Late Weichselian depositional processes, £uxes, and sediment volumes on the margins of the Norwegian Sea (62^75‡N)

Late Weichselian depositional processes, £uxes, and sedimentvolumes on the margins of the Norwegian Sea (62^75‡N)

J. Taylor �, J.A. Dowdeswell 1, M.J. SiegertBristol Glaciology Centre, School of Geographical Sciences, University of Bristol, Bristol BS8 1SS, UK

Accepted 25 February 2002

Abstract

Quaternary continental margin sedimentation in the Norwegian^Greenland Sea has been a response to acombination of tectonic, oceanographic, and glacial activity. Using geophysical survey data (6.5 kHz side-scan sonarand 3.5 kHz penetration echo sounding), in association with bathymetric datasets, the magnitude and frequency of themain forms of margin sedimentation into the Lofoten and Norwegian Basins of the Norwegian Sea during the LateWeichselian and Holocene are estimated. By comparing these geophysically determined estimates of sediment fluxeswith numerically modelled ice-sheet sediment delivery, the role of ice-sheets in shaping the processes acting on thismargin is assessed. Sediment flux at the shelf edge during the Late Quaternary is dominated by ice-sheet delivery ofsediments, focused at the mouths of fast-flowing ice streams formed in bathymetric troughs. Interglacials are, bycontrast, characterised by comparatively little sediment accumulation. Submarine fans, at the margins of cross-shelftroughs, are major depocentres on this glacier-influenced margin. However, occasional large-scale failures of thecontinental slope account for approximately 75% of the basin sediment accumulation during the past 30 000 yr, byeroding substantial quantities of pre-Quaternary deposits. Trough mouth fan and a combination of hemipelagic andglacimarine processes are responsible for the accumulation of the remaining 15% and 7% of basin sediments,respectively. ; 2002 Elsevier Science B.V. All rights reserved.

Keywords: Quaternary; glaciation; Norwegian Sea; ice-sheet numerical modelling; sediment budgets

1. Introduction

Glaciation is acknowledged as a fundamentalcontrol on the morphology of, and processes act-ing on, high-latitude continental margins. How-

ever, the issue of the relative importance of rese-dimentation processes, for example slope failuresand channel systems, in sediment delivery to theabyssal plain has seldom been addressed in Qua-ternary glacimarine sequences, and their relation-ship to glaciation often remains unclear. Quanti-tative studies of the di¡ering contributions to thetotal sediment storage on the ocean £oor havebeen limited traditionally to seismic pro¢le-basedmeasurement of stratigraphic units (e.g. Alibe¤s etal., 1999). With the gradual collation of large-scale geophysical surveys of the European mar-

0025-3227 / 02 / $ ^ see front matter ; 2002 Elsevier Science B.V. All rights reserved.PII: S 0 0 2 5 - 3 2 2 7 ( 0 2 ) 0 0 2 7 5 - X

1 Present address: Scott Polar Research Institute and De-partment of Geology, Universiy of Cambridge, Lens¢eldRoad, Cambridge, CB2 1ER, UK.* Corresponding author. Tel. : +44-117-928-8186;

Fax: +44-117-928-7575.E-mail address: [email protected] (J. Taylor).

MARGO 3120 23-7-02

Marine Geology 188 (2002) 61^77

www.elsevier.com/locate/margeo

gins of the Norwegian^Greenland Sea, we cannow produce relatively well-constrained LateQuaternary budgets of sediment transfer fromthe continental shelf and upper slope into theabyssal plain for two high-latitude glacially in£u-enced basins.

This study presents geophysically derived esti-mates of the magnitude and frequency of sedi-ment transfer processes from the continental shelfand upper slope to the abyssal plain within theLofoten and Norwegian Basins, in the PolarNorth Atlantic, during the Late Weichselian(30^10 ka) and Holocene (Fig. 1). The sourceand sink areas of sediments derived from key pro-cesses are mapped directly using two large data-sets of GLORIA 6.5 kHz long-range side-scansonar imagery and contemporaneously collected3.5 kHz sub-bottom pro¢ler records, coveringmost of the margins below 500^750 m waterdepth of the 290 000 km2 Norwegian and the485 000 km2 Lofoten Basins, respectively (Fig.1). The mapped sedimentation processes arethen combined with published bathymetric data,in order to derive an estimate of the sedimentvolumes transferred into the deep ocean. Sedi-ment £uxes are derived by applying known chro-nological controls. We then discuss how the spa-tial patterns of resedimentation may be linked toice-sheet dynamics and glacial history, using nu-merically modelled predictions of the patterns andvolumes of sediment delivered from Weichselianice-sheets on Fennoscandia and in the Barents Sea(Dowdeswell and Siegert, 1999).

2. Phsyiographic setting and sediment deliverymechanisms

The Norwegian and Lofoten Basins form thefocus of this study (Fig. 1). The basins are de¢nedfrom the features associated with the dominanttectonic controls on oceanic basalt emplacement.The seaward extent of the basins is therefore de-¢ned by the Mid-Atlantic Ridge, comprising theKnipovich and Mohns Ridges for the LofotenBasin. The Norwegian Basin is de¢ned by thenow extinct Aegir Ridge spreading axis. Theseridges and spreading troughs form substantial

barriers to sediment transfer on the abyssal plainsfrom the Norwegian and Faeroes margins. TheVQring Plateau and Jan Mayen Fracture Zoneseparate the Norwegian and Lofoten Basins, andtwo independent sedimentary basins are thus ex-amined. The major physiographic features of thecontinental shelves around the Lofoten and Nor-wegian Basins are cross-shelf troughs, primarilythe Norwegian Channel, the BjQrnQya Trough,and a series of cross-shelf troughs on the rela-tively narrow north Norwegian shelf (Fig. 1).

During interglacials, sediment delivery acrossthe shelf is minor. Overdeepened fjords form tem-porary stores of terrestrially derived sediment,trapping the majority of sediment eroded andtransported from glacially in£uenced land massesduring interglacials (Sejrup et al., 1996). Sedimentsupply to the deep ocean is often redirected bycontour currents, becoming deposited as majorsediment drifts on the North Faeroes and Norwe-gian margins (Nielsen et al., 1998; van Weering etal., 1998; Laberg et al., 1999) and winnowing ofNorwegian shelf sediments (Kenyon, 1986; Vor-ren et al., 1998). Hemipelagic and pelagic sedi-mentation rates, as recorded in sediment cores,are generally low (Fig. 2; Taylor et al., 2002).

Sediment delivery to this continental margin inthe Quaternary is therefore suggested to be dom-inated by the growth and decay of ice-sheets onFennoscandinavia. The control exerted by ba-thymetry is seen through the development offast-£owing ice streams in cross-shelf troughsand the resulting enhancement of sediment supplyto the shelf edge at their margins. Hemipelagic,pelagic, and glacimarine rates of sedimentationduring glacial conditions are generally much high-er than interglacial values, and are controlledstrongly by the proximity and character of theice-sheet (e.g. Jansen et al., 1983; Dowdeswell etal., 1996; Dowdeswell and Siegert, 1999; Tayloret al., 2002) (Fig. 2).

Subsequent resedimentation of material deliv-ered to the continental margins occurs throughthree principal mechanisms (e.g. Vorren et al.,1998): (a) large-scale mass failures, (b) debris£ows which are found mainly on submarinefans, and (c) sedimentation through channelsand canyons. The division is based on di¡erences

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^7762

Fig. 1. The study area in the Norwegian^Greenland Sea, showing Lofoten and Norwegian Basins (black polygons) and the GLO-RIA and 3.5 kHz survey area (white outline). Contours are at 200, 500, and 3000 m water depth.

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^77 63

in process, interpreted, for example, in terms ofsediment £uid content, levels and pervasiveness ofshear stress, and degree of sorting of the ¢naldeposit (Table 1) (Mulder and Cochonat, 1996).The geophysical expression and character of theproducts of these processes is outlined in Table 1and forms the basis for the mapping of theirthree-dimensional extent from long-range side-scan sonar and 3.5 kHz records.

The major data sources in examining regionalsedimentation regimes are the PONAM geo-physical survey conducted aboard the R.R.S.

James Clark Ross during summer 1994, and theENAM II survey from the R.V. Siren duringJuly^August 1996. In both cruises, the Southamp-ton Oceanography Centre (SOC) 6.5 kHz GLO-RIA long-range side-scan sonar was used in con-junction with an SOC 3.5 kHz sub-bottompro¢ler for bathymetric control and shallow sub-surface architecture. An area of 30 000 km2 of theNorwegian Basin and 60 000 km2 of the LofotenBasin, respectively, were imaged by GLORIA,and over 15 000 trackline km of geophysicaldata were acquired (Dowdeswell et al., 1996).

Fig. 2. Comparing (a) Holocene and (b) Late Weichselian linear sedimentation rates from core records in the Norwegian^Green-land Sea, it is readily evident that (hemi)pelagic accumulation and the delivery of sediment from continents to the deep ocean ba-sins is much less during interglacial periods than during interglacials. Adapted from Taylor et al. (2002). High sedimentationrates o¡ Greenland during the Holocene are attributable to elevated glacimarine sedimentation near Shannon Island on the conti-nental shelf by the ice-sheet. The same pattern of elevated ice-sheet-in£uenced sedimentation is seen during the Late Weichselian,especially between Svalbard and Norway.

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^7764

GLORIA imagery was processed and mosaiceddigitally. The numerically modelled predictionsof ice-sheet history, behaviour, dynamics, andsediment delivery used in this study are describedfully in Dowdeswell and Siegert (1999).

3. Margin sedimentation processes and volumeestimates

The extent and locations of the sediments de-rived from each of the three major resedimenta-tion processes, mapped from the extensive geo-physical datasets outlined above, are shown inFig. 3. Whilst the map is similar in outline toseveral others published recently, there are signi¢-cant di¡erences between this map and those pre-sented in Vorren et al. (1998) and Vogt et al.(1999). Di¡erences occur because these previousmaps are based largely on preliminary interpreta-tions of GLORIA and 3.5 kHz data (Dowdeswelland Kenyon, 1995), which have since been re¢ned(Taylor, 2000). Several features are mapped de¢n-itively in their full extent for the ¢rst time, and

signi¢cant alterations have been made to theunderstanding of the extent and behaviour of sev-eral key processes in this revised mapping. Exam-ples of the data used to map each of the processesare shown in Fig. 4.

3.1. Trough mouth fans

The trough mouth fans (TMFs) in the studyarea are represented by the BjQrnQya and theNorth Sea Fans, located at the mouths of theBjQrnQya Trough and Norwegian Channel, re-spectively (Fig. 3). The full extent of the NorthSea Fan is shown for the ¢rst time, reaching tothe foot of the Aegir Ridge, as originally sug-gested by Vogt (1997), and inter¢ngering withmass movement deposits from the North Faeroesmargin in the west (Taylor et al., 2000b). TheGLORIA and 3.5 kHz records demonstrate thepresence of cohesive glacigenic debris £ows togreater than 3000 m water depth (Taylor, 2000).

The area most recently active (Late Quater-nary) on the BjQrnQya TMF in terms of debris£ow activity is dramatically smaller than that sug-

Table 1Marine mass-wasting classi¢cation is based on the continuum of processes de¢ned by end-points associated with £uid content ofthe sediment, the level and pervasiveness of shear stress, and the amount of sorting

Also shown are the generalised geophysical appearance of the three principal resedimentation processes on 3.5 kHz sub-bottompro¢ler records and GLORIA side-scan sonar imagery. The former is described in terms of Damuth’s (1978) acoustic facies clas-si¢cation, to which further reference should be made.

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^77 65

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^7766

gested by previous work (Laberg and Vorren,1995, 1996) (Fig. 3). This is clear from both theGLORIA records, which show debris £ow activityto be concentrated on the central and northernsections of the fan (Fig. 4) and 3.5 kHz data,which indicate that debris £ows have been absentfrom the southern fan for a substantial (105 yr)period of time (Taylor, 2000). Glacigenic debris£ows are buried deeply beneath a thick sequence(s 50^75 m) of strati¢ed sediments on the south-ern fan, imaged on the 3.5 kHz records. Thisstrati¢ed sequence has been eroded subsequentlyby the BjQrnQyrenna Slide, dated to at least 200 kaBP (Laberg and Vorren, 1993; Vorren et al.,1998). Furthermore, in many areas on the abyssalplain, mapped previously as part of the BjQrnQyaTMF, the GLORIA imagery shows that thedownslope transport paths of many of these de-posits are inconsistent with a TMF origin. Thedistal deposits of mass movements, sourced fromthe Norwegian margin, are very similar in appear-ance to debris £ows from the fan on 3.5 kHzrecords, explaining previous overestimates of de-bris £ow activity. The area of the most recently(Weichselian) active BjQrnQya TMF is thereforede¢ned here as 125 000 km2, approximately halfthat of previous estimates (215 000 km2) (Fig. 3).Older fan deposits may cover a larger area, how-ever (Vorren et al., 1998). The BjQrnQya TMF hasa run-out distance of 490 km, and upper and low-er widths of 250 and V350 km, respectively. TheNorth Sea Fan covers 108 000 km2, with a run-out distance of 490 km, and upper and lowerwidths of 190 and 250 km.

The volume of sediments stored is derivedstraightforwardly from the trough mouth fandata (Table 2); the uppermost sequence of £owsare imaged fully on GLORIA records (Fig. 4a),and their three-dimensional geometry is well-de-¢ned from 3.5 kHz data (Fig. 4b and c). Approx-imating glacigenic debris £ow cross-section as anellipse, cross-sectional area is given as (ZW(1/2)dW(1/

2)w), where d and w are maximum thickness andwidth, respectively. Flow volume is de¢ned as theaverage cross-sectional area multiplied by length,and is justi¢ed by the observation that debris£ows do not vary signi¢cantly in width downslope(Taylor, 2000). Debris £ows are stacked three tofour lenses deep on 3.5 kHz records from theBjQrnQya TMF (Fig. 4b and c), above a regionallyextensive strati¢ed re£ector that is interpreted torepresent sedimentation prior to the onset of LateWeichselian glaciation at V28 ka (Dowdeswelland Siegert, 1999). The total Late Weichselianvolume of sediment stored on the BjQrnQyaTMF is set, therefore, at three times the volumein the uppermost package of debris £ows, or 820^1100 km3. The discrepancy between this ¢gureand the previous estimate of 4000 km3 (Labergand Vorren, 1996) is a direct result of our mea-surement of the much smaller area of the fanactively accreting sediment through glacigenic de-bris £ows. The same principles are applied to theNorth Sea Fan. Seismic sections from the NorthSea Fan suggest the presence of debris £owsstacked four lenses deep in the Late Weichselianunit (e.g. Fig. 2, King et al., 1998), the upper layerof which is mapped from GLORIA and 3.5 kHzdata as containing an estimated 200 km3 of sedi-ment (Taylor, 2000). A sediment volume of ap-proximately 800 km3 is therefore suggested tohave been deposited during the Late Weichselianon the North Sea Fan.

3.2. Large-scale failure

Large-scale failures on the eastern margin ofthe Norwegian^Greenland Sea are widespread,and several of these failures are well-de¢ned al-ready (e.g. Vorren et al., 1998). However, ourmapping identi¢es the full extent and nature ofseveral slides for the ¢rst time (Fig. 3; Table 2).The most well-understood large-scale failure is theStoregga Slide. The extent of sliding in this area is

Fig. 3. Major sedimentary processes mapped within the Lofoten and Norwegian Basins on the basis of GLORIA and 3.5 kHzdata. The association between trough mouth fans and cross-shelf troughs is clear. The Bellesund and Storfjorden trough mouthfans are not covered in su⁄cient detail within the database used here and are therefore not shaded or included in further discus-sion. Contours are at 200, 500, and 3000 m water depth.

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^77 67

Fig. 4. GLORIA 6.5 kHz long-range side-scan sonar and 3.5 kHz sub-bottom pro¢ler examples of processes mapped and used toconstruct £uxes estimated in this study. (a) Debris £ows on the BjQrnQya Trough Mouth Fan are elongate, lobate, and (b and c)characterised by extensive stacking. (d) Slides are generally imaged as extensive lobate rough deposits downslope of lower back-scatter evacuation zones and steep headwalls, as demonstrated at the AndQya Slide (Laberg et al., 2000). Channels are easilyidenti¢ed on both (e) GLORIA and (f) 3.5 kHz records. Examples located in (g).

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^7768

derived from Bugge et al. (1988) and requires onlyvery minor adaptation to match the middle slopeof its western margin, as seen on the Siren GLO-RIA mosaic (Taylor, 2000). The extent of theAndQya Slide is that presented in Laberg et al.(2000) (Fig. 4d), whilst the extent of and datingto the Weichselian of mass failure on the NorthFaeroes margin is taken from Taylor et al.(2000b). Both of the latter extents are derivedprimarily from GLORIA and 3.5 kHz records.Previously unrecognised areas of mass movementare identi¢ed, from GLORIA and 3.5 kHz data,located immediately north of the AndQya Slide(termed here the Fugloy Bank Slide complex)and between the BjQrnQya and StorfjordenTMFs. The upslope limits of failure are not visi-ble in the geophysical data available, however(Fig. 3). The downslope limit of the BjQrnQyrennaSlide is de¢ned fully, the extent of the slide lo-

cated on the BjQrnQya TMF, as mapped fromour present dataset, matching well with the limitsde¢ned by Laberg and Vorren (1993) from seismicrecords. The existence of sliding on the northernmargin of the VQring Plateau, the TrUnadjupetSlide, was ¢rst indicated by Damuth (1978). Thecombination of data available here allows thecomplete downslope extent of failure products as-sociated with the TrUnadjupet Slide to be identi-¢ed. A relatively small area of failure is also iden-ti¢ed between the Inbis Channel and StorfjordenTMF, immediately downslope of Vestbakken(Fig. 3). However, the data outlined above donot su⁄ciently resolve this area of instability toinclude it further in investigations, although itsexclusion is not thought to a¡ect our conclusionssigni¢cantly.

Volume estimates derived for large-scale fail-ures must be treated with some caution. The

Table 2Volumes of sediment and time-dependent £uxes associated with individual mapped units within the Lofoten and Norwegian Ba-sins

Sedimentary unit 30^0 ka sedimentvolume

30^0 ka sediment£ux

Notes Source

(km3) (km3/kyr)

BjQrnQya TMF 820^1100 V200 3 or 4 units of 274 km3 over 5 kyr this studyNorth Sea Fan 800 160 4 units of 200 km3 over 5 kyr this studyStoregga Slide 5580 instantaneous (0.5W(34 000/290)WV0.3)W290 Bugge et al. (1988)AndQya Slide 900 instantaneous (0.5W(3600/100)WV0.5)W100 cf. Laberg et al. (2000)North Faeroes Slide 6 1700 instantaneous (0.5W(11 200/40)WV0.3)W400 Taylor et al. (2000b)BjQrnQyrenna Slide 1400 instantaneous (0.5W(7000/550)WV0.4)W55 this studyTrUnadjupet Slide 6 1900 instantaneous (0.5W(7500/130)WV0.5)W130 this studyFugloy Bank Slide ?? instantaneous insu⁄cient dataUnlabelled Slide ?? instantaneous insu⁄cient dataNorwegian margincanyons

35 instantaneous 8 canyons at 2^5 km3 Taylor et al. (2000a)

Lofoten Channel system s 800 s 0.4 s 30 m thick over 25 000 km2 fors 2 Myr

this study

North Faeroes Channelsystem

?? ?? insu⁄cient data Taylor et al. (2000b)

Inbis Channel system ?? ?? insu⁄cient data (Vorren et al., 1998)InterglacialNorwegian Basin 58 5.8 2 cm/kyrW10 kyrW290 000 km2 this studyLofoten Basin 97 9.7 2 cm/kyrW10 kyrW485 000 km2 this studyGlacialNorwegian Basin 290 15 5 cm/kyrW20 kyrW290 000 km2 this studyLofoten Basin 485 25 5 cm/kyrW20 kyrW485 000 km2 this study

Processes are divided into debris £ows on trough mouth fans, large-scale failures, channel and canyon systems, and interglacialand glacial pelagic-related processes. The notes provide details on how sediment volumes are calculated (see text for more generalexplanations of methods). These data are based on mapping from this study, unless indicated otherwise. Large-scale failure domi-nates both sediment volume and £ux within both basins.

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^77 69

Storegga Slide volume includes distal mega-turbi-dite deposits. Similar distal deposits are likely toexist for other failures on the margin, but theyremain unidenti¢ed as yet. The only other slidesystem which is mapped to a similar level of com-pleteness is that on the North Faeroes margin,where distal mega-turbidites are con¢ned to achannel system and regional deep within the Nor-wegian Basin (Taylor et al., 2000b). However, thisproblem concerning depositional volume may becircumvented by calculating the volume lost with-in the slide scar, as represented in Fig. 5. Such acalculation does make the crude assumption thateach slide scar is a result of a single large-scalefailure event and evacuation. Such a scenario isunlikely, given the reported long-term instabilityrecorded at the site of the Storegga Slide, for ex-ample (Evans et al., 1996), two or three slidingepisodes normally being implicated. However,we are constrained by the limitations of thepresent chronology. Even assuming that each vol-ume transported, as calculated in this study, is aresult of several failure events does not alter thefundamental conclusions of our study. The vol-ume derived in this instance from the slide scarfor the Storegga Slide (5100 km3) matches wellwith the estimate from Bugge et al. (1988) (5580km3), providing con¢dence in the technique andsuggesting that, if anything, volume is underesti-mated. Assuming that the present Storegga Slide

scar is a result of two or three major episodes ofcollapse, upwards of 1800 km3 of sediment istransported to the Norwegian Basin in each fail-ure event. Calculated (single episode) volumes andspatial characteristics for each of the slides on theNorwegian and Faeroes margins are shown inTable 2 and Fig. 6.

3.3. Channel and canyon systems

Four channel and canyon systems are locatedwithin the Norwegian and Lofoten Basins. TheNorth Faeroes margin channel system is associ-ated with the large-scale failure located immedi-ately upslope (Fig. 3). However, the geophysicaldata are insu⁄cient to characterise fully the vol-ume of sediments held within the channel anddeposited within the regional deep (Taylor et al.,2000b). The canyon system on the north Norwe-gian margin has been described previously byTaylor et al. (2000a), who estimate that the can-yons have transferred 35 km3 of sediment fromthe edge of the continental shelf to the continentalrise during the Late Weichselian (Fig. 3).

The location and extent of the Lofoten Channelon the £oor of the Lofoten Basin have been rec-ognised for some time now (Dowdeswell andKenyon, 1995), but the nature of the channel sys-tem remains poorly understood. The channel issourced from the AndQya Canyon (Laberg et

Fig. 5. The volume of sediment a¡ected by sliding can be derived from the volume of the slide scar. This is approximated by awedge shape, triangular in cross-section. Erosional area, headwall width, and headwall height may all be measured directly. Nom-inal erosional length is then determined from total plan area divided by headwall width. Vertical cross-sectional area is then cal-culated and multiplied by headwall width to give slide volume. Errors from the cross-sectional area not being a perfect right-angled triangle are small (cross-sectional area= 0.5WlWh vs. 0.5WlWhWsina, where a=V89‡).

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^7770

al., 2000) and extends across the Lofoten Basininto the abyssal plain (Figs. 3 and 4e and f). Theentire channel system covers approximately 25 000km2, and sediments are generally at least 30 mthick across much of the rise and abyssal plain.This gives a minimum volume of 800 km3 ofsandy sediments stored within the system sinceits inception, probably some time prior to thePleistocene (cf. Laberg et al., 2000). The InbisChannel, located adjacent to the northern marginof the BjQrnQya TMF (Fig. 3), is described byVorren et al. (1998) and is also seen fully inboth GLORIA and 3.5 kHz data (Taylor, 2000).However, these data are not suitable in resolutionor well enough calibrated sedimentologically tocharacterise the system in terms of the volumeof sediment within it.

3.4. Pelagic, hemipelagic, and glacimarine £uxes

Estimates of hemipelagic, pelagic, and glaci-marine sedimentary £uxes are estimated veryroughly from available core records. Average sed-imentation rates are estimated for the Late Weich-selian glacial (30^10 ka) and Holocene interglacial(10^0 ka) periods, a single value being used forboth the Lofoten and Norwegian Basins. Overall,glacial sedimentation rates are much greater thanthose of any given location during interglacials(Fig. 2; Taylor et al., 2002). However, exceptionsare found in some sedimentary environments,

such as areas of contourite deposition (e.g. Ras-mussen et al., 1996). In this study, therefore, theglacial period value is set higher than the inter-glacial value, to re£ect the greatly elevated ratesof glacimarine sedimentation on parts of the mar-gin adjacent to ice-sheets (e.g. King et al., 1996,1998). The values used here are 2 cm/kyr and5 cm/kyr for interglacial and glacial time periods,respectively. We feel that these values re£ect thedecrease of sedimentation rate with water depth,the dominance of continental rise and abyssalplains in basin hypsometries and, thus, to be rep-resentative of the entire basin.

Therefore, total interglacial sediment thicknessdelivered to the Lofoten and Norwegian Basins isestimated at 20 cm (10 kyr at 2 cm/kyr; Table 2).Less than 50 cm of sediment is consistent with thevertical resolution of the 3.5 kHz records, whichdo not resolve a Holocene sediment cover acrosseither the North Sea Fan or BjQrnQya TMF, forexample. These values give sediment volumes of58 km3 for the Norwegian Basin and 97 km3 forthe Lofoten Basin (Table 2). Glacial sediment vol-umes for the Norwegian and Lofoten Basins are290 and 485 km3, respectively.

4. Basin accumulation £uxes and relation toice-sheet history and dynamics

From the data presented above, summarised in

Fig. 6. Summary diagram showing morphology of ¢ve of the major slides in the Norwegian^Greenland Sea. All slide representa-tions are proportional, using headwall width, erosion and accumulation area, and slide volume as the de¢ning parameters. NFMrefers to North Faeroes margin slides. Su⁄cient data are not available for the Fugloy Bank Slide (Fig. 2) and an unlabelled slidebetween the BjQrnQya and Storfjorden trough mouth fans to be included in this analysis.

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^77 71

Table 2 and Fig. 7, the major sources and therelative importance of sediment input into the gla-cially in£uenced Norwegian and Lofoten Basinsmay be assessed during the period 30^0 ka. Im-portantly, this period does not include the depo-sition of the BjQrnQyrenna Slide, on the basis ofits assignment to the pre-Weichselian (Laberg andVorren, 1993). However, the slide is an importantindicator that large-scale failure is a widespreadfeature beyond the last 30 000 yr.

Sediment £uxes are determined from absolute

volume estimates by assigning timespans for theoperation of each sedimentation process (Table2). In the case of large-scale failure, depositionis nominally instantaneous; the £ux presented,therefore, necessarily represents a large volumeof sediment transported to the deep ocean overthe 30 000 yr timespan in question. Fluxes forthe fans are calculated for the time interval overwhich ice streams are understood to have beenpresent at the shelf break (V5 kyr for boththe Norwegian Channel and BjQrnQya Trough)

Fig. 7. Sediment volumes (km3) delivered to the continental rise and abyssal plain during the period 30^0 ka for both the Norwe-gian and Lofoten Basins (a) and individually (b and c). Fan, slide, channel, and pelagic-related processes are boxed by heavylines. Sedimentation is dominated by large-scale failures, which account for up to 75% of all reworked sediment. Only approxi-mately 10% of sediment is delivered to the basins by pelagic, hemipelagic, and glacimarine processes.

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^7772

(Dowdeswell and Siegert, 1999). The £ux throughthe Lofoten Channel is estimated as total sedi-ment deposited over its likely minimum age (800km3 in s 2 Myr) and thus represents a value thatis not sensitive to glacial/interglacial contrasts.

By applying a tentative timescale to these £uxdata, it is possible to estimate the nature of sedi-ment delivery to glacially in£uenced basins overtime. Such a record through time is presented forthe Lofoten Basin (Fig. 8), with some cautionarynotes. The timescale represents the last glacialcycle (isotope stage 5d to present) by applyingmeasured Late Weichselian £uxes to the previoustwo stadial events during the Weichselian. Thechronology of the AndQya and TrUnadjupetSlides, like many slides on the northern Europeanmargin, is not constrained tightly; both are sug-gested to be Holocene in age (Laberg et al., 2000;Laberg and Vorren, 2000). It nonetheless illus-trates the point that large-scale failure and glaci-genic debris £ow processes have completely dom-inated the input of sediment to this glacier-in£uenced margin during at least the last 30 000,

and probably several 100 000 yr. There are alsosigni¢cant glacial/interglacial contrasts in therate of sedimentation, directly related to themode of sediment supply, and most of the sedi-ment is delivered within a very short timespan,even within periods of glaciation (Fig. 8).

The majority of sediment input to both theLofoten and Norwegian deep sea basins is derivedfrom large-scale failures (60 and 85%, respec-tively). Overall, approximately 75% of all sedi-ment in the two basins are deposited throughthis process. These conclusions are una¡ected bythe dominance of the Storegga Slide in the data,because evacuated slide volume has been used inthe analysis, rather than measured deposit vol-ume. Submarine fan deposition through glacigenicdebris £ows is about 15^20% of the total. It isroughly equivalent to any one large slide (Table2; Fig. 7), although deposition occurs over a lon-ger time period and represents almost exclusivelythe reworking of sediment transported to the shelfedge and upper slope during the Late Weichseli-an, because of its subglacial origin (Laberg and

Fig. 8. Hypothesised sedimentation rates for the Lofoten Basin since oxygen isotope stage 5d (dashed line), incorporating mapped£uxes for the Late Weichselian (30^10 ka) (solid line). Rates are km3 per kyr delivered to the Lofoten Basin continental rise.Variations in rates re£ect the growth and decay of the Eurasian Ice-Sheet and the associated switching on and o¡ of sedimenta-tion on the BjQrnQya TMF, elevated glacial hemipelagic rates, and the activation of the Lofoten Channel. The ends of glacial pe-riods are associated with the emptying of the Norwegian margin canyons, although the additional volume of sediment deliveredis small. The TrUnadjupet and AndQya Slides represent major but short-lived rises in sedimentation rates, although the timingsare poorly constrained. Note breaks in sedimentation scale. Smoothed GRIP N

18O record and marine oxygen isotope stages fortimescale (Dansgaard et al., 1993).

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^77 73

Fig. 9. The relation between glaciation limits (dark dotted line), ice-sheet sediment £ux predicted by numerical modelling (darkcircles) (Dowdeswell and Siegert, 1999) and total delivery of sediment to the deep ocean (grey circles) associated with the majorlarge-scale processes identi¢ed in this study for the period 30^0 ka BP. Ice-sheet dynamics are represented through the velocity ofice predicted from numerical modelling. Major slides are associated with medium ice-sheet sediment £ux and high sediment deliv-ery rates to ocean basins, whilst trough mouth fans are characterised by very high £ux and medium delivery rates. Channels andcanyons are characterised by both low £ux and delivery. Addition of sediment from major slides is geologically instantaneous,giving the apparently extremely high deep ocean transfer £uxes during the Late Weichselian.

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^7774

Vorren, 1995). Deposition by channel systemsrepresents a very small proportion of total depo-sition, generally less than 1%. Pelagic, hemipela-gic, and glacimarine depositions account for theremaining 6 1^4% of basin sedimentation.

The timing of sediment delivery is related to thedominant control of ice-sheet history. By compar-ing sediment £uxes to the Lofoten and NorwegianBasins with numerically modelled sediment £uxdelivered to the edge of the continental shelf dur-ing the Late Weichselian (Fig. 9) (Dowdeswelland Siegert, 1999), the role of ice-sheet dynamicsin controlling sediment transfer can also be ad-dressed. Ice-sheet delivery of sediment at the topof the continental shelf is concentrated at themouths of fast-£owing ice streams formed incross-shelf troughs. Rapidly delivered sedimentis reworked e⁄ciently into the basins by glacigen-ic debris £ows that form the building blocks ofthe large submarine fans. In contrast, low sedi-mentation rates associated with channel and can-yon systems are related to relatively stagnantparts of the ice-sheet margin (Dowdeswell et al.,1996; Taylor et al., 2000a). Intermediate rates ofice-sheet delivery of sediment are associated withthe extremely large, geologically ‘instantaneous’£uxes of sediment from slides on the margin.

We can de¢ne a range of sediment £uxes fromour ice-sheet numerical model (Dowdeswell andSiegert, 1999) that appear to be related to speci¢ctypes of subsequent sediment-reworking processesat the continental margin. Therefore, rapid sedi-ment delivery at the shelf break (0.6 km3/kyr/kmice front) results in glacigenic debris £ows and thebuilding of fan depocentres. Extremely low sedi-ment delivery rates (6 0.05 km3/kyr/km ice front)are associated with channel and canyon systems,that would otherwise be swamped by glacial ma-terial. Intermediate rates of ice-sheet sediment £ux(0.05^0.1 km3/kyr/km ice front) are linked tolarge-scale failure. This represents perhaps themost important aspect of this study, because themodelled rate of sediment delivery can be used topredict the sedimentary processes likely to be ac-tive within the basin downslope of any givenpoint. More particularly, the intermediate ratesof sediment delivery by ice-sheets appear to pre-condition the margin for large-scale failure in

some fashion. This is probably because, in con-trast to the higher £ux trough mouth fans, debrisis not reworked immediately. Sliding is not foundwhere sediment £ux to the ice-sheet margin is ei-ther extremely low or very high. The relativelyfast deposition of glacigenic sediment over weakand poorly consolidated hemipelagic and glaci-marine deposits may lead to the creation of planesof weakness, for example, or the loading of pre-existing potential failure planes. Observations onsediment supply rates to source areas of failureare thus important in constraining the role ofthe build-up of excess pore water pressures as acause of sliding, for example. Intermediate sedi-ment supply rates seem to represent a conditionduring the Late Quaternary conducive to large-scale failure, although other factors (gas hydratedecomposition, seismic activity, etc.) will providethe ultimate, direct trigger for sliding.

5. Conclusions

1. Sediment delivery into the Norwegian andLofoten Basins during the Late Weichselian (30^0 ka) is dominated by glacial processes (Dowdes-well et al., 1996; Vorren et al., 1998; Dowdeswelland Siegert, 1999). Much of this sediment is re-worked subsequently, a¡ecting the majority of thearea of both basins (Fig. 3).

2. Approximately 75% of sediment deliveredinto basins is in the form of large-scale marginfailure (Figs. 6 and 7). Sediment input in thisform is geologically instantaneous and over-whelms sediment supply from other sources atthe timescale considered (Table 2).

3. Trough mouth fans are the other major sedi-mentary system, to which 16% of all Late Weich-selian sediment is delivered, mainly by glacigenicdebris £ows (Fig. 7). Only V7% of the sedimentstored in the Lofoten and Norwegian Basins isderived from undisturbed pelagic, hemipelagic,or glacimarine processes (Fig. 7).

4. Time-dependent £ux of sediments to theNorwegian margin is suggested to be cyclicaland related directly or indirectly to the growth,decay, and dynamics of ice-sheets during glacialand interglacial periods (Fig. 8).

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^77 75

5. Margin sedimentation processes are relatedclosely to the delivery £ux of sediments to theshelf break and upper slope. Sediment deliveredrapidly to the shelf break is reworked as glacigen-ic debris £ows on trough mouth fans. Low sedi-ment delivery rates result in the generation orpreservation of channel and canyon systems(Fig. 9).

6. Intermediate sediment delivery rates appearto prime the continental margin for large-scalemass failure (Fig. 9). This association is impor-tant, because de¢ning the range of sediment accu-mulation rates at the top of the slope at locationsof sliding is critical in furthering understanding ofthe processes resulting in failure.

7. Ice-sheets and their dynamics play a quanti-¢able and fundamental role in the pattern andrate of sediment accumulation in high-latitudedeep ocean basins. The majority of the sedimentis delivered and reworked by mechanisms whichare related directly or indirectly to ice-sheet histo-ry and behaviour.

Acknowledgements

The R.R.S. James Clark Ross survey wasfunded by the U.K. NERC grant GR3/8508 toJ.A.D. and N.H. Kenyon, and is a contributionto the ESF programme on the ‘Polar North At-lantic Margins: Late Cenozoic Evolution’ (PO-NAM). We thank Captain C.R. Elliott and hiscrew, and the scienti¢c support sta¡ from theBritish Antarctic Survey and SouthamptonOceanography Centre (SOC). The Siren surveywas carried out with funding from EU MASTIII, ENAM II grant MAS3-CT95-0003. We thankCaptain I. Simpson and his crew for logisticalsupport, and the SOC technicians for e¡ectivehandling of the GLORIA and 3.5 kHz systems.Additional thanks are due to Dr. T.P. Le Bas,SOC, for his training of J.T. in GLORIA imageprocessing. J.T. acknowledges the ¢nancial sup-port of the University of Wales, Aberystwythand the University of Bristol. We also thank D.Evans, C.F. Forsberg, and J. Scourse for usefuland constructive reviews.

References

Alibe¤s, B., Rothwell, R.G., Canals, M., Weaver, P.P.E., Alon-so, B., 1999. Determination of sediment volumes, accumu-lation rates and turbidite emplacement frequencies on theMadeira Abyssal Plain (NE Atlantic): a correlation betweenseismic and borehole data. Mar. Geol. 160, 225^250.

Bugge, T., Belderson, R.H., Kenyon, N.H., 1988. The Sto-regga Slide. Philos. Trans. R. Soc. London A 325, 357^388.

Damuth, J.E., 1978. Echo character of the Norwegian-Green-land Sea: relationship to Quaternary sedimentation. MarineGeology 28, 1^36.

Dansgaard, W., Johnsen, S.J., Clausen, H.B., Dahl-Jensen, D.,Gundestrup, N.S., Hammer, C.U., Hvidberg, C.S., Ste¡en-sen, J.P., Sveinbjo«rnsdottir, A.E., Jouzel, J., Bond, G., 1993.Evidence for general instability of past climate from a 250-kyr ice-core record. Nature 364, 218^220.

Dowdeswell, J.A., Kenyon, N.H., 1995. Long-Range Side-Scan Sonar Investigations of the Polar North Atlantic: Pat-terns and Processes of Sedimentation on a Glaciated PassiveContinental Margin. Cruise Report: R.R.S. James ClarkRoss ^ Cruise 08. Report, University of Wales, Aberyst-wyth.

Dowdeswell, J.A., Kenyon, N.H., ElverhQi, A., Laberg, J.S.,Hollender, F.-J., Mienert, J., Siegert, M.J., 1996. Large-scalesedimentation on the glacier-in£uenced Polar North Atlanticmargins: Long-range side-scan sonar evidence. Geophys.Res. Lett. 23, 3535^3538.

Dowdeswell, J.A., Siegert, M.J., 1999. Ice-sheet numericalmodeling and marine geophysical measurements of glacier-derived sedimentation on the Eurasian Arctic continentalmargins. Bull. Geol. Soc. Am. 111, 1080^1097.

Evans, D., King, E.L., Kenyon, N.H., Brett, C., Wallis, D.,1996. Evidence for long-term instability in the Storegga Slideregion o¡ western Norway. Marine Geology 130, 281^292.

Jansen, E., Sejrup, H.P., FjUran, T., Hald, M., Holtedahl, H.,SkarbQ, O., 1983. Late Weichselian paleoceanography of thesoutheastern Norwegian Sea. Nor. Geol. Tidsskr. 63, 117^146.

Kenyon, N.H., 1986. Evidence from bedforms for a strongpolewards current along the upper continental slope ofNorthwest Europe. Mar. Geol. 72, 187^198.

King, E.L., Ha£idason, H., Sejrup, H.P., LQvlie, R., 1998.Glacigenic debris £ows on the North Sea Trough Mouthfan during ice stream maxima. Mar. Geol. 152, 217^246.

King, E.L., Sejrup, H.P., Ha£idason, H., ElverhQi, A., Aar-seth, I., 1996. Quaternary seismic stratigraphy of the NorthSea Fan: glacially-fed gravity £ow aprons, hemipelagic sedi-ments, and large submarine slides. Mar. Geol. 130, 293^315.

Laberg, J.S., Vorren, T.O., 1993. A Late Pleistocene submarineslide on the Bear Island Trough Mouth Fan. Geo-Mar. Lett.13, 227^234.

Laberg, J.S., Vorren, T.O., 1995. Late Weichselian submarinedebris £ow deposits on the Bear Island Trough Mouth Fan.Mar. Geol. 127, 45^72.

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^7776

Laberg, J.S., Vorren, T.O., 1996. The Middle and Late Pleis-tocene evolution of the Bear Island Trough Mouth Fan.Glob. Planet. Change 12, 309^330.

Laberg, J.S., Vorren, T.O., 2000. The TrUnadjupet Slide, o¡-shore Norway ^ morphology, evacuation and triggeringmechanisms. Mar. Geol. 171, 95^114.

Laberg, J.S., Vorren, T.O., Dowdeswell, J.A., Kenyon, N.H.,Taylor, J., 2000. The AndQya Slide and AndQya Canyon,north-eastern Norwegian^Greenland Sea. Mar. Geol. 162,259^275.

Laberg, J.S., Vorren, T.O., Knutsen, S.-M., 1999. The Lofotencontourite drift o¡ Norway. Mar. Geol. 159, 1^6.

Mulder, T., Cochonat, P., 1996. Classi¢cation of O¡shoreMass Movements. Journal of Sedimentary Research 66,43^57.

Nielsen, T., Van Weering, T.C.E., Andersen, M.S., 1998. Ce-nozoic changes in the sedimentary regime on the northeast-ern Faeroes margin. In: Stoker, M.S., Evans, D., Cramp, A.(Eds.), Geological Processes on Continental Margins: Sed-imentation, Mass-Wasting and Stability. Geol. Soc. LondonSpec. Publ. 129, 167^171.

Rasmussen, T.L., Thomsen, E., van Weering, T.C.E., Labey-rie, L., 1996. Rapid changes in surface and deep water con-ditions at the Faeroe Margin during the last 58 000 years.Paleoceanography 11, 757^771.

Sejrup, H.P., King, E.L., Aarseth, I., Ha£idason, H., ElverhQi,A., 1996. Quaternary erosion and depositional processes:western Norwegian fjords, Norwegian Channel and NorthSea Fan. In: De Batist, M., Jacobs, P. (Eds.), Geology ofSiliciclastic Shelf Seas. Geol. Soc. London Spec. Publ. 117,187^202.

Taylor, J., 2000, Large-scale sedimentation and ice-sheet dy-namics in the Polar North Atlantic. Unpublished Ph.D. the-sis, University of Bristol, 2 volumes.

Taylor, J., Dowdeswell, J.A., Kenyon, N.H., 2000a. Canyonsand Late Quaternary sedimentation on the North Norwe-gian margin. Mar. Geol. 166, 1^9.

Taylor, J., Dowdeswell, J.A., Kenyon, N.H., Whittington,R.J., van Weering, T.C.E., Mienert, J., 2000b. Morphologyand Late Quaternary sedimentation on the North Faeroesslope and abyssal plain, North Atlantic. Mar. Geol. 168,1^24.

Taylor, J., Tranter, M., Munhoven, G., 2002. Carbon cyclingand burial in the glacially in£uenced Polar North Atlantic.Paleoceanography, 17, 10.1029/2001PA000644.

van Weering, T.C.E., Nielsen, T., Kenyon, N.H., Akentieva,K., Kuijpers, A.H., 1998. Sediments and sedimentation atthe NE Faeroe continental margin; contourites and large-scale sliding. Mar. Geol. 152, 159^176.

Vogt, P.R., 1997. Hummock ¢elds in the Norway Basin andEastern Iceland Plateau: Rayleigh^Taylor instabilities?Geology 25, 531^534.

Vogt, P.R., Gardner, J., Crane, K., 1999. The Norwegian^Barents^Svalbard (NBS) continental margin: Introducing anatural laboratory of mass wasting, hydrates, and ascent ofsediment, pore water, and methane. Geo-Mar. Lett. 19, 2^21.

Vorren, T.O., Laberg, J.S., Blaumme, F., Dowdeswell, J.A.,Kenyon, N.H., Mienert, J., Rumohr, J., Werner, F., 1998.The Norwegian^Greenland Sea continental margins: mor-phology and Late Quaternary sedimentary processes andenvironment. Quat. Sci. Rev. 17, 273^302.

MARGO 3120 23-7-02

J. Taylor et al. /Marine Geology 188 (2002) 61^77 77