The role of tidal mixing in Rupert and Holbcrg
K. F. Drinkwatt+
Institute
Inlets, Vancouver Island
and T. R. Osborn
of Oceanography,
University
of British
Columbia,
Vancouver
V6T
lW5
Abstract
Analysis of monthly observations
of temperature,
salinity, and dissolved oxygen content
in the basin formed by Rupert and Holbcrg
Inlets reveals greater vertical mixing than
in most British Columbia
inlets. The water temperature
correlates with solar radiation
while the salinity changes follow river runoff, in turn controlled
by precipitation.
The
variation in dissolved oxygen appears due to a combination
of biological influences and the
influx of water. A model has been developed which ascribes the monthly fluctuations
and
vertical homogeneity
to an accumulation
of irregular mixing events associated with the tidal
flow through Quatsino Narrows, a shallow connecting channel,
Rupert and IIolberg Inlets ( Fig. 1) near
the northern end of Vancouver Island, possess a shape typical of British Columbia
fjords ( Pickard 1956)) being elongated,
narrow bodies of water with roughly parallel sides. Rupert Inlet is 10 km long, 1.8
km wide, and has a mean midchannel depth
of 110 m. IIolberg Inlet is 34 km long, 1.3
km wide, and has a mean midchannel
depth of 80 m (Pickard 1963). Together
they form a basin of 170 m maximum
depth, separated from Quatsino Sound and
Neroutsos Inlet by Quatsino Narrows, a
long, narrow channel with a sharp bend at
its southern extreme. Its depth decreases
northward, culminating
in a sill of depth
18 m west of Makwazniht Island (Fig. 2).
The principal river affecting the water
properties of the Rupert-IIolberg
basin is
the Marble River. It has a drainage area
of 512 km2 and flows from Alice Lake entering Rupert Inlet near the junction with
IIolberg and Quatsino Narrows. This discharge location presents a situation different from that of the typical inlet in which
the freshwater enters near the head.
Industrial, scientific, and public interest
in Rupert and Holbcrg Inlets was generated during 1970 with the application and
subsequent granting of a permit to discharge wastes from a copper-molybdenum
mine into the bottom of these inlets.
Although some data have been published
for the Rupert-IIolberg
system, discussion
1 Present address : Bedford Institute
rphy,
Dartmouth,
Nova Scotia.
LIMNOLOGY
AND
OCEANOGHAPIIY
of Occanog-
of its physical oceanography has been limited to brief comments in a general description of Vancouver Island inlets by Pickard
(1963), who mentioned the influence of
freshwater inflow, the uniformity
of the
water propertics, and the high oxygen values in the deep water. It is in these respects that Rupert and IIolberg differ most
significantly
from other inlets in British
Columbia.
Here we attempt to explain the unusual
uniformity
of the water properties and
their monthly changes.
We are indebted to J. 13. Evans who was
helpful in making available the data collected by Utah Mines. The help of C. Pelletier is also appreciated.
Information
about the Marble River was supplied by
both the Canadian Fisheries Service and
Inland Waters Branch of the Department
of Energy, Mines and Resources. We also
thank G. L. Pickard for his comments and
criticism of this manuscript.
This work
was supported by grants from the National
Research Council and a contract from Marine Science Branch, Pacific Region, Department of the Environment.
Data
The oceanographic
data used in this
study have been collected from several
of Oceanography
sources. The Institute
has conducted three cruises into the Rupert-Holberg
area: a survey including
Quatsino Sound-Ncroutsos
Inlet in May
1959 (Inst. Oceanogr. Data Rep. 15) and
518
JULY
1975,
V.
20(4)
Tidal
Fig.
Inlets.
1.
Index
map
of
Rupcrt
and
BUCHOLZ
Fig.
CHANNEL
2.
Detail
of Quatsino
Narrows.
F
I-Iolbcrg
two cruises investigating only Rupert and
Holberg during 1971, one in March when
thermal microstructure measurements were
taken and the other in April (Inst. Oceanogr. Data Rep. 33). During 1957-1967, the
Pacific Oceanographic Group of the Fisheries Research Board of Canada at Nanaimo (Waldichuk et al. 1968) conducted
an extensive survey of several inlets along
the west coast of Vancouver Island including the Quatsino Sound region. Surveys
during August 1957, November 1962, and
August 1966 included stations in Rupert
and Holberg Inlets.
A program to determine the effects of
the copper-molybdenum
mine was begun
,
519
mixing
MAMJJASO’NDJFMAMJ
1971
I M1-l..
I.-J I
A
M
1972
Fig. 3. Monthly values of temperature,
salinity,
and dissolved
oxygen content
at station
B in
IIolbcrg
Inlet.
in spring 1971 and is to run until Septcmber 1976. Monthly surveys at six stations
(A through F: Fig. 1)) included observations of temperature, salinity, and dissolved
oxygen. After June 1972, salinity and dissolved oxygen were measured once every
3 months. Temperature was measured until
March 1972 with an Applied Research
model ET 100 marine thermometer, then
with reversing thermometers. Salinity and
dissolved oxygen wcrc determined by titration ( Strickland and Parsons 1968).
Monthly values of temperature, salinity,
and dissolved oxygen for several depths at
station B are plotted in Fig. 3. The data
collected during the surveys conducted by
the mine show uniform distribution of these
properties within Rupert and Holbcrg Inlets, and hence this station is representative
of the entire basin.
The temperature throughout the water
column changes steadily, with a minimum
in early spring and a maximum in late sum-
520
Drinkwater
71
’
’
I
2MAMJJASONDJ
1
“0
i
ti
#
MAMJJASOND
1971
I-
L -,.
I -1 -I-.--L--.L
FMAMJ
3I
1 -LA-
I
I
,
.
.. .
I
,
L
-I.
L-I_
FMAMJ
.-I
I%72
l?ig. 4. Monthly values of temperature, salinity,
and dissolved oxygen content at station D in Quatsino Sound.
mer or early autumn, Within Holberg the
maximum temperature in the deep water
occurs during early October whereas in
Rupert and at station E the maximum is
found at the beginning of September, Below 30 m, the water in Rupert Inlet during
September is about 0.2”C higher than at
an equivalent depth at station B; during
October it is 0.2”C lower.
Salinity is more variable, generally high
in midsummer and late winter and low in
early spring and late autumn. Surface salinities at station B (not plotted) vary between 12%0 and 31%0; below 30 m, salinities
vary from 29-32%0 with the difference between maximum and minimum
salinity,
3%0, independent of depth.
and Osborn
Monthly changes of the order of l%,
are common. As density is primarily
dependent on salinity, the decreases in salinity correspond to decreases in density.
Dissolved oxygen tends toward high values in winter and spring, with lower values during summer and autumn. Significant short term increases in oxygen appear
during August 1971 and December 1971.
Monthly data for Quatsino Sound, station D, are displayed in Fig. 4. The variability of the upper layers ( surface to 60
m) is similar to that within Rupert and
IIolberg; the bottom water ( 60 to 116 m)
has diffcrcnt characteristics, with a maximum in January and a smaller temperature
range.
Except during autumn, the salinity trends
at station D for all depths are alike and
follow a pattern close to that of the basin.
The difference in salinities between the top
and bottom layers is greater at station D
than in the deeper Rupert-IIolberg
basin.
The magnitude of the salinity fluctuations
decreases with depth at station D.
Although the range of oxygen values is
wider within the water column at station
D, the trends are similar to those in the
basin. A notable increase, especially in the
upper layers, occurs during August 1971.
Oxygen values at 30 and 46 m are frequently higher than at 9 m; the cause of
this unusual occurrence may be an influx
of low oxygen water in the surface waters
of Neroutsos Inlet from a pulp mill at Port
Alice. The mill is still in operation.
Figure 5 shows the temperature, salinity,
and dissolved oxygen profiles at stations 13
and D in September 1971 and February
1972, drawn from data collected through
bottle casts; these months are representative of summer and winter structure. The
oxygen minimum directly above 150 m at
station B during February results from a
No corresponding
minisingle reading.
mum
is detected elsewhere within
the
b asin, hence the value might be in error.
All stations in the Rupert-IIolberg
basin
show profiles similar to those of station B,
excepting
station E, where a slightly
greater degree of vertical uniformity nor-
Tidal
FkB
1972
mixing
521
SEP
1971
SALINITY
(%o)
28
011L
I
I
304
0
12 16 2024;6
OCT 1971
I
I
I
I
I
I
I
I
5
9
13 17 21
NOV 1971
I--
I
1 -.
25
29
Fig. 6. Daily
precipitation
recorded
at the
mint site and the daily discharge from the Marble
River.
FEB
1972
DISSOLVED
OXYGEN
SEP
1971
(ml/l)
Fig. 5. Depth profiles of tcmpcraturc,
salinity,
and dissolved
oxygen content for station B in
IIolberg
Inlet and station D in Quatsino Sound
dnring Scptcmbcr 1971 and February 1972.
mally exists. These profiles indicate a more
homogcncous water body below 30 m in
the basin than in Quatsino Sound. Studies
by Pickard ( 1961, 1963) indicate that the
propertics of most .Rritish Columbian inlets
are not as uniform as those in the RupertIIolbcrg basin.
The correspondence between prccipitation and the Marble River runoff is close
( Figs. 6, 7, and 8). The daily values of precipitation and river discharge during October and November 1971 are displayed in
Fig. 6. Precipitation was recorded at the
mine site. Generally, fluctuations
in precipitation produce corresponding fluctuations in the discharge, but the actual relationship between these two features is
nonlinear. The discharge data reflect a filtcring effect on the precipitation; for examplc, several of the high frequency fluctuations in precipitation are not evident in the
discharge data. A lag of a day or two exists
between peaks in rainfall and river discharge.
Figures 7 and 8 show the weekly and
monthly means of precipitation
and Marble River discharge for all data between
March 1971 and June 1972. The magnitude
of the peak in precipitation
during early
September is questionable. As in the daily
data, both weekly and monthly means show
corresponding precipitation
and river runoff.
The basin volume is compared to the
522
Drinkwater
and Osborn
. WEEKLY MEAN
o MONTHLY MEAN
150-
OL-M
A
1371
Fig. 7. Weekly and monthly means of precipitation recorded at the mint site.
tidal prism and freshwater inflow in Table
1. Volumes of Rupert and IIolberg Inlets
are calculated on the basis of a rectangular
basin with the mean values previously
quoted used for the dimensions. The tidal
prism is defined as the surface area of the
inlets multiplied by the increase in depth
due to the tide. The average tidal prism is
determined from a tide of 2.8 m, with an
increase of 4.2 m for maximum conditions;
these are the ranges for Coal Harbour (Can.
IIydrogr. Serv. 1971) .
A clrainagc area of about 580 km2 surrounds Rupert and Holberg Inlets, excluding the Marble River drainage of 512 km2.
No major river system exists within the
former area. Thus, based entirely on the
relative sizes of the drainage areas, the
total volume of the freshwater flow into
the basin is assumed to be about twice that
of the Marble River. The average discharge
is tabulated from the total of all daily measurements divided by the number of days.
The maximum inflow was on 11 November
1971.
Fig.
Marble
monthly
means
of
the
into the basin of denser water from Quatsino Sound. IIowever, the periods of decreasing bottom densities require an influx
of energy to raise the potential energy of
the water within the basin.
Sea surface temperatures at station B
are compared with the monthly mean air
temperatures and the mean air temperatures 3 days immediately preceding the sea
measurcmcnts in Fig. 9. The close resemblance between the sea surface temperatures and the 3-day average of the air temperatures implies that the surface layers
are influenced by short term changes of air
Table 1. Comparison
of the volumes of the
Ruperl-Holberg
basin to the estimated tidal prism
and fresh.water inflow.
(4
Volume
Narrows
of
Rupert-Ilolberg
Basin
and
Volume
Rupert
Inlet
Holberg
Inlet
Rupert-Holberg
Basin
Quatsino
Narrows
Discussion
The vertical uniformity
of temperature,
salinity, and dissolved oxygen profiles, accompanied by their month-to-month
variability, suggests a process operating on a
time scale of less than a month which is
capable of increasing or decreasing the salinity, temperature, and density of the entire basin. Increasing deep salinities and
densi ties can be attributed
to intrusions
8. Weekly
and
River outflow.
(b)
Volume
of
tidal
inflow
(c)
Volume
(tidal
Average
Maximum
conditions
conditions
of
freshwater
conditions
conditions
(106m3)
2,000
3,400
5,400
30
Volume
tidal
Average
Maximum
Quatsino
prism)
( 106m3/
cycle
170
260
inflow
Volume
(106m3/day)
12
40
523
Tidal mixing
o-- -0 MONTHLY AIR TEMPEATURE
I
a
D--O
MEAN AIR TEMPERATURE
A “TXRl@R,&,DAYS PRIOR
ABOVE
THERMOCLINE
IMMEDIATELY
THERMOCLINE
SEA SURFACE TEMPERATURE
FlFTE;;Ty;;ERS
o1ti
1971
F
‘A-‘M
M
A-
M
BELOW
FROM
-
J
1972
of air tcmpcratures
Fig. 9. Comparison
sea surface temperatures.
with
The similarity
in temperatemperature.
ture trends between the deep water and the
surface layers (Fig. 3) suggests that these
atmospheric influcnccs are felt throughout
the water column.
The salinity fluctuations (Fig. 3) appear
to be inversely related to the fluctuations
in river discharge ( Fig. S), and hence the
precipitation
(Fig. 7). The time response
of the basjn water to discharge is between
1 and 4 weeks.
Pickard (1961) has described seasonal
variations in Bute Inlet and Indian Arm
(see Fig. 14) where salinity fluctuations
are observable to 30 and 50 m respectively,
while temperature changes are apparent to
100 m in both. The extent of the seasonal
variability
in the Rupert-Holberg
basin
thus indicates a different mixing process
than is normally found in mainland British
Columbian inlets.
The near uniformity of the oxygen content throughout the water column implies
frequent mixing within the basin. The
lower values during summer and autumn
may bc attributed to an influx of low-oxygen water from the Pacific Ocean into
Quatsino Sound. Lane ( 1962) has shown
upwelling to occur during summer along
the coast of Vancouver Island (off Amphitrite Point) and then move onto the shelf.
Pickard (1963) reports evidence of upwelling off Quatsino Sound but with no distinct seasonal pattern, The rise in salinity
during Tune, 1971, and the low tempera-
0
JUL
AUG
SEP
1971
Fig. 10. Chlorophyll
A in Rupert Inlet.
a measurements
at station
tures within the deep water during summer at station D are additional evidence
for upwelling.
Also, during summer, sinking organic matter, caused by an increase
in biological activity, may use up oxygen
in the deeper waters. An increase of chlorophyll (Fig. 10) indicates that the high
August oxygen values are associated with a
plankton bloom.
Investigation
into tidal effects leads to
a possible explanation of the mixing. Assuming complete mixing with an average
tidal volume per tidal cycle, in 1 month
about 15% of the original water will remain
in the basin. The average monthly freshwater discharge is 7% of the entire volume
of the basin, During a high runoff month,
this volume inflow doubles. Mixing the
average freshwater inflow into the basin
throughout the water column would produce a salinity change of about 2%0.
524
Drinkwater
Excellent
conditions
for mixing exist
within Quatsino Narrows (Fig. 2). Most
of the water flowing into or out of Quatsino
Narrows must pass through a narrow channel north of Quattische Island, as shallow
banks prohibit the movement of large volumes of water south of the island. On a
floodtide, a narrow eastward flowing current passes north of Quattische Island; a
large deflection is needed to redirect it
northward towards Rupert and Holberg.
Surface current data (Can. Hydrogr. Serv.
1972) show that the current splits near the
shore, part moving south to create a large
eddy southeast of Quattische Island and
part hugging the cast shoreline as it heads
north through the narrows. Several smaller
eddies, produced by indentations along the
eastern shorclinc, are also seen farther up
the channel. In August 1957 over the 20-m
depth surveyed, the density changed by
only 0.27 of a crt unit and an instability was
present between 9 and 14 m (Waldichuk
et al. 1968). The density of the upper 10 m
was greater than that at equivalent depths
outside the narrows; at 20 m, however, the
density was slightly less in the narrows than
in Quatsino Sound.
During an ebbtide, conditions similar to
those of the floodtide
produce mixing
within the same region. The upper layers
of the Rupert-IIolberg
basin flow southward through the channel and beyond Ohlsen Point. The required redirection of the
flow westward into Quatsino Sound is at
first prohibited by the momentum of the
flow. The current studies (Can. Hydrogr.
Serv. 1972) show the formation of a large
anticlockwisc eddy southeast of Quattische
Island and a westward flow past the northcm end of this island. We have seen very
turbulent water throughout
the narrows
during both flood- and ebbtides.
The volume of Quatsino Narrows between Makwazniht Island and Quattische
Islancl ( Table 1) is about a sixth of the
average tidal prism. As the tides are scmidiurnal, the water at the tidal prism spends
about 1 h within the narrows. This allows
sufficient time for mixing to occur.
and Osborn
=Q---
.,.I/_
--- -1
Fig. 11. Plots of constant surfaces for (upper)
UBC data (Pickard)
and (lower)
Pacific Oceanographic Group data ( Waldichuk
et al. ).
The fresher and the denser upper layers
which enter the channel are mixed in the
narrows, so that the density of the water
within the narrows is greater than that of
the surface layer outside. Upon entering
the Rupert-IIolberg
basin on a floodtide,
this water will sink beneath the less dense,
low salinity layer formed by the Marble
River; less energy is then required to mix
it into the deeper water of the inlet as it is
already in or perhaps even partially
through the pycnocline.
Isopleths of ut are shown in Fig. 11. The
bottom waters in the basin are less dense
than, or at most of equal density to the upper 40 m of water in Quatsino Sound, indicating that changes in the upper layers will
likely be reflected by changes in the bottom waters. Although the isopleths are
drawn smooth through the narrows, the
water is probably more homogeneous than
indicated.
Energetics-We
calculated the rate at
which available energy flows into the Rupert-IIolberg
basin using the method of
Taylor ( 1919). C onsider a given volume of
water which at time f is enclosed by a surface area A (extending the entire depth
under consideration).
The rate of change
of the energy ( W) inside the volume due
to tidal effects is given by
Tidal
525
mixing
For the Rupert-IIolberg
system, the tidal
height and velocity are, to the first approximation, of the form
h = I-1 ~0s F
+ v = -V sin 2”‘ty
To) ,
-----M&
where V is the maximum current velocity,
11 is half of the tidal range, T is the tidal
period, and To is the time difference between high water and turn to ebb at the
northern end of Quatsino Narrows.
Averaging over one tidal cycle, the last
three terms disappear and the average rate
of energy crossing the sill, (W), assuming
that conditions are uniform across the channel, is given by
LEVEL
Fig.
12.
Diagram
for
dW= pgD$-h
T(v
dt
Eq.
1.
!iI
cos e) (D + h)]dS
+ s pg(D+h)
qv
+
s
+D
cos e> dS
+ h) (v cos 6) dS,
(W) = ++gHV(sin
(1)
where h is the tidal height above mean sea
level, p is the density of the fluid (assumcd homogeneous ), g is the acceleration
due to gravity, t is the time measured after
an arbitrary turn of the tide to flood, D is
the depth of the sea floor below mean sea
level, v is the velocity, 8 is the angle between the current direction and the elemcnt dS, and dS is the element of length
along A ( Fig. 12).
The first term on the right-hand side
represents the rate at which work is done
by the mean hydrostatic pressure on that
portion of fluid originally within the volume; the second term is the rate of gravitational potential energy transport across A,
with zero potential at mean sea level; and
the last term is the rate at which kinetic
energy crosses A.
Combining terms,
dW
x
= pg 1 Dh v(cos 8) dS
+ 51
v(cos e) (2gh2 + Dv2 + hv2) ds. (2)
Assuming no net increase of kinetic energy
within the basin, the time average of Eq.
2 over a tidal cycle equals the energy dissipated by friction and the increase in gravitational potential energy due to mixing,
F)
(cos 0) DL, (3)
where L is the width of the sill.
The numerical values of the constituents
are p = 1.023 g crns3, g = 9.8 m s2, E-I= 1.4
m, v = 7.0 knots = 3.6 m s-l, T = 12 h 25
min = 745 min, To = 7 min, sin 2vTo/T
= sin 3’32’ = 0.059, 0 = 0”. T and To were
obtained from Canadian Hydrographic Service (1971) tables. cos 8 = 1, D = 20 m,
L = 250 m.’
Substituting into Eq. 3, we find the average energy which must be dissipated
within the-basin by friction and mixing to
bc of the order of 7 X lo6 J s-l. This value
multiplied
by the tidal period ( T) gives
3 x 1O1l J as the total energy to be dissipated during one tidal cycle. The average
energy per unit volume of the incoming
tidal water is then the total energy divided
by the tidal prism and has a value of 1.8
x lo3 J m-3.
When waters of unequal density mix,
work is done against the buoyancy force.
To mix a layer of thickness hI and density
pl into a layer of thickness b and density
~2, the amount of work done per unit area
is
-(plgh) x ++
which reduces to
(pgh) x $,
526
Drinkwater
and Osborn
Table 2. Energy required for mixing of water
from Quatsino Narrows
into the Rupert-Holberg
basin during the times of the surveys conducted
by the mine.
-TL-c-z_
-___
-
~----z
.-----_
_ _-
----a-
1971-1972
PI (s/cm3)
8-10 Mar
5-6 Apr
14-18 Jun
6- 8 Jul
2- 4 Aug
l- 2 Sep
4- 5 Ott
7-11 Dee
3- 6 Jan
l- 3 Feb
17-22 Mar
5- 7 Apr
l- 3 May
~-.- 5- 7 Jun
1.02333
1.02326
1.02424
1.02332
1.02387
1.02319
1.02306
1.02223
1.02319
1.02389
1.02200
1.02238
1.02245
---~1.02291
p2 (g/cm3)
1.02344
1.02353
1.02488
1.02407
1.02487
1.02440
1.02360
1.02297
1.02368
1.02405
1.02325
1.02306
1.02257
1.02359
-=-
~_
_.
.-
_-_
Joules
(J/m31
80
200
400
550
740
900
400
530
360
50
920
500
90
500
To simulate flood conditions in Rupert
and IIolberg
Inlets, we choose a 10-m
layer, whose density corresponds to the 9-m
depth at station D, over a 150-m layer
whose density equals that found at the bottom of Rupert and Holberg basins, Table 2
shows the mixing energy per unit volume
of the upper layer, values obtained by dividing the work done per unit area by the
thickness of the upper layer.
On an average, each cubic meter of water within the tidal prism contains excess
energy of the order of 1,800 J. If mixing is
similar to that assumed in the model, then
between 50 and 1,000 J mm3 of incoming
water is needed. Much energy is dissipated
by molecular viscosity. IIowever even if
only a small fraction of! the available cnergy per tidal cycle is used in raising the
potential
energy, its continual
injection
into the basin can still produce significant
mixing. If not for the mixing in Quatsino
Narrows, an amount of energy between
250 and 3,000 J m-3 of incoming water
would be needed to mix the surface water
from Quatsino Sound into the basin. Thus,
mixing within the narrows decreases the
energy required for mixing in the basin by
about 2 to 3 .times.
A descriptive model can be developed
to visualize the mixing processes, outlined
in Fig. 13. A three-layer system is considercd in Quatsino Sound: A-an
upper
layer; B-the pycnocline region; and C-
Fig. 13. Schematic diagram showing flow conditions between Quatsino Sound and the RupcrtHolberg basin during floodtides and ebbtidcs.
the bottom water. Layer D represents the
homogeneous water within Quatsino Narrows. Rupert and IIolberg consist of a low
salinity layer, E, above a layer extending
to the basin floor, F. The order of increasing density is E<A<D<F<B<C.
Slight modifications to this order are discussed below.
During flood tide, water from layers A
and B flows into Qautsino Narrows where
it is thoroughly
mixed, thus producing
more of layer D. At the opposite end of
the narrows, layer D enters into the Rupert-IIolberg
basin, sinking below the low
salinity layer E and mixing into region F.
The degree of mixing depends on the density difference between D and F and the
available energy. On an ebbtide, layer E
and part of F move into the narrows to become well mixed. This water eventually
passes into Quatsino Sound at a depth dependent on the density difference between
it and layers A and B.
Tidal
mixing
This model illustrates how water from
the narrows is injected into the basin. The
momentum of the incoming water forces
aside the low salinity surface layer in Rupert and IIolberg between the narrows and
Hankin Point. Turbulent
water can be
seen within this area during a floodtide, in
contrast with the apparently calm conditions surrounding it. The density difference eventually causes the incoming water
to sink below the fresher layer. Thermal
microstructure
measurements
in March
1971 suggest that most of the tidal inflow
is then stirred into the basin just beyond the
sill; this causes the greater vertical uniformity at station E. Transport proceeds
by pressure gradient currents due to the
density difference between this region and
the rest of the inlet.
During ebb conditions, the freshwater
layer which originates from the Marble
River flows out toward the narrows, entraining the deeper, more saline water into
it. A flow is required to replace the saline
water. During floodtides, injection of the
water beneath the low salinity layer just
beyond the narrows necessitates a subsurface flow to spread it throughout the basin.
Normal estuarine circulation
requires a
subsurface up-inlet flow to replace the saline water entrained into the continually
outflowing
surf ace layer ( Tully
1949 ) ,
IIowever, the major freshwater inflow here
is concentrated near the mouth, and thus
the circulation is different from that of the
“typical” inlet.
Comparison with other shallow silled inlets-A
high degree of mixing within an
inlet is revealed by near uniform conditions
of temperature and salinity in conjunction
with high oxygen values. The mixing may
be continuous or the result of recent flushing of the inlet. Either nonuniform conditions or low oxygen values would eliminate
the possibility of continuous mixing.
We examined data collected by the University of British Columbia since 1951 to
determine the mixing charactcristice of inlets with geographical configurations similar to those of the Rupert-Ilolberg
basin,
527
1-Amphitritc
Point.
Fig. 14. Location map:
Sound (see Fig. 15
2Tofino
Inlet; 3-Bedwell
fols detail).
4-Quatsino
Sound; 5-Rupert
Inlet;
6-Holberg
Inlet (see Figs. 1 and 2 for detail).
7-Indian
Arm.
8-Sechelt
Inlet.
9-Bute
Inlet. lo-Drury
Inlet.
ll-Seymour
Inlet.
12Belize Inlet. Inset map: 13-Porchcr
Inlet; 14Work Channel.
looking for the factors essential to continuous mixing. We restricted the investigation
to those British Columbia inlets described
by Pickard (1961, lQ63) which are connected by long narrow channels and have
sill depths less than 30 m.
Indian Arm, Sechelt, Belize, Seymour
Inlets, and Work Channel all have the
necessary geographic features (Fig. 14).
They do, however, exhibit nonuniform waIndian Arm and Work
ter properties.
Channel show a gradual increase in depth
away from the sill, which may explain the
absence of thorough mixing. In Sechelt,
Belize, and Seymour Inlets, the ratio of the
tidal prism to the total volume is less than
1 : 100; such small tidal flow would produce correspondingly
small effects on the
water properties.
Drury Inlet and Porchcr Inlet both show
near-uniform
water properties, but only
one survey has been conducted in each.
Therefore, further data are required to determine if tidally induced vertical mixing
does occur on a regular basis.
Uniform conditions of temperature, salinity, and oxygen exist below 20 m in Fortune Channel, between Bedwell Sound and
Tofino Inlet on the west coast of Vancouver
Island (Fig. 15). Coote (1964) attributes
528
Drinkwater
and Osborn
7
Fig. 15. Detail
rounding area.
of Fortune
Channel
and sur-
these conditions to tidal action. A sill of
28 m in Matlset Narrows, which separates
Fortune Channel from Bedwell
Sound,
causes tidal currents of 3-4 knots. The
maximum depth in Fortune Channel is 140
m and lies just beyond the sill. The ratio
of the tidal prism to the total volume is
about 1 : 26 which compares closely with
1 : 30 for the Rupert-Holberg
basin. Since
the basin lies perpendicular to the narrows,
the shoreline directly opposite the narrows
may force the incoming tidal waters into
the deeper sections of the basin. RupertHolberg also has a shore opposite its narrows, and this may play a significant role
in the mixing process. The major freshwater inflow for Fortune Channel originates from the Kennedy River, which enters halfway up Tofino Inlet and flows
through Dawley Passage. Thus a much
less saline upper layer exists in Fortune
Channel than in Rupert Inlet.
Conclusions
Our study of the Rupcrt-Holberg
system
reveals a well mixed water body with large
monthly variations in density. Two processcs account for these observations. Decreases in density are caused by the mixing
of lower salinity water into the basin. Vigorous stirring, due to tidal action, mixes
the water just beyond the narrows. The
effectiveness of this mixing is enhanced because the turbulent
tidal intrusions are
denser than the existing low-salinity
surface laver. Enerev for this Drocess comes
from the tide. The depth and degree of
mixing changes with each tidal cycle. Increases in density are caused by the gravitational flow of dense water over the sill
directly into the deep part of the basin.
Both processes contribute to the high oxygen values in the basin. Continuous mixing exists by the action of one or the other
of these processes. Currents transport the
water from the region of the narrows to the
rest of the basin but we do not yet have a
detailed picture of the mixing away from
the narrows.
Fortune Channel exhibits like mixing
conditions, suggesting that a shoreline opposite the narrows may be important to dcfleet oncoming tidal water into the basin.
A classification of all inlets according to
their mixing properties would be useful in
the location of industries that plan to deposit wastes in inlets. Depending on the
type of disposal system required, such as
high oxygen demand, dilution,
etc., the
most suitable inlet could be chosen. If an
industry must be located on a specific inlet, knowledge of the mixing properties of
that inlet would help to determine where
the waste disposal system should be placed.
References
CANADIAN HYDROGRAPHIC SERVICX. 197 1. Canadian tide and current tables, v. 6. Queen’s
Printer.
1972. Tidal currents, Quatsino Narrows,
-.
British
Columbia.
Queen’s Printer.
and chemical
COOTE, A. R. 1964. A physical
study of Tofino
Inlet,
Vancouver
Island,
British Columbia.
M.S. thesis, Univ. British
Columbia,
Vancouver.
74 p.
LANE, R. K. 1962. A review of temperature and
salinity structures in the approaches to Vancouver Island, British Columbia.
J. Fish. Res.
Bd. Can. 19: 45-91.
PICKARD,
G. L. 1956. Physical fcatruxs of British
Columbia inlets. Trans. R. Sot. Can. Ser. 3
50: 47-58.
1961. Oceanographic
fcaturcs of inlets
-.
in the British Columbian
coast. J. Fish. Rcs.
Bd. Can. 18: 907-999.
1963. Oceanographic
characteristics
of
-.
inlets of Vancouver Island, I3ri tish Columbia.
J, Fish. Rcs. Bd. Can. 20: 1109-1144.
STRICKLAND, J. D. H., AND T. R. PARSONS. 1968.
A practical
handbook
of seawater analysis.
Bull. Fish. Res. Bd. Can. 167.
Tidal
TAYLOR, C. I. 1919.
Scn. Phil. Trans.
1-33.
TULLY, J. P. 1949.
tion of pulp mill
Bull. Fish. Rcs.
WAT,DICIIUK, M., J.
MIGKLE.
1968.
Tidal friction
R. Sot. Lond.
529
mixing
in the Irish
Ser. A 220:
Oceanography
and prcdicpolkion
in Albcrni
Inlet.
Bd. Can. 83.
R. MAIIICERT, AND J. II.
Physical
and
chemical
oceanographic
data from the west coast of
Vancouver
Island and the northern
British
Columbia
coast, 1957-1967.
Fish. Rcs. Bd.
Can. Ms. Rep. Ser. 990. 1681 p.
Submitted:
Accepted:
G May 1974
31 March 1975