Lit Review7

OSX 4016: Literature Review
1
Canyons: Sewage Pipes of the Shelf?
Meghan Rochford
Abstract
Shelf seas are regions of great importance both physically and biogeochemically. Many physical
processes affect these regions, such as topography, wind stress, tidal currents and stratification. In
temperate regions topography is a controlling factor of shelf seas, it causes an along-slope geostrophic
flow to form, which cannot cross isobaths. Biogeochemically shelf seas are hotspots for phytoplankton
growth and nutrient recycling. The Celtic Sea is a broad shelf sea known for its strong tidal forcing and
seasonal stratification. During spring and summer it becomes stratified due an increase in surface
temperatures, allowing the thermocline to shoal, bringing nutrients up from the deeper ocean,
causing the spring bloom. The conditions which allow for dense water cascades to occur have been
documented in the Celtic Sea. Submarine canyons are areas of enhanced cross-shelf exchange and are
important topographic features which have been observed to cause enhanced
upwelling/downwelling, mixing due to internal tides, and act as conduits for dense water cascades. In
this review, the difference between long and short canyons will be discussed with reference to
changing Rossby numbers and the effect it has on the flow regime of a canyon.
Shelf Seas
The coastal ocean and open ocean each have diverse physical and biological processes. The area
where they meet, the shelf-edge, has its own unique processes. Exchange between the open ocean
and shelf seas have important implications for shelf-sea currents, flushing and the supply of nutrients
(Huthnance et al. 2009), which in turn, have implications for phytoplankton production (Rees et al.
1999). It is also believed that shelf processes exercise some control on open ocean circulation in ocean
basins and mixing over slopes, which are known to contribute to the main oceanic density structure
(Munk & Wunsch 1998). Topography is an important controlling factor in shelf seas, as it constrains
large-scale flows (geostrophic) from crossing the slope (Huthnance et al. 2009), causing unique
exchange processes at the shelf-edge. This geostrophic constraint is not as severe at the equator, and
is more relaxed, especially in Ekman layers, due to friction (Huthnance 1995). The transport of
nutrients and carbon between shelf seas and the open ocean has important significance to the
nutrient and carbon cycles, however, it has not been adequately quantified (Huthnance et al. 2009).

2
Physical Processes
It has been documented many times that along-isobath flow along the continental slope is primarily
steered by steep topography (Huthnance 1995; Allen 2004), meaning that cross-slope flow is hindered,
and the exchange between coastal and open ocean waters is limited (Allen & Durrieu de Madron
2009). This means that a homogeneous, geostrophic flow has no divergence, and therefore moves as
rigid columns of water, unable to change its length (Taylor 1923). Because of the rigidity of the water,
it cannot change depth, confining it between isobaths. In the same way as above, a stratified flow is
limited to a depth with no flow, and anything above that depth can flow across isobaths. Therefore, a
shelf-break current, from the surface to a depth, will block even a stratified geostrophic flow from
crossing isobaths (Allen & Durrieu de Madron 2009). Deep ocean shelf exchange (DOSE) therefore
occurs when there are ageostrophic flow dynamics. This occurs in the presence of large frictional
processes, time dependence, or advection (Allen 2004). Fig. 1 shows the main environmental physical
processes associated with shelf seas: wind stress, tidal currents, stratification, and shelf geometry
(Estrade et al. 2008).
Fig. 1. Summary of the physical processes which occur in a shelf sea (Huthnance et al. 2009).

3
The Ekman (1905) paper is the cornerstone concept of upwelling processes. Ekman (1905) states that
wind-forced oceanic mass transport occurs to the right of wind in a shallow layer of water. Northerly
winds drive surface Ekman transport offshore along a western boundary. These conditions are
equatorward at an eastern boundary, and polewards at a western boundary (in the northern
hemisphere)(see Fig. 1 and Huthnance 1995). This causes upwelling of deep, cold water, from a depth
no deeper than 200-300 metres (Pond & Pickard 1983). Sea surface temperatures (SST) taken off the
Northwest African coast shows that there is a temperature minimum close to the coast, showing this
cold water upwelling (Fig. 2)(Estrade et al. 2008).
Fig. 2. Seven-day composite of MODIS SST
images off the northwest African coast.
Shows a temperature minimum close to the
coast (Estrade et al. 2008).
Estrade et al. (2008) used a two-dimensional model to study Ekman’s 1905 theory further, showing
that upwelling occurs offshore of a ‘kinematic barrier’ to the cross-shelf flow (dependant on the level
of stratification), resulting from the upper and lower Ekman layers merging. Their results showed that
90% of Ekman transport upwells for h/D (h being depth and D being the Ekman depth) between 1.25
and 0.5 (alongshore wind) (Fig. 3). Wenju & Lennon (1988) suggested that the seasonal upwelling
studied in the Taiwan Strait was dependant on surface wind stress, driving Ekman pumping.

4
Fig. 3. Conceptual schematic of the mechanism of upwelling separation from the coast (Estrade et al. 2008).
Tidal energy, especially internal tides, and subsequently internal waves are important physical
processes needed for mixing in the water column. At the upper shelf slope and shelf edge there is
enhanced internal mixing due to the formation of non-linear internal waves (often tidal) (Fig. 4). These
waves can then break into internal solitons, causing significant fluxes across the shelf edge (Inall et al.
2001). This is all made possible by the relatively steep slope. These solitons cause an increase in
current shear, leading to increased internal turbulence and mixing. This leads to a dissipation of the
energy from the internal tide (Sharples et al. 2007).

5
Fig. 4. Schematic showing the formation
and dissipation of internal waves at the
shelf edge. A) During off-shelf ebb flow
the thermocline is distorted, forming a
depression over the shelf edge. B) As the
ebb tidal flow decreases shorter internal
waves, with increased amplitude. C) As
the waves flow onto the shelf, they are
further shortened and steepened, causing
a higher shear and an increase in vertical
mixing (Sharples et al. 2007).
There are different levels of stratification experienced throughout shelf seas (Fig. 1); for some, like the
Irish Sea, seasonal stratification is common. This is when an increased heat flux in spring and summer
out-competes mixing due to wind and tides, and causes the formation of a warmer surface layer, while
the water column remains vertically homogeneous in winter. In other regions the water column
remains vertically homogeneous throughout the year, caused by higher levels of turbulent energy in
the system, due to wind and tidal mixing (van Aken 1986).
Tidal mixing fronts are boundaries between stratified and well-mixed seas (Fig. 1). They occur in
regions with high tidal dissipation adjacent to regions of large seasonal heat exchange (Simpson &
Bowers 1981). The average temperature of the stratified region rises more slowly than the mixed, due
to the heat being kept at the surface. Therefore, the surface layer of the stratified regime has a higher
temperature than that of the mixed regime. This causes a density difference between the two regimes,
with the mixed side having a lower depth-mean density. At the bottom layer depth the density
increases from the mixed regime to the stratified. Overall, this causes a pressure slope at depth and
at the surface from the mixed to the stratified regime. This produces a pressure gradient force (PGF)
driven flow, which is deflected by the Coriolis force, balancing as a geostrophic flow. This flow is
perpendicular to the PGF; therefore, there is an along front flow parallel to the mixing front flow.

6
Biogeochemical Processes
Primary production in temperate regions is characterised by seasonal intermittency, and starts with a
spring bloom. This occurs when the light-determined critical depth descends into the mixed layer
depth, allowing for further algal growth. As heating increases and wind stress decreases the mixed
layer depth shoals. The timing of the spring bloom varies locally and inter-annually. For example, the
bloom may occur earlier if there are larger freshwater inputs, causing a shallower surface layer, or it
may be later if there is an increase in suspended sediments, reducing light penetration, and thus
phytoplankton growth. It ends once the initial near surface nutrient concentration becomes limited
(Huthnance et al. 2009). Summer growth depends on upwelling of biologically fixed regenerated
nitrogen. It is exchanged through the thermocline by turbulence from winds, waves and internal
waves. The presence of a subsurface chlorophyll maximum occurs when the surface water becomes
nutrient limited, forcing the advection of nutrients from the bottom layer through the thermocline,
where it is immediately consumed by phytoplankton.
The role of shelf seas in the carbon cycle is of significant importance, yet they are sometimes over-
looked in global estimates of CO2 uptake and production. Using the North Sea as an example, the
difference in CO2 uptake will be discussed. The North Sea can be divided into two regimes: in the north
the water column is stratified, while in the south the water column is shallower, and therefore
vertically homogeneous (mixed) (Fig. 6). In the northern North Sea the uptake of dissolved inorganic
carbon (DIC) occurs in the surface mixed layer, organic material sinks into the bottom layer, where
respiration takes place. This allows the surface layer to have a low concentration of carbon, allowing
for the continual uptake of DIC. This allows the northern part of the sea to act as a sink (Thomas et al.
2004). In the southern North Sea production and respiration occur in the same ‘compartment’. This
means that the uptake and release of DIC is in equilibrium with the atmosphere, causing a higher
concentration of DIC in the water column, causing it to act as a source (Prowe et al. 2009).

7
Fig. 6. Schematic of the North Sea: The difference in regimes in the north (sink) and south (source).
The Northwest European Shelf
The Northwest European Shelf encompasses the Hebrides and Malin shelves, the English Channel, the
Irish Sea, the Celtic Sea, and Irish Shelf (Fig. 6). It is approximately 2000 kilometres long, from the
Amorican Shelf in the south, up to the North Sea in the north (46-60°N), at the eastern boundary of
the North Atlantic Ocean. Due to the presence of the British Isles, the region experiences a range of
complex topography.
The region experiences strong tidal forcing at the ocean boundary. The English Channel, Irish Sea and
Bristol Channel are characterised by strong tidal responses, with the largest ranges occurring on the
eastern side of the basins. The largest tidal ranges (>8 metres at M2 tides) have been recorded near
the port of St. Malo, at the head of the Bay of Seine, and Avonmouth, Bristol (Simpson 1998).
The M2 tide enduces frictional stresses at the seabed, over much of the region, with a maximum stress
of 0.25 Pa, which is the equivalent of a sea surface wind stress of 13 ms-1
. In the Irish Sea, extreme
stresses have been recorded at 4 Pa, which is on the scale of hurricane force wind stresses. Due to
this, the European Shelf is believed to cause approzimately 12.5% of the global tidal energy loss
(Simpson & Bowers 1981).

8
Fig. 6. Map of the Northwest
European Shelf, with the locations
of the Celtic Sea, Irish Sea, Irish
Shelf, English Channel, the
Hebrides and Malin Shelf (200
metre depth contours, red lines
separating shelves) (Huthnance et
al. 2009).
The area is subject to strong seasonal forcing, with surface heating and cooling changing the structure
of the water column. In regions with strong tidal flows (the English Channel and eastern Irish Sea) the
water column is continually mixed, while regions such as the Celtic Sea and Hebridean Shelf
experiences stong seasonal stratification. In the North Channel, between the Irish Sea and the Malin
Shelf there is complete vertical homogeneity to a depth of 200 metres, due to a strong tidal current
of 1.5 ms-1
at spring tides (Simpson 1998).
The Celtic Sea
The Celtic Sea is a 500 kilometre wide (approximately), 100-200 metre deep shelf sea with a highly
dynamical environment (Fig. 6)(Huthnance et al. 2009; Green et al. 2008). It has a large tidal energy
input originating from the Atlantic Ocean. It is characterised by strong tidal currents, which are known
to be the dominant source of mechanical energy (Simpson 1998). The area is subject to strong
seasonal variations in surface heating and cooling. Freshwater supply to the area is limited, meaning
stratification is dominated by temperature (Green et al. 2008). This stratification becomes established
over summer, where buoyancy input out-competes stirring by wind and tidal stirring.
At the shelf edge, internal tides generate internal waves due to the forcing of the barotropic tide. It
causes the water column up and down the steep slope, generating waves (Fig. 4)(Sharples et al. 2007).
This also produces a baroclinic energy flux (Green et al. 2008). These internal waves cause mixing and

9
diffusion across the thermocline, bringing cooler, deeper water to the near surface. This water is then
exposed by wind mixing (Fig. 1). These processes lead to a large ocean-shelf exchange of 3 m2
s-1
(Huthnance et al. 2009).
North Atlantic water forms a poleward slope current, that is warm and saline, flowing along the
continental slope from Portugal, past Ireland to (Cooper 1952). This barotropic current is centred at
approximately 500 metres on the slope (Cooper 1949; Pingree & Le Cann 1989; Huthnance et al. 2009).
The depth of the slope current is suggested to be forced by the dynamic height of warmer subtropical
waters (Huthnance 1984). Below this current is the bottom Ekman layer, where the current is reduced
to zero, due to friction. Off-shore Ekman transport in the region is believed to be of the order of 1
m2
s-1
(Huthnance 1995).
In the Celtic Sea, low-frequency circulation is generally weak, except at the upper slope, and when
channelled through topographic features (e.g. canyons). This localised exchange is equal to the slope
current transport (of order of 1 Sv.) (Huthnance et al. 2009). The discontinuous coastline allows wind-
driven flow eastward through the English Channel (0.1 Sv.: Prandle et al. (1996)) and northward
through the Irish Sea (0.1 Sv.: Knight & Howarth (1999)). At the shelf edge, tidal currents exceed 0.5
ms-1
, with tidally reflected flow reaching 0.1 ms-1
(Huthnance et al. 2009). Internal tides are particularly
strong in the region of the shelf edge. At spring tide they have wavelengths >50 metres, and propagate
as decaying waves both off- and on-slope. This means that off-shore exchange occurs in wave-form of
approximately 1.3 m2
s-1
(Huthnance et al. 2001).
The enhanced mixing that is associated with internal waves, causes a flux of nutrients and chlorophyll
over the thermocline, associated with a cooler band of water (Sharples et al. 2007; Green et al. 2008;
Huthnance et al. 2009). This flux of nutrients allows the subsurface chlorophyll maximum to be
sustained. Primary production is therefore localised. It has also been documented that there is an
increase in chlorophyll near seamounts and banks, possibly due to intensified mixing (Green et al.
2008).

10
Fig. 7. Satellite imagery showing the average sea surface temperatures (left), and chlorophyll concentrations
(right) in the Celtic Sea, between 3rd
July and 19th
July 2004 (Green et al. 2008).
Historical surveys (Cooper 1949) of the Celtic Sea show that there are cascading-favourable conditions
over the shelf in winter, and traces of cascades at the continental slope. In simplest terms cascading
can be split into two locations. Both the locations in Fig. 8 have the same initial salinity, temperature
and density. During the onset of winter, the water columns are cooled, with location A cooling faster,
due to the reduced depth. This causes higher density water further up the slope, causing it to ‘cascade’
down past location B. When looking at an actual continental slope, the same concept can be used. In
Fig. 9 there are three locations; the shelf, the shelf break and the slope. Both locations A and B cool
faster than C, with A cooling faster than B, causing both to cascade down the slope due to higher
densities (Cooper 1949). Ivanov et al. (2004) also determined that there were two distinct areas of
dense water cascading in the Celtic Sea: in the South Celtic Sea and over the adjoining Armorican Shelf.
Fig. 8. Simplised diagram showing the basis of water
cascading at two locations: A and B (twice as deep)
(Cooper 1949).
Fig. 9. Simplised diagram of dense water cascading at
the continental slope (Cooper 1949).

11
Canyons
Submarine canyons are abundant features along continental shelves to deep ocean basins. Shepard &
Dill (1966) mapped just 96 major submarine canyons globally. Using modern satellite bathymetry data,
more than 660 submarine canyons have been mapped globally (Fig. 10)(De Leo et al. 2010). They are
areas of enhanced cross-shelf exchange, primarily because they are regions of large Rossby numbers.
In such regions, the effects of planetary rotation are secondary to the effects of advection of
momentum. Because canyons are much smaller than the surrounding slope, the Rossby number gets
increasingly larger (Wåhlin 2002; Allen 2004).
Fig. 10. Global distribution of submarine canyons: red (circles) named, white un-named, and yellow and orange
are from studies (De Leo et al. 2010).
Canyons are features with complex topography, hydrography, flow, and sediment transport and
accumulation. Because of these complexities, they are known for their distinctive characteristics, such
as accelerated currents, enhanced mixing, and dense water cascades. These can be forced by
topographic or climatic forcings (Klinck 1996; Mulder et al. 2012). Submarine canyons are important
features for physical and biogeochemical reasons. They have been observed to be hotspots for
enhanced upwelling and downwelling, allowing for an increased exchange between shelf waters and
open ocean (Allen & Durrieu de Madron 2009; Allen & Hickey 2010).
Geostrophic flows cannot cross isobaths, restricting cross-shelf flows, forcing along-slope continental
slope flows. Due to geostrophy, this causes an across-slope pressure gradient (Allen & Durrieu de
Madron 2009). In a submarine canyon, flow cannot be along-slope due to the restrictions of the
canyon walls. This means that the Coriolis force cannot balance the pressure gradient force allowing
for flow down the pressure gradient (Freeland & Denman 1982). Therefore, flow is dominated by the

12
pressure gradient at the canyon rim. There is no direct impact on the near surface flow. Canyon flow
can be broken down into two types (Allen & Durrieu de Madron 2009):
1. A wind-driven shelf-break or slope current, with the strongest effect felt at the canyon rim;
2. On-shelf deep water formation with an equally strong cross-slope pressure gradient.
However, this water cascades deep into the canyon, making it independent of the wind-driven
flows.
A typical upwelling or downwelling event can be divided into three main phases (Allen & Durrieu de
Madron 2009):
1. An initial transient phase;
2. A near steady advection-dominated phase;
3. A relaxation phase.
The first phase, the initial transient phase is a time-dependent response, as the shelf-break flow
increases. It is generally quite strong and occurs quickly, normally within an inertial period (Allen &
Durrieu de Madron 2009). If there is a steady wind, causing the along-current to continue, density
advection within the canyon reduces the time dependent upwelling after about five days (She & Klinck
2000). It is essentially linear, with similar responses for both positive and negative flows (see Fig. 11).
The second phase, the advection-dominated phase, is not linear, and therefore more complicated. It
is dependent on the canyon topography and flow strength. This phase occurs when the shelf-break
flow is reasonably steady. In this phase, upwelling is generally stronger than downwelling (She & Klinck
2000). Upwelling is driven by negative flows (Fig. 11), thus opposing the shelf waves and arresting
them, leading to strong across isobaths flow. Downwelling is driven by positive flows, moving in the
same direction as wave propagation, allowing along-isobath flows to be established around the
canyon and onto the shelf (Allen & Durrieu de Madron 2009).
Oscillatory flows over the canyon have been suggested to create mean flow over the canyon, due to
the asymmetry of upwelling and downwelling. In the positive phase, the flow leaves the canyon via
the downstream wall, having diverged from the upstream wall (Fig. 11). In the negative phase, flow
follows the upstream wall into the canyon, and leaves via the downstream wall (Allen & Durrieu de
Madron 2009).
In the final phase, the relaxation phase, shelf-break flow reduces (Allen & Durrieu de Madron 2009).
Hickey (1997) suggested that upwelled water leaves the canyon laterally in this phase, rather than
horizontally.

13
Fig. 11. Schematic showing the flow of
water through a submarine canyon. The
negative phase (upwelling conditions) is
flow in the opposite direction of Kelvin
wave propagation, while the positive
phase (downwelling conditions) is in the
same direction as wave propagation
(Edited: Allen & Durrieu de Madron
2009).
Upwelling
Submarine canyons are important mechanisms for coastal upwelling, with a high concentration of
zooplankton seen around them (Allen et al. 2001). However, there is a difference in upwelling between
short canyons and long canyons. Short canyons are those that the head of the canyon reaches the
continental slope, long before it reaches the coastline, for example, Astoria (off the west coast of the
USA) and Barkley canyons (west of Vancouver Island)(Hickey 1997; Allen 2000). In a long canyon, the
head of the canyon does not reach the continental slope before the coastline, but rather it extends
far into the coastal region, usually into an estuary (Hickey 1995; Allen 2000). Examples of long canyons
are Juan de Fuca, Mackenzie and Monterey canyons (Waterhouse et al. 2009). Flow in short canyons
has been well studied and documented, while long canyons remain largely unstudied.
In short canyons, as a geostrophic flow passes over the canyon, water is driven up the canyon (Fig.
12). This occurs due to a pressure gradient imbalance caused by restrictions in the topography
(Freeland & Denman 1982). This imbalance is what causes enhanced mixing and upwelling (Hickey
1995). Water columns, originating upstream of the canyon, flow over the top, becoming stretched.
This is due to an increase in bottom depth downstream of the canyon rim. This stretching creates a
cyclonic vorticity in the flow (Hickey 1997; Allen 2004). This has been linked to flow separation at the
mouth of the canyon, which is then advected into the canyon. The flow then turns towards the canyon
head and is advected onto the shelf. Due to the vortex stretching, a cyclonic eddy is formed from the
shelf break down to a depth in the canyon mouth (She & Klinck 2000). Flow above the canyon (<100
metres) does not feel the effects of the canyon, except for a possible elevation of isopycnals (Hickey
1997; Allen 2004; Waterhouse et al. 2009).

14
Fig. 12. Schematic showing the processes which lead to upwelling within a short submarine canyon (Allen &
Hickey 2010).
In long canyons, upwelling has been observed to happen throughout the canyon, in all regions (Allen
et al. 2001; Waterhouse et al. 2009). Flow dynamics in long canyons are dependent on Rossby
numbers, time-dependence and advection. Some of the flow characteristics are similar to that of short
canyons, while some are different (Skliris et al. 2001). Waterhouse et al. (2009) used modelling to
determine the effects of changing Rossby numbers on a long canyon, such as the Juan de Fuca.
When low Rossby numbers were input into the model it was determined that there were two
characteristic stages of flow, due to the restriction of isobath convergence (Waterhouse et al. 2009):
1. The generation of vorticity through isopycnals stretching on the upstream side of the canyon,
upwelling occurring downstream of the mouth rim, a slow cyclonic flow within the canyon
walls, and a slowing of flow on the shelf, upstream of the canyon.
2. The formation of an eddy at the canyon mouth, continuation of the flow cyclonic flow in the
earlier stage, as well as the continuation of upwelling at the downstream rim of the mouth. In
this stage the vorticity of the canyon mouth eddy was dependent on stratification. However,
the upstream flow within the canyon was not dependent on stratification.
When a moderate Rossby number was used, it was observed that the pattern of upwelling was
different between short and long canyons. In long canyons, upwelling occurred at the mouth of the
canyon, while in short canyons it occurs through the head of the canyon and at the downstream rim.
At high Rossby numbers, there were similar upwelling processes between the short and long canyons.
Both showed advective regimes, with upwelling at the head of the canyons. As the Rossby number

15
decreases in short canyons, upwelling tends to occur in the first phase. Therefore, it was determined
that upwelling was dependent on flow strength and a high Rossby number. In long canyons, upwelling
occurred during all Rossby number simulations, even during quasi-ready conditions. This is due to high
isobath convergence over the canyon (Waterhouse et al. 2009; Allen 2000).
Dense Water Cascading
As discussed previously, dense water cascades form on the shelf, where water is cooled and cascades
down the slope (Fig. 9). Submarine canyons have been documented to be conduits for this process
(Allen & Durrieu de Madron 2009). Dense water (DW) cascading contributes to the ventilation of
intermediate and deep water of the open ocean, which has a substantial impact on biogeochemical
cycles. The effect of canyons on DW cascades varies with the length, width (Wåhlin 2004) and
orientation (Chapman 2000) of the canyon, as well as the topographic features present (Wåhlin et al.
2008). In a uniformly sloping shelf with a canyon cutting into it, a portion of DW will cascade into the
canyon, forming a plume, flowing offshore along the canyon axis, to the right side of the canyon. The
formation of eddies, due to a density front, have been documented to slump into the canyon,
disrupting this DW plume (Chapman & Gawarkiewicz 1995). Wåhlin (2002) looked at the steering
influence of canyons on DW cascades. Dependent on the magnitude of the along-slope current, DW
was observed to cascade through the canyon, with an accumulation of this DW in the canyon.
Wåhlin (2004) looked at the influence of length and width of canyons on DW cascading. It was
determined that the transport capacity of deep channels was larger than that of shallow channels.
When gently sloping topography was used in the model, there was a maximum downward flow
through a wide canyon (>10 kilometres), however, steeper regions were the most active, when the
canyon was a few kilometres wide. Wåhlin et al. (2008) looked at the influence of the overall shape of
the canyon (v-shaped) with respect to different flow regimes and topographic features. They
determined that small scale topography has a bigger influence on mixing than large scale topography.
Chapman (2000) looked at three different canyon orientations: normal, diagonal and parallel to the
shelf. It was found that little DW enters the normal and diagonal canyons. This is due to the fact that
along-slope flow follows isobaths, and cannot flow down into the canyon. In regards to the parallel
canyon, there was a higher portion of the flow in this canyon. The amount of DW in the canyon was
dependent on the rate of flow over the canyon. A slower along-slope flow meant that there was more
DW in the canyon, due to the reduced speed and therefore increased meandering. It was also found
that DW cascading has shown to induce localised upwelling of deep water onto the shelf (Kämpf 2005).

16
Internal Tides and Mixing
Upwelling and dense water cascading are both processes of advection, which cause the movement of
water through the canyon. Another process of transport is the mixing of deep water, due to tides.
Submarine canyons act as conduits for deep ocean water further onto the shelf then if it were of
uniform length. This means that the head of canyons is an area of enhanced tidal mixing (Allen &
Durrieu de Madron 2009).
Long canyons (as discussed above) are particularly strong areas of enhanced tidal mixing. This is due
to them stretching from the shelf break, along the slope and up to large estuaries. Deep ocean water
circulates in these estuaries, and with the axis and head of the canyon being subjected to large tidal
currents, mixing is enhanced. Long canyons are therefore areas of strong nitrate concentrations (Allen
& Durrieu de Madron 2009).
In long canyons that do not reach into large estuaries, tidal currents do not penetrate into the canyon,
but rather across the canyon, parallel to the shelf-break (Allen & Durrieu de Madron 2009). It has been
documented that many of these canyons are areas of very high internal tidal energy, such as
Hydrography Canyon (Wunsch & Webb 1979) and Monterey Canyon (Kunze et al. 2002). They are
areas of enhanced tidal and internal wave energy due to focusing (Wunsch & Webb 1979) or due to
them being large regions of critical slope (Hotchkiss & Wunsch 1982). It has also been suggested that
small scale roughness causes enhanced mixing (Wåhlin et al. 2008) and internal tidal energy (Kunze et
al. 2002). Diffusivity values in such canyons are very large at the canyon axis (0.05 m2
s-1
), with the
canyon rim have values just a factor of ten smaller, recorded in the Monterey Canyon.
Conclusion
Shelf seas are important regions for both physical and biogeochemical reasons. They are regions
where the exchange of open ocean and shelf water occurs. At the shelf break, a poleward (in the
northern hemisphere, eastern side of basin) along-isobath slope current flows as rigid water columns,
unable to change its depth (Taylor 1923). Because of this rigidity, exchange between the open ocean
and shelf sea is limited. Wind stress drives Ekman transport offshore, allowing for deeper water to be
upwelled close to the coast, bringing with it nutrient rich water (Ekman 1905). Shelf seas are areas of
high tidal energy. This energy enhances mixing at the shelf edge due to the formation of internal waves
(Green et al. 2008), which have been observed to move onshore. Shelf seas, such as the Irish Sea,
become seasonally stratified during the spring and summer months, due to an increase in buoyancy,
which out-competes stirring by wind and tides (van Aken 1986). Tidal mixing fronts form where
stratified water meets mixed water. During spring, the water column becomes stratified due to an
increase in temperature at the surface. As the light determined critical depth descends below the

17
thermocline (due to an increase in day length) more nutrients become available, leading to the spring
bloom. It varies locally and inter-annually due to freshwater and sediment inputs.
The Northwest European Shelf is characterised by its complex topography, due to the presence of the
British Isles. It is approximately 2000 kilometres long, from Portugal to Norway. The area is known for
its strong tidal forcing, with the M2 tide causing an increase in frictional stress at the seabed, causing
much mixing. The region is subject to strong seasonal forcings, being heated in spring and summer,
and cooled in winter. The Celtic Sea is a region located within the Northwest European Shelf. It is 500
kilometres wide and 100-200 metres deep. It has a large tidal energy range, originating from the North
Atlantic, which is the dominant mechanical energy input (Simpson 1998). Stratification in the region
is dominated by temperature, which is established during summer. A barotropic poleward current is
centred around 500 metres, made up of North Atlantic water (Cooper 1949). Ocean-shelf exchange in
the region is approximately 3 m2
s-1
(Huthnance et al. 2009). The region has also been documented as
having favourable conditions for dense water cascades, in winter (Cooper 1949).
Submarine canyons are regions of enhanced upwelling and downwelling on the continental shelf.
There have been more than 660 mapped globally using satellite bathymetry (De Leo et al. 2010). They
are characterised as regions with complex topography, flow and hydrography, and large Rossby
numbers. They are known to have two forms of water transport: advection (upwelling/downwelling
and dense water cascades) and mixing (internal tides). Upwelling/downwelling has three phases: the
first is an initial transient phase when shelf break flow increases, the second is an advection-
dominated phase during steady state, and the third is a relaxation phase, when flow begins to slow.
Upwelling occurs predominantly in the first two phases (Allen & Durrieu de Madron 2009).
In short canyons, there is a pressure gradient imbalanced caused by topographic restrictions, which
allows for upwelling to occur. Cyclonic vorticity is a dominate feature in short canyons (Hickey 1997;
Allen 2004). In long canyons, which protrude onto the shelf, reaching a large estuary, different Rossby
numbers allow for different flow regimes. When a moderate Rossby number is considered upwelling
occurs at the mouth of the canyon, whereas in short canyons it occurs at the head and downstream
of the rim. When a high Rossby number is considered the flow regime is similar to that of the short
canyon. Unlike short canyons, upwelling/downwelling can occur during any of the three phases in a
long canyon (Waterhouse et al. 2009).
Dense water cascading has been observed to occur through canyons. The magnitude of the cascade is
dependent on the width, length (Wåhlin 2004), orientation (Chapman 2000) and topographic features
(Wåhlin et al. 2008) of the canyon. Canyons which are orientated parallel to the coastline are believed
to allow higher volumes of water to pass through them, forming a plume. The speed of the along-
slope current also affects the amount of dense water cascading through the canyon. Slower currents

18
will allow for more to cascade through the canyon, while faster currents will flow directly passed it
(Chapman 2000).
Mixing, due to internal waves and tidal energy, has also been suggested as a mechanism for water
transport through a canyon. In long canyons, where the head reaches into an estuary, there are large
tidal currents, allowing for enhanced mixing, bringing a high level of nutrients to the region. In long
canyons which do not extend up to the coastline internal tides are the dominant mixing force, due to
focusing and a large critical slope. In these canyons, diffusivity values are of order 0.05 m2
s-1
at the
axis (Allen & Durrieu de Madron 2009).
References
Allen, S.E., 2000. On subinertial flow in submarine canyons: Effect of geometry. Journal of Geophysical
Research, 105, pp.1285–1297.
Allen, S.E. et al., 2001. Physical and biological processes over a submarine canyon during an upwelling
event. Canadian Journal of Fisheries and Aquatic Sciences, 58, pp.671–684.
Allen, S.E., 2004. Restrictions on deep flow across the shelf-break and the role of submarine canyons
in facilitating such flow. Surveys in Geophysics, 25, pp.221–247.
Allen, S.E. & Durrieu de Madron, X., 2009. A review of the role of submarine canyons in deep-ocean
exchange with the shelf. Ocean Science Discussions, 6(2), pp.1369–1406.
Allen, S.E. & Hickey, B.M., 2010. Dynamics of advection-driven upwelling over a shelf break submarine
canyon. Journal of Geophysical Research: Oceans, 115(March), pp.1–20.
Chapman, D.C., 2000. The influence of an alongshelf current on the formation and offshore transport
of dense water from a coastal polynya. Journal of Geophysical Research, 105(C10), p.24007.
Chapman, D. C., & Gawarkiewicz, G., 1995. Offshore transport of dense shelf water in the presence of
a submarine canyon. Journal of Geophysical Research: Oceans (1978–2012), 100(C7), 13373-
13387.
Cooper, L.H.N., 1949. Cascading over the continental slope of water from the Celtic Sea. , pp.719–750.
Cooper, L.H.N., 1952. The physical and chemical oceanography of the waters bathing the continental
slope of the Celtic Sea.
Ekman, V. W., 1905. On the influence of the earth's rotation on ocean currents, Arkiv. Mat, Astron,
Fysik, 2, pp.1–52.
Estrade, P. et al., 2008. Cross-shelf structure of coastal upwelling: A two — dimensional extension of
Ekman’s theory and a mechanism for inner shelf upwelling shut down. Journal of Marine
Research, 66, pp.589–616.
Freeland, H.J. & Denman, K.L., 1982. A topographically controlled upwelling center off southern
Vancouver Island. Journal of Marine Research, 40(4), pp.1069–1093.
Green, J. a. M. et al., 2008. Internal waves, baroclinic energy fluxes and mixing at the European shelf
edge. Continental Shelf Research, 28, pp.937–950.
Hickey, B.M., 1995. Coastal Submarine Canyons. Topographic Effects in the Ocean. SOEST Special
Publications, pp.95–110.

19
Hickey, B.M., 1997. The Response of a Steep-Sided, Narrow Canyon to Time-Variable Wind Forcing.
Journal of Physical Oceanography, 27, pp.697–726.
Hotchkiss, F.S. & Wunsch, C., 1982. Internal waves in Hudson Canyon with possible geological
implications. Deep Sea Research Part A. Oceanographic Research Papers, 29(4), pp.415–442.
Huthnance, J.M., 1995. Circulation, exchange and water masses at the ocean margin: the role of
physical processes at the shelf edge. Progress in Oceanography, 35(95), pp.353–431.
Huthnance, J.M. et al., 2001. Physical structures, advection and mixing in the region of Goban Spur.
Deep-Sea Research Part II: Topical Studies in Oceanography, 48(14-15), pp.2979–3021.
Huthnance, J.M., 1984. Slope Currents and “JEBAR.” American Meteorological Society, 14, pp.795–
810.
Huthnance, J.M., Holt, J.T. & Wakelin, S.L., 2009. Deep ocean exchange with west-European shelf seas.
Ocean Science, 5, pp.621–634.
Inall, M.E., Shapiro, G.I. & Sherwin, T.J., 2001. Mass transport by non-linear internal waves on the
Malin Shelf. Continental Shelf Research, 21(13-14), pp.1449–1472.
Ivanov, V. V. et al., 2004. Cascades of dense water around the world ocean,
Kämpf, J., 2005. Cascading-driven upwelling in submarine canyons at high latitudes. Journal of
Geophysical Research C: Oceans, 110, pp.1–10.
Klinck, J.M., 1996. Circulation near submarine canyons: A modeling study. Journal of Geophysical
Research, 101(95), p.1211.
Knight, P.J. & Howarth, M.J., 1999. The flow through the north channel of the Irish Sea. Continental
Shelf Research, 19(5), pp.693–716.
Kunze, E. et al., 2002. Internal Waves in Monterey Submarine Canyon. Journal of Physical
Oceanography, 32(6), pp.1890–1913.
De Leo, F.C. et al., 2010. Submarine canyons: hotspots of benthic biomass and productivity in the deep
sea. Proceedings. Biological sciences / The Royal Society, 277(May), pp.2783–2792.
Mulder, T. et al., 2012. Present deep-submarine canyons activity in the Bay of Biscay (NE Atlantic).
Marine Geology, 295-298, pp.113–127.
Munk, W. & Wunsch, C., 1998. Abyssal recipes II: Energetics of tidal and wind mixing. Deep-Sea
Research Part I: Oceanographic Research Papers, 45(1998), pp.1977–2010.
Pingree, R.D. & Le Cann, B., 1989. Celtic and Armorican slope and shelf residual currents. Progress in
Oceanography, 23, pp.303–338.
Pond, S. & Pickard, G. L., 1983. Introductory Dynamical Oceanography, Pergamon Press, 329.
Prandle, D. et al., 1996. Combining modelling and monitoring to determine fluxes of water, dissolved
and particulate metals through the Dover Strait. Continental Shelf Research, 16(2), pp.237–257.
Prowe, A.E.F. et al., 2009. Mechanisms controlling the air-sea CO2 flux in the North Sea. Continental
Shelf Research, 29(15), pp.1801–1808.
Rees, A.P., Joint, I. & Donald, K.M., 1999. Early spring bloom phytoplankton-nutrient dynamics at the
Celtic Sea shelf edge. Deep-Sea Research Part I: Oceanographic Research Papers, 46, pp.483–
510.
Sharples, J. et al., 2007. Spring-neap modulation of internal tide mixing and vertical nitrate fluxes at a
shelf edge in summer. Limnology and Oceanography, 52(5), pp.1735–1747.

20
She, J. & Klinck, J.M., 2000. Flow near submarine canyons driven by constant winds. Journal of
Geophysical Research, 105(C12), pp.28671–28694.
Shepard, F. P. & Dill, R. F., 1966. Submarine canyons and other sea valleys. Rand McNally.
Simpson, J. H., 1998. The Celtic Seas coastal segment. In: Brink, K. H. & Robinson, A. R. (Eds.), The Sea,
11, Wiley, New York, pp.659-698.
Simpson, J.H. & Bowers, D., 1981. Models of stratification and frontal movement in shelf seas. Deep
Sea Research Part A. Oceanographic Research Papers, 28(7), pp.727–738.
Skliris, N. et al., 2001. Shelf-slope exchanges associated with a steep submarine canyon off Calvi
(Corsica, NW Mediterranean Sea): A modeling approach. Journal of Geophysical Research,
106(C9), p.19883.
Taylor, G.I., 1923. Stability of a viscous liquid contained between two rotating cylinders. Proc. Roy. Soc.
A, 223, pp.289–343.
Thomas, H. et al., 2004. Enhanced open ocean storage of CO2 from shelf sea pumping. Science (New
York, N.Y.), 304(5673), pp.1005–1008.
van Aken, H.M., 1986. The onset of seasonal stratification in shelf seas due to differential advection in
the presence of a salinity gradient. Continental Shelf Research, 5(4), pp.475–485.
Wåhlin, a. K., 2004. Downward channeling of dense water in topographic corrugations. Deep-Sea
Research Part I: Oceanographic Research Papers, 51(4), pp.577–590.
Wåhlin, a. K. et al., 2008. Laboratory observations of enhanced entrainment in dense overflows in the
presence of submarine canyons and ridges. Deep-Sea Research Part I: Oceanographic Research
Papers, 55(6), pp.737–750.
Wåhlin, a. K., 2002. Topographic steering of dense currents with application to submarine canyons.
Deep-Sea Research Part I: Oceanographic Research Papers, 49, pp.305–320.
Waterhouse, A.F., Allen, S.E. & Bowie, A.W., 2009. Upwelling flow dynamics in long canyons at low
Rossby number. Journal of Geophysical Research: Oceans, 114(5), pp.1–18.
Wenju, C. & Lennon, G.W., 1988. Upwelling in the Taiwan Strait in response to wind stress, ocean
circulation and topography. Estuarine, Coastal and Shelf Science, 26, pp.15–31.
Wunsch, C. & Webb, S., 1979. The Climatology of Deep Ocean Internal Waves. Journal of Physical
Oceanography, 9(2), pp.235–243.

Lit Review7

More Related Content

What's hot (20)

Viewers also liked (18)

Similar to Lit Review7 (20)

Lit Review7