energy and mixing length. The model reproduced observed areas of locally-intensified tidal currents near the shelf edge (for example). Modelled flow fields produced by a westerly wind, a southerly wind or by a northward flow out of the Irish Sea through the North Channel showed very similar patterns, and were in good agreement with the long-term flow fields derived from radioactive tracer studies. The direction of the flow field at depth is clearly influenced by the effects of bottom topography.
The 3D model was enhanced to include baroclinic effects. It was used in the form of a cross- shelf section to examine the generation of internal tides and associated turbulence energy and mixing at the shelf break; the model contains all the non-linear terms necessary to generate short internal waves. To resolve internal tides accurately over the shelf break, a fine cross- shelf grid (about 1 km or less) was found necessary. Results showed that the intensity and spatial variability of the internal tide are sensitive to the (summer, winter, up- or down- welling) density field and to the shelf-slope depth profile.
[A comparison was made with a UWB semi-analytic internal tide generation model. The models were run with the same summer stratification and M2 tidal forcing, over similar depth profiles. The results showed similar distributions of on/off-shelf velocity and vertical displacement. However, amplitudes of the calculated currents differed and showed great sensitivity to the exact depth profile.]
Near the shelf break, short internal waves are formed; their wavelength is influenced by the model's horizontal diffusion. At the sea surface, above the shelf break, internal tidal shear enhances mixing, which can influence sea-surface temperatures (cooling by this means is observed near the Celtic Sea shelf edge); the mixing may also supply nutrients upwards, near enough to the surface (in the presence of light) to fuel production.
The model simulates upwelling in a bottom boundary layer with enhanced mixing over the shelf, giving a shelf-edge front. In downwelling, where lighter water is drawn below heavier water, instantaneous vertical mixing takes place in the model, giving a region of vertically homogeneous water which moves down the slope, separated from the offshore water by a frontal system.
These model calculations of the scales and locations of expected flow structures have helped in designing the SES array for moorings, surveys and near-bed sampling. For example, high concentrations of suspended particulate matter (SPM) are expected where bed stresses are large.
In its 3D form, the model has been used to study the generation of internal lee-waves, behind sea-mounts (under-sea mountains ) as the flow passes over the sea-mount. The model showed enhanced mixing just behind the sea-mount, a Taylor column over the sea-mount, and eddies moving downstream (all according to the size of sea-mount, strength of flow and vertical stratification).
The grid in the Hebrides shelf-edge model is being refined, to simulate these varied processes in 3D with realistic topography. An Irish Sea model with the same characteristics is being combined with the shelf-edge model to study the important input through the North Channel. Improved account is also being taken of inputs from west of Ireland, by southward extension of the model there.
A 2D model of the cross-slope section was run for the seasonal cycle of circulation in response to tidal, wind and thermal forcing. Appropriate formulations for fine near-surface resolution, advection and dynamical pressure gradient over the slope showed encouraging improvements. Data from 25 years of DML CTD sections across the Hebrides shelf near 57°N were analysed to provide a mean annual cycle (and variance) for comparison.
Towards the incorporation of microbiological models, there has been collaboration with UWB. The seasonal cycle of stratification on the Hebrides Shelf has been modelled, together with the primary production cycle in 1D (vertical) using the UWB model SEDBIOL, which couples physics, SPM transports (including deposition and resuspension) and microbiology. Results predict a spring bloom at the end of May, in an average (meteorological) season, to be compared with the May 1995 SES cruise results.