Mixing in the Barents Sea Polar Front near Hopen in spring

1
Mixing in the Barents Sea Polar Front near Hopen in spring
2
Ilker Fer 1, 2 and Ken Drinkwater 3, 2
3
4
5
1 Geophysical Institute, University of Bergen, Bergen, Norway
2 Bjerknes Centre for Climate Research, Bergen, Norway
3 Institute of Marine Research, Bergen, Norway
Manuscript for:
Corresponding author:
Journal of Marine Systems
Dr. Ilker Fer
Geophysical Institute
University of Bergen
Allegaten 70
N-5007, Bergen
Norway
E-mail: Ilker.Fer@gfi.uib.no
Tel: +47 55 58 25 80
Fax: +47 55 58 98 83
1

6
Abstract
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Observations were made of hydrography, nutrients and shear microstructure in the water
column across the Barents Sea Polar Front near Hopen. Profiles were collected in early May
2008 along a 125-km section extending from the Hopen Trench, across the front between the
Arctic and Atlantic origin waters, and on toward the Spitsbergen Bank. Additionally a 10-h
time series station was undertaken near the front. The Polar Front was identified near the 150
m isobath, coinciding with the 1C isotherm, with strong horizontal gradients in the
temperature and salinity fields, which compensated in density. A second tidally-generated
front with a strong horizontal density gradient was detected on the bank, dominated by
salinity gradients due to ice melt. Biologic activity was elevated between the two fronts.
Nutrients were depleted in the euphotic zone where the chlorophyll concentrations were
significantly enhanced relative to the stations on the cold and warm side of the front. Below a
turbulent surface layer, the water column on the warm side of the front was quiescent. Farther
on the bank the turbulent boundary layers near the surface and seabed merged and the entire
water column was turbulent. Energy arguments show that tidal straining together with stirring
due to wind and tides can overcome the stratification induced by melting and heating,
maintaining the tidal front. A mechanism is proposed whereby high nitrate concentrations on
the warm side of the front are transported along the isopycnals on to the bank where tidal
mixing then effectively mixes the nutrient-rich deep water upward, sustaining the
phytoplankton bloom.
26
27
2

28
29
30
31
32
33
34
35
36
37
38
39
40
1. Introduction
The confluence of cold, low saline Polar waters of Arctic origin and warm, high saline waters
of Atlantic origin in the Barents Sea forms the Barents Sea Polar Front (BSPF). The physical
oceanographic conditions of the Barents Sea, including the main water masses and the
circulation patterns contributing to the formation of BSPF, have been described in Loeng
(1991) and Pfirman et al. (1994). The position of this thermohaline front is largely controlled
by the topography (Gawarkiewicz and Plueddemann, 1995). The summertime conditions
associated with the BSPF near Bear Island were described in Parsons et al. (1996). The
surface expression of the front in summer tends to be dominated by a surface salinity gradient
as a result of summer ice melt (Harris et al., 1998) and moves laterally on the order of 10 km
associated with tidal oscillations. Beneath this shallow density front and the mixed layer, is a
nearly barotropic, density compensating front with a moderate temperature gradient colocated
with the 2C isotherm.
41
42
43
44
45
46
47
48
49
Frontal systems, owing to the different water masses, are home to transitions in physical and
biogeochemical parameters as well as in the composition and productivity of species. Fronts
are often associated with enhanced primary production (Le Fèvre, 1987) and can be hot-spots
of marine life (Belkin et al., 2009). Key ecosystem components and basic food web structure
of the Barents Sea have been reviewed by Wassmann et al. (2006) and Loeng and Drinkwater
(2007). The BSPF plays an important role as a faunal boundary separating boreal and arctic
flora and fauna (Fossheim et al., 2006). In addition, certain fish species such as capelin
(Gjøsæter, 1998), as well as marine mammals, have been associated with the Barents Sea
Polar Front.
50
51
Cross-frontal transport and vertical mixing are essential to supply nutrients from the deeper
parts of the water column to the euphotic zone, and given sufficient illumination, these
3

52
53
54
55
56
57
58
59
60
61
62
63
64
processes might control the primary productivity. Cross-frontal exchange depends critically
on frontal instability and relaxation of geostrophic constraints through friction (Huthnance,
1995). The very presence of a front suggests a local reduction in turbulent diffusivity as the
density gradient suppresses mixing. Locally enhanced mixing rates, however, have been
observed in frontal regions and in the marginal ice zone of the Barents Sea (Fer and
Sundfjord, 2007; Sundfjord et al., 2007). Detailed knowledge of the mixing processes and
reliable measurements of turbulent fluxes are necessary to quantify the rate at which nutrients
become available for primary production. The International Polar Year (IPY) project entitled
Norwegian Ecosystem Studies of Subarctic and Arctic Regions (NESSAR) focused on the
physical and biological processes at the fronts separating the warmer Atlantic origin waters
and the colder waters of Arctic origin in both the Norwegian and Barents Sea. In this study
investigations of the dynamical frontal processes in the Barents Sea Polar Front near Hopen
are reported with focus on mixing and utilizing microstructure measurements.
65
66
67
68
69
70
71
72
73
74
75
2. Measurements and Data
Observations were made from the research vessel Jan Mayen between 24 April and 18 May
2008. The dataset reported here is a subset of the cruise data and includes vertical profiles of
hydrography from a conductivity-temperature-depth (CTD, Sea-Bird Electronics SBE911+)
package and of turbulence from 180 casts of a vertical microstructure profiler (MSS, ISW
Wassermesstechnik, Germany). The CTD was equipped with a fluorometer (Aqua-III) and a
SBE32 rosette with 12, 5-liter Niskin bottles. Nutrient samples were collected during the
upcast at discrete depths from the water bottles on the rosette. Chloroform was added to the
nutrient samples and the samples placed in a refrigerator on board the ship. Upon return to
port, they were transported to the Institute of Marine Research in Bergen to undergo analyses
for dissolved inorganic nutrients (nitrite, nitrate, orthophosphate and silicate) according to
4

76
standard methods (Parsons et al., 1992) adapted to an auto-analyser (Rey et al., 2000).
77
78
79
80
81
82
83
84
85
86
87
88
A station location map is shown in Figure 1. The sampling was made along a line extending
from the Hopen Trench in about 250 m depth, across the BSPF, and on toward the shallow
Spitsbergenbanken (Spitsbergen Bank) southwest of Hopen. The line was approximately 125
km long oriented along 145 from due East (i.e. -55T), approximately perpendicular to the
orientation of the 100 m isobath. Although we did not undertake a spatial survey of the
hydrography in the region, we assume that the position of BSPF is guided by the topography
(Gawarkiewicz and Plueddemann, 1995) and interpret the orientation of the transect as the
cross-front direction. The occupation time of stations with respect to the tides and the springneap
cycle is shown Figure 2. The sampling was undertaken during spring tides and in
transition to neaps. Sea ice was present on top of Spitsbergenbanken during the cruise;
however, the tendency was for a gradual reduction in the amount of ice over the duration of
the study.
89
90
91
92
93
94
Section A (the main sampling line) was occupied twice; in the first run, indicated by A0 in
Figure 2, the shipboard CTD was deployed at selected stations where nutrients were also
sampled. In the second run, indicated by A, station spacing was denser and the microstructure
profiler was deployed following the shipboard CTD at each station. Upon completing the A
section, a time series station (TS) was occupied for about 10 h at a site in the vicinity of the
BSPF with a depth of approximately 150 m.
95
96
97
98
The MSS was equipped with accurate CTD sensors, a fast response thermistor (FP07) and a
pair of microstructure airfoil shear probes used in measuring the viscous dissipation rate of
turbulent kinetic energy per unit mass (, dissipation rate hereafter). High resolution
temperature gradients were also sampled from the FP07 through a pre-emphasized channel.
5

99
100
101
102
103
104
105
106
Data from all channels were sampled during the downcast at 1024 Hz at a typical profiling
speed of 0.6-0.7 m s -1 . The MSS was deployed along Section A and as a time series close to
the front (station TS) (Figure 1). Profiles of temperature and salinity measured by MSS were
averaged in 10-cm vertical bins; turbulent variables were obtained at 1-m vertical bins.
Turbulence is inherently intermittent and averaging increases the statistical reliability of the
results. We therefore chose a sampling strategy where we collected 3 (Section A) to 5 (station
TS) continuous repeat casts at each station. All profiles presented in this study are ensemble
averages (isobaric) of the repeat casts.
107
108
109
110
111
112
113
114
Salinity data from the shipboard CTD were calibrated using water samples collected at the
bottom of the CTD cast from Niskin bottles attached to the rosette. The corrected CTD
profiles were then compared to the MSS-derived salinity. Using all MSS-SBE CTD profile
pairs close in time and space, 121 portions of data of at least a 5-m vertical stretch were
selected where the vertical salinity gradient was less than 110 -3 psu m -1 . The comparison of
average temperature and salinity values from these segments were in excellent agreement
within the measurement uncertainty (not shown) and no correction was deemed necessary for
the MSS-CTD profiles.
115
3. Calculation of dissipation rate and eddy diffusivity
116
The dissipation rate is calculated using the isotropic relation 7.5 uz
2
, where is the
117
viscosity of seawater. Small scale shear variance
2
u
z
is obtained by iteratively integrating
118
119
120
121
the low wavenumber portion of the shear spectrum of half-overlapping 1-second segments
(Fer, 2006). Unresolved shear variance in the high wavenumber portions, affected by noise,
are corrected for using the empirical theoretical shape (Oakey, 1982). The profiles of ε are
produced as 1 m vertical averages to a noise level of 10 -10 W kg -1 . The diapycnal eddy
6

122
diffusivity is then calculated using
K
2
0.2
N (Osborn, 1980) assuming 17% mixing
123
124
125
126
127
efficiency, and using the buoyancy frequency, 1/2
N g/
/ z
, from sorted
potential density anomaly, , profiles. The sorting of the density profile ensures a stable,
background stratification against which the turbulence works. The vertical gradient is
approximated by central first differencing and further smoothing the profile by a 10-m phasepreserving
moving average filter.
128
129
130
131
132
133
134
135
136
137
Diapycnal diffusion is assumed to be representative of that in the vertical direction, and for
sufficiently turbulent stratified flow, all passive scalars including temperature, salinity and
nutrients are assumed to be diffused vertically by the mixing coefficient K . The vertical flux,
positive upward, of a scalar concentration c is then F K c/
z. The eddy diffusivity
inferred from attains unrealistically large values for weak stratification as N approaches
zero. Due to the measurement error in density, N is not resolved to better than about 0.001 s -1
equivalent to 0.6 cycle per hour (cph). We therefore exclude the values of K from the
analysis for segments with N < 0.6 cph. In a stratified environment, turbulent stirring must be
strong enough to overcome the stratification to be able to produce a significant net buoyancy
flux resulting in turbulent mixing. Turbulence produces negligible vertical mixing when the
c
138
activity index
AI
1 2
N
is less than about 20 (Rohr et al., 1988). For the context of the
139
140
present study this implies that vertical nutrient fluxes can only occur for sufficiently large
values of A I .
141
4. Tides
142
143
Tidal velocity components were predicted from the 5-km resolution Arctic Ocean Tidal
Inverse Model (AOTIM-5, Padman and Erofeeva, 2004) at 25 stations along section A,
7

144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
separated approximately by 5 km, consistent with the model’s horizontal resolution. All 8
constituents available in the model were used (K 1 , O 1 , P 1 , Q 1 , M 2 , S 2 , N 2 , K 2 ). The water
depth in the model, at the extracted stations, was typically within 10% of the actual water
depth with an average of 7% error. Tidal currents were re-scaled using the actual water depth.
The time series of the tidal velocity components inferred at station TS, close to the front, is
shown in Figure 2. The M 2 constituent dominated, and among the diurnal constituents, K 1
was strongest. When averaged within ±5 km of the front (3 stations) the horizontal velocity
components in units of 10 -2 m s -1 ± one standard deviation are (u, v)= (14.4 ± 0.6, 6.6 ± 0.5)
for M 2 and (1.9 ± 0.07, 1.1 ± 0.05) for K 1 ; on the Polar Water (PW) side of the front,
averaged within 60 x 80 km (5 stations), (28.2 ± 2.5 , 17.9 ± 2.8) for M 2 and (2.7 ± 0.1,
3.8 ± 0.9) for K 1 . The maximum tidal excursion across the front, inferred from the
progressive vector diagram over one spring-neap cycle at the front position, is about ± 1.5
km. Repeating the same analysis for the mooring locations of the BSPF-experiment reported
in Parsons et al. (1996), we obtain similar results. This is less than ± 5 km lateral variability
of the surface expression of the BSPF observed near Bear Island in summer (Parsons et al.,
1996), suggesting that tidally induced lateral variability can be larger than the tidal excursion.
This advection will result in deformation of the frontal structure and variations in scalar
properties measured in the Eulerian frame. The data presented in this work are not corrected
for the tidal excursion.
163
5. Results and Discussion
164
165
166
167
5.1. Frontal structure
Horizontal distance versus depth distributions of the hydrographic properties and the
fluorescence sampled by the shipboard CTD at Section A show the presence of warm and
saline Atlantic Water (AW) and the cold and relatively fresh Polar Water (PW) separated by
8

168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
a thermohaline front (Figure 3). We referenced the horizontal distance, x, to the position of
the front, located at station 277, which was inferred from the horizontal temperature and
salinity gradient extrema (Figure 4). The front (x = 0) was near the 150 m isobath, extended
through the water column coinciding approximately with the 1C isotherm, and was density
compensated (Figure 3); i.e. the contributions of temperature and salinity anomalies to
density compensate such that there is not a significant horizontal density gradient across the
front. The CTD section showed the presence of a relatively dense layer near the bottom at the
location of the front. This layer was associated with positive temperature and salinity
anomalies relative to its surroundings. Such anomalously saline water can be a result of
brines rejected during ice formation on Spitsbergenbanken (Lien and Ådlansvik, 2011).
Similar to the dense water originated in the Storfjorden polynya (Fer et al., 2003; Fer and
Ådlandsvik, 2008), high salinity water formed on the bank is originally at freezing
temperature, but will entrain warm surrounding waters en-route from the formation region to
deeper water, explaining the anomalous temperature – salinity properties near the seabed,
near the front. Dense water formation on the bank can have consequences for the primary
production in the region, triggering an earlier start of the spring bloom (Lien and Ådlansvik,
2011).
185
186
187
188
189
190
The density, potential temperature, , salinity, S, and their horizontal gradients at 5 m and 50
m depth are shown in Figure 4. For a truly density compensating front, the horizontal density
ratio R T / S is unity. Here the temperature, T, and salinity, S, differences are taken
in the horizontal direction, is the coefficient of thermal expansion, and is the coefficient
of haline contraction. When averaged within 5 km of the front, R = 0.65 0.15 at 5 m and
0.74 0.09 at 50 m, i.e. salinity anomalies slightly dominate the density gradient. The density
9

191
192
ratio near the surface is approximately equal to that measured in the surface front at
Storbanken (Great Bank) in August 2007 (Våge et al., 2011).
193
194
195
196
197
198
199
200
On the PW side of the front on Spitsbergenbanken a second surface front existed at x ~ 72 km
which had a strong signature in density, particularly in the upper water column (Figure 3).
The variability in density was dominated by negative salinity anomalies as a result of ice
melt: averaged between x = 70 and 80 km R = -0.25 0.01 at 5 m and -0.11 0.01 at 50 m.
Fluorescence, a proxy for the biological activity, was near zero in the AW side, in the lower
part of the water column over the shelf, and also, rather surprisingly, at the front. The
elevated values of fluorescence were observed in the upper 40 m between x = 8 km and 60
km, i.e. between the two fronts.
201
202
203
204
205
206
207
208
209
210
211
212
213
5.2. Nutrients
Section A was thoroughly sampled for nutrients in its first occupation (A0). Distribution of
nitrate and chlorophyll-a are shown in Figure 5. For consistency, the front position is still that
identified for Section A (station 277) and may not be correct in this realization of the section
due to the tidal excursion and other temporal variability. Nitrate (NO 3 ), on average,
accounted for 95 (10)% of the total oxidized nitrogen (nitrate + nitrite). Chlorophyll
concentration is an approximate indicator of phytoplankton biomass. At about x = 15 km, on
the cold side of the front, chlorophyll concentration was significantly enhanced relative to the
two end stations on the warm and cold sides of the front with near zero concentrations.
Chlorophyll was severely undersampled by only three stations in this section, however, the
data suggest that nutrients were depleted in the upper layer through phytoplankton
production. A fraction of the high chlorophyll-a concentration between the two fronts can be
due to the winter time nutrient concentrations in the upper water column; the entire
10

214
enhancement can not be attributed to solely the vertical transport of nutrients.
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
The reservoir of nitrate was the warm AW, and the concentrations were large throughout the
entire water column away from the front. This was the case in all occupations of the stations
at x < -5 km, and was independent of the tides (see the profiles for stations 213 and 239 in
Figure 13a, introduced later). AW was characterized by nitrate concentrations above 10 mol
L -1 . Vertical distribution of the high nutrient concentration was typically homogenous in the
water column following the winter convection in the central Barents Sea. The weak
stratification on the AW side and on the PW side on the bank makes it difficult to achieve
high phytoplankton biomass in the euphotic zone, because even when production is initiated
it is distributed throughout the entire water column. The melting ice, on the other hand,
creates strong vertical stratification as observed between the two fronts that allows a build-up
of high phytoplankton biomass near surface. Cross-frontal transfer accumulated nutrients in
the deeper half of the water column on the cold side, and physical processes driving vertical
mixing supplied the nutrients to the upper part of the water column. Given sufficient
illumination in the euphotic zone, the phytoplankton blooms are controlled by the rate of
stirring. Microstructure measurements reveal the turbulent structure on the shelf and are
discussed in the following section.
231
5.3. Turbulence and mixing levels
232
233
234
235
236
5.3.1 Section A
Below a turbulent surface layer, the water column in the Polar Front and on the warm side of
the Front was found to be quiescent (Figure 6). Dissipation rates were close to the noise level.
From station 271 to about x = 23 km, the mixed layer depth, D mixed , was deeper than the
mixing depth, D mixing , by 22 m on the average over the 11 stations. On the shelf, the
11

237
238
239
240
241
242
243
244
difference D= D mixed - D mixing was variable, between -12 m to +8 m, but D

260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
stratified with the cold and low salinity upper layer separated from the warmer and saltier
water below in the bottom 50 m. The vertical separation of the isopycnals and the thickness
of AW- and PW-origin layers varied with the tides. The evolution of the warm layer in the
upper 20 m and the intrusions of cold water in the middle and warm water at the bottom were
likely a result of the tidal excursion. At about 3-4 h into the record, was enhanced near the
seabed above the background levels of O(10 -9 ) W kg -1 (Figure 7c). Fluctuating tidal currents
at semi-diurnal periodicity will modulate the bottom stress leading to one-quarter diurnal
variation in shear-production and dissipation rate with maxima at maximum flood and ebb
and minima around slack water (Simpson et al., 1996). Using the AOTIM-5 derived tidal
currents, a rough estimate of the dissipation rate had a maximum at about t = 3 h, a minimum
at about 6 h, followed by another maximum at 9 h. Assuming a drag coefficient of 3.510 -3 ,
and that the dissipation occurred in a 20-m thick bottom boundary layer, the corresponding
dissipation rate should have varied between 10 -6 and 310 -8 W kg -1 , consistent with the
observations. In order to avoid landing the microstructure profiler on the bottom, the profiles
were terminated above the bottom boundary layer, and the second dissipation maximum near
the bottom at t = 8-9 h could not be confirmed with the observations.
276
277
278
279
280
281
282
283
5.3.3 Average Profiles
Average profiles were calculated with respect to the horizontal distance x from the front.
Profiles collected at x < -10 km, -10 < x < 10 km, and 20 < x

284
285
time-averaged conditions at each location, but rather a single realization of the temporal and
spatial variability.
286
287
288
289
290
291
292
It is not trivial to obtain representative average profiles from a site with strong topographic
variability and lateral gradients in hydrography. We therefore present two alternatives:
isopycnal averages and averages in depth coordinate, z, normalized with the total water
depth, H. Isopycnally-averaged variables were obtained in 50 equally spaced bins of the
average profile and are assigned the average isopycnal depth. Each profile was sorted
before calculations. The normalized depth coordinate averaging was done over vertical bins
of z/H = 0.05 thickness.
293
294
295
296
297
298
299
Average dissipation rates were lowest at mid-depth on the AW side, independent of the
averaging method. Isopycnally-averaged profiles showed more than one order of magnitude
increase in dissipation below the upper layer from the AW to PW side of the front. Vertical
gradients of temperature and salinity that appeared in the isopycnally-averaged profiles at the
front were smeared out in z/H coordinates. A clear pattern emerged in z/H-averaged profiles
of : relative to mid-depth the dissipation rate was enhanced by up to two orders of
magnitudes in about z/H = 0.2 unit thick layers in the upper and bottom layers.
300
301
302
303
304
305
306
5.4. Isopycnal mixing
There was significant variability in temperature and salinity properties along isopycnal
surfaces sampled at the station TS. Fluctuations of T and S along the isopycnals relative to the
10-h mean are displayed in Figure 9. Due to the density compensating nature of the
thermohaline BSPF T S, i.e. their contributions to density are similar. The anomalies
appeared to be modulated at a time scale consistent with the tidal frequency suggesting that
the tides were driving along-isopycnal intrusions, whereby allowing the water on either side
14

307
308
309
310
311
of the front to be exchanged. Isopycnal frontal mixing was suggested to contribute to Polar
Front Water formation near Storbanken in August 2007 (Våge et al., 2011), manifested as
mixing lines along isopycnals accompanied with temperature-salinity inversions. Isopycnal
mixing was also observed in the BSPF at the southern flank of Spitsbergenbanken near Bear
Island in summer 1992 (Parsons et al., 1996).
312
313
314
5.5. Tidal stirring and straining
The amount of work per unit volume required to bring the water column of depth h to
complete mixing is
315
0
1
z
gzdz
h
h
316
where (z) is the density profile and
denotes averaging with respect to depth, z. The
z
317
318
319
320
321
potential energy anomaly, , in units of J m -3 has positive contribution from processes
increasing the stratification (e.g. heating and melting) and negative contributions from
vertical mixing (e.g. wind and tidal stirring) (Simpson and Bowers, 1981). The mean rate of
change of can be split into the following components, neglecting ice freezing, freshwater
runoff and horizontal advection
322
d
dt t t t t t
heat melt stir wind strain
(1)
323
324
325
Here, the terms are in units of W m -3 ( kg m -1 s -3 ) and are due to surface heat flux, ice
melting, tidal stirring, wind stress and tidal straining. A fraction, e t , of the depth integrated
dissipation of energy from tidal currents, D T , is available for work against gravity:
15

326
t
stir
eD
t
h
T
. (2)
327
Using a drag coefficient at bottom, C D , the tidal dissipation is
328
2 2
3/2
D C u v ,
T D z z
t
329
and
denotes averaging with respect to time, here over one spring-neap cycle. Studies of
t
330
331
tidal mixing in shelf seas suggest e t ~ 410 -3 (Simpson and Bowers, 1981). Wind of speed W
will lead to
332
t
wind
aewCDW
W
h
3
t
,
333
334
335
336
337
338
where C DW is the drag coefficient at the sea surface, e w is the efficiency of wind stirring, is
the ratio of wind-induced surface current to the wind speed, and a is the air density. The
study by Simpson and Bowers (1981) suggests e w =0.910 -3 . While typically the tidal stirring
dominates over the wind term over the European shelves, the role of wind mixing was found
to be important in predicting the position of tidal fronts in the Gulf of Maine region (Loder
and Greenberg, 1986).
339
The stabilizing components increasing the potential energy anomaly are
340
gQ
gdm
S S
;
t 2C t
2
heat
P
melt
i
,
341
342
where C P is the specific heat, Q is surface heat flux, d m is the melt rate, and S and S i are the
salinity of the water and ice, respectively.
16

343
344
345
346
347
In the presence of significant horizontal (longitudinal) density gradients and shear in the tidal
current, tidal straining (Simpson et al., 1990) will induce periodic modulation of
stratification, which is increased in ebbs and reduced in floods. The strain term contribution
to Eq.(1) has thus variable sign. For rectilinear tidal streams in the x direction, potential
energy is released (by convective mixing) or gained by tidal straining according to
348
0
g
u u zdz 0.03gh u
z
z
t h x . (3)
x
strain
h
349
The latter term is obtained for a typical quadratic tidal velocity profile (Simpson et al., 1990).
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
In calculating the different contributions to Eq.(1) the following parameters, all in SI units,
are chosen: = 510 -5 , = 7.810 -4 (for water at atmospheric pressure with S = 35 and =
0C), S i = 5, e t = 410 -3 , C D = 310 -3 , C DW = 1.510 -3 ; a = 1.2; = 1025, C P = 3990,
e w =0.910 -3 . We further assume a typical melt rate of 410 -7 m s -1 (i.e. approximately 1 m
thick ice melt in one month), and use monthly mean values of wind speed and solar radiation,
for May, of 6 m s -1 and 180 W m -2 , respectively (see Table 1 of Smedsrud et al. (2010)). The
assumed melt rate is typical for June, but is an overestimate for May conditions. The melting
contribution term is linear in melt rate, and we also report results for 0.5 m per month melt
rate. For tidal straining, in Eq.(3) the horizontal density gradient /x = 10 -5 kg m -4 (i.e.,
0.01 kg m -3 per 1 km, see Figure 4b) at x = 72 km is used. On the bank for x > 70 km, the
horizontal density gradient is relatively homogeneous in the vertical (Figure 10c). AOTIM-5
derived tidal velocity along section A, averaged over one spring-neap cycle, is used in Eqs.
(2) and (3). Averaged over 6 data points on the bank between 70-95 km, tidal strain and
stirring terms are found to be 5.4 (±0.06)10 -5 and 4.8 (±1.7)10 -6 W m -3 , respectively. The
ratio of the straining and stirring terms is R strain-stir = 11.3 using the average values or R strain-stir
17

365
= 12.6 ± 4.5 as the mean and standard deviation over the 6 data points.
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
Of the stirring terms, wind dominates for x< 45 km where the ratio of the mean tide and wind
stirring terms is R wind-stir = 0.1. Toward the bank the contribution due to tidal stirring
increases, and on the bank for x>70 km they are of comparable magnitude with R wind-stir = 0.9.
Melting increases the potential energy anomaly by 4.710 -5 W m -3 , approximately 4 times
that due to heating. For a reduced melt rate of 0.5 m per month, the melting term still
dominates with a contribution twice as much as the heating term. The stabilizing components
due to heat and melting are about 5 times the mixing contributions from wind and tidal
stirring (3.2 times for a reduced melt rate of 0.5 m per month). When tidal straining is
included, assuming that it destabilizes the water column in the ebb half cycle, the positive and
negative contributions nearly balance: the ratio of stabilizing and destabilizing terms are 1.5
and 0.9, respectively, for melt rates of 1 and 0.5 m per month. We conclude that the ice
melting dominates over heating in increasing the potential density anomaly. Stirring by winds
and tides, on the bank, is of comparable magnitude but not sufficient to completely mix the
water column. Tidal straining, periodically acting to reduce the stratification, dominates over
the wind and tide stirring, complements the stirring to overcome the positive induced by
melting and heating. A tidal front is thus maintained.
382
383
384
385
386
387
6. Implications for nutrient fluxes
Nitrate profiles collected at different times but at the same location show significant
variability that can be explained by tidal stirring and straining. Figure 10 shows the
occupation time, relative to the tide, of selected stations near the bank that were visited three
times, twice for nutrient sampling and the third time for microstructure profiling. The tidal
velocity was inferred at about x = 70 km, close to the station 222/249/259, and had a peak-to-
18

388
389
390
391
392
393
peak amplitude reaching 1 m s -1 during spring tides. The station locations on the section are
shown in Figure 10c together with the isopycnals. The water column on the bank, for x>80 m
was characterized by a strong horizontal density gradient, rather homogeneous throughout the
water column. Stations 218 to 222 were collected on floods during spring tides. While most
of the remaining profiles were collected also during flood periods later in transition to neap
tides, stations 249 and 259 were taken during ebb periods.
394
395
396
397
398
399
400
401
402
403
404
Pairs of nutrient profiles for each repeat station are shown in Figure 11. There was no
significant difference in the vertical structure of profiles at 220 and 247, occupied during
floods. At 247, occupied about 4 days layer, the nitrate concentration was higher below the
depleted zone, suggesting vertical transport of nutrients consistent with D mixing > D mixed
episodes observed at this site. This, however, could have been a result of temporal variability,
advection or internal wave heaving of density surfaces. The difference between the profiles
of 218 and 245 was more significant. The location of this station is closer to the region where
the mixing depth did not penetrate below the mixed layer depth (x < 23 km). We propose a
mechanism whereby high nitrate concentrations at the base of the pycnocline are transported
with the T and S anomalies along the isopycnals which deepen toward the PW side of the
front, which are then effectively mixed upward and then depleted by the phytoplankton.
405
406
407
408
409
410
The most striking difference was the well-mixed profile at 222 in contrast to significantly
stratified vertical structure of 249. Since station 249 was occupied during an ebb, tidal
straining will act to enhance stratification and suppress vertical mixing. Recall that profile
222 was taken at spring tides during flood. Tidal stirring together with the vertical mixing
favourable tidal straining will act to mix the nitrate profile. Nutrients brought on to the shelf
at deeper layers of the water column will thus be differentially mixed at different phases of
19

411
the tide, continually supplying the nutrients needed for phytoplankton growth.
412
413
414
415
416
417
418
419
420
421
422
423
424
425
The eddy diffusivity profiles collected at co-located stations confirm these inferences. Note,
however, that the profiles of nitrate concentration and K were not taken simultaneously (see
Figure 10a). The diffusivity at 253, collected during flood, was the largest with the turbulent
activity index 4 orders of magnitude larger than that required to achieve mixing in the entire
water column (Figure 12b). This station is also where we expect the influence of destabilizing
tidal straining and convective mixing during floods. The other profiles show attenuation in
diffusivity in the pycnocline where A I was not significantly above the threshold required for
vertical mixing. Although profiles 263 and 267 were collected close to flood the vertical
structure of density was not homogenous and tidal straining was not at play. Tidal stirring
effectively mixed the deeper layer, but mixing across the stratification was not achieved.
Profile 259 was taken during ebb when tidal straining enhanced the stratification. The phase
of the tide was comparable when the nitrate profile was taken at station 249; the distinct
vertical structure in the nitrate profiles between 222 and 249 can thus be related to tidal
straining.
426
427
428
429
430
431
432
433
434
Using selected co-located nutrient and eddy diffusivity profiles vertical flux of nitrate is
calculated along section A. Stations are chosen on the AW side, near the front and on the PW
side. Profiles of nitrate and the vertical flux of nitrate in units of mg m -2 s -1 are shown in
Figure 13. The stations span x = -27 to x = 52 km, before the tidal front. The vertical
concentration gradient on the AW side was not significant, and together with the weak eddy
diffusivity, resulted in negligible nutrient fluxes. Close to the front and on the PW side of the
front vertical fluxes were significantly enhanced. At x = - 9 km vertical fluxes reached
3.310 -3 mg m -2 s -1 , or about 290 mg m -2 per day, at 35 m. Typical range on the PW side of
the front was between 50 – 1200 mg m -2 per day in the upper 50 m. Maximum values were
20

435
436
437
438
439
440
441
442
443
250 mg m -2 per day at x = 36 km, and in excess of 1200 mg m -2 per day at x = 52 km, both
close to the surface at 13 m. These nutrient fluxes can be compared to those inferred from the
microstructure measurements undertaken in late July in the marginal ice zone in the northern
Barents Sea (Sundfjord et al., 2007). The largest vertical flux of about 150 mg m -2 per day
reported by Sundfjord et al. (2007) is nearly one order of magnitude less than the maximum
measured in our study. This, however, is a single data point whereas Sundfjord et al.’s result
is representative of the average flux over the pycnocline at a station with elevated levels of
vertical mixing. When averaged in the upper 50 m, the vertical nutrient flux at x = 52 km was
290 mg m -2 per day.
444
445
446
447
448
449
At station 245 (x = 36 km) the nitrate flux was elevated below the pycnocline characterized
by weak vertical fluxes between 20-30 m depth. Distributions of fluorescence (Figure 3d) and
chlorophyll-a concentration (Figure 5b) suggest negligible phytoplankton biomass below the
pycnocline. Because it is not expected that the nutrients are depleted below the pycnocline,
the divergence in the vertical flux suggests lateral export of nutrients. We hypothesize that
this export can be established along the isopycnals.
450
451
452
453
454
455
456
457
Figure 13c shows the - S diagram for the water samples taken for nutrient analysis. The
data points are color coded for the nitrate concentration. At stations on the AW side of the
front (231, 239 and 243), the nitrate concentration is large and the data points trace a curve
close to the = 28 isopycnal. On the PW side of the front, high nitrate concentration in the
deeper part of the water column is then vertically transported upward by turbulent mixing.
The vertical mixing is most pronounced well into the cold side of the front, on the bank,
where the turbulent activity index is large and tidal stirring contributes to complete mixing of
the entire water column.
21

458
459
460
461
462
463
464
465
466
467
468
7. Summary
Measurements of hydrography, nutrients and turbulence were undertaken in early May 2008
along a section across the Barents Sea Polar Front (BSPF) southwest of Hopen. The BSPF
was characterized by a density-compensated thermohaline front near the 150 m isobath
coinciding approximately with the 1C isotherm. Approximately 70 km into the cold side of
the BSPF a well-defined tidal front was detected where the horizontal density gradient,
homogeneous in the water column, exceeded 0.01 kg m -3 per km, dominated by negative
salinity anomalies as a result of ice melt. Biological activity was observed to be elevated
between the two fronts. Nutrients were depleted in the euphotic zone where the chlorophyll
concentrations were significantly enhanced relative to the stations at the cold and warm side
of the BSPF.
469
470
471
472
473
474
475
476
Below the turbulent surface layer, the water column in the BSPF and on the warm Atlantic
side was quiescent. Farther onto the Spitsbergenbanken, in half of the stations the mixing
layer at the surface was typically deeper than the mixed layer depth by about 9 m, suggesting
and the euphotic zone intermittently entrained nutrients from the deeper layer. On the bank
the turbulent boundary layers near the surface, caused by wind mixing, and near the seabed,
caused by tidal mixing, merged and the entire water column was turbulent. Near the BSPF,
temperature and salinity anomalies along the isopycnals, relative to 10-h average, revealed
along isopycnal intrusions with variability at a time scale consistent with the tidal period.
477
478
479
480
481
Ice melting was found to dominate over heating in increasing the potential density anomaly
on the shelf. Stirring by winds and tides on the bank were of comparable magnitude but not
sufficient to completely mix the water column. Tidal straining dominated over the wind and
tide stirring, and contributed to overcome the stratification induced by melting and heating.
Nutrient profiles sampled at different phases of the tide at the same location show
22

482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
significant variability that can be explained by tidal stirring and straining. Stations occupied
during ebb tide reveal weak diffusivity levels; tidal straining enhances the stratification and
suppresses vertical mixing. During flood tide, on the other hand, tidal stirring together with
vertical mixing favourable tidal straining mix the nutrient concentrations in the water column.
While vertical fluxes of nitrate were insignificant on the warm side of the front, on the bank,
they were calculated to be in the range of 50 – 1200 mg m -2 per day. Near the bank, vertical
profiles of nitrate flux showed mid-depth minimum increasing toward the bottom and
surface, the divergence of the flux thus suggest lateral export of nutrients. A mechanism is
proposed whereby high nitrate concentrations on the warm side of the BSPF are transported
with the T and S anomalies along the isopycnals which deepen toward the tidal front. Tidal
mixing then effectively mixes the nutrient rich deep water upward, sustaining the
phytoplankton bloom. The low turbulence levels in the Polar Front coupled with the low
chlorophyll-a concentrations suggest that the Front was not a region of high primary
production. This is consistent with the results found for the density compensating Arctic
Front in the Norwegian Sea (Erga et al., 2011).
23

497
498
Acknowledgments
499
500
501
The field work and KD received funding from the Research Council of Norway through the
NESSAR project (# 176057). This is publication number XXX from the Bjerknes Centre for
Climate Research.
24

502
REFERENCES
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
Belkin, I.M., Cornillon, P.C. and Sherman, K., 2009. Fronts in Large Marine Ecosystems.
Progress In Oceanography, 81(1-4): 223-236, doi: 10.1016/j.pocean.2009.04.015.
Erga, S.R., Ssebiyonga, N., Hamre, B., Frette, Ø., Hovland, E., Drinkwater, K. and Rey, F.,
2011. Environmental control of phytoplankton distribution and photosynthetic
capacity at the Jan Mayen Front in the Norwegian Sea. Journal of Marine Systems,
submitted (this issue).
Fer, I., 2006. Scaling turbulent dissipation in an Arctic fjord. Deep-Sea Research Part II-
Topical Studies in Oceanography, 53(1-2): 77-95.
Fer, I., Skogseth, R., Haugan, P.M. and Jaccard, P., 2003. Observations of the Storfjorden
overflow. Deep-Sea Research Part I-Oceanographic Research Papers, 50(10-11):
1283-1303, doi: 10.1016/s0967-0637(03)00124-9.
Fer, I. and Sundfjord, A., 2007. Observations of upper ocean boundary layer dynamics in the
marginal ice zone. Journal of Geophysical Research, 112(C4): C04012, doi:
10.1029/2005jc003428.
Fer, I. and Ådlandsvik, B., 2008. Descent and mixing of the overflow plume from Storfjord
in Svalbard: An idealized numerical model study Ocean Science, 4: 115-132.
Fossheim, M., Nilssen, E.M. and Aschan, M., 2006. Fish assemblages in the Barents Sea.
Marine Biology Research, 2: 260-269.
Gawarkiewicz, G. and Plueddemann, A.J., 1995. Topographic control of thermohaline frontal
structure in the Barents Sea Polar Front on the south flank of Spitsbergen Bank.
Journal of Geophysical Research-Oceans, 100(C3): 4509-4524, doi:
10.1029/94jc02427.
25

525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
Gjøsæter, H., 1998. The population biology and exploitation of capelin (Mallotus villosus) in
the Barents Sea. Sarsia, 83: 453-496.
Harris, C.L., Plueddemann, A.J. and Gawarkiewicz, G.G., 1998. Water mass distribution and
polar front structure in the western Barents Sea. Journal of Geophysical Research-
Oceans, 103(C2): 2905-2917, doi: 10.1029/97jc02790.
Huthnance, J.M., 1995. Circulation, exchange and water masses at the ocean margin: the role
of physical processes at the shelf edge. Progress In Oceanography, 35(4): 353-431.
Le Fèvre, J., 1987. Aspects of the biology of frontal systems, Advances in Marine Biology.
Academic Press, pp. 163-299.
Lien, V.S. and Ådlansvik, B., 2011. Bottom water formation as a primer for spring-blooms
on Spitsbergenbanken? Journal of Marine Systems, submitted (this issue).
Loder, J.W. and Greenberg, D.A., 1986. Predicted positions of tidal fronts in the Gulf of
Maine region. Continental Shelf Research, 6(3): 397-414.
Loeng, H., 1991. Features of the physical oceanographic conditions of the Barents Sea. Polar
Research, 10: 5-18.
Loeng, H. and Drinkwater, K., 2007. An overview of the ecosystems of the Barents and
Norwegian Seas and their response to climate variability. Deep Sea Research Part II:
Topical Studies in Oceanography, 54(23-26): 2478-2500, doi:
10.1016/j.dsr2.2007.08.013.
Oakey, N.S., 1982. Determination of the rate of dissipation of turbulent energy from
simultaneous temperature and velocity shear microstructure measurements. Journal of
Physical Oceanography, 12(3): 256-271.
26

547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
Osborn, T.R., 1980. Estimates of the local rate of vertical diffusion from dissipation
measurements. Journal of Physical Oceanography, 10(1): 83-89.
Padman, L. and Erofeeva, S., 2004. A barotropic inverse tidal model for the Arctic Ocean.
Geophysical Research Letters, 31(2): L02303, doi: 10.1029/2003gl019003.
Parsons, A.R., Bourke, R.H., Muench, R.D., Chiu, C.S., Lynch, J.F., Miller, J.H.,
Plueddemann, A.J. and Pawlowicz, R., 1996. The Barents Sea Polar Front in summer.
Journal of Geophysical Research, 101(C6): 14201-14221.
Parsons, T.R., Maita, Y. and Lalli, C.M., 1992. A manual of chemical and biological methods
for sea water analysis. Pergamon Press, New York, 173 pp.
Pfirman, S.L., Bauch, D. and Gammelsrød, T., 1994. The Northern Barents Sea: Water mass
distribution and modification. In: O.M. Johannessen, R.D. Muench and J.E. Overland
(Editors), The polar oceans and their role in shaping the global environment: the
Nansen centennial volume. Geophysical Monograph 85, American Geophysical
Union, pp. 77-94.
Rey, F., Noji, T.T. and Miller, L.A., 2000. Seasonal phytoplankton development and new
production in the central Greenland Sea. Sarsia, 85: 329-344.
Rohr, J.J., Itsweire, E.C., Helland, K.N. and Van Atta, C.W., 1988. Growth and decay of
turbulence in a stably stratified shear flow. Journal of Fluid Mechanics, 195: 77-111.
Simpson, J., Brown, J., Matthews, J. and Allen, G., 1990. Tidal straining, density currents,
and stirring in the control of estuarine stratification. Estuaries and Coasts, 13(2): 125-
132, doi: 10.2307/1351581.
Simpson, J.H. and Bowers, D., 1981. Models of stratification and frontal movement in shelf
seas. Deep Sea Research Part A. Oceanographic Research Papers, 28(7): 727-738.
27

570
571
572
573
574
575
576
577
578
579
580
581
582
583
Simpson, J.H., Crawford, W.R., Rippeth, T.P., Campbell, A.R. and Cheok, J.V.S., 1996. The
vertical structure of turbulent dissipation in shelf seas. Journal of Physical
Oceanography, 26(8): 1579-1590.
Smedsrud, L.H., Ingvaldsen, R., Nilsen, J.E.Ø. and Skagseth, Ø., 2010. Heat in the Barents
Sea: transport, storage, and surface fluxes. Ocean Science, 6(1): 219-234, doi:
10.5194/os-6-219-2010.
Sundfjord, A., Fer, I., Kasajima, Y. and Svendsen, H., 2007. Observations of turbulent
mixing and hydrography in the marginal ice zone of the Barents Sea. Journal of
Geophysical Research, 112(C5): C05008, doi: 10.1029/2006jc003524.
Våge, S., Basedow, S.L., Tande, K.S. and Zhou, M., 2011. Physical structure of the Barents
Sea Polar Front near Storbanken in August 2007. Journal of Marine Systems,
submitted (this issue).
Wassmann, P. et al., 2006. Food webs and carbon flux in the Barents Sea. Progress In
Oceanography, 71(2-4): 232-287, doi: 10.1016/j.pocean.2006.10.003.
584
28

585
Figure captions
586
587
588
589
590
591
592
593
Figure 1. Location map of the study region together with the place names and the positions of
the stations. Bathymetric contours are drawn at 25 m intervals. The inset identifies the region
(box) in the Barents Sea shown in detail and the 1000-m isobath (gray). Microstructure
profiles were collected along section A (dots), at the time series station TS (square) and at
stations 231 and 239, marked by diamonds, occupied prior to section A. Shipboard CTD
measurements were made at each station of section A, and also during an earlier occupation
of the section (A0) with coarser station spacing (open circles). Nutrient sampling was
conducted at all stations indicated by circles. 33
594
595
596
597
598
599
Figure 2. Occupation of stations relative to the spring-neap cycle. The east, u, and north, v,
components of the tidal velocity are inferred from AOTIM-5 (Padman and Erofeeva, 2004) at
the position of station TS. Gray shading marks the duration of the first occupation of section
A (A0, stations 213 to 225) when only CTD was deployed, and Section A and station TS
when the microstructure profiler was deployed. Occupation of individual stations is indicated
by arrows and station numbers. 34
600
601
602
603
604
605
Figure 3. Contours of a) potential temperature, , b) potential density anomaly , c) salinity,
S, and d) fluorescence along Section A measured by the SBE911+ CTD unit. The section is
extended to include station 231. The horizontal distance is referenced to station 277 which is
identified as the front position (vertical dashed line). Arrow heads mark the stations with the
station number from the cruise log. Bottom depth is inferred from the ship’s echo sounder.
Data are not available in the gray portions near the bottom. 35
606
607
Figure 4. Properties along section A evaluated at 5 m (black) and 50 m (red) depth, averaged
within 1-m vertical extend of the target depth. Horizontal distance is referenced to the front at
29

608
609
610
station 277 identified as the extrema in the horizontal salinity and temperature gradients.
Panels show a) the potential density anomaly , c) potential temperature, , e) salinity, and
their horizontal gradients in property unit per km in panels b, d and f. 36
611
612
613
614
615
616
617
Figure 5. Contours of a) nitrate (color) and potential temperature, (black) and b)
chlorophyll-a (color) and potential density anomaly, (black) for the first occupation of
Section A (A0). Note that the nitrate is sampled at all stations identified by arrowheads but
chlorophyll is sampled in only 3 stations (213, 216 and 225) and interpolated linearly
between the stations. Nitrate is given in units of mol/L, equivalent to 10 -3 mol m -3 or 62 mg
m -3 . Gray dots mark the sampling depths for the corresponding parameter. Temperature and
density values are available at all sampling depths. 37
618
619
620
621
622
623
624
Figure 6. Properties along Section A measured by MSS: contours of a) potential temperature
(black), , and log 10 of dissipation rate (color), b) potential density anomaly (black) and
log 10 of eddy diffusivity, K (color). The section is extended to include station 231. The white
contours in b) are N = 0.6 cph and show where the stratification is weak. The horizontal
distance is referenced to the front position (vertical dashed line). Arrow heads mark the
stations with the station number from the cruise log. Bottom depth is inferred from the ship’s
echo sounder. Data are not available in the gray portions near the bottom. 38
625
626
627
628
629
Figure 7. Properties at station TS measured by MSS: contours of a) potential temperature, ,
b) salinity, and c) log 10 of dissipation rate . In all panels contours of the potential density
anomaly are also shown in black. The time is referenced to the start of the first cast on 15
May 2008 20:50 UTC. Arrow heads mark individual casts. Each batch of 5 cast is averaged
(vertical dashed lines) and used in producing the figures. In c) vertical profiles of
30

630
631
fluorescence (red) are shown in units of g/l with scale at the bottom. The deep parts of the
fluorescence profiles have values near zero and mark the time when the profile is taken. 39
632
633
634
635
636
Figure 8. Profiles of potential temperature, , salinity, S, potential density anomaly ,
dissipation rate , and eddy diffusivity, K averaged over stations in the AW side (red), PW
side (blue) of the front and at the front (FR, black). Upper row shows profiles averaged
isopycnally. Lower row shows profiles averaged with respect to depth, z, normalized by the
total water depth, H. 40
637
638
639
640
Figure 9. Anomalies of (color) a) temperature and b) salinity along isopycnals at station TS.
The anomalies are multiplied by the coefficient of thermal expansion and haline contraction,
respectively, to indicate their contribution to density. Contours of temperature and salinity
along isopycnals are shown in black. 41
641
642
643
644
645
646
Figure 10. Summary of the time and location of revisited stations. a-b) Time series of the east
component of the tidal velocity together with the time of occupation of the stations marked
with dashed lines and station numbers c) Contours of (thick: 0.1 unit intervals, gray, thin:
0.02 unit intervals) along the PW side of the front. The co-located stations are indicated with
station numbers. Nitrate profiles from 218/245, 220/247 and 222/249 are shown in Figure 11.
Turbulent parameters for stations 253, 259, 263 and 267 are shown in Figure 12. 42
647
648
649
650
651
Figure 11. Profiles of nitrate at three stations on the PW side of the front at approximately a)
x = 37 km, b) x = 53 km, and c) x = 67 km. See Figure 10 for location on Section A and time
relative to tides. Each location is sampled twice, during springs (218, 220 and 222) and about
3.7 day later in transition to neaps. All profiles are collected during flood, except station 249
during the ebb tide. 43
31

652
653
654
655
656
657
Figure 12. Profiles of a) eddy diffusivity, K , and b) turbulent activity index, A I for the
selected stations on the cold side of the front. See Figure 10 for location on Section A and
time relative to tides. Stations 267, 263 and 259 are co-located with 245/218, 247/220 and
249/222, respectively, where nutrients are sampled. No nutrient sampling was made when the
microstructure was deployed. Vertical dashed line in b) shows the approximate threshold
below which no vertical mixing is expected. 44
658
659
660
661
662
663
Figure 13. Profiles of a) nitrate and b) vertical flux of nitrate at stations indicated by station
numbers and distance from the front. Nitrate concentration is converted to mg m -3 before
calculating the vertical flux. Temperature-salinity diagram for the same water samples colorcoded
for nitrate concentration. Contours of are drawn at 0.1 unit intervals. Arrows
indicated suggested pathways of mixing, along isopycnals until well into the cold side of the
front, across the isopycnals where tidal stirring and straining becomes important. 45
664
665
32

Greenland
77 o N
Svalbard
Barents
Sea
Hopen
Hopen−
banken
200
Norway
100
76 o N
100
225
Spitsbergenbanken
100
223
75 o N
20 o E 25 o E 30 o E
A
249
247
245
219
200
TS
216
243
239
213
231
300
Hopen Trench
666
667
668
669
670
671
672
673
674
Figure 1. Location map of the study region together with the place names and the positions of
the stations. Bathymetric contours are drawn at 25 m intervals. The inset identifies the region
(box) in the Barents Sea shown in detail and the 1000-m isobath (gray). Microstructure
profiles were collected along section A (dots), at the time series station TS (square) and at
stations 231 and 239, marked by diamonds, occupied prior to section A. Shipboard CTD
measurements were made at each station of section A, and also during an earlier occupation
of the section (A0) with coarser station spacing (open circles). Nutrient sampling was
conducted at all stations indicated by circles.
675
33

A0 ↓ ↓↓↓
A
TS
Tidal Velocity [cm s −1 ]
25
20
15
10
5
0
−5
−10
−15
−20
−25
u
v
231
239
243
245
247
676
01 03 05 07 09 11 13 15 17 19
Day of May 2008
677
678
679
680
681
682
Figure 2. Occupation of stations relative to the spring-neap cycle. The east, u, and north, v,
components of the tidal velocity are inferred from AOTIM-5 (Padman and Erofeeva, 2004) at
the position of station TS. Gray shading marks the duration of the first occupation of section
A (A0, stations 213 to 225) when only CTD was deployed, and Section A and station TS
when the microstructure profiler was deployed. Occupation of individual stations is indicated
by arrows and station numbers.
34

683
0
231
280
278
276
274
272
270
268
266
264
262
260
258
251
253
255
3
231
280
278
276
274
272
270
268
266
264
262
260
258
251
253
255
28.05
Depth [m]
50
100
150
200
250
0
2.5
1.5
0
−1.5
a) θ [°C]
2
1
0
−1
−2
35.1
28
27.95
b) σ θ
28
27.95
27.9
27.85
27.8
27.75
27.7
12
684
Depth [m]
50
100
34.8
34.9
35
150
200
250
c) S
−20 0 20 40 60 80
Distance [km]
35
34.9
34.8
34.7
34.6
34.5
34.4
d) Fluor [μg/l]
−20 0 20 40 60 80
Distance [km]
10
8
6
4
2
0
685
686
687
688
689
690
Figure 3. Contours of a) potential temperature, , b) potential density anomaly , c) salinity,
S, and d) fluorescence along Section A measured by the SBE911+ CTD unit. The section is
extended to include station 231. The horizontal distance is referenced to station 277 which is
identified as the front position (vertical dashed line). Arrow heads mark the stations with the
station number from the cruise log. Bottom depth is inferred from the ship’s echo sounder.
Data are not available in the gray portions near the bottom.
35

σ θ
28
27.9
27.8
a
50 m
5 m
b
0.01
0
−0.01
∂σ θ
/∂x
θ
2
1
0
−1
c
d
0.1
0
−0.1
∂θ/∂x
35
e
f
0.01
S
34.8
34.6
0
−0.01
∂S/∂x
691
−25 0 50 100
Distance (km)
−25 0 50 100
Distance (km)
692
693
694
695
696
Figure 4. Properties along section A evaluated at 5 m (black) and 50 m (red) depth, averaged
within 1-m vertical extend of the target depth. Horizontal distance is referenced to the front at
station 277 identified as the extrema in the horizontal salinity and temperature gradients.
Panels show a) the potential density anomaly , c) potential temperature, , e) salinity, and
their horizontal gradients in property unit per km in panels b, d and f.
36

697
0
213
215
217
219
221
223
225
Nitrate
10
Depth [m]
50
100
2
1
0
0
0
−0.5
−1.5
8
6
4
Depth [m]
150
200
0
50
100
150
a) Nitrate [μmol/L], θ [°C]
27.75
27.95 27.85 27.7
27.95
2
0
Chl−A
9
8
7
6
5
4
3
2
698
200
b) Chlorophyll [mg/m 3 ], σ θ
0 20 40 60 80 100
Distance [km]
1
0
699
700
701
702
703
704
705
Figure 5. Contours of a) nitrate (color) and potential temperature, (black) and b)
chlorophyll-a (color) and potential density anomaly, (black) for the first occupation of
Section A (A0). Note that the nitrate is sampled at all stations identified by arrowheads but
chlorophyll is sampled in only 3 stations (213, 216 and 225) and interpolated linearly
between the stations. Nitrate is given in units of mol/L, equivalent to 10 -3 mol m -3 or 62 mg
m -3 . Gray dots mark the sampling depths for the corresponding parameter. Temperature and
density values are available at all sampling depths.
37

706
Depth [m]
0
50
100
150
231
2.5
2
280
278
276
274
272
270
268
266
264
262
260
258
0.5 0 −1.5
251
253
255
log 10
(ε)
−6
−6.5
−7
−7.5
−8
−8.5
200
−9
250
0
a) θ [°C], ε [W kg −1 ]
−9.5
−10
log 10
(K ρ
)
0
707
Depth [m]
50
100
150
200
250
27.95
27.8
28
b) σ , K [m 2 s −1 ]
θ ρ
−20 0 20 40 60 80
Distance [km]
−1
−2
−3
−4
−5
−6
708
709
710
711
712
713
714
Figure 6. Properties along Section A measured by MSS: contours of a) potential temperature
(black), , and log 10 of dissipation rate (color), b) potential density anomaly (black) and
log 10 of eddy diffusivity, K (color). The section is extended to include station 231. The white
contours in b) are N = 0.6 cph and show where the stratification is weak. The horizontal
distance is referenced to the front position (vertical dashed line). Arrow heads mark the
stations with the station number from the cruise log. Bottom depth is inferred from the ship’s
echo sounder. Data are not available in the gray portions near the bottom.
38

0
27.94
θ
0.4
Depth [m]
50
100
27.99
27.98
0.2
0
−0.2
715
Depth [m]
Depth [m]
150
0
50
100
150
0
50
100
150
a)
b)
c)
27.99
27.99
27.98
Fluor.
0 10
27.94
0 1 2 3 4 5 6 7 8 9
Time [h]
−0.4
−0.6
S
34.9
34.88
34.86
34.84
34.82
34.8
34.78
34.76
log 10
(ε)
−6
−6.5
−7
−7.5
−8
−8.5
−9
−9.5
−10
716
717
718
719
720
721
722
Figure 7. Properties at station TS measured by MSS: contours of a) potential temperature, ,
b) salinity, and c) log 10 of dissipation rate . In all panels contours of the potential density
anomaly are also shown in black. The time is referenced to the start of the first cast on 15
May 2008 20:50 UTC. Arrow heads mark individual casts. Each batch of 5 cast is averaged
(vertical dashed lines) and used in producing the figures. In c) vertical profiles of
fluorescence (red) are shown in units of g/l with scale at the bottom. The deep parts of the
fluorescence profiles have values near zero and mark the time when the profile is taken.
39

723
Depth [m]
0
50
100
150
0
0.2
AW
FR
PW
a b c d e
i
j
PW
724
−z/H
0.4
0.6
0.8
1
−1 0 1 2
θ [°C]
FR
AW
f g h
34.6 34.8 35 27.9 28 10 −9 10 −8 10 −7 10 −6
10 −5 10 −3 10 −1
S
σ θ ε [W kg −1 ] K ρ
[m 2 s −1 ]
725
726
727
728
729
730
Figure 8. Profiles of potential temperature, , salinity, S, potential density anomaly ,
dissipation rate , and eddy diffusivity, K averaged over stations in the AW side (red), PW
side (blue) of the front and at the front (FR, black). Upper row shows profiles averaged
isopycnally. Lower row shows profiles averaged with respect to depth, z, normalized by the
total water depth, H.
731
40

732
Mean Isopycnal Depth [m] Mean Isopycnal Depth [m]
0
20
0.2
0.1
40
60
80
−0.4
−0.1
0.1
0.2
100
a)
0
20
34.8
34.78
34.79
40
34.81
60
34.83
34.82
34.86
34.86
80
100
b)
0 1 2 3 4 5 6 7 8 9
Time [h]
10 5 αΔT
2
1.2
0.4
−0.4
−1.2
−2
10 5 βΔS
2
1.2
0.4
−0.4
−1.2
−2
733
734
735
736
Figure 9. Anomalies of (color) a) temperature and b) salinity along isopycnals at station TS.
The anomalies are multiplied by the coefficient of thermal expansion and haline contraction,
respectively, to indicate their contribution to density. Contours of temperature and salinity
along isopycnals are shown in black.
737
41

u [cm s −1 ]
50
0
−50
0
218
220
222
a)
09 Day of May 2008 10
245
247
249
253
259
263
267
b)
13 14
Day of May 2008
27.9
27.8
738
Depth [m]
50
100
218−245−267
220−247−263
222−249−259
253
30 40 50 60 70 80 90
Distance [km]
c)
739
740
741
742
743
744
745
Figure 10. Summary of the time and location of revisited stations. a-b) Time series of the east
component of the tidal velocity together with the time of occupation of the stations marked
with dashed lines and station numbers c) Contours of (thick: 0.1 unit intervals, gray, thin:
0.02 unit intervals) along the PW side of the front. The co-located stations are indicated with
station numbers. Nitrate profiles from 218/245, 220/247 and 222/249 are shown in Figure 11.
Turbulent parameters for stations 253, 259, 263 and 267 are shown in Figure 12.
746
42

0
a b c
25
Depth [m]
50
75
100
218
245
220
247
222
249
747
0 2 4 6 8 10
Nitrate [μmol/L]
0 2 4 6 8 10
Nitrate [μmol/L]
0 2 4 6 8 10
Nitrate [μmol/L]
748
749
750
751
752
753
Figure 11. Profiles of nitrate at three stations on the PW side of the front at approximately a)
x = 37 km, b) x = 53 km, and c) x = 67 km. See Figure 10 for location on Section A and time
relative to tides. Each location is sampled twice, during springs (218, 220 and 222) and about
3.7 day later in transition to neaps. All profiles are collected during flood, except station 249
during the ebb tide.
43

754
0
a
b
25
Depth [m]
50
75
100
267
263
259
253
755
10 −6 10 −4 10 −2
K ρ
[m 2 s −1 ]
10 0 10 2 10 4 10 6
Activity Index, A I
756
757
758
759
760
761
762
Figure 12. Profiles of a) eddy diffusivity, K , and b) turbulent activity index, A I for the
selected stations on the cold side of the front. See Figure 10 for location on Section A and
time relative to tides. Stations 267, 263 and 259 are co-located with 245/218, 247/220 and
249/222, respectively, where nutrients are sampled. No nutrient sampling was made when the
microstructure was deployed. Vertical dashed line in b) shows the approximate threshold
below which no vertical mixing is expected.
44

763
Depth [m]
0
20
40
θ [°C]
2
1
0
c
27.7 27.9
28.1
NO 3
10
8
6
4
2
100
120
-1
34.6 34.8 35
S
0
764
140
a
160
0 5 10
Nitrate [µmol/L]
b
10 −4 10 −3 10 −2
Nitrate Flux [mg m -2 s -1 ]
231, x = -27 km
239, x = -10 km
243, x = -9 km
245, x = 36 km
247, x = 52 km
765
766
767
768
769
770
771
Figure 13. Profiles of a) nitrate and b) vertical flux of nitrate at stations indicated by station
numbers and distance from the front. Nitrate concentration is converted to mg m -3 before
calculating the vertical flux. Temperature-salinity diagram for the same water samples colorcoded
for nitrate concentration. Contours of are drawn at 0.1 unit intervals. Arrows
indicated suggested pathways of mixing, along isopycnals until well into the cold side of the
front, across the isopycnals where tidal stirring and straining becomes important.
45