(to 1300 m from 700 m) and 1200 m (to 1800 m
from ϳ 600 m), respectively. The stronger entrainment in the western basin may also be responsible
for the longer period variability seen at the
Angmagssalik site (2.5–60 days) compared with
that at the southeast Iceland site (4–10 days).
At Cape Farewell, the southernmost tip of
Greenland, a 60-day deployment of three moorings has also furnished a transport estimate
(Clarke, 1984). However, only three current meter
records (on two moorings) were involved in calculating the dense water overflow. These measurements were combined with a hydrographic survey
to yield a transport of overflow water in the range
13–16 Sv. However, the offshore limit of flow was
arbitrary and it is quite impossible to associate any
kind of uncertainty with the estimate. This figure
is widely quoted in the literature and has been,
unfortunately I believe, linked with those estimates
for which error bounds can be derived. Nevertheless, the site is so well defined geographically that
estimates there are almost inevitable (see Section
5.6.6). Clearly array measurements targeting the
overflow should be made at Cape Farewell – if
possible soon.
To summarize this section: the principal
observed characteristics of the overflows in the
Iceland and Irminger basins are presented in Tables
5.6.1–5.6.3. Nowhere has a seasonal signal been
seen and variability (energy and period) generally
increases downstream. Questions remain concerning steep downslope flows, the southward penetration of the overflow in the Iceland basin, and the
merging of the two overflows in the Irminger basin.
5.6.4 Processes in the overflows
Currently the subjects of the location, intensity
and impact of mixing within the northern overflows are active fields of research and it is likely
that when this chapter is in print new results will
already be appearing. All this information will be
used to test and improve the parameterization of
mixing in both analytical and numerical models;
the latter are of particular importance in describing the overflows and hence the meridional overturning of the North Atlantic.
The vertical structure of overflows using expendable current profilers has been studied by Johnson
and Sanford (1992) in the Faroe Bank Channel, in
the region downstream of the Denmark Strait, as
well as in the Mediterranean outflow. Based on
all these studies a turbulent planetary boundary
layer (Ekman layer and logarithmic layer) is found
to exist below the current maximum. A bottom
mixed layer is also found, often considerably
thicker than the planetary boundary layer. In the
interface between the cold overflow and overlying
warm water the shear is sufficiently strong to
overcome the stratification and generate Kelvin–
Helmholtz billows and mixing. As a result a crossstream flow occurs opposite in sign and in terms
of transport roughly equal in magnitude to that
found near the bottom.
The author’s unpublished observations from a
35-day deployment of moored instrumentation
made in October 1989 near the sill of the Faroe
Bank Channel support this picture (see Fig. 5.6.8).
Assuming the downstream flow is in the direction
308–310°, the turning of the flow to the left 10 m
above the bottom (ϳ20°), and the turning to the
right 300 m above the bottom (ϳ5–7°) is shown in
the mean flow: note that the warm water 400 m
above the bottom has the direction 128°, i.e. in
opposition to 308°. The thickness of the bottom
mixed layer exceeds 100 m about 60% of the time
and about 10% of the time it exceeds 200 m.
Given the existence of a well-defined temperature–salinity relation, we can calculate the bulk
Richardson numbers over a depth interval of
100 m centred about 350 m above the bottom. A
time series from this calculation is shown in
Figure 5.6.9. Richardson numbers are seen to be
strongly modulated by semidiurnal tides and
inertial oscillations and critical values are frequently achieved, the latter arising almost entirely
from maxima in the shear (not shown here).
The impact of mixing is seen most clearly where
the overflow is in contact with only warmer
(lighter) water; then the warming that results can
be quantitatively linked via mass flux divergence
to average diapycnic diffusivity. Over a 75-km
path within and beyond the exit of the Faroe Bank
Channel a diffusivity across the 3°C isotherm
has been estimated as 100<50 cm
2 s
92 (Saunders,
1990).
Although to date no detailed profile observations and study of mixing processes have been
reported from either the Denmark Strait Overflow
or the Faroe Bank Overflow, we are aware that
some have been made and await their publication
with interest. An unexpected discovery of the flow
5.6 The Dense Northern Overflows
411
Saunders
from ϳ 600 m), respectively. The stronger entrainment in the western basin may also be responsible
for the longer period variability seen at the
Angmagssalik site (2.5–60 days) compared with
that at the southeast Iceland site (4–10 days).
At Cape Farewell, the southernmost tip of
Greenland, a 60-day deployment of three moorings has also furnished a transport estimate
(Clarke, 1984). However, only three current meter
records (on two moorings) were involved in calculating the dense water overflow. These measurements were combined with a hydrographic survey
to yield a transport of overflow water in the range
13–16 Sv. However, the offshore limit of flow was
arbitrary and it is quite impossible to associate any
kind of uncertainty with the estimate. This figure
is widely quoted in the literature and has been,
unfortunately I believe, linked with those estimates
for which error bounds can be derived. Nevertheless, the site is so well defined geographically that
estimates there are almost inevitable (see Section
5.6.6). Clearly array measurements targeting the
overflow should be made at Cape Farewell – if
possible soon.
To summarize this section: the principal
observed characteristics of the overflows in the
Iceland and Irminger basins are presented in Tables
5.6.1–5.6.3. Nowhere has a seasonal signal been
seen and variability (energy and period) generally
increases downstream. Questions remain concerning steep downslope flows, the southward penetration of the overflow in the Iceland basin, and the
merging of the two overflows in the Irminger basin.
5.6.4 Processes in the overflows
Currently the subjects of the location, intensity
and impact of mixing within the northern overflows are active fields of research and it is likely
that when this chapter is in print new results will
already be appearing. All this information will be
used to test and improve the parameterization of
mixing in both analytical and numerical models;
the latter are of particular importance in describing the overflows and hence the meridional overturning of the North Atlantic.
The vertical structure of overflows using expendable current profilers has been studied by Johnson
and Sanford (1992) in the Faroe Bank Channel, in
the region downstream of the Denmark Strait, as
well as in the Mediterranean outflow. Based on
all these studies a turbulent planetary boundary
layer (Ekman layer and logarithmic layer) is found
to exist below the current maximum. A bottom
mixed layer is also found, often considerably
thicker than the planetary boundary layer. In the
interface between the cold overflow and overlying
warm water the shear is sufficiently strong to
overcome the stratification and generate Kelvin–
Helmholtz billows and mixing. As a result a crossstream flow occurs opposite in sign and in terms
of transport roughly equal in magnitude to that
found near the bottom.
The author’s unpublished observations from a
35-day deployment of moored instrumentation
made in October 1989 near the sill of the Faroe
Bank Channel support this picture (see Fig. 5.6.8).
Assuming the downstream flow is in the direction
308–310°, the turning of the flow to the left 10 m
above the bottom (ϳ20°), and the turning to the
right 300 m above the bottom (ϳ5–7°) is shown in
the mean flow: note that the warm water 400 m
above the bottom has the direction 128°, i.e. in
opposition to 308°. The thickness of the bottom
mixed layer exceeds 100 m about 60% of the time
and about 10% of the time it exceeds 200 m.
Given the existence of a well-defined temperature–salinity relation, we can calculate the bulk
Richardson numbers over a depth interval of
100 m centred about 350 m above the bottom. A
time series from this calculation is shown in
Figure 5.6.9. Richardson numbers are seen to be
strongly modulated by semidiurnal tides and
inertial oscillations and critical values are frequently achieved, the latter arising almost entirely
from maxima in the shear (not shown here).
The impact of mixing is seen most clearly where
the overflow is in contact with only warmer
(lighter) water; then the warming that results can
be quantitatively linked via mass flux divergence
to average diapycnic diffusivity. Over a 75-km
path within and beyond the exit of the Faroe Bank
Channel a diffusivity across the 3°C isotherm
has been estimated as 100<50 cm
2 s
92 (Saunders,
1990).
Although to date no detailed profile observations and study of mixing processes have been
reported from either the Denmark Strait Overflow
or the Faroe Bank Overflow, we are aware that
some have been made and await their publication
with interest. An unexpected discovery of the flow
5.6 The Dense Northern Overflows
411
Saunders