THE DISTRIBUTION OF DISSOLVED SILICA IN THE
DEEP WESTERN NORTH ATLANTIC OCEAN
by
Gerald J. Needell
B.S., Northeastern University
(1975)
SUBMITTED IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE
at the
MASSACHUSETTS INSTITUTE OF TECHNOLOGY
and the
WOODS HOLE OCEANOGRAPHIC INSTITUTION
November, 1978
Signature
of
Author.....
.-.....
.
.............
Joint Program in Oceanographr, Massachusetts Institute
of Technology - Woods Hole Oceanographic Institution,
Department of Meteorology, Massachusetts Institute of
Technolog
November,C/L9f78.
.
Certified by..........................
...
Thesis Supervisor
Thesis Supervisor
Accepted by..................................................
Chairman, Joint Oceanography Committee in the Earth
Sciences, Massachusetts Institute of Technology Woods Hole Oceanographic Institution.
MASSACHUSES
rwK
OF TECH
IT
LVI~ARES
To Sally - who is my inspiration
THE DISTRIBUTION OF DISSOLVED SILICA IN THE
DEEP WESTERN NORTH ATLANTIC OCEAN
by
Gerald J. Needell
Submitted to the Department of Meteorology
on November 13, 1978 in partial fulfillment of the requirements
for the Degree of Master of Science
ABSTRACT
The distribution of dissolved silica in the deep
western North Atlantic Ocean is presented.
The potential
temperature-dissolved silica relationship is compared
with the potential temperature salinity relationship in
the North Atlantic Deep Water.
Geographical variations
in the potential temperature-dissolved silica relationship are discussed with particular emphasis on the low
silica signal of the Western Boundary Undercurrent (WBUC).
The WBUC is shown to have a significant influence on the
potential temperature-dissolved silica relationship from
the tail of the Grand Banks of Newfoundland to Cape
Hatteras.
It is suggested that a region of enhanced mixing is
present west of 65*W that is responsible for the observed
changes in the dissolved silica distribtution.
Thesis Supervisor:
Dr. J. R. Luyten, Associate Scientist
ABSTRACT
The distribution of dissolved silica in the deep western
North Atlantic Ocean is presented.
The potential temperature-
dissolved silica relationship is compared with the potential
temperature salinity relationship in the North Atlantic Deep:
Water.
Geographical variations in the potential temperature-
dissolved silica relationship are discussed with particular
emphasis on the low silica signal of the Western Boundary
Undercurrent (WBUC).
The WBUC is shown to have a significant
influence on the potential temperature-dissolved silica
relationship from the tail of the Grand Banks of Newfoundland
to Cape Hatteras.
It is suggested that a region of enhanced mixing is present west of 65*W that is responsible for the observed changes
in the dissolved silica distribution.
Table of Contents
Abstract
I.
II.
Introduction
Background
2.1
Historical uses of dissolved silica as a water
mass tracer.
2.2
III.
IV.
The deep western North Atlantic Ocean.
Data and methods.
Water mass characteristics.
4.1
Potential temperature-salinity characteristics.
4.2
Potential temperature-dissolved silica
characteristics.
4.2.1
V.
Polynomial fits to the O-SiO 2 relationship
The distribution of dissolved silica on surfaces of
constant potential temperature.
VI.
VII.
5.1
The 1.9*C potential temperature surface.
5.2
The 2.20 C potential temperature surface.
5.3
The 2.40 C potential temperature surface.
5.4
The 3.0*C potential temperature surface.
Discussion
6.1
The O-SiO 2 relationship
6.2
Geostrophic transport measurements.
6.3
Moored current meter observations.
6.4
Enhanced mixing west of 65*W.
Summary
Appendix I
Acknowledgements
Bibliography
I.
Introduction
The distribution of dissolved silica in the deep western
North Atlantic Ocean may be used to identify water masses
with distinct characteristics within the potential temperature and salinity range of the ubiquitous North Atlantic deep
water.
The most prominent feature of the silica distribution is
a narrow band of low silica along the continental rise.
This
low silica feature shall be identified as the Western Boundary
Undercurrent CWBUC) in keeping with the literature (Richardson (1977)).
It shall be demonstrated that the WBUC can be
traced from the tail of the Grand Banks of Newfoundland to
Cape Hatteras.
With the data presented below, a "global"
view of the WBUC shall be provided.
In the past, observations of the WBUC have consisted of
hydrographic sections across the continental rise for which
the geostrophic transport was computed.
Either neutrally
buoyant floats or moored current meters were used to set absolute velocities (Swallow and Worthington (1961), Volkmann
(1962), Richardson (1977), Clarke, Hill, Reiniger, and Warren
(1978)).
Occasionally, dissolved silica was measured and it
could be demonstrated that the water observed did have a
northern origin (Richardson (1977), Clarke, Hill, Reiniger,
and Warren (1978)).
Eight month long current meter records along 700 W show
a steady westward flow on the upper continental rise (Luyten,
1977).
Unfortunately, one cannot be sure whether the west-
ward flow is the WBUC (i.e., low silica) or is part of a
broad westward flow in the slope water region (Webster (1969)).
This brings up an important problem with the historical
observations of WBUC.
Stommel (1957) postulated the exis-
tence of a deep southward flowing western boundary current
in the Atlantic Ocean.
His model was based on the hypothesis
that some thermohaline process caused sinking of water in
arctic regions and upwelling in antarctic regions.
To con-
serve mass, he argued that a deep flow would be present along
the western boundary.
In the following year, Swallow and Worthington (1961)
occupied hydrographic sections and tracked neutrally buoyant
floats near Cape Romaine, North Carolina.
In the hydrographic
data, they found that the deep isotherms sloped downward off
shore indicating the presence of westward shear.
When they
observed that the neutrally buoyant floats moved westward, it
was assumed that the WBUC had been found.
The current meter
data from 70*W shows that even where the flow is relatively
steady toward the west, it is not always so.
versals in the flow are not uncommon.
O-ccasional re-
Thus one must be
cautious in the interpretation of few day-long float tracks.
It should also be noted that the spatial coverage of the
data from which we car observe the WBUC is not uniform.
The
only long-term current meters on the continental rise were
at 70*W.
Hydrographic sections wiLh some method of determin-
ing absolute velocities are only availaS.e near 70*W and Cape
7
Hatteras.
The results of a section at 50*W are not considered
realistic for reasons discussed in the text.
Thus we are left
with only a limited set of observations of the WBUC.
There is
no direct evidence for a continuously flowing current that
passes arcund the tail of the Grand Banks of Newfoundland and
follows the cortinental rise toward the west.
To the contrary,
the current meterz data and the variations in the transport
calculations suggest that whatever flow there is, is not
steady.
The distribution of dissolved silica will be used to
demonstrate that the WBUC can be identified in terms of its
water mass characteristics.
The effects of the WBUC can be
traced along the continental rise from the tail of the Grand
Banks of Newfoundland to Cape Hatteras.
However, the struc-
ture of the flow cannot be determined from the available data.
Evidence is presented below which suggests that a
region of enhanced mixing is present west of 65 0 W and has a
pronounced effect upon the dissolved silica distribution.
II.
Background
2.1
Historical uses of dissolved silica as a water mass
tracer
It has only been in the past three decades that one has
been able to map out the distribution of dissolved silica.
On the basis of sparse data, Sverdrup, Johnson, and Fleming
(1942) concluded that the concentration of dissolved silica
increased monotonically with depth everywhere in the world
oceans.
They pointed out that the silica values in the
Atlantic Ocean were in general, lower than those of the other
oceans.
Cooper (1952) examined the sources and sinks of dis-
solved silica and suggested that the concentration of silica
in the deep water should be in equilibrium and might be useful as a water mass tracer.
Without any observations to rely
on, Cooper assumed that the tundra drainage in the Arctic
would result in a high silica content of the waters of the
North Polar Sea.
He then reasoned that there would be a flow
of high silica water in the Greenland and Irminger Currents
which could be easily detected by its silica content.
While
recent observations have indeed shown that there exists high
silica water inthe Baffin Basin (SiO2 > 80pg A/k - Corwin and
McGill (1963), Grant (1970)), there is no evidence of it
getting over the sill and into the Labrador Basin.
In fact,
the Norwegian and Labrador Seas are characterized by a low
silica signal (SiO 2 nu 10-15 pg A/k).
As it shall be demon-
strated, these waters are easily detected along the western
boundary of the western North Atlantic Ocean.
Armstrong (1965) showed the distinctly lower concentrations of silica in the deep Atlantic Ocean, when compared to
the world oceans.
He concluded that this was a result of
the unique exchange of bottom waters that occurs in the
Atlantic Ocean and suggested that "...our understanding of
water movements and of mixing processes in the Atlantic Ocean
would be greatly increased by more knowledge of the silica
content of Arctic waters."
Metcalf (1969) showed that the concentration of dissolved silica could be used to identify Antarctic Bottom
Water, North Atlantic Deep Water, and Antarctic Intermediate
Water.
He also suggested that silica might be useful in
studying the NADW where little variation in the potential
temperature-salinity relationship is found.
Carmack (1973) found that silica was a valuable tracer
in the Antarctic.
He found that he could distinguish be-
tween water masses that appeared indistinguishable on the
basis of 0-S data alone.
2.2
The deep western North Atlantic Ocean
It is well known that there is a tight correlation between potential temperature and salinity in the deep western
North Atlantic Ocean for potential temperatures below 3.0*C
(Worthington and Metcalf (1961), Wright and Worthington
(1970)).
Most of the water characterized by this 0-S corre-
lation for 0 > 1.9*C have been classified as North Atlantic
Deep Water (NADW) (Wright and Worthington (1970)).
There is
little geographic variation in the 0-S distribution of the
western North Atlantic in this potential temperature range
(Worthington and Wright (1970)).
This has made it difficult
for physical oceanographers to study the deep general circulation in terms of potential temperature and salinity
characteristics.
Near the continental rise, geostrophic calculations and
direct velocity measurements (neutrally buoyant float and
moored current meters) allowed observers to deduce the existence of a deep southward flowing current (Swallow and
Worthington (1961); Volkmann (1962); Richardson (1977)).
This flow has become known as the Western Boundary Undercurrent and is responsible for the equatorward movement of
deep water from the Norwegian and Labrador Seas (Stommel
(1957); Richardson (1977); Clarke, Reiniger, Hill, and
Warren (1978)).
Metcalf (1969) demonstrated the usefulness of the dissolved silica content of deep water as a water mass tracer.
Using the available data, he constructed a map of the concentration of dissolved silica on the 2.0*C potential temperature surface.
His figure is reproduced as Fig. 1.
The
principal feature of this map is the general north-south
gradient of dissolved silica.
The paucity of data prevents
a detailed analysis of the distribution but this map does
suggest that dissolved silica is a useful tracer of the
general circulation of the North Atlantic Ocean.
Richardson (1977) and Worthington (personal communication) found that there is a drop in the silica concentration
along potential isotherms as one approaches the continental
rise.
Worthington (1976) has proposed a model of the general
circulation of the North Atlantic Ocean, in which he postulated the existence of two anti-cyclonic gyres.
One to the
north of the Grand Banks of Newfoundland, and the second
comprised of the Gulf Stream and its return flow bounded to
the south by the Mediterranean salt tongue.
In the deep
water Worthington based his interpretation partly on the
change in the dissolved silica distribution in the deep water
across the proposed boundary between the two gyres.
His fig-
ure showing characteristic deep O-SiO 2 plots for the two
gyres is reproduced as Fig. 2.
Worthington's model also
allows for a WBUC passing around the tail of the Grand Banks
of Newfoundland and along the continental rise until it
crosses beneath the Gulf Stream near Cape Hatteras.
He al-
lows for no exchange of water between the WBUC and the southern gyre.
Clarke, Hill, Reiniger, and Warren (1978) have criticized Worthington's model.
They claim that in the deep water,
no evidence exists for two distinct gyres, and that the
silica distribution (from a survey of their own) indicates
that there is an exchange of water between the "Northern
Gyre" area and the Sargasso Sea.
In this work the distribution of dissolved silica in
the deep western North Atlantic Ocean, based on several sections occupied during the last two decades, shall be presented.
years.
A large portion of the data is from the last six
After presenting a summary of the data and methods
of analysis used, we shall describe the water mass characteristics present and form a consistent picture of the geographical variations in the silica distribution.
It shall
be demonstrated that there are significant variations of
the silica distribution within the NADW.
These variations
will allow us to trace the origins of some components of the
NADW which cannot be distinguished on the basis of 0-S characteristics.
The distribution of silica on potential temper-
ature surfaces shall be examined and these distributions
compared with models of the general circulation.
In addition
the WBUC shall be examined by following its silica signal
along the continental rise from the tail of the Grand Banks
to Cape Hatteras.
Evidence shall be presented for isopycnal
mixing along the WBUC path, resulting in downstream dilution
of the low silica WBUC water.
The data coverage of the western North Atlantic Ocean
has expanded greatly over that which was available to Metcalf
(1969) and Worthington and Wright (1972).
More structure in
the silica distribution can be discerned here than was possible previously.
The time span over which the data was
collected must certainly be kept in mind.
Little is known
of the temporal variations of the deep water properties.
That a consistent picture of the deep silica distribution
can be presented is taken to indicate that these temporal
variations are indeed small.
It must, of course, be kept in mind that local anomalies
do exist.
They may be either due to temporal, spatial, or
possibly instrumental effects.
As discussed in Appendix I
large anomalies which contradict surrounding data must be
suspect, but more reasonable local features are retained in
the data set and are accepted as essential characteristics
of the ocean.
noted.
Where these features stand out they will be
III.
Data and Methods
The data used in this report was obtained by searching
the WHOI and NODC oceanographic files for cruises in which
temperature, salinity, and dissolved silica were measured.
Only those cruises where salinities were measured by a conductivity bridge (i.e., three decimal place precision) were
used.
This basically restricted the data set to cruises
since 1958.
The data used is
summarized in
Table 1 (Fig., 3).
data from one cruise, KNORR 12, was rejected for reasons
given in Appendix I.
In a few cases, duplicate sections (i.e., two or more
sections at the same location) existed in the data.
The
section with more closely spaced stations was used.
In add-
ition, cruises for which samples were analyzed at sea were
preferred to those during which the samples were frozen.
All
of the data used in this report was analyzed at sea with
either a spectrophotometer or autoanalyzer.
For all cruises except AII 100 (see below) the data were
accepted as they existed in the files.
Observations denoted
as questionable by the observers were discarded.
Plots of
O-SiO2 and S-SiO2 were formed and observations which disagreed with the main body of data were rejected (see discussion of O-SiO2 data).
All of the data (again with the
exception of AII 100 as noted below) were from previously
published sources (see Table I) and had been edited on the
basis of the mean 6-S distribution of the North Atlantic
Ocean.
The
14
Table I.
Year
Ship
Cruise
Location
Previous Publication
1958
DISCOVERY II
I.G.Y. 2
240 30' N*
320 15' N*
Metcalf (1969)
Worthington and Wright (1970)
1961
ATLANTIS
265
40*
1961
CHAIN
20
Woods Hole-Bermuda
Worthington and Wright (1970)
1966
ATLANTIS
18
near 360 N,66*W
and 36*N,63*w
Metcalf(1969)
1971
TRIDENT
Cape Hatteras
Richardson (1977)
1972
HUDSON
50*W and around
the tail of the
Grand Banks
Clarke, Hill, Reiniger and
Warren (1978)
1972
CHAIN
104
around the tail of
the Grand Banks
Clarke, Hill, Reiniger and
Warren (1978)
1972
KNORR
30
GEOSECS stations*
27-34
Broeker, Takahashi and Li(1976)
1975
KNORR
48
640
Worthington(in preparation)
1977
KNORR
66
55*W
1978
ATLANTIS
100
70*W
15' N*
30'W
Worthington(in preparation)
* used only in the preparation of the maps of dissolved silica on potential
temperature surfaces
The section from cruise AII 100 was occupied by the
author in the spring of 1978.
The samples were collected
with a 24-bottle rosette of Niskin bottles on a Neil Brown
CTD.
For this work only the temperature data from the CTD
was used.
The salinities were determined at sea on a Guild-
line Autosal salinometer and the dissolved silica was determined with a spectrophotometer.
The data were edited by
plotting both salinity and silica versus pressure and removing those points from which both salinity and silica values
indicated contamination of the sample.
For the thirteen
stations occupied, the worst had eight "contaminated" samples
and the best had none (there were 24 samples per station).
The final 0-S data set was then compared to the WorthingtonMetcalf (1969) mean 0-S curve and no other questionable salinities were obvious below 4.0*C.
The contour plots presented in Section V were prepared
by linearly interpolating the 0-SiO2 data for each station
to the 0 values used.
IV.
Water Mass Characteristics
4.1
Potential temperature salinity characteristics
As previously stated, the deep western North Atlantic
Ocean is characterized by a tight potential temperature salinity correlation.
Figure 4 is a composite plot of deep 6-S
observations for those stations indicated.
The boundaries
of the North Atlantic Deep Water are indicated (Worthington
and Wright (1970)).
Below the inflection point at 6
"
1.85*C
a low salinity tail is present indicating the effects of
Antarctic Bottom Water (Wright and Worthington (1970); Broeker,
Takahashi, and Li (1976)).
4.2
Potential temperature-silica characteristics
Figure 5 is a composite plot of deep observations of
potential temperature and dissolved silica for those stations
indicated.
It should be noted that within the potential tem-
perature range of the North Atlantic Deep Water (6 > 1.9*C),
we find the maximum scatter in the 8-Sio2 plot.
Note also that for 0 < 1.85 0 C there is a sharply increasing silica tail corresponding to the low salinity tail of the
Antarctic Bottom Water (AABW) (Broeker, Takahashi, and Li
(1976)).
This high silica branch is confined to the south
of the tail of the Grand Banks of Newfoundland.
One can see
the difference clearly when the data are separated into groups
of data from either side of the tail of the Grand Banks (see
Figures 6a and 6b).
Those stations east of 50*W show no AABW.
This geographical limit to the northward influence of the AABW
is similar to that used by Wright and Worthington (1970) in
their volumetric potential temperature, salinity census of
the North Atlantic Ocean.
The AABW was seen clearly in the
North American Basin, but not in the Labrador Basin (Figures
7a and 7b were reproduced from Wright and Worthington (1970)).
Wright and Worthington (1970) also found that in the
Labrador Basin (but absent in the North American Basin) there
was a relatively saline (with respect to the Antarctic Bottom
Water) and cold (6 < 1.9*C) water mass (Denmark Straits,
Overflow Water (DSOW)).
In agreement with this we find that
for 0 < 1.9*C, east of 50*W (Figure 6b),
branch on the O-SiO 2 plot.
there is a low silica
This water is not generally ob-
served west of 50*W (see exception noted below).
The two branches of the 6-SiO2 plot come together at
o
".
1.9C as did the two branches of the 0-S distribution
(Wright and Worthington (1970)).
Worthington and Wright (1970)
showed that at 1.80 C and below there is no direct path between
the AABW and the DSOW.
path first exists
It is between 1.80 C and 1.90 C a direct
(Figures 8a and 8b).
The data presented above suggest that for 0 < 1.9*C both
salinity and dissolved silica can be used to identify the AABW
and the DSOW.
In the range of the NADW (0 > 1.9C) one may take advantage of the wide variation of dissolved silica (whereas there
is very little variation of salinity) to distinguish between
water masses of different origin within the NADW.
In the data from east of 50*W (Figure 6b) there is a
distinct gap in the data from 0 = 1.9* to 3.4*C.
The higher
silica branch runs from 0 Ix,3.40 C and SiO2 'x 18 pg A/
0
"u
1.9C and SiO2 n 32 pg A/
AABW and the DSOW.
to
where it joins in with the
In this same potential temperature range,
there exists a lower silica branch running from 6 % 3.4C,
SiO 2 "- 16 pg A/9
to 0 % 1.9 0 C, SiO 2 % 18 pg A/2.
This lower
silica water may be traced back into the Labrador and
Norwegian Seas (Grant (1970)), and is presumed to be the
source of the Western Boundary Undercurrent (WBUC) (Richardson (1976); Clarke, Hill, Reiniger, and Warren (1978)).
West of 50*W, the distribution is quite different
(Figure 6a) in the NADW range of potential temperature.
There are only a few observations to the low side of the
majority of the observations.
east is not evident here.
The distinct gap seen to the
In fact, most of the "anomalous"
observations come from two cruises.
The section along 50*W
shows the two branched distributions clearly and also shows
some of the DSOW water at 0
%
1.8*C (Figure 9).
This sec-
tion was grouped with the western group because of the AABW
seen for e < 1.9 0 C and SiO 2 > 30 pg A/2.
of the anomalous data is AII 18
The other source
(Figure 10).
These data
also show evidence of DSOW (0 % 1.85*C, SiO 2 'u 24 pg A/k).
These are the only observations of the water mass at a point
so far west.
Since these observations of the DSOW are so
isolated, it is assumed that there was an intrusion of this
water mass into the area at the time these data were
collected.
With the two sections discussed above deleted, the
o-
SiO 2 distribution west of 50*W consists of a broad line
of values from 0 = 3.5*C and SiO 2 = 14 pg A/
SiO2 = 30 pg A/
to 0 = 1.9*C
where it meets the AABW (Figure 11).
We stated previously that it was, in general, the lower
values of silica that were found closer inshore.
This will
become more clear when the charts of silica on constant
potential temperature surfaces are discussed in Section
V.
The reason for the lower values inshore is assumed to be the
influence of the WBUC (Richardson (1976); Worthington (1976)).
Examination of the individual sections which extend from
the abyssal plain up onto the continental rise reveals that
while there is a general tendency toward lower values of
silica inshore, the strength of the gradient and the extreme
values vary considerably (Figures 12a-e).
divided into two categories.
The figures may be
The first are those sections
which show a distinct gap between the WBUC and the remainder
of the NADW (HUDSON,
KN 48).
The second are those which show
a continuous band of observations but are limited to a range
of values within the extremes of HUDSON
CH 20, TRIDENT).
geographically,
and KN 48 (AII 100,
These types of distributions are grouped
the
two branched distributions being to the
east (east of 65*W) and the more uniform distributions being
to the west.
4.2.1
Polynomial fits to the potential temperature
silica relationship
To quantify the differences between these distributions, polynomial fits to the data were made.
It was found
that little was gained for fits of higher than fourth order.
The fits were obtained with a least square polynomial regression program for the Hewlett-Packard 9830 calculator.
fits and residuals are presented in Figures 13-17.
The
Root mean
square deviations for each of the fits were also computed
(Table 2).
It should be noted that the r.m.s. deviations are
lower to the west than to the east, the plots of the residuals
show that for CH 20, AII 100, and TRIDENT, the residuals are
evenly distributed between ±2a with an increase of the scatter
toward lower potential temperatures.
For both KN 48 and
HUDSON, we find that the residuals are evenly distributed
above 0 = 3.0*C but break into two branches for 0 < 3.0*C.
The low residuals can readily be identified in the WBUC water
which forms the low silica branch of the 0 - SiO 2 plots.
To
illustrate the effects of the WBUC on the fits, the inshore
stations were removed (containing the WBUC observations) from
the KN 48 and HUDSON sections and re-computed the fits
(Figures 18, 19).
The r.m.s. deviations are now 1.10 for
HUDSON and 1.01 for KN 48.
The plots of the residuals now
show a more uniform distribution than did the fits including
the WBUC.
The curves fitted to the KN 48 (65*W) and HUDSON
(50*W) sections with the WBUC removed are nearly identical
(Figure 20).
To the west of 65*W, the three fits produce
similar curves.
Above 20*C those for TRIDENT and AII 100 are
nearly identical and CHAIN 20 shows a deviation from the
other two between 2.50 and 40 C (Figure 21).
In Figure 21 a
2 standard deviation envelope has been included to show that
the discrepancies are not statistically significant even from
2.5 C
+
4C.
To demonstrate that the WBUC has an effect on the distribution of silica west of 650 C the fits for AII 100, CH 20,
21
Table II.
Cruise
Location
HUDSON
50*W
KNORR 48
650 W
CHAIN 20
Woods Hole-Bermuda
ATLANTISII 100 70*W
TRIDENT
Cape Hatteras
Stand4rd Deviation
(ugA/l)
2.72
1.46
1.41
1.10
1.21
and TRIDENT were plotted along with HUDSON and KN 48 (With WBUC
removed)
(Figure 22).
The 2a envelope is
from KN 48.
It can
be shown that for 2.00 < 0 < 2.40*C there is a significant difference between the AII,
CH 20, TRIDENT curves and the HUDSON
aknd KN 48, indicating that the water of the WBUC has a significant effect west of 650 W.
It should be emphasized that the data being compared were
collected over a span of several years.
The three sections
which show- the effect of the WBUC west of 65 0 W were occupied in
1961, 1971,
and 1978 (CH 20, TRIDENT and AII 100, respectively).
Th-e cons-istency of the data suggests that the conditions were
not anomalous at the time of the observations.
For comparison, the e
'
SiO 2 plots for the sections dis-
cussed above are presented again with the KN 48 (without WBUC)
curve and 2a envelope (Figures 23a-h).
Thus it has been shown while there are dramatic changes
in the 0 - SiO
distribution from the tail of the Grand Banks
2
of Newfoundland to Cape Hatteras, one can demonstrate that the
low silica signal of the WBUC can be traced throughout the
region.
V.
The Distribution of DIssolved Silica on Surface of
Constant Potential Temperature
In the previous section, the plots of silica versus po-
tential temperature demonstrated the existence of geographic
variations in the silica distribution of the deep western
North Atlantic Ocean.
In particular, it was found that near
the tail of the Grand Banks of Newfoundland and along the
continental rise as far west as 650 W, the North Atlantic deep
water consisted of two distinct water masses, easily distinguished by their silica content.
Further west, the water masses
could not be so easily separated on the O-SiO 2 plots, however, it was shown that the low silica water did have a significant effect on Q-SiO 2 relationship as far west as Cape
Hatteras.
To illustrate further the geographic variability of the
silica distribution, four potential temperature* surfaces
were chosen upon which to prepare contour maps of the silica
distribution (Figures 24-27).
The data base used in preparing these maps is presented
in Table I.
For potential temperatures less than 1.9*C one cannot
trace water continuously throughout the North Atlantic Ocean.
Figures 8a,b (Worthington and Wright (1970)).
Between 1.8*C and 1.9*C the bottom waters from northern
and southern origin come in direct contact.
*Due to the tight 6-S relationship in the deep western
North Atlantic Ocean for 0 < 3.0C (Wright and Worthington (1970); Worthington and Metcalf (1961)) potential temperature surfaces nearly coincide with potential
density surfaces.
5.1
The 1.90 potential temperature surface
The distributions of dissolved silica on the 1.9*C surface (Figure 24) shows the general southward increase in silica
noted by Metcalf (1969) on the 2.0*C surface (see Figure 1).
Along the mid-Atlantic Ridge, a tongue of high silica
water is found to extend northward.
This is AABW and was
identified by Metcalf (1969) using the same IGY sections.
In the vicinity of the tail of the Grand Banks of Newfoundland, there is a sharp silica gradient, with the lowest
values inshore, following the bottom topography.
One should
note that this sharp gradient cannot be traced westward around
the tail of the Grand Banks.
The lowest contours (<22.5 pgA/L)
are only resolved by a few observations and our interpretation
of this as a continuous feature may not be justified.
Since
none of the very low silica water penetrates around the tail of
the Grand Banks, it is not associated with the WBUC, but rather
with the Denmark Straits overflow water that is usually observed
to be confined to the Labrador Basin (Figure 2, Wright and
Worthington (1970); Worthington (1976)).
Away from the contin-
ental rise the silica values are uniformly high (>35 yg A/i) and
join with the plume of AABW extending north.
It should be
noted that 1.9*C is the potential temperature at which the
O-SiO2 and e-S (Wright and Worthington (1970)) plots show the
joining of the DSOW and AABW (Figure 5).
West of 50*W, the silica distribution takes on a different character.
There is a decrease in the silica values as
one approaches the continental rise but the sharp gradient of
the DSOW is not present.
After passing around the tail of the
Grand Banks, the 30 and 32.5 pg A/Z contours diverge leaving a
large area with little change in the silica concentration.
This is the location of the deep anti-cyclonic gyre proposed
by Worthington (1976).
It should also be noted that the 30 pg A/t contour
follows the topography of the continental rise until it disappears west of 70*W.
It is commonly observed that the lowest
silica contours will follow the topography until they end by
running into the coast indicating some mixing process
(Figures 24-27).
5.2
The 2.2*C potential temperature surface
On the 2.2*C potential temperature surface (Figure 25)
the general north-south gradient is again present with a zonal
gradient with the higher values of silica being found along
the mid-Atlantic Ridge.
There is a sharp gradient near the
tail of the Grand Banks with the lowest value nearest the
continental rise.
This gradient may be traced along the
continental rise and westward around the tail of the Grand
Banks.
In contrast to the 1.90 C surface, the sharp gradient
along the continental rise may be traced further west to
650 W.
As it was previously defined, this is the WBUC trans-
porting water southward from the deep Labrador Sea (.Grant
(1970)).
West of 65*W the gradient weakens.
The 20 pg A/Z con-
tour runs into the coast and the 22.5 and 25 pg A/t contours
diverge.
Similar to the 1.9*C surface, the silica contours
on the 2.2*C surface, away from the continental rise, near
the Grand Banks diverge as one proceeds westward.
25 and 27.5 pg A/k contours diverge.
Here the
The waters above the
abyssal plain are characterized by a nearly uniform distribution of silica.
5.3
The 2.4*C potential temperature surface
The 2.4*C potential temperature surface, Figure 26, is
similar to the 2.2*C surface.
We see once again the northwest-
southeast gradient with the highest values of silica found
along the mid-Atlantic Ridge and the lowest along the continental rise.
One can trace a sharp silica gradient from the tail of
the Grand Banks westward to 65*W where the 20 and 22.5 pg A/
contours then diverge in a similar fashion to the 22.5 and
25 pg A/Z contours on the 2.2*C surface.
Also similar to the
2.2*C surface, the 22.5 and 25 pg A/Z are close together near
the tail of the Grand Banks but diverge when traced westward
beyond 50*W.
Thus it is found that the waters above the
abyssal plain nearly uniform in silica.
The inshore gradient on the 2.4*C surface is not as sharp
as that seen on the 2.20 C surface.
As shown on the e-SiO 2
plots (Figure 23a for example) the spread between the WBUC
water and the remainder of the NADW increases with decreasing
potential temperature.
5.4
The 3,0*C potential temperature surface
The 3.0* potential temperature surface (Figure 27)
shows a departure from the contribution on the previous surfaces.
There is the expected southward increase in silica
but at this potential temperature, there is open communication across the mid-Atlantic Ridge.
The data indicate a
sharp gradient near 32 0 N with abrupt changes above the midAtlantic RIdge.
These features are not well resolved and
shall only be pointed out here without further comment.
Near the tail of the Grand Banks, and along the continental rise, the sharp gradients seen on the previous surfaces
are no longer present.
The distribution is nearly uniform
throughout the entire region north of 35*N from the Grand
Banks to Cape Hatteras.
This was shown also in the O-SiO
2
plots (see Figure 5) by the narrowing of the Q-Sio2 scatter
for 6 > 3.0*C.
VI.
Discussion
6.1
The 0-SiO 2 relationship
In the preceding sections a silica minimum (on surfaces
of constant potential temperature) was found along the continental rise.
This feature has been associated with the
WBUC (Richardson (1977); Clark, Hill, Reiniger, and Warren
(1978)).
It has been demonstrated that the low silica
water can be traced from the tail of the Grand Banks of Newfoundland to Cape Hatteras.
However, it was also found
that the 6-SiO2 characteristics of the WBUC do not remain unchanged throughout the region of study.
There is a marked
change in the 6-SiO2 relationship to the west of 650 W.
The
data suggest that a narrow band of low silica water comes around
the tail of the Grand Banks and extends to 65*W along the continental rise.
nounced.
Further west, the silica minimum is less pro-
The two branches of the O-SiO2 plot seen at 50*W
are no longer distinguishable at 70*W where a more diffuse
silica pattern exists.
At 50*W the low silica branch was
8-10 pg A/k lower in silica (for 0 = 20 -2.2*C) than the high
silica branch.
When compared to the high silica branch at
the KN 48 (650 W) station (or equivalently the high silica
branch of the HUDSON section, see Figure 22),
the TRIDENT,
AII 100, and CH 20 sections all show a 2-4 pg A/
silica
deficit in the same potential temperature range.
The process(es) responsible for the alteration of the
0-SiO2 relationship west of 65 0 W is not a simple one to identify.
The distribution of dissolved silica has been described,
but it has not been related to the flow field.
It has been
suggested that since the low silica water undoubtedly
originated in the northern regions, its presence in a narrow
band along the continental rise implies the existence of a
WBUC.
6.2
Geostrophic transport measurements
Several observers have made hydrographic sections and
computed the geostrophic transport of the WBUC.
Richardson
(1977) compared the results and Table III is reproduced from
his work.
A wide range of values is found.
How much of the
variability represents changes in the mass flux of the WBUC
and how much is due to the choice of stations or the method
used to establish the absolute velocities is not known.
Richardson (1977) has described the methods used in each computation and suggested possible error sources.
All of the
data listed in Table III came from sections to the west of
650 W.
There are few available calculations of the transport to
the east.
Clarke, Hill, Reiniger, and Warren (1978) found
no evidence for the WBUC at 50*W but they suggested that this
was due to unrealistic reference velocities used.
The current
meters used to adjust the geostrophic velocities were probably
outside the WBUC (see further discussion below).
Worthington (1970), in his model of the circulation of
the North Atlantic Ocean, allows only 6 x 1016m3 /s
to flow in
the WBUC, although he suggests that the flow may be unsteady.
Although Worthington's 6 x 10
+6 3
m /s is smaller than most
of the observations to the west of 65*W, it is not felt that
any of the measurements are precise enough to be compared.
Collectively, these data do suggest that the WBUC is present
from the Grand Banks to Cape Hatteras.
It cannot, from the
available data, be determined if any changes in the transport
are occurring to accompany the changes in the 0-SiO
2
relationship.
Table III.
(taken from Richardson (1977))
Volume transport estimates of the Western Boundary Undercurrent.
The value in parenthesis is the r.m.s. deviation of the innividual
transports about the mean.
Measured by
Swallow & Worthington(1961)
Transport
6 3
(10 m /sec)
Latitude
Date
7
33
Mar 1957
Volkmann (1962)
50
17
38
38
Jul 1959
Jul 1960
Barrett (1965)
4
12
35
35
Oct 1962
Oct 1962
Worthington & Kawai (1972)
2
35
Nov 1966
Richardson & Knauss (1971)
12
35
Jul 1967
Amos,Gordon & Schneider(1971)
22
31
May 1968
Average
16 (14)
6.3
Moored current meter observations
There have been several current meters moored in the
WBUC.
Along 70*W steady westward flow was observed for
eight months on the upper continental rise (bottom depth
<4000) (Luyten (1977)).
Off Cape Hatteras, records less than
one month long show flow to the southwest directly beneath
the Gulf Stream (Richardson (1977)).
Recent data from the
Blake-Bahama Outer Ridge indicate the presence of a strong
westward flow (Rhines, personal communication).
To the east,
at 50*W, a two-month record just to the south of the tail of
the Grand Banks shows a predominently westward flow.
In gen-
eral, however, the current meter data along 50*W and along
another line to the north of the Grand Banks running southeast,
show little evidence of the WBUC.
Clarke, Hill, Reiniger,
and Warren (1978) remarked that "...our velocity vectors do
not show a definite westward flow along the continental slope
in conformity with the unequivocal evidence in the deep silica
field for the western boundary undercurrent."
They suggested
that the disparity between the current meter and hydrographic data
was due to a difference in the time scale to which the two
types of data respond.
While this may in fact be true, an
alternative explanation is offered.
In this case, it appears
that the current meters were placed outside the WBUC.
Some
idea of the location of the WBUC may be gained from the deep
silica data (Figure 28).
The 2,0*C potential isotherm has
been added to and one can identify the WBUC where the silica
contours sharply turn downward and cross the isotherm.
Note
also the slope of the 2.0*C isotherm indicating the presence of
deep shear.
Along the section running southeast (Figure 29)
none of the instruments were in the low silica water.
Along
50*W (Figure 28) one instrument is located clearly within the
low silica water.
The record from this instrument shows a
predominantly westward flow for two months (Figure 30).
Thus
we feel that the disparity between the current meter and hydrographic data is due to the location of the instruments outside
the WBUC.
Fortunately, it is along 50*W that the silica data provides the clearcut evidence for the WBUC.
Along 70*W and
near Cape Hatteras, the silica data alone suggests only that
the WBUC has had some influence but the current meter data and
the geostrophic transport calculations offer evidence that the
undercurrent is present.
It is from this observation that one
may anticipate that the change in the silica distribution is
not due to a fundamental change in the character of the flow,
but rather is caused by some external effect which results in
an exchange of water between the WBUC and the central basin.
6.4
Enhanced mixing west of 650 W
The current meter data may be used to offer a suggestion
as to the nature of the process which provokes this exchange
of water particles.
Current meter measurements along both
70*W and 550W indicate that the
floT.?
ovor the abyssal plain is
characterized by energetic eddy activity (Schmitz (1976)(1978);
Luyten (1976)).
The effect of these eddies can be seen in the
silica distribution by the nearly uniform silica level over
the abyssal plain (Figure 24).
When the WBUC rounds the tail of the Grand Banks, it follows the topography westward and effectively skirts around the
deep eddy field.
After passing through the New England sea-
mount chain (near 65*W) the topography steers the WBUC southward and forces it to flow directly under the Gulf Stream and
presumably into more vigorous contact with the deep eddy field.
Thus the topography steers the flow into a region of enhanced
mixing which exchanges the low silica water along isopycnal
surfaces with the surrounding North Atlantic deep water, thus
eroding the silica signal and providing the smoother distribution observed.
VII.
Summary
The distribution of dissolved silica in the deep western
North Atlantic Ocean has been presented.
It has been found
that the dissolved silica content can be used, in conjunction
with potential temperature, to identify water masses.
For potential temperatures less than 1.9*C the dissolved
silica potential temperature relationship is similar to the
salinity potential temperature relationship.
Both show two
distinct branches which identify the Denmark Straits overflow
and the Antarctic bottom water.
In the North Atlantic deep water, where the 6-S relationship forms a tight correlation, the 6-SiO2 relationship shows
a wide range of values.
It has been shown that a silica mini-
mum, on surfaces of constant potential temperature, can be
traced along the continental rise from the tail of the Grand
Banks of Newfoundland to Cape Hatteras.
The low silica water has been termed the Western Boundary
Undercurrent (for historical reasons), which has been interpreted as the mechanism by which water is transported from the
Norwegian and Labrador Seas toward the equator.
Given this observed silica distribution, one is left with
a difficult task of interpretation.
The WBUC is the most
prominent feature in the distribution yet little is known about
the structure of the flow.
What roles do advection and diffu-
sion play in determining the observed silica distribution?
It has been hypothesized that a region of enhanced mixing is
present west of 65*W.
It has been suggested that this enhanced
mixing is due to the presence of the deep eddy field observed
with moored current meters.
Appendix I.
Rejected Data
The data collected on KNORR 12 has been rejected because
of anomalous 0-S and 6-SiO 2 characteristics.
This was unfortu-
nate since the stations were located in the region between 55*W
and 650 W where there is no other data (Figure 31).
The
-S relationship from KN 12 shows a consistent bias
toward lower salinities when compared to neighboring sections
(Figure 32).
The 6-SiO2 relationship shows a consistent bias
toward higher silica values (Figure 33).
35
It may be argued that the low salinity and high silica
values imply an intrusion of water from the south.
However,
since these anomalous conditions were observed in the data
from only a single cruise, it is felt that they should be
suspect.
ACKNOWLEDGEMENTS
An advisor who is willing to give advice when sought and
support when needed is extremely important to a graduate
student.
Jim Luyten has been such an advisor.
There are many people who have helped in the preparation
I
of this thesis and in my training as an oceanographer.
would like to thank Val Worthington and Mike McCartney for
teaching me the joys of collecting hydrographic data.
Thanks
are also due to all those who helped on ATLANTIS II 100.
Val Worthington and Bruce Warren generously allowed me to
use their unpublished data.
I would like to thank Doris Haight for typing this
manuscript.
Finally, I would like to express my thanks to the other
students in the WHOI/MIT Joint Program in Oceanography who
have been a constant source of encouragement and enjoyment.
I am particularly indebted to Nan Bray, Dean Roemmich,
Neal Pettigrew, and John Toole for being patient whenever I
took to roaming the halls.
BIBLIOGRAPHY
Armstrong, F. A. J. (1965).
raphy, Vol.
I.
Eds.
J.
Silicon.
P.
In:
Chemical Oceanog-
Riley and G. Skirrow.
Broecker, W. S., T. Takahashi, and Y.-H. Li (1976).
of the central Atlantic - I:
Hydrography
The two degree discontinuity.
Deep-Sea Research 23, pp. 1083-1104.
Carmack, E. C. (1973)
Silicate and potential temperature in
the deep and bottom waters of the western Weddell Sea.
Deep-Sea Research 20, 927-932.
Clarke, R. A., H. W. Hill, R. F. Reiniger, and B. A. Warren
(1978).
Current system south and east of the Grand Banks
of Newfoundland, in press.
Cooper, L. H. N. (1952)
Factors affecting the distribution of
silicate in the North Atlantic Ocean and the formation of
North Atlantic deep water.
J. Mar. Biol. Ass. U. K. 30,
511-526.
Nutrient Distribution
Corwin, N. A. and D. A. McGill (1963).
In the Labrador Sea and Baffin Bay.
U. S. Coast Guard
Bulletin #48.
Luyten, J. R. (1977).
Scales of motion in the deep Gulf Stream
and across the Continental Rise.
Metcalf, W. G. (1969).
Atlantic.
JMR 35, 49-74.
Dissolved silicate in the deep North
Deep-Sea Research 16 supplement, 139-145.
Richards, F. A. (1958),
Dissolved silicate and related prop-
erties of some western North Atlantic Caribvean waters.
JMR 17, 449-465.
Richardson, P. L. (1977).
On the crossover between the Gulf
stream and the western boundary undercurrent.
Deep-Sea
Research 24, 139-159.
Schmitz, W. J. (1977).
On the deep general circulation in the
western North Atlantic.
Schmitz, W. J. (1978)
JMR 35, 21-28.
Observations of the vertical distribu-
tion of low frequency kinetic energy in the western North
Atlantic.
JMR 36, 295-310.
Stommel, H. M. (1957).
A survey of ocean current theory.
Deep-Sea Research 4, 149-184.
Sverdrup, H. U., M. W. Johnson, and R. H. Fleming (1942).
The Oceans: Their physics, chemistry and general biology.
Prentice Hall, 1087 pp.
Swallow, J. C. and L. V. Worthington (1961).
An observation of
a deep countercurrent in the western North Atlantic.
Deep-Sea Research 8, 1-19.
Volkmann, G. (1962).
North Atlantic.
Webster, F. (1969).
currents.
Deep current observations in the western
Deep-Sea Research 9, 493-500.
Vertical profiles of horizontal ocean
Deep-Sea Research 15, 85-98.
Worthington, L. V. (1976).
On the North Atlantic circulation.
The Johns Hopkins Oceanographic Studies 6, 110 p.
Worthington, L. V. and W. R. Wright (1970).
North Atlantic
Ocean atlas of potential temperature and salinity in the
deep water including temperature, salinity, and oxygen
profiles from the Erika Dan cruise of 1962.
Oceanographic Institution Atlas Series, 2.
Woods Hole
58 plates.
39
Worthington, L. V. and W. R. Wright (1971).
a paper by X. LePichon, S. Eittreirr
Discussion of
and J. Ewing, "A
sedimentary channel along Gibbs Fracture Zone."
JGR 76,
6606-6608.
Wright, W. R. and L. V. Worthington (1970).
The water masses
of the North Atlantic Ocean; a volumetric census of
temperature and salinity.
8 p.,
7 plates.
Ser. Atlas Mar. Envir. 19,
00040
300
60..
10
r
--25
20.
30
45
55
5o
460
Fig. 1. Contours of dissolved silica (ugA/l)
on the 2.0' c. potential temperature surface.
(taken from Metcalf (1969))
I~
0
1
I
I
I
I
I
I
I
I
3.0
I
I
00
0o
2.8
0
2.8
0o
000
2.6
2.6
*0
Se
2.4
0
2.4
0
0
00
22
2.2 0 1
00
0
0
.0
0
0 0
2.0 -
0
0
00
*
C0O
o 0
1.8
0
0
I
34.88
I
.90
S
S
0 1
0 000
0
1
0
o
I
*
0
0
0
*
000
I
.92
Salinity %,
I
.94
I
.96
12
16
20
24
28
32
Dissolved SIlicate pig / atoms / /
Fig. 2. Plots of potential temperature vs.
salinity and potential temperature vs.
dissolved silica for stations north
(open circles) and south(filled circles) of
the tail of the Grand Banks.
(taken from Worthington(1976))
0
* S
36
~
40
-.8
Fig. 3. Locations of stations used in this
work. Stations denoted by * were used only
in the preparation of the charts of silica on
potential temperature surfaces.
80*
70*
60*
-'_
-
1
I
40*.
.50
_ _-_'__
.l__
_ _ __
L
*
**** *
0
0 *.iy
6
**
.0
*
0
*
40
*.
0
qk.
2.
'~0*
.*
.0*
*
0*
*
.5
j:
0
*
0
*0*
4'.
*SS0*~
F
*
.1.
'et'*
I
3
*.~*.**
02.
00
'C .5.
~.
.- 4.
I-.
20
ATLANTIS 265
0
5
5
*OofSO
.
~
Os
0
*
,~
'Oq
**
0
:F
*
* * 0-.
*a *4
0
I'
S
*
*~Al
'S
*
S
..~
0
~
S
0
*
0
0
0
*
'
$*
*.
.'
0
'
*
*
.--
0
0
*
V
.
DisO.VE
RYi G.Y 2
I....
JAJ
*
*2
30*
*0
0
e
e
0
*
e
30'
-'I,
-'0
DISCOVERY I
LGY.
2'
0
t..O
SO'
7.'.6.
70*
60*
0
0
'50*-
.Q...*.0
-40*0
20*
043
Cc
4.0
.
-
t..J
C-O
3.5
.
.
zO
3.0
I.-
e
-2----5-
-
.-.
2.5J
2.0
.
X
a
--
.-
bt
.*.-.
*
Fig. 4. Potential temperature vs. salinity
from those stations indicated in the inset.
: Heavy lines ipdicate the bcr.ndaries of the
North Atlantic Deep Water (Wright and
Worthington (1970)).
3485
34.95
34.90
- SALINITY
(%o)
s
3500
000C44
.
.,
.. **-
..
-*
s-.
-.
*S~b
,
..
o
3.0.
-
.*
CL.
81b:.*
,
..
-
-
..
-
-
:..-
.
3.5.
4 z2
.
:-
-
-
*
-
*--.
.
.-
-.
.-
..
-
.
-
o
.
silca
ro
inet ...
-*
2
t.-
-.
.
-.
-
...
.:
.
--
.*-..-.-
:. . .I.-.----
.. --.
..........
LOi 3.0
.
-
.-.
-"
.
,
P%
4.0
-
.-..
.
.
-
ths4ttosidctdi
h
.
-.-.
*
I
.'.
5.
*Fig.
Potential temperature vs. dissolved
-sili.ca from those stations indicated in the
0~~~
*.*
-
07. 0
*~
00~
o..
DISOVE
4.
30
4.
20
SIIC
(gA1
t.... * S
...
,
.
ee
....
0
3.51
*..
1,
~1-I--j
Ii I [I
Al
. ....* :
.*
1.,
0
LiJ
eS. ...
.
-
.
-*..--..-
goe*
*:.
CL
i
.
-
e..-.
.
*
*
. *
-
-*
0
0 to
...
... .
.
. - *
.
'a
*..
*0.
.
*.*
*,..
I-:
es*..
*
-.-.
-
.*..
--
1*
-
CL
.a
.....
2.5
-
0
a
-
.
. 0
..
.
0.
Fig. 6a. Potential temperature vs . dissolved
silica from stations south and wes of the tail
of the Grand Banks. (see inset)
0
2.0
.
*.
a
5
0
*
-;
*
SS
0
.I.7
DISSOLVED
SILICA (ugA/I)
PLATE 4
34.70
34.80
34.90
4.0
2,e".
-13.9
35.00
69
27
35.10
' I 19
.t i.59
Z;
-2.4
.]
g2.0
0
a- L5
-1.0
0.5
18
12
2426
/0476
14991
.X6
553 5 S960 3
0.0
452 0
34.90
34.80
SALINITY (e/*a)
NORTH AMERICAN BASIN
A Volumetric Census
.
The
nurher
n
.Thefiatc
. m')by
fletcaf to*) at the rightof exh
numt-%er
rin'
f
0.S. *3..4
7 3
-*.J
diagraem
gve tre
vorum,subterilsx
41to
rv. a
d.h .ag~ ta nd thk
of eacht
"n 20* Isoesin the warmwater diagram. g-se the volumesub-otals
no.s aboveIle
(0 10, hr'") bySa'-mryrrv'
hr.Ior-fa
10'
tor
. ,.
per
ore
c,,d rms
ton
(ci!!
of .,..e (, 10' k')
onthe
)
rm
On
tr d
i each dgrm aCs shown
m 4ta14s at
ithe
e reresent
cold
ese
-35
w~a'er
diagram
the
curves
4*-3.
r
ght
m
34*Ce-76%0
.34 4'!.
r.. 4*-30'
...
90%of total volume
16es'-30'
"
I3968 7
64
3
940530
42251
cententers
showi
potential
thermoveic
Frc~a
y
Voulo1i
ENVKON%
CENT
ATLASOFTHf.MAR.%F
SERIAL
Fig.
the-l-
4*-30.
50%of total volume
75%
of total volume
anomaly
thermosterc
16'-Y-281081
26/02'
" 0
at the tutto-n
*Thefi.C mumbe.,in thehCOr.znlrow
.The
7t
, thoandsof cu.ctkom!ers, in at
of .ster
iach bo*gvs rhevolume
7a.
(taken from Wright and Worthington (1970))
tPv .
G.aGAMUA.
. P'"
Ci.T
NJ
ASPA.J01.A C LORAPCIUCAZ.6'ACUXY.
I"
PLATE 2
34.70
4.0 1
u
34.80
.
2
34.90
.
35.00
.
v
35.10
,
-
1
'
.
1
.
.
4.0
*2A
- 2.2
2.0
z
0
IaI
~7f/~JX
LABRADOR BASIN
A Volumetric Census
be we 1chbo.
.Th n
.I
at the
a numes(eiclrw
u-
he
h,j
bly
.
krr)-
tne
t
t ro
the hoN
10* and20' I---. . th. Aamea
Illy
sa,
by
WIa "-."m
of
brees
Vr-Ior Ict'tI)
ach
onl, the
Olt
Srfit AL. TLASOf I ti
thousands f
sc i
d
.
too
1,.'
.
r.
o,,eI.
aN~d
-Thec.ed
e
rght o'
n
w.ter
of
gve
cub.
e s
.1-e
fle
1ometers
m th
:~
the
L
--
Mlle
(0
*
water
co1
=A4
.'A:%I
F
.ce d
16'
d
1
U
of
r.
.h
the
eVch
d
r
voi.
d..,:-,J
thel
%if
.I
ST
curvs
e
te
sbt.ots
e,
75%
In
shown
L.Lj
50%of total volurne
horzortal
0-4*. 34 10%71-35.16%1
a-A'.
-34 70-
~
4'-30.
3h 4%.
-30.
of total volume
stt.:s a*
the,
lower
9011 oftotalvolum-
1693
106564
7..' 0-4*
~
10487/i
-37
33?8
6%.
*34 4%.
4
175
Ters s'- xr33145
9
140X
3
t'ht
1er4 o
-3X7
ceriftaf-S
.srne'ectemseramay
d.gram
a-d
):.
10
(.
Ato'a-s
tervah.
temfreature
.c
."
rwsabove th,
(1 10
vlme
res tt.e
Sh.
PotNb1.s'
tl
-rr
s'in
a
Im
POLIOIa
Fig. 7b.
(taken from Wright and Worthington(1970))
ALRLP1CA7CLIbRCAJ1CA1
MDITI
,
I9n
V'IA 1: V)
34
.11
42
.. 4,
c
s
SALINITY (S..)
-,
-
-
--
AT TlE I
C POTr.NTIAL 1!.MPIRA-1URV
St-l
ACV
Fig. 8a.
(taken from Worthington and Wright
(1970))
00049
PI A I
e5K 9)7
92
040
7<..
Yf.
SA LINITY (/,)AT T HE
Fig.
1.9'C
r
POTENTIA L 'lLMiPi RATUIP.E. SURl'A CE
8b.
(taken from Worthington and Wright
(1970))
F 21
I
4.
**
*
.
-
*-
.
-
--
*
--
* .*a
*
.
.y
-
~000~l
1
0
0
*00
0
--
*.. ....
3.5
0
04
.
*
LiiJ
*
-
0
*00*
0-
-..--
.
-
-
--
0
.0
S
000
~0
cr.
*
0%
Lii
.0
0
3.0
0
-LJ
0
0
0
0
a
*
00
00
0
0
0
0
0
0
03
0
0.
0.
0
.0
0
00
'0
0
***
0
CL
tail of the Grand Banks.
a-0
1*..
*
:
.
DISSOLVD
.
2.0
170
*
.
.
Fi.6.0oetiltmprtrev.disle
**
.*
DI
S
*'
00
o~~
frIC stain)ot
e
0
-
.
-
1-
0
.
(see inset)
u0
n
atoh
(A
.
s
.
0
0
0
e
S
00
M
0 0
I-..
00
0
00
0
'7!
*
?6
CA
iD
OM
oCrt*
o.~
et*
09e
,
0
*
0
0
00
0
*0
000
0
0
tu
*or.
0
I
0
...
SO*.
0
,
*
'
2
~.
1'
e
I.
70
0
0
~'
iij~!
.
0
60-
*
49.
0
0
LII I
*e
0
*wg.*#vtIawg
*
I I
~4Li
A
It,
~
.6*
0
Og
0
0
**
0
Ij
0
I
o
.
~00
0
.
%10ot0
IV
0
6
6
.0
.0 a
*
0
I II
~;*~40~
*i
;:*
0
0
40*.
9
79L
II II I I I III
.50
*00**
0
.I
.3
I II I
e
0
0
POTENTIAL TEMP.ERATURE. (DEG. C.).
1-
I I
0
3V.
e
0
0.
-. e
0
0
,
90
*
0
I
1
i.Ai h%
0oe
..
4.0
*
ATLANTIS U[ 18
gO
-
S*.
*
55
0
#4
St..
S
*
e.
-- *....
:..
eS
3.5 t
5
.........
*,..
..
-e
.
5iy1so
e4-e
-*.
e%
*~4u~m4*.
*
e
*
0
,0
5O(
b
e
L
.
--
**
.5-
w..:
3.0
1.
...-
w.
9.1*%.
*
-I#
S
SQ
0
-j -
Z
0
-.* ".
0
Q..
0
-2.5
*
S
Fig. 10. Potential temperature vs.
dissolved silica from ATLANTIS II 18.
*a%-
S
*
~
0
OS'
~0
OS.,.
2.0 t
*
1.7
-
1o
.3
S
S
1
5
*
~
.
*0
20
***~*
:I.I:,~IE~.s:~S
40
30
oIS SOIiVFD
SILICA
(uA/I)
0
4.0
,,
s
.
..
..
-.
.
e
-I
* ,.5
I I I j I I I I*I I 1.1
Ij
I I I
1 IL
55
a
..
.**.
.*....
--.-.
0
.............
-
40
.*
0
s
.
0
o
*
.55
Fig. 11. Potential temperature vs. dissolved
silica from those stations indicated in the
inset.
0
--.
0
555
S
.s
5.
St
S...
~
24ft
145
5
1.71
K0
a
I
20
DISSOLVED
4U
30
SILICA
(ugA/l)
55
to
,-)
In
W
0
U
1
'5*0
0
0
9
*
9
@0
0~~
9
*er. 9
0
.*
e.
9
-4q
*
0.9
99
e
9
99
.
6
9
0*
CD
30*
-
e9se
e
.
V.
e9
U'
I\
99.9.
-
-
55,,
T'
.
.-
.
-
.
**
T-70o .
5 ',
-
*--
700
e
*-0
;
0
5
5
-
.
i43
00*
.
'
o.
1 ~
*
.
**'.
*
*S*
* S.
** .,:
3
*~
*
.5
6
54
5..
~'*
-~
i'i:
;:
0
*9
.5
S
*
*~*
.1.
*Pi44lS~14Vj
0
2
to
0
.
,
e0
*
0
*
*
5
5
0*~
U.
5..
I I~1~~I111111 1
*~*
.55;.~
5,*,*:
~..
2
*5
.so'.I-I
a
.
.
300
.
.
.9
*
-4o.
ti lli fl~ I I I I I I I I 'ti I I I i
.
'.
9
0
TEMP.ERATURE. (DEG. C.).
- -
Cw
so",
*
go-.
0
POTENTIAL
.
N.
..
.
*
U
/
- It
a
9
SO
~p.
S.
S
S
KNORR 48
t0e
4.0
* %
alt.
a
S
S
t
*
.
1*
*
..
16*
hilL
I I I Ii 1 1.1T* j.A I Iii
10
-I I'I I
.Si
0
0
0
5* - ....
,
S
..
-@
.
*
*..
*. .
.
.
.
.
.
.
*0
3.5
.
*0
I
..
*
-5
C-)
CD
Q
S.-,
* e
3.0 t
e
-.
.3.
*
*
...... S.0
Ta.w.SSaC
-
t
*
- - --
.
-3.
C
w
aw
S
e
z
S
~0
0
.Y.
--
-S
w
F0
a_
-
1
S
e0
S.
...
Fig. 12b. Potential temperature vs. dissolved
from KNORR 48.
*e
esilica
*
S
S
2.0 P
S
4~3*
S.,
S
0***
s-.
S
1.71
-10
20
30.
DISSOl VFD -SIL I CA
40
(uaA/1)
Cr>
ni
C
1~
(dl)
Cl)
'Q
oI
-9
.9
9
0
9
9 .
8
0
Hd.
(D
Zrt
H
Z tQ
"0
0~
0
0
0
9
-
0
9
9
,
'
POTENTIAL TEMPERATURE
9Ot0
*
Od.
0
*
'0
g9 *0 *o9499
*9
0.00
9
9
(DEG. C.)
0e
So
.
0*
009
z
0
0*
N
S.
,
.
007
ATLANTIS IE 100
S*
L~FVYT}FFITh[II
1 A I L I I I I I I I.-
. f.1 l I i .ii
S*
.
S..
;*.
3.5ti
..
,
.
. ..
Q)
...
S
*55
S
S
S
go
WI
@5
S
55
S
S
-
b
-
.
-
tu
i 6F I-.IP
-~
1
Jr
.
.1
-~
%
..
*
*
S
-I
>-i
WJI
i
0
S
S
9
*
..
-
C
Fig. 12d. Potential temperature vs. dissolved
silica from ATLANTIS II 100.
5
S
S.
9
*
5
Se
*
S
*
9555
S
2.0k
S..
5
*
S
*e
-*
*
1.71
- 0
201
DIS SOLVED
4U i
SILICA (ugA/1)
N
C
0
r
Ci)
r
-
Of
of
y
C---o
-0!
.9e
0
9
9
9
*
*
0
90
0
*
0
0
O
H
Vt
'
(ne
o D
(D
'S
O
Sm
H
**,
.N
-30'
*
9,
so'--
0,@
0~
*0
e,,
-
o-
-
.'
-
0
3
9
.
0
0
'V.
'"
9
s'
*0
*
9
S
gee
(DEG. C)
*
0
*9
5
so-.
0
'I
0*~,
"
'
11111111111 I Iii
- So-
0
0
POTENTIAL TEMPERATURE
S*
.'.''.
IIIj
40
*
eS~
0
0.v0
*
1
.~
liii
*--
...
0
4~
0
0
9.
90
'o
*
o
-4
?0
0,
*
0S
*0
48.1
Fig. 13a. Fourth order polynomial fit to
data from HUDSON (50*W).
32.11.
*
2411.
2I.',
U)
.J
O
U)
++
++
+
0+s
++
++
12.0
OI
4.0
-..
go*
C449%4
POTENTIAL
TE MPERATURE (DEG. C.)
rC4
V!
c"
0"
c
DRRD
ERN
ZED
Fig. 13b
* -*
Nrmalized
REM I DL R .L
r-eiuals
t~o the fourth
order pblynomial from HUDSON (50*W).
Residuals are normalized by the standard
deviation. (see Table 2)
3.+
2.3
+ +
-t
s.,+
+
+
++
+
'm+.
++
#
++
+
+
TE M
++
+
ri
onrftW
"
POTENTIAL
&A
TEMPERATURE (DEG. C.)
-'
Fig. 14a. Fourth order polynomial fit to
the data from XNORR 48.
40.03
32.5
20.0
+
+
0
(I)
20.0
++
0
Ur)
.Cf
12.0
3.5 4~
3~.
0-
g~4
POTENTIAL
94
01
%
ms
1"
g"
T EMP ER13ATURE
(DEG. C.)
ft
I ZED
ETRNDRRD
RES
I DURLS
Fig. 14b. Normalized 'residuals to the fourth
order polynaomial fit from KNORR 48.
2.0
1.44
±-t.+-
pop.~
+
6.3
4+
*p
04.1
Im .
04
1%
0"
A
0
POTENTIAL TEMPERATURE
(DEG. C.)
S
+$.
Fig. 15a.
Fourth order polynomial fit to
the data from CHAIN 20.
20.0
32.3
211.0
."0. 23.3
on
0"*.0i
.POTIENTIAL
TEMPERATURE
(DEG.
C.)
ES"o RN
E>RRC
I Z E
D U
=1L
Fig. 15b. Normalized residuals to the fourth
order polynomial fit from CHAIN 20.
*
3.0
a.0
3.3
+
+
+
0
++--i
I)
-2.0,
-4.0,
4.oa
0-4
roo.
=I
8%d
PC) TEN TI AL TE MPE RATURE
(ugA/1) .
Fig. 16a. Fourth order polynomial fit to
the data from ATLANTIS II 100.
35.
0
23..
24.5,
-J
Id
0
r)n
ClD
a
15.8
12.0
S
1'10
Pd
Id'0
Pd
POTENTIAL
0
m0
1'4
0
19
lYl0
19
Id'0
19
TEMPERATURE (DEG. C.)
0
19
m-i
19
0
STRNDRRD I ZED
RE
I DURL
Fig. 16b. Normalized residuals to the fourth
order polynomial fit from ATLANTIS II 100.
".3g,
3.1.
2.1.
I .3
.. +
p.'.
-1.3
-6.33
I'.
3~w
V.;
r4I
ATURE
N1TFALET
PO-N T AL c E MPER
(DEG. C.)
P
Fig. 17a. Fourth order polynomial fit
the data from TRIDENT.
:zz
0'
to
32.34
:3
0
7A. N.
+
U)
0
Lu
20.'
-J
0
U)
"1)
0
son
4.R
311
I~.
POTENTIAL
r'4
TEMPERATURE
(DEG. C.)
I ZE>
RES
I
>LJFR
Fig. 17b. Normalized residuals to the fourth
order polynomial fit from TRIDENT.
+
+
1.6
+
+
+
+
++
4-3.
-T~. a.1I
0-
r"
";
94
9
"
POTE NTI AL TEMPERATURE
(DEG. C.)
48.
42.8.
Fig. 18a. Fourth order polynomial fit to
the data from HUDSON (50*W) with the WBUC
observations removed. (see text)
30.3
0i
25.3.
0
CD
C)
(I)
12.3
IOU,
".3
3.3].
'I
_0-
W1
0-
rra3%
POTENTIAL
rq
r
TEMPERATURE
(DEG. C.)
m
I ZEC
STHNDRD
*4.
S
I DJFI
Fig. l8b. Normalized residuals to the fourth
order polynomial fit from HUDSON (500W) with
the WBUC observations removed.
3.3
+LA
+
+
++
+
+
+
-I
+
*i~!T~
+
+++
-2.0
-.
-
.
.
PC-TENTIAL
TE MPERATURE
(DEG. C.)
.
+
4L
Sixs.
218.3,
Fig. 19a. Fourth order polynomial fit to
the data from KNORR 48 with the WBUC
observations removed. (see text)
22.0
26.0
L0
0
CJ
28.3.
14.0
3.3 .1.
"I,
r4
4
ro
r"14
0"I
sus
POTENTIAL TEMPERATURE
(DEG. C.)
IZED
~TH~AR
-i-
14.g
MI
DUJRL.M
Fig. 19b. Normalized residuals to the fourth
order polynomial fit from KNORR 48 with the
WBUC observations removed.
+i
1.3.
3.3
+
+++
+
-
+
-14.0
-r.
a
V10
W-0
r4
10%AamN
p.
to",
.
94
r"
POTENTIAL TEMPERATURE
PI
(DEG. C)
84
c
Fig. 20. Fourth order polynomial fits to
the data from HUDSON (504W) and KNORR 48
with the WBUC observations removed. Two
standard deviation envelopes have been
included.
40
30
LU
0
-- J
20-
(I)
1.7
2.0
25p
0
3'5
2.5
POTENTIAL
TEMPERATURE (DEG. C.)
4.0
Fig. 21. Fourth order polynomial fits to
the data from CHAIN 20, ATLANTIS II 100,
and TRIDENT. The two standard deviation
envelope for CHAIN 20 is included.
40
30
C)
LU
20-
CfO
CD) 01II
.
17
-
2.0
-
'
-
2.5
3.0
1
--
4
3.5
POTENTIAL TEMPERATURE (DEG. C.)
10
-
Fig. 22. Fourth order polynomial fits from
.TRIDENT, CHAIN 20 and ATLANTIS II 100
compared with those from HUDSON (50*W) and
KNORR 48 with the WBUC observations removed.
The two standard deviation envelope from
KNORR 48 is shown.
30
30
0D
-LJ
0
20 0)f
1.7
2.0
2.5
POTENTIAL
3.0
.
4.0
3.5
TEMPERATURE
(DEG. C)
db
4.0
4?4
:3.5
.
Ld
ee.. . *
*. ..
H.-
.
.
CL.
2.5S
30
.
-
H
0.L
%.-
...*...ig.
I-
o
.
.a
o-,.--
-ig.a.-.
ntil
?*~~.
H
.*
2.5
*
*
.*
from those statior
SI0 The
60inset.
.
- .-**order
0
fourth
S
.~.silica
*- ' .
.....
e*
*
(with WBUC observations
two standard deviation er
-0
: %-
0-O
~0 i i i | '
Fig 23.Poetaltm
siiafrmtoe
tto
2.00
003
2.00
20
30
DI-SSLVEDSILIC
WWI
000?V 2
-
.
..
..-
Ij.I I I I Ii Ii I-
,
.
f.
.
.
cr_
O'
CL
Z
w
-I I I I I t I -I .I
- ..
0
CL
I I I r
- *
.
Fig. 23b. Potential temperature vs. dissolved
silica from those stations indicated in the
inset. The fourth order polynomial fit and
two standard deviation envelope from KNORR 48'
(with WBUC observations removed) is shown.
U
2.0l
e
.0
*-.'
o
DISSOLVED
-
30
SILICA (ugA/1)
4.0
-
23.5
.
.
.
.....
(j.
g
'
.3.5
c.r...
Uie
e
c>
n..
3.0
2.. .
- .
* -
,.**..-
-
*.
silica
.inset.
*
e
e .
*
.
-.-.
tp
.*.
*.
..
*..
-
.
two standard deviation envelope from KNORR 48
removed)
is shown.
-
*-
2.0
-
z
1.710
20
DISSOLVED
+.
from those stations indicated in the
The fourth order polynomial fit and
-(with WBUC observations
*.-
ue
.-
30
SILICA (ugA/)
-4
000
n0ise
0
HUDSON (50*W)
4.0
0
.
-
.-.
..
e. .
*.*
3.5*
*..
I
QZ.
Q..*.
H
DISO
* .-
5-
1.71]
0*
.-
SILIA
f....
o-
'
3
.
Potntial
F
2.5
0
3.
-
.5
-
,-
e
0*
4
* .
*-
*
.0~
tmperatue
eprtr
-s s
to stadard evito
(with~WBCosrain
iSshown
rem~ed)
0
6
--
0g
.0
*
-
-0f
-0
--
..
e ~
0
-0a
*0.*-
0
'
.
e-
*
.
-
-2-.-0
.
DSSLVD
isle
isle
KNORR 48
0.
,
20.
Poeta
olynoial
*
3-
.e..
-
.;-.
-*
.0.
Fig.23d
23d.
.i.
0
Q..
1
.
.2.0e
1.
-ILIA
-ug*I
-
e0
KNORR 48
*0
4.0
0
%0
. 0
s.
.
so*FF[
f..
e
.*
.
.
*3
*..
0
~
%
**
-1**.*.0
.0
0
3.0
0e...
Le
0
0
.
**.
g
*o
0
O
.0.
~
*.
..
***.
,
.
.0
L.2.5
0
elo..
~. removed)
Fig.
W
2.5
0..
-
>
S
Potential
temperature
from KNORR 48.
The
vs.
fourth
dissolved
order
polynomial fit and two standard deviation*
envelope from KNORR 48 (with WBUC observations
0
ff1
.-
is shown.
23e.
silica
w
KmhSe
20
00
.~DISSOLVE
3
ft
If~~*
.1ffffff~s
CA PotntAl)meauevs
Fig. 2
0
30
D*
S
40-3
.
g
-
isle
-.
.4.0
.CHAIN
20
e
00
*,
e
.
a-35
3.0.
0:.
.
*.
_j
*
<
--
.*4****
Z
O
2.5
2Fig.
.
*.
0
*
*
.*..
Potential
23f.
isilica
from
temperature
CHAIN
20.
The
vS.
fourth
dissolved
order
polynomial fit and two standard deviation'
0
S
*
envelope from KNORR 48 (with WBUC observations
removed) is shown.
e
*
.-..
*
2.0
.
*
*.
'10
20
-DISSOLVED
.
*7..
e
30
SILICA (ugA/)
40
09
ATLANTIS IE 100
4.0
SL
.*
3.5
00
*
.
-
*.C.
4.
3.0
0
*
2.5*
0.
Fig.~~~C
23g.
.
0.
1
Poeta
emeauevs
isle
.
%
KNRR4g( itABU/o s rv ti n
fromC
envelopeD
h.e.i*e
0
04
O
0
0..
%
**
-.*a
.0
i
iii
.0
.
CL
2.0
tj.J
wO
*.
*0
.ee
s
.1
si
.
.
fo
AN
If
,
* -
*4
*
Elope
envSOL
from (ugA/R)
48(ih.
Bosrain
*
0
TRIDENT
4.0
Ct'
.
00
0.
..
3 .
*..
,
*0g*
..
w
S
Q.
*
*
**
j-0
:~.
-6.7.
0
2.5
we
2.0
D*
10
Fig.
L...4
*
10
23h.
Potential
temperatures.
dissolved
.
silica from TRIDENT. The fourth orderpolynomial fit and two standard deviation
envelope from KNORR 48 (with WBUC observations
e
.
removed) is shown.
DISSOLVED
SILICA
.
4O0,
30
-20
-.-
(ugA/1)'
600
700
-1-1-17 71-1 1 1 1 1 1
..............
.........
..................
.....
...........
.............................
.....
Boo
..............
30
.........................
0,;
40"
5 0"
N
..........
................
..........
................
........
. .................
CD
.............
.... ..........
................
.
... .
..........
.........
...........
.
.......
........
..
V..
...........
............
..........
....
..........
.............
.........
........
.... . .. . . . . .. . . . .. . . . .. .
............
.................
.
.......... ......
. . . . . .. . . . . . . . .. ....
-00
................
........ .....
............................
.....
.............
...
............
......
..................
...
........
............
..............................
........
. . . . . . . . I. . . . .. .
..
..
. ..
.. . .. .. . . ....... x
..
..
..
................
..
......
............
............
........... . .
............
..............
...............................
...........................................
...
...............
.......................
..............
...................
...........................
................................
20
-.
'
80*
( * ,
Fig. 24.
*
60
Dissolved silica (ugA/l) on the
l.9C potential temperature surface.
50*
40
1
Poo
K
40yV
TFT0FTTyFFEF~FFFfTT
-0050
70
....
..7 ....
H.
00
. ... .. ......
...
.
.~~~~~~~
......
.........
.u./.).on
siic
Disle
. Fig.. ...25
..... ... poenia.t.praur.srfce
the
4
.0
9L4 110
oo.0
(T/V5n) VOTTTS pDATOSSTGj
09Z
E-b
___________________
09.z
0
.00 ,0
aot ---....
0.
101
.~~
*
~ ~~
-.
.
.
.
.
..
:
;>..
:::::::
.::::-.
009
tp .
-x.
008
K.-
- - - . . , .. .
0
50*
60*
70*
80*
.. ..... -.
..*...<
-1
e
-r
400
440
1
300L.
30
KX
I
*5
e x*
-
~7.5
3
*0
ii1
V-t
200
* II
80-
7u -
I I I I
600
I I I I II I I I I I I I
50*
Fig. 27. Dissolved silica (ugA/l) on the
3.0*C potential temperature surface.
I
40*
5M
1/9
02(.
TA/9
.
1500
'?% 5
5'?
59
4 5~5
'
41,
5,
41
d
* .1-
.
-
-
r
--
---
-o
2000.
2500{
ow.350c[)
5500
4500
4f
The.
potential
istem
50W
h
otdln
large..
dot
niae
indicate..
h
.0
potential isotherm. The large dots indicate
current meter locations (see text).
(taken from Clarke, Hill, Reiniger and
Warren (1978))
-
V
7,
oo.
9A
---
1;
C-.1
r'. ' 9
17
169
I C
;r
-0
I&
16
I?
-
-
2000
Md-
-
-
soo
3503-
4
0,001
0
Fig. 29. Section of dissolved silica (ugA/l)
running southeast fron the tail of the Grand
Banks. The dotted line indicates the 2.0*C
potential isotherm. The heavy dots indicate
current meter positions (see text).
(taken from Clarke, Hill, Reiniger and
Warren (1978))
11
-'
Fig. 30. Current meter record from the
instrument along 50*W that was in the low
silica water (see text).
Sso.
so
Cw...
n
a.0,
-5-150
I-So xx/st
T
3500.
-Jsoo
0
-00on
us
o
1600
-150
- -15
270
/SC70.
soI
3
o
f
Aoo
i
o
too
too
550
45*
450
-
40*
40*
35'
750
70*
Fig. 31.
65*
Locations of stations from KNORR 12.
600
55*
.
*
.
0
*
.0.0
,
4.0
0
*.
*
*0
3.5
0
LU
0
*J0
*
*S
0e
0
0.
*0
0
-**OR
9.*
.
0-
0
&
00
. 0L
.0
*%
e
0
%e
*.
y*
..
4.3 *.
**
e . 3.*
LU.
*.
..--.
00
.*
.
:
t..
2
* *
*e
3.0
*0.9
S
0
*
. .0-
M*o*,
.04000.
*
z
LU*
0
ee
*
e*0*
*0..0*0-
2.5
0
s00*
.
.0*.*
.
i.32. Potential temperature vs. salinity
from KNORR 12, HUDSON (50'W), KNORR 66 and
*
e*
*
.
4
2.0
.
HUDSON (50*W)
*KNORR 66
KNORR 48
*
*,
3.785
,KNORR 48.-
*KNORR
.
34.90
-
34.95
.*
SALINITY
(%/o)
35.00
12
5
5
HUDSON (50*W)
*KNORR 48
KNORR 66
0.;
4.0
Or
0.
* KNORR
0
12
.g
3.51
*
LiJ
.
S4 ..
.
.0.e
,
Fig. 33. Potential temperature vs. dissolved
silica from KNORR 12, HUDSON (50*W), KNORR 66
and KNORR 48.
0
w
*
*0
.
..
%
w
00
4p
w
:..
.1
*
*
*
w
I0
(L
ojo
gO
0
z
e0
0
0
*
6
*
0
*0
4I *o~~
*
*
0
*iiP~.**
*
2.5
0
.
~*
0
*0*~
*
0
*O0
*00000
*0
0
:.
.?
a~*
0
*
0.
0
0
*
9
'.
.
. .
0
2.-0t
.*
.
e* ..
g *.,
*
.*.
%.
%
*
1.71
Ic
~O
*
I
44 V1
DISSOLVED
SILICA (ugA/1)