Iceland Basin Research Articles

Effect of the overflows on the circulation in the subpolar North Atlantic: A regional model study

An ocean circulation model for process studies of the Subpolar North Atlantic is developed based on the Geophysical Fluid Dynamics Laboratory (GFDL) Modular Ocean Model (MOM) code. The basic model configuration is identical with that of the high‐resolution model (with a grid size of 1/3° × 2/5°) of the World Ocean Circulation Experiment (WOCE) Community Modeling Effort (CME), except that the domain of integration is confined to the area from 43° to 65°N. Open boundary conditions are used for the inflows and outflows across the northern and southern boundaries. A comparison with the CME model covering the whole North Atlantic (from 15°S to 65°N) shows that the regional model, with inflow conditions at 43°N from a CME solution, is able to reproduce the CME results for the subpolar area. Thus the potential of a regional model lies in its use as an efficient tool for numerical experiments aiming at an identification of the key physical processes that determine the circulation and water mass transformations in the subpolar gyre. This study deals primarily with the representation and role of the overflow waters that enter the domain at the northern boundary. Sensitivity experiments show the effect of closed versus open boundaries, of different hydrographic conditions at inflow points, and of the representation of the narrow Faeroe Bank Channel. The representation of overflow processes in the Denmark Strait is the main controlling mechanism for the net transport of the deep boundary current along the Greenland continental slope and further downstream. Changes in the Faeroe Bank Channel throughflow conditions have a comparatively smaller effect on the deep transport in the western basin but strongly affect the water mass characteristics in the eastern North Atlantic. The deep water transport at Cape Farewell and further downstream is enhanced compared to the combined Denmark Strait and Iceland‐Scotland overflows. This enhancement can be attributed to a barotropic recirculation in the Irminger Basin which is very sensitive to the outflow conditions in the Denmark Strait. The representation of both overflow regions determine the upper layer circulation in the Irminger and Iceland Basins, in particular the path of the North Atlantic Current.

A Comparison of Methods to Determine Mass Transports from Hydrographic Measurements

Abstract The purpose of this paper is to discuss the possibility of determining mass transports in a relatively small ocean region, using a hydrographic dataset and some general physical principles. A new hydrographic dataset of the Iceland Basin is used as an example. The physical principles imply geostrophy for the baroclinic component of the velocity field, whereas the barotropic component is determined in the first instance by assuming a uniform level of no motion. It is attempted to eliminate the subjectivity inherent in this method by using inverse modeling techniques, which describe the physical principles (conservation of tracers) in a mathematical form. In this paper it is shown that some of this subjectivity is misleading, by presenting in detail which choices are to be made and how they influence the results. These choices include weighting parameters, omission of planned measurements, and smoothing parameters. From a simulation study on the Levitus dataset it is concluded that the stability of...

Diets and standing stocks of benthopelagic fishes at two bathymetrically different midoceanic localities in the northeast Atlantic

Deep-sea benthopelagic fishes were sampled by trawl and free-fall longline at two latitudinally and bathymetrically separated stations in the northeastern Atlantic. The catch structure and the contents of the fish stomachs were analysed. The northern, 3000 m deep station in the Iceland Basin exceeded the deeper (4500 m) site in terms of demersal fish biomass and abundance by factors of 5 and 10, respectively. Synaphobranchidae of the species Histiobranchus bathybius and several species of Macrouridae contributed to the samples, the latter dominating with Coryphaenoides (Nematonurus) armatus most abundant in the north and with Coryphaenoides (Chalinura) leptolepis in the south. Diet composition was studied for numerical abundance and wet weight of prey organisms. In the Iceland Basin, where two size classes of C. (N.) armatuscould be analysed separately, an ontogenetic change in diet was noted. Individuals shorter than 30 cm in length fed mainly on benthic and epibenthic Crustacea whereas the larger specimens showed a high proportion of fish and Cephalopoda in their stomachs. At the southern site (BIOTRANS station in the West European Basin) Amphipoda, fish and Cephalopoda were the main food source. Coryphaenoides (Chalinura) leptolepisfed on Malacostraca and fish at the northern site, but on Polychaeta and Cephalopoda at the southern station. Stomachs of H. bathybius contained food only in the Iceland Basin. Their diet consisted mainly of Malacostraca and Cephalopoda. Despite their state of digestion, most Cephalopoda could be identified by means of their beaks, which also allowed a recalculation of their original size. The taxonomic composition and the size structure led to the assumption that most Cephalopoda were epi- or mesopelagic and had been preyed upon near the bottom as carcasses. The possible trophic position of the predominant benthopelagic fishes in the benthic boundary layer is discussed.

Changes in the Western Boundary Undercurrent Outflow since the Last Glacial Maximum, from smectite/illite ratios in deep Labrador Sea sediments

High‐resolution mineralogical studies were performed on late glacial and deglacial sediments from two deep piston cores from the Labrador Sea, located at the inlet (SW Greenland Rise) and outlet (Labrador Rise) of the Western Boundary Undercurrent (WBUC) gyre. At the two sites, smectites transported from the eastern Iceland and Irminger basins by the WBUC are observed. Clay mineral changes are used as proxies for the paleointensity reconstruction of the WBUC. On the Greenland Rise, a clay mineral index (smectite/illite (S/I) ratio) is defined. A S/I ratio of ∼1 characterized the Last Glacial Maximum. It increased after ∼17 ka. and reached a maximum value of 4 during the early Holocene. The mineralogical changes are gradual and do not show any reversal during the Younger Dryas. This pattern, which is confirmed by first‐order estimations of smectite and illite fluxes, suggests gradually increasing sedimentary fluxes and WBUC intensity since the Last Glacial Maximum. A peak in the velocity of the WBUC at ∼9 ka, as recorded by clay assemblages, is consistent with other regional studies based on pollen, foraminifera, or grain‐size measurements. A massive dilution of smectites by illite and chlorite (S/I ≈ 3) occurs at ∼8.5 ka. It corresponds to a period of rapid sediment accumulation and reflects an intensified illite‐rich detrital supply by meltwaters from the southern Greenland Ice Margin. On the Labrador Rise, the smectite content varies between 20 and 60% with no obvious trend through time. The mineralogical composition is strongly influenced by ice‐rafted deposition and by the abundance of fast deposit units (cf. Heinrich layers in the North Atlantic) which contain abundant detrital carbonates spilled‐over from the North‐West Atlantic Mid‐Ocean Channel. In such layers, smectites are present but are diluted by the addition of illites, chlorites, and kaolinites. This provides evidence for a discrete and continuous WBUC supply of fine particles from the Irminger and Iceland Basins as far as the southeastern part of the Labrador Basin. Early deglacial smectite‐rich layers (up to 60%) are also observed at this site. They indicate an increase in the outflow of the WBUC at ∼13.5 ka. (Bölling‐Alleröd), as previously reported from grain size or foraminiferal assemblage studies.

A UV absorbing compound in HPLC pigment chromatograms obtained from Icelandic Basin phytoplankton

A UV absorbing compound was observed in surface waters of the Iceland Basin during the decline of a phytoplankton bloom in June 1989. The compound elutes early during reverse-phase HPLC analyses of phytoplankton chlorophylls and carotenoids and has a broad absorption band from 300 to 470 nm with a n absorption maximum a t 380 nm. The compound (subsequently referred to a s P380) is characterised by similar, but not identical, elution properties and absorption spectrum to scytonemin, an ultraviolet sunscreen pigment not previously found in the phytoplankton. The sinularity of P380 to mycosporine-like amino acids (MAAs) is also discussed. P380 concentrations are highest in surface waters, dec l~ne sharply within the upper eupho t~c zone, and are linearly correlated (r' = 0.68) with the photoprotective carotenoid diadinoxanthin.

Carbon dioxide characteristics of water masses in the northern North Atlantic Ocean

Abstract Accurate measurements of total dissolved carbon dioxide (TCO 2 ) together with other tracers (Θ, S , O 2 , and nutrients) have been carried out during a hydrographic survey of a zonal section between Ireland and Greenland and a number of sections in the Iceland Basin. The resulting data, with > 1300 TCO 2 data points, have allowed the determination of the TCO 2 signature of the water masses in the northern North Atlantic Ocean in terms of seven water types. The structure of TCO 2 and dissolved phosphate and nitrate appears to be well correlated with the apparent oxygen utilization. This indicates that ageing of the water types, by mineralization of organic matter, is more important for the observed TCO 2 structure than differences in the preformed TCO 2 when the water types were formed. The overturning of water in the “oceanic conveyor belt” will support meridional TCO 2 transport. This transport was estimated at 0.16 · 10 −1 mol s −1 (−0.06 Gt a −1 C) from the differences of TCO 2 content between the northward flowing Surface and Sub-Polar Mode water and the southward flowing deep Iceland-Scotland and Denmark Strait Overflow Waters.

Grain-size and SrNd isotopes as tracer of paleo-bottom current strength, Northeast Atlantic Ocean

A crucial question in modern paleoceanographic research is how vigourous was the bottom current in the glacial North Atlantic Ocean. Distribution of sediment grain size can provide such information, if the characteristics of the source functions are known. The interest of this paper is to use the SrNd isotopic composition of lithic particles as a tracer of origin, in order to determine possible changes in the source function, over the last 185 kyr. Grain-size distribution data are derived from a sediment core in the Icelandic basin (Gardar Drift). The distribution pattern reveals a well sorted silt size (modal size 20 μm) in the glacial deposits. Sr and Nd isotopic composition measured on separate size fractions indicates a more constant local origin (Iceland) for the silt fraction than for the fine fraction. These data suggest that the increase in the silt fraction during glacial periods is related primarily to an increase in the bottom current strength, rather than to changes in sediment supply (i.e. ice-rafting, turbidity current). The increase in vigour of the bottom current, recorded during the isotopic stages 6, 4 and the last post glacial periods, could have been initiated by a southward shift of the zone of deep water formation in the northern part of the Icelandic basin.

Surface circulation in the northeast Atlantic as observed with drifters

ARGOS surface drifter data from the northern parts of the northeast Atlantic over the years 1990–1993 have been analysed. The drifters had a drogue at a depth of 15 or 30 m. These data cover well over 10 drifter years. The overall drift appeared to be towards the northeast. Analysis by geographic area and by season revealed regional and temporal variations of both the mean flow and the eddy statistics. In winter the drifter velocities had a tendency towards higher values, probably due to increased wind speeds. Highest eddy kinetic energy was found in the deep Iceland Basin, where an extension of the Sub-Arctic front was observed, while the lowest eddy kinetic energy was observed over the shallower Rockall Plateau. The strongest mean surface velocities were found in the Iceland Basin, just west of the Rockall Plateau. They were connected with the Sub-Arctic front in this region. No evidence was found of westward transport of surface water across the Reykjanes Ridge towards the Irminger Sea. Two drifters were observed to leave the area across the Iceland-Faroe Ridge, flowing eastwards over the northern Faroes slope, whereas two drifters left the area through the Faroe-Shetland Channel. The small banks in the area appeared to generate anti-cyclonic surface circulation on the scale of these banks. Over the larger Rockall Bank no preference for cyclonic circulation was found. The eddy kinetic energy was highest over the deep Iceland Basin where transient eddies were found with scalar velocities well over 20 cm s −1, while the Rockall Plateau seems to be an “eddy desert”. The dispersion due to the temporal variability of the Lagrangian surface velocity could well be modelled with the simple Taylor's theory, with a good fit of the data to the theoretical lines for timescales of 0.25–40 days.

Clay mineral signature of the NW Atlantic Boundary Undercurrent

Surface sediments were sampled along 2 transects, across the Iceland and Irminger basins and in the Labrador Sea. Clay mineral assemblages (deduced from X-ray analyses of the carbonate-free < 2 μm fraction) are largely dominated by smectites (about 60%) in the Iceland and Irminger basins. In the Labrador Sea, smectites are present along the Greenland Slope, but absent or rare in sediments from the Labrador Slope. They may, however, represent up to 50% of the clay fraction at depths between 2800 and 3400 m along both margins of the Labrador Sea, i.e., along the axis of the Western Boundary Undercurrent (WBUC). A detrital supply from the adjacent continents is unlikely. The WBUC is thought to be responsible for erosion and transport of fine particles from the smectite-rich Irminger and Iceland Basins, then for their redeposition in the Labrador Sea. These results suggest that clay minerals can be used as paleocurrent indicators in the Northwest Atlantic.

Palaeosecular variation recorded by lava flows over the past five million years

We present a new global palaeomagnetic database, comprising lava flows and thin intrusive bodies, suitable for studying palaeosecular variation and the time-averaged field. The database is presented in some detail in the appendix and is available on-line from the authors. We review palaeosecular variation models to date, emphasizing the assumptions required and the rather arbitrary construction of many of these models. Preliminary studies of the statistical properties of the new database suggest that existing palaeosecular variation models are inadequate to explain the long-term temporal variations in the field. It is increasingly apparent that data distribution and quality are pivotal in determining the characteristics of the secular variation. The work presented here demonstrates the need for revised models of the time-averaged field structure for both normal and reverse polarities before reliable models for palaeosecular variation can be made.

North Atlantic deep-sea sedimentation of Late Quaternary tephra from the Iceland hotspot

Piston cores recovered from the North Atlantic were used to study the sedimentation of Holocene and Pleistocene volcanic ash in the Irminger and Iceland Basins. Ash Zones 1 (≈ 11,100 yr B.P.), 2 (≈ 55,000 yr B.P.) and 3 (≈ 305,000 yr B.P.) were identified from their major element glass composition. The silicic and alkalic Ash Zones 1 and 2 originate from the Southeastern Volcanic Zone of Iceland, where they correlate with the Sólheimar ignimbrite from Katla volcano and the Thórsmörk ignimbrite from Tindfjallajökull volcano, respectively. The low-alkali composition of silicic Ash Zone 3 indicates a source from one of the silicic centers in the active rift system. Ash Zones 2 and 3 occur in the Irminger Basin as dispersed glass shards over a depth interval of several tens of centimeters. Their compositional and granulometric characteristics reflect an initial fallout on pack-ice north of Iceland, followed by ice-rafting sedimentation in the Denmark Strait, prior to bioturbation. δ 18O stratigraphy of foraminifera in the cores indicates that the ash zones were deposited during a cold interval, at the time when seas north of Iceland were ice-covered. Sedimentary features indicate that turbidity currents were also involved in the dispersal of Ash Zones 1 and 2 south of Iceland. The initiation of these gravity currents from the shelf can be attributed to either glacier bursts (jökulhlaups) carrying tephra, or the entrance of pyroclastic flows into the ocean.

The Flux of Dense Cold Overflow Water Southeast of Iceland

From the period commencing midsummer 1990 and lasting for 13 months, 13 current meters were deployed on seven moorings at depths between 1000 and 2300 m southeast of Iceland (63°09′N, 17°18′W to 61°44′N, 15°24′W). The purpose of the array (designated WOCE ACM8) was to measure the overflow of dense water, downstream of the source in the Faeroe Bank Channel and downstream of any sources on the Iceland–Faeroe Ridge. In the depth range 1300–2100 m three of the moorings showed a bottom intensified current that was persistent and directed to the southwest; at the other moorings the current was erratic and weak. The records were combined to form a transport time series for water denser than σ0>27.8 kg m−3 in the direction 240°. A mean value of 3.2 ± 0.5 Sv was found with considerable variability near a period of seven days, but there was no convincing evidence for a seasonal signal. The amount of pure Norwegian Sea water present in the overflow was estimated at 1.6 ± 0.15 Sv, close to previous determinations and comparable to that measured in the Faeroe Bank Channel. A small contribution is therefore deduced as crossing the Iceland–Faeroe Ridge. A general freshening of the water column in the northern Iceland Basin in the period 1960–1990 is demonstrated.

Hydrographic Variability in the Bottom Layer of the Iceland Basin

Abstract The variability of the lowest 400 dbar in the Deep Iceland Basin has been studied for the years 1988–1991. It appears that hydrographic parameters of the Iceland–Scotland Overflow Water in the lowest 100 dbar mainly vary on longer timescales (interannual). Centered around 140 dbar from the bottom, at a more or less constant density level, salinity, dissolved silica, oxygen, and potential vorticity show a maximum in variability. This is attributed to the isopycnal advection of lenses of water with distinct, coherent hydrographic properties. These lenses are assumed to form a mechanism for lateral isopycnal advection of deep water, enriched with either Lower Deep Water or Subpolar Mode Water. The vertical extent of the lenses is determined by a density window between Labrador Sea Water and Iceland–Scotland Overflow Water, centered around σ2.5 = 39.27 kg m−3. The lateral, isopycnal advection, followed by subsequent diapycnal mixing, is proposed as the mechanism to incorporate Lower Deep Water into t...

Currents and mixing in the Irminger Sea and in the Iceland Basin

On the basis of a hydrographic section from the Faeroe Islands toward Greenland along 62°N and the deployment of about 20 satellite‐tracked buoys, it is shown that the Irminger Current branches at the Reykjanes Ridge. Its total transport amounts to 17–18 Sv. The eastern branch of the Irminger Current (8 Sv) turns east and enters the Norwegian Sea both through the Faeroe‐Shetland Channel (3 Sv) and by passing northeast of the Faeroe Islands. Maximum speeds of 70 cm s−1 are reached at the outer edge of the East and West Greenland Current. Very intensive mixing is observed between the four water masses that flow into the Irminger Basin. Stirring by eddies plays a dominant role for strong interleaving, subsequent salt fingering, and double‐diffusive layering. Conditions for these processes are optimal nearly everywhere in the Irminger Sea.

Early diagenesis of lipid biomarker compounds in North Atlantic sediments

Near‐surface sediments (0–45 mm depth interval) from the Iceland Basin (59°N, 21°W, 3070‐m water depth) and Biscay Abyssal Plain (48°N, 17°W, 4105‐m water depth) were sectioned at millimeter‐scale resolution to assess alterations in key lipid biomarkers during early diagenesis. Inventories (µg cm−2 in the topmost 45 mm) of biomarkers from presumed terrestrial sources (&gt;C20 linear n‐alkanes, n‐alkanoic acids, n‐alkanols) and from phytoplankton sources (phytosterols, long‐chain alkenones and alkenoates) were 1.5–4 times more abundant in the Iceland Basin core than in the Biscay Abyssal Plain core, indicating greater sedimentary inputs and/or preservation at the Iceland Basin site. Biomarker concentrations varied significantly over millimeter‐scale depth intervals. Pronounced subsurface concentration maxima were present at both sites. At the Biscay Abyssal Plain site, steep concentration gradients in the uppermost few mm of the sediment column indicate extensive diagenetic losses at or near the sediment/water interface. Strong covariance was seen in the concentrations of various compounds within each class (e.g., long‐chain n‐alkanes, n‐alkanols, n‐alkanoic acids and long‐chain alkenones and alkyl alkenoates), indicating that individual compounds within each class are associated with the same organic matrix and have similar degradation and mixing rates. Thus biomarker indices for assessment of land‐derived inputs (CPI, C29/C31, ACL) and sea surface temperature (IPT) remain largely unaltered by early diagenetic processes in oxic, abyssal sediments in spite of extensive bioturbation and degradative losses within the uppermost few centimeters of the sediment column.

Origin and fate of organic biomarker compounds in the water column and sediments of the eastern North Atlantic

This paper focuses upon lipid biomarkers as tracers of the biological carbon cycle and our efforts to derive and valid ate ‘molecular tools’ for oceanography and palaeoceanography as part of the 1989—1991 UK -JGOFS Biogeochemical Ocean Flux Study (BOFS). Biomarker concentrations and composition in water column particulates and bottom sediments in the North Atlantic show a strong correspondence to seasonal and interannual patterns of productivity. Biomarkers document the rapidity with which vertical flux processes operate in the high latitude North Atlantic: for example, the massive sedimentation of phytodetritus following a coccolithophorid bloom in the Iceland Basin and its subsequent resuspension, and the benthic biological response to this pulse of biologically available carbon. Sedimentary biomarker distributions indicate that organic material decomposition in sediments is dominated by processes at or near the sediment water interface and that downmixing of labile material into the sediments is largely controlled by advective rather than diffusive-like processes.

Magnetic susceptibility variations in Upper Pleistocene deep‐sea sediments of the NE Atlantic: Implications for ice rafting and paleocirculation at the Last Glacial Maximum

Magnetic susceptibility (MS) variations are used to intercorrelate 17 Upper Pleistocene sediment cores taken from the NE Atlantic, between 40° and 60°N. The MS‐based correlation depends on regionally consistent patterns of variation in the deposition of ice‐rafted detritus (IRD) in response to Pleistocene glaciations, and especially to high‐frequency ice‐rafting episodes referred to in recent studies as “Heinrich events.” The sedimentological and rock‐magnetic basis for the apparent relationship between the MS signal and IRD content of NE Atlantic sediments is examined by (1) comparing the MS profiles of selected cores with their records of coarse fraction (&gt; 150 µm) lithic fragment abundance and Neogloboquadrina pachyderma (sin) percentages, and (2) normalizing MS by expressing it both on a carbonate‐free basis, and as a quotient with anhysteretic remanent magnetization (a parameter sensitive to magnetic mineral grain size variations). These comparisons show that variations in bulk‐sediment MS are only partly driven by simple carbonate dilution (± productivity and dissolution) effects. Changes in both the concentration and grain size of magnetic minerals within the lithogenic noncarbonate fraction also impose a significant influence on bulk MS values. In particular, horizons rich in IRD are associated with significant increases in the relative proportion of coarse grained (multidomain) ferrimagnetic particles in the sediment. This is because ice‐rafting, in contrast to most other mechanisms capable of transporting detrital magnetic minerals to pelagic environments, has a high potential for delivering large ferrimagnetic grains as components of sand‐sized, polycrystalline lithic fragments. This fundamental linkage between the IRD content and MS signal of NE Atlantic sediments is used to reconstruct the patterns of variation in IRD deposition and, by inference, surface currents of the last glacial maximum (LGM, ∼18–19 ka) relative to the present‐day NE Atlantic, using the time‐slice mapping approach developed by the CLIMAP project group. Our LGM/Holocene MS ratio map, based on sample pairs from over 80 deep‐sea cores, confirms that there was a weak, cyclonic gyre north of the polar front in the LGM North Atlantic. The gyre comprised a sluggish warm current in the NE Atlantic flowing north between latitudes 47° and 62°N, partly fed by subtropical waters from south of the polar front, and carrying large numbers of icebergs derived from several sources, most of which melted between Latitudes 45° and 52°N. The warm current probably continued its flow into the Iceland Basin, where it fed into a south‐flowing current which transported melting icebergs from Iceland and Scandinavia along the western flank of the Reykjanes Ridge.

Labrador Sea Water in the Eastern North Atlantic. Part I: A Synoptic Circulation Inferred from a Minimum in Potential Vorticity

Abstract A synoptic distribution of Labrador Sea Water (LSW) in the eastern North Atlantic is determined from a regularly sampled, but sparse (3° resolution), survey covering 39° to 54°N, 11° to 34°W during spring 1991. The core of LSW can be defined by a minimum in potential vorticity (PV). Using property values at this minimum the authors infer the circulation of LSW. In addition, using a known source function for salinities at the core of LSW, estimates are able to be made of LSW vintages. The authors then compared the synoptic circulation to historical data. Youngest, 1986 vintage, LSW crosses the Mid-Atlantic Ridge to the eastern basin between 48° and 51°N at 34°W. This water then flows north to the Iceland Basin and eastward to the Rockall Trough, where it was found to be of 1978 vintage. Tongues of low salinity, low temperature, and high oxygen extend southward on the eastern side of the Mid-Atlantic Ridge, indicating that LSW also flows southward in the eastern basin. At the southern edge of the s...

Mean currents and current variability in the iceland basin

Long-term (>6 months) current measurements from five moorings in the Iceland Basin have been analysed for the mean currents and the structure of the variable current components. The time-averaged flow at all five moorings had a strong baroclinic character. The mean circulation in the upper layers with relatively warm Sub-Polar Mode Water appears to have a general north-eastward direction with maximum mean velocities of 6 to 7 cm·s −1. In the bottom layer south of Iceland, where the cold Iceland-Scotland Overflow Water flows westwards along the topography in a Deep Northern Boundary Current, mean velocities of the order of 10 to 20 cm·s −1 have been observed. Over the deep slope of the Hatton Bank, water enters the Iceland Basin in a branch of the Deep Northern Boundary Current which has a cyclonic rotation sense in the Iceland Basin. The variable part of the current has been analysed by means of principal-component analysis. The current variations in the central Iceland Basin appear to have a mainly barotropic character while variations in the baroclinic flow of Iceland-Scotland Overflow Water contributed 10% or less to the total energy of the variable deep flow. Over the slope of the Hatton Bank the variable currents had a mainly baroclinic character with shear in both current speed and direction. Comparison of the geostrophic velocity with the mean Eulerian velocity has revealed that the σ θ =27.725 kg·m −3 surface can be used adequately as level of non-motion for the geostrophic modelling of the flow along the Iceland and Hatton slopes. The mean westward geostrophic transport of ISOW south of Iceland relative to this reference surface amounted to 3.5 Sv, in agreement with existing independent estimates.

A study of objective methods for water mass analysis, applied to the Iceland Basin

Cluster analysis and principal component analysis have been applied to multi-parameter hydrographic data from the Iceland Basin. The results indicate that both objective methods form a useful addition to the classic water mass analysis methods. Only 3 clusters, representing surface water, the permanent thermocline and the deep water complex, appear to explain about 84% of the observed variance of the hydrographic parameters. The first two empirical orthogonal functions, representing similar contrasts between surface water, thermocline water, and deep water, are able to explain over 95% of the variance. The main water mass structure can be described with a 4 dimensional sub-space of the 5 dimensional parameter, space in agreement with the assumption that only 5 independent water types are responsible for this structure.