BasinResearch Newclassificationsystemformasstransport ... · 11 Big 95 0 DebrisFlowEbroSpainw Pleistocene 26 Lastrasetal.(2005) 12 PeachSlide w Pleistocene 135 Holmesetal.(1998)andKnutzetal

New classification system formass transportcomplexes in offshoreTrinidadLorena Moscardelli and Lesli Wood

Bureau of Economic Geology, Jackson School of Geosciences, University of Texas at Austin, Austin,TX, USA

ABSTRACT

This paper delineates our use of10 708 km2 of three-dimensional (3D) seismic data from thecontinental margin of Trinidad and TobagoWest Indies to describe a series of mass transportcomplexes (MTCs) that were deposited during the Plio-Pleistocene.This area, situated along theobliquely converging boundary of the Caribbean/SouthAmerican plates and proximal to theOrinocoDelta, is characterized by catastrophic shelf-margin processes, intrusive/extrusive mobile shales andactive tectonism. Extensive mapping of di¡erent stratigraphic intervals of the 3D seismic surveyreveals severalMTCs that range in area from11.3 to 2017 km2.Three types ofMTCs are identi¢ed: (1)shelf-attached systems that were fed by shelf-edge deltas whose sediment input is controlled by sea-level £uctuations and sedimentation rates; (2) slope-attached systems, which occur when upper-slope sediments catastrophically fail owing to gas-hydrate disruptions and/or earthquakes and (3)locally detached systems, formedwhen local instabilities in the sea£oor trigger relatively smallcollapses. Such classi¢cation of the relationship between slope mass failures and sourcing regionsenables a better understanding of the nature of initiation, length of development history andpetrography of suchMTCs. 3D seismic enables more accurate calculation of deposit volumes,improves deposit imaging, and, thus, increases the accuracy of physical and computer simulationsof mass failure processes.

INTRODUCTION

Mass transport complexes (MTCs) play a large role in thestratigraphic ¢ll of basins in o¡shore northeastern SouthAmerica, as well as along other continental marginsaround the world (Table 1). An incredible variety of Plio-cene- and Pleistocene-age MTCs have been documentedin deep-water areas around o¡shoreTrinidad (Moscardelliet al., 2006).These deposits provide an opportunity to ex-amine the variety of MTCs that occur in deep-water mar-gins, document their architecture and propose a newclassi¢cation system that is based on causal mechanisms,dimensions and source areas ofMTCs.

The area of this study, 10 708 km2 in size, is located inwater depths of 300^1500m along the SE margin ofthe tectonically active Caribbean Plate Boundary Zone(Fig.1). Earthquakes up tomagnitude 7 are frequent occur-rences in the area, and the region is characterized by bothcompressional and extensional tectonic forces. Avariety offaults are found in the area, including, from south tonorth, extensional down-to-the-NE listric faulting, trans-pressional south-verging thrusts and right-lateral strike^slip faults (Heppard et al., 1998; Wood, 2000; Boettcheret al., 2003). High sediment accumulation rates, with sedi-

ment dominantly supplied by the Orinoco River andDelta,high-frequency Plio-Pleistocene sea-level £uctuations,abundant gas-hydrate deposits and tectonic instability cre-ate the right geological preconditions for mass shelf-edgefailures and the subsequent deposition of gravity-induceddeposits (MTCs and turbidites).

The predominance of mass-wasting processes in mostcontinental margins around the world (Table 1) makesMTCs a potentially important element of any deep-waterstratigraphic succession. In addition, their in£uence in theshaping of sea£oor bathymetry and continental margin ar-chitecture is well documented (Norem et al., 1990;Massonetal., 1998; Gee etal., 1999, 2001, 2006; Frey-Martinez etal.,2005; Moscardelli et al., 2006). The sometimes cata-strophic, erosive nature ofMTCdeposits represents a po-tential geotechnical hazard for o¡shore infrastructuresowing to their ability to compromise the integrity of un-derwater equipment such as cables, communication linesand pipelines (Ho¡man et al., 2004; Shipp et al., 2004). Itis also well known that a sudden displacement of the sea-£oor through catastrophic slumping or sliding can poten-tially disrupt the water column above the failure andgenerate tsunamigenic waves that can a¡ect coastal areas(Pelinovsky & Poplavsky, 1996; Nisbet & Piper, 1998; Daw-son et al., 2004; Fryer et al., 2004). Deep-water MTC de-posits possess physical characteristics that often di¡ergreatly from deep-water submarine-fan or slope-leveedchannel systems (Shipp etal., 2004).The frequent presence

Correspondence: Lorena Moscardelli, Bureau of EconomicGeology, Jackson School of Geosciences, University of Texas atAustin, PO Box X, Austin,TX 78713-8924, USA. E-mail: [email protected]

BasinResearch (2007) doi: 10.1111/j.1365-2117.2007.00340.x

r 2007 The Authors. Journal compilation r 2007 Blackwell Publishing Ltd 1

of these deposits in deep-water basins means that e⁄cientoil and gas exploration and development in such settingshinge on an understanding of the nature, processes of for-mation, distribution and relationship that these depositsshare with surrounding deep-water environments.

Terminology and classification of MTCs

More than four decades have elapsed since Dott (1963)pointed out that there was a considerable amount of con-fusion in the application of terms that were used to de-scribe mass movement phenomena. Despite the greatadvances that have been made so far in areas such as faciesmodel analysis, sequence stratigraphic studies and seismic

acquisition and imaging techniques of deep-water depos-its, a review of the ‘modern’ literature shows that terminol-ogy in MTC study remains a confusing mix of termscompiled by scientists working in a variety of geologicalenvironments (Nardin et al., 1979; Shanmugam, 2000,2002). Part of the problem is that submarine mass move-ments are broadly de¢ned as movement of sediment dri-ven by gravity (gravity-induced deposits) and, therefore, ahuge spectrum of deposits can fall into the category,including slides, slumps, debris £ows and turbidites.

Nardin et al. (1979) reviewed the terminology andcompiled a comprehensive classi¢cation system for sub-marine mass movement processes. The descriptive partof the classi¢cation is based on the recognition of certain

Table1. Worldwide occurrence and distribution of o¡shoreMTCs

No. Location and type Age Volume (km3) Reference

1 Angola Slidesn Miocene 20 Gee et al. (2006)2 NigerMTCsw Pliocene ^ Shanmugam et al. (1997) andNissen

et al. (1999)3 CapeVerde Slidesn Pleistocene ^ Wynn et al. (2000)4 Sahara Debris Floww Pleistocene 600^1.100 Gee et al. (1999)5 Canary Debris Flowsw Pleistocene 400 Masson et al. (1998)6a El Golfo Debris Avalanchew Pleistocene 150^350 Wynn etal. (2000) andGee etal. (2001)6b El Julan Debris Avalanchew Pleistocene 130 Wynn etal. (2000) andGee etal. (2001)6c Las Playas Debris Avalanchew Pleistocene 30 Wynn etal. (2000) andGee etal. (2001)7 Orotava-Icod-Tino Avalanchew Pleistocene 1000 Wynn et al. (2000)8 MoroccoMTCsn,w Tertiary ^ Lee et al., (2004a)9 NileMTCsn,w Quaternary 670 Newton et al. (2004)10 Israel Slump Complexesn,w Quaternary 1000 Frey-Martinez et al. (2005)11 Big 950 Debris Flow Ebro Spainw Pleistocene 26 Lastras et al. (2005)12 Peach Slidew Pleistocene 135 Holmes et al. (1998) and Knutz et al.

(2001)13 Faeroe Sliden,w Pleistocene 135 VanWeering et al. (1998)14 Storegga Slidew Holocene 2400^3500 Jansen et al. (1987), Evans et al. (1996),

Bouriak et al. (2000), Ha£idason et al.(2005)

15 Traenadjupet Slidew Holocene 900 Laberg & Vorren (2000)16 Andoya Slidew Holocene 485 Laberg et al. (2000)17 Ba⁄n Bay Slide/Debris Flown,w Plio-Pleistocene ^ Aksu &Hiscott (1989)18 Nova ScotiaMTCsn,w Eocene to Holocene ^ Campbell et al. (2004)19 New JerseyMTCsw Quaternary 1.7 McAdoo et al. (2000)20 Baltimore Canyon Slidew Holocene 200 Embley & Jacobi (1986)21 Cape Fear Slidew Pleistocene 1400 Popenoe et al. (1993)22 Gulf ofMexicoMTCsn,w Tertiary 27.5 McAdoo et al. (2000)23 TrinidadMTCsn,w Plio-Pleistocene 11.3^2017 Moscardelli et al. (2006)24 Amazon FanMTCsn,w Pleistocene 1500^2000 Piper et al. (1997)25 Ranger Slide Baja Californian Pleistocene 20 Normark (1974)26 PalosVerdesMTC^Californiaw Holocene 0.34^0.72 Lee et al. (2004b)27 SantaMonicaMTC^Californiaw Quaternary 0.0002 Lee et al. (2004b)28 GoletaMTC^Californiaw Holocene 0.5 Lee et al. (2004b)29 GaviotaMTC^Californiaw Holocene 0.01^0.02 Lee et al. (2004b)30 OregonMTCw Quaternary 3.8 McAdoo et al. (2000)31 BruneiMTCn,w Quaternary 80 McGilvery et al. (2004)32 Eastern KalimantanMTCn,w Pleistocene ^ Posamentier &Kolla (2003)

Numbers correspond to modern and ancientMTCs identi¢ed in o¡shore seismic. Letters correspond to ancientMTCs identi¢ed in outcrop.nBuriedMTCs.wMTCs that have sea£oor expression.MTCs, mass transport complexes.

r 2007 The Authors. Journal compilation r 2007 Blackwell Publishing Ltd,Basin Research, 10.1111/j.1365-2117.2007.00340.x2

L. Moscardelli andL.Wood

Blo

ck 2

7

Blo

ck 2

6

Blo

ck 2

5b

Blo

ck 2

5a

Blo

ck 4

a

10 k

m

(b)

Mu

d V

olc

ano

MTC

s

N(a

)

MTC

_2.1

MTC

_2

MTC

_2.2

MTC

_2.3

MTC

_2.4

Dar

ien

Rid

ge

Hay

dn

Rid

ge

Cal

lico

re R

idg

e

Hel

ico

niu

s R

idg

e

Cat

fish

Rid

ge

East

Pro

spec

t R

idg

e

Mu

d V

olc

ano

Rid

ge

LEG

END

Dar

ien

Ridg

e2000

1000

Co

lum

bu

s B

asin

Are

a o

f Stu

dy

59°

61°

10°

11°

N

Trin

idadCar

ibb

ean

pla

te m

oti

on

Sou

th A

mer

ica

pla

te m

oti

on

23 4

5

1

Y'6

Y

Cent

ral

Rang

eX

X'

Blo

ck 4

b

200

Fig.1.

(a)M

apshow

ingthearea

ofstud

ylocatedinthearea

ofno

rtheastern

Sou

thAmericaalon

gtheC

aribbean

PlateB

ound

aryZon

ewheretheC

aribbean

plate(totheno

rth)andtheS

outh

American

plate(tothesouth)meet.The

area

ofthree-dimension

alseismicdataishighlig

hted

ingrey.T

heshelfb

reak

isshow

nasaheavyblackdash

anddotted

linealon

gthe2

00to300m(660

to1000

ft)contou

r.Noticethealignm

entofm

ud-volcano

ridges

inthedeep

-marinesetting,withtheNE^S

W-orientedanticlineslocatedin

theshelfalpartofthe

Colum

busB

asin.(b)Se

ismicsurveysinthestud

yarea

correspo

ndto

blocknames.S

hadedareascorrespo

ndtothelocation

ofMTCsw

ithinseismicun

itVI.

r 2007 The Authors. Journal compilation r 2007 Blackwell Publishing Ltd,Basin Research, 10.1111/j.1365-2117.2007.00340.x 3

New classification system forMTCs in offshoreTrinidad

sedimentary features of the deposits, including sorting ofsediments, size distribution, strati¢cation, grading and se-dimentary fabric. On the other hand, the genetic part ofthe classi¢cation is based on interpretation of transportand mechanical behaviour from which characteristics ofthe sedimentary deposits can be inferred. This approachdoes not address the issues of grain-size distribution, se-dimentation rate or slope gradient, which ultimately willin£uence boundary conditions for mechanical behaviour(Nardin et al., 1979) (Fig. 2). For purposes of this discus-sion, slides, slumps and debris £ows will be consideredconstituents ofMTCs and can co-occur in the same eventor depositional unit (Dott, 1963).

The existing classi¢cation system (Nardin et al., 1979) isuseful in recognizingMTCs from a process sedimentolo-gical perspective, but it fails to address the broader issuesof causal mechanisms and pre-failure conditions ofMTCs. This lack of di¡erentiation of mechanisms fromconditions is critical because it is thought that for a givenset of geological conditions and environments, a variety ofcausal mechanisms can generate MTCs having similargeomorphological characteristics. However, previous stu-dies in o¡shore Trinidad (Brami et al., 2000; Moscardelliet al., 2006) have shown that there are signi¢cant geomor-phological di¡erences betweenMTCs that were depositedin this region during the Plio-Pleistocene. These di¡er-ences are particularly notorious in the updip (proximal)parts of theMTCs, where initial interpretations have sug-gested that there is a connection between the geomorphol-ogy of individual MTCs and their potential causalmechanisms (Moscardelli et al., 2006). Despite the prolif-eration of papers describingMTC architecture around theworld (Table 1), none has categorized the geomorphologi-cal relationships that exist between the updip (proximal)and downdip (distal) parts of MTCs. We think that a

detailed examination of these geomorphological relation-ships (updip vs. downdip), in conjunction with a compara-tive analysis of the scales and dimensions of MTCs, couldhelp in deducing the causal mechanisms associated withindividualMTC events.

The main aim of this paper is to propose a new classi¢ -cation system for MTCs in o¡shoreTrinidad on the basisof the relationships encountered between slope mass fail-ures, sourcing regions, dimensions and geometries ofMTCs. Establishment of this new classi¢cation systemseeks to enable a better understanding of the nature of in-itiation, length of development history and causal me-chanisms of MTCs.The new classi¢cation scheme that isproposed herein is not intended to con£ict with butto complement the existing classi¢cation system forgravity-induced deposits (Nardin et al., 1979). The newclassi¢cation will be established by our answering thefollowing questions: (1) What are the dimensions andgeomorphological characteristics of various MTCs inthe deep-marine margin of o¡shore Trinidad? (2) Whatgeomorphological elements in£uence the distribution ofgravity-induced deposits in the deep-marine margin ofo¡shoreTrinidad? (3) What geomorphological di¡erencesbetween MTCs have di¡erent sourcing areas? (4) Can weuse the seismic geomorphology of ancient MTCs to un-derstand the processes active at their time of deposition?(5) Does each force trigger the same process for the givencritical conditions of sediment stability?

METHODOLOGYAND DATA SETS

The primary data set for this study of the deep-water ar-chitecture and nature of the o¡shoreTrinidad continentalmargin are three-dimensional (3D) seismic re£ection data.

GRAVITY INDUCED DEPOSITS

Mas

s Tr

ansp

ort

Co

mp

lex

Turb

idit

y C

urr

ent

Turb

idit

eD

ebri

s Fl

ow

Slu

mp

Slid

e

Descriptive ClassificationGenetic Classification

Shear failure along discrete shear planes with little or no internal deformation or rotation

Shear failure accompanied by rotation along discrete shear surfaces with various degrees of internal deformation

Shear distributed throughout the sediment mass. Strength is principally from cohesion due to clay content. Additional matrix support may come from buoyancy. Plastic rheology and laminar state.

Transport Mechanism

Supported by fluid turbulence (newtonian rheology)

Sedimentary Structures

Essentially undeformed, continuous bedding

Plastic deformation particularly at the toe or base. Plow structures, folds, tension faults, joints, slickensides, grooves, rotational blocks

Matrix supported, random fabric, clast size variable, matrix variable. Rip ups, rafts, inverse grading and flow structures possible.

Normal size grading, sharp basal contacts, gradational upper contacts.

Seismically Recognizable Features

(Moscardelli et al., 2006; this work)

Continuous blocks without apparent internal deformation. High-amplitude, continuous reflections.

Compressional ridges, imbricate slides, irregular upper bedding contacts, duplex structures, contorted layers. Low- and high-amplitude reflections geometrically arranged as though deformed through compressive stresses.

Mega rafted and/or detached blocks, irregular upper bedding contacts, lateral pinch-out geometries, oriented ridges and scours. Low-amplitude, semitransparent chaotic reflections.

Lobate features

Laterally continuous

Fig. 2. Classi¢cation of gravity-induced deposits. Compiled fromDott (1963), Nardin et al. (1979) andMoscardelli et al. (2006).



Seismic coverage is large enough to encompass several en-tireMTC staging areas and depocentres (Fig.1). Five sepa-rate 3D seismic data volumes were available and weremerged into a single continuous volume spanning10 708 km2. Imaging depths vary between individual sur-veys but range from a minimum of 1400ms (milliseconds)two-way traveltime (TWTT) in the westernmost part ofthe area (blocks 4ab) to a maximum of 5000msTWTTofcoverage over central and eastern areas (blocks 25b and26) (Fig. 3a).These data are all late-1990s vintage, and theiracquisition parameters and quality are similar. Bin spacingof the time-migrated volumes is 25 � 12.5m, and they allhave a 4ms vertical sampling rate. Seismic volumes wereprocessed to zero phase, and the average frequency con-tent of the full-stack seismic data was 30^35Hz. For ap-proximate time-depth conversions at shallow depths,1ms is equivalent to 0.75m.

The quality of the 3D seismic data used in this studymakes it possible to correlate seismic units with a high de-gree of con¢dence. Eight regional unconformities and thesea£oor were mapped and correlated across the slope intothe deep-marine basin (Fig. 3a). The correlation alsocrossed several mud-volcano provinces (Fig. 4).The eightinterpreted unconformities and the sea£oor were named,from base to top, P60, P50, P40, P30, P20, P15, P10, P4 andsea£oor.Mapping of these horizonswas the result of a col-laborative e¡ort between di¡erent researchers working forthe Quantitative Clastic Laboratory Consortium at theBureau ofEconomicGeology (Sullivan etal., 2004;Garcia-caro, 2006;Moscardelli etal., 2006) (Figs 3a and 4). In eachcase, all surfaces were seeded using the auto-picking tooland then interpolated along key amplitude horizons; theinterpolation method honoured the geometry and ampli-tude values of individual re£ections. These surfaces werealso adjusted by hand where the picking tool strayed fromthe horizon. The P10 unconformity de¢nes the base ofMTC_1, a unit that was described extensively by Moscar-delli et al. (2006).The P4 unconformity de¢nes the top ofMTC_1 and the base of the youngest levee-channel com-plex in this region (Moscardelli et al., 2006).The latest ex-pression of this channel complex on the modern sea£oorand its main geomorphological characteristics were de-scribed byMize et al. (2004) and Sullivan et al. (2004). Sur-faces mappedwere used to de¢ne the gross architecture ofthe basin and to assess its sedimentary- ¢ll history. Eightmajor seismic units (seismic units I^VIII) were de¢nedusing the mapped unconformities and the sea£oor(Fig. 3a).However, the main focus of this research has beenconcentrated in describing and understandingMTCs thatare contained in seismic units VI (MTC_2, MTC_2.2,MTC_2.3 andMTC_2.4) andVII (MTC_1) (Fig. 3a). Isochronmaps of the youngest seismic units (seismic unitsVI^VIII)(Figs 3a and 5) helped us understand the character of thecontinental margin ¢ll in the last stages of basin evolution.Documentation ofvariations in the location and geometryof major depocentres and sedimentary conduits thatconnected the outer shelf/upper-slope region with thedeep-marine parts of the basin was emphasized. These

isopach maps were also useful in tracking changes in theareal extension and geometry of mud-volcano ridgesidenti¢ed in the study area (Sullivan et al., 2004). Track-ing changes in the general morphology of the mud-volcano ridges was important because these featuresacted as palaeobathymetric barriers that helped controlsediment transport and distribution in the deep-waterregion.

MTC architecture was imaged by using root-mean-squared (RMS) amplitude extractions. RMS is a seismicattribute that calculates the square root of the sum oftime-domain energy (square of amplitude), a¡ording highamplitudes the maximum opportunity to stand out frombackground contamination (Chen & Sidney, 1997). TheMTCs that are described in this workwere imaged bygen-erating RMS amplitude extractions in successive 20ms(�15 -m) time windows above the P15 unconformity inblocks 25 and 26 (Fig. 6).The P15 unconformity representsthe erosional surface that de¢nes the base of the MTCsthat are contained within seismic unit VI (MTC_2,MTC_2.1, MTC_2.2, MTC_2.3 and MTC_2.4) (Fig. 7). TheP15 unconformity was picked in blocks 25 and 26 as the‘key horizon’ that guided the amplitude extractions withinthe unit of interest (seismic unit VI) because the internalarchitecture of MTCs is in£uenced greatly by the initialerosional events that generate their basal surfaces. Unfor-tunately, a lack of seismic coverage below1400ms in blocks4ab (Fig. 3a) prevented us from mapping the equivalent ofthe P15 unconformity in proximal areas (updip) of the 3Dmegamerge (Figs 3a and 7). Owing to this data limitation,an alternative methodology had to be implemented forimaging the equivalent interval of seismic unitVI in blocks4ab. The ¢rst step was to £atten the 3D seismic volumeusing the P10 unconformity (upper boundary of seismicunit VI) as the hanging horizon (Fig. 3b).The main objec-tive of £attening the volume was to minimize stratigraphicdistortion caused by post-depositional structural defor-mation in this area (notice the abundance of faults inblocks 4ab) (Fig. 3a). The second step was to generate anRMS amplitude extraction in a 20ms window below theP10 unconformity. This particular time window withinthe upper part of seismic unit VI was chosen because itwas the only one that could provide us with complete arealcoverage of blocks 4ab. Finally, the amplitude extractionsthat were obtained for seismic unit VI in blocks 4, 25 and26 were merged to obtain a regional picture of the strati-graphic elements that were contained in this unit (Fig.6a). There is a vertical stratigraphic shift (200^400ms)between the intervals that were imaged updip (blocks 4ab)vs. the ones that were imaged downdip (blocks 25 and 26)in the 3D seismic megamerge; however, the composite im-age of the multiple RMS amplitude extractions shows analmost perfect match of geomorphological architecturesbetween the di¡erent blocks (Fig. 6). This match occursbecause the seismic character of seismic unit VI in blocks4ab does not present dramatic vertical variations, thus pre-serving a similar geomorphological expression throughthe entire interval.



500

1000

1500

2000

2500

3000

3500

4000

45000

4000

m

Blo

ck 4

a

Blo

ck 2

5bB

lock

25a

NW

SEN

E

XX

'

X

X'

N

Blo

ck 4

bSW

?

?

mse

cTW

TT 0

−500 50

0

1000

4000

m

NW

SESW

NE

??

Am

plit

ud

e Ex

trac

tio

n20

mse

c w

ind

ow

Inte

rval

of I

nte

rest

Shel

f Bre

ak

Inte

rval

of I

nte

rest La

ck o

f Dat

a

(a)

(b)

I

Surf

aces

Seismic Units

IIIIIIVVVI

VII

VIII

Y

Y'

Seis

mic

Un

it V

I

Fig 4

Fig.3.

(a)C

ompo

siteNE^S

W-orientedregion

alseismiclin

eshow

ingadipview

acrossthestud

yarea.N

inekeyho

rizons,including

thesea£oor,wereinterpretedto

explainthearchitectureofthe

Tertiary

sedimentary

in¢llintheregion.(b)Se

ismiclin

eon

blocks4abshow

ing£atteningoftheseismicon

theP

10MTC_1surface.In

yellow^20

mstim

eintervalused

togeneratetheR

MSam

plitud

eextraction

show

ninFig.6(blocks4ab).



CY NW

Y'

SE

Hay

dn

Rid

ge

Cal

lico

reR

idg

e

Hel

ico

niu

sR

idg

e

Cat

fish

Rid

ge

2

34

5

a

Hay

dn

Ch

ute

sD

arie

n C

hu

tes

Cat

fish

Ch

ute

Cal

lico

re C

hu

te

Fig.4.

NW^S

E-orientedregion

alseismiclin

eshow

ingastrike

view

acrossthestud

yandlocation

ofthemainmud

-volcano

ridges

andchutespresentintheregion.



10 km

N

Phase VIII

N

Phase VI

10 km

10 km

N

Darien RidgeEast Prospect Ridge

Callicore Ridge

Heliconius Ridge

Catfish Ridge

Haydn Ridge

Thinning of Sedimentary Cover

Paleo Shelf Break

N

10 km

Darien Ridge

Haydn Rid

ge

East Prospect Ridge

Callicore Ridge

Heliconius Ridge

Catfish Ridge

Paleo Shelf Break

10 km

N

Darien RidgeEast Prospect Ridge

Callicore Ridge

Haydn Rid

ge

Heliconius Ridge

Catfish Ridge

Catfish Mega-Chute

CallicoreMega-Chute

Catfish Mega-Chute

CallicoreMega-Chute

HaydnMega-Chute

CallicoreMega-Chute

HaydnMega-Chute

DarienMega-Chute

D

D

D

D

D

D

D Depocenters

(a)

(c)

Reactivation andUplifting

no seismic coveragebelow 1.4 sec


D

D

10 km

N

Phase VII

D

D

D

Paleocanyon

MTC_2

MTC_2.1

MTC_1

(b)

Fig. 5. Isochron maps between key stratigraphic surfaces. (a) PhaseVI ^ sedimentationwas constrained mainly to theNW,Darien andHaydn megachutes were delivering sediments from the outer shelf/upper slope area (main feeders ofMTC_2.1).The Callicore megachutestarted to deliver sediments towards theNE (main feederMTC_2). (b) PhaseVII ^ the main axis of deposition shifted towards the SE.TheCallicore andCat¢sh megachutes were the only active sedimentary pathways delivering extrabasinal sediments towards the deep-marineenvironments.During this time,MTC_1was deposited (c)PhaseVIII ^ last stage of basin in¢lling, theCallicoremegachute appears to be theonly active sedimentary pathway at this time. Uplifting of the northern region of the study area caused de£ection of the main depositionalaxis towards the east.This interval was dominated by deposition of levee-channel complexes.MTCs, mass transport complexes.



10 km

N

10 km

N

Darien Ridge

Darien Ridge

Haydn Ridge

Haydn Ridge

East Prospect Ridge

Callicore Ridge

Heliconius Ridge

Heliconius Ridge

Catfis

h Rid

ge

East Prospect Ridge

Catfish

Rid

gePaleo Shelf Break

Paleo Shelf Break

Callicore Mega Chute

Catfish Mega Chute

Haydn Mega Chute

Slope- AttachedMTC_2.1

DetachedMTC_2.2

Slope- AttachedMTC_2

DetachedMTC_2.3

DetachedMTC_2.4

Callicore Ridge

Faults

Depletion Zone

(a)

(b)

Mud Volcano Ridges

Attached MTCs

Detached MTCs

Faults

Slumps

Channels

LEGEND

Fig. 6. (a) Root-mean-squared (RMS) amplitude extraction map showing the geomorphological expression in plan view of attached(MTC_2 andMTC_2.1) and detached (MTC_2.2,MTC_2.3 andMTC_2.4)MTCs in the study area. RMS amplitude extraction maps inblocks 25ab and 26 were extracted from the basal part of seismic unit VI. RMS amplitude extraction maps in blocks 4ab were extractedfromwithin seismic unit VI (interval highlighted in Fig. 3). (b) Geomorphological interpretation of RMS amplitude extraction mapswithin seismic unit VI. Sedimentary sources of attachedMTCs are associatedwith the outer shelf/upper slope area, whereassedimentary sources for detachedMTCs can be linked to the £anks of the mud-volcano ridges.MTCs, mass transport complexes.



Several exploratory wells have been drilled in the studyarea, but direct access to these datawas not available to theauthors. Several drop coreswere collected in the region forgeotechnical tests on near-sea£oor sediments, but nonewas deep enough to penetrate the intervals of interest.Data from exploration drilling in the study area were usedby Shipp et al. (2004) to document lithologic and physicalcharacteristics of the north part of the shallowest MTC inthe area (MTC_1 contained in seismic unit VII) (Moscar-delli et al., 2006).The MTCs that are described herein areslightly older thanMTC_1; however, it is thought that theirlithologic and physical characteristics are similar.

GEOLOGICAL SETTING

Avariety of palaeobathymetric and structural features havein£uenced sediment transport and distribution in theproximal and distal parts of the deep-marine margin ofo¡shoreTrinidad and Venezuela.This bathymetry a¡ectedevolution of the sedimentary succession during Plioceneand Pleistocene times (Moscardelli etal., 2006). In the out-

er-shelf/upper-slope area, listric normal and counternor-mal faults occur. Additionally, normal and inverse faultsde¢ne the lateral boundaries of shelf-edge horst and gra-ben palaeocanyons that funnel sediments from the outer-shelf/upper-slope area towards the deep-marine basin(Moscardelli et al., 2006).

In distal parts of the basin, mud volcanoes have been animportant bathymetric element in£uencing deep-watersedimentation (Figs 4 and 5). Sullivan et al. (2004) docu-mented 161 mobile mud-associated structures on thesea£oor within the study area that vary greatly in size,shape and distribution. Some of these mud diapirs andvolcanoes form visible linear clusters that are de¢ned asmud-volcano ridges and are associated with deeper anti-clinal structures (Figs 4 and 5).These mud-volcano ridgesare part of the palaeobathymetric highs that bound largesediment pathways or megachutes (Fig.5). Inwestern partsof the study area, these ridges funnel sediments directlydownslope from shelf-edge deltas and from the upperslope collapses towards basinward depocentres that areuncon¢ned (Figs 5 and 6). Alternatively, in some areas ofthe basin and parts of the stratigraphic record,mud ridges

10 km

Cat FishMega-Chute

N

HeliconiusRidge

a

b

c

d

a

b

c

d

Western Erosional Edge

Eastern Erosional Edge

Mega Scours

Terraces

MTC_2

CallicoreMega-Chute

HaydnMega-Chute

DarienMega-Chute

CallicoreRidge

East Prospect Ridge

Darien Ridge

Haydn Ridge

Block 4a Block 4b

Block 26

Block 25b

Block 25a


Scale in Time (TWTT)

CatfishRidge

Fig.7. Stratigraphic surface P15MTC_2 ^ regional erosional surface de¢ning the base of seismic unit VI, which contains theMTCs thatare described in this work. Slope-attachedMTC_2, located in the central part of the data set, generated more than 60m of verticalincision in the basal part of seismic unit VI. Some of the main geomorphological elements that are part of the basal parts ofMTC_2include (a, b) east andwest erosional edges, (c) mega scours and (d) several levels of terraces. Colours represent depth in milliseconds,with shallower depths in brighter colours and deeper depths in darker colours. MTCs, mass transport complexes.



act to isolate and con¢ne some parts of the basin fromupslope sedimentation. For purposes of discussion, fromthe NW to the SE, the ridges associated with these mudvolcanoes have been named (1) Darien Ridge (2) HaydnRidge, (3) East Prospect Ridge, (4) Callicore Ridge,(5) Heliconius Ridge and (6) Cat¢sh Ridge (Figs 4^6).

Summary of basin evolution

The last three stages of basin evolution (seismic unitsVI^VIII) are discussed brie£y in this section to explain someof the elements that in£uenced the distribution of gravity-induced deposits in the deep-marine margin of o¡shoreTrinidad. Isopach maps of seismic unitsVI^VIII show sig-ni¢cant variations in the location, geometry and extensionof the depocentres that were developed in the study areaduring the last stages of basin in¢lling (Fig. 5). Generallyspeaking, these depocentres have an elongated geometryand present a NE^SWorientation that parallels the axis ofthe mud-volcano ridges. The depocentres were connectedto the outer-shelf and upper-slope region by megachutesthat were responsible for sediments funneling downslope(Figs 5 and 6) (see also, Moscardelli et al., 2006).The mapsalso reveal an overall migration of the depocentres fromNW to SE, as well as progressive thinning of the strati-graphic succession near theDarien Ridge area (Fig. 5).

The isopach map of seismic unit VI shows that the maindepocentre during this phase was located in the NWcor-ner of the study area nearDarienRidge.Darien andHaydnmegachutes were most likely acting as main feeders of theNW depocentre, whereas the Callicore megachute wasfunneling sediments from the upper slope towards thecentral part of the basin (Figs 5a and 6). The MTCs thatare described in this paper are contained in the NWandcentral parts of seismic unit VI (MTC_2, MTC_2.1,MTC_2.2,MTC_2.3 andMTC_2.4) (Fig. 6).

Location and distribution of the main depocentres inseismic unit VII (Fig. 5b) greatly di¡er from those of theprevious basin con¢guration (Fig. 5a). The isopach mapof seismic unit VII (Fig. 5b) shows that the areal extensionof the NW depocentre was signi¢cantly reduced and newdepocentres were relocated to the central and SE parts ofthe basin.The central depocentre was located to the southof Haydn Ridge, where a 2017 km2 MTC (MTC_1) was de-posited (Moscardelli et al., 2006).The second depocentrewas located in the SE corner of the data set, to the southof Heliconius Ridge, and its axis was parallel to the Cat¢shmegachute.The isopach map of this unit also shows thin-ning of the stratigraphic succession towards the NW (neartheDarienRidge area), aswell as an increase in areal exten-sion of the relative highs associatedwithDarien, East Pro-spect andHaydnRidges.During the deposition of seismicunitVII (Fig. 5b),most sedimentation started to shift fromthe NW (seismic unit VI) (Fig. 5a) to the central and SEparts of the study area, as revealed by the SE migration ofthe main depocentres.

Finally, the isopach map of seismic unit VIII shows thatmost sedimentation was taking place in the central part of

the study area (Fig. 5c) between Haydn and HeliconiusRidges.The Callicore megachute was the main sedimen-tary conduit for sediments to reach the deep-marineregion from outer-shelf and upper-slope regions. Devel-opment of levee-channel complexes during this time wasdominant in the basin (Mize et al., 2004; Moscardelliet al., 2006). In the SE, the main depositional axis of thechannelized features had an initial NE^SW orientationthatwas later diverted towards the east (an almost 901 shift)near East Prospect Ridge (Fig. 5c). Additional channel sys-tems were also developed during this time to the south ofCallicore and Helliconius Ridges. The modern sea£oorshows the latest expression of this levee-channel complex(Mize et al., 2004). Today, hemipelagic sedimentation isdominant in the area, as well as local collapses from mud-volcano ridges and individual mud volcanoes. Occasionalcollapses on the shelf edge and upper-slope region couldtrigger deposition of gravity-induced deposits.

Boettcher et al. (2003) pointed out in a regional studythat covered the SE corner of the Caribbean (includingour study area), that the deep-water basin of eastern o¡-shoreTrinidad had been developed as a product of tectonicsubsidence in response to emplacement of a fold-and-thrust belt associated with the Caribbean deformationalfront. According to this model, the origin of the NE^SW-trending mud-volcano ridges (Darien, East Prospect,Haydn, Callicore, Heliconius and Cat¢sh Ridges) (Figs 4and 5) can be related to successive thrust imbricationsgenerated by the advances of the Caribbean deformationalfront towards the SE. In this regional context, DarienRidge represents the eastern continuation of the CentralRange in onshoreTrinidad and the o¡shore equivalent ofthe fold-and-thrust belt associatedwith theCaribbean de-formational front (Fig. 1a) (Babb & Mann, 1999; Wood,2000). The progressive advances of the Caribbean defor-mational front towards the SE were probably responsiblefor the thinning and relative uplifting that was recordedby isopach maps in the north parts of the study area (Dar-ien Ridge region) (Fig. 5).These tectonic events were alsoresponsible for the SE migration of the main sedimentarydepocenters (Fig. 5).

From this summary, we can conclude that the morpho-logical variations that the mud-volcano ridges experi-enced through time played an important role inin£uencing the distribution of gravity-induced depositsin the deep-marine parts of the eastern o¡shoreTrinidadbasin.These morphological variations not only controlledthe location of sedimentary pathways that connected theouter-shelf/upper-slope regions with the deep-marineparts of the basin but also controlled the location and mi-gration of the main depocentres in the region.

SEISMIC AND LITHOLOGICALCHARACTERISTICS OF MTCs

Brami et al. (2000), using shallow 3D seismic data, charac-terized seven principal architectural elements in the most



recent deep-water sequence of o¡shore Trinidad: masstransport, con¢ned channel, levee-channel and distribu-tary-channel complexes; slope with minor channels un-di¡erentiated slopes; and mud diatremes. MTCs andlevee-channel complexes seem to be the dominant faciesin the basin. In cross-sectional seismic lines, MTCs arerecognizable by the presence of seismic facies that showlow-amplitude, semitransparent, chaotic re£ections (Fig.8c and d) (Posamentier & Kolla, 2003). However, internalto theMTC may exist regions of semi-deformed mixturesof low- and high-amplitude re£ections showing thrustand embrication geometries (Frey-Martinez et al., 2005,2006;Moscardelli et al., 2006) (Fig.9b).

The upper and lower boundaries of MTCs tend to bemarked by high-amplitude seismic re£ections (Fig. 8cand d).The lower boundary ofMTCs often shows a varietyof erosive features (Figs 6 and 8). The largest of these,termed megascours by Moscardelli et al. (2006), can showscour into the substratum that can be tens of metresdeep, several kilometres wide and hundreds of kilometreslong (Fig. 8) (McGilvery & Cook, 2003; Posamentier &Kolla, 2003; Weimer & Shipp, 2004; Moscardelli et al.,2006). These basal erosional surfaces can also containfeatures termed cat- claw scours (Moscardelli et al.,2006) or monkey ¢ngers (McGilvery & Cook, 2003),which are composed of a series of smaller, shallow, radiat-ing scours, whose apex is usually close to the downdiptermination of individual megascours. Secondary scoursand scratches (Moscardelli et al., 2006) or glide tracks(Nissen et al., 1999) are also common elements observedon the basal erosional surfaces of MTCs and are asso-ciated with out-runner blocks that most likely tooled thescours on the sea£oor. Condensed sections acting as theinitial sliding surface of MTCs can also de¢ne the basalboundary of mass-wasting events (Lee et al., 2004a)(Fig. 9b). Condensed sections can sometimes be identi-¢ed in 3D seismic data as continuous high-amplitudeseismic re£ectors (Fig.9b). In o¡shoreTrinidad, both ero -sional surfaces (Fig. 8) and condensed sections (Fig. 9b)can be used to identify the basal boundaries of MTCs.However, the upper bounding surfaces of MTCs in o¡-shoreTrinidad commonly show erosion, which is thoughtto be associated with processes along the base of overly-ing levee- channel complexes.These channel-levee com-plexes appear to preferentially deposit in depo-axescontrolled by alterations in bathymetry created by com-paction and ¢nal geometry of underlying MTC deposits(Figs 5c and 8c) (Moscardelli et al., 2006). MTCs tend tobe amalgamated, and surfaces between successive,stacked MTCs can be di⁄cult to recognize in seismic(Fig. 8) (Posamentier, 2004).

Internally, MTCs can contain thrusted blocks (Weimer& Shipp, 2004) or syndepositional thrusts (Fig. 9b) (Frey-Martinez et al., 2005, 2006; Moscardelli et al., 2006),re£ecting the ongoing deformation due to compressionassociated with deceleration of the £ow (Marr et al., 2001).Such deceleration could occur at the terminal end of deb-ris lobes (Fig. 9b) or anywhere the £ow is constricted and

slows. The presence of bathymetric barriers on the sea-£oor at the time of deposition, such as mud volcanoes(Figs 4^6) or salt diapirs, can contribute signi¢cantly tothe reduction of the cross-sectional area in regions experi-encing gravitational sedimentation (Moscardelli et al.,2006). Another distinctive internal characteristic ofMTCsis the presence of internally undeformed transportedblocks that lay directly on top of or partly submerged inthe surrounding chaotic material (Lee et al., 2004a).Theseblocks can vary in size from several hundred to severalthousand square metres across and150m thick. In contrastto these transported olistholiths, there occur intact butburied erosional remnants that are surrounded by chaoticMTC material (Moscardelli etal., 2006). InTrinidad, thesefeatures usually form in the shadow of some barrier thatinhibits MTC downslope movement, causing it to £owaround the obstacle, thus preserving the older strata lyingdownslope of the barrier. Such deposits, termed erosionalshadow remnants by Moscardelli et al. (2006), are asso-ciated closelywith physiographic barriers such as mudvol-canoes and salt diapirs.

The lithology ofMTCs generally consists of muddy se-diments showing extensive deformation, dipping beds andblocks of varying sizes (Macdonald et al., 1993; Piper et al.,1997; Shipp et al., 2004). However, it is important to pointout that rheological characteristics ofMTC sediments de-pend on the nature of the original material that failed(Shipp et al., 2004; Pickering & Corregidor, 2005). TheshallowestMTC in o¡shoreTrinidad (MTC_1) is composedof muddy sediments (Shipp et al., 2004) that were derivedfromPalaeo-Orinoco shelf-edge deltas that fed sedimentsinto canyons that impinged upon the lowstand shelf break(Carr-Brown, 1972; Diaz de Gamero, 1996; Sydow et al.,2003). Some MTCs in o¡shoreTrinidad were also gener-ated by failures in the upper slope region (MTC_2 andMTC_2.1) (Moscardelli etal., 2006).These deposits are alsothought to be composed of muddy sediments. Log re-sponses of the shallowest MTC in o¡shore Trinidad(MTC_1) show an increase in resistivity, compressional ve-locity, density and porosity (Shipp et al., 2004). Geotechni-cal measurements also indicate an increase in shearstrength, with a corresponding decrease in void ratio andwater content (Pirmez et al., 2004; Shipp et al., 2004).

GEOMORPHOLOGICAL DESCRIPTIONOF MTCs IN OFFSHORE TRINIDAD

Five separate MTCs have been examined within seismicunit VI: MTC_2, MTC_2.1, MTC_2.2, MTC_2.3 andMTC_2.4 (Fig. 6). As shown in Fig. 6, these ¢ve MTCspresent very di¡erent geomorphological characteristicsand dimensions. General descriptions of the MTCs con-tained in seismic unit VI and a brief overview of MTC_1(seismic unit VII) (Moscardelli et al., 2006) are providedin this section.



5 km

5 km

5 km

5 km

N N

N N

Phas

e 1

Hay

dn

Meg

a-C

hu

te

Cal

lico

reM

ega-

Ch

ute

Hay

dn

Meg

a-C

hu

te

Cal

lico

reM

ega-

Ch

ute

MTC

_2.1

MTC

_2

MTC

_2.3

MTC

_2.1

MTC

_2

MTC

_2.4

MTC

_2.3

Haydn Rid

ge

Haydn Rid

ge

Hay

dn

Meg

a-C

hu

te

Cal

lico

reM

ega-

Ch

ute

Cal

lico

re R

idg

e

Cal

lico

re R

idg

e

MTC

_2

MTC

_1LCC

51

3 m

NW

SE

X

X'

X'

X

Hay

dn

Meg

a-C

hu

te

Cal

lico

reM

ega-

Ch

ute

XX

'

54

2 m

MTC

_2

MTC

_1

LCC

_1

Y

Y' Y

'

Y

NW

SEY

Y'

127

0

Am

plit

ud

e −128

0 2439

5

RMS

Am

plit

ud

e

(a)

(b) (c)

(d)

Meg

a Sc

ou

r

Meg

a Sc

ou

r

East

Pro

spec

t R

idg

e

East

Pro

spec

t R

idg

e

Hel

lico

niu

s R

idg

e

LCC

_2LC

C_2

Phas

e 2

Phas

e 1

Phas

e 2

Fig.8.

Root-mean-squ

ared

(RMS)

amplitud

eextraction

map

show

inggeom

orph

ologicalexpression

inplan

view

ofattached

(MTC_2andMTC_2.1)anddetached

(MTC_2.2,MTC_2.3andMTC_2.4)

MTCsinthestud

yarea.T

hemap

isfrom

thebasalpartofseism

icun

itVI.Se

dimentary

sourceso

fattachedMTCsareassociated

withtheoutershelf/up

perslope

area,w

hereassedimentarysourcesfor

detached

MTCscanbe

linkedtothe£anksofmud

-volcano

ridges.L

CCs,leveechannelcom

plexes;M

TCs,masstranspo

rtcomplexes.



General description of MTC_1

MTC_1 is the shallowest MTC in the study area (seismicunit VII).Younger than any of theMTCs describedwithinseismic unit VI, it was described in detail by Moscardelliet al. (2006). MTC_1 was point-source fed by a single pa-laeocanyon graben nearly 200m deep that was bounded

along both margins by normal faults. The palaeocanyon,o2 km wide on its upper section (updip), widens down-slope, reaching more than 10 km on its lower end (toeof slope) (see detailed discussion in Moscardelli et al.,2006). MTC_1 covers an approximate area of 2017 km2; itsaverage length from the upper-slope area down to the eastboundary of the seismic coverage (deep basin) is

N(a)

(b)

DarienMega-Chute

HaydnMega-Chute

E. ProspectRidge

0

24395

RMS Amplitude

Phase 1

10 km

Darien Ridge

A

A'

1026 m

NESWA A'

MTC_2.1 (Dip)

mse

cTW

TT

HaydnRidge

Syndepositional Thrusts

MTC_2.1

MTC_2Haydn Ridge

MTC_2.2

Fig.9. (a) Root-mean-squared (RMS) amplitude extraction map from the basal part of seismic unit VI.MTC_2.1was fed by extrabasinalsediments that were transported downdip by the Haydn megachute. On the other hand, the sedimentary source ofMTC_2.2 wasassociatedwith the collapsed south £ank of the Darien Ridge. (b) NE^SW-oriented seismic line parallel to the depositional dip ofMTC_2.1. Basal and upper parts ofMTC_2.1are separated by a seismically prominent re£ector (condensed section) that onlaps against theHaydn Ridge.MTCs, mass transport complexes.



140 km. The estimated volume of remobilized sedimentsis 242 km3, and the maximum thickness in its core area is250m. MTC_1 presents a series of geomorphologicalfeatures that include an erosional edge that de¢nes itswestern lateral boundary, deep basal scours and pressureridges. In addition, multiple cat-claw scours, erosionalshadow remnants (ESRs), smaller scale scratch marksand imbricated thrust complexes have been reportedwith-in MTC_1 (see detailed description in Moscardelli et al.2006).

General description and evolution MTC_2

Moscardelli etal. (2006) described the main geomorpholo-gical characteristics ofMTC_2 in the upper slope area nearthe palaeoshelf break (source area). According to this de-scription, at least three consecutive collapse features canbe seen along the palaeoshelf break that have a semicircu-lar shape and average diameters of 4.5 km and average areasof 14 km2 (their Fig. 18b; Fig. 6a this work). The collapsedsediments were transported downslope through the Calli-core and Haydn megachutes as gravity-induced depositsthat were ¢nally deposited asMTC_2 andMTC_2.1 (Figs 1band 6).

The core of MTC_2 is located in the central part of thestudy area (Fig. 6).The Callicore megachute was the mainfeeder of MTC_2 (Figs 6 and 8), feeding sediment directlyand consistently from the upper-slope source into thedeep-water basin. MTC_2 covers an area of approximately626 km2; its average length from the upper-slope areadown to the east boundary of the seismic coverage (deepbasin) is 40 km. MTC_2 is likely to be near its terminus atthis eastern point, with evidence of signi¢cant thinningand lateral dispersion of the deposit. Its width varies froma minimum of 10 km in the proximal area, widening pro-gressively towards the NE, until it reaches a maximumwidth of 25 km in the distal parts of the unit. According toseismic data, the average thickness ofMTC_2 is 60m; how-ever, it is considered to be a minimumvalue of the originalthickness because of reductions in thickness caused bypost-depositional compaction and erosion. On the basisof these values, the minimumvolume of remobilized sedi-ments calculated forMTC_2 is around 35 km3.The volumeis a rough estimation that was calculated by multiplyingthe total area occupied by MTC_2 and its minimum esti-mated thickness (60m).

In the central and distal parts of the data set, the base ofMTC_2 is de¢ned by an erosional surface that contains aseries of distinctive megastratigraphic features that are ty-pical of regional MTCs (Fig. 7). These features includeerosional edges, megascours and multiple levels of terra-cing. Steep erosional edges that can reach 60m in relief de-¢ne the west and east lateral boundaries of MTC_2 (Figs 7and 8c).Various terrace levels, located in the south part ofthewestern erosional edge, suggest that several episodes ofincision took place during the emplacement of MTC_2(Figs 7 and 8c). Elongated megascours that are more than50m deep and 4 km long surround individual mud volca-

noes in the east part of the unit (Figs 6^8). These mega-scours were formed when passing debris £ows werecon¢ned by individual mudvolcanoes, causing a reductionin the £ow’s cross-sectional area and a short-term increasein the erosive power of the £ow (Marr et al., 2001;Moscar-delli et al., 2006). Similar features have been described inseveralMTCs around theworld (Lee etal., 2004a;McGilv-ery et al., 2004; Posamentier, 2004).

RMS amplitude extractions thatwere taken from di¡er-ent stratigraphic levels within MTC_2 (Fig. 8) reveal twomain phases of this unit’s evolution. Phase1 (Fig. 8a) is as-sociatedwith initial erosional processes that generated thebasal surface ofMTC_2.During this phase vertical and lat-eral erosionwas the dominant process, several megascourswere generated around individual mudvolcanoes (Fig. 8a),and successive episodes of incision were responsible forgeneration of di¡erent levels of terraces on the SWmarginof the Callicore megachute (Fig. 8a and c). Sediments weretransported from the SW to the NE, parallel to the mainaxis of Haydn and Callicore Ridges. Most sediments weredeposited between these two major mud-volcano ridges(Figs 6 and 8). However, during Phase 1 part of the sedi-ments were diverted towards the NW through a relativelow that was located between Haydn and East ProspectRidges that allowed a ‘tongue’ of sediment to spreadthrough this opening (Fig. 8a).This elongate £ow of sedi-ments is 15 km long and extends from the west margin ofthe main body of MTC_2 to its terminal end to the NW(Fig.8a).Near the main body ofMTC_2, the tongue of sedi-ments widens from 1km towards the NW to a width of4 km. In the NW termination of the tongue of sediments,a series of pressure ridges can be observed on the RMSamplitude extraction map, which are typical in terminallobes of debris £ows (Fig. 8a). Similar pressure ridgeshave been described at the terminal ends of MTCs in o¡-shore Morocco (Lee et al., 2004a), o¡shore Brunei(McGilvery & Cook, 2003), o¡shore Israel (Frey-Martinezet al., 2006) and in the Nigerian continental slope (Nissenet al., 1999).

Phase 2 represents the ¢nal stages of evolution ofMTC_2. RMS amplitude extractions were taken throughthe top part of MTC_2 (Fig. 8b) and show an increase inthe areal extent occupied by the amplitude anomalies thathave been associated with MTC seismofacies. The RMSamplitude extraction map of this interval (Fig. 8b) showshigh-amplitude lineations that diverge from one anotheras one follows them to the north and NE. Such patternsare interpreted as indicating lateral spreading of the sedi-ments in the north parts of Callicore Ridge.The ‘gateway’that allowed the ‘tongue’of sediments to move towards theNW during Phase 1 (Fig. 8a) has become inactive at thistime. The interruption of sediment supply towards theNW through the ‘sediment gateway’could have been trig-gered by a series of factors, including (1) reactivation ofHaydn Ridge and subsequent uplifting of its northeasterntermination, resulting in a closing of the gateway and/or (2)obstruction of the gateway entrance between Haydn andEast Prospect Ridges by the sediment excess deposited



during Phase 1. Of these two hypotheses, reactivation andrelative uplifting of Haydn Ridge because of migration ofthe Caribbean deformational front towards the SE (Babb& Mann, 1999; Boettcher et al., 2003) is most likely thecause of this interruption of sediment supply towards theNW(Fig. 8b).Therefore, during Phase 2 an important partof the active sedimentation that had occurred in this areashifted towards the NE, where these sediments started to¢ll up the region located to the north of Callicore Ridge(Fig. 8b). Internally, additional megascours formed be-tween the west erosional edge and individual mud volca-noes in the central part of MTC_2. Similar to themegascours that were described in Phase 1, these featureswere generated as a result of the reduction in cross-sectional area between the erosional edge and the indivi-dual mud volcanoes.This reduction caused an increase ofthe frictional forces that were a¡ecting the basal parts ofthe £ow, prompting vertical incision and developmentof drag marks from the movement of huge chunks of lithi-¢ed rock in the base of the £ow (Moscardelli et al., 2006)(Fig. 8c and d).

General description and evolution of MTC_2.1

MTC_2.1 is located in the NWcorner of the study area, be-tweenDarien andHaydnRidges (Figs 6, 9 and10a). Similarto MTC_2, sediments from the upper-slope area weretransported from the SW to the NE through the Haydnmegachute into deep-marine settings (Fig. 1b).The maingeomorphological di¡erence between this system(MTC_2.1) and MTC_2 is that the former presents a moreelongated morphology in plan view (Figs 6 and 9). Theelongate geometry of MTC_2.1 suggests a major degree oflateral con¢nement during deposition.

The area covered byMTC_2.1varies from 60 to100 km2,its average length from the upper-slope area down to theNE boundary of seismic data coverage is 29 km, and it ismore than 60m thick. Basal and upper parts of MTC_2.1are separated by a seismically prominent re£ector that iswell preserved in the downslope parts of the basin butabruptly onlaps Haydn Ridge and disappears in upslope(Fig. 9b).This re£ector is interpreted to be the seismic ex-pression of a condensed section, similar to that describedbyLee etal. (2004a) in theTejas AMTC in o¡shoreMoroc-co. The presence of the condensed section means that asigni¢cant time break occurred between deposition of thebasal and upper parts of MTC_2.1. According to Lee et al.(2004a), these condensed sections have the capacity to pro-vide the initial surfaces across which mass failure destabi-lizes and slides. The surfaces are often destroyed by theinitial erosional processes associated with MTC develop-ment, but preservation of the basal condensed sections in-dicates the predominance of laminar £ow in the basal partsof theMTCs.

RMS amplitude extractions taken from di¡erent seis-mic time intervals within theMTC_2.1unit reveal two mainstages of development.An amplitude extraction taken overthe basal part of MTC_2.1 (Phase 1) (Fig. 9a) revealed the

geomorphology ofMTC_2.1during its ¢rst stages of devel-opment. Figure 9a shows MTC_2.1 during this phase as aNE^SW-elongated body that internally presents a seriesof closely spaced pressure ridges. During this phase theunitwas approximately 20 km long and 3 kmwide and cov-ered an approximate area of 60 km2.

An amplitude extraction near the top of MTC_2.1(Phase 2) (Fig. 10a) shows that this interval has a slightlydi¡erent geomorphology. The upper deposits are moreaerially extensive than the lower deposits, and althoughits elongated geometry is preserved in the proximal (SW)parts of the deposit, this unit shows a more lobate shape inthe distal (NE) terminal locations (Fig.10a).The length ofthe upper unit (phase 2) is 38 km, showing an increase of18 km compared with the length of the underlying lowerunit (phase 1) (Fig. 9). The width of the upper depositis 3.5 km in proximal areas and 10 km in distal areas(Fig. 10a). Once again, the deposit is widening in itslate stages (upper unit, NE locations). The total areacovered by the upper unit is approximately 100 km2, andthe total remobilized volume of sediments was calculatedto be 7 km3.

The ¢rst stages of in¢lling forMTC_2.1were in£uencedgreatly by the presence of the mud-volcano-cored HaydnRidge. This ridge acted as a strong physiographic barrierthat constrained gravity £ows to theNW,within theHaydnmegachute (Fig. 9a). However, during Phase 2 the £owreached the NE terminus of this relatively con¢ned mini-basin, where more accommodation space was available,thus allowing the lateral spreading of sediments (Fig.10).

General description and evolution of MTC_2.2/MTC_2.3/MTC_2.4

These localized MTCs (MTC_2.2, MTC_2.3 and MTC_2.4)cover signi¢cantly smaller areas than the previous MTCs(MTC_1, MTC_2 and MTC_2.1) (Table 3), and they are fedby localized source areas that do not have a continuousin£ux of shelf- or slope-derived, extrabasinal sediments(Table 2). MTC_2.2 is located along the south £ank ofDarien Ridge in the NWcorner of the study area (Figs 9and 10). MTC_2.3 and MTC_2.4 are both located along thenorth £ank of Heliconius Ridge (Figs 6 and 11a) in the SEcorner of the study area. Given its closely spaced ampli-tude extractions (Fig. 8), MTC_2.3 appears to be slightlyolder than MTC_ 2.4, with MTC_2.3 appearing along the£anks of Heliconius Ridge earlier than MTC_2.4. Theareas covered by these smaller MTCs range from 11 to22 km2, and their thicknesses range from 98 to 150m.Estimated volume of remobilized sediments that col-lapsed from £anks of the mud-volcano ridges is between1and 3 km3 (Table 3).

RMS amplitude extractions guided by the P15 horizon(Fig. 7) in the basal parts of seismic unit VI revealed thatMTC_2.2 is lunate in plan view (Fig. 9), whereas MTC_2.3and MTC_2.4 have a more elongate geometry (Fig. 11).These units also have a series of common geomorphologi-cal characteristics. Seismic lines parallel the depositional



strike axis of these localized MTCs (Figs 10 and 11) andshow vertical incision and erosion a¡ecting their basalsurfaces. An RMS amplitude extraction map of the toppart ofMTC_2.3 (Fig.11a) also shows a series of concentricamplitude anomalies located at the NE termination ofthe MTC_2.3 lobe. These concentric anomalies are theplan view expression of terminal syndepositional thrusts(Fig. 11b). The seismic lines also show that the proximalpart of MTC_2.3 is composed of a series of slumps (Fig.11b). Similar syndepositional thrusts and slumps havebeen interpreted in updip and downdip parts of MTC_2.2andMTC_2.4 (Figs 10b and11d).

Sediments that form these localized MTCs are col-lapsed sediments derived from the £anks of mud-volcanoridges. Extensional forces associatedwith gravitational in-stabilities that a¡ected the upper parts of the £anks of themud-volcano ridges caused the initial slumping that trig-gered formation of these units. The collapse materialsmoved downslope and reached a transitional state where£ow transformations started to occur. Decrease in slopeangle, loss of cohesiveness and lack of space to accommo-date the incoming volume of sediments eventually causeddeceleration of £ow and formation of syndepositionalthrusts that compensate for lack of space in the distal ends

Phase 2

A

A'

1025 m2500

2000

1500

DarienRidge

NW SEA A'

msecTWTT

MTC_2.1

MTC_2

MTC_2.2

(a)

(b)

Fig.10. (a) Root-mean-squared (RMS) amplitude extraction map near the top of theMTC_2.1.The upper part ofMTC_2.1 is moreaerially extensive than its lower part.The elongated geometry ofMTC_2.1 is preserved in the proximal part of the deposit, but it shows amore lobate shape in its distal termination. (b)NW^SE seismic line parallel to the depositional dip ofMTC_2.2.The line shows slumpedmaterial that failed downslope ^ the top half of the unit is composed of low-amplitude, chaotic re£ectors, but the internal character ofthe lower half of the unit shows small fragments of continuous high-amplitude re£ectors that appear to be downsloping onto theP15MTC_2 basal surface.This seismic character and geometry is typical of slumps.MTCs, mass transport complexes.



of theseMTCs (Farrell,1984; Farrell &Eaton,1988;Martel,2004). The £ow direction followed by the collapsed sedi-ments that form these localized MTCs was controlledmainly by the geometry and changes in slope gradientsthat occurred in the £anks of the mud-volcano ridges.

NEW CLASSIFICATION SYSTEMFORMTCs

Defining the classification system

The newclassi¢cation system that is used herein takes intoaccount a series of factors, including (1) geomorphologicaland morphometric dimensions of the MTCs, (2) causalmechanisms, (3) location of their source area and (4) rela-tionship of theMTCs to their source area.This classi¢ca-tion system de¢nes two categories: attached MTCsand detached MTCs. Attached MTCs have also beensubdivided into two subcategories: shelf attached andslope attached (Tables 2 and 3).

In terms of scales anddimensions, regionally extensive at-tachedMTCs can occupy hundreds to thousands of square

kilometres in area and tens of kilometres in width andlength. DetachedMTCs are smaller; these deposits are lessthan tens of square kilometres in area and only a few kilo-metres in width and length (Tables 2 and 3). In this paper,length is a measurement based on the total area that hasbeen a¡ected by the mass wasting event, and it represents apossible maximum value for run-out distance (given theirareal extension and complexity, di¡erent parts of the £owwithin a single MTC event can present di¡erent run-outdistances) (Fig. 12). The thickness of MTCs is not a usefulparameter fordi¡erentiating between attached anddetachedsystems because our data show that both types ofMTCs canreach thicknesses in excess of 100m (Table 3). However, thesediment volume associated with detached MTCs does notseem to exceed10km3 (Table 3).The length/width (L/W) ra-tio of these deposits can also be used to di¡erentiate at-tached from detached MTCs. According to our data, L/Wratios44 are characteristic of attachedMTCs, and L/Wra-tioso4 indicate detachedMTCs (Tables 2 and 3).

Attached MTCs are triggered by regional events thathave the capacity to destabilize outer-shelf/upper-slopeareas, causing massive collapses that are responsible for

Table 2. Classi¢cation, causal mechanisms, and source areas ofMTCs

Classi¢cation Causal mechanisms Source area

AttachedMTCsArea5100s to1000s km2

Width and Length510s km

Shelf-attachedMTCMTC_1

Relative sea-level £uctuationsHigh sedimentation rates

Palaeoshelf edge deltas

Slope-attachedMTCMTC_2MTC_2.1

Tectonism (earthquakes)VolcanismGas hydrate dissociationLongshore currentsStorms and hurricanes

Upper-slope collapses (mega-cookie bites)

DetachedMTCsArea510s km2

Width and length5 few km

MTC_2.2MTC_2.3MTC_2.4

Gravitational instabilities on the£anks of mud-volcano ridges/saltmasses and levee-channelcomplexes:Tectonic pulsesOversteepeningMud-volcano activity/methanerelease

Flanks mud-volcano ridges/saltdiapirs/leveesFlanks of collapsing mid-oceanicridges

MTCs, mass transport complexes.

Table 3. Classi¢cation and morphometry ofMTCs in o¡shoreTrinidad

Classi¢cation MTC Area (km2) Length (km) Width (km) Thickness (m) Volume (km3)n Length/width

AttachedMTCsShelf-attachedMTC MTC_1 2017 140 25 o250 �242 5.6Slope-attachedMTC MTC_2 626 40 10^25 460 435 4^1.6

MTC_2.1 60^100 20^38 3^10 460 47 6.6^3.8

DetachedMTCs MTC_2.2 22 5.6 5.5 150 3.3 1.08MTC_2.3 21.4 8.7 3 136 3 2.9MTC_2.4 11.3 5.4 2.5 98 1 2.16

nEstimated.MTC, mass transport complexes.



860 m

XSW NE

X'

HeliconiusRidge

MTC_2.3 (Dip)

P10P10

P15P15

P4P4

MTC_2.3

MTC_2.4

X

X'

W

W'

Phase 2

Heliconius Ridge

(a)

(b)

(c)

(d)

(e)

Y

Y'

Z

Z'

860 m

552 m

P10P10

P15

MTC_2.3 (Strike)

NW W SEW'

P4

662 m

SE NWY Y'

MTC_2.4 (Dip)

Syndepositional Thrusts

P4P10

P15

683 m

NW SE

Z Z'

MTC_2.4 (Strike)

P4

P10P15

0

24395

RMS Amplitude

N

127

0

Amplitude

-128

Fig.11. (a)Root-mean-squared (RMS)amplitude extractionmapnear the top ofMTC_2.3 andMTC_2.4.Both detachedMTCs are derivedfrom the north £ank of theHeliconius Ridge. (b) NE^SW seismic line parallel to the depositional dip ofMTC_2.3. Notice the slumps onthe SWpart of theMTC and the syndepositional thrusts on itsNE termination. (c)NW^SE seismic line parallel to the depositional strikeofMTC_2.3. (d) NW^SE seismic line parallel to the depositional dip ofMTC_2.4. Syndepositional thrusts can be interpreted on its NWtermination. (e) NW^SE seismic line parallel to the depositional strike ofMTC_2.4.MTCs, mass transport complexes.



the initiation of mass-wasting processes (Table 2). Geo-morphologically speaking, shelf-attached and slope-attachedMTCs are quite similar in distal area (downslopepart of the deposits). However, their geomorphologies dif-fer greatly in the staging area near the palaeoshelf break(upslope part of the deposits). Geomorphological ele-ments common in the proximal part of shelf-attachedMTCs include palaeocanyons that funneled sedimentsdownslope, as well as aggradational or progradationalclinoforms near the palaeoshelf break that are recogniz-able in 3D seismic lines. On the other hand, regionally ex-tensive scarps that are tens of kilometres long andhundreds of metres deep de¢ne the updip boundaries ofslope-attached MTCs in the shelf-break/upper-slope re-gion. Seismic re£ections associated with the collapsedmaterials that are contained in the updip parts of slope-at-tachedMTCs tend to be discontinuous and chaotic.

Previously described geomorphological di¡erences be-tween shelf-attached and slope-attached MTCs re£ectthe di¡erent causal mechanisms that can trigger each sys-tem, as well as the variety of source areas that can be asso-ciated with them. Sedimentary sources of shelf-attachedMTCs are usually associated with shelf-edge deltas thatare controlled by sea-level £uctuations and high sedimen-tation rates. Oversteepening in the distal parts of shelf-edge deltas can trigger gravitational instabilities in theouter-shelf region that can generate debris £ows, slidesand slumps. Seasonal increments in sediment and waterdischarge a¡ecting the £uvio-deltaic systems associatedwith shelf-edge deltas can also prompt the developmentof mass-wasting events in the delta front. In contrast, se-dimentary sources of slope-attachedMTCs are associatedwith catastrophic and extensive collapses of the uppercontinental slope that are usually triggered by earth-quakes, long-shore currents, hydrate dissociation, orstrong storms and/or hurricanes. Accelerated rates of sea-

level fall can also trigger slope-attachedMTCs bydestabi-lizing gas-hydrate reservoirs on the continental slope,causing subsequent failure of the upper-slope region(Maslin et al., 2004, 2005).

Alternatively, triggering mechanisms of localized de-tachedMTCs dependmore on local gravitational instabil-ities that occur on the £anks of mud or salt masses, alongthe £anks of levees in levee-channel complexes, or on the£anks of mid-oceanic ridges.DetachedMTCs are de¢nedin this paper as deposits that are formedwhen instabilitiesthat a¡ect localized bathymetric highs trigger partial col-lapse of their £anks, generating cascading debris £ows orslumps. Most of the sediments of detachedMTCs are de-rived locally within the minibasin of deposition (Table 2).

Classification of MTCs in offshoreTrinidadandworldwide analogues

According to the MTC classi¢cation scheme proposedherein, MTC_1 (Moscardelli et al., 2006) is classi¢ed as ashelf-attached system, MTC_2 and MTC_2.1 (this paper)are classi¢ed as slope-attached systems and MTC_2.2,MTC_2.3 and MTC_2.4 (this paper) are classi¢ed as de-tached systems (Fig.13) (Tables 2 and 3).

MTC_1, MTC2 and MTC_2.1 all occupy regional exten-sive areas (from100 to 2017 km2) (Table 3), which is a char-acteristic of attachedMTCs. In addition, previous studiessuggest that MTC_1 was fed by aggradational clinoformsassociated with a Palaeo-Orinoco shelf-edge delta, whosedelta-front facies had become increasingly unstable anda¡ected by gravitational tectonics and slumping duringstill- stand conditions (Carr-Brown, 1972; Sydow et al.,2003; Moscardelli et al., 2006). According to Moscardelliet al. (2006), high sedimentation rates associated with thePalaeo-Orinoco delta are a key factor in the formation ofthe shelf-edge delta that fedMTC_1. Previous observations

N

0

0 15 km

10 mi

Erosional shadow remnants (ESR's)

Mud diapirs

Scours

Run out distance foreastern portion of the flow

SyndepositionalThrust

Run out distance forwestern portion of the flowand length of the deposit

Rafted Blocks

MTC

LEGEND

PaleocanyonEastern

Mega Scour

Fig.12. Geomorphological interpretation ofMTC_1 in o¡shoreTrinidad. Dotted line represents length of the deposit and themaximum possible run-out distance for the northern portion of the £ow. Continuous line parallel to the main axis of the palaeocanyonand to the eastern megascour represents the run-out distance of the southern portion of the £ow.MTCs, mass transport complexes.



support the classi¢cation of MTC_1 as a shelf-attachedsystem whose triggering mechanisms are associated withrelative sea-level £uctuations and high sedimentationrates (Fig. 13a). In contrast, the areally extensive nature ofscarps observed in the updip parts of MTC_2 andMTC_2.1 near the palaeoshelf break suggests that earth-quakes, caused by ongoing tectonic activity along theCaribbean plate boundary, most likely triggered multiplefailure events (Moscardelli et al., 2006). Because MTC_2and MTC_2.1 were probably triggered by an earthquake,these MTCs have been classi¢ed as slope-attached sys-tems (Fig.13b).

MTC_2.2, MTC_2.3 andMTC_2.4 are areally constrainedto the £anks of the mud-volcano ridges in o¡shoreTrinidad, and their areas do not exceed more than 22 km2.Their formation is clearly tied to gravitational instabilitiesa¡ecting the £anks of the mud-volcano ridges that are pre-sent in the area. The localized nature of these MTCs(MTC_2.2,MTC_2.3 andMTC_2.4) and their disconnectionfrom source areas located in the shelf or upper-slope re-gion were considered relevant factors that allowed us toclassify these deposits as detachedMTCs (Fig.13c).

Validation of this new classi¢cation system depends onits applicability to similar deposits around the world, aswell as to the use of di¡erent types of data sets (e.g. outcropstudies) for class identi¢cation. The eastern and westerndebris £ows in o¡shore Brazil (Maslin etal., 2005) are goodexamples of shelf-attached systems. These MTCs arethousands of square kilometres in area and hundreds ofmetres thick. Both MTCs were generated during relativesea-level rise or still- stand conditions. Increase in the se-diment discharge of the palaeo-Amazon River probablyboosted sedimentation rates in the palaeo-Amazon shelf-edge delta (Maslin et al., 2005).This increase in water andsediment discharge could have been caused by deglacia-tion of the Andes during greenhouse times (Thompsonetal., 1995;Maslin etal., 2005).The extensive areal coverageof eastern and western debris £ows in o¡shore Brazil(�10 000 km2) and their connection with the palaeo-Amazon shelf-edge delta as their primary sedimentarysource suggest that these systems can be classi¢ed asshelf-attachedMTCs (Fig.13a).

TheDeepEarthMTDandUnit R MTDalso located ino¡shore Brazil (Maslin etal., 2005) are thousands of squarekilometres in area and were deposited during icehouseperiods of rapidly falling sea level that occurred between35 and 37 ka and between 41and 45 ka, respectively (Shack-leton, 1987; Flood et al., 1995). However, the triggeringmechanisms associated with these mass-wasting eventswere not directly related to relative fall in sea level. Instead,removal of water over the shelf and subsequent reductionof hydrostatic pressure during this time caused destabili-zation of the gas-hydrate reservoirs on the continentalslope, triggering the subsequent failure of the North Bra-zilian margin (Maslin et al., 2004, 2005). Identi¢cation ofgas-hydrate dissociation as the main causal mechanismthat ¢nally triggered deposition of the Deep Earth MTDand Unit R MTD in o¡shore Brazil was the key criterion

that we used to classify both units as slope-attached sys-tems.This particular case study highlights the di⁄cultiesthat interpreters can encounter when trying to pinpointspeci¢c causal mechanisms for individual MTC events.These complications increase when several mechanismsact in a short time before the occurrence of the mass-wast-ing event.However, it is critical to make the di¡erentiationbetween shelf-attached MTCs and slope-attached MTCsbecause the former can be generated in successive events(e.g. multiple clinoform collapses over a period of time),but the latter generate as the result of a catastrophic event(e.g. earthquake).

This new classi¢cation system for MTCs shown hereincan also be applied to MTCs that have been identi¢ed inoutcrop studies. It is di⁄cult to appreciate the real spatialdimensions of MTCs in outcrops because their thick-nesses often are in excess of the dimensions of the outcrop(Kleverlaan, 1987). Moreover, compaction and erosion arealso factors that need to be considered when trying tocompare outcrop thicknesses of MTCs using ‘equivalent’measurements collected in modern continental marginsusing 3D seismic re£ection data. Despite all these ‘scale’complications, it is more accurate to identify sedimentprovenance (source area) of MTCs if outcrop studies canbe integratedwith seismic observations.

Pickering & Corregidor (2005) documented anMTC inan outcrop of the Ainsa Basin in the Spanish Pyreneesthat, according to the classi¢cation system de¢ned herein,¢ts into the shelf-attached category. This particular unitwas described as a multiphase granular £ow associatedwith extraformational material derived from the shelfand from £uvio-deltaic input. This unit contains well-rounded pebbles, shallow-marine shells and abundantreworked nummulites. The thickness of the Ainsa MTCranges from a few metres to tens of metres and may showcumulative basal erosion of up to tens of metres.The basalpart of these multiphase granular £ows is de¢ned by re-gional erosional surfaces that contain £utes and grooves(Pickering & Corregidor, 2005). This MTC exhibits inthe outcrop some of the characteristics described in thebasal parts of MTCs that have been identi¢ed in o¡shoreTrinidad, o¡shore Brunei and o¡shore Indonesia using3D seismic re£ection data (McGilvery & Cook, 2003;Posamentier & Kolla, 2003; Moscardelli et al., 2006).Identi¢cation of £uvio-deltaic input in the Ainsa MTCand geomorphological elements described in the basalpart of the unit allowed us to classify the multiphase AinsaMTC (Pickering & Corregidor, 2005) as a shelf-attachedsystem.

In o¡shoreTrinidad, detached MTCs are derived fromthe £anks of mud-volcano ridges (MTC_2.2, MTC_2.3 andMTC_2.4), although according to the new classi¢cationscheme, detached MTCs can be originated from anypalaeobathymetric high in the sea£oor (Fig. 13). In theo¡shore region of the Gulf of Mexico, detached MTCscan originate from the £anks of salt masses and from therelative bathymetric highs that de¢ne the boundaries ofdistal minibasins (Fig. 13c and d) (Posamentier, 2004;



TYPES OF DETACHED MASS TRANSPORT COMPLEXES

Slumps

Upper Slope Collapse

"Cookie Bite Features"

Debris Flow

Slope Attached Mass Transport Complex

TYPES OF ATTACHED MASS TRANSPORT COMPLEXES

Debris Flow

Slumps

Paleocanyon

Shelf Edge Delta Shelf Attached Mass Transport Complex

Mud / Salt Volcano Ridge

Debris Flow

Slides

SyndepositionalThrusts

Mini-Basin

Debris Flow

Slides

SyndepositionalThrusts

SaltSalt

Debris Flow

Channel

LeveeLevee

Levee Channel Complex

Evacuated Strata

Fluvio deltaic input

(a)

(b)

(c)

(d)

(e)

Fig.13. Schematic depiction of the di¡erent kinds ofMTCs and the processes associatedwith their genesis. (a) Slope-attached masstransport complex ^ sediments are derived from the catastrophic collapse of the upper slope area. (b) Shelf-attached mass transportcomplex sediments are provided by shelf-edge deltas and are dumped into the deep-marine basin. (c) Detached mass transportcomplex whose genesis is associatedwith the collapsing £ank of a mud-volcano ridge. (d) Detached mass transport complex whosegenesis is associatedwith oversteepening of one of the margins of a deep-water minibasin. (e) Detached mass transport complex whosegenesis is associatedwith a levee-channel complex.MTCs, mass transport complexes.



Weimer & Shipp, 2004; Montoya, 2006). Local slides asso-ciated with collapses in channel margins have also beenidenti¢ed in outcrop studies in the Spanish Pyrenees(Pickering & Corregidor, 2005); these localized units areherein classi¢ed as detached MTCs (Fig. 13e). Previouslyunknown, localized MTCs have also been documented intheMid-Atlantic Ridge at the KaneTransform using deep-tow side-scan sonar imagery (Gao, 2006); these slumps anddebris £ows are composed of gabbros, basalt and brecciasthat were derived from theMid-Atlantic Ridge.

DISCUSSION

According to Posamentier & Kolla (2003), the occurrenceof MTCs has been traditionally associated with lowstandconditions, when sedimentation on the shelf edge is sup-posed to be at its peak and water overburden weight isbeing reduced over shelf regions.These same authors pre-sented a schematic depiction of traditional relationshipsthat have been de¢ned between relative sea level anddominant mass £ow processes. According to this scheme,a typical deep-water stratigraphic succession comprises(1) debris £ow deposits at the base corresponding to theinitial period of relative sea-level fall (traditional MTCsassociated with lowstand), (2) frontal- splay-dominatedsections and then (3) leveed-channel-dominated sectionsthat correspond to the subsequent period of late lowstandand early relative sea-level rise, respectively. However, thisscheme has an addition in which the succession is ulti-mately capped by deposition of debris £ow and condensedsection deposits, corresponding to periods of rapid sea-level rise and highstand. This relatively new conceptualmodel suggests that MTCs can be deposited at any timeduring a margin’s history (Masson etal., 1997; Posamentier&Kolla, 2003). Observations made by a series of research-ers in di¡erent continental margins seem to support thisidea (Wynn et al., 2000; McMurtry et al., 2004; Maslinet al., 2005; Moscardelli et al., 2006; Sutter, 2006; Twichellet al., 2006;Wynn, 2006; Gao, 2006).

The existing MTC classi¢cation system (Nardin et al.,1979) was designed mainly from a process sedimentologi-cal perspective (Fig. 2) and cannot answer questions asso-ciatedwith potential triggering mechanisms or the overallgeological conditions thatwere predominant at the time ofMTC deposition. The new classi¢cation system that isproposed in thiswork provides an alternative classi¢cationapproach that goes beyond the traditional subdivision ofMTCs into slides, slumps and debris £ows. This newMTCclassi¢cation system de-emphasizes the role of rela-tive sea-level £uctuations as the dominant control inMTC generation and in so doing explicitly expands therange of geological conditions that can be interpreted aspotential causal mechanisms of these deposits. It insteademphasizes the relationship between mass failures andthe geomorphology of their source areas.Despite the vari-ety of triggering mechanisms that can create MTCs, thenew classi¢cation scheme avoids introduction of further

complications by narrowing classi¢cation criteria into threemain aspects: (1) sourcing regions of MTCs, (2) geomor-phological expression ofMTCs on their updip part (sourcearea) and (3) dimensions and geometries ofMTCs (Table 2).The new classi¢cation system provides a simple schemethat can be used to qualitatively determine pre-failure geo-logical conditions and triggering mechanisms for MTCs(Fig. 13). These conditions in turn have implications forthe material that comprises the mass failure deposit. Suchunderstanding is critical to assess the impact these depositsmay have as £uid reservoir or seal elements in a basin’s ¢ll.

From the oil and gas exploration perspective, shelf-attached MTCs are more likely to preserve parts ofstrati¢ed sediments in the updip parts of the system,where potential reservoirs could be encapsulated withinslides. However, preservation of reservoir quality inter-vals in the updip parts of slope-attached MTCs, wheremost of the sediments are catastrophically evacuatedand where internal stratigraphic architectures are com-pletely lost, would be very unlikely. Detached MTCs arealso unlikely to contain viable reservoirs because the ori-ginal material that fails is composed mainly of mud (e.g.mud-volcano £anks in o¡shore Trinidad) or by poorreservoir quality rocks (e.g. detached MTCs derivedfrom the Mid-Atlantic Ridge). Although both attachedand detached MTCs may not prove to be top-qualityreservoirs, the erosive processes and subsequent strati-graphic relationships surrounding MTC developmentprovide the opportunity for stratigraphic trapping tooccur (Moscardelli et al., 2006).

Additionally, determination of MTC triggering mechan-isms and pre-failure conditions is a valuable tool that can beused to assess the tsunamigenic and geotechnical hazardsthat are associated with MTC occurrences. Modelling hasshown an apparent relationship between depth atwhich fail-ure occurs, the volume of material that fails, the speed atwhich the original slide moves and the height of the tsuna-migenic waves that are generated by the mass failure event(Pelinovsky & Poplavsky, 1996; Bryant, 2001).This classi¢ca-tion scheme allows us to predictwater depth of origin.Whencombinedwith the apparent depositedvolume of theMTCs,prediction of palaeo-tsunamigenic hazards is possible.

Classi¢cation of MTCs as shelf-attached, slope-attached or detached systems allows geoscientists to de¢nethe geological context in which the deposits originate, aswell as to analyse the implications associated with theircausal mechanisms, dimensions, geomorphology andphy-siographic location. Success in the application of this clas-si¢cation system resides in the availability of data tocomplete detailed geomorphological descriptions of theMTCs, aswell as the capacity to determinewhere the sour-cing areas ofMTCs are located.

CONCLUSIONS

Three di¡erent types of MTCs have been identi¢ed inthe study area of o¡shoreTrinidad and mapped using an



extensive 3D seismic data volume. These three types in-clude (1) shelf-attached MTCs (MTC_1) (Moscardelliet al., 2006), (2) slope-attached MTCs (MTC_2 andMTC_2.1) and (3) detached MTCs (MTC_2.2, MTC_2.3and MTC_2.4). Attached MTCs are more likely to beregional units that can reach thousands of squarekilometres in area and hundreds of metres in thickness;their source areas are usually associated with extrabasinalsystems such as shelf-edge deltas and zones of collapse inthe upper-slope region. Causal mechanisms associatedwith regional attached MTCs include large-scaleearthquakes, relative sea-level £uctuations, gas-hydratedissociation, high sedimentation rates, storms and hurri-canes and longshore currents. All these elements canpotentially destabilize the upper-slope area and triggermass-wasting processes. Regional attached MTCs havebeen subdivided into shelf-attached and slope-attachedsystems. This classi¢cation was made on the basis ofdi¡erences associated with sourcing areas, geomorphol-ogy and morphometric parameters that characterize eachsystem.

Detached MTCs are smaller; they can occupy tens ofsquare kilometres in area and their source areas are usuallyconstrained within the margins of minibasins or from lo-calized bathymetric highs. Collapsed £anks of mud-volca-no ridges, unstable margins of deep-water mini basins,collapsed £anks of midoceanic ridges and oversteepenedlevees in levee-channel complexes are common sourcesof sediments for detached MTCs.These systems are trig-gered by local gravitational instabilities that can be asso-ciated with seismicity, oversteepening of slope or diapir/volcano reactivation.

The length/width ratio (L/W) ofMTCs is an importantmorphometric parameter that can be used to determinethe level of con¢nement of these deposits. L/Wratios thatare44 indicate regional attachedMTCs, whereas L/Wra-tios that are o4 are characteristic of con¢ned detachedMTCs.

Within the study area, large, buckle-fold ridges repre-sent a major control in the de¢nition of sedimentary path-ways that are responsible for funneling sediments from theouter-shelf/upper-slope area to the deep-marine basin.These ridges often show mobile muds associated withtheir core anddisplay signi¢cant mud-volcano trains alongtheir crests. In the study area, six mud-volcano ridges havebeen identi¢ed as the main palaeobathymetric featuresthat controlled sedimentation pathways in this region dur-ing the Tertiary: (1) Darien Ridge, (2) Haydn Ridge, (3)Callicore Ridge, (4) Heliconius Ridge, (5) Cat¢sh Ridgeand (6) East Prospect Ridge. Four major sedimentarypathways associatedwith these ridges have been identi¢edas main conduits of extrabasinal sediments to the deep-marine basin: (1) Darien, (2) Haydn, (3) Callicore and (4)Cat¢sh megachutes. Such ridges and intervening chutesare important components in deep-marine transport sys-tems, a¡ecting the nature of MTC run-out pathways,thickness of deposits and their links to shelf-sedimentsources.

ACKNOWLEDGEMENTSResearch in the quantitative seismic geomorphology ofdeep-marine systems is made possible through the generous mem-bers of the Quantitative Clastics Laboratory consortium, in-cluding BHP Billiton, Anadarko, BG, ChevronTexaco,ConocoPhillips, ExxonMobil, Norsk Hydro, IMP, TalismanEnergy, Noble Energy, Shell Oil and Marathon. We wouldespecially like to thank theMinistry ofEnergy andEnergy In-dustries forTrinidad and Tobago for ongoing partnership inthe study of the geology of Trinidad and Tobago.We also ap-preciate the generous donation of seismic data from Exxon-Mobil, ChevronTexaco, Shell, BP, BHP Billiton and theirpartner companies, as well as the ongoing support fromLandmarkUniversityGrant software.We also thank the Jack-son School of Geosciences and the Geology Foundation forproviding partial funding to cover publication costs, and si-milar funding provided by theDirector of theBureau of Eco-nomicGeology.Thisworkwas possible thanks to the excellenttechnical support providedby theComputing andMediaDe-partments of the Bureau of Economic Geology. Publicationauthorized by theDirector, Bureau of EconomicGeology.

REFERENCESAksu, A.E. &Hiscott, R.N. (1989) Slides and debris £ows on the

high-latitude continental slopes ofBa⁄nBay.Geology,17,885^888.Babb, S. &Mann, P. (1999) Structural and sedimentary devel-

opment of a neogene transpressional plate boundary betweenthe Caribbean and South American Plates inTrinidad and theGulf of Paria. Sediment. BasinsWorld, 4, 495^557.

Boettcher, S.S., Jackson, J.L., Quinn, M.J. & Neal, J.E.(2003) Lithospheric structure and supracrustal hydrocarbonsystems, O¡shore EasternTrinidad.AAPGMem., 79, 529^544.

Bouriak,S.,Vanneste,M.&Saoutkine,A. (2000) Inferredgas hy-drates andclaydiapirs near the storegga slide on theSouthern edgeof theVoring Plateau, O¡shoreNorway.Mar.Geol., 163,125^148.

Brami,T.R., Pirmez, C., Archie, C.,Holman,K.W.P., Slatt,R.M., Coleman, J., Rosen, N.C., Nelson, H., Bouma,A.H.,Styzen,M.J.&Lawrence,D.T. (2000)LatePleistocenedeep-water stratigraphy and depositional processes, O¡shoreTrinidad andTobago.Deep-WaterReservoirs of theWorld,GulfCoast Section SEPM20th Annual Research Conference.

Bryant, E. (2001) Tsunami; The Underrated Hazard. CambridgeUniversity Press, Cambridge.

Campbell,D.C.,Shimeld, J.W.,Mosher,D.C.&Piper,D.J.W.

(2004) Relationships between sediment mass-failure modesand magnitudes in the evolution of the Scotian slope, O¡shoreNova Scotia. O¡shoreTechnology Conference, Houston,TX.

Carr-Brown,B. (1972)The holocene/pleistocene contact in theo¡shore area east ofGaleota Point,Trinidad,West Indies.Trans.Caribb. Geol. Conf., 6, 381.

Chen, Q. & Sidney, S. (1997) Seismic attribute technology forreservoir forecasting and monitoring. The Leading Edge, 16,445^450.

Dawson, A.G., Lockett, P. & Shi, S. (2004) Tsunami hazardsin Europe. Environment Int., 30, 577^585.

Diaz de Gamero,M.L. (1996) The changing course of the Ori-noco river during the Neogene: a review. Palaeogeogr. Palaeocli-matol. Palaeoecol., 123, 385^402.

Dott,R.H. (1963)Dynamics of subaqueous gravity depositionalprocesses.AAPGBull., 47, 104^128.



Embley, R.M. & Jacobi, R. (1986) Mass Wasting in the WesternNorth Atlantic. Geological Society of America, Boulder, CO.

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.Mar. Geol., 130, 281^292.

Farrell, S.G. (1984) A dislocation model applied to slumpstructures, Ainsa Basin, South Central Pyrenees. J. Struct.Geol., 6, 727^736.

Farrell, S.G. & Eaton, S. (1988) Foliations developed duringslump deformation of Miocene Marine Sediments, cyprus. J.Struct. Geol., 10, 567^576.

Flood, R.D., Piper, D.J.W., Klaus, A., Burns, S.J., Busch,W.H., Cisowski, S.M., Cramp, A., Damuth, J.E., Goni,M.A., Haberle, S.G., Hall, F.R., Hinrichs, K.-U.,Hiscott, R.N.,Kowsmann, R.O.,Kronen, J.D., Long,D.,Lopez, M., McDaniel, D.K., Manley, P.L., Maslin,M.A., Mikkelsen, N., Nanayama, F., Normark, W.R.,Pirmez, C., dos Santos, J.R., Schneider, R.R., Showers,W.J., Soh,W. & Thibal, J. (1995) Proceedings of the OceanDrilling Program, Part A: Initial Reports, 155.

Fryer, G.J.,Watts, P. & Pratson, L.F. (2004) Source of thegreat tsunami of 1April 1946; a landslide in the upper aleutianforearc.Mar. Geol., 203, 201^218.

Frey-Martinez, J.,Cartwright, J.&Hall,B. (2005) 3D seis-mic interpretation of slump complexes: examples from thecontinental margin of Israel.Basin Res., 17, 83^108.

Frey-Martinez, J., Cartwright, J. & James, D. (2006) Fron-tally con¢nedversus frontally emergent submarine landslides:a 3D seismic characterization.Mar. Petrol. Geol., 23, 585^604.

Gao,D. (2006)Gravitational sliding on theMid-Atlantic ridge atthe kane transform: implications for submarine basin-slopedegradation and deformation.AAPGBull., 90, 159^176.

Garciacaro, E.J. (2006) Stratigraphic architecture and basin ¢llevolution of a plate margin basin, Eastern O¡shoreTrinidad andVenezuela.Master’sThesis,TheUniversity of Texas at Austin.

Gee,M.J.R.,Gawthorpe,R.L.&Friedmann,S.J. (2006)Trig-gering and evolution of a Giant Submarine Landslide, O¡-shore Angola, revealed by 3D seismic stratigraphy andgeomorphology. J. Sediment. Res., 76, 9^19.

Gee,M.J.R.,Masson, D.G,Watts, A.B. & Allen, P.A. (1999)TheSaharan debris £ow: an insight into the mechanics of longrunout submarine debris £ows. Sedimentology, 46, 317^335.

Gee, M.J.R.,Watts, A.B., Masson, D.G. & Mitchell, N.C.

(2001) Landslides and the evolution of El hierro in the CanaryIslands.Mar. Geol., 177, 271^293.

Haflidason, H.L., Reidar Sejrup, H.P., Forsberg, C.F. &Bryn, P. (2005) The dating and morphometry of the storeggaslide.Mar. Petrol. Geol., 22, 123^136.

Heppard, P.D., Cander, H.S. & Eggertson, E.B. (1998) Ab-normal pressure and the occurrence of hydrocarbons in O¡-shore EasternTrinidad,West Indies.AAPGMem., 70, 215^246.

Hoffman, J.S.,Kaluza,M.J.,Griffiths,R.,McCullough,G.,Hall, J. &Nguyen,T. (2004) Addressing the challenges in theplacement of sea£oor infrastructure on the east breaks slide -a case study:The Falcon Field (Eb 579/623): Northwestern GulfofMexico. O¡shoreTechnology Conference, Houston,TX.

Holmes, R., Long, D. & Dodd, L.R. (1998) Large-scale deb-rites and submarine landslides on the Barra Fan,West of Brit-ain.Geol. Soc. Spec. Publ., 129, 67^79.

Jansen,E.,Befring,S.,Bugge,T.,Eidvin,T.,Holtedahl,H.

& Sejrup, H.P. (1987) Large submarine slides on the Norwe-gian Continental Margin; sediments, transport and timing.Mar. Geol., 78, 77^107.

Kleverlaan,K. (1987)GordoMegabed; a possible seismite in aTortonian Submarine Fan,Tabernas Basin, Province Almeria,Southeast Spain. Sediment. Geol., 51, 165.

Knutz, P.C., Austin,W.E.N. & Jones, E.J. (2001) Millenial-scale depositional cycles related to British ice sheet variabilityand North Atlantic Paleocirculation since 45Kyr B.P., BarraFan, U.K.Margin. Paleoceanography, 16, 53^64.

Laberg, J.S.&Vorren,T.O. (2000) TheTraenadjupet slide, O¡-shore Norway; morphology, evacuation and triggering me-chanisms.Mar. Geol., 171, 95^114.

Laberg, J.S.,Vorren,T.O.,Dowdeswell, J.A.,Kenyon,N.H.

&Taylor, J. (2000)TheAndoyaSlide and theAndoyaCanyon,North-Eastern Norwegian-Greenland Sea. Mar. Geol., 162,259^275.

Lastras,G.,DeBlasio,F.V.,Canals,M.&Elverhoi,A. (2005)Conceptual and numerical modeling of the big ‘95 debris £ow,WesternMediterranean Sea. J. Sediment. Res., 75, 784^797.

Lee,C.,Nott, J.A.,Keller,F.B.&Parrish,A.R. (2004a) Seis-mic Expression of the Cenozoic Mass Transport Complexes,DeepwaterTarfaya-Agadir Basin, O¡shore Morocco. O¡shoreTechnology Conference, Houston,TX.

Lee, H.J., Normark,W.R., Fisher, M.A., Greene, G., Ed-wards, B.D. & Locat, J. (2004b) Timing and extent of sub-marine landslides in Southern California. O¡shoreTechnology Conference, Houston,TX.

Macdonald, D.I.M., Moncrieff, A.C.M. & Butterworth,P.J. (1993)Giant slide deposits from aMesozoic fore-arc basin,Alexander Island, Antarctica.Geology, 21, 1047^1050.

Marr, J.G.,Harff, P.A., Shanmugam,G.& Parker,G. (2001)Experiments on subaqueous sandy gravity £ows: the role ofclay and water content in £ow dynamics and depositionalstructures.GSABull., 113, 1377^1386.

Martel, S.J. (2004) Mechanics of landslide initiation as a shearfracture phenomenon.Mar. Geol., 203, 319^339.

Maslin, M., Owen, M., Day, S. & Long, D. (2004) Linkingcontinental-slope failures and climate change: testing theclathrate gun hypothesis.Geology, 32, 53^56.

Maslin,M.,Vilela, C.,Mikkelsen,N. &Grootes, P. (2005)Causes of catastrophic sediment failures of the Amazon fan.Quatern. Sci. Rev., 24, 2180^2193.

Masson,D.G.,Canals,M.,Alonso,B.,Urgeles,R.&Huh-

nerbach,V. (1998) The canary debris £ow; source area mor-phology and failure mechanisms. Sedimentology, 45, 411.

Masson,D.G.,van Niel, B. &Weaver, P.P.E. (1997) Flow pro-cesses and sediment deformation in the canary debris £ow onthe NwAfrican Continental Rise. Sediment. Geol., 110, 163^179.

McAdoo,B.G.,Pratson,L.F.&Orange,D.L. (2000)Submar-ine landslide geomorphology, U.S. Continental slope. Mar.Geol., 169, 103.

McGilvery,T.A. & Cook,D.L.R. (2003) The in£uence of localgradients on accommodation space and linked depositionalelements across a stepped slope pro¢le, O¡shore Brunei. In:Shelf Margin Deltas and Linked Down Slope Petroleum Systems:Global Signi¢cance and Future Exploration Potential: Gulf CoastSection SEPM 23rd Annual Research Conference (Ed. by RobertsH.H., Rosen N.C., Filon R.H. & Anderson J.B.), pp. 387^419.SEPM,Houston,TX.

McGilvery,T.A.,Haddad,G.&Cook,D.L. (2004)Sea£oor andshallow subsurface examples of mass transport complexes, O¡-shore Brunei. O¡shoreTechnology Conference, Houston,TX.

McMurtry, G.M., Watts, P., Fryer, G.J., Smith, J.R. &Imamura,F. (2004)Giant landslides,Mega-Tsunamis, andPa-leo-Sea Level in theHawaiian Islands.Mar.Geol., 203, 219^233.



Mize,K.L.,Wood,L.J.&Mann,P. (2004)Controls on the mor-phology and development of deep-marine channels, EasternO¡shore Trinidad and Venezuela. AAPG Annu. Meeting Pro-gram, 13, A98.

Montoya, P. (2006) Salt tectonics and sequence-stratigraphichistory of minibasins near the Sigsbee Escarpment, Gulf ofMexico. PhDThesis,The University of Texas at Austin.

Moscardelli,L.,Wood,L.&Mann,P. (2006)Mass-transportcomplexes and associated processes in the O¡shore Area ofTrinidad and Venezuela.AAPGBull., 90, 1059^1088.

Nardin, T.R., Hein, F.J., Gorsline, D.S. & Edwards, B.D.(1979) A review of mass movement processes, sediment, andacoustic characteristics and contrasts in slope and base-of-slope systems versus canyon-fan-basin £oor systems. Spec.Publ. ^ Soc. Economic Paleontologists Mineralogists, 27, 61^73.

Newton, C.S., Shipp, R.C., Mosher, D.C. & Wach, G.D.

(2004) Importance of mass transport complexes in the qua-ternary development of theNile Fan, Egypt. O¡shoreTechnol-ogy Conference, Houston,TX.

Nisbet, E. & Piper, D.J.W. (1998) Giant submarine landslides.Nature (London), 392, 329^330.

Nissen, S.E.,Haskell,N.L., Steiner,C.T. &Coterill,K.L.

(1999)Debris £owoutrunner blocks, glide tracks, andpressureridges identi¢ed on theNigerianContinental Slope using 3-Dseismic coherency.The Leading Edge, 18, 595^599.

Norem,H., Locat, J. & Schieldrop, B. (1990) An approach tothe physics and the modeling of submarine £owslides. Mar.Geotechnol., 9, 93^111.

Normark,W.R. (1974) Ranger submarine slide, Northern Sebas-tianVizcainoBay,BajaCalifornia,Mexico.GSABull., 85, 781^784.

Pelinovsky,E.&Poplavsky,A. (1996)Simpli¢edmodel of tsunamigeneration by submarine landslides.Phys. Chem. Earth, 21, 13^17.

Pickering,K.T.&Corregidor, J. (2005)Mass-transport com-plexes (mtcs) and tectonic control on basin- £oor submarinefans, Middle Eocene, South Spanish Pyrenees. J. Sediment.Res., 75, 761.

Piper, D.J.W., Pirmez, C., Manley, P.L., Long, D., Flood,R.D., Normark,W.R., Showers,W.J., Flood, R.D., Piper,D.J.W., Klaus, A., Burns, S.J., Busch, W.H., Cisowski,S.M., Cramp, A., Damuth, J.E., Goni, M.A., Haberle,S.G., Hall, F.R., Hinrichs, K.-U., Hiscott, R.N., Kows-

mann, R.O., Kronen, J.D., Long, D., Lopez, M., McDa-

niel, D.K., Manley, P.L., Maslin, M.A., Mikkelsen, N.,Nanayama, F., Normark,W.R., Pirmez, C., dos Santos,J.R., Schneider, R.R., Showers,W.J., Soh,W. &Thibal, J.(1997)Mass-transport deposits of the Amazon fan.Proc. OceanDrilling Progr. Sci. Res., 155, 109^146.

Pirmez, C.,Marr, J., Shipp, C. & Kopp, F. (2004) Observationsand numerical modeling of debris £ows in the Na Kika Basin,Gulf ofMexico. O¡shoreTechnologyConference,Houston,TX.

Popenoe,P.,Schmuck,E.A.&Dillon,W.P. (1993)The cape fearlandslide; slope failure associatedwith salt diapirism and gas hy-drate decomposition. In: Submarine Landslides; Selected Studies inthe U.S. Exclusive Economic Zone. U.S. Geological Survey Bulletin(Ed. byW.C. Schwab,H.J. Lee &D.C.Twichell), pp. 40^53.

Posamentier, H. (2004) Stratigraphy and geomorphology ofdeep-water mass transport complexes based on 3D seismicdata. O¡shoreTechnology Conference, Houston,TX.

Posamentier,H.W.&Kolla,V. (2003) Seismic geomorphologyand stratigraphy of depositional elements in deep-water set-tings. J. Sediment. Res., 73, 367^388.

Shackleton, N.J. (1987) Oxygen isotopes, ice volume and sealevel.Quatern. Sci. Rev., 6, 183^190.

Shanmugam, G. (2000) A 50 years of the turbidite paradigm(1950s^1990s); deep-water processes and facies models; a criti-cal perspective.Mar. Petrol. Geol., 17, 285^342.

Shanmugam,G. (2002) Ten turbidite myths. Earth-Sci. Rev., 58,311^341.

Shanmugam,G.,Famakinwa,S.B.,Hodgkinson,R.J. Blun-dell, L.C. & Anonymous (1997) Deep-marine slump anddebris- £ow dominated reservoirs of the Za¢ro Field, O¡shoreEquatorial Guinea.AAPGBull., 81, 1411.

Shipp,C.,Nott, J.A.&Newlin, J.A. (2004). Physical character-istics and impact of mass transport complexes on deepwaterjetted conductors and suction anchor piles. O¡shoreTechnol-ogy Conference, Houston,TX.

Sullivan, S., Wood, L.J. & Mann, P. (2004) Distribution,Nature and origin of mobile mud features O¡shoreTrinidad.Salt-Sediment Interactions and Hydrocarbon Prospectivity:Concepts, Applications, andCase Studies for the 21st Century,Houston,Texas, Gulf Coast Section SEPM.

Sutter, J.R. (2006) Shelf margin systems: interface betweenshallowwater sediment sources and deepwater sinks. ExternalControls on DeepWater Depositional Systems: Climate, Sea-Level and Sediment Flux, London, UK.

Sydow, J.C., Finneran, J. & Bowman, A.P.R. (2003) Stackedshelf-edge reservoirs of the Columbus Basin, Trinidad,WestIndies. In: Shelf Margin Deltas and Linked Down Slope PetroleumSystems: Global Signi¢cance and Future Exploration Potential: GulfCoast Section SEPM 23rd Annual Research Conference (Ed. byH.H. Roberts, N.C. Rosen, R.H. Fillon & J.B. Anderson), pp.411^465.

Thompson, L.G., Mosley-Thompson, E., Davis, M.E., Lin,P.N., Henderson, K.A., Cole-Dai, J., Bolzan, J.F. & Liu,K.B. (1995)Late glacial stage and holocene tropical ice core re-cords from huscaran, peru. Science, 269, 46^50.

Twichell, D.C., Schwab,W.C., Nelson, C.H. & Kenyon,N.H. (2006) The e¡ects of submarine canyon and proximalfan processes on the depositional systems of the distalMissis-sippi fan. External Controls on DeepWater Depositional Sys-tems: Climate, Sea-Level and Sediment Flux, London, UK.

vanWeering,T.C.E.,Nielsen,T.,Kenyon,N.H.,Akentieva,K.&Kuijpers,A.H. (1998) Large submarine slides on the NeFaeroe continental margin.Geol. Soc. Spec. Publ., 129, 5^17.

Weimer, P.& Shipp,C. (2004)Mass transport complex: musingon the past uses and suggestions for future directions.O¡shoreTechnology Conference, Houston,Texas.

Wood, L.J. (2000) Chronostratigraphy and tectonostratigraphyof the Columbus Basin, Eastern O¡shore Trinidad. AAPGBull., 84, 1905^1928.

Wynn, R.B. (2006) Timing and relation to climate/sea level ofgiant landslides and turbidity currents on theNorthwest Afri-can continental margin, from Morocco to Senegal. ExternalControls on Deep Water Depositional Systems: Climate, SeaLevel and Sediment Flux, London, UK.

Wynn, R.B.,Masson,D.G., Stow,D.A. &Weaver, P.P. (2000)The Northwest African slope apron: a modern analogue fordeep-water systems with complex sea£oor topography. Mar.Petrol. Geol., 17, 253^265.

Manuscript received 6 January 2007; Manuscript accepted 7September 2007



Documents

BasinResearch Newclassificationsystemformasstransport ... · 11 Big 95 0 DebrisFlowEbroSpainw Pleistocene 26 Lastrasetal.(2005) 12 PeachSlide w Pleistocene 135 Holmesetal.(1998)andKnutzetal