Long-term hydrographic variation in the Skagerrak based on the section Torungen Hirtshals

ICES Journal of ine Science, 3: 1 2. 1 Long-term hydrographic variation in the Skagerrak based on the section Torungen Hirtshals Didrik S. Danielssen, Einar Svendsen, and ek Ostrowski Danielssen, D. S., Svendsen, E., and Ostrowski, M. 1. Long-term hydrographic variation in the Skagerrak based on the section Torungen Hirtshals. ICES Journal of ine Science, 3: 1 2. Measurements of temperature and salinity taken about once a month across the Skagerrak between Norway and Denmark have been analysed for long-term variations in the seasonal cycle in the surface water masses from 12 onwards. Since most of the southern and central North Sea water passes through the Skagerrak before leaving the North Sea along the Norwegian coast, the results are expected to indicate possible variations in the hydrographical conditions of the North Sea. There is a persistent seasonal pattern in the upper layer across the section. Higher salinities are observed at the Danish shelf break during winter which is attributed to the intensification of direct inflow from the North Sea. During summer, salinity on the Danish side is always lower than during winter. However, the opposite situation exists in the centre of the Skagerrak with lower salinities in winter than during the summer. A long-term anomaly beginning in the late 1s with a decrease in both temperature and salinity was coupled to the Great Salinity Anomaly in the North Atlantic. A second anomaly started in the late 1s with an increase in both temperature and salinity. 1 International Council for the Exploration of the Sea Key words: circulation, North Sea, salinity, Skagerrak, temperature, time series. D. S. Danielssen: Flødevigen ine Research Station, IMR, N-1 His, Norway. E. Svendsen and M. Ostrowski: Institute of ine Research (IMR), PO Box 1 Nordnes, N-2 Bergen, Norway. Introduction The hydrographical conditions across the Skagerrak along the section Torungen Hirtshals have been measured at about montly intervals since 11, but regular observations have only been carried out since the 1s. Since about % of the water entering the North Sea is assumed to pass through the Skagerrak before it leaves the North Sea, many of the hydrographic events taking place in the North Sea will be reflected in this area. A review of the circulation pattern and water masses in the area is given by Svansson (1), Danielssen et al. (11), and Anon. (13). This paper summarizes information on the temporal variability of the seasonal cycle for the period 12 to 1 based on the spatial distribution of temperature and salinity statistics. Materials and methods Figure 1 shows the hydrographic stations along the section from Torungen to Hirtshals. The innermost station (Stn 2) on the Danish coast has only been sampled since ch 1 and has therefore been omitted from the present analysis. Means and standard deviations of temperature and salinity as well as Spearman rank-order correlation coefficients between temperature and salinity have been calculated for the ch and ust data separately, based on algorithms published by Press et al. (12). Subsets of data forming time series covering the years 12 1 for each position across the section and at standard ICES depth levels were extracted from the hydrographical database. Statistical calculations were applied to each time series separately. The tools from the SKAGEX Atlas as described by Ostrowski (1) were used for presentation of transect data. In addition, modelled volume transports into the North Sea between western Norway and the Orkney Islands using the NORWegian ECOlogical Models system (NORWECOM) (Skogen, 13; Skogen et al., 1; Svendsen et al., 1) are included. Results and discussion From studies of depth versus time plots for individual stations, data from 2 m and 3 m depth were selected as being representative of the long-term hydrographic variability of the deep Atlantic water masses from the northern North Sea and of the upper water layers, 1 313//1+ $1./ 1 International Council for the Exploration of the Sea

1 D. S. Danielssen et al. N Norway Kragerø Sweden 3' Kristiansand Arendal Z 21 Z 2 Z 21 Z 21 Z 22 3' Z 23 Z 23 Z 21 Z 2 Z 2 Z 2 Hirtshals Skagen Læsø Denmark 3' 3' 3' 1 3' 11 3' ' E Figure 1. Map of the stations on the hydrographical section Torungen Hirtshals in the Skagerrak. The last two digits of the station numbers indicate the distance in nautical miles from the Norwegian coast. respectively. Near-surface data were avoided because of enhanced short-term variability on time scales of a few days and associated problems of under-sampling. Long-term (12 1) variability of the seasonal cycle of temperature (Fig. 2) at 3 m depth in the middle of the Skagerrak (Stn 22) shows that maximum temperatures are reached in ober ember. This is in contrast to what is happening at the surface (not shown), where the maximum occurs in y ust. Minimum temperatures are reached in ch il, approximately one month later than at the surface. Large year-to-year variations as well as long-term (multi-year) anomalies have occurred during the period. Relatively warm years include 13 and the period 1 13. At 2 m depth, the seasonal cycle shows only a weak signal, temperature being slightly lower during summer ( C) than during winter ( C). However, the strong positive multi-year anomaly during 1 13 was pronounced throughout the entire year, temperatures being roughly 1 1. C higher than normal. Long-term variability in the salinity (Fig. 3) at 3 m depth at the same station also shows a seasonal cycle with highest values during the autumn. During the first part of the year (uary ), salinities are more variable between years. There are several periods with less than 3 (e.g. 1 11), while the period 1 13 is characterized by unusually high salinities throughout the entire year. At 2 m depth, no clear seasonal cycle can be seen. However, a negative salinity anomaly appears to be present during most of the period 1 11, and also a clear positive anomaly is observed in the years 1 13. The relatively cold and low salinity period of 1 11 appears to be associated with the large-scale Great Salinity Anomaly in the North Atlantic (Dickson et al., 1; Svendsen and Magnusson, 12; Svendsen et al., 1). These multi-year anomalies are also present at most of the other stations along the Torungen Hirtshals section. In particular, the warm and saline period 1 13 was observed throughout the Skagerrak and also in the

Hydrographic variation in the Skagerrak 1 1 11 11 1 1 11 11 1 1 11 3 1 3 2 3 1 1 1 1 3 1 1 1 1 1 1 1 Year Figure 2. Temperature at station 22 at depths of 3 m (top) and 2 m (bottom), 12 1. North Sea (Svendsen et al., 1). This positive anomaly can be directly coupled to large inflows of Atlantic Water to the North Sea (and also to the Skagerrak), especially during the winters of 1 1. The seasonal variability in the period 1 12 of the monthly mean volume transports southward through the

2 D. S. Danielssen et al. 33. 3. 33. 32. >3. 1 33. 3. 33. 3. 33. 3. 3. 3. 33. 3. 3. 33. >3. 33. 1 1 1 1 3. 33. 33. 32. 3. 3. 33. 1 3. 3. > > < 1 1 1 1 1 1 Year Figure 3. Salinity at station 22 at depths of 3 m (top) and 2 m (bottom), 12 1. Orkney Shetland Feie section and the Orkney Utsira section as estimated by means of NORWECOM (Fig. ) shows that relatively large transports occurred in uary ruary (and partly in ch) in 1 1, but also transports in 11 and 12 were enhanced. The very mild winters of these years were associated with

Hydrographic variation in the Skagerrak 21 Month 2. 2.2 1. 2.2 1. Orkneys-Shetland-Feie 1. 2.2 2. 2.2 2. 1. 2. 1. 1. 1. 1. Orkneys-Utsira 2. 1. 1. 1. 1. 2.2 1. 1. 1 2. 2.2 3.1 3. 2. 3. 2. 1 1 1 11 12 Year 1 1 3.1 1. 2. 2.2 1 1 11 Figure. Inter-annual variability (1 12) of the seasonal cycle of the monthly mean volume transports (1 m 3 s 1 ) through the Orkney Shetland Feie (left) and the Orkney Utsira (right) sections in the North Sea estimated with the NORWegian ECOlogical Model system (NORWECOM). 12 strong south-westerly winds, increased cloudiness, and increased inflow of relatively warm Atlantic Water (Svendsen and Magnusson, 12), resulting in an extremely warm ocean climate. According to Norwegian coastal sea surface temperature data, the winter of 1 was probably the warmest since measurements started in 1 (Anon., 13). According to the model, transport through the English Channel was also enhanced during the winter of 1 1 (Svendsen, unpubl. data). The largest transports into both the northern and the southern North Sea took place in 1, resulting in the highest salinities ever recorded in both inflow regions (Heath et al., 11; Ellett and Turrell, 12). The mean temperature (Fig. ) and salinity (Fig. ) distributions along the Torungen Hirtshals section during the period 12 1 for ch and ust indicate a thicker layer of low salinity surface water on the Norwegian side. This is mostly due to the influence of Baltic Water, which is the predominant source of fresh water to the Skagerrak (Anon., 13). In ust, the water in the middle of the trench at depths of about 1 2 m is less saline (Fig. ) and has a lower standard deviation (Fig. ) than along the Danish and Norwegian shelf, resulting in the feature of twin peaks. This feature has earlier been described by Ljøen and Svansson (12) in il ; they concluded that it resembles a cyclonic vortex with the lower and denser layer rotating more rapidly. However, recent numerical and physical model simulations suggest that periods with anticyclonic circulation might give a similar twin peaks picture. The large, slightly warmer, sub-surface volume in the middle of the trench during winter indicates that this water mass has a much longer residence time than the water masses along the slopes. The coldest deep water is found during the summer. The distribution of warmer (. C), high salinity ( 3 psu) water along both slopes at approximately 1 2 m depth during summer indicates that this water is of Atlantic origin. The higher temperature is due to enhanced downward mixing of warm surface water along the coasts. The convex shape of the long-term mean isotherms at depths between and 2 m, especially during summer, represents a dominant feature, with a sharp and persistent thermocline between 3 and m in the middle of the Skagerrak (Figs and ). This feature has been observed by several authors and is generally referred to as the dome or ridge (Pingree et al., 12; Richardson, 1; Danielssen et al., 11). The long-term mean isohalines reveal a rather different picture (Fig. ). The isohalines of 3 and are higher up in the water column in ust than in ch and standard deviation

22 D. S. Danielssen et al. 1 1 2 3.. 3... 1 1 2 3 1. 1. 12. 1...... 1 2 2...2 1 2 2..2 3 3. 3 3. 22 2 21 23 23 22 21 21 2 21 22 2 21 23 23 22 21 21 2 21 Figure. Mean temperature (12 1) along the Torungen Hirtshals section in ch (left) and ust (right). 1 1 1 2 3 33. 32. 31. 3. 1 1 1 2 3 33. 32. 31. 33. 3. 2 2 2 3 3 2 3 3 22 2 21 23 23 22 21 21 2 21 22 2 21 23 23 22 21 21 2 21 Figure. Mean salinity (12 1) along the Torungen Hirtshals section in ch (left) and ust (right). is lower (Fig. ), indicating that there is consistently more water of Atlantic origin during summer than during winter. In contrast, the water mass of central and northern North Sea origin is smaller during summer. This observation is in agreement with the conclusion by Ljøen (11a) that the Atlantic water mass reached its

Hydrographic variation in the Skagerrak 23 1 1 1 2 3 1. 1. 1.2 1. 1. 1.2. 1. 1 1 1 2 3 1.2 1. 2. 2.. 1.. 1. 2. 2 2. 2 2. 3 3.3 3 3. 22 2 21 23 23 22 21 21 2 21 22 2 21 23 23 22 21 21 2 21 Figure. Standard deviation of the temperature (12 1) along the Torungen Hirtshals section in ch (left) and ust (right). 1 1 2 3 2..2. 1..1.2 2. 1 1 2 3 1. 1...1.2.1 2. 1 2..1 1 2.. 2 2 3 3 3 3. 22 2 21 23 23 22 21 21 2 21 22 2 21 23 23 22 21 21 2 21 Figure. Standard deviation of the salinity (12 1) along the Torungen Hirtshals section in ch (left) and ust (right).

2 D. S. Danielssen et al. 1 1 2 2 3 3 1 2 3.1.1.2.3.1.1.2..3...3.2 1 1 2 2 3 3 1 2 3.1.2.1.3.2.2.3.2.3.....1.. 22 2 21 23 23 22 21 21 2 21 22 2 21 23 23 22 21 21 2 21 Figure. Spatial distribution of the Spearman rank correlation between temperature and salinity (12 1) along the Torungen Hirtshals section in ch (left) and ust (right). maximum extension in the summer. Low salinity waters (with slightly lower standard deviations) also seem to spread further out at the surface from the Norwegian coast into the Skagerrak during summer than in winter, as stated earlier by Sætre et al. (1). The variability in surface salinity increases towards the Norwegian coast in relation to decreasing salinities (Fig. ). This is assumed to reflect the influence of varying amounts of Baltic Water. The maxima in subsurface temperature variability (Fig. ) during summer on both the Danish and the Norwegian slope at about m depth are probably caused by enhanced variation in the vertical position of the thermocline in these areas under the influence of local weather conditions. Temperature variability reaches minima in the deepest part of the trench in both ch and ust, and at intermediate depths on the slopes during summer. This may be due to bottom friction. The spatial distribution of Spearman rank correlation coefficients between the salinity and temperature time series (Fig. ) shows values in excess of. at depths exceeding 2 m during summer, indicating the existence of a weak relationship. Elsewhere, the rather low correlations indicate that the monthly temperature and salinity time series are uncorrelated over the entire -year period. Nevertheless, Figure does reveal a clear pattern in the spatial distribution of this parameter above 1 m, which does not change significantly between summer and winter. During summer, the improvement in the correlation towards the deeper parts may be related to the exchange processes of the deep water in the Skagerrak basin. During some winters, the northern North Sea plateau water is cooled and salinities may be slightly reduced due to mixing with surface water, as described by Ljøen and Svansson (12) and Ljøen (11b). This heavy water flows as a plume into the Skagerrak with a time-lag of a few months, reaching the Torungen Hirtshals section in spring/summer. In other years, the Atlantic Water has a density which is high enough to directly replace the deep water. Conclusions The years 1 13 were characterized by an exceptionally warm multi-year ocean climate event associated with a large inflow of Atlantic Water during the winter months. Clear long-term variations with a time scale in the order of a decade have occurred in the deeper parts of the Skagerrak. Lowest temperatures and salinities were observed during the Great Salinity Anomaly of 1 11 and highest temperatures and salinities during 1 13. The data confirm a persistent dome or ridge shape of the temperature profile during summer.

Hydrographic variation in the Skagerrak 2 The intermediate and deep water below the seasonal pycnocline is generally characterized by higher temperatures and lower salinities during winter than during summer. Central and northern North Sea Water contribute less, and Atlantic Water more, to the volume of water in the Skagerrak during summer than during winter. References Anon. 13. North Sea Subregion. Assessment Report 13. North Sea Task Force. State Pollution Control Authority (SFT), Oslo, Norway. pp. Danielssen, D. S., Davidsson, L., Edler, L., Fogelquist, E., Fonselius, S. H., Føyn, L., Hernroth, L., Håkansson, B., Olsson, I., and Svendsen, E. 11. SKAGEX: Some preliminary results. ICES CM 11/C: 2. 1 pp. Dickson, R. R., Meincke, J., Malmberg, S. A., and Lee, A. 1. The Great Salinity Anomaly in the Northern North Atlantic 1 12. Progressive Oceanography, 2: 13 11. Ellett, D. J., and Turrell, W. R. 12. Increased salinity levels in the NE Atlantic. ICES CM 12/C: 2. Heath, M. R., Henderson, E. W., Slesser, C., and Woodward, E. M. S. 11. High salinity in the North Sea. Nature, 32: No. 331, p. 11. Ljøen, R. 11a. Seasonal variations in the flow of different water masses to the Skagerrak. In The Norwegian Coastal Current, Vol. 1, 3 3. Ed. by R. Sætre and M. Mork. University of Bergen. Ljøen, R. 11b. On the exchange of deep waters in the Skagerrak basin. In The Norwegian Coastal Current, Vol. 1, 3 3. Ed. by R. Sætre and M. Mork. University of Bergen. Ljøen, R., and Svansson, A. 12. Long-term variations of sub-surface temperatures in the Skagerrak. Deep-Sea Research, 1: 2 2. Pingree, R., Holligan, P., dell, G., and Harris, R. 12. Vertical distribution of plankton in the Skagerrak in relation to doming of the seasonal thermocline. Continental Shelf Research, 1: 2 21. Press, W., Teukolsky, S., Vetterling, W., and Flannery, B. 12. Numerical Recipes in C. Cambridge University Press, Cambridge. pp. Ostrowski, M. 1. The Skagex Atlas, Thema Nord,, Part II: 33. Richardson, K. 1. Phytoplankton distribution activity in the Skagerrak: a review. ICES CM 1/L: 2. 1 pp. Skogen, M. D. 13. A user s guide to NORWECOM, the NORWegian ECOlogical Model system. Institute of ine Research, Division of ine Environment, Technical Report No., Bergen, Norway. 23 pp. Skogen, M. D., Svendsen, E., Berntsen, J., Aksnes, D., and Ulvestad, K. B. 1. Modelling the primary production in the North Sea using a coupled three-dimensional physicalchemical-biological ocean model. Estuarine, Coastal and Shelf Science, 1:. Svansson, A. 1. Physical and chemical oceanography of the Skagerrak and the Kattegat. I. Open sea conditions. Fishery Board of Sweden, Institute of ine Research, Report no. 1: 1. Svendsen, E., Aglen, A., Iversen, S. A., Skagen, D. W., and Smedstad, O. 1. Influence of climate on recruitment and migration of fish stocks in the North Sea. In Climate change and northern fish populations, pp. 1 3. Ed. by R. J. Beamish. Canadian Special Publication of Fisheries and Aquatic Sciences, 121. Svendsen, E., Berntsen, J., Skogen, M., Ådlandsvik, B., and tinsen, E. In press. Model simulation of the Skagerrak circulation and hydrography during Skagex. Journal of ine Systems, 1. Svendsen, E., and Magnusson, A. K. 12. Climate variability in the North Sea. ICES ine Science Symposia, 1: 1 1. Sætre, R., Aure, J., and Ljøen, R. 1. Wind effects on the lateral extension of the Norwegian Coastal Water. Continental Shelf Research, : 23 23.