at SciVerse ScienceDirect

Marine and Petroleum Geology 32 (2012) 36e49

Contents lists available

Marine and Petroleum Geology

journal homepage: www.elsevier .com/locate/marpetgeo

Acoustic evidence for a gas migration and release system in Arctic glaciatedcontinental margins offshore NW-Svalbard

Anupama Rajan*, Jürgen Mienert, Stefan BünzDepartment of Geology, University of Tromsø, Dramsveien-201, 9037-Tromsø, Norway

a r t i c l e i n f o

Article history:Received 23 August 2011Received in revised form18 November 2011Accepted 27 December 2011Available online 9 January 2012

Keywords:Gas hydrateFluid migrationGas expulsionGlacigenic sequenceGlacial debris flowAcoustic anomaliesPockmarks

* Corresponding author. Tel.: þ47 776 46479; fax:E-mail addresses: anupamaraja@gmail.com, anup

jurgen.mienert@uit.no (J. Mienert), stefan.buenz@uit.

0264-8172/$ e see front matter � 2012 Elsevier Ltd.doi:10.1016/j.marpetgeo.2011.12.008

a b s t r a c t

High-resolution 3D and 2D seismic data offshore NW-Svalbard, west of Prins Karls Forland, providegeophysical evidence for geologically controlled fluid migration pathways, gas hydrate occurrence, andan active seabed gas expulsion system. The investigated seabed area coversw1600 km2 and lies betweenKongsfjorden cross-shelf trough in the north and Isfjorden cross-shelf trough in the south ranging inwater depths from w200 m on the shelf to 800 m on the upper continental slope. Acoustic evidence forpresent day methane release from the seabed to the water column comes from more than 220 gas flaresat the outer shelf while past methane release activity at the mid-shelf area is evident from pockmarkswithout flares. The fluid migration pathways towards the seabed can be drawn from sub-seabed acousticanomalies. Fluid migration towards the upper slope occurs mostly along strata in upslope direction andlargely prevails over vertical focused migration. Fluids accumulate in the uppermost part of the slope justwestward of the shelf break, where they are trapped beneath the prograding glacigenic sequence, whichis not permeable enough for fluids to migrate through. Fluids are expelled on the shelf where the base ofthe glacigenic sequence outcrops. Some gas-charged fluids may originate from deep-seated hydrocarbonreservoirs and can be temporally stored in gas hydrates in the shallow subsurface. Though evidence forhydrates on the uppermost slope is missing the seismic data from the lower slope clearly suggest theoccurrence of gas hydrate.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Focused sub-seabed fluid flow is a widespread phenomenon incontinental margins worldwide (e.g. Berndt, 2005). Here, methanemay migrate as free gas through permeable sediments oftenfollowing lateral extensions of strata boundaries or vertical exten-sion of faults. In shallower water, such as the North or Barents Sea,the release of methane from the seabed has the potential to enterthe atmosphere (Rehder et al., 1998; Reshetnikov et al., 2000)where it is 25 times more effective than the same amount of thegreen house gas CO2 (Lelieveld et al., 1993).

In the sub-seabed methane can form hydrate, an ice-like crys-talline solid, formed from water molecules and natural gas mainlymethane under low temperature and high pressure condition(Sloan, 1998). The zone where gas hydrates are stable is known asgas hydrate stability zone (GHSZ) (Sloan, 1998). An increase inocean bottom temperature can cause gas hydrate to dissociate

þ47 776 45600.ama.rajan@uit.no (A. Rajan),no (S. Bünz).

All rights reserved.

(e.g. Jung and Vogt, 2004;Mienert et al., 2005; Vogt and Jung, 2002)and might release free gas (gas bubbles) into the water column.Offshore NW-Svalbard, such gas releases are widely observed (e.g.Hustoft et al., 2009; Knies et al., 2004; Westbrook et al., 2009),though, a direct link to gas hydrate dissociation has yet to beproven.

Identifying and quantifying the gas hydrate reservoirs, thepotential methane release from the reservoir and the state of theircurrent stability in the prevailing condition of global warming istherefore crucial for understanding their potential impact onclimate change. Global warming is particularly evident in theNordic Seas and the Arctic where climate change appears to takeplace more rapidly than previously predicted by ocean warmingmodels (e.g. Graversen et al., 2008).

Risingmethane bubbles from the seafloor have been detected byvarious acoustic systems (Greinert et al., 2006). Flares discovered inthe water column along the shelf and slope of NW-Svalbardcorroborate the active release of methane from the seabed(Hustoft et al., 2009; Knies et al., 2004; Westbrook et al., 2009). Gasanalysis from water samples along the SW and NW-Svalbard shelfsuggests higher amount of methane concentrations in the watercolumn that are elevatedwith respect to background concentration

A. Rajan et al. / Marine and Petroleum Geology 32 (2012) 36e49 37

(Damm et al., 2005; Knies et al., 2004). Traces of thermogenic gas insediments offshore NW-Svalbard suggests emanations fromhydrocarbon prolific provinces in deeper sediment formations(Knies et al., 2004). Westbrook et al. (2009) hypothesize that gasoriginates from dissociating gas hydrates as a consequence of oceanwarming.

It appears that lithological variations, e.g. the transition fromhemipelagic sediments to glacigenic debris flows (GDF), has aneffect on gas hydrate formation (Bünz et al., 2003). Observationsfrom the major gas hydrate field of the mid Norwegian marginnorth of the giant Storegga submarine slide (Bryn et al., 2005;Haflidason et al., 2004; Mienert et al., 2005) show that the GDFmayhinder gas hydrate growth because it clearly interrupts a contin-uous bottom simulating reflector at the base of the GHSZ (Bünzet al., 2003). However, gas hydrates may also exist without a BSRas documented by Blake Ridge drilling results (e.g. Holbrook et al.,2002).

The main objective of this study is to enhance our under-standing of fluid flow pathways in the gas-hydrate prone Arcticcontinental margins offshore NW-Svalbard. A research cruise on R/V Jan Mayen in July 2009 provided a high-resolution 3D-PCableseismic data set, 2D seismic data, swath bathymetry data and CTDcasts to investigate the distribution of gas hydrates indicated by

Figure 1. (a) Location of study area offshore NW-Svalbard (map from IBCAO bathymetry, Jresolution 3D P-Cable seismic cube (white rectangular box), 2D seismic lines (black solidfigure shows the zoomed location of the acoustic flares). The working area lies between Konranging from less than 20 m up tp 270 m height. 10% of the recorded flares correspond to groto group III (20e90 m, green circle). Abbreviations: KF, Kongsfjorden; HFZ, Hornsund Fault

a bottom simulating reflector (BSR) in relation to fluid flow path-ways (enhanced seismic anomalies) and methane release from theseabed (pockmarks, flares).

2. Geological settings

The working area offshore NW-Svalbard lies west of Prins KarlsForland and extends from w200 m water depth at the shelf to800 m on the upper continental slope (Fig. 1aeb). The area inves-tigated covers approx. 1600 km2. The Svalbard continental marginformed as a result of several rifting episodes in the Late Paleocene/Early Eocene, continental breakup and the development of theNorwegian-Greenland Sea (Sundvor and Austegard, 1990). Duringthe Pleistocene ice ages, ice streams carved cross-shelf troughs intothe existing sediment formations (Ottesen et al., 2007). As a resultKongsfjorden in the north and Isfjorden in the south of Prins KarlsForland connect through deep troughs to the shelf break. The fastflowing ice streams causedmajor erosion on the shelf and sedimenttransport and deposition in prograding glacigenic sequences (e.g.Laberg and Vorren, 1995; Sættem et al., 1992; Solheim et al., 1998)on the adjacent continental slopes building up the glaciatedmargin. Glacigenic debris flows (GDF) originated during peakglaciations by sediment release along ice stream fronts at the shelf

akobsson et al., 2008), the stippled line show the COT (Hustoft et al., 2009). (b) High-lines) and mapped acoustic flares are indicated in the swath bathymetry map (insetgsfjorden in the north and Isfjorden in the south. The flares are on average 100 m highup I (180e270 m height, yellow circles), 44% to group II (90e180 m, blue circle) and 46%Zone; COT, Continental Ocean Transition.

A. Rajan et al. / Marine and Petroleum Geology 32 (2012) 36e4938

break (e.g. Laberg and Vorren, 1995). Trough mouth fans on thecontinental slope are the result of these glacier driven sedimentaryprocesses (e.g. Sarkar et al., 2011; Vorren et al., 1998).

The western Svalbard shelf reflects the glacial sedimentaryevolution of the margin during the Plio-Pleistocene (Solheim et al.,1998). Glacigenic sediments primarily built the outer shelf ofwestern Svalbard. Highly compacted till deposits are generallyabsent within the troughs on the inner shelf but they are abundanton the shelf bank areas (Andersson et al., 2000). A previous high-resolution study by Landvik et al. (2005) from the area west ofPrins Karls Forland shows a glacigenic unit crossing the ForlandetMoraine Complex and a pronounced ridge in the western part,which was interpreted as a terminal moraine. The lithostratigraphycomprises a matrix-supported diamiction unit interpreted assubglacial till deposited by a grounded glacier (Landvik et al., 2005).

In addition to characteristic glacial sediment formations such asprograding glacigenic sequences, till sediments on shelf banks anddebris flows at the continental slope the study area is highlyinfluenced by the West Spitsbergen Current (WSC), a shallownorthward flowing warm water branch of the North AtlanticCurrent (e.g. Slubowska-Woldengen et al., 2007; Spielhagen et al.,2011), which transports heat into the Arctic Ocean. The north-ward flowing warm current is responsible for maintaining meanannual temperatures around 10 �C higher than locations at similarlatitudes, e.g. Arctic Canada (e.g. Slubowska-Woldengen et al.,2007).

Westbrook et al. (2009) suggest that the warming of the WSCduring the last 30 years reduced the size of the gas hydrate stabilityzone (GHSZ) at the uppermost continental slope and outer shelfarea, leading to the dissociation of gas hydrate and the annualrelease ofw900 kg of methane per meter length of the margin. It is

Figure 2. CTD station integrated into 2D seismic profile line 009 illustrate the presence of thecold current, the East Spitsbergen Current (ESC) flows also northward along the coast anstratigraphic sequences, the glacial debris flow (GDF) deposits, acoustic blanking, acousticmodeled and projected gas hydrate stability zone (GHSZ).

notable that the area was subjected to a number of oceanographictemperature increases and decreases during the past 17,500 years,as reported by shallow sediment core analysis (Slubowska-Woldengen et al., 2007).

3. Data and methods

3.1. Multi-beam bathymetry data for seabed mapping

Swath bathymetry data (Fig. 1b) were acquired using theKongsberg Simrad EM-300 system (R/V Jan Mayen, July 2009 andprevious RV JanMayen cruises; Hustoft et al., 2009). Themultibeamsystem operates at a sonar frequency of 30 kHz with an angularcoverage sector of up to 150�. Multibeam datawere processed usingNeptune software. Gridding and imaging of the acquired data weredone using GMT (Wessel and Smith, 1998) and data visualizationwas performed in IVS Fledermaus. Sound velocity profiles forcalibration of beams in the water columnwere extracted from nineCTD stations acquired during the survey and processed using “SBEData processing”.

3.2. CTD stations for oceanography

Nine CTD casts at 164 m, 193 m, 235 m, 365 m, 423 m, 445 m,520 m, 616 m and 800 m water depths from Forlandbankentowards the continental slope provided temperature and salinityversus depth measurements (Fig. 2). Furthermore, the use of AWI’sOcean Data View software (Schlitzer, 2008) allowed to compile andintegrate the CTD data into transects for identifying the watermasses in the study area (Fig. 2).

warm northward flowing Atlantic water marked as West Spitsbergen Current (WSC). Ad lies east of the WSC. The 2D seismic profile 009 indicates the sub-seabed seismicpipe and enhanced seismic reflection amplitudes (ER). The stippled line shows the

A. Rajan et al. / Marine and Petroleum Geology 32 (2012) 36e49 39

3.3. Echo sounder system (EK60) for acoustic flare (bubble)detection in the water column

The Simrad EK60 scientific echo sounder system is designed forfishery research, yet it represents a useful instrument formeasuring ocean water density contrasts including objects in thewater. The system installed on R/V Jan Mayen consists of a splitbeam transducer at three different operating frequencies: 18, 38and 120 kHz. During the survey the equipment was used to detectgas bubble streams in form of acoustic flares using 38 kHz and18 kHz frequencies. An acoustic flare indicates release of free gas inform of bubbles from the seafloor and it is a first order remote prooffor active gas seepage at the respective site (Greinert, 2008). Thebasic principle for detecting bubbles with an echo sounder is thestrong backscattering of the transmitted pressure wave at distinctimpedance contrasts between water and gas bubbles. It createscharacteristic flare-shaped backscatter features in echograms(Greinert et al., 2006). The visualization of the recovered data wasdone with the Echoview software and mapped with GMT (Wesseland Smith, 1998). For a better overview of the different seeps,three groups of flares have been used in a map as a clusteraccording to their flare height. Group 1 contains all flares which are270e180 m high, group 2 comprises flares from 180 to 90 m highand finally group 3 are the smallest flares with a height range from90 to 20 m.

3.4. High-resolution 3D-PCable seismic data for imaging sub-seabed/seabed seismic architecture and attributes

The University of Tromsø’s high-resolution 3D seismic acquisi-tion system (P-Cable) consists of up to 20 streamers, each 25m longwith 8 recording channels (Petersen et al., 2010). In 2009, thesurvey was carried out with 12 streamers together with a Granzowhigh pressure (140 bar) compressor that provided the pressure forone GI gun with a total volume of 150 in3 (G ¼ 45 in3 and I ¼ 105in3) shot at a rate of 5 s. The 3D seismic cube covers an area of6.1 km � 1.41 km (Fig. 1c).

Data processing was done using the RadExPro 3.96 softwarepackage and consisted of navigational correction, binning, staticand tidal correction, bandpass filtering (35e300 Hz), amplitudecorrection, trace editing, normal move out, 3D stack and 3D stoltmigration (velocity 1500 m/s). Two seismic attribute maps (RMSamplitude and Variance attribute) were produced using Petrel 2010interpretation software. The RMS amplitudemap indicates elevatedgas concentrations or geological features with increased amplituderesponse. The Variance attribute map measures the dissimilaritybetween adjacent seismic traces, and can enhance structuralfeatures such as faults but it also can enhance for example depo-sitional/erosional features such as channels (Bemmel et al., 2000;Chopra and Marfurt, 2007).

3.5. 2D seismic data for imaging sub-seabed structures

Four regional 2D seismic lines are used to complement the 3Dseismic data (Fig. 1b). 2D single channel reflection seismic dataacross the margin off Prins Karls Forland were acquired using thesame GI gun system and a single channel streamer. A DelphSeismicsystemwas used to record the 2D single channel seismic data. Dataprocessing included bandpass filtering (35e300 Hz), trace editing,muting and stolt migration (velocity 1500 m/s).

3.6. Modeling of the base of the gas hydrate stability zone (BGHSZ)

In this study we modeled the GHSZ thickness using site specificparameters and the methods explained by Sloan (1990). The

specific parameters of the study area encompass water depth(pressure) derived from the bathymetry, bottomwater temperaturederived from CTD measurements (Figs. 1 and 2), geothermalgradients derived from Vanneste et al. (2005) and observed BSRdepth, pore water salinity from fresh to seawater and gas compo-sitions from pure methane to heavier carbon contributions. Theresults demonstrate how changes in the parameters impact on thecalculated thickness of the GHSZ.

The Svalbard-Barents Sea seabed lies close to the gas hydratestability zone which makes it particularly sensitive to changes insea level and bottom water temperature (e.g. Chand et al., 2008;Mienert and Posewang, 1999; Mienert et al., 2005). Often, oneassumes a constant geothermal gradient for a region, and the gashydrate stability modeling is carried out accordingly. However,observations from the Gulf of Mexico (Ruppel et al., 2005) and theBarents Sea (Bugge et al., 2002) suggest that the geothermalgradient can be highly variable due to changes in the underlyingsediment formations.

The average geothermal gradient observed in NW-Svalbardregion is w70 �C/km (e.g. Hustoft et al., 2009; Vanneste et al.,2005). We estimated the geothermal gradient from a clearlyobservable BSR and found the geothermal gradient to be 55 �C/kmwith an inferred decrease towards the upper slope to 44 �C/km.

Different water masses control the bottom water temperaturesat the study area showing a major boundary between cold waterfrom the eastern Svalbard margin occupying the shelf and thewarm water from the Atlantic influencing the upper water columnacross the continental slope (Fig. 2). The bottomwater temperaturein the study area varies from 1 �C at the lower continental slope atw700 m water depth to 3.5 �C close to the shelf edge at w250 mwater depth (Fig. 2). We choose an average bottom water temper-ature of 3.3 �C for the gas hydrate stability modeling since most ofthe seabed is bathed in such temperature.

The calculations of the gas hydrate stability curve (Sloan, 1990)and the implementation for areal prediction (Chand et al., 2008)allowed estimating BSR depths for different gas composition andsalinities. Carbon isotope studies from the NW-Svalbard regionshow that higher order hydrocarbons in bottom waters occurpreferably in the vicinity of active seepage sites on the shelf (Knieset al., 2004). Hence, wemodeled the GHSZ within the study area forboth pure methane structure I hydrates and structure II hydrateswith small amounts (4%) of higher order hydrocarbon (propane,ethane) (Appendix, Fig A.1). To estimate the BSR depth from seismicdata we used an average sediment velocity of 1630 m/s (Vannesteet al., 2005) from the seabed to the observed BSR and an averagevelocity of 1450 m/s in the water column above. The modeledBGHSZ was plotted on the 2D and high-resolution 3D seismicdataset using the geothermal gradient (55 �C/km) and BWT (3.3 �C).

4. Results and observations

4.1. Stratigraphic interpretation and seismic anomalies

4.1.1. High-resolution 3Dseismic reflection dataIn general, 3D seismic data (Fig. 3a) show increasing acoustic

penetrations from the shelf to the upper slope due to decreasingacoustic attenuation. The decreases in amplitude may be partly dueto the less compacted sediments further downslope (i.e. at distalareas from the former ice sheet margin) and comparatively lowergas content in the pore space of sediments. At the Svalbard shelfseaward dipping reflectors are visible (Fig. 3a). They mark seismicsequences and can be divided into six units (Fig. 3a) based onhigher seismic amplitudes at the sequence boundaries, the seismiccharacter such as internal seismic reflection pattern, and the overallshape of the adjacent units so that they become distinctive

Figure 3. (a) High-resolution 3D P-Cable seismic data (inline 56) shows prograding glacigenic sequences at the shelf, a major GDF, enhance seismic reflection amplitude (ER)beneath the GDF and a phase reversal within the well-stratified seismic reflector sequence at the continental slope. The phase reversal occurs at 0.67 TWT. The interpretedsedimentary sequences I (youngest)eVI (oldest) are marked along the seismic profile. The stippled line shows the modeled GHSZ. The green box in the seismic section indicates thehigh RMS amplitude area in the region of phase reversal (black rectangular box). (b) The RMS amplitude map extracted from a search window of 10 ms at the area of enhancedreflections and a phase reversal.

A. Rajan et al. / Marine and Petroleum Geology 32 (2012) 36e4940

(e.g. Miall, 2010; Sarkar et al., 2011). They are interpreted to beprograding glacigenic sequences, deposited in front of an advancingice sheet (e.g. Vanneste et al., 2007). The six prograding glacigenicsequences I (youngest) to VI (oldest) (Fig. 3a) suggest major glacieradvances at the Svalbard continental margin at peak glaciationsduring the last 3.2 million years (Dowdeswell and Elverhøi, 2002;Elverhøi et al., 1995; Landvik et al., 1998; Sarkar et al., 2011).Sedimentary sequences IV and VI may correspond to ages

w1.5e1.0Ma andw2.8Ma respectively (e.g. Sarkar et al., 2011). Thesequences indicate significant shelf progradations (Fig. 3a) buildingthe continental margin westward by several tens of kilometers.Downslope, the sequences are thinning, poorly stratified anddiscontinuous with semi-transparent to opaque reflections(Fig. 3a). Chaotic, lens-shaped seismic features are noticeable in theprograding glacigenic sequences II, V and VI, which show internallychaotic reflection pattern enveloped by high amplitude reflections

A. Rajan et al. / Marine and Petroleum Geology 32 (2012) 36e49 41

(Fig. 3a). The wedges are thinning towards the west (continentalslope) (Fig. 3a). Enhanced seismic reflection anomalies (ER) areobserved at the upper continental slope below the base of theprograding glacigenic sequence I encompassing the GDF (Fig. 3a).Glacigenic sequence I consists mainly of a major GDF deposit that isacoustically defined by a chaotic and partly transparent seismiccharacter (Hjelstuen et al., 2004). Enhanced seismic reflectionanomalies and a phase reversal at 0.67 s TWTover a lateral distanceof w1 km suggest a potential gas accumulation area (Fig. 3a). The3D seismic profile does not show evidence for a BSR within the GDFdeposits or beneath them (Fig. 3a) indicating that gas hydratemight be lacking in the area covered by the 3D seismic data at theupper slope of the W-Svalbard continental margin.

The RMS attribute map (Fig. 3b) derived from a time window of10 ms centered around the horizon (phase reversal marked as blueline in Fig. 3a) shows the distribution of the high amplitudeanomalies in the region of the polarity reversal indicating gasaccumulation at the lower continental slope.

The volume (a short window of 15ms) attribute map of varianceat a depth of w750 ms TWT shows a number of north-southtrending lineations (Fig. 4). In accordance with interpretationselsewhere (Chopra and Marfurt, 2007) these spatially coherentlineations indicate the presence of small-offset faults. The faultsinterrupt the well-stratified seismic reflection pattern in the uppercontinental slope, though they are more difficult to identify on the3D seismic profile in Figure 4. The faults occur at depths below thegas hydrate stability zone (GHSZ), but the data does not allow us todetermine the depth to which they extend.

The RMS amplitude map of the seabed extracted from a timewindow of 10 ms clearly shows high amplitude anomalies sug-gesting gas accumulation and migration at the upper slope andshelf area (Fig. 5a). The amplitudes relate to the numerousenhanced reflections (ER) and the phase reversal observed in the

Figure 4. High-resolution 3D P-Cable (inline 96) shows prograding glacigenic sequences atstratified seismic reflector sequences in the continental slope. The volume (a short windowoffsets of the faults below the theoretical GHSZ.

high-resolution 3D seismic data (Fig. 3a). Negative relief features,interpreted as pockmarks are also observed in the RMS amplitudemap for the seabed (Fig. 5a). The variance surfacemap of the seabedcalculated from a time window of 10 ms clearly indicate the cropout zone of the glacigenic sequence I boundary (Fig. 5b), whichcoincides well with the flare locationsmapped on the shelf (Fig.1b).Most of the flares are concentrated along the glacigenic sequence Ioutcrop zone whereas only a few flares are scattered within closevicinity.

4.1.2. 2D seismic reflection data2D seismic data (Fig. 1b) encompass part of the shelf that lies

immediately south of the Kongsfjorden cross-shelf trough, the shelfbreak and part of the upper continental slope (Fig. 1a and b). All 2Dseismic profiles (line-002, line-008, line-009 and line-011) showacoustic transparent and/or chaotic sub-seabed reflections from theKongsfjorden cross-shelf trough over a distance of w4 km beyondthe shelf break followed by well-stratified seismic sequencesdownslope (Figs. 2, 6e8). The modeled GHSZ depth is projected onthe seismic data (Figs. 2, 7 and 8).

2D seismic line-002 (see Fig. 1b for location) runs from north tosouth crossing a shelf bank (Fig. 6). High amplitude reflections arevisible at w50 ms TWT below the seabed indicating the possiblepresence of shallow gas or sediment facies changes causingenhanced acoustic impedance contrasts. Below the enhancedreflections a distinct acoustic turbidity zone exists (Fig. 6).

2D seismic line-008 (see Fig. 1b for location) has a length ofapprox. 32 km (Fig. 7). The subsurface strata shows fairly well-stratified units beyond the shelf break (Fig. 7). At the western endof the seismic line a high amplitude reflection at w240 ms TWTbelow the seafloor cuts across the strata and mimics the seafloor.This high amplitude reflection matches well with the predicteddepth of the base of the GHSZ. Beneath the BGHSZ indicated by

the shelf, enhanced seismic reflections beneath the GDF and fault zones interruptingof 15 ms) attribute of variance map at a depth of w750 ms TWT highlights the small

Figure 5. (a) High-resolution 3D P-Cable RMS amplitude map calculated from a search window of 10 ms integrated with mapped flare locations and pockmarks. High amplitudeanomalies indicate shallow gas accumulations. Note that the pockmarks are not associated with flares. Within the pockmark field no evidence from active fluid expulsion wasdetected during the time of the survey (July 2009). (b) Seafloor variance map extracted from a search window of 10 ms show the crop out zone of the glacigenic sequence Iboundary. The color dots indicate the flare locations based on 2009 survey.

A. Rajan et al. / Marine and Petroleum Geology 32 (2012) 36e4942

a BSR, a number of enhanced reflections (ER) are visible indicatingthe presence of appreciable amounts of gas in the sub-seabed(Fig. 7). The BSR cannot be traced further upslope towards the shelf.

A zone of acoustic masking is visible beneath the seabed atapprox. 300ms TWT (Fig. 7). It is either related to the attenuation ofseismic energy caused by the overlying ER or the occurrence of gasin the masking zone. The masking zone shows individual occur-rences of lens-shaped ER with reversed polarity compared to theseafloor (Fig. 7). The observations suggest that free gas is trappedlocally beneath the BSR but also further down at greater depth inareas outside the GHSZ and upper slopewhere a BSR is absent. Few,more than 260 ms TWT long vertical and seismically transparentfeatures similar to pipes or chimneys exist beneath the seafloor.They are not connected to pockmarks at the seabed (Fig. 7). The

transparent pipe structures terminate beneath the seafloor some-where within the GHSZ.

Seismic profile line-009 (see Fig. 1b for location) (Fig. 2) runsalong the centre of the 3D seismic data. The seismic profile ischaracterized by well-stratified continuous reflections downslope(westward) and prograding glacigenic sequences in the shelf area(Fig. 2). Polarity-reversed and enhanced seismic reflections withinan otherwise low amplitude zone approx. 200 mbsf suggest thewidespread occurrence of gas (Fig. 2). The seismic data alsoconfirms the presence of a lens-shaped acoustic transparent unit atthe seafloor between 5 and 14 km comparable to a low permeableglacigenic debris flow (GDF lobe) (Laberg and Vorren, 1995; Nielsenet al., 2005; Ottesen et al., 2008; Vorren and Laberg, 1997)extending from the shelf break to the upper continental slope

Figure 6. 2-D seismic profile line 002 runs south-north (for location see Fig. 1aeb) and shows seismic features such as a dipping reflector from a shallow bank towards a depressionand enhanced reflections (ER) in the shallow sub-seabed. The location of acoustic flares is indicated by arrows.

A. Rajan et al. / Marine and Petroleum Geology 32 (2012) 36e49 43

(Fig. 2). No BSR was observed on this seismic line. However, thepredicted base of the GHSZ lies within the GDF and theoreticallyoutcrops at 370 m water depth approx. at the foot of the uppercontinental slope (Fig. 2). Faults marked by small offsets of highamplitude reflections are clearly visible at the western end of theline (Fig. 2). This observation could indicate migration of fluidsalong faults similar to previous observations from this area (Hustoftet al., 2009; Vanneste et al., 2005).

2D seismic profile line-011 (Fig. 8, see Fig. 1b for location) islocated w7 km south of the 3D seismic area and also confirms thepresence of a prograding glacigenic sequence comprising GDFdeposits in the upper slope area. At the lower slope area, a prom-inent cross-cutting BSR exists above an area of enhanced seismicamplitudes. The BSR can be traced along the slope at w190 mbsfbetween SP 30 and 100 (Fig. 8). Enhanced seismic amplitudesreflections beneath the BSR are probably due to the accumulation ofgas while the reflection amplitude just above the BSR is subdueddue to gas hydrates (Hustoft et al., 2009). No distinct BSR isobserved towards the shelf but several acoustic anomalies (ER)inferred to be caused by shallow gas accumulations can be noticed(Fig. 8). Close to the shelf edge very chaotic seismic reflectionpatterns occur beneath the GDF deposit and the prograding glacialsequence I to a water depth of approx. 400 m (Fig. 8). This zone ofbrightened and dimmed amplitude anomalies resemble those ofvery wide hydrocarbon leakage features (Løseth et al., 2009) andcould be a gas cloud.

4.2. Vertical acoustic pipes and sub-seabed acoustic anomalies

Geological structureswithin the area covered by 2D seismic dataprovide evidence for vertical focused fluid flow. Acoustic pipe andacoustic turbidity areas suggest various forms of vertical gasmigration. The presence of gas is characterized by amplitudedimming, amplitude enhancement or complete disruption ofreflection as seismic energy gets scattered by gas (Gay et al., 2007;Judd and Hovland, 1992). The most prominent features are thevertical and narrow zones of acoustic wipe out zones with upward

bending marginal reflections (Fig. 7). These structures, referred toas acoustic pipes or gas chimneys, are often connected to pock-marks at the present day seafloor (e.g. Hovland and Judd, 1988;Plaza-Faverola et al., 2010). The high-resolution bathymetry from3D seismic survey shows clearly round to oval depressions (Fig. 9aand b) at the seafloor on the shelf, which are also observed in the2D seismic profiles on the slope (Figs. 2 and 7). The largest prom-inent pipe structure within the 3D seismic dataset (Inline-163) onthe shelf (Fig. 9a) can be followed down to 0.5 s TWT. This structureshows only high amplitudes directly beneath the seafloor. Thediameters of the pipes on the slope vary between 20 and 450 m(Fig. 7). Some pipe structures indicate seismic reflection pull-up byas much as 8e10 ms (Fig. 7) towards their central zone, whichprogressively decreases with decreasing depth. The depth deter-mination of the base of the acoustic pipe structures are difficult dueto poor signal strength with increasing depths. However, theprominent pipe structures in the 2D seismic profiles are traceabledown to 1.5 s TWT (Fig. 7).

4.3. BGHSZ (BSR) depth prediction and observed BSR depthdistribution

The BGHSZ was modeled for the available 2D and high-resolution 3D dataset using a bottom water temperature of 3.3 �C,a geothermal gradient of 55 �C/km and the pressure (water depth).The predicted BGHSZ for a composition of 100% methane is shownin Figures 2, 7 and 8 and the predicted BSR for 96% methane, 3%ethane, 1% propane is shown in the appendix (Fig. A.1). Changes inthe salinity from fresh water to sea water resulted in a shoaling ofthe BGHSZ while a contribution of ethane and propane causeda deepening of the BGHSZ (Fig A.1). Noteworthy, the BHGSZ for puremethane and fresh water matches well with the present dayobserved BSR position. The modeled GHSZ varies laterally inthickness fromw200m below the seafloor at the lower continentalslope at w800 m water depth and gradually decreases to a theo-retical pinch out zone towards the uppermost continental slope at370 m water depth (Figs. 2, 7 and 8). The GHSZ extends from the

Figure 7. 2-D seismic profile line 008 runs east-west (for location see Fig. 1) indicating acoustic pipes, seismic amplitude reversal, a continuous and limit BSR extension at the slope,a gas cloud at the most upper slope and outer shelf area, and an enhanced reflection (ER) in the sub-seabed at w0.25 TWT. The acoustic pipes are marked by pull up at their uppertermination in the sub-seabed. The predicted BGHSZ fits the observed BSR location.

A. Rajan et al. / Marine and Petroleum Geology 32 (2012) 36e4944

Vestnesa Ridge, a major sediment drift on young oceanic crust(Fig. 1, Hustoft et al., 2009) to the upper continental slope of theNW-Svalbard continental margin resting on continental crust,thereby linking the gas hydrate field to an area of active fluid escapefeatures on the outer shelf. No BSR has been observed near the shelfor upper slope at the theoretical BGHSZ pinch out zone.

4.4. Seabed fluid flow expressions

The high-resolution bathymetry covers the shelf bank area,28 km west of Prins-Karls Forland (Fig. 1b). Echosounder informa-tion reveals the presence of flares in the water column associatedwith a vigorous release of gas from sub-seafloor sediments at theouter shelf high (Figs. 5a and 9a-c). However, the flare sites do nothave any seafloor expressions (pockmarks) normally associatedwith the focused expulsion of gas. Approximately 3 km eastward ofthe flare area, themid-shelf region shows round to oval depressions

interpreted as pockmarks (Fig. 9a and b). Active gas release fromthese pockmarks is lacking. However, their formation may becaused by past seepage of thermogenic gas as suggested fromobservations from NW-Svalbard fjords (e.g. Forwick et al., 2009).The pockmarks on the shelf are w20e30 m deep and w80e100 mwide (Fig. 9b). Their occurrence on the shelf bank suggests that thisarea must have experienced active fluid expulsion after the lastglaciation.

The flares are the ultimate proof for active fluid venting (Fig. 9c).Figure 1b shows clusters of flares in which more than 220 flaresoccur (Figs. 5a and 9a-c). The flares are concentrated on the NW-Svalbard continental margin at a water depth of w253 m. Thesize or height of the flares varies from 20 m to 270 m where thehighest flares may reach the sea surface. The highest concentrationof flares lies in the northern area at the outer shelf, clearly outsideand east of the pinch out of the projected gas hydrate stability zone(Figs. 2, 6 and 8).

Figure 8. 2-D seismic profile line 0011 runs westeeast (for location see Fig. 1) and shows high amplitude reflections beneath the BSR and predicted BGHSZ. The acoustically maskedarea indicates a unique gas cloud at the most upper slope and outer shelf area between the theoretical GHSZ pinch out and the location of the observed acoustic flare area.

A. Rajan et al. / Marine and Petroleum Geology 32 (2012) 36e49 45

4.5. Ocean water masses across the shelf-slope transect

The temperature transect (Fig. 2) illustrates the presence ofwarm and saline Atlantic water belonging to WSC at the shelf edge.This water mass is particularly evident in the uppermost w300 mof the water column at the outer shelf areawhile cold water massesof <1 �C exists beloww600 mwater depth at the continental slope(Fig. 2). The oceanographic observations together with seismicevidence for the existence of gas hydrate along the continentalslope suggest that the NW-Svalbard margin oceanic gas hydratefields may be impacted by deep and to a lesser extent by increasesin shallow ocean warming.

5. Discussion

High concentrations of methane in the water column offshoreSvalbard are most probably caused by the release of sub-seabedmethane at seepage sites on the SW-Svalbard shelf (Damm et al.,2005). In the Svalbard shelf and slope area deep-seated hydro-carbon sources (Knies and Mann, 2002), shallow gas hydrates(Hustoft et al., 2009; Posewang andMienert, 1999; Vogt et al., 1994)and natural seabed seeps (Knies et al., 2004; Solheim and ElverhøI,1985; Vogt et al., 1999; Westbrook et al., 2009) all may contributevia natural seep processes at still unknown rates over time toelevated dissolved methane concentrations in the water column.Methane gas released from sub-seabed environments as a result ofdestabilization of gas hydrates due to ocean warming in the last10e30 years (Westbrook et al., 2009) is considered as one of thepossible processes that may contribute methane to the ocean.However, warmwatermasses need to be inmore or less continuouscontact with the seabed in order to send awarming pulse to the gashydrates causing gas hydrate dissociation. Our seismic data provide

evidence for the presence of gas within the upper slope and shelfsediments but not for the occurrence of gas hydrate. Amongst themost prominent indicators for active gas release are gas flares in thewater column. While high amplitude seismic anomalies and a BSRexist at the lower slope, vertical acoustic pipes and pockmarksoccur upslope. The locations of fluid flow indicators point to verycomplex gas migration systems that can be split up into the lowerand upper continental slope zone, where gas hydrate can exists,and into the shelf break zone with water depth too shallow for gas(methane) hydrate to exist. This following discussion focuses onelucidating the pathways and mechanisms by which gas migrates,and on evaluating the source of the gas expelled from the seabed ofthe shelf.

5.1. Fluid migration pathways

Most evidence for fluid and gas accumulations stems from highamplitude anomalies in the mostly well-stratified upper slope.Some of the gas accumulations occur beneath a BSR at the lowerslope in a very similar way as observed at other locations of theNW-Svalbard area (Hustoft et al., 2009; Vanneste et al., 2005).However, here a BSR can only be observed inwater depth exceedingapprox. 800 m (Figs. 7 and 8), which is w20e25 km westward ofthe shelf break. The theoretical GHSZ thins eastward towards theshelf but without any evidence for a BSR. Only patches of enhancedseismic reflections exist beneath the base of the predicted GHSZ(Fig. 8) at a depth of w300 mbsf. They may indicate upslope gasmigration along strata. Strata and sub-seabed gradient controlledfluid migration (Weibull et al., 2010) apparently deviates gas awayfrom the thinning GHSZ instead of leading the gas into the BGHSZas is the case for the area, where a BSR exists (Figs. 7 and 8). Thismechanism may allow gas to bypass to the shelf break without

Figure 9. (a) High-resolution 3D P-Cable (inline 163) shows a prominent pipe structure, pockmarks on the shelf, prograding glacigenic sequences at shelf and GDF deposit on theslope. (b) Swath bathymetry shows the pockmarks field. The cross profile (yellow line) depicts the depth (w10 m) and the width (w50 m) of the pockmarks. (c) EK 60 acousticrecord shows examples of flares observed in the flare field areas west of the pockmarks.

A. Rajan et al. / Marine and Petroleum Geology 32 (2012) 36e4946

entering the GHSZ. As a consequence, the area beneath the shelfbreak is a major gas transport and gas release zone, which is sup-ported by the seismic data, which show the strongest enhancedreflections stacked over up to 200 ms TWT (Figs. 3a-b and 4)interpreted to be a gas cloud (Figs. 7 and 8).

Acoustic pipe structures are commonly associated with focusedmigration of fluids (e.g. Hustoft et al., 2007). Acoustic pipes in theseismic profiles line-008 and line 009 (Figs. 2 and 7) show pull-upeffects. These may be associated with locally high velocity zonescaused by either authigenic carbonate build ups or gas hydrateplugs (e.g. Hustoft et al., 2007; Plaza-Faverola et al., 2010) ora combination of both. However, the pipe structures terminatebeneath the seafloor, which suggests that they are either a paleo-

feature or that there was not enough overpressure build up forgas to reach the seafloor.

Other possible pathways for gas migration are through faultsthat have been detected in the 3D seismic data on the upper slope(Fig. 4). These small-offset faults are very typical for this setting andhave been associated with fluid flow (Hustoft et al., 2009; Vannesteet al., 2005). The origin of the faulting is difficult to identify basedon the available data but it might be related to the stretching of thecontinental crust since the opening of the northern North Atlanticalong the Hornsund fault complex (Fig. 1a) (Faleide et al., 1991).

3D and 2D seismic data provide clear evidence for well-developed prograding glacigenic sequences and encompassingGDF deposits that determine the sedimentary architecture of the

Figure 10. Conceptual model for geologically controlled gas migration pathways and gas release at the seabed into the water column. Major vertical and lateral fluid flow pathwaysare indicated by arrows. The stippled and the solid line show the BGHSZ before and after ocean warming respectively.

A. Rajan et al. / Marine and Petroleum Geology 32 (2012) 36e49 47

outer shelf and uppermost slope area (Figs. 2 and 3a). GDF depositsconcentrate at slope areas (Figs. 2, 3a, 4 and 8) and their distribu-tion varies along the uppermost slope in the prograding glacigenicsequence I. On some seismic lines the BGHSZ falls within the GDF(Figs. 2e4) while on others it seems to connect to the base of theGDF (Fig. 8). The physical properties of GDFs (Sættem et al., 1992)are not seen to be conducive for gas hydrate growth and theirreduced permeability may provide an effective trapping mecha-nism for gas (Bünz et al., 2003). The prograding glacigenicsequences control fluid migration pathways (Figs. 3a, 4 and 8). Thelocation, where the prograding glacigenic sequence I intersects theseafloor coincides well with the location of the gas flares on theshelf (Fig. 5b), clearly showing that gas mostly migrates along itsbase (Figs. 3a, 4 and 9a).

Acoustic flares on the outer shelf appear not to be connected topockmarks at the seabed (Fig. 9a, b and c). In contrast, the pock-mark field at the inner shelf shows no flares. Evidence for gasconduits below the widespread pockmark field is relatively scarce(Figs. 5 and 9a). The fluid escape features on the seabed within thezone of acoustic flares are thought to be in their infancy becausethere are no recognizable pockmark expressions in the very high-resolution multibeam data. Seafloor expressions of pockmarksmay develop if both the fluid flow increases and its activity occurover longer time periods.

The location of the pockmarks further eastward from the gasflares is difficult to explain based on our seismic data. One possiblemechanism obviously is that gas hasmigrated along the boundariesof some of the older prograding glacigenic sequences (Fig. 9a).Another explanation for the origin of the pockmark field could bethat they were fed from a yet unknown source area. The fact thatgas flares are absent at the pockmark field suggests that activity hasbeen significantly reduced or has ceased. This might be furthercorroborated by the absence of seismic amplitude anomalies in thesub-seabed indicating the absence of appreciable amounts of gas.

The high-resolution 3D seismic data image one distinct acousticpipe in shelf sediments that connects to one of the pockmarksobserved in the swath bathymetry and the 3D seismic data (Fig. 9a).The seismic profile shows a high amplitude anomaly directlybeneath the seafloor indicating trapped shallow gas accumulationsand possibly a plugged pockmark (Fig. 9a). Since the pockmark liesoutside the GHSZ, plugging by gas hydrates has to be ruled out butauthigenic carbonate formation may be a possible long-termplugging mechanism (Pierre et al., 2010; Zhu et al., 2003).

5.2. Origin of fluids

Whether the source of the gas expelled at the seafloor on theuppermost slope and shelf break area is from deep-seated hydro-carbon systems or dissociating gas hydrates is obviously of envi-ronmental importance given the ongoing climate change debate(Kerr, 2010). The seismic observations presented herein supporta deep-seated source as there are many locations where gas accu-mulations occur at significant depth beneath and outside theBGHSZ (Figs. 2, 7 and 8). This is in good agreement with observa-tions by Knies et al. (2004), which showed that traces of bothbiogenic and thermogenic methane gas exists in the sediments ofthe Kongsfjorden shelf. Also, high carbon content measured inMiocene deposits of the Fram Strait further north suggest a highpotential for hydrocarbon source rocks along the continentalmargin of NW Svalbard (Knies and Mann, 2002; Knies et al., 2004).

However, we cannot rule out that some of the gas that isexpelled comes from dissociating gas hydrate, though, a clearevidence for gas hydrate is lacking in sediments of the uppermostslope. But it is certainly possible that some of the gas that ismigrating from depth enters the GHSZ in the vicinity of its outcropzone. There, it might be temporally stored as gas hydrates andreleased during changes in the stability field as e.g. changes inbottom-water temperatures due to seasonal changes or due to

A. Rajan et al. / Marine and Petroleum Geology 32 (2012) 36e4948

long-lasting temperature changes as a consequence of globalwarming. The released gas could then migrate further upslopetrapped beneath a GDF deposit until it can be expelled through oneof the many gas expulsion sites.

5.3. Conceptual fluid flow

Based on our seismic observations, we develop a conceptualmodel that intends to explain and summarize the geologically-controlled gas migration system on the uppermost continentalslope and shelf on the NW-Svalbard margin (Fig. 10). In theuppermost 300e400 mbsf seismic evidence for migration of gas-charged fluids along strata greatly prevails over focused verticalfluid flow migration. Typical expressions of vertical fluid flow suchas pipes or chimneys are scarce. Gas migrates along permeablestrata upslope thereby eluding the theoretical GHSZ, which pinchesout at the uppermost part of the slope at 370mwater depth (Figs. 2,7 and 8). A clear seismic marker for the BGHSZ such as a BSR is onlyevident at the lower continental slope in water depth exceeding800 m, where the GHSZ becomes w190 m thick (Fig. 8). Gas seemsto be trapped beneath the prograding glacigenic sequences Iencompassing GDF deposits, which generally have greatly reducedpermeability and which also inhibit gas hydrate and BSR formation(Bünz et al., 2003). Fluids continue to migrate upslope alongpermeable layers at the base of the prograding glacigenic sequenceI until the layers intercept with the seafloor and fluids are expelledon the shelf.

Studies of glacial landforms at continental margins of the NordicSea by Vorren et al. (1989) demonstrated that during the timewhenglaciers reach the shelf edge high sedimentation rates prevail dueto glacigenic sediment input in form of debris flows. If sedimen-tation rates are high enough, the speed of sediment burial exceedstheir rate of compaction. Thus if the drainage of gas is hindered bylow permeability sediments on top, overpressures may develop(e.g. Hustoft et al., 2009). The area beneath the upper slope maytherefore still be overpressurized. If the pressure cannot be reducedand increases further, the overpressure build up within sedimentlayers may initiate lateral fluid migration and/or vertical focusedfluid migration through micro fracturing (Judd and Hovland, 2007).When fluids rise, they experience a decrease in pressure. This willallow the gas to come out of solution, resulting in the formation ofbubbles. The free bubbles will expand and reduce the bulk densityof the material even further. Once initiated, gas migration may leadto a self perpetuating situation wherein gas bubbles develop, andtherefore increase the buoyancy and gasmaymigrate upward alongthe area beneath the upper slope (Judd and Hovland, 2007).

6. Conclusion

The integration of high-resolution 3D and 2D seismic datareveals the presence of active fluid escape from sediments of theW-Svalbard continental margin. High concentrations of flares(>220) at the shelf break area provide a first order proof for activefluid seepage. The westward prograding shelf consists of glacigenicsequences and a glacial debris flow deposit (GDF), which arespatially confined. Their geological setting controls gas migrationpathways and hence, the location of gas leakage zones at theseafloor on the shelf.

Gas migration occurs towards the upper slope along strata,whereas vertical focused fluid flow indicators are scarce. Verticalmigration of thermogenic methane from deep hydrocarbon reser-voirs may be caused by diffusive processes.

Seismic evidence for gas hydrates exists only at the lowercontinental slope in water depth exceeding 800 m. The hypothesis

that gas hydrates may also occur in close vicinity to the theoreticaloutcrop zone of the GHSZ has still to be proven.

Widespread occurrence of gas release from the shelf area maybe driven by free gas supply from lower continental slope regions.

Acknowledgements

The authors acknowledge the support by Deco Geophysical,Russia for the seismic processing software RadExpro 3.9. We arethankful to Schlumberger for providing the Petrel softwarepackage. Dr. Shyam Chand of NGU is thanked for his assistancewiththe modeling of the base of gas hydrate stability zone. We aregrateful to the captain, crew and the scientific participants onboardRV Jan Mayen for their help during seismic acquisition. We arethankful to the University of Tromsø for financing the researchproject under the PhD program, and to Prof. Christian Berndt andan unknown reviewer for providing helpful comments thatimproved the quality of the paper. Part of this research is a contri-bution to the NFR funded Geosystem 3-D seismic imaging (G3)project (195379) and to the HERMIONE EU funded project(226354).

Appendix. Supplementary data

Supplementary data related to this article can be found online atdoi:10.1016/j.marpetgeo.2011.12.008.

References

Andersson, T., Forman, S.L., Ingolfsson, O., Manley, W.F., 2000. Stratigraphic andmorphologic constraints on the Weichselian glacial history of northern PrinsKarls Forland, western Svalbard. Geografiska Annaler: Series A, Physical Geog-raphy 82A, 455e470.

Bemmel, V., Peter, P. (Houston, TX)., Pepper, Randolph, E.F. (Sugar Land, TX)., 2000.Seismic Signal Processing Method and Apparatus for Generating a Cube ofVariance Values: United States, Schlumberger Technology Corporation (Hous-ton, TX).

Berndt, C., 2005. Focused fluid flow in passive continental margins. PhilosophicalTransactions of the Royal Society A: Mathematical, Physical and EngineeringSciences 363, 2855e2871.

Bryn, P., Berg, K., Forsberg, C.F., Solheim, A., Kvalstad, T.J., 2005. Explaining theStoregga slide. Marine and Petroleum Geology 22, 11e19.

Bugge, T., Elvebakk, G., Fanavoll, S., Mangerud, G., Smelror, M., Weiss, H.M.,Gjelberg, J., Kristensen, S.E., Nilsen, K., 2002. Shallow stratigraphic drillingapplied in hydrocarbon exploration of the Nordkapp Basin, Barents Sea. Marineand Petroleum Geology 19, 13e37.

Bünz, S., Mienert, J., Berndt, C., 2003. Geological controls on the Storegga gas-hydrate system of the mid-Norwegian continental margin. Earth and Plane-tary Science Letters 209, 291e307.

Chand, S., Mienert, J., Andreassen, K., Knies, J., Plassen, L., Fotland, B., 2008. Gashydrate stability zone modelling in areas of salt tectonics and pockmarks of theBarents Sea suggests an active hydrocarbon venting system. Marine andPetroleum Geology 25, 625e636.

Chopra, S., Marfurt, K.J., 2007. Seismic Attributes for Prospect Identification andReservoir Characterization, vol. 11. Society of Exploration Geophysicists,Geophysical Developments, Tulsa, USA, 464 pp.

Damm, E., Mackensen, A., Budéus, G., Faber, E., Hanfland, C., 2005. Pathways ofmethane in seawater: plume spreading in an Arctic shelf environment (SW-Spitsbergen). Continental Shelf Research 25, 1453e1472.

Dowdeswell, J.A., Elverhøi, A., 2002. The timing of initiation of fast-flowing icestreams during a glacial cycle inferred from glacimarine sedimentation. MarineGeology 188, 3e14.

Elverhøi, A., Andersen, E.S., Dokken, T., Hebbeln, D., Spielhagen, R., Svendsen, J.I.,Sørflaten, M., Rørnes, A., Hald, M., Forsberg, C.F., 1995. The growth and decay ofthe late Weichselian ice sheet in western Svalbard and adjacent areas based onProvenance studies of marine sediments. Quaternary Research 44, 303e316.

Faleide, J.I., Gudlaugsson, S.T., Eldholm, O., Myhre, A.M., Jackson, H.R., 1991. Deepseismic transects across the sheared western Barents Sea-Svalbard continentalmargin. Tectonophysics 189, 73e89.

Forwick, M., Baeten, N.J., Vorren, T.O., 2009. Pockmarks in Spitsbergen fjords.Norwegian Journal of Geology vol. 89, 65e77. Trondheim 2009. ISSN 029-196X.

Gay, A., Lopez, M., Berndt, C., Séranne, M., 2007. Geological controls on focused fluidflow associated with seafloor seeps in the Lower Congo Basin. Marine Geology244, 68e92.

Graversen, R.G., Mauritsen, T., Tjernstrom, M., Kallen, E., Svensson, G., 2008. Verticalstructure of recent Arctic warming. Nature 451, 53e56.

A. Rajan et al. / Marine and Petroleum Geology 32 (2012) 36e49 49

Greinert, J., Artemov, Y., Egorov, V., De Batist, M., McGinnis, D., 2006. 1300-m-highrising bubbles from mud volcanoes at 2080 m in the Black Sea: hydroacousticcharacteristics and temporal variability. Earth and Planetary Science Letters244, 1e15.

Greinert, J., 2008. Monitoring temporal variability of bubble release at seeps: thehydroacoustic swath system GasQuant. Journal of Geophysical Research 113,C07048.

Haflidason, H., Sejrup, H.P., Nygård, A., Mienert, J., Bryn, P., Lien, R., Forsberg, C.F.,Berg, K., Masson, D., 2004. The Storegga Slide: architecture, geometry and slidedevelopment. Marine Geology 213, 201e234.

Hjelstuen, B.O., Sejrup, H.P., Haflidason, H., Nygård, A., Berstad, I.M., Knorr, G., 2004.Late Quaternary seismic stratigraphy and geological development of the southVøring margin, Norwegian Sea. Quaternary Science Reviews 23, 1847e1865.

Holbrook, W.S., Lizarralde, D., Pecher, I.A., Gorman, A.R., Hackwith, K.L.,Hornbach, M., Saffer, D., 2002. Escape of methane gas through sediment wavesin a large methane hydrate province. Geology 30 (5), 467e470.

Hovland, M., Judd, A.G., 1988. Seabed Pockmarks and Seepages: Impact on Geology,Biology and the Marine Environment. Graham and Trotman Ltd., London. 293 p.

Hustoft, S., Mienert, J., Bünz, S., Nouzé, H., 2007. High-resolution 3D-seismic dataindicate focussed fluid migration pathways above polygonal fault systems ofthe mid-Norwegian margin. Marine Geology 245, 89e106.

Hustoft, S., Dugan, B., Mienert, J., 2009. Effects of rapid sedimentation on devel-oping the Nyegga pockmark field: constraints from hydrological modeling and3-D seismic data, offshore mid-Norway. Geochemistry Geophysics Geosystems10, Q06012.

Jakobsson, M., Macnab, R., Mayer, L., Anderson, R., Edwards, M., Hatzky, J.,Schenke, H.W., Johnson, P., 2008. An improved bathymetric portrayal of theArctic Ocean: implications for ocean modeling and geological, geophysical andoceanographic analyses. Geophysical Research Letters 35, L07602.

Judd, A.G., Hovland, M., 1992. The evidence of shallow gas in marine sediments.Continental Shelf Research 12, 1081e1095.

Judd, A.G., Hovland, M., 2007. Seabed Fluid Flow, the Impact on Geology, Biologyand the Marine Environment. Cambridge University Press, 475 pp.

Jung, W.-Y., Vogt, P.R., 2004. Effects of bottom water warming and sea level rise onHolocene hydrate dissociation and mass wasting along the Norwegian-BarentsContinental Margin. Journal of Geophysical Research 109, B06104.

Kerr, R.A., 2010. ‘Arctic Armageddon’ needs more science, less hype. Science 329,620e621.

Knies, J., Mann, U., 2002. Depositional environment and source rock potential ofMiocene strata from the central Fram Strait: introduction of a new computingtool for simulating organic facies variations. Marine and Petroleum Geology 19,811e828.

Knies, J., Damm, E., Gutt, J., Mann, U., Pinturier, L., 2004. Near-surface hydrocarbonanomalies in shelf sediments off Spitsbergen: evidences for past seepages.Geochemistry Geophysics Geosystems 5, Q06003.

Laberg, J.S., Vorren, T.O., 1995. Late Weichselian submarine debris flow deposits onthe Bear Island Trough Mouth Fan. Marine Geology 127, 45e72.

Landvik, J.Y., Bondevik, S., Elverhøi, A., Fjeldskaar, W., Mangerud, J., Salvigsen, O.,Siegert, M.J., Svendsen, J.I., Vorren, T.O., 1998. The last glacial maximum ofSvalbard and the Barents Sea area: ice sheet extent and configuration.Quaternary Science Reviews 17, 43e75.

Landvik, J.Y., Ingolfsson, O., Mienert, J., Lehman, S.J., Solheim, A., Elverhoi, A.,Ottesen, D., 2005. Rethinking Late Weichselian ice-sheet dynamics in coastalNW Svalbard. Boreas 34, 7e24.

Lelieveld, J., Crutzen, P.J., Brühl, C., 1993. Climate effects of atmospheric methane.Chemosphere 26, 739e768.

Løseth, H., Gading, M., Wensaas, L., 2009. Hydrocarbon leakage interpreted onseismic data. Marine and Petroleum Geology 26, 1304e1319.

Miall, A., 2010. The Geology of Stratigraphic Sequences, second ed. Springer-Verlag,Berlin, p. 522.

Mienert, J., Posewang, J., 1999. Evidence of shallow- and deep-water gas hydratedestabilizations in North Atlantic polar continental margin sediments. Geo-Marine Letters 19, 143e149.

Mienert, J., Vanneste, M., Bünz, S., Andreassen, K., Haflidason, H., Sejrup, H.P., 2005.Ocean warming and gas hydrate stability on the mid-Norwegian margin at theStoregga Slide. Marine and Petroleum Geology 22, 233e244.

Nielsen, T., De Santis, L., Dahlgren, K.I.T., Kuijpers, A., Laberg, J.S., Nygård, A., Praeg, D.,Stoker,M.S., 2005. Acomparisonof theNWEuropeanglaciatedmarginwithotherglaciated margins. Marine and Petroleum Geology 22, 1149e1183.

Ottesen, D.A.G., Dowdeswell, J.A., Landvik, J.Y., Mienert, J., 2007. Dynamics of theLate Weichselian ice sheet on Svalbard inferred from high-resolution sea-floormorphology. Boreas 36, 286e306.

Ottesen, D., Dowdeswell, J.A., Benn, D.I., Kristensen, L., Christiansen, H.H.,Christensen, O., Hansen, L., Lebesbye, E., Forwick, M., Vorren, T.O., 2008.Submarine landforms characteristic of glacier surges in two Spitsbergen fjords.Quaternary Science Reviews 27, 1583e1599.

Petersen, C.J., Bünz, S., Hustoft, S., Mienert, J., Klaeschen, D., 2010. High-resolution P-Cable 3D seismic imaging of gas chimney structures in gas hydrated sedimentsof an Arctic sediment drift. Marine and Petroleum Geology 27, 1981e1994.

Pierre, D.F., Martire, L., Natalicchio, M., Clari, P., Petrea, C., 2010. Authigeniccarbonates in Upper Miocene sediments of the Tertiary Piedmont Basin (NW

Italy): vestiges of an ancient gas hydrate stability zone? Geological Society ofAmerica Bulletin 122, 994e1010.

Plaza-Faverola, A., Bünz, S., Mienert, J., 2010. Fluid distributions inferred fromP-wave velocity and reflection seismic amplitude anomalies beneath theNyegga pockmark field of the mid-Norwegian margin. Marine and PetroleumGeology 27, 46e60.

Posewang, J., Mienert, J., 1999. The enigma of double BSRs: indicators for changes inthe hydrate stability field? Geo-Marine Letters 19, 157e163.

Rehder, G., Keir, R.S., Suess, E., Pohlmann, T., 1998. The Multiple sources andpatterns of methane in North Sea waters. Aquatic Geochemistry 4, 403e427.

Reshetnikov, A.I., Paramonova, N.N., Shashkov, A.A., 2000. An evaluation of histor-ical methane emissions from the Soviet gas industry. Journal of GeophysicalResearch 105, 3517e3529.

Ruppel, C., Dickens, G.R., Castellini, D.G., Gilhooly, W., Lizarralde, D., 2005. Heat andsalt inhibition of gas hydrate formation in the northern Gulf of Mexico.Geophysical Research Letters 32, L04605.

Sarkar, S., Berndt, C., Chabert, A., Masson, D.G., Minshull, T.A., Westbrook, G.K., 2011.Switching of a paleo-ice stream in northwest Svalbard. Quaternary ScienceReviews 30, 1710e1725.

Schlitzer, R., 2008. Ocean Data View", 4.3.7, software. Available at: http://odv.awi.de.Sloan Jr., E.D., 1998. Gas hydrates: review of physical/chemical properties. Energy

and Fuels 12, 191e196.Sloan, E.D., 1990. Clathrate Hydrates of Natural Gases. Marcel Dekker, New York.Slubowska-Woldengen, M., Rasmussen, T.L., Koç, N., Klitgaard-Kristensen, D.,

Nilsen, F., Solheim, A., 2007. Advection of Atlantic Water to the western andnorthern Svalbard shelf since 17,500 cal yr BP. Quaternary Science Reviews 26,463e478.

Solheim, A., ElverhøI, A., 1985. A pockmark field in the Central Barents Sea; gas froma petrogenic source? Polar Research 3, 11e19.

Solheim, A., Faleide, J.I., Andersen, E.S., Elverhoi, A., Forsberg, C.F., Vanneste, K.,Uenzelmann-Neben, G., Channell, J.E.T., 1998. Late Cenozoic seismic stratig-raphy and glacial geological development of the East Greenland and Svalbard-Barents sea continental margins. Quaternary Science Reviews 17, 155e184.

Spielhagen, R.F., Werner, K., Sørensen, S.A., Zamelczyk, K., Kandiano, E., Budeus, G.,Husum, K., Marchitto, T.M., Hald, M., 2011. Enhanced modern heat transfer tothe Arctic by warm Atlantic water. Science 331, 450e453.

Sundvor, E., Austegard, A., 1990. The evolution of the Svalbard margins: synthesisand new results. In: Bleil, U., Thiede, J. (Eds.), Geological History of the PolarOceans: Arctic vs. Antarctic. Kluwer Academic Publishers, Bremen, pp. 77e94.

Sættem, J., Poole, D.A.R., Ellingsen, L., Sejrup, H.P., 1992. Glacial geology of outerBjornoyrenna, Southwestern Barents Sea. Marine Geology 103, 15e51.

Vanneste, M., Guidard, S., Mienert, J., 2005. Bottom-simulating reflections andgeothermal gradients across the western Svalbard margin. Terra Nova 17,510e516.

Vanneste, M., Berndt, C., Sverre Laberg, J., Mienert, J., 2007. On the origin of largeshelf embayments on glaciated marginseeffects of lateral ice flux variationsand glacio-dynamics west of Svalbard. Quaternary Science Reviews 26,2406e2419.

Vogt, P.R., Crane, K., Sundvor, E., Max, M.D., Pfirman, S.L., 1994. Methane-gener-ated(?) pockmarks on young, thickly sedimented oceanic crust in the Arctic:Vestnesa ridge, Fram strait. Geology 22, 255e258.

Vogt, P.R., Gardner, J., Crane, K., 1999. The NorwegianeBarentseSvalbard (NBS)continental margin: introducing a natural laboratory of mass wasting, hydrates,and ascent of sediment, pore water, and methane. Geo-Marine Letters 19, 2e21.

Vogt, P.R., Jung, W.-Y., 2002. Holocene mass wasting on upper non-Polar continentalslopes; due to post-Glacial ocean warming and hydrate dissociation?Geophysical Research Letters 29, 1341.

Vorren, T.O., Lebesbye, E., Andreassen, K., Larsen, K.B., 1989. Glacigenic sedimentson a passive continental margin as exemplified by the Barents Sea. MarineGeology 85, 251e272.

Vorren, T.O., Laberg, J.S., 1997. Trough mouth fansdpalaeoclimate and ice-sheetmonitors. Quaternary Science Reviews 16, 865e881.

Vorren, T.O., Laberg, J.S., Blaume, F., Dowdeswell, J.A., Kenyon, N.H., Mienert, J.,Rumohr, J., Werner, F., 1998. The Norwegian-Greenland Sea continentalmargins: morphology and late quaternary sedimentary processes and envi-ronment. Quaternary Science Reviews 17, 273e302.

Weibull, W., Mienert, J., Bunz, S., Hustoft, S., 2010. Fluid migration directionsinferred from gradient of time surfaces of the sub seabed. Marine and Petro-leum Geology 27, 1898e1909.

Wessel, P., Smith, W.H.F., 1998. Improved version of the generic mapping toolsreleased. EOS Transactions American Geological Union 79, 579.

Westbrook, G.K., Thatcher, K.E., Rohling, E.J., Piotrowski, A.M., Pälike, H.,Osborne, A.H., Nisbet, E.G., Minshull, T.A., Lanoisellé, M., James, R.H.,Hühnerbach, V., Green, D., Fisher, R.E., Crocker, A.J., Chabert, A., Bolton, C.,Beszczynska-Möller, A., Berndt, C., Aquilina, A., 2009. Escape of methane gasfrom the seabed along the West Spitsbergen continental margin. GeophysicalResearch Letters 36, L15608.

Zhu, Y., Huang, Y., Matsumoto, R., Wu, B., 2003. Geochemical and stable isotopiccompositions of pore fluids and authigenic siderite concentrations from site1146. ODP leg 184: implications for gas hydrate. In: Proceedings of the OceanDrilling Program, vol. 184.