Deutsche Hydrographische Zeitschrift German Journal of Hydrography Volume 49 (1997) Number 2/3
ISSN 0012-0308 9 BSH, Hamburg und Rostock
Interaction between Baltic Sea and North Sea Bo GUSTAFSSON
Summary The water exchange between the Baltic Sea and the North Sea involves many processes, some more investigated and understood than others. However, the processes that are of major importance are quite well known. The aim of the present paper is to review current knowledge about the water exchange. It is known that freshwater input to the Baltic Sea and the strength of the oscillating flows across the sills of the Danish Straits both have a large impact on the salinity of the Baltic Sea. This is demonstrated using a simple analytical model. The dynamics of the outflow from the Skagerrak by the Norwegian Coastal Current are discussed in some detail. Especially the importance of wind forced currents on the magnitude of the outflow is demonstrated.
Wechselwirkungen zwischen Ostsee und Nordsee (Zusammenfassung) Der Wasseraustausch zwischen Ostsee und Nordsee enth< viele Prozesse, von denen einige mehr erforscht und verstanden sind als andere. Jedoch k6nnen Prozesse von grol3er Bedeutung als bekannt vorausgesetzt werden. Das Anliegen dieser Arbeit ist es, einen Uberblick 0ber den gegenw&rtigen Wissensstand 0ber den Wasseraustausch zu geben. Es ist bekannt, dal3 der SQI3wassereintrag in die Ostsee und die Intensit~.t wechselnder Str6mungen 0ber die Schwellen in den d&nischen Gew~tssern einen groi3en Einflul3 auf den Salzgehalt der Ostsee haben. Dieses wird mit Hilfe eines einfachen analytischen Modells gezeigt. Die Dynamik des Ausstroms vom Skagerak mit dem norwegischen KL)stenstrom wird im Detail er6rtert. Insbesondere wird die Beduetung der winderzeugten Str6mungen fQr die St&rke des Ausstroms gezeigt.
1
Introduction
The exchange of water and dissolved constituents between the Baltic Sea and the North Sea is rather complicated. Water exchange is hampered by topographic restrictions the form in of narrow and shallow straits and by hydrodynamic restrictions such as fronts and mixing. Also the spatial and temporal variability is high in a wide spectral range which makes quantification of the exchange from observations rather uncertain. In the present paper, an attempt is made to summarize current knowledge of the dynamics of the coupling of the Baltic Sea and North Sea. The highest priority is not to give a complete review of the many contributions to the subject, but to give the reader a basic understanding of the most important physical processes in the transition zone between the Baltic Sea and North Sea. These processes do
not only produce the observed circulation and variability of the stratification, but are also essential to salt exchange in the Baltic Sea and the variability in volume flow and salinity of the freshwater-influenced water exported to the North Sea. To illustrate these processes some extremely simplified model examples are given. In sections 2 and 3 of this paper, some background on the topographic and hydrographic features of the area is given. In section 4 the coupling between Baltic Sea and Kattegat/Belt Sea is discussed in more detail and an analytical model is presented which shows important dynamic features of the exchange through the Danish Straits. In section 5 different aspects of the flow of low saline water into the North Sea are analyzed, especially the effect of wind forcing. The paper is concluded with a short discussion of important but not yet fully understood processes involved in the interaction between the Baltic and North Seas.
165
Deutsche Hydrographische Zeitschrift - German Journal of Hydrography
2
Topographic features
The transition zone between North Sea and Baltic Sea is divided into four basins, each with a very different topography; the Skagerrak, the Kattegat, the Oresund and the Belt Sea, see maps in Figs. 1 and 2. The Skagerrak, often classified as a part of the North Sea, with a mean depth of 210 m, e. g. SVANSSON [1975], is rather deep compared with the adjacent seas. The deepest part is the Norwegian Trench, located in the northern and central parts. The trench continues, following the Norwegian coast, up to the Norwegian Sea. The maximum depth is 710 m, located near the center of the Skagerrak, and the sill depth is 270 m off Utsira on the Norwegian west coast. On the Danish side, the Skagerrak is rather shallow and its depth is less than 40 m as far as 40 nautical miles off the coastline.
The connection between the Baltic and the Kattegat goes across a shallow sill. The Danish islands separate the sill into three straits; Little Belt, Great Belt, and (Dresund. The deepest sill is the Darss Sill in the Belt Sea with a depth of 18 m and a vertical cross-sectional area of about 300 000 m 2, e. g. JACOBSEN [1980]. The sill in Oresund, the Drogden sill between Copenhagen and MaimS, is only 8 m deep and the vertical cross-sectional area is 60 000 m 2. At the northern end the Oresund is quite narrow with an even smaller cross-sectional area, but considerably deeper, e. g., MATTSSON [1996a].
10 ~
11 ~
12 ~
13 ~
14 ~
]5~ 58 ~ N
57 ~
0~
6~
12=
18 ~
24~
65 ~
56 ~
62 ~
59' 55 ~
56 ~
Fig. 1:
54 =
Map of the Baltic Sea and North Sea. Fig. 2:
An appendix of the Norwegian Trench, the Deep Trench, continues along the Swedish coast into the Kattegat, but the depth decreases rather rapidly. At the border to the Kattegat the depth is about 100 m and the Deep Trench, disappears near the island Anholt. Apart from the Deep Trench, the Kattegat is quite shallow, its mean depth being 23 m.
166
3
Map of the Baltic entrance area.
Descriptive hydrography of the area
3. 1 The Baltic Sea
The Baltic Sea can be regarded as a fjord estuary as the parallel sills in the Danish sounds are very much shallower than both the average depth
Volume 49 (1997) Number 2/3
(54 m) and the maximum depth (460 m). River runoff plus precipitation strongly exceeds evaporation. The excess varies with the seasons, with a pronounced maximum in spring. The river run-off is rather accurately known; the average river discharge to the Baltic Sea in 1950-90 was 14151 m3/s plus an additional 1159 m3/s to the Belt Sea and Kattegat according to BERGSTR(3M AND CARLSSON [1994]. However, there are uncertainties about both precipitation in and evaporation from the Baltic Sea because of which an accurate estimate of net freshwater supply from the atmosphere does not yet exist. It has recently been shown that the variability between years is very high. A calculation for the years 1981-1994 showed that in some years the atmospheric net supply was some 4000 m3/s while in others it was as low as 500 m3/s (OMSTEDT et al. [1997]). The investigation points to a long-term average of some 2000 m3/s, but the large variability makes the average rather uncertain. The closure of the freshwater budget of the Baltic drainage basin is a main objective of the on-going BALTEX-program (ANON.[1995]), and so improved estimates are likely to become available in the near future. Because of the large freshwater supply and the limitation of water exchange by shallow sills and channels with a large frictional resistance in the ()resund and the Belt Sea, the surface salinity is only 7-8 psu in the Baltic Proper and even less in the Bothnian Sea, Bothnian Bay and Gulf of Riga. The halocline of the Baltic Proper is located at some 60-80 m depth and below this the salinity increases to 11-13 psu, see e.g. SAMUELSSON [1996] for a discussion of salinity variations during the last 40 years. The source of salt for the Baltic Sea is inflows of salty water through the Danish Sounds, and since the halocline is much deeper than the sill depth there is no possibility for the Baltic deep water to return directly through the Danish Sounds. Due to the long residence time in deep water in combination with a high rate of deposition of organic material, oxygen is rapidly consumed. Long periods with anoxic conditions are common below 150 m (e. g. FONSELIUS[1969], STIGEBRANDT[1995]). However, at depths above approx. 120 m water exchange is more continuous and not restricted to extreme
events of highly saline inflows as in deeper layers. Therefore, anoxic conditions very seldom reach as high as 120 m (STIGEBRANDTAND WULFF [1987]).
3.2 The Danish Sounds
As stated above, water exchange through the Danish Sounds is very important for hydrographic conditions in the Baltic Sea. Net water flow through the Sounds is equal to the net freshwater supply to the Baltic drainage basin, i.e., about 16.000 m3/s. However, the instantaneous flows across the sills are an order of magnitude greater. The variability of the flows is of crucial importance for the effective salt transport across the sills. A number of authors, probably beginning with KNUDSEN[1899], have tried to quantify the exchange through the Danish Straits. A quite complete review of studies up to 1980 is given in JACOBSEN[1980]. Therefore, focus essentially on more recent approaches. The Belt Sea and Oresund are usually very strongly salt stratified because of the large supply of low saline water from the Baltic Sea and the supply of highly saline waters from Kattegat to the lower layers (see Fig. 3). The tides are rather weak (only about 10 cm amplitude) but still appear to contribute to mixing in the Great Belt (STIGEBRANDT[1983]) because of the rather high speeds of the tidally induced currents (JACOBSEN [1980]). At the shallow sills of Drogden and Darss, turbulence generated by bottom friction is of importance, especially at Drogden where the water column is almost always thoroughly mixed. Out of phase variations of sea levels in the Baltic Sea and Kattegat force large instantaneous flows (of the order of 10~ m3/s) through the Danish Sounds (e. g. JACOBSEN [1980]). The flow will alter direction as the sea level of the Kattegat varies with air pressure and wind set-up. The main flow resistance is due to friction in the narrow and shallow Sounds. It has been shown that the flow is readily calculated from sea level variation if a frictional law is fitted to observations (STIGEBRANDT[1980, 1992], OMSTEDT [1990]), but also effects of earth rotation contribute to the resistance (LASS[1988], MATTSSON
167
Deutsche Hydrographische Zeitschrift - German Journal of Hydrography
[1995]). During severe ice winters, the ice cover increases the frictional flow resistance considerably (OMSTEDT AND NYBERG [1996]). Momentary conditions in the vicinity of the sills are strongly dependent on the direction, magnitude and duration of flow. During outflows from the Baltic the low saline water propagates northward through the Oresund and the Belt Sea, stabilizing the stratification. Even-
KATTEGAT
tually the fronts reach the mouths of the Great Belt and Oresund and enter into the open Kattegat. Since mixing in the Belt Sea is greater than in the C)resund the salinity of water reaching the Kattegat has increased to at least 14 psu for flow through the former compared to about 10 psu for the latter strait
(e. g. PEDERSEN[1993]).
BELT SEA
ARKONA BASIN Darss Si[[
Fig. 3:
Conceptual picture of stratification in the Kattegat, Belt Sea and Arkona Basin showing typical volumes V (in km3) and salinities S of the different layers. From STIGEBRANDT[1995].
3.3 The Kattegat The hydrography of Kattegat is characterized by a very sharp halocline at about 15 m depth dividing the surface mixed layer of salinity 15-25 psu from the deep water of salinity 32-35 psu (see Fig. 3), e.g. SVANSSON [1975]. The deep water layers in the north are filled with Skagerrak water that is successively emptied by wind forced vertical entrainment into the surface layer. The strength of the deep water inflow varies between 28000 m3/s in summer and 64000 mS/s in winter, ANDERSSONAND RYDBERG[1993], because of the strong dependence on wind forcing. The average magnitude of outflow from the surface layer to the Skagerrak is calculated by adding the total freshwater supply (i.e. 17000 m3/s) to the deep water inflow, the result being 45000 m3/s during summer and 81000 m3/s during winter. Outflow mainly occurs in the salinity interval of 23-27 psu (ANDERSSON AND RYDBERG [1993]).
168
Bornholm Channel
The salinity of the southern Kattegat typically is some 18 psu. Outflow from the Great Belt and Oresund initially forms two jets propagating northwards, separated by a small area of higher salinity (PEDERSEN [1993]). A wind shift towards west, usually consistent with flows towards the Baltic, will cause a destruction of the jets by spreading the low saline water across the central Kattegat where it is subsequently mixed by entrainment. However, the buoyancy flux due to the spreading of low saline water across the Kattegat surface represses much of the wind driven turbulence at the main halocline (RASMUSSEN[1995, 1997]). Usually the surface salinity gradually increases in northerly direction in the Kattegat due to wind forced vertical entrainment. However, the gradient is small compared to the large salinity gradient associated with the KattegatSkagerrak front that puts a clear marking line between the seas. The position of the front is somewhat variable, but it usually runs from Skagen to the northeast (e.g. GUSTAFSSON AND STIGEBRANDT
Volume 49 (1997) Number 2/3
[1996]). It is seldom found south of the island of Las(3 (ANDERSSONAND RYDBERG[1993]), but may occasionally extend far into the Skagerrak during northeasterly winds (AURE AND SfETRE [1981]). In general, frontal dynamics force outflow from the surface layer of the Kattegat into the Skagerrak along the Swedish coast.
with a very high flow rate and nutrient content. When reaching the Skagerrak the freshwater is highly diluted and a salinity range of 30-33 psu is typical.
~9~
3.4 The Skagerrak The Skagerrak, like the Kattegat, has a strong saline stratification with a rather shallow pycnocline above Atlantic water of salinity greater than 35 psu that fills the rest of the basin. However, freshwater in the Skagerrak is mainly confined to the near coastal region where baroclinic coastal currents are running in a cyclonic direction (Fig. 4) (e.g. GUSTAFSSON AND STIGEBRANDT[1996]). The magnitude of flow in the coastal currents increases drastically in the cyclonic direction, from some 45000-80000 m3/s leaving the Kattegat to 400000 m3/s leaving the Skagerrak in the Norwegian coastal current (e.g., STIGEBRANDT [1984], RODHE [1989]). This implies strong mixing with the highly saline waters of the lower layers. The mixing occurs mainly in the central region of the Skagerrak where the halocline is only some 10-20 m below the sea surface. A baroclinic flow is also present along the coast of Jutland, carrying low saline water emerging from the southern North Sea where large rivers reduce the salinity. The long-term average freshwater supply from these rivers is some 4500 m3/s and probably a large portion of this freshwater passes through the Skagerrak. However, RYDBERG et al. [1996] estimated a freshwater transport of only 1500 m3/s along the Jutland west coast. According to that investigation, the supply of freshwater-influenced water from the southern North Sea (Fig. 5) appears to be smaller than that found in previous investigations (e. g. North Sea Task Force [1993], JENSEN AND JONSSON [1987]). Even though the flow is lOW, RYDBERGet al. [1996] suggest that the diffusive transport of nutrients from the southern North Sea is substantial, and they also observed events
~8~
57~
S~
Fig. 4:
s
10~
11~
12OE
The time averaged freshwater height distribution (in meters) in the Skagerrak (from GUSTAFSSON AND STIGEBRANDT[1996]).
The measurements presented by RYDBERG et al. [1996] show that the volume transport of low saline water increases eastward along the Jutland coast from Tybor6n, on the west coast, via Hanstholm to Hirtshals (Fig. 5). This clearly shows the importance of recirculation within the Skagerrak. This recirculation was also observed during SKAGEX 1 (DANIELSSEN et al. [1997]) and it appears quite clearly in the investigation of historical hydrographic measurements in GUSTAFSSON AND STIGEBRANDT [1996]. Geostrophic calculations showed that while the mean volume flow of the Norwegian Coastal Current rapidly increases westward along the coast of southern Norway, the mean geostrophic freshwater flow decreases along the same path. Later in this paper, the recirculation of freshwater is explained in terms of Ekman transports across the Skagerrak.
169
Deutsche Hydrographische Zeitschrift - German Journal of Hydrography
There is usually a substantial inflow, on average 4.105 m3/s (RYDBERGet al. [1996]), of highly saline waters (S > 35 psu) along the southern slope of the Norwegian Trench (Fig. 5). The origin of these waters is the northern North Sea and the North Atlantic. Only very recently has a comprehensible theoretical discussion of the dynamics of the barotropic circulation in the Norwegian trench been presented (RODHE [1996]). Earlier investigations (e.g. DOOLEY AND FURNES [1981], FURNES [1983], RODHE [1987]) showed that local winds explain the large variability (of the order of 106 mS/s) of the barotropic circulation in the Trench, but not the average circulation. In RODHE [1996] it is shown that
HIRTSHALS
the average circulation is related to the estuarine circulation associated with the Norwegian Coastal Current. He suggests that the inflow should be forced by mixing of highly saline water into the freshwater-influenced surface layers. ROHDE does not explicitly discuss the forcing of the Norwegian Coastal Current. In GUSTAFSSON AND STIGEBRANDT [1996] it was concluded that the energy gain due to wind driven diapycnic mixing is not sufficient to explain the large increase of potential energy of the Norwegian Coastal Current. Instead, an isopycnal process such as downwelling due to wind forced Ekman transports is proposed as the major driving force.
HANSTHOLM
a
M(S)
an M(S}
1990- 9&.
'71 1990-96.
6
61
5
51
b
ul
xt 2
11
f
01 -1
'
i
29
10
.'%1
i
I0 ~
s (PSU)
31 '
3'2
S(PSU)
t'3
it
TYBOR()N M{S)
TYBORON MIS) 1990-94
1990-9/,
G
w
x
f
0 -1
I
i,
t,
,
,
A
t2 5 (PgU}
Fig. 5:
170
ts
i
J
s ips
The accumulated flow distribution with respect to salinity for sections perpendicular to Hirtshals (a), Hanstholm (b) and Tybor6n (c) and magnified in (d). From RYDBERGet al. [1996].
Volume 49 (1997) Number 2/3
4
Coupling between the Baltic Sea and Kattegat/Belt Sea
The strong stratification connected to the front in the surface layer of the northern Kattegat effectively protects the Kattegat surface waters from direct influence from the Skagerrak. The properties salinity, temperature and any other state-variables north of the front will be those imported to the deep water of the Kattegat. As will be shown below, coupling between the Baltic Sea and the Kattegat/Belt Sea is very tight, and it is not possible to find a section closer to the Baltic Proper where simple decoupling is possible. Thus, the existence of the Kattegat-Skagerrak front is of utter importance to the Baltic Sea system and has to be included in any model of the long-term water exchange in the Baltic Sea. In the following examples the very important role of the Kattegat/Belt Sea on the Baltic Sea will become evident. The general assumption is that the front is maintained by a geostrophic balance as suggested by STIGEBRANBT [1983] although an assumption of a non-rotating hydraulic control gave good results in the model by PEDERSEN AND MOLLER [1981]. JAKOBSEN [1997] has shown that the geostrophic assumption is consistent with observations.
4. 1 Long-term analysis of the coupling between Baltic Sea and Kattegat/Belt Sea The time-scale for changes of the overall salinity in the Baltic Sea is roughly 25 years. STIGEBRANDT [1983] used a time-dependent model to examine the sensitivity of the Baltic Sea salinity to changes in the barotropic flow through the Danish Sounds, freshwater supply to the Baltic Sea and mixing in the Kattegat-Belt Sea. PEDERSEN AND MQLLER [1981] used a steady-state model of the Baltic Sea and the Baltic entrance area to investigate the sensitivity of the stratification to changes in the freshwater supply. In this section the mechanisms of the long-term salt balance of the Baltic Sea are illustrated with a steady-state model that is even simpler than the one mentioned above.
It appears that the outflow of surface water from the Kattegat into the Skagerrak at the Kattegat-Skagerrak front in general is in geostrophic balance, STIGEBRANDT[1983] and JAKOBSEN[1997]. Thus, the outflow can be easily determined. This assumption has been used in time-dependent models by STIGEBRANDT [1983] and OMSTEDT[1987, 1990] and it will also be used here in a steady-state model. In this model we only take into consideration the upper layer of the Kattegat and Belt Sea, which are treated as a single water mass (box), while the lower layer is present as a dynamically passive source of water and salt. In steady-state the volume and salt conservation of the box yields. -QG + QE + QF = 0
(1)
- S Q G + S2Q E + SoQ F + SoQ B - SQB = 0
(2)
The first equation expresses volume conservation of the Kattegat/Belt Sea box, QG being geostrophic outflow to the Skagerrak, QE entrainment flow from the lower layer and QF the net supply from the Baltic Sea (freshwater input). The second equation expresses conservation of salt in the upper layer of salinity S. $2 is the salinity of the lower layer, So the salinity of the Baltic, QB is the effective fluctuating flow across the sills. Conservation of sali in the Baltic Sea is also required, SoQ~ + SoQB - SQB = 0
(3)
Both freshwater supply and fluctuating flow are due to external forcing, for which equation (3) gives a relation between S and So. If So is eliminated from Eq. (2) and volume conservation, Eq. (1), is used to eliminate the entrainment velocity the following expression is obtained, ($2 - S)QG - S2QF = 0
(4)
The geostrophic flow and the entrainment velocity, wE, are calculated from Qo _
g~(S~ - S)h ~ 2s
(5)
171
Deutsche Hydrographische Zeitschrift - German Journal of Hydrography
2m~ WE gfl( S2 - S)h
(6)
where h is the depth to the pycnocline in the Kattegat-Belt Sea, f t h e Coriolis parameter, g the constant of gravity, fl the expansion coefficient of sea water due to salinity, m0 the efficiency of turbulent mixing with respect to work against the buoyancy forces and u, the friction velocity. The hypsographic function of the Kattegat-Belt Sea shows that the horizontal area A(h) decreases approximately linearly with depth and that at about 15.5 m depth the area is 50% of the area at the sea surface. Thus,
and does therefore only vary as a function of wind stress and freshwater supply and not independently of pycnocline depth and salinity. However, the variation of the area of the pycnocline with depth makes the expression for entrainment flow more complex, see Eq. (8). If the flow parametrizations are inserted into the volume conservation, Eq. (1), the following expression for the pycnocline depth h is deduced.
h-
(7)
A(h)=A~ 1 -- ~m
2mo.:AoI h}
gfl-~2Z~h
1
2h m
(10)
The upper layer salinity is given by
Here hm = 15.5 m is the topographical scale height. The entrainment flow is the entrainment velocity times the area at the depth of the pycnocline.
QE =
hm(2mo":Ao + g,flSzhmQF + mou3Ao
(g~&e~+_mou3Ao)
]
(11)
hm(4 2 fgt~S2Q F + 2mou2Ao)]
(8) Finally the Baltic Sea salinity is given by
From Eqs. (4) and (5) we see that the freshwater height in the Kattegat-Belt Sea ( F - $2 -mS h) $2 is solely determined by the freshwater supply and the lower layer salinity, thus
F = ; 2fQF gflS2
(12)
or
rl-
(9)
This is a direct consequence of the assumption of a geostrophic front as pointed out in STIGEBRANDT [1987a]. Note that the entrainment velocity, Eq. (6), is inversely proportional to the freshwater height
172
SQB
So - - QF + Q B
Q~ + Q~ L ~ gpS=
x
(gflS2h~QF + mou3.Ao) ] hm(42 Sgl~g2Q 3 q.-2mou3.Ao)
(13)
Using the parameter values in Table 1, the pycnocline depth becomes 16.9 m, the surface salinity 25.9 psu and the Baltic Sea surface salinity 7.5 psu.
Volume 49 (1997) Number 2/3
The geostrophic outflow from the Kattegat becomes approximately 70.000 m3/s, which is also in close agreement with observations. However, the major drawback of the model is that the effective flow fluctuations, QB, have to be set as low as 6500 m3/s in order to get the right salinity of the Baltic Sea, which is very much smaller than the actual average of the flow fluctuations. From a simulation of the daily barotropic flow with the channel model by STIGEBRANDT [1992] for the period 1970-76, it is found that the average magnitude of the oscillating part of the flow
Table
1
Parameter values used to compute the longterm exchange between the Baltic and the Kattegat/Belt Sea Parameter
Value
f g
1.2 910 -4 s -1 9.8 m s -2
mo /3 u* Ao h,n $2
0.6 8 . 1 0 -4 0.013 m s -1 42000 km 2 15.5 m 33.5 psu 16000 m3s 1
Qv
model layer only through entrainment of deep-water and this is solely wind-driven. However, the fluctuating flow has a prominent influence on the salinity of the Baltic Sea (Fig. 6). The reason for this is that the net salt flux through the system is required to be zero. Thus, a change of'the fluctuating flow will be compensated by the salinity of the Baltic Sea in such a way that the salt flow remains zero. An increase of the freshwater supply has a radical influence on the whole system, decreasing the salinity of the Baltic Sea (Fig. 6) and the pycnocline depth and salinity of the Kattegat-Belt Sea (Fig. 7). Changes in the mixing winds directly influence the stratification in both the Kattegat-Belt Sea and in the Baltic Sea. However, the absolute salinity change in the Baltic Sea is less than in the Kattegat-Belt Sea. These conclusions were also drawn by STIGEBRANDT [1983]. AS mentioned above, the oscillating flow between the Kattegat-Belt Sea and the Baltic Sea is forced by out-of-phase sea level variations. It has been demonstrated that the sea level variability on time-scales longer than five days is forced by largescale zonal wind variations (LASS et al. [1987]). Thus, the large-scale wind climate should, through its influence on QB, have a considerable effect on average salinity in the Baltic Sea. 12
\
was 43000 m3/s. The explanation of the discrepancy is twofold; the lack of a horizontal gradient in salinity makes the salinity of the Belt Sea and Oresund highly overestimated, and the exclusion of the effect of moving fronts hampers salt exchange across the sills. It has been demonstrated that the effective flow is some 10000-14000 m3/s (STIGEBRANDT [ 1 9 8 3 ] ) . However, as long as the horizontal salinity gradient within the Belt Sea and Kattegat does not change dramatically, the response to changes in the magnitude of the oscillating flow should be correct to the lowest order. It is notable that the fluctuating flow does not have any influence on the long-term average stratification in the Kattegat-Belt Sea (Eqs. (10) and (11)). The reason for this is that salt comes into the
\%
%%
10
%% %% ..~ %% ...,."* %% .....'* % .~
~= 8 --=
~176176176
6
[~
4
w]
Change of freshwater supply Change of effective oscillating rio
2 i I -50 -40 Fig. 6:
Change of friction velocity
i
i
i
i
i
I
r
i
-30
-20
-10
0
10
20
30
40
50
Change (%) Model results for the response of Baltic Sea salinity (So) with respect to changes, in wind stress (solid line), freshwater input (dashed line) and effective flow oscillations (dotted line).
173
Deutsche Hydrographische Zeitschrift - German Journal of Hydrography
25
'
'
'
'
. . . . . . . . . . . . .
25
30
,
i
,
L
,
i
i
"-,,
-<
E
i
,
I-- ~
~"" -.,...
~" 20
,
i
,
i
,
E .c 20 v
&
"o
/ / / / / /
Fig. 7:
T , ~ , ~ -20 -10 0 Change
,
~ 10 (%)
,
~ 20
,
~ 30
u3
o15
b
n
~ 40
20 50
10 -50
r , ~ , ~ -40 -30 -20
~ ' ~ -10 0 Change
~ 10 (%)
'
~ 20
~ 30
'
~ 40
20 50
Model results for the response of the pycnocline depth (h) and salinity (S) of the Kattegat-Belt Sea with respect to changes in wind stress (a) and freshwater input (b).
4.2 Impact of water and salt exchange on vertical stratification in the Baltic Sea The simplified model of long-term salt exchange given above does not give any information on the mechanisms maintaining vertical stratification in the Baltic Sea. It only shows that the requirement of zero net salt flux to the Baltic sets limits to the long-term average surface salinity. On timescales less than a few years, the Baltic Sea surface salinity is almost independent of salt flow variations through the Danish Sounds due to the long residence time. The variations are instead due to mixing and variations of freshwater supply over the year, e.g. STIGEBRANDT [1985a] and EILOLA [1997]. Heating-cooling of the surface layer of the Baltic Sea is also of great importance for the near surface stratification. Vertical circulation in the Baltic is maintained by an advection-diffusion balance, where the source of advection is dense gravity currents (e.g. STIGEBRANDT[1987b]). The gravity currents are emerging from the overflow of saline waters across the sills of the Danish Sounds. In order to explain the variability of dense gravity currents in the Baltic Sea, some additional information about the Kattegat/Belt Sea must be included.
174
2 5 >"
O~ ~.'E
a
~ , ~ -40 -30
30
Pycnecline depth Salinity
25 ~>~
10 -50
,
The barotropic flow through the Danish Sounds depends on the sea level difference between Kattegat and Baltic Sea and not on the stratification in the area. However, the correlation between baretropic flow and salinity at the sills is very strong. For the Oresund it has been shown that most of the variance of the surface salinity is due to frontal movements (MATTSSON[1996b]), i. e. when the barotropic flow is directed towards the Kattegat, low saline water from the Baltic will move northward. During southward flow the front will move back, but also some water from below the sill depth will cross the sill. Salinity at the sill will increase with time and strength of the flow as the surface layer is emptied. The water and salt flow across the Drogden sill in Oresund has been measured directly during the past years, see JAKOBSEN AND CASTEJON [1995] and JAKOBSEN
AND
LINTRUP
[1996].
At the Darss sill a two-layer stratification is usually present which makes the dynamics of water exchange across this sill somewhat different from that across the Drogden sill. Thus, there exists a more or less continuous flow of highly saline water from below the halocline of the Belt Sea to the Arkona Basin. Most probably the flow is regulated by geostrophy (STIGEBRANDT[1983], LASS et al. [1987]). Further, it appears that vertical mixing is a much
Volume 49 (1997) Number 2/3
more important factor in determining the stratificaby direct measurement, and it has been subjected tion of the Belt Sea than it is in the Oresund north of to several thorough investigations, HAKANSSONet al. the sill. However, also in the Belt Sea fronts are pro[1993], MATTH,&USet al. [1993], HUBER et al. [1994], minent and the duration and strength of an inflow to JAKOBSEN [1995], MATTHAUSAND LASS [1995] and LILthe Baltic will affect the salinity of the inflowing water JEBLADH AND STIGEBRANDT [1996]. as is easily seen in Fig. 3. If an inflow carries very The effect of a gravity current is twofold; it large amounts of water the stratification of the Kattransports saline water to depths greater than the tegat will become important as the surface layers of sill depth and entrains water from the ambient straOresund and Belt Sea are emptied. tification. The latter process results in significant Conclusively, to model these processes one mixing of deep waters in the Baltic Sea (e. g. KOUTS needs a time-dependent model that resolves the AND OMSTEDT [1993]). The vertical diffusivity in a horizontal movements of fronts and describes the deep basin is probably due to the breaking of inmixing processes correctly. Such models are rather ternal waves (cf. STIGEBRANDT[1987b]) although the complicated and beyond the scope of the present main forcing of internal wave activity in the ocean, paper, but the reader is referred to STIGEBRANDT i.e. tides, is almost absent in the Baltic Sea. Thus [1983], OMSTEDT[1987] and MATTSSON [1996b]. the most probable forcing of internal wave activity is After crossing the sills, highly saline inflowing the wind, but the mechanisms of energy transfer water will sink and form a dense pool at the bottom from wind forcing to internal waves and subseof the Arkona Basin. In STIGEBRANDT[1987a] it was quently to turbulence and diffusion are not yet unargued that the flow from this pool into the Bornholm derstood. channel is determined by geostrophic control in the Arkona Basin. On this assumption it was possible to reconstruct the statistical properties of the upstream 5 The dynamics of the Skagerrak- Exchange condition of the gravity current from salinity obserwith the Central North Sea vations in the Arkona Basin. The hypothesis has been successfully used in models for the Baltic Sea Circulation in the Skagerrak is quite different circulation (e.g. STIGEBRANDT [1987b], OMSTEDT from that in the Kattegat. Time-variable narrow [1987, 1990]). Only very recently has a comprehencoastal currents transport the freshwater influenced sive set of field observations confirmed the hyposurface water, while the stratification of the central thesis of geostrophic control (LILJEBLADH AND STIGE- part is either very weak or characterized by a quite BRANDT [1996]). The statistics presented by thin mixed layer of relatively low salinity. As disSTIGEBRANDT [1987a] show that the dense gravity cussed in Section 3 above, the major source of current into the Baltic is almost always present but freshwater is outflow from the Kattegat, about that events with large inflows are extremely rare as 17000 m3s -1, but Norwegian rivers contribute subis also shown by direct observations of deep water stantially, some 2000 m3s-1 according to SVANSSON flow in the Bornholm Channel presented by WALIN [1975]. RYDBERGet al. [1996] claim that 1500 m3s-~ [1981]. The statistics of rare events have been anaof river discharge to the southern North Sea passes lyzed by MATTH.~.USAND FRANCK [1992] using data through the Skagerrak. from the Darss sill. However, since the highly saline The freshwater-influenced surface waters water that flows across the sills is stored in the leave the Skagerrak with the Norwegian Coastal Arkona Basin for a considerable time, it may be exCurrent which can be traced along the whole Norposed to strong wind mixing entraining it into the wegian coast. This current is highly variable, and it surface layer and thus preventing it from forming is fair to say that the governing dynamics is highly new deep water in the Baltic Proper. dependent on events. The major inflow to the Baltic in 1993 is the first inflow for which it is possible to quantify transports
175
Deutsche Hydrographische Zeitschrift - German Journal of Hydrography
5. 1 Factors controlling the strength of the Norwegian Coastal Current The Skagerrak is extremely sensitive to transversal wind-forced Ekman flows. Strong southwesterly winds establish an Ekman flow from the Norwegian coast towards the Jutland coast. Thus, winds from this direction will decrease or even completely block outflow of the Norwegian Coastal Current and enhance the circulation of the coastal currents along Jutland and Sweden (MOLLER AND SVANSSON [1 978]). This blocking phenomenon is described by AURE AND SAETRE [1981]. When the wind shifts direction the blocking is released and the Norwegian Coastal Current increases in strength, see Fig. 8. A further increase in the flow is possible if a northeasterly wind follows the blocking since this will cause additional transport towards and downwelling along the Norwegian coast. An additional effect increasing the efficiency of the blocking is due to the influence of the Kattegat and Baltic Sea. During southwesterly winds the sea level of the Kattegat is generally rising due to direct influence of the wind and, since westerly winds usually are associated with low pressure systems, by the inverse barometric law. The rising sea level in the Kattegat implies inflow to the Baltic Sea and consequently southward net transport in the Kat2o
3o
4o
5o
6=
T
8o
9o
lo~
tegat. In addition, winds with a westerly component will directly force the surface waters of the Kattegat southward. Northeasterly winds have the opposite effect on the Kattegat and Baltic Sea, with the sea level generally dropping in the Kattegat implying outflow from the Baltic Sea, and Ekman currents within the Kattegat forcing surface water northward. The importance of the Kattegat and Baltic Sea for the storage of low saline water was also pointed out by AURE AND S/ETRE [1981]. In order to illustrate the large potential of a blocking event in inducing strong outbreaks of Skagerrak surface water into the northeastern North Sea, an order of magnitude calculation will be done. Assume that a southwesterly wind starts to act upon an initially undisturbed surface layer, the Ekman flow will soon after the onset of wind stress transport surface water towards the southeast and Jutland Coast at a flow rate of M~w. The accumulated transport of water by the Ekman flow V~ at subsequent times are given by
V~(t) = Mswt
(14)
The salinity of the accumulated Ekman transport would in reality increase as more highly saline water is mixed into the surface layer of the central Skagerrak. However, the effect of increasing salinity
11~
2o
3=
4o
5~
6o
7o
8~
9o
10 ~
11 ~
ao
59 ~
57 ~ .
Fig. 8:
176
Surface temperature charts showing an example of blocking of the Skagerrak outflow (a) and the subsequent release of blocking enabling a large outflow (b). Arrows indicate the direction of the wind. From AURE AND SAETRE[1981 ].
Volume 49 (1997) Number 2/3
on the resulting flow rates in the coastal current is an order of magnitude smaller than the effect of volume increase due to Ekman transport. Therefore, a constant salinity $1 is assumed. The Ekman flow may be calculated from
M.w - T"wLE pf
Jutland and Swedish coasts in the form of a coastal current. Assume that the coastal current takes the shape of a tube with a triangular cross-section. The typical width of a coastal current is an internal Rossby radius (R = c/J), for which the pycnocline depth at the coast is given by
(15)
where r,w is the southwesterly wind stress, L E is the length over which the Ekman current is acting, ,o is density a n d f t h e Coriolis parameter, see Table 2.
Table
~g/f'(SA - S, )L
(16)
were SA is the salinity of underlying Atlantic water. The flow of the coastal current is given by
2
Parameter values used in the calculation of Skagerrak outflow
Parameter
Value
S~ SA TSW LE f p
31 psu 35 pSU 0.5 Pa 150 km 1.2.10 4 s-1 1027 kg/m 3
Equation (14) gives the accumulation in the southeastern Skagerrak for a limited time only; convergence at the coast will generate a cyclonic coastal current that will propagate around Skagerrak. After the time L/c, where L is the perimeter of the eastern Skagerrak and r the phase speed of an internal wave, the nose of the coastal current will reach the Norwegian coast were the current is emptied by the Ekman flow opposite to the coast and a closed circulation will evolve in the eastern Skagerrak. Assuming L = 200 km and r = 1 m/s gives a time scale of 2.3 days. The time scale may become longer if the Kattegat and Baltic Sea are taken into consideration because some or all of the water transported by Ekman currents in the Skagerrak may flow into the Kattegat under certain conditions. Assuming values as in Table 2 the volume V1 becomes 121 km 3 after 2.3 days. At this point the water transported by the Ekman flow is located along the
Q=
g,8(S A - S,)h
2
2f
(17)
and combining these
Q= I2 fgfl(SLz S~)~4
(18)
Using the above results the flow becomes 106 m3/s. If it starts to blow from the northeast the outflow will be even greater as Ekman transport of surface water from the central Skagerrak further increases the volume transport of the coastal current. The time evolution of the potential for geostrophic outflow, i. e. Eq. (18), is shown in Fig. 9. The assumption of narrow coastal currents is the basis for the simple time-dependent numerical model of the Skagerrak surface layer presented by STIGEBRANDT [1984]. The model consists of twelve dynamically interacting boxes and includes the basic processes discussed above, i.e. Ekman transport, baroclinic geostrophic transport and vertical entrainment. The model is coupled with upstream models of the Kattegat/Belt Sea (the model of STIGEBRANDT [1983]) and the Jutland Coastal Current and forced with observed winds and river runoff. The model reproduces observations of stratification in the coastal currents quite well (STIGEBRANDT [1985b]). The results from this model probably give the best available statistics of outflow
177
Deutsche Hydrographische Zeitschrift - German Journal of Hydrography
2.0
250
200
•
1.5
E
%
_~150 0,.)
N
E
E
"6 100 >
0 Q..
1.0
::3
o 0.5
b
50
0
Fig. 9:
i
0
1
I
i
i
2 3 Time (days)
4
5
0.0 0
1
2 3 Time (days)
4
5
Model results for the accumulation of water in the eastern Skagerrak during blocking. The volume increase is shown in (a) and the potential geostrophic flow (b).
from the Skagerrak into the Norwegian Coastal Current. The histogram of the flow magnitude based on a simulation of the years 1980-84 (Fig. 10) is reproduced from a table in STIGEBRANDT [1985b]. The
model simulations also show a clear relation between outflow from the Skagerrak and southwestnortheasterly wind for periods longer than a few days.
4000
\ ~1 "~
'~ 3000
E
r272"2t 1980 ~ 1981 ~ 1982 E~q 1983
/ ~ ~
o
~
1984
2000
O E
0
i
,
,o
,o
0.0-0.2 0.2-0.4 0.4-0.6 0.6-0.8 0.8-1.0 1.0-1.2 1.2-1.4 1.4-1.6 Flow intervals (10 6 m3s -1)
Fig. 10: The annually accumulated outflow from the Skagerrak by the Norwegian Coastal Current for the years 1980-84 divided according to the magnitude of the daily mean flow. Drawn from results, computed with a numerical model, presented in STIGEBRANDT[1985b].
178
Volume 49 (1997) Number 2/3
5.2 Impact on the northern North Sea due to outflow from the Skagerrak
The width of the Norwegian Coastal Current has a pronounced annual cycle with a substantial increase during summer. Outside Bergen on the Norwegian west coast the average width increases from 6 km in February to 50 km in July (SAETREet al. [1988]) while the baroclinic Rossby radius only changes from 4 to 5 km, see Fig. 11. Quite a similar annual cycle has been found within the Skagerrak
where the width of the Norwegian Coastal Current outside Torungen increases from about 10-15 km in winter to some 40-50 km during spring (Fig. 12), while the baroclinic Rossby radius here, too, remains almost constant (GUSTAFSSON AND STIGEBRANDT [1996]). In the latter case, it was found that the widening of the Norwegian Coastal Current was coherent with a rapid and large decrease in potential energy of the current and it was argued that the primary reason is the rapid decrease of southwesterly winds in spring.
61 ~
60
59
58
57
56 3~
2~
1~
0~
1~
2~
3~
4~
5~
6~
7~
8~
9~
10 ~
~3
o
2~
1~
0~
1~
2~
3~
4~
5~
6~
7~
8~
9~
10 ~
Fig. 11 : Surface salinity charts showing the Norwegian Coastal Current before (a) and after (b) the large spreading into the northern North Sea in 1984. Fully drawn arrows indicate winds during the first 5 days and dotted arrows the winds during the last 5 days. From S/:ETREet al. [1988].
Dec
Oec
Nov
Nov
Oct
2o
Oct
Sep
Sep
Aug
Aug
Jut
Jul
Jun
Jun
Hay
Hay
Apr
Apr
Hat
Mar
Feb
Feb
10 ~
20
r
Jan
Jan 20
30
/+0
50
Distance
60
70
80
from Hirtsha[s (kin)
90
100
110
10
20
30
4-0
50
60
70
80
90
100
110
Distance from Hirtshats (kin)
Fig. 12: The annual cycles of (a) freshwater height (m) and (b) potential energy (m3s-2) in a section across Skagerrak from Hirtshals to Torungen. From GUSTAFSSONAND STIGEBRANDT[1996].
179
Deutsche Hydrographische Zeitschrift - German Journal of Hydrography
In the case of the Norwegian Coastal Current knowledge which should be briefly discussed. along the Norwegian west coast, it appears that the An interesting finding is the large mean freshwidening of the current is closely related to the frewater inflow (about 10000 m3s-1) to the Skagerrak quency and strength of winds from the north (S/ETRE from the North Sea found by RYDBERGet al. [1996] et al. [1988]). Northerly winds are quite common duon a section outside Hanstholm. This is approximaring the summer along the Norwegian west coast. tely twice the river discharge into the southern North Thus, the widening of the Norwegian Coastal CurSea and very much larger than the observed freshrent in summer should, therefore, be a direct consewater flow further to the south along the Jutland quence of Ekman transport away from the coast. If west coast (some 1500 m3s-1 according to RYDBERG the northerly wind is strong and persistent enough, et al. [1996]), so that there must be another explalow-saline waters may be spread over a considernation than transport from the southern North Sea able area of the northern North Sea (Fig. 11b) in along the Jutland west coast. A plausible mechaform of a thin surface layer (LJQEN AND SfETRE nism is recirculation of Skagerrak surface water [1985]). It is notable that the year this was found which could be explained by lateral displacement of (1984) was a year with exceptionally large outflows the Norwegian Coastal Current off western Norway from the Skagerrak (see Fig. 10), STIGEBRANDT during northerly winds. In that case the water would [1985b]. be returned to the Skagerrak by Ekman transport or The Norwegian Coastal Current appears to be perhaps by the barotropic currents of the Norwegian rather stable within the Skagerrak, but west of Trench. An indication of the possible importance of Norway frequent and intense eddy formation is ob- the displacement of the Norwegian Coastal Current served (MORK[1981], JOHANNESSENet al. [1989]). into the North Sea to the inflow of freshwater along The eddies may be quite energetic with current the northern Jutland coast may be found in the anspeeds of up to about 2 m/s. The eddies most pronual cycle of the freshwater height outside Hirtshals bably occur in conjunction with large outflows from (see Fig. 12a). GUSTAFSSONAND STIGEBRANDT[1996] the Skagerrak as suggested in the instructive video found that the freshwater height has a pronounced movie of laboratory simulations by MCCLIMANS maximum in July-September which is then consi[1985]. However, the actual generating mechanism stent with the seasonal cycle of the displacement of is a rather complex combination of topographic the Norwegian Coastal Current presented by SAETRE steering, vortex stretching and barotropic instability et al. [1988]. To the present author's knowledge no (JOHANNESSENet al. [1989], IKEDAet al. [1989]). The attempt to correlate these processes has been importance of the eddies as carriers of low-saline made yet. However, during the SKAGEX 1 experiwater from the Norwegian Coastal Current into the ment, Skagerrak surface waters were observed central North Sea is not known. However, from the west of Hanstholm (DANIELSSENet al. [1991]) and it observations it appears that the eddies are mainly seems reasonable to assume that this water sooner propagating northward, and so a substantial contrior later would return the Skagerrak. bution from eddies to the spreading of low-saline The large influence of the dynamics of the Katcoastal water into the central North Sea is not protegat/Belt Sea on the Baltic Sea points to the imporbable. tance of further studies in this region. An example of a study object is the Skagerrak-Kattegat front which only very recently has been the subject of a dedi6 Discussion cated investigation, JAKOBSEN[1997]. Although net flow along the front now is known to be mainly in geIn the present paper, some essential findings ostrophic balance there are still several interesting on the exchange between Baltic Sea and North Sea questions about the dynamics of the strong front, for have been discussed. However, there are some example its north-south migration and interaction major gaps both in this presentation and in our with the usually quite strong eastward currents from
180
Volume 49 (1997) Number 2/3
north of Jutland. Other examples of interesting study objects are the fronts in the southern Kattegat, Belt Sea and Oresund which are of major importance for the exchange of salt and nutrients with the Baltic Sea, e. g. PEDERSEN[1993] and MATTSSON [1996b]. Much effort during the last few years has been put into quantifying transports through the Danish Sounds, with less emphasis on the baroclinic properties of the exchange, an exception being LASS et al. [1987]. Numerical modelling of the area has grown rapidly during the last few years including several investigations using three-dimensional models, e.g. WINKEL-STEINBERG et al. [1991], LEHMANN[1995] and SVENDSEN et al. [1996]. However, so far these modeling efforts have not increased our knowledge on the dynamics of the area. An exception is the investigation by SAYIN AND KRAUSS [1996], who examine some aspects of the water exchange through the Danish straits using a three-dimensional model in a process-oriented manner. A remaining problem of grid models of the area is the enormous computational effort needed to resolve the short lateral scales of baroclinic currents and straits. We can, however, expect more progress in this field as computer capacity increases and new, more physically correct, parametrizations of unresolved processes are developed.
Acknowledgements This work has been financially supported by the Swedish Environmental Protection Agency and the European Commission via MAST contract no. MAS3-CT96-0058.
References
ANDERSSON,L. AND L. RYDBERG,1993: Exchange of water and nutrients between the Skagerrak and the Kattegat. Estuarine, Coastal and Shelf ScL, 36, 159-181. ANONYMOUS,1995: Baltic Sea Experiment (BALTEX), Initial Implementation Plan. International Baltex Secretariat, Pub. No 2.
AURE, J. AND R. SAETRE,1981: Wind effects on the Skagerrak outflow, In: R. S~etre and M. Mork. The Norwegian Coastal Current. University Press, Bergen, Norway, 263-293. BERGSTR(~M,S. ANDB. CARLSSON,1994: River runoff to the Baltic Sea: 1950-1990. Ambio, 23, No. 4-5, 280-287. DANIELSSEN,D. S., L. DAVIDSSON,L. EDLER,E. FOGELQVIST, S. FONSELIUS,L. FOYN, L. HEBNROTH,g. HAKANSSON, I. OLSSONAND E. SVENDSEN,1991: SKAGEX: Some preliminary results. Ices Statuary Meeting, C.M. 1991/C:2. DANIELSSEN, D.S., L. EDLER, S. FONSELIUS,L. HERNROTH, M. OSTROWSKIAND E. SVENDSEN,1997: Oceanographic variability in Skagerrak/northern Kattegat, May-June 1990. ICES J. Mar. ScL, in press. DOOLEY, H. g. AND G. K. FURNES, 1981: Influence of the wind field on the transport of the northern North Sea, In: R. Saetre and M. Mork. The Norwegian Coastal Current. University Press, Bergen, Norway, 57-71. EILOLA, K., 1997: Development of a spring thermocline at temperatures below the temperature of maximum density with application to the Baltic Sea. J. Geophys. Res., 102:C4, 8657-8662. FONSELIUS,S. H., 1969: Hydrography of the Baltic Deep Basins II1. Fishery Board of Sweden. Series Hydrography, Report 23, 97 pp. FURNES,G. K., 1983: A three-dimensional numerical sea model with eddy viscosity varying piecewise linearly in the vertical. Cont. Shelf Res., 2, 231-241. GUSTAFSSON,g. AND A. STIGEBRANDT,1996: Dynamics of the freshwater-influenced surface layers in the Skagerrak. J. Sea Res., 35, 39-53. HUBER, K., E. KLEINE, H-U LASSAND W. MATTHAUS,1994: The major Baltic inflow in January 1993 - Measurements and modelling results. Dt. Hydrogr. Z., 46, 2, 103-114. H,&KANSSON,B., B. BROMANAND H. DAHLIN, 1993: The inflow of water and salt in the sound during the Baltic major inflow event in January 1993. Ices Statuary Meeting, C.M. 1993/C:57. IKEBA, M., J. A. JOHANNESSEN,K. LYGREAND S. SANDVEN, 1989: A process study of mesoscale meanders and eddies in the Norwegian Coastal Current, J. Phys. Oceanogr., 19, 20-35. JACOBSEN, T. S., 1980: The Belt Project: Sea water exchange of the Baltic - measurements and methods. National Agency of Environmental Protection, Denmark, 106 pp. JAKOBSEN,F., 1995: The major inflow to the Baltic Sea during January 1993. J. Mar. Sys., 6, 227-240. JAKOBSEN, F., 1997: Hydrographic investigation of the Northern Kattegat front, Cont. Shelf Res., 17, No. 5, 533-554. JAKOBSEN, F. AND S. CASTEJON, 1995: Calculation of the discharge through Oresund at the Drogden Sill by measurements at two fixed stations. Nordic Hydrology, 26, 237-258.
181
Deutsche Hydrographische Zeitschrift - German Journal of Hydrography
JAKOBSEN,F. AND M. J. LINTRUP, 1996: The exchange of water and salt across the Drogden Sill in Qresund September 1993 - November 1994. Nordic Hydrology, 27, 351-368. JENSEN, T. AND S. JONSSON, 1987: Measurements and analyses of currents along the Danish west coast. Dt. Hydrogr. Z., 40, 193-213. JOHANNESSEN, J. A., E. SVENDSEN, S. SANDVEN, O. M. JOHANNESSEN AND K. LYGRE, 1989: Three-dimensional structure of mesoscale eddies in the Norwegian Coastal Current. J. Phys. Oceanogr., 19, 3-19. KNUDSEN, M., 1899: De hydrografiske forhold i de danske farvande indenfor Skagen i 1894-98.(The hydrographic conditions in the Danish waters inside Skagen in 189498), Komm. For Vidensk. Unders. I de danske farvande, 2:2, 19-79. (In Danish) K6UTS, T. AND A. OMSTEDT,1993: Deep water exchange in the Baltic Proper. Tellus, 45A, 311-324. LAss, H. U., 1988: A theoretical study of the barotropic water exchange between the North Sea and the Baltic and the sea level variations of the Baltic. Beitr. Meereskd., Berlin, 58, 19-33. LASS, H. U., R. SCHWABE, W. MATrH~,US AND E. FRANCKE, 1987: On the dynamics of water exchange between Baltic and North Sea. Beitr. Meereskd., Berlin, 56, 27-49. LEHMANN, A., 1995: A three-dimensional baroclinic eddyresolving model of the Baltic Sea. Tellus, 47A, 1013-1031. LILJEBLADH,B. AND A. STIGEBRANDT,1996: Observations of the deepwater flow into the Baltic Sea. J. Geophys. Res., 101 :C4, 8895-8911. LJ~IEN, a. AND R. SfETRE, 1985: Outflow of Skagerrak water to the North Sea during the summer 1984. Ices Statuary Meeting, C.M. 1985/C:21,6pp. MATTH.&US, W. AND H. FRANCK, 1992: Characteristics of major Baltic inflows - a statistical analysis. Cont. Shelf Res., 12, No. 12, 1375-1400. MATTHAUS, W., H. U. LASSAND R. TIESEL, 1993: The major Baltic inflow in January 1993. Ices Statuary Meeting, C.M. 1993/C:51. MATTH.&US,W., AND H. U. LASS, 1995: The recent salt inflow into the Baltic Sea. J. Phys. Oceanogr., 25,280-286. MATTSSON, J., 1995: Observed linear flow resistance in the Oresund due to rotation. J. Geophys. Res., 100:C10, 20779-20791. MATTSSON, J., 1996a: Some comments on the barotropic flow through the Danish Straits and the division of the flow between the Belt Sea and the Oresund. Tellus, 48A, 456-464. MATTSSON, J., 1996b: Analysis of the exchange of salt between the Baltic and the Kattegat through the Oresund using a three-layer model. J. Geophys. Res., 101:C7, 16571-16584. MCCLIMANS, T. A., 1985: Forecasting ocean currents of the Northern North Sea. Norwegian Hydrotechnic Laboratory, Trondheim, Norway, Video, Proj. no. 603600.
182
MORK, M., 1981 : Circulation phenomena and frontal dynamics of the Norwegian coastal current. Phil. Trans. R. Soc. Lond., A 302, 635-647. MOLLER, P. AND A. SVANSSON, 1978. Investigations in the Northern Kattegat during the International JONSDAP76 period INOUT, March-April 1976.- Meddelande fr&n Havsfiskelaboratoriet - Lysekil, Report no. 243: 20pp. NORTH SEA TASK FORCE, 1993: North Sea. Subregion 8. Assessment Report 1993. State Pollution Control Monitoring. Oslo, 79 pp. OMSTEDT,A., 1987: Water cooling in the entrance of the Baltic Sea. Tellus, 39A, 254-265. OMSTEDT, A., 1990 Modelling the Baltic Sea as thirteen sub-basins with vertical resolution. Tellus, 42A, 286-301. OMSTEDT,A. AND L. NYBERG,1996: Response of Baltic Sea ice to seasonal, interannual forcing and climate change. Tellus, 48A, 644-662. OMSTEDT, A., L. MEULLER AND L. NYBERG, 1997: Interannual, seasonal and regional variations of precipitation and evaporation over the Baltic Sea. Ambio, 26, 8, 484-492. PEDERSEN, F. B. AND J. S. MeLLER, 1981: Diversion of the river Neva, How will it influence the Baltic Sea, the Belts and Cattegat. Nordic Hydrology, 12, 1-20. PEDERSEN,F. B., 1993: Fronts in the Kattegat: The hydrodynamic regulating factor for biology. Estuaries, 16, 1, 104-112. RASMUSSEN, B., 1995: Stratification and wind mixing in the Southern Kattegat. Ophelia, 42, 319-334. RASMUSSEN, B., 1997: The near-surface horizontal buoyancy flux in a highly stratified region, Kattegat. Estuarine, Coastal and Shelf Sci., in press. RODHE, J., 1987: The large-scale circulation in the Skagerrak, interpretation of some observations. Tellus, 39A, 245-253. RODHE, J., 1989: The large-scale mixing and estuarine circulation in the Skagerrak; calculations from observations of salinity and velocity fields. Tellus, 41A, 436-446. RODHE, J., 1996: On the dynamics of the large-scale circulation of the Skagerrak. J. Sea Res., 35, 9-21. RYDBERG, L., J. HAAMER, O. LIUNGMAN, 1996: Fluxes of water and nutrients within and into the Skagerrak. J. Sea Res., 35, 23-38. SAMUELSSON,M., 1996: Interannual salinity variations in the Baltic Sea during the period 1954-1990. Cont. Shelf Res., 16, No. 11, 1463-1477. SAYIN, E. AND W. KRAUSS,1996: A numerical study of the water exchange through the Danish Straits. Tellus, 48A, 324-341. STIGEBRANDT, A., 1980: Barotropic and baroclinic response of a semi-enclosed basin to barotropic forcing from the sea. In: H. J. Freeland, D. M. Farmer and C. D. Levings. Fjord Oceanography, New York: Plenum, 151-164.
Volume 49 (1997) Number 2/3
STIGEBRANDT,A., 1983: A model for the exchange of water and salt between the Baltic and the Skagerrak. J. Phys. Oceanogr., 13, No. 3, 411-427. STIGEBRANDT,A., 1984: A model for the estuarine circulation in the Skagerrak, Troll field engineering studies: Oceanographic support work - Task Report 3, OTTER Report STF88 F84017. Proprietary NHL, Trondheim: 43pp. STrGEBRANDT,A., 1985a: A model for the seasonal pycnocline in rotating systems with application to the Baltic proper. J. Phys. Oceanogr., 15, No. 11, 1392-1404. STIGEBRANDT,A., 1985b: On the outflow of brackish water from the Skagerrak, Troll field engineering studies: Oceanographic support work stage 2 - Task Report 2, OTTER Report STF60 F85002. Proprietary NHL, Trondheim: 50pp. STIGEBRANDT, A., 1987a: Computations of the flow of dense water into the Baltic Sea from hydrographical measurements in the Arkona Basin. Tellus, 39A, 170-177.
STIGEBRANDT,A., 1987b: A model for the vertical circulation of the Baltic deep water. J. Phys. Oceanogr., 17, 1772-1785.
STIGEBRANDT,A., 1992: Bridge-induced flow reduction in sea straits with reference to effects of a planned bridge across Oresund. Ambio, 21, 130-134.
STIGEBRANDT,A., 1995: The large-scale vertical circulation of the Baltic Sea. In: A. Omstedt (editor) First study conference on BALTEX, Conference proceedings, International Baltex secretariat, Pub. 3. STIGEBRANDT,A. AND F. WULFF, 1987: A model for the dynamics of nutrients and oxygen in the Baltic proper. J. Mar. Res., 45, 729-759. SVANSSON, A., 1975: Physical and chemical oceanography of the Skagerrak and the Kattegat, 1. Open sea conditions. Fishery Board of Sweden, Mar. Res. Inst., Report 1: 1-88. SVENDSEN, E., J. BERNTSEN, M. SKOGEN, B. ,~LANDSVIKAND E. MARTINSEN,1996: Model simulation of the Skagerrak circulation and hydrography during Skagex. J. Mar. System, 8, 219-236. S/ETRE, R., J. AURE AND R. LJ~EN, 1988. Wind effects on the lateral extension of the Norwegian Coastal Water. Cont. Sheff Res., 8, 239-253. WALIN, e., 1981: On the deep water flow into the Baltic. Geophysica, 17, 75-95. WlNKEL-STEINBERG, N., J. O. BACKHAUSAND T. POHLMANN, 1991: On the estuarine circulation within the Kattegat. In: Prandle, D. (editor): Dynamics and exchanges in estuaries and the coastal zone. New York: Springer, 231-251.
Submitted: 12. 02. 1997 Accepted: 24. 10. 1997
Address of author: Bo Gustafsson Department of Oceanography, Earth Sciences Center, G6teborg University S-413 81 G6teborg
183