N PS ARCHIVE
1965
CERES, R.
mmWBL ,
imSIi
m
$S!2
THE SEASONAL VARIATION OF PERMANENT
CUU:-rR'JI"S VYSlKfN SAFMsJ RAY
CERES
■ -■'■.■
,J ■ :-v. ■..:•■.■• ■
ft
fliiiHHiil
i' v-,, ■...■■:.;■ .■'.','.■ -i,'! .... .■ ■:
in
■>'-&■;■'■■■'■■■.•■■::":
flKiHH
HaMSSssag*Hft
fe«S»*«
s^
THE SEASONAL VARIATION OF PERMANENT
CURRENTS WITHIN BAFFIN BAY
*****
Robert L. Ceres
THE SEASONAL VARIATION OF PERMANENT
CURRENTS WITHIN BAFFIN BAY
by
Robert L. Ceres
Lieutenant, United States Navy
Submitted in partial fulfillment of
the requirements for the degree of
MASTER OF SCIENCE
United States Naval Postgraduate School
Monterey, California
1 9 6 S
DUDLEY KNOX LIBRARY
NAVAL POSTGRADUATE SCHOOL
MONTEREY, CA 93943-5101
THE SEASONAL VARIATION OF PERMANENT
CURRENTS WITHIN BAFFIN BAY
by
Robert L. Ceres
This work is accepted as fulfilling
the thesis requirements for the degree of
MASTER OF SCIENCE
from the
United States Naval Postgraduate School
I
ABSTRACT
The seasonal variation of permanent currents within
Baffin Bay is investigated. A model based on physical reason-
ing is developed, whereby the speed of surface currents may be
calculated during winter months, when observations are not
available due to ice coverage. Verification is made where possi-
ble utilizing late fall, and early spring data. Limitations,
applications, and variability of results are discussed. Two
important applications are: first, to add to the oceanographic
knowledge of the Baffin Bay area, and second, to provide an
additional input to computer programs already in existence for
forecasting the drift of sea ice within Baffin Bay.
11
TABLE OF CONTENTS
Section Title Page
1. Introduction 1
2 . Background 3
3 . Wintertime Regime 5
4. Application 19
5. Limitations 23
6. Current Profile Study 25
7 . Verification 29
8. Conclusions 33
9. Bibliography 34
Appendix A 35
in
LIST OF ILLUSTRATIONS
Figure Page
1. Summertime Surface Circulation Pattern M-
Baffin Bay
2. Wind Driven Velocity Profile (Schematic) 8
3 . Winter Profiles within the Friction Layer 11
M-. Transition from Summer to Winter Profile 18
5. Mean Seasonal Variation of Surface Currents 20
Within Baffin Bay
6. Spring and Fall Velocity Profiles 22
7. Ice Edge; forecast and observed 30
June II and 16, 1953
8. Small Scale Forecast of Ice Lead 31
June 30, July 1, 1960
9. Ice Edge; forecast and observed 32
May 12 and 17, 1959
IV
1. Introduction
Numerous oceanographic cruises have been made into Baffin Bay
in recent years primarily by Canadian and United States oceano-
graphic vessels. The purposes of these cruises are to extend
man's knowledge of this region, and to explore its commercial
and military potential. This research, and other traffic as well,
is hindered in large part by the yearly cycle of winter ice growth
which generally leaves Baffin Bay accessible to shipping only dur-
ing the short summer season, July through September. Thus, while
a great deal is known about Baffin Bay during these summer months,
the winter oceanographic regime is almost totally unexplored. It
is, then, the first intent of this paper to consider one aspect of
this winter regime, that is, the surface circulation pattern.
One important aspect of extending knowledge of this area is
the problem of forecasting the drift of sea ice. The purpose and
usefulness of such forecasts is well established; they are present-
ly being made in the Baffin Bay area on a routine basis by hand
calculation. Both the Fleet Weather Facility, Monterey, California,
and the U. S. Naval Oceanographic Office, are in the process of
developing computer programs to accomplish this task; and, more
specifically, Knodle (1) in 1964 developed a computer program for
forecasting the wind drift of sea ice. Verification of test data
utilizing this program indicated that, if the additional input of
permanent currents could be introduced into the program, the accur-
acy of results might be significantly improved. This input of
permanent currents would be in the form of vectors at grid
points of any scale desired and would be added vectorially to
the present output, which considers only wind effects. While
this final step of programming was not completed in this paper,
the basic data for its completion are developed and examined in
some detail, together with a discussion of the variability of
the currents and the limitations of the methods employed.
2 . Background
The first extensive attempt to determine the circulation pattern
in Baffin Bay was made in 1928 • and,while much has been done since
that time, the work of Kiilerich (1939) (2) on the Danish Godthaab
expedition of 1928, still stands as the most authoritative account
of the summertime surface currents that exist within the Bay. The
results of this work, as modified by the more recent observations,
is shown in figure 1. Exhaustive descriptive accounts of this
region are available from any number of sources (3, M-) ; so only a
few distinctive features which are pertinent to the problem at hand
will be mentioned. Again, with reference to figure 1, it should be
noted that Baffin Bay is largely an isolated body of water with
limited exchange taking place through the narrow channels to the
north, and across the Davis Strait to the south. A ridge there
forms a definite sill, with a limiting depth of less than 730m.
The continental shelf is narrow on both the Greenland and Canadian
sides of the bay, with steep slopes falling off rapidly into the one
large basin, whose depths are in excess of 2000m.
The circulation is cyclonic with one small gyral to the immedi-
ate north of Davis Strait and one large gyral centered over the
Baffin Basin. Mean velocities are shown in cm/sec.
Various clima to logical atlases show that. from mid-November
through late May, Baffin Bay is virtually ice-bound; and thus in the
following section a model is presented to show how the surface current,
at any given location within Baffin Bay, is diminished with the advent
of total ice coverage.
3 . Wintertime Regime
Due to the fact that no winter measurements have been made,
this section is devoted to the development of a model, which. when
applied, will describe the wintertime variation of the surface
currents within Baffin Bay.
Two assumptions will be made initially" first, that the
summertime circulation is primarily maintained by the wind, and
further, that the contribution made to the surface circulation from
the Artie Basin and the North Atlantic is at a maximum during the
summer, decreasing to a minimum by mid-winter. The latter part of
this first assumption will be discussed in section 6. The former,
is justified qualitatively on the basis of climatological wind data,
and descriptive accounts of the area. For example, mean sea pres-
sure charts for the summer months show Baffin Bay to be under the
influence of a cold low pressure center. Further, the Polar Basin
is under the influence of a cold high which contributes to steady
easterly winds across the northern portion of Baffin Bay throughout
the summer. The second assumption is that, once Baffin Bay becomes
completely ice covered, no wind stress is transmitted through the
ice. In other words, the contribution the wind makes towards main-
taining the current is zero. Naturally, there will be some shifting
of the ice due to tides, freezing effects, etc.; but in the mean the
ice will be, for all practical purposes, stationary.
If the assumptions above hold, the only real forces acting
to alter the basic currents are first, the force of friction be-
tween ice and water, and second, the frictional forces between
the sea floor and the water. The latter is assumed to be rela-
tively small and is not considered in this model. This is reason-
able when one considers that the continental shelf is narrow, and
that several investigators, including Kiilerich, found the deep-
basin water to be near motionless. Further, if one assumes that
the energy gained from currents entering the bay, is approximately
equal to bottom and boundary losses, then the omission of these
losses is further justified, at least qualitatively.
The fundamental winter relationship can now be seen: the
energy of the currents is diminished at the rate at which energy
is dissipated by frictional stress between ice and water. If this
rate is known, together with the vertical distribution of velocity,
it follows that the rate at which the current velocity decreases
can be calculated. In order to find this rate, first let the stress
be given by
where k = coefficient of friction between ice and water,
u = surface current velocity, defined to be the current
found a few meters beneath the sea surface, and
©w = density of the water at the same level
The rate of work per unit area is "£u> which is to be
equated to a rate of energy dissipation per unit area:
^U- " ~$E t*<lnetlc Ener§y) or,
Area
where u . = mean current speed in the friction layer,
u, = the deviation from the mean u,
0,= the mean density in the friction layer,
and, ku= depth of the frictional layer.
The corresponding parameters in the layer from the depth of frict-
i
ion, to the level of no motion are ^a.^ 9-j. ,and £"£• as shown in
figure 2. Note that U. ~ tt. -f- -^ lX •* is applicable in either
layer.
Z *z
Lam.
Schematic representation of a
wind-driven velocity profile
showing quantities represented
in equation 2 .
FIGURE 2
From the observations made of velocity profiles shown in
Appendix A, two initial observations can be made. First, the
profile can change radically within a short period of time, and
within short distances; and, second, these changes take place
throughout the entire column of water. While these profiles may
contain many inaccuracies as will be discussed in section 6, the
observations do suggest that a loss or gain of energy at the sur-
face influences the entire column of water that is in motion.
Thus, the assumption is made that the kinetic energy loss is dis-
tributed from the surface to the level of no motion nearly uni-
formly. This means that the total loss, that is, the right hand
side of equation (2) , may be expressed solely in terms of the
loss within the friction layer, or more specifically
where the constant C is dependent upon the level of no
motion (LNM) and &h . The level of no motion applicable to
Baffin Bay is discussed in section M-; but, if for example, the
LNM is found at 1000m, and An = 50m, then C would equal twenty
etc. (or C = LNM/&K )• This doubtless is an oversimplification.
This assumption, however, appears to fit the observations, and
gives realistic results.
Thus, in equation (3) , the total kinetic energy loss is ex-
pressed solely in terms of the loss that occurs within the fric-
tion layer. This layer will be examined in some detail in the
remainder of this section. If several basic relationships, taken
from Shuleikin (5) , relating to wind drift of sea ice are intro-
duced, it will be shown that the thickness of the friction layer,
kl\ , may be expressed purely as a function of the surface current
velocity, and further that this same expression may be used for the
model under consideration. Thus, for a current flowing beneath
stationary ice:
where
(d) t = J>a . &y* )
T = tangential stress between air and ice
/{ = eddy viscosity coefficient,
Oou = density of the air,
oo =
cosind, where «> is the angular velocity
of the earth's rotation, and (D = latitude,
-nL = coefficient of friction between air and
water
A = the wind factor,
V = the wind velocity, and
aK , dW) U.} are as previously defined
10
From (b) and (c) is obtained Ai= ' -L -Jy •
which when substituted into (a) yields
Finally, solving (c) for V and substituting into the above 9
fb^NlL. where N^ff £ffl jfofo. '
With sufficient accuracy ^ = 2x10 , and on the basis of studies
r-.2
by several authors, A = 1.27x10 . Substituting these, and other
known values, into the above expression yields
kK = i+7 3U (4)
where Ar\ is in meters if u is in meters/sec.
In the above development the tangential stress T operating
on the sea surface was caused by wind. Shuleikin concluded, how-
ever, that with ice coverage the frictional layer would also be
present, but would be generated by the surface friction between ice
and water. Further, he concludes that the numerical value of the
coefficient N will not vary perceptibly because "it is known that
the turbulent regime affecting the value &K becomes established
under the influence of any velocity of the surface current u,
independent of whether this velocity is generated by the tangential
stress of wind origin, or by the stress caused by the friction of
ice on the water". Thus, in thte model equation (4) will be used to
relate u and An
11
With the magnitude of the currents that we are dealing with^
it is easily seen that this depth of friction will rarely exceed
fifty or sixty meters' and within this thin upper layer it can be
assumed with a good degree of accuracy that <?w=Pi •> where t) w is
the density of the water a few meters beneath the surface, and (?)
is the mean density within the friction layer.
The numerical value of the coefficient of friction, kw, has
been investigated in some detail by several authors, and in part-
icular by Brown and Crary (6) , Fukutomi (7) , and Shuleikin (S) .
The values varied with location and with the age of the ice.
Wittman and MacDowell (8) modified ShulekinTs coefficients for use
specifically in Baffin Bay, and Knodle (1) concluded after further
investigation, that a mean value of .013 was the best approximation.
This value of J^ then will be used for this model.
Now, from equation (4) and the density approximation above *
equation (3) takes the form,
la3 - \ ('.'■ ■' '■'■I < <-/.> M; ''V C5)
It will be shown in the following paragraphs that the expression
pti. -4- <. U. J> J may be treated, with some approximation, solely
as a function of the surface velocity, u, if it is assumed that the
shape of the velocity profile, once established, does not alter
significantly. While the wind-driven profiles given in appendix A
show a great variety of shape, it should be noted that the greatest
12
variability occurs when the currents are weak and in shallow
water. Profiles associated with stronger currents in deep water
(i.e., from S to 10 cm/sec. at the surface) show a definite
uniformity of shape. It is argued here that most of the profile
variability comes about as the result of non-uniform influences,
such as wind gradients at the surface, summer run-off, inflowing
currents, tides, or internal waves, and that these disturbing in-
fluences are either totally absent or at a minimum during the
winter season.
Considering the above, we will now construct a model for the
current velocity profile. First, it is assumed that during the
winter, the frictional component of the velocity (VV) is analogous
to that given by Ekman; then
(e) \ff« -n*e-,,v*
where, u ^ = velocity found at the depth of friction, and
VrCz) = frictional component of the velocity found at
depth i within the friction layer.
Further, it is assumed that. due to the proximity of land within
Baffin Bay, there is no rotation of this frictional component with
depth. It follows that:
(f) u* - u± - %W ,
where ua = the magnitude of the Current found at depth ^.
within the friction layer. If, for example, at the time the ice
becomes stationary, the depth of friction is found at 50m, then. by
equation (4) , u ^ may be calculated and a velocity profile plotted
(see figure 3) .
13
11+
Several comments are needed concerning this profile. First,
the velocity at a = bh is the proper velocity to give the rela-
tion between u and &\\ , i.e., Lr\ = 473u i . Second, equations
(e) and (f) are being utilized as the winter profile model since no
direct and detailed profile studies for flow beneath stationary ice
were available to the author. Finally, the introduction of this
profile necessitates redefining u, the surface current velocity,
as will become apparent in the following paragraphs .
— *r
To find the relation between the surface velocity u and u in
the friction layer, several velocity profiles were constructed utiliz-
ing equations (e) , (f) , and (4) for assumed values of uA , from
12 cm/sec to M- cm/sec. For ease of computation each profile was
approximated by a straight line as shown in figure 3. The velocity
gradient with depth could then be treated as a constant (du/da = C) ,
and
(a) TL = «V + 0 ^fi-
(h) U.a="-<r + C*
where u^ is the velocity at depth 5 meters, defined as the
surface current velocity.
By definition
(i) a- = *»-*:•
Thus .
or
and hence
CO ^K^ |^
15
By utilizing equations (g) and (j) to calculate the quantities
<U,,> and u for each velocity profile approximation, two important
conclusions were reached! first, that the quantity (JOL +<M ^J
could be approximated quite accurately by LU. +* < U >JS ^ ^ i
and secondly, that the same expression can be described in terms of
the surface current velocity u, again without significant loss of
accuracy
where the constant ki equals 2.32. If equation (5) is modified
in accordance with equation (5a). then
MM u\ = - fa (i Ctf £] m u ^ . (6)
With the rapid decrease in current velocity near the ice surface
associated with the winter profile of figure 3, it becomes appar-
ent that u no longer adequately defines the velocity within the
friction layer, but rather that CV^k is the representative current
velocity. Utilizing the same winter profiles as in the above
development of u , it was found that the approximation U.^ = 3.18ur
was quite accurate, providing the range of velocity profiles
chosen is not too large. That is, within the range utilized, 4-12
cm/sec for u^ , the accuracy is quite good, and would be the
range of values expected for the major portion of Baffin Bay, Hence,
if we define u = u^ , and substitute the above approximation into
equation 6, then its solution is
where
16
It is seen then, that at the time the ice becomes stationary the
current velocity u of equation (7) represents the current found at
the depth of friction, and that this depth will decrease as the
winter season progresses approaching the surface by the time of
spring ice break -tp (see figure 4) . Thus, u at the depth of frict-
ion is applicable only during the time the ice is stationary, while
u several meters beneath the sea surface (i.e., at 5m) is applicable
for the remainder of the year.
Thus, equation (7), shows the time variation of the current at
depth th , where u(t0) represents the current at the initial time
of ice coverage and is known if the assumption is made that the
current velocities of figure 1 are representative of u ,L at the
time the ice becomes stationary. Hence, we may calculate the current
at any time u(t) between initial ice coverages and the time of break-
up of the ice . The application of this equation is discussed in the
following section, and a curve showing the seasonal variation of the
surface current is developed utilizing this equation.
The transition from the summer to the winter profile, is be-
lieved to take place within a short time after the ice becomes sta-
tionary. This is analogous, for example, to tidal currents near
bottom, as given by Defant (9 ). In this transition period the K. E.
loss is confined almost entirely to the friction layer, and the re-
duction of K. E. with depth does not occur until after the winter
profile becomes established. This transition period is shown schemat-
ically in figure 4, together with a profile, say several months later.
17
u
inrrL--
nm,-
Schematic velocity profiles
showing transition and decline
of u vs . a .
ice initially becomes stationary
t-, : several hrs . later
tp: several months later
(Note - not to scale)
FIGURE 4
18
*+. Application
In this section the application of equation (7) developed in
the previous section will be discussed, together with a qualita-
tive treatment of the spring and summer current build-up, thus
completing the mean yearly cycle.
Two facets of equation (7) must be dealt with prior to its
application. First, the time the ice becomes stationary and the
time break-up of the ice occurs must be determined; and, second,
the value of (^discussed briefly in section 3, needs to be calcul-
ated. The latter will be treated first. Recall that C = LNM/&K
and that the assumption has been made that kinetic energy is dissi-
pated from the surface to the level of no motion nearly uniformly.
This implies that the level of no motion descends with time . By
equation (M-) it is seen that the depth of friction &h is a
function Of the surface current velocity u and thus is likewise a
function of time (see figure 4). Hence, a profile study was under-
taken, one purpose of which was to investigate the variation of
these quantities (see appendix A) . From it, some approximations
could be made. From the fall data, it was estimated that the
level of no motion in deep water occurred at approximately 1000m
and3as previously discussed, bJ\ = 50m. Only a very few spring
soundings were available to the author, and only one in deep water.
The velocity profile constructed from them is as shown in figure 5,
together with a fall velocity profile taken at approximately the
19
U (cm /sec)
0
100
200-
300
400
Z
Cm)
500^
600-
700
eoo
900
1000
^•zero u
to ref. L
at 1500m
Spring and Fall Velocity Profiles
At: Lat. 69° 30'N, Long. 62° 40 TW
(constructed from data
contained in (10) & (12) )
FIGURE 5
Decrease in u
to zero at
ref. L. at 1200m
20
samfe location. While this indeed is sketchy information, both
the spring and the fall profiles ^ in this instance, lend support
to earlier assumptions, and in particular to the assumption that
an energy loss is dissipated with depth nearly uniformly. From
this profile it was estimated that at the time of break-up of the
ice LNM = 500m and kn = 20m. Thus, C varies from the fall to
the spring only from 20 to 25; a mean value of 22.5 is chosen for
C in equation (7) .
The span of time during which this model would be applicable
was determined on the basis of climatological data to be from
October through April. The limitations of utilizing these dates
are discussed in section 5; however, they should be approximately
correct for an average year. Now all the quantities of equation
(7) are known, and u(t) may be calculated utilizing the velocities
of figure 1 as u(t0) . The expression from equation (7) g. »M»l*^f
represents the fraction of the summer time current velocity and
may be plotted versus time with tQ = 0 on 1 October and t (in
seconds) up to 1 May. The plot is shown by the solid line portion
of figure 6 .
The assumed spring and summer build-up of the current is
shown by the dashed portion of figure 6 . This build-up was
handled in a purely qualitative fashion due to time limitations.
Now, figures 1 and 6 together permit determination of the mean
current velocity for any time during the year or location within
Baffin Bay.
21
■*->
a.
R
\
\
CD
3
\
<
\ =
\ O
>»
\ <r*
. 3
\ 1
->
x L
\M
V
X
c
. 3
^
\
->
\
X
\
(0
7
TT
'2
>
/
*£ co
/
°c0
j
L
*
J.
a.
1
'<
c c
j
o —
• — »♦—
CO
r
to
ft
ZCQ
UJ
1
°> .-
co ,
in
/
- ■>
durin
appi
. to
Z)
/t
U.
05 C
L.
A
of time
lod el is
C
-01
->
eason
Currc
/ <
/ 3
9 *-
eriod
ich rr
(j
to
•Q
00 u
u
c co
/ °
0. x:
cot
/ **
:
:
/
>
o
"2
2c/)
/
/
+j
j
o
1
j 1
1
jL
■o
Q-
■ to
§ss^ss$^
o
eg
0
V)
( t a jnBu ass) uoi;e|n dj|o
lie;
l<>%
'
22
S. Limitations
There are three major limitations to the winter-time model
developed in the preceding sections which require discussion.
First, the assumption was made, based on climatological data,
that the ice cover became approximately stationary on 1 October,
and that break-up occurred on 1 May. This statement implies that
'these times occur simultaneously throughout the Bay, when in fact
the freezing and break-up occur over a period of time with vary-
ing locations . An attempt was made to divide Baffin Bay in four
approximately equal areas and develop separate curves similar to
figure 6 for each quadrant. This trial was abandoned jhowever, as
it introduced more uncertainties and complications into the pro-
blem, and these disadvantages seemed to outweigh whatever benefits
might be derived. For example, a declining current in one quadrant
would certainly effect the next quadrant into which the current was
flowing. No ready solution to this problem is proposed; it is felt
that whatever modification is introduced would lead to only small
changes in figure 6. The use of climatological data also implies
that any one year is like the next, and this obviously is not the
case. Figure 6 could, however, be quite easily reconstructed,
based on information from local reconnaissance flights for any
given year.
The second major limitation concerns the variation in the
summer time build-up of currents. Evidence indicates that this
23
variation in some years might be large. For example, in 1964-
residual ice was observed in Melville Bugt throughout the summer,
because the mean pressure patterns normally associated with this
season were displaced enough from their normal positions that the
winds, which normally would advect ice out of Baffin Bay and con-
tribute to the current build-up, were simply not present- During
such an anomolous year, the model would obviously not work. How-
ever, this unusual situation is a rarity.
The final limitation concerns the sum of the approximations
and assumptions made in the development of the model itself.
While these assumptions and approximations have all been discussed
individually and justified, the cumulative effects could be addi-
tive and result in serious error. Particularly in the assumption
of the shape of winter velocity profile and in the treatment of
C=LNM/a|-\ could large errors arise. However, in the absence of
winter data, it is not possible to assign a number to this cumula-
tive error. In section 7, verification of a forecast utilizing
this model is quite good. This in itself is not proof that all the
approximations and assumptions are valid, but it does at least lend
support.
24
6. Current Profile Study
In order to verify the assumptions made in earlier sections,
as well as to verify figures 1 and 6, an attempt was made to ob-
tain the published results from the latest fall and earliest
spring oceanographic cruises that had been made into Baffin Bay.
These data were received from the Bedford Institute of Oceano-
graphy, Darthmouth, N.S., via the Canadian Oceanographic Data
Center (10, 11) . Utilizing the standard techniques of computing
and summing dynamic height anomolies, a number of velocity profiles
was constructed. These profiles, together with their locations,
are shown in appendix A. While, in large part, this study failed
to yield the verification that was hoped for, some interesting ob-
servations can be made. The remainder of this section is devoted
to a discussion of what was hoped to be gained by this study,
whether success or failure was met, and whenever possible why.
First, it was hoped that the shape of the velocity profiles,
at least for the fall data, would fit the preconceived notion of
a wind-driven velocity profile (see figure 2) . While in a number
of instances, particularly in deep water, this was the case, a
great variety of shapes resulted. While causes for variability
were mentioned briefly in section 3 (tides, internal waves, etc.),
it is believed by the author that a major limitation of the dynamic
method of computing currents is that oceanographic station data can-
not be taken simultaneously (that is, with only one ship) . For the
25
profiles shown in appendix A, a lapse time of from two to six hours
occurred between adjacent stations; if reorientation of the dynamic
topography were to take place during this interval, distortion in
the magnitude of the currbnt at a given depth would occur.
It was hoped, secondly, that the level of no motion could be
firmly established for the spring and the fall, as various investi-
gators, and particularly Kiilerich, had stated that this level will
vary from season to season, perhaps even irregularly from place to
place. While some substantiation of the figures used in this model
were gained from this study (see figure 5) , it can be noted from
appendix A that this level cannot be fixed with confidence. Here
also, one must be careful to distinguish between the reference
level and the level of no motion. For this study, the reference
level was chosen in every instance at the greatest depth that the
adjacent soundings would permit. However, in some instances the
soundings did not approach the bottom* and, hence, the profiles give
only a velocity that is relative to whatever current exists at this
reference level. This of course means also that the true level of
no motion could be at quite a different depth from that implied by
the profiles .
One of the initial assumptions made was that exchange of water
between Baffin Bay and the bodies of water to the North and South was
at a maximum during the summer, decreasing to a minimum by mid-winter.
As can be seen from appendix A, only the exchange across Davis Strait
26
could be examined. While the assumption made cannot be validated
on the basis of this study, it is interesting to note that, if the
assumed level of no motion is correct, outflow greatly exceeded in-
flow in both seasons with no inflow occurring with the fall data,
and very little with the spring data. From previous calculations
of current flow across Davis Strait, Dunbar (4) had stated that
the volume transport is approximately in the ratio of 2 to 1 with
outflow being twice as great. He concluded therefore, that Baffin
Bay received its waters in approximately equal quantities from the
north and the south, but failed to consider runoff, which on the
basis of this study would appear to make an important contribution.
Hence, while the ratio of 2 to 1 would appear to be an underestima-
tion, if , in fact, water does enter Baffin Bay in approximately
equal quantities from the North and South, one could draw the tenta-
tive conclusion that the amount of water entering Baffin Bay is
small during all seasons.
On the basis of the fall data, an effort was made to validate
the current velocities given in figure 1. Of the twenty-five fall
velocity profiles plotted in appendix A, fourteen, or 56%?were in
good agreement with figure 1, that is. within + 3cm/sec at the sur-
face. One further limitation of the dynamic anomoly technique is
that it gives only that component of the current which is normal to
a lijn^ drawn between adjacent stations. This surely accounts in
large part for the variation from figure 1 in several of the profiles
27
The remainder^ however, and some are the exact opposite of what
one would expect, can only be explained in terms of the limita-
tions of the methods employed outlined in previous paragraphs.
In conclusion, there was only fair agreement between the obser-
vations and figure 1, with no consistent error being observed
that would indicate that it should be changed.
A verification of figure 6 was hoped to be accomplished on
the basis of the spring data. These data, however, due to ice
coverage were restricted primarily to Labrador waters, with only
eight soundings taken within Baffin Bay. Of these eight, all
but one showed a definite decrease in the velocity as the model
predicts. The verification presented in the next section is an-
other test of the model, and of figure 6.
28
7. Verification
In Section 1, it was stated that one of the primary reasons
for investigating the seasonal variation of the permanent cur-
rents within Baffin Bay was that the results of the study could
be used as an additional input to Knodle's computer program for
forecasting the wind drift of ice, developed specifically for
Baffin Bay. In this section, the results of Knodle's three veri-
fication runs are shown in figures 7, 8, and 9, together with a
modified forecast which includes the effects of permanent currents.
This latter forecast was arrived at by adding to the wind-drift
calculations of Knodle's, hand calculated, advection of the ice
edge by currents in accordance with figures 1 and 6. While this
is a laborious and time consuming task, the results were gratify-
ing; and the job would be a simple one for the computer.
As can be seen, there was indeed a significant improvement
over Knodle's results in each of the three cases. One might con-
clude, however, that the magnitude of the currents was too small,
or5in effect, that figure 6 overestimates the winter seasonal de-
cline of the current. It is possible that melting might account
for a large part of the remaining error. It is concluded on the
*
basis of the foregoing that, while figures 1 and 6 may be altered
in the light of later findings, their use in forecasting the drift
of sea ice within Baffin Bay tends to improve the drift forecasts.
29
obse rved-T^
BAFFIN
BAY
EENLAND
— Input
Obse rved
Forecast (Kno die)
Forecast (Modified
for permanent
c u rrents )
.0 Limit of ice edge
0 Hand 16 June 1953
//
FIGURE 7
30
31
r=»
i nput
observed
f o recast ( K nodle)
forecast (modif i ed for
permanent currents) I f
DAVIS
Limit of ice edge
12and17 May 1959
FIGURE 9
32
8. Conclusion
In the preceding sections, a model has been developed which
predicts- the winter decline of the permanent currents within
Baffin Bay. In the absence of winter data, it was necessary to
make a number of assumptions, on the basis of physical reasoning,
about the nature of the flow. The spring and summer build-up was
handled qualitatively thus completing the yearly cycle. A limited
verification of the results was obtained.
It is concluded that, despite the limitations of the model,
steps have been taken toward adding to the oceanographic knowledge
of the Baffin Bay area and providing the basic material for a com-
puter input to forecasting the component of sea ice associated
with permanent currents. Finally, it is recommended that the re-
sults of this paper be combined with the results of Knodle . It is
felt that there would result a fully operational program, far
faster and more accurate than is presently done by hand calculation.
33
9 . BIBLIOGRAPHY
1. Knodle, W. C, A Computer Program For Forecasting the
Wind Drift of Sea Ice. U. S. Naval Postgraduate
School, May 1964.
2. Kiilerich, A. B., The Godthaab Expedition 1928.
Kobenhaven, C. A. Reitzels Forlag. 1939.
3. Oceanogr. Mar. Biol. Ann. Rev., 1964, 2, 45-75 Publ.
George Allen and Unwin Ltd., London. Physical Oceano-
graphy in Arctic Canada by AJEL Collin and Mw J.Dunbar.
4. Canadian Fisheries Research Board Bulletin (1949-52)
Bulletin #88. Eastern Arctic Waters by M,JJJunbar
5. Shuleikin, V. V. Fizika Moria (Physics of the Sea)
Moscow, 1953.
6. Browne, I. and Crary, A. P., The Movement of Ice in the
Arctic Ocean. Arctic Sea Ice Conference, Easton, Md.,
Feb. 1958.
7. Fukutomi, T. On the Steady Drift Current and Steady
Drift of Sea Ice, Due to Wind in the Frozen Sea.
Low Temperature Science Research :abpratpru. Hokkaido
University. Study of Sea Ice, report #14. October,
1948. #123.
8. U. S. Naval Oceanographic Office. Manual of Short-term
Sea Ice Forecasting, by W. I. Wittmann and G. P. Mac
Dowell, July 1963: SP - 82.
9. Defant, A. Volume II of Physical Oceanography.
The Macmillan Company New York, 1961.
10. Eastern Arctic -- I960; Data Record #18.
Canadian Oceanographic Data Centre 1964.
11. ICNAF Norwestlant - 2 Survey Canada, Data Record #14.
Canadian Oceanographic Data Centre 1964.
12. Report of the International Ice Patrol Service in the
North Atlantic Ocean. Season of 1962. Bulletin #48.
U. S. Government Printing Office, Washington. 1963.
34
APPENDIX A
VELOCITY PROFILES
On the following pages are shown the velocity profiles as
constructed from the data contained in (EO) and (11), together
with the locations for the fall profiles (figure A. Sept. 17 -
24- , 1960) and the spring profiles (figure B. June 1 to 11, 1963).
The computations were made in accordance with the Sandstrom-
Holland-Hansen method of computing currents. It is assumed that
the reader has some familiarity with the technique, so that only
■
a brief outline will be required here.
1. Dynamic height anomolies were computed for the standard
levels and summed upwards from the reference level (R.L.)
to the surface in accordance with: l\0 ~ ) , v Ap
was
with the specific volume anomoly 0 . The R.L.
picked in all cases at the maximum depth that adjacent
soundings would allow.
Relative currents for each level were computed by the
formula:
where
W|= current found at a given level
W^ = current found at the reference level
assumed equal to zero
OcecLn Oc&a_n
Sto..A
J^
^
R.L.
I W«. ■ O
35
The results of these computations, as stated, are shown in
the following pages together with other pertinent data such as
the depth of the water, the time the sounding was taken, etc.
36
FIGURE A
Fall Data: 17 to 22 Sept. 1960
Location of Ocean Stations
57 through 89 . Direction of surface
current between stations as indicated
37
i
^^™" ™T"™ 1 ™^™ ^^^ ^T^ ^^^
K
^2r\
47V. i /
'•51 /*
> GREENLAND
c
)
\J
K V27 1
V X
V \
'
\ ' ©
^J
\ 1 1
I ill i i i I I I /l -i ^^^
60
55
50
FIGURE B
Spring Data: 1 to 11 June 1963
Location of Ocean Stations 24 through 27,
and 41 through 51. Direction of surface
current between stations as indicated
38
CD
i
O
CD
O
CD
i
CD
ID
■v
G)
ID
i
00
ID
CD-
ID
ft
z
O
i-
i<
W
_)
U.
O
OH
Q_
>-
u
o
_l
LU
>
z
O
LU
i—
, , ^
<
UJ 3 ■
u
o
u_
_l
1
>— '
Q.
UJ
o
C\J
CD
o
CD
m
o
0)
CD
Q
2
J
O
o
o
O
u:
LU.
O
o
o
CJ
m
ID
in
in
££
h
2
in
CM
*
S
cp
o
c\j
m
m
f\\
^r
d
m
K
T—
^;
o d
10 z
o>
00
0)
o
,-
, ,z
in
lO
CD
ID
- Q
i
i
00
i
0)
i
o
(-• 1-
in
in
ill
(0
0_ <
UJ i-
i/) ^
39
i+O
VELOCITY PROFILES AT STATION
67-68, 68-69,69-70,70-71
-UCCM./SEC.) -
6 e 10
S TA 69-70 •
U.
r^~
■STA 68-69
: fi ? f
J § *-
-STA 67-68
STA 70-71
Z(M)
SEPT. 1 9, 1960
STATION NO.
G.M.T.
REF.L.(M)
DE PTH (M)
67 68
61 83
1 75
544
68 69
8 3 12.1
700
1392
69 70
12.1 16.5
1500
2062
70 71
16.5 23.0
2000
2309
LOCATION. SEE FIG J RE A
41
VELOCITY PROFILES AT STATION
73-74,74-7 5,75-76,76-77,7 7-78
■IKCM./SEC.) *
STA 7 3-74-
SEPT 20-21 , 1960
STATION NO.
G.M.T.
REF. L. (M )
DE PTH (M)
LOCAT I ON
7 3-74
11.8 -15.6
700
1 504
SEE
FIGUR E
A .
74 -7 5
15.6-19.3
700
2080
75-76
19.3-23.6
7 00
150 6
76 "77
23.6 -2.0
500
777
77 -78
2.0 "4.3
100
350
42
VELOCITY PROFILES AT STATION
80-81, 81-82,82-83,83-84
-U(CM/S EC) —
STA 80-81
V
-STA 82-83
SEPT 21-22, 1960
STATION NO
G.M.T.
REF. L. (M )
DE PTH (M )
80-81
18 3 -22
75
681
81 "82
22 -2.6
■1200
1415
82-83
26-8.2
12 00
1525
83 -84
- i
200
904
LOCATION : SEE F, - Fi
43
VELOCITY PROFILES AT STATION
85-86,86-87,87-8 8 ,88-89
ST A 85 -8 6
STA. 86-8 7
U(CM /SEC)
2 4
SEPT. 23-24 ,1960
STA* NO.
\'0 G.M.T.
REF L.(M)
DEPTH(M)
85-86
5.8-9.9
75
139
86-87
9.9-15.7
125
389
87-88
1 5.7- 208
500
576
88-89
208- 0.2
150
361
LOCATION:SEE FIGURE A
J
44
VELOCITY PROFILES AT STATION
24-25 , 25-26 26-27
-IKCM/SEO-
STA 25-26-
Z(M)
STA 24-25
JUNE 1 -2
STATION
1963
NO.
G M. T
14.4 18.8
REF. L. (M)
DE PTH(M)
24 25
1200
1527
25 26
188 22 8
1000
1308
26 27
22.8 02.2
1000
11 61
LOCATION : SEE FIGURE B
M-5
VELOCITY PROFILES AT STATION
41-42,42-43,43-44 ,44-45 ,45-46
10
U(CM/SEO-
2
STA 45-46
_£
STA 43-44'
ST A 41-42-
STA 44-45
100
Z(M)
(STA 42-43:ZERO)
200
300 <)
JUNE 10.1963
STA NO.
G. M.T.
REFL.(M)
DEPTH(M)
41-42
1.6-3.5
30
49
42-43
3.5-6.2
30
79
4 3-44
6.2-8.5
75
133
44-45
8.5-10.7
125
253
45-46
10.7-14.8
300
514
LOCATION: SEE FIGURE B
46
VELOCITY PROFILES AT STATION
47-48 48-49 49-50 50-51
U(CM/SEC ) ►
0 2 4
i i_
STA 4 7-43
Z(M)
100
200
STA 50-51
STA4 9-50
-STA 4 6-4 9
JUNE 11 1963
STA. NO.
G. M. T.
REE L.(M)
DEPTH(M)
47-48
3.3-5.2
175
223
4 8-49
5.2-8.2
100
149
49-50
8.2-10.5
50
88
50-51
10.5-12.6
30
52
LOCATION: SEE FIGURE B
47
BlUOE
4 AUO -
1691?
„4-UH-
L
Thesis
C33838 Cere
aJ!;e sfasonal vari-
ation of permanent
currents wi nl
79938
thin Baffin
<-4 .c
C WIDER
16 9]
Thesis
C33838 Ceres 79338
The seasonal vari-
ation of permanent
currents within Baffin
Bay.