JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 104, NO. B7, PAGES 15,227-15,245,JULY 10, 1999
Structure and evolution of the continental crust of northern
eastGreenlandfrom integratedgeophysicalstudies Vera Schlindwein 1 andWilfried Jokat AlfredWegenerInstitutefor PolarandMarineResearch, Bremerhaven, Germany
Abstract. EastGreenland's continental crustis littleexploredcompared to thegeologically related Norway.As partof a comprehensive geophysical studyof eastGreenland,theAlfred WegenerInstitutefor PolarandMarine Researchacquiredeightseismicrefractionprofilesbetween70øNand 76øN.We presenttheresultsof fourprofilescombined withmagneticandgeological data.Thedata setrevealedpronounced differences in thecrustalarchitecture northandsouthof KongOscarFjord (72øN).The crustnorthof KongOscarFjordshowsthreestructuralunits,a Caledonian,a Devonian, anda Mesozoicto Tertiary,whereasonly a Caledoniananda Devonianto Tertiaryunit canbe distinguished in thesouthernarea.A seismichigh-velocitylayerwasdiscovered in thelowercrustnorth of KongOscarFjord whichis absentin the southernarea.This high-velocitylayer coincideswith a pronounced magneticanomalyrelatedto Tertiarymagmatismandis interpretedasTertiarymagmatic underplate.By integratinggeophysicalandgeologicaldata,we showedthat a differentevolution of theareasnorthandsouthof KongOscarFjord startedafterthe Devonianextensionalcollapseof the Caledonides.Mesozoicextensionshiftedto the eastandpreservedDevonianstructuresin the northern area, whereas the Devonian crust was further thinned and weakened in the southern area.
We proposethat it allowedan easierascentof Tertiary meltsto the surfacethanthe crustnorthof Kong OscarFjord, wheremostof the meltsgot trappedin the lower crust.The model explainsthe distributionof extrusivesandintrusivesin eastGreenland'sTertiaryigneousprovince.
1.
Introduction
lack of information
on the crustal structure of the east Greenland
Caledonides.
EastGreenland's continentallithospherehasundergonea complex evolutionsinceArcheantimes[see,e.g.,Escherand Pulvertaft, 1995]. Major geologicalevents,documented in spectacular outcrops,influencedthe architectureof eastGreenland's crust:the Silurianformationof theCaledonianorogendueto thecollisionof Laurentiaand Baltica and its extensionalcollapsein Devonian times;thecontinuedextensionandbasinformationthroughout the Mesozoic;andextensiveTertiarymagmatism relatedto theIceland hotspotand the breakupof the NorthAtlantic.Theseeventsmake eastGreenlanda worthwhiletargetfor studieson lithospheric evolution.However,previousgeophysical work in eastGreenlandtbcusedon hydrocarbonexploration(see Larsen and Marcussen
Since 1990, the Alfred WegenerInstitutefor Polar andMarine Research(AWI) has been investigatingthe crustof northerneast Greenlandin a multidisciplinarygeophysicalapproach(Figure1). Seismicrefractionand gravity data were acquiredin the large fjords of east Greenland [Jokat et al., 1995, 1996]. In addition, a large-scaleaeromagneticsurvey covering northeastGreenland yields informationon regional structuralunits [Schlindweinand Meyer, 1999]. RegionalBouguergravitydata,providedby theNationalSurveyandCadastre,Denmark[e.g.,Forsberg,1986, 1991], completethe dataset.This paperfocuseson theresultsof a wideangleseismicexperimentin eastGreenlandin 1994. We present the data of this experimentand introduceseismicmodelsfor the [ 1992] for a review). Few studieswere dedicatedto the structureof crustalarchitectureof eastGreenland.The subsequent interpretathe crustonshoreof eastGreenlandand the adjacentcontinental tion of the seismicdata integratesmagneticand geologicaldata. shelf [e.g., Larsen, 1990; Weigel et al., 1995; Mandler, 1995; We developa modelfor the post-Caledonian evolutionof the conFechnerand Jokat, 1996]. tinentalcrustof eastGreenland.Emphasisin this paperis placed Knowledgeof the architectureandevolutionof the crustof east on the analysisof the mainresult,thediscoveryof a TertiarymagGreenlandis vital for modelsfor the extensionalcollapseof the maticunderplate.Jointanalysisof the seismicandgravitydataand Caledonian orogenandfor theunderstanding of theNorthAtlantic their geologicalimplementations yieldeda regionalmodelfor the Tertiaryigneousprovince[e.g.,Uptonet al., 1995;Saunderset al., Devonianextensionalcollapseof the eastGreenlandCaledonides 1997]. For example, controversialviews of the extensionalcolwhich is presentedby V. Schlindweinand W. Jokat(Extensional lapseof the Caledonianorogenin recentmodels[Milnes et al., collapseof the east GreenlandCaledonides:new insightsfrom a 1997;Reyet al., 1997;Andresenet al., 1998]resultpartlyfromthe synthesis of geoscience data, manuscript in preparation, 1999)(hereinafterreferredto as SchlindweinandJokat,manuscript I NowatDepartment of Geological Sciences, University ofDurhmn,in preparation,1999). ScienceLaboratories,Durham,England.
Copyright1999by theAmericanGeophysical Union.
2. Acquisition of SeismicData
Papernumber1999JB900101.
During cruiseArktis X/2 of R/V Polarstem in summer1994, eight seismicrefractionprofileswere shotin eastGreenlandbe-
0148-0227/99/1999JB900101 $09.00
15,227
15,228
SCHLINDWEIN AND JOKAT:CRUSTAL STRUCTURE OF EAST GREENLAND km
0
160
260 -76ø
?5ø
?4ø
-73ø
-72ø
71 ø
70 ø
i
-32 ø
.30 ø
.28 ø
69 ø
i
.26 ø
.24 ø
.22 ø
.20 ø
Figure 1. Locationof the studyareain eastGreenland.Aeromagneticdata exist for the shadedarea. Seismiclines shotduringcruiseArktis X/2 in 1994areindicatedby solidandwhitelines.ProfilesNVF andGAA wereshotduring cruiseArktis VII/3 in 1990 and extendedin 1994. The white dottedlinesshowfurtherwide-angleprofilesshotin 1990. The solidlines mark the profilespresentedhere.For profile KFJ (bold line), the receiverpositions(white dots)
areshownalongwiththeprojected profileplaneandreceiverlocations(dashedline withdots).The profilesarenamed after the fjordswhichthey follow: KFJ, KejserFranzJosephFjord; KOF, Kong OscarFjord; DKS, DicksonFjord;
GAA, Gfisefjord; andNVF, Nordve•tfjord. Scaleis validat72øN.
tween70øNand76øN(Figure1) [Jokatetal., 1995].Twoprofiles firedat anaveragespacingof about150m. Generally,theprofiles extendedformerwide-angleprofilesin the ScoresbySundarea wereshotstartingat theheadof thefjords,followingthefjordeastwhichwereacquiredduringcruiseArktisVII/3 in 1990[Jokatet warduntilseaice prevented furtherprogress. As theexperiment al., 1996].Onlyoneof theseprofiles wasusedin thisstudy,which depends onthefjordsasnaturalseaways through thesurveyarea, focuses ontheseismic refraction dataacquired in thelargefjords no straightline profilegeometry is possible. The seismic signals between 72øNand74øN,referred to in thefollowingasthe"fjord wererecorded onlandby upto 11receivers. Therecording equipregionof eastGreenland"(Figure1). ment was deployedat appropriatelocations.Seismicstationson The seismicsurveywasdesignedasa combinedland-seaseis- nunataks formthe westernendof theprofiles.Alongthefjords, micexperiment [Jokatetal., 1995].Seismic energywasgenerated reasonably flat outcropsof bedrockweresoughtfor seismometer by two 321 Bolt air gunstowedbehindR/V Polarstern.Shotswere
sites.
SCHLINDWEINAND JOKAT:CRUSTALSTRUCTUREOF EAST GREENLAND
3. Modeling Methodology P wave arrivalswerepickedfrom all seismicsections.An error wasassignedto eachpick, dependingon the noiselevel andthe reliability of the phaseidentification(0.025 - 0.25 s). The P waves
15,229
fixed. The processis time consuming,soonly a few representative nodesper layerweretested.Althoughthismethodyieldsa quantitativeestimateof absoluteerrors,it is subjectivelyassessed whether the perturbedmodel still yields an acceptablefit of the travel time data. Therefore
the absolute uncertainties
have to be consid-
comprise refractedwavesthroughtheuppercrust(Pg), themiddle eredasroughestimates. crust(P1),thelowercrust(P2),andtheuppermost mantle(Pn),and reflected arrivalsfromupper(P1P)andlower(P2P) crustalreflectors and the Moho (ProP).
Prior to ray tracing,additionalprocessing is necessary to ac- 4. Seismic Models of the Continental Crust countfor the curvedgeometryof the profiles and the combined of East Greenland land-seaseismictechnique.Figure1 illustratesthetwo-dimensionWe present the crustal structureof the fjord region of east al (2-D) approximation for theexampleof profileKFJ.A straight line wasfitted to the shotpositions. The receiverlocationswere Greenlandasderivedfrom seismicmodelsfor thekey profilesKethenprojectedintothisplane.However,thetrueshot-receiver dis- jser Franz JosephFjord (KFJ), DicksonFjord (DKS), and Kong tances, that is, the offsets of the observedP wave arrivals, re- OscarFjord (KOF) (Figure2). Detailson furtherseismicprofiles mainedunchanged. For a laterallyinhomogeneous subsurface, this of cruise Arktis X/2 (Figure 1) are included in the thesis by straightline approximation involveserrors.Owing to the curved Schlindwein[ 1998]. The followingdescriptionfocuseson the 375 geometryof the profile,the seismicraysdo not travelin a single km long E-W trendingprofile KFJ asa representativeexamplefor the crustalstructureof the main studyarea.For comparisonof the plane.Henceaveragingof crustalstructurewill result. crustal structureof the fjord regionandthe ScoresbySundarea,the In the 2-D approximation, shotsin the fjord andreceiverson seismic profile G•tsefjord(GAA) [Mandiet, 1995] wasremodeled land cometo lie in oneplane(Figure 1). A layer of waterneedsto including the newly acquireddata(Figure2). be modeledbeneaththe shots,and a layer of bedrockneedsto be modeledbeneaththe receivers.Bothconditions cannotbe implementedsimultaneously in a 2-D model.To solvethis conflict,the receiversin the modelwerelocatedat thefjord bottom,andthe observedtravel times to each receiverwere reducedaccordingly. Vertical ray incidencewas assumed,and a correctionvelocityof 5.5 km/s was chosenbasedon the averageseismicvelocityobservedfor the uppermostcrust.This assumption involveserrors. However,the traveltime errorsdueto incorrectcorrectionvelocity and nonverticalray incidenceare estimatedto be smallerthan the
typicalpickuncertainties.
4.1. Fjord Region Along profile KFJ, 11 receiverswere deployedwhich recorded high-qualityseismicsignals.Figure 3 showsexamplesof seismic sections.The majority of the seismicreceiversin the fjord region recordedsignalof thisquality.In Figure4, observedandcalculated travel time curvesare compared;Figure 5 showsthe ray tracing throughall modellayers.The 3-D effectsdueto the curvedprofile geometryandthe pronouncedMoho topographycausedifficulties. For example,the compromisepositionof theMoho slopesdoesnot satisfyall observedPmP or Pn phaseswithin the errorlimits (compare station323 in Figure 4), as theseslopeareasare especially prone to 3-D effects. Figure 6 showsthe seismicvelocity structureof the crust for profile KFJ. Three geologicaland structuralunits in the fjord region of eastGreenlandcan be distinguishedfrom W to E as fol-
The observedtravel time datawere modeledby ray tracingusing thesoftwareby Zelt andSmith[ 1992].Crustalmodelswereobtained by forward modeling proceedinglayerwise from top to bottom.The 2-D inversionwas only usedto refine the models.The intersectingprofilesin the northernpartof the surveyarea(Figure 2) werefirst modeledseparately.The modelswere comparedat the intersections,and velocitiesand boundarieswere adjustedto fit within the respectiveuncertainties.Poorly constrainedvelocities, particularlyin the lower crust,were standardizedon all profilesusing the best definedvalue. The modelsfor eachprofile were then refinedholding the adjustedparametersfixed. Signal amplitudes couldonlyqualitativelybeincorporated intothemodels.Crosstalk of the time signalchannel(Figure3) affectedmostrecordingsand considerablyhindereddeterminationof absolutesignalamplitudes or visual comparisonof syntheticand observedtrue-amplitude
The westernpart of the seismicprofile betweenkm 0 and 158 crossesthe Caledonianfold belt of eastGreenland(Figure2). Seismic velocitiesat the surfacerapidly increasefrom about5.7 krn/s to 6.0 krn/s at about 2 km depth.Between 2 and 14 km depth, a large numberof rays (Figure 5) constrainthe seismicvelocityto 6.1 + 0.1 km/s, increasingonly slowlywith depth.A few reflected arrivalsP1P markthelowerboundaryof thislayer.Generally,the Caledoniancrustis devoidof significantreflectedenergyfrom the
seismic sections.
middle and lower crust. Therefore
The programby Zelt and Smith[ 1992]providesa methodto calculatethe resolutionof the parameterswhich specifythe final velocity model. However, the resolutionstronglydependsuponthe parameterization of the model.We thereforepreferredto explicitly showthe ray coveragefor eachlayer in orderto give a transparent qualitativeestimateof the model resolution. Uncertaintiesof the boundaryandvelocitynodesare estimated followingBarton and White [ 1997]. The positionof a specificinterfacenodeis increasinglyperturbeduntil the traveltime fit of the perturbedmodel clearly deviates from that of the unperturbed model.The maximumperturbationthat allowsa comparablefit is a measureof the absoluteuncertainty[Zelt and Smith,1992]. The equivalentprocedurewasappliedto determineuncertainties of the velocitynodes,while holdingthe velocitygradientwithin thelayer
creasecontinuouslywith depthfrom about6.3 + 0.2 krn/sat 14 km depthto 6.9 + 0.2 km/s at the baseof the crust.Thesevaluesare within the rangeof a global averageof crustalvelocitiesin orogenic belts[Christensen andMooney,1995].The Moho is a prominent reflectorin the westernpart of the profile. It risestowardthe east from a maximumof 43 km to about30 km depth. The middlepart of theprofilebetweenkm 158 and210 is dominated by Devonian geology(Figures2 and 6). At km 158, Pg phasesmaybe distortedor interrupted(comparestation322, Figure 4) as the profile crossesthe WesternFault Zone which separatesthe Caledonianorogenicbelt from theDevoniansedimentary basin [Escherand Pulvertaft, 1995]. Velocities in the Devonian basinvary between5.3 km/s and5.7 km/s at the surfaceandreach 6 km/s at about5 km depth.The crustbeneaththe basinis charac-
lows:
velocities
were assumed to in-
15,230
SCHLINDWEIN AND JOKAT: CRUSTAI, STRUCTURE OF EAST GREENLAND km
0
160
260
76 ø
*
o •'
•-
.... '::':::, ..................
+.,-, •i
:5.,..,.:> ........
e
5!
e
/
m
+
o /
I
....
+
L ::.:,:..
2,
">2 5'0
160
sw
1,•0
200
o
250
Station 90302
o
NE
lO.
,,i
.,.,/." Pg •0
sw
160
1•0
260
2,•0
Station 90303
NE
lO
.•.8.
.•.8
PI.•,Y
•_ 6
PmP
IIl•j•U FL 6
ß • 4
4 ,,•, .... pg
o o
50
100
sw
150
2b0
o
2,•0
Station 90304
NE
lO
1•)0
o
sw
1•0
260
2,•0
Station 401
NE
lO
.•.8
-•-8. PmP
•_ 6
6.
p2 P•"•'•
.=E4
ß • 4
iiiiii1• IIIII
•>2,
,,,,,/•",1 IPg ,•
o
o
5'0
SW
'
.
160
1•0
260
o
2,•0
Station 90316
o
NE
lO
sw
160
'
Station
1•0
'
200
90319
'
2•0 NE
lO.
.•.8
.•.8.
•_ 6
6.
.=E4
4.
0 0
,•0
160
1,•0
200
Station 90310
SW
o
250
o
,•0
sw
NE
10
1b0
1,•0
260
Station 90323
2,•0 NE
lO.
.•.8 I-
,•0
.•.8 6
6
.=E4.
4
•>2, o o
50
sw
100
1•0
260
Station 90325
2•0 NE
o o
50
100
150
200
250
Distance X [km]
lO.
.•.8. i-
6.
=E4.
•>2,
O0
50
100
150
200
250
Distance X [km]
Figure11. Observed andcalculated P wavearrivals forprofileGAA.Notetheabsence of high-velocity P2arrivals. Prominent Proparrivalsarerecordedby moststations. SeeFigure4 for explanation of linesandsymbols.
SCHLINDWEINAND JOKAT:CRUSTALSTRUCTURE OFEASTGREENLAND
15,240
[lu] AleW0UV 0
0
[w>l]41docI
0
i
i
!
0
!
0
! ! ! ! ! ! i i i i
,,
0
i
t-
0
•
0
•
g
ii
0
0
0
0
0
0
0
[ñu]•lemOU•/
0
0
0
0
0
0
0
to
0
to
0
to
0
[w>l]41docI
to
0
•0
0
SCHLINDWEIN AND JOKAT:CRUSTAL STRUCTURE OF EAST GREENLAND
15,241
km
0
1•0
2•0 76'
75'
74'
============================== :•:
.;.
....
...,*** ;- ß
73'
.:?::;.
.
.: ............
::'2.•ii::•,:.!i,!•i.::;z.,.:i.:.;:;i!::.:.:':•fi
.
72'
.....
+
71' 4-
Sund
.......
'.,!::i?•i;•
ß-.:--.'
70'
.4::;?•.j•
!
i
-30' -28 -26' -24' -22' -20' -300
-200
- 1 O0
0
1 O0
i
-18'
200
-16' 300
MagneticAnomaly[nT] Figure 13. Positionof the seismichigh-velocitylayer (dashedline) on the seismicprofileplanes(whitelines)in the fjord regionof eastGreenland.The high-velocitylayer correlateswith a prominentnegativemagneticanomaly(arrows)andis interpreted asTertiarymagmaticunderplate. It extendsfromKongOscarFjordto Shannon. Its eastward extentis speculative. Scaleis valid at 72øN.
is probablyreachedat depths> 20 km. At leastthe upperpart of the underplatecan thereforebe expectedto lie abovethe Curie depthandhencecarrya reversedpolarityremanence.The low net magnetizationof the underplateobtainedfrom our modelmay indicatethat the underplatedoesnot consistentirelyof magmatic material,partlyliesbelowCurietemperature, or hasa Q factornot significantly greater than one, such that induced and remanent magnetizationalmostcanceleachother. The extentof the underplatedareain the fjord regionof east Greenlandcanbe derivedfrom geophysicaldata(Figure 13). Magneticandseismicdatacontrolthewesternmarginof theunderplate. The southlimit can be inferred from the magneticanomalywhich fadesout just southof Kong OscarFjord. A changein magnetic signatureof the anomalyeastof Shannon(Figure 13) on the magnetic anomalymap by Thorning[1988] may mark the northern
margin. The east extent of the magmaticunderplateis debatable. The boundaryshownin the seismicmodelshasbeententativelyinferred from the seismicdata. Magnetically,the easternborderis obscuredby a strongpositiveanomaly(compareprofile KOF on Figure 13). Southof Kong OscarFjord, the absenceof a lower crustalhighvelocitylayeron theprofileGAA andthelackof a relatednegative magneticanomalyindicatethatno detectablemagmaticunderplate formed during Tertiary magmatism. This is supported by geochemicalstudiesof Fram and Lesher[ 1997],who suggestthat the ScoresbySundbasaltsexperiencedextensivefractionationin shallowmagmachambersas representedby sill intrusionsin the JamesonLand Basin. Our data set allows no conclusionson magmaticunderplating of theareasouthof Scoresby Sundtoward68øN (Figure2), wherefractionationof magmastookplaceat 20-25 km
15,242
SCHLINDWEIN AND JOKAT: CRUSTAL STRUCTURE OF EAST GREENLAND
depth[Fram andLesher,1997],a depthrangeagreeingwell with
the Devonianbasinswas thrther thinneddestroyingthe probably
our seismicresultsfor the fjord region. The ScoresbySundarea is characterizedby significantextru-
existingDevonianMohoplateauat 30 km depth.The crustalthinningfinally led to the smoothandcontinuous Moho sloperising
siveactivity. Values of 160,000 km3 aregiven forthevolume of
from the Caledonianorogento the JamesonLand Basin,wherethe Devoniansedimentarysequenceis buried beneathMesozoicdeScoresby Sundandabout68øN(Figure2) [Saunders et al., 1997]. posits[Surlyk,1990;LarsenandMarcussen,1992]. A lineamenttollowing Kong OscarFjord (Figure 14a) repreIn addition,theareabetweenScoresbySundandKongOscarFjord mayhave hadacover ofabout 10,000 km3ofbasalts which was sentsa majorfault zonewhosepost-Devonianactivityis well doccompletely eroded[LarsenandMarcussen, 1992;Saunders etal., umentedin the sedimentaryrecord[Surlyk,1990].This weakzone 1997]. Conversely, onlyabout 24,000 km3ofbasalt reached the in the crustmay have accommodatedthe differentialmovementof surfacenorthof KongOscarFjordin theareaof themagmatic un- the two crustalblocksnorthandsouthof Kong OscarFjord. Owing to Tertiary magmaticoverprintingof the Mesozoicbaderplate[Uptonet al., 1984].Compared to theheavilyintruded Jameson LandBasin,the sedimentary basinsnorthof KongOscar sins of east Greenland, no evidence for the mechanism of Late JuFjordincludesignificantly lessTertiaryintrusives. Hencethearea rassicto Early Cretaceouscrustalthinningis visiblefor eitherarea; northof KongOscarFjordischaracterized bymagmatic underplat- however,the tectonicstyle of the southernarea differs from the resu. lting in the ing andlittle extrusive activity,whereas southof KongOscar northernareain beingmore smooth.The processes development of new rift-related structures north of Kong Oscar Fjord,extrusive activitydominates anda magmatic underplate is Fjord andthe reactivationof older structures southof Kong Oscar absentin the studyarea. Fromtheintegrated interpretation of geophysical andgeologi- Fjord remainunknown.In addition,nothingis knownaboutthenacal datawe propose a modelfor theevolution of thecontinental tureof the Kong OscarFjord fault zone,asit is not exposed. Finally, in Tertiary times, melts formedunderthe influenceof crustof eastGreenlandwhichshedssomelight on the differences in theeffectsof Tertiarymagmatism northandsouthof KongOs- the Icelandhotspotbeneaththe thinnedcontinentalcrustalonga line from at leastKangerlussuaq (68øN)to Shannon(75øN)(Figure car Fjord. 15a). Our geophysicalstudyaswell asthe geochemicalandpetrologicalstudiesby Fram andLesher[1997] andUptonet al. [ 1995] suggestthattheTertiaryigneousprovinceof eastGreenlandhasto 6. Model for Crustal Evolution of East Greenland be subdividedinto subprovinceswith different stylesof magmaIn Devonian times, the overthickenedCaledonianorogenic tism.North of Kong OscarFjord (Figure 15b), the majorityof the wedgecollapsed [McClayet al., 1986].Devonian extensional molten material becametrappedat the crust-mantleboundaryto structures arewidespread in eastGreenland (Figures2 and14a) crystallizeas a magmaticunderplate.In contrast,southof Kong and compriseextensional detachments [Hartz and Andresen, OscarFjord (Figure 15c), the melts easily ascendedthroughthe basins 1995],fault-controlled Devonian basins[LarsenandBengaard, crust,andlargevolumesof meltsintrudedthe sedimentary 1991],andextension-related graniticintrusions [HartzandAn- anderuptedasflood basalts.The abruptchangeof the characterof dresen,1995].In addition,thewesternMohoslopeandtheMoho TertiarymagmatismacrossKong OscarFjord is unlikelyto result plateau(Figure6) correlate withDevonian extensional structures from differencesin the structureof the plume.Instead,the magcorrelatewith areasof contrastingcrustalevoat the surfaceand are assumedto have evolvedduringthe exten- matic subprovinces sionalcollapse (Schlindwein andJokat,manuscript in preparation, lution.A differentialdevelopmentof the areassouthandnorthof 1999).Fromtheregionalextentof theextensional structures and Kong OscarFjord startedafterDevonianextensionalcollapseand thewesternMohoslope,theseauthorssuggest thatDevonianex- culminatedduringLate Jurassicto Early Cretaceousrifting. We tensionaffectedboththe fjord regionof eastGreenland andthe thereforesuggestthatthepreexistinglithosphericstructurecriticalScoresby Sundareaandreduced theoverthickened crustto a more ly controlledthe magmaticactivity. However, the differences in crustal structurenorth and southof stablethickness of 30-40 km (Figure14a). Kong OscarFjord whichmighthavecontrolledthebehaviorof the Thesedimentary recordof thelatePaleozoic toMesozoic basins crust toward the Tertiary meltsare difficult to determinefrom our of eastGreenland [Surlyk,1990]andourgeophysical dataallowreconstruction of thepost-Devonian extensional history (Figure14). dataset.The thicknessof the crystallinebasementunderneaththe Riftinginitiatedby theDevonian extensional collapse gradually sedimentarybasinsnorthandsouthof KongOscarFjordprobably diedoutin mid-Permiantimes.In LateJurassicto Early Cretaceous differs. From seismic reflections of the basement beneath northern times,a majorriftingeventbegan[Surlyk,1990].Northof Kong JamesonLand, Larsen and Marcussen[1992] proposedthat the OscarFjord(Figure14b),significant tectonic activitytookplace, JamesonLandbasinis up to 16 km thick andundefiainby an only andriftingshiftedto theeastwithrespect to theDevonian struc- 6 km thickcrystallinecrust.Our seismicretractiondatafor thearea turesandpreserved a Devonian Mohoat30kmdepth.Farthereast, northof Kong OscarFjord suggesta basinthicknessof about6-8 Althoughthe a newMesozoicMoho developed. BetweentheMohoplateaus, a km anda crystallinecrustof about16 km thickness. newMohoslopeformedwhichis matched at thesurface by the crustalthicknessis similarin both areas,the lithostaticpressureat faultcontactbetweenDevonianandMesozoicstrata(Figures2 and the base of the crustin the southernarea might have been lower 14b).Theneweastern Mohoslopeextends in N-Sdirection in the thanin the northernareabecauseof the largerproportionof light
tholeiitic basaltsin the Tertiary flood basaltprovincebetween
fjordregion ofeastGreenland andterminates abruptly atKongOs-
sediments in the crust. This could have facilitated the ascent of the
car Fjord.
melts into the JamesonLand Basin. In addition,tectonicactivity,
stretching, andsubsidence remainedlocalizedin theJameson Land areasinceDevoniantimes and may have weakenedthe thin crystalline crustmore thannorthof Kong OscarFjord, wherethe tectonic activity migratedand affecteda differentarea. All these factorsmayhaveaffectedthecrustin theJameson Land areasuch blockhosting theDevonian sedimentary basin.Thecrustbeneath thatit couldnot retainsignificantvolumesof melt at its base.
Southof KongOscarFjord(Figure14c),themislittleevidence for fault activityduringLateJurassic to EarlyCretaceous extension,and the JamesonLand Basinsubsided regularly[Surlyk, 1990].Riftingdid not shiftto theeastrelativeto theDevonian structures asfoundin thefjordregionbutoverprinted thecrustal
SCHLINDWEIN ANDJOKAT: CRUSTAL STRUCTURE OFEASTGREENLAND
15,243
extensional detachment Westem FaultZone(WFZ)
a
.,•"-
NJ
Lowercrustalreflector
Oscar Fjord:
•
zone offaulting
D
WFZ
-'
Profile KFJ lO
20
!
km
Caledonian Moho
30
\
/
•
,
_
.
•"•••
L0Wer crus[/U--" sozo•(:•• e '""' '••hC
40
5o
0
km
C
300
WFZ•
•,
Mesozoic
Profile GAA
Devonian
lO 20 km 30 40 50
km
300
Figure14. Model forpost-Devonian extension ineast Greenland. (a)Situation afterDevonian extensional collapse. Theposition ofthecrustal sections ABandCDrelative toKongOscar Fjordisindicated. TheKongOscar Fjordacts asa zoneof weakness accommodating thedifferential evolution northandsouthof it. (b) Post-Devonian extension northofKongOscar Fjord.Deepcrustal structures areinferred fromseismic profileKFJ.Extension shifted totheeast
andpreserved theDevonian structures. (c)Post-Devonian extension south ofKongOscar Fjord.Deepcrustal structuresareinferredfrom seismicprofileGAA. ExtensionaffectsthesamecrustalblockasDevonianextension.
7.
Conclusions
The integrated interpretation of geophysical datarevealedthe little-knownarchitectureand evolutionof the continentalcrustof
countfor this strikingdifferencein crustalarchitecture.In Devo-
niantimes,theCaledonian orogencollapsed. A steepwestdipping Mohoslopeformedin bothareas.Extension continued throughout theMesozoic.Northof KongOscarFjord,extension shiftedto the
northern eastGreenland. Seismic andgeological datasuggest that east,preservingthe DevonianMoho andforminga secondMoho thecrustin the fjordregionof eastGreenlandcanbe dividedinto threestructural unitsaffected mainlybyCaledonian, Devonian and
slope.Southof KongOscarFjord,theDevoniancrustalblockwas continuously furtherthinned,andDevonianstructures wereburied
Mesozoicto Tertiarygeological events,respectively. In contrast, seismic datafromtheScoresby Sundareashowthatonlytwostructuralunitscanbe distinguished in thisarea,namely,a Caledonian crustal blockanda significantly thinned crustal blockaffected by
beneath Mesozoicsediments. ThustheKongOscarFjordactedas majorfaultzoneduringthisdifferential evolution. Thecontrasting development musthaveresulted in differentphysical properties of thecrust.Thecrustbeneath theMesozoic basins northof KongOscarFjordtrappedatitsbasethemajorityof themeltsproduced duringTertiarymagmatic activity,whereas southof KongOscarFjord
Devonianto Tertiarygeologicalevents.The dominantfeatureof the crustal structureof east Greenland is the existenceof a seismic
high-velocitylayer in the lower crustbeneaththe Mesozoicsedimentarybasinsnorthof KongOscarFjordandits absence southof
the crustallowedan easyascentof the meltsto the surface,where largevolumesof thesemeltserupted. KongOscarFjord.The combined analysis of seismic, magnetic, Thedetection of a Tertiarymagmatic underplate in thefjordreandgeological datarevealed thenatureof theseismic high-veloci- gionof eastGreenlandandits explanation in termsof crustalevoty layer:It represents a Tertiarymagmatic underplate. Wepropose lutionyieldsanimportant contribution to theunderstanding of the a model for the crustal evolution of east Greenland based on the
distribution of extrusives in theNorthAtlanticTertiaryigneous
joint analysisof geophysical andgeologicaldatawhichcan ac-
province. Rifting along the axis between ShannonIsland and
15,244
SCHL!NDWEIN ANDJOKAT: CRUSTAL STRUCTURE OFEAST GREENLAND km
0
100
200
+ +
SH+.+." '/5ß
+
?4'
ß
+
+
,.t,
+
ß
ß
+
.
-/3ß
+ .
+ ,
+
,
+
+
+
+
,
ß•
ß c•
7o'
sco .'..• + +
.+,•
,.•
•
'• +
+ .... • .... '4 '
69'
+ .' .•+ .
....' .
+
+
+
65'
'36•'33'.30 ß -27'-24'-21 ß -18'-15'-12'-9' b
0
A
Profile KFJ
•.__Devon•an •,* Mese*zoic. • -- *"•""• • I
lO
Magmatic underplate •, •1
20
Caledonian
km
'
•'-••
30 40 50
0
km
300 floodbasalts(S of sCO)
c
,
ProfileGAA
• • ,•
zD
i•• / ß' _• '
lO
km
30
Lower crust
40 Caledonian Moho 0
Figure 15. Model forTertiary magmatism ineast Greenland. (a)Distribution ofmagmatic activity. Ruled area indi-
cates floodbasalts; shaded-hatched areaindicates intrusives in Mesozoic basins; dashed linerepresents KongOscar
Fjord fault zone (KOF); andbold oval indicates magmatic underplating. Abbreviations areasfollows: KA,Kangerlussuaq; SCO, Scoresby Sund; and SH,Shannon. Scale isvalid at72øN. (b)Model forunderplated crust north ofKong Oscar Fjord. Themajority ofTertiary melts areretained atdepth. Structures arederived from seismic profile KFJ. (c) Model fornonunderplated crust south ofKong Oscar Fjord. Significant volumes ofextrusives and intrusives formed. Structures arederivedfrom seismicprofileGAA.
of crystalline Kangerlussuaq didnotleadto continental breakup andformation basisof the currentdataset.The relativethickness and sedimentary crust and the tectonic history of the lithosphere of a continental marginbutdiedoutin Tertiarytimes.Therefore themagmatic underplate ispreserved in a continental setting andis mightbekeysto theproblem. largelyunaffected by laterprocesses. HenceeastGreenland providesan excellentnaturallaboratoryto examinemagmaticunder-
platingandtheinfluence oflithospheric setting onmagmatism. We couldmapunderplated andnonunderplated regions andrelatethem to areasof contrasting lithospheric evolution. However, thecritical parameters of crustalstructure, whichcontrolif meltscanbe retainedin magmachambers at thecrust-mantle boundary or easily ascend through thecrystalline crust,couldnotbedetermined onthe
Acknowledgments. TheNational Survey andCadastre, Denmark, is thanked forproviding thegravity data.H. Nilbold modeled seismic profile GAAduring hispractical training atAWI.WethankB.UptonandL. Lars-
enfor fruitfuldiscussions onTertiarymagmatism in eastGreenland. The excellent support of thecrewandthehelicopter pilotsof R/V Polarstern
during dataacquisition isgratefully acknowledged. Themanuscript greatly profited fromthecomments ofW. Mooney, T. Abramovitz, andananonymousreviewer.Thisis AlfredWegenerInstitutecontribution 1430.
SCHLINDWEINAND JOKAT:CRUSTALSTRUCTUREOF EASTGREENLAND
15,245
References
jord transect,in OrogenyThroughTime,editedby .I.-P.Burg m•dM. Ford, Geol. Soc.Spec.Publ., 121, 123-148, 1997. Caledonidesandtheir Andresen,A., E. H. Hartz, andJ. Void, A late orogenicextensionalorigin Rey, P., J.-P.Burg and M. Casey,The Scandinavian relationshipto the Variscanbelt, in OrogenyThroughTime,editedby for the infracrustalgneissdomesof the east GreenlinedCaledonides J.-P.Burg and M. Ford, Geol. Soc.Spec.Publ., 121, 179-200, 1997. (72-74øN), Tectonophysics, 285, 353-369, 1998. Saunders,A.D., J. G. Fitton, A. C. Kerr, M. J. Norry, andR. W. Kent,The Barton, A. J., and R. S. White, Crustal structureof the Edoras Bm•k contiNorth Atlanticigneousprovince,in LargeIgneousProvinces,editedby nentalmargin and mantle thermalanomaliesbeneaththe North Atlm•J. MahoneyandM. Coffin,Geophys.Monogr.Ser.,vol. 100,pp. 45-94, tic, J. Geophys.Res.,102, 3109-3129, 1997. AGU, WashingtomD.C., 1997. Christensen, N. I., m•dW. D. Mooney,SeismicvelocitystructureandcomSchlindwein, V., Architecture and evolution of the continental crust of east positionof the continentalcrust:A globalview, J. Geophys.Res.,100, Greenlandfrom integratedgeophysicalstudies,Rep. Polar Res.,270, 9761-9788, 1995. 1-148, 1998.
Escher,J. C., and T. C. R. Pulvertaft,Geologicalmap of Greenlined,scale 1: 2,500,000, Geol. Surv. of Denmark and Greenland,Copenhagen, Schlindwein,V., m•d U. Meyer, Aeromagneticstudy of the continental crustof northeastGreenland,J. Geophys.Res.,104, 7527-7537, 1999. 1995. Fechner,N., and W. Jokat,Seismicrefractioninvestigationson the crustal Shire, P. N., R. J. Blakely, B. R. Frost, and D. M. Fountain,Magnetic propertiesof the lower continentalcrust,in ContinentalLower Crust, structure of the western Jameson Land Basin, east Greenlined, J. GeoDevel. Geotecton.,vol. 23,editedby D. M. Fountain,R. Arculus,and phys.Res.,101, 15867-15881,1996. R. W. Kay, pp. 145-177, Elsevier,New York, 1992. Forsberg,R., Gravity measurements in JamesonLand and neighbouring Surlyk, F., Timing, style and sedimentaryevolutionof late Palaeozoicpartsof eastGreenland,Medd. omGrOnl.,Geosci.,15, 24 pp., 1986.
Forsberg,R., Gravity measurements in east Greenland 1986-88, Tech. Rep.,4, 32 pp., Nat. Surv.andCadastre,Copenhagen, 1991. Fram,M. S., and C. E. Lesher,Generationandpolybaricdifferentiationof east Greenlandearly Tertiary flood basalts,J. Petrol., 38, 231-275, 1997.
Hartz, E., and A. Andresen,Caledoniansole thrust of central east Greenland: a crustal-scaleDevonian extensionaldetachment?,Geology,23, 637-640, 1995. Jokat,W., et al., Seismicrefraction,in The ExpeditionARKTIS-X/2 qfR/V "Polarstern" in 1994, edited by H.-W. Hubberten,Rep. Polar Res., 174, 14-22, 1995.
Mesozoic extensional basins of east Greenland, in Tectonic Events
Responsible .for Britain's Oil and Gas Reserves,edited by R. F. P. HardmanandJ. Brooks,Geol. Soc.Spec.Publ., 55, 107-125, 1990. Thorning,L., Aeromagneticanomalymap of Greenland,sheet5 (69ø30'76ø00'N, 12ø-24øW), scale 1: 1,000,000, Geol. Surv. of Greenlined, Copenhagen,1988. Upton, B. G. J., C. H. Emeleus,and R. D. Beckinsale,Petrologyof the northerneast GreenlinedTertiary flood basalts:Evidence from Hold with Hope andWollastonForelined,J. Petrol.,25, 151-184, 1984. Upton, B. G. J., C. H. Emeleus,D.C. Rex, and M. F. Thirlwall, Early Tertiary magmatismin northeastGreenland,J. Geol. Soc. London, 152,
959-964, 1995. Jokat,W., et al., MarineGeophysik(in German),in TheExpeditionARKTIS-VII/3 o.fR/V "Polarstern"in 1990, editedby H. Miller and H. Weigel, W., et al., Investigationsof the eastGreenlandcontinentalmargin between70ø and 72øNby deepseismicsoundingm•dgravitystudies, Grobe,Rep. Polar Res.,189, 17-22, 1996. Mar. Geophys.Res., 17, 167-199, 1995. Larsen,H. C., The eastGreenlandShelf, in The Geologyof North America, vol. L, The Arctic Oceanregion,editedby A. Grm•tz,L. Johnson, White, R. S., Magmatismduring and after continentalbreak-up,in Magmatismand the Causes.forContinentalBreak-up,editedby B. C. Stoand J. F. Sweeney,pp. 185-210,Geol. Soc. of Am., Boulder,Colo., 1990. rey,T. Alabaster,andR. J. Pankhurst,Geol.Soc.Spec.Publ., 68, 1-16, 1992. Larsen,H. C., and C. Marcussen,Sill-intrusion,flood basaltemplacement anddeepcrustalstructure of theScoresby Sundregion,eastGreenland, White, R. S., and D. P. McKenzie, Magmatismat rift zones:The generation of volcanic continentalmargins and flood basalts,J. Geophys. in Magmatismand the Causes.fbrContinentalBreak-up,editedby B. Res., 94, 7685-7729, 1989. C. Storey,T. Alabaster,andR. J. Pankhurst, Geol.Soc.Spec.Publ.,68,
365-386, 1992.
Zelt, C. A., and R. B. Smith, Seismic traveltime inversion for 2-D crustal
velocitystructure,Geophys.J. lnt., 108, 16-34, 1992. Larsen,P.-H., andH.-J. Bengaard,Devonim•basininitiationin eastGreenlined:A resultof sinistralwrenchfaultingand Caledonim•extensional W. Jokat, Alfred WegenerInstitute for Polar and Marine Research, collapse,J. Geol.Soc.London,148, 355-368, 1991. D-27568 Bremerhaven,Germany.(jokat@awi-bremerMm•dler,H., The crustalstructureof the ScoresbySundarea, east Green- Columbusstrasse, land: Resultsfrom refractionseismicm•d gravity measurements (in haven.de) V. Schlindwein,Departmentof GeologicalSciences,Universityof German),Rep. Polar Res.,172, 1-116, 1995. McClay, K. R., M. G. Norton,P. Coney,and G. H. Davis, Collapseof the Durham, ScienceLaboratories,SouthRoad, Durham DHI 3LE, Englm•d.
Caledonian orogen andtheOldRedSandstone, Nature,323,147-149, (
[email protected]) 1986.
Milnes, A. G.,O. P.Wennberg, 0. Skfir, m•dA. G. Koestler, Contraction, (Received May18,1998;revised February 1, 1999; extension andtimingin theSouthNorwegian Caledonides: theSognef.- accepted March 1, 1999.)