GEOPHYSICAL RESEARCH LETTERS, VOL. 25, NO. 19, PAGES 3631-3634, OCTOBER 1, 1998 Direct measurement of magnetic reversal polarity boundaries in a cross-section of oceanic crust Maurice A. Tivey, • H. PaulJohnson, 2Corinne Fleutelot, 3Stefan Hussenoeder, • Roisin Lawrence, 4CherylWaters, • andBeecher Wooding • Abstract. Magnetic field measurements made by submersible define the cross-sectional geometryof a magneticpolarityreversal boundaryand the verticalvariationof crustalmagnetizationin upperoceaniccrust. Measuredpolarityboundariesshowa systematicpatternof shallowdip towardsthe spreadingaxis within the upper extrusivelavas,and steeperdip in the lower extrusive lavas. This geometryis a consequence of the emplacementof extrusivelava at a midoceanridge. Reversalboundarygeometry and magnetizationestimatesare usedto calculatethe magnetic contributionof the extrusivelava sequenceto the overlyingmarine magneticanomalysignal. From the forwardmodeling,the highly magnetizedextrusivelavas contributethe majority (5075%) of the observedsea surfacemagneticanomaly,consistent with the extrusive crust forming the primary sourcelayer for youngmarinemagneticanomalies. anomaliesprovide informationon the lateral extent of magnetization and have been used to hypothesize dipping polarity boundarieswithin the crust [Atwaterand Mudie, 1973; Blakely and Lynn, 1977; Cande and Kent, 1976; Arkani-Hamed, 1988; Wilsonand Hey, 1981], thesemeasurements are nonuniqueand provide no constraintson the vertical polarity structureof the crust. By combining magnetic field measurementsat several levels, including outcrop polarity measurements,a strong case was madefor the dip of polarity boundariestowardsthe spreading axiswithin the extrusives[Macdonaldet al., 1983]. Conceptual modelsof the geometryof a polarity boundaryprovide an interpretational framework[Kidd, 1977; Schoutenand Denham, 1979], but prior to this study,no direct observationshave been made of the three-dimensionalcross-sectionalgeometry of a magneticpolarityreversalboundaryin oceancrust. Data Collection and Analysis Introduction We used the submersibleALVIN to collect near-bottommag- A cornerstone of plate tectonictheoryis the Vine and Mat- netic profiles up a steepscarpface that exposesupper oceanic crust of Jaramillochron age (0.99-1.07 Ma [Cande and Kent, thews[ 1963] hypothesis,which statesthat coherentmarinemagnetic anomaliesare a recordof the recurringpolarity reversalsof 1995]).Thesurvey focused ontheBlar•co Scarp atthewestern Earth's magnetic field preservedin oceanic crust by seafloor end of the Blanco fracture zone (BFZ) in the northeast Pacific spreadingat midoceanridges. While the spatialvariationin (Figure 1), which offsetsJuande FucaRidgefrom GordaRidge. magneticanomalieshasbeenusedwith greatsuccess, thevertical Between0.2 and 0.4 Ma, a northwardjump of 20 to 40 km ocdistributionof magnetizationwithin the crust and the relative curred at the western end of the BFZ that resulted in the formacontributionof each stratigraphiclayer to the anomalysignalhas tion of WestBlancodeepandexposureof relativelynormal,non- remaineda topicof considerable discussion [e.g., Talwaniet al., 1979; Harrison, 1987]. Oceanic crustal drilling directly samplesthe vertical dimension, but hasbeen limited by a lack of significantbasementpenetration, poor core recovery,and ambiguousdirectionalinformation [Talwani et al., 1979]. Early drilling resultsfrom the Atlantic found multiple magneticreversalswithin the extrusivelavas 44' 30'N 130' 15'W 130' 10'W I I 130' 5'W ! suggestingsignificantoverlap of magneticreversalboundaries that reducethe effective contributionof the extrusivelayer to the overlyinganomalies[Johnsonand Merrill, 1978; Talwaniet al., 1979]. Overlappingpolarityboundaries highlightthe importance of spatialvariationsin crustalmagnetization,a circumstance for which singlehole drilling is poorly suited. While the shapeand 44' 25'N transition widths of sea surface and near-bottom magnetic •Woods HoleOceanographic Institution, Woods Hole,Massachusetts. 2Sch. ofOceanography, University ofWashington, Seattle, Washington. 3Dept. GEOTER, Universita diUdine, Udine, Italy. 4Dept. ofGeology, DukeUniversity, Durham, NorthCarolina. 5Dept. ofGeology, University ofNorthCarolina, Chapel Hill,North Carolina. Copyright1998 by the AmericanGeophysicalUnion Papernumber98GL02752. 0094-8534/98/98GL-02752505.00 -32-28-24-20-16-12 -8 -4 0 4 8 12 16 20 24 48 Figure 1. Bathymetry map of the BLANCOVIN survey area (contourinterval 250 m) with color-shadedmagnetization. Submersibledive tracksare shownby bold black lines annotatedwith dive numbers. A deep-towedmagnetic survey line (DTM) is shown by bold line along the top of the scarp. The trace of the Blanco transformfault is shownby the dashedline. 3631 3632 TIVEY ET AL.' MEASUREMENT OF MAGNETIC BLANCO SCARP - Phase shiftedTotal Magnetic Field Profiles -1.5 o• cb m o• (o >. m o i ,• • (o cu i , , , , , , , , I -3.5 -4 Figure 2. Observedtotal field magneticanomalyprofilesplotted versusdepthand phase-shifted (70ø) to removethe skewhess due to the slopingscarpgeometryandmagneticinclination. REVERSAL POLARITY BOUNDARIES can be correctedto an accuracyof-200 nT, trivial relative to -10,000 nT anomaliesthat mark the polarity boundaries. Total field was then calculatedfrom the componentdata and corrected for the 1995 InternationalGeophysicalReferenceField [IAGA, 1996] for the area. Subsequentanalysisof the ALVIN magnetic profilestreatedthe horizontallayeringof oceaniccrustasdipping 40ø relative to the observationplane and subparallelscarpface [Tivey, 1996]. This angleof the sourcebodies,treatedas a semiinfinite slabs,generatesa phaseshiftexactlyasif the observation planehadbeenhorizontalandsourceslabshadbeendipping. Inversionfor magnetizationuseda modifiedFouriertransformapproachfor the analysisof dippingtabularbodies,assumingthe sourceextendsto infinite depthwith an averagedistancefrom the sensorto the scarpface of 5 meters(Figure3) [Tivey, 1996]. A short-wavelength filter cutoffof 100 m wasusedto filter out high frequencyanomaliescontaminatedby bathymetrysources. No corrections were made for talus, which is assumed to be non- fracturezone crustby a steep(-40ø), south-facingscarp[Embley magneticdue to the randomorientationof blocks,however,talus and Wilson,1992]. The 2.3 km high BlancoScarpcleanlytrun- doesattenuatethe magneticsignalby placingthe sensorfurther cateswell-defined, highly-lineatedBrunhesand Jaramillomag- from the source.Figure3 showsthe upperlimit of the main talus netic anomaliesthat formed at the Juande Fuca spreadingcenter fan, whichis thoughtto progressively maskmoreof the extrusive (Figure 1). Submersibleobservations of the scarpfacerevealthat layer towardsthe westernedgeof the survey. The zero-levelof the exposedextrusivecrustcan be dividedinto a relativelyintact, the profilesaredefinedby adjustingtheDC-levelof the magnetiundeformedsectionof pillow lavas that overlies more massive zation contrastat the top of the scarpto obtainzero magnetizaand stronglyjointed basaltlava with cross-cuttingfeederdikes. tion for seawater. The zero-crossingsdefine the boundariesof An intrusivediabasesequenceunderliesthe extrusiverocksbut is the main positiveandreversepolaritychronsin the magnetization obscuredby talus, excepttowardsthe easternend of the study profiles (Figure 3). We have interpretedonly the main reversal area [Juteau et al., 1995]. patternsthat correlate laterally between profiles, becausesmall Magneticfield datawere obtainedusinga 3-axis fluxgatemag- local anomalies are either associated with dikes, which can have netometerattachedto ALVIN (Figure 2). A deep-towedmagne- oppositepolarity to the host extrusives,or are due to tectonic tometerprofile was also obtainedacrossthe top of the scarp,to complexitysuchas faulting. We testedfor two-dimensionalityof magneticfield tie the ALVIN profiles to the overlying sea surfaceanomalies the main anomaliesutilizing the three-component (Figure 1). The observedthree-component magneticfield data data corrected for submersible orientation [Rutten, 1975]. A were first correctedfor the magnetismof ALVIN usinga Nelder- minimum amplitudein the componentsparallel to scarpface is Meade minimization approach[Tivey, 1996]. Calibration by obtainedparallel to the strike of the assumedsourcebodiesand spinningALVIN in the water coluntoindicatesits 2000 nT field scarp azimuth (110ø). Also, a good match between the scarp- BLANCOSCARP- Magnetization Profiles -1.5 -2 E -2.5 • -3 -4 20 22 24 26 28 30 32 Distance from axis [km) Figure 3. Magnetizationprofilesversusdepthplottedat approximate lateraldistancefrom the spreadingaxis. Dive numbersrefer to ALVIN or NAUTILE (BN prefix) dives. Positivelymagnetizedcrustis shadeddark gray with profiles filled in black. Reverselymagnetizedcrustis shownin light gray and profiles are filled with white (J: Jaramilloand B: Brunheschron). Bold dashedline indicatesthe approximateboundarybetweenupperpillow lavas (UPL) and lower lava (LL) sequence.Bold solid line indicatesupper limit of talus below which the polarity units and extrusivecrust are unshadedand the boundariesand base of extrusivesdashedto indicate greaterlevel of uncertainty. 34 ,. TIVEY ET AL.: MEASUREMENT OF MAGNETIC REVERSAL POLARITY BOUNDARIES 3633 perpendicularcomponentand the transformedtrack-parallel [Johnsonand Salem, 1994]. Also, while lateral variability in componentconfirmsthat the major anomaliesare generallytwo- polarity structureis correlatableat -0.5-1 km profile spacing, dimensional perpendicular to the profiles(i.e. alongthe scarp magnetizationintensityappearsto vary on a smallerlength-scale face). Results The magneticresultsshowa remarkablycoherentpictureof the horizontaland verticalmagneticstructureof upperoceaniccrust. Magneticpolarityboundaries of the normalJaramillochronsystematicallydip towardsthe spreading axiswith a shallowdip (< 5ø) in theupper150 m of extrusivecrustandsteeperdip (30ø to 45ø) in thelowerextrusives (Figure3). The shallow-dipping polarity boundariesoccurwithin the undeformedand recognizable pillow lava section,while the steeply-dipping boundariesoccur within the more massive and tectonized lower extrusives. This suggestsa geneticlink betweenlava emplacementand polarity boundarygeometryas suggestedby others [e.g. Atwater and Mudie, 1973; Kidd, 1977; Schoutenand Denham, 1979; Mac- donaldet al., 1983]. The totalhorizontalextentof the dipping polarity boundariesfor both the young and old edge of the Jaramillochron is -3-4 km, with the largestproportion(3 km) occurringwithin the shallowupperlava sequence. Magnetizationis highly variable within the extrusive lavas, varyingfrom zero to -20 A/m, but averages-12 A/m over the entire extrusivelava thickness,consistentwith paleomagnetic rock measurements [Tivey, 1996]. Our magnetization estimates of the underlyingintrusivedikesare not well-constrained because of talus cover, but appear to be an order of magnitudeless 1000 ........ a) •' Seasurface level - • -1000 b) • andis likely to be ultimatelyof a fractalnature[Tivey, 1996]. The mean magnetizationand mapped polarity boundarygeometry(Figure 3) can be usedto model the magneticfield anomaly and estimate the contribution of the extrusive lavas to the overlying sea surfacemagneticanomalysignal. We used a twodimensional(2D) polygonroutine[Won and Bevis,1987] to calculatethe magneticanomaliesat both deeptow(-2.1 km) and sea surfacelevelsusingthe mappedgeometryand estimatedmagnetization intensity. The forward model assumessourcebodiesthat extendto infinity perpendicularto the profile. The boundariesof the Jaramillo polarity are well-definedby our submersibleprofiles, but the adjacentBrunhes anomaly boundaryis not wellknown. We used two previous profiles (BN18, BN19) from Tivey [ 1996] and assumedthe Brunhesboundaryhad a geometry similar to the Jaramillofor the forward modeling(Figure 3). We comparedthe calculatedanomalieswith the observeddeeptow anomalyand a representativesea surfacemagneticanomalyprofile 5 km northof the scarp,well awayfrom anytopographic edge effectsof the scarp(Figure4). Initial calculations,usingonly the extrusivecrustalsectionexposedabove the talus fan (Figure 4), producedanomaliesonly 50% of the observedseasurfaceamplitude. To correctfor this discrepancy,we assumedthat talusconcealsthe extrusivebaseat the westernend of the surveyand extrapolatedthe reversalboundariesto a sourcedepth of 3400 m depth, producingapprox. 75% of the amplitudeof the Jaramillo anomalyobservedat both the deeptowand sea surfacelevel and -........... ,...... , , , , I , [I ----' Observed -Model full 2A thickness 2000Near-boffomlevel-2.1 km 0 -1.5 C) ........ Mognetizotion Geometry -2 ' -3.5............ -4 ]6 I I I I I I I I 18 20 22 24 26 28 30 32 34 Distance from spreading •xis [km) Figure 4. (a) Magnetizationpolaritygeometry(Fibre 3) usedin forw•d modelof magneticfield. Magnetization is 12 Nm, J is J•a•11o andB is Bmnheschron. Computedmagneticfield for exposedlayer 2A thickness(dotted) andfull layer2A thickness(solid)comp•ed to (b) the observeddeeptowmagneticprofile at -2.1 km depth(dashed) and(c) observedseasurfacemagnetic•omaly (dashed). 3634 TIVEY ET AL.:MEASUREMENT OFMAGNETICREVERSAL POLARITYBOUNDARIES closely approximating theposition andshape of theobservedReferences anomalies (Figure 4). Thisresult indicates thattheextrusive la- Arkani-Hamed, J.,Remanent magnetization of theoceanic uppermantle, Geophys. Res.Lett.,15,48-51,1988. vas,although highlyvariable in magnetization intensity, neverT., andJ.D.Mudie,Detailed near-bottom geophysical studyof theless constitute theprimary source of theseasurface marine Atwater, GordaRise,J. Geophys. Res.,78, 8665-8686,1973. magnetic anomaly signal. Theoldedge oftheJaramillo anomalyBlakely,R.J.,andW.S.Lynn,Reversal transition widthsandfastshows a slightdiscrepancy in shape compared to theobserved spreading centers, EarthandPlanet. Sci.Lett.,33,321-330, 1977. imposed bytheshape ofmarine anomaly (Figure 4). Thismayberelated tocomplexities in the Cande,S.C.,andD.V. Kent,Constraints magnetic anomalies onthemagnetic source, J. Geophys. Res.,81, vertical magnetization pattern andfaulting atthislocation. The 4157-4162, 1976. dipofthepolarity boundaries suggests thatthelower extrusiveCande,S.C.,andD.V. Kent,Revised calibration of thegeomagnetic posection contributes thegreater proportion of themagnetic anom- laritytimescale forthelateCretaceous andCenozoic, J. Geophys. alysource because thepolarity boundaries have steeper dipand Res.,100, 6093-6095, 1995. G.L., G.M. Purdy,andG.J.Fryer,Seismicconstraints on greater effective magnetization contrast compared totheshallowChristeson, shallow crustal emplacement processes at thefastspreading EastPadipping polarity boundary intheuppermost extrusives. cificRise,J. Geophys. Res.,99, 17,957-17,973, 1994. Thechange in slope of thepolarity boundary fromshallow to Embley, R.W.,andD. Wilson, Morphology of theBlanco Transform steeply dipping andthecorresponding lithologic break between faultzone-NE Pacific:Implications for its tectonic evolution, Mar. Geophys. Res.,14, 25-45,1992. theupper andlowerlavascanbeexplained bya bimodal lava emplacement process [e.g.,Schouten andDenham, 1995; Hooft Harrison, C.G.A.,Marinemagnetic anomalies - Theoriginof thestripes, Ann. Rev. Earth Planet. Sci., 15, 505, 1987. et al., 1996]. Lavaserupted anddeposited directly on-axis are Hooft,E.E.E.,H. Schouten, andR.S.Detrick,Constraining crustal emrapidly buried andintruded bydikefeeding subsequent eruptions placement processes fromthevariation inseismic layer2Athickness andformthetectonized lowerlavasequence. Lavadeposited 1 to at the EastPacificRise,Earth and Planet.Sci.Lett., 142, 289-309, 1996. 2 kmoff-axis, either byeruption off-axis orbylateral transport of Association of Geomagnetism andAeronomy (IAGA),Dilavasfromtheaxisarerelatively undeformed andcomprise the International vision V, Working Group 8,International geomagnetic reference field, upper lavasequence. Thedeposition ofoff-axis lavas hasbeen 1995revision, Geophys. J. Int,125,318-321,1996. suggested asa mechanism forthenear-axis increase inthickness Johnson, H.P.,andR.T. Merrill,A directtestof theVine-Matthews hypothesis, EarthandPlanet. Sci.Lett.,40,263-269, 1978. oflayer2A[Christeson etal.,1994].If thismodel iscorrect, the H.P.,andB.L. Salem,Magnetic properties of dikesfromthe transition widthof theshallowdippingpolarityzone(3 km) can Johnson, oceanic upper crustal section, J. Geophys. Res.,99,21,733-21,740, beinterpreted astheapproximate maximum distance thatlavas 1994. canflowfromthespreading axisorcanbeerupted off-axis. The Juteau, T., D. Bideau, O. Dauteuil, G. Manac'h, D.D.Naidoo, P. Nehlig, lowerlavashavea narrower transition width(1 to 2 km) which reflectsthehalf-widthof the axialrift valley,consistent withthe H. Ondreas, M.A. Tivey,K.X. Whipple, andJ.R.Delaney,A submersible studyin thewestern BlancoFracture Zone,N.E. Pacific: Lithostratigraphy, magnetic structure, andmagmatic andtectonic evolution duringthelast1.6Ma, Mar. Geophys. Res.,17, 399-430, present-day JuandeFucaspreading axis[Tivey, 1994].The transition widthof thelowerlavasis alsocomparable to previous 1995. studies at similarspreading ratemidocean ridges[e.g.Atwater Kidd,R.G.W.,A modelfortheprocess of formation of theupperoceanic crust,Geophys. J. R.Astr.Soc.,50, 149-183,1977. and Mudie,1973;Macdonald et al., 1983],whilethe wider K.C., S.P.Miller, B.P. Luyendyk, T.M. Atwater,andL. transition in theupper lavassuggests a greater variability in the Macdonald, emplacement ofoff-axis flows possibly duetolocal topography. Conclusions Shure,Investigation of a VineMatthews magnetic lineation froma submersible: The sourceandcharacter of marinemagneticanomalies, J. Geophys. Res.,88, 3403-3418,1983. Rutten,K., Two-dimensionality of magnetic anomalies overIceland and Reykjanes ridge, Mar.Geophys. Res.,2, 243-263, 1975. H., andC.R.Denham, Modeling theoceanic magnetic source Submersible magnetic profiles of an exposed crustal section Schouten, showthat the horizontalvariationin crustalmagnetization as a function ofdepth canbemapped anddirectly related totheover- lyinglineated anomaly signal measured attheseasurface. Dippingpolarity boundaries withintheextrusive sequence arecom- layer,in DSDPResults in theAtlantic Ocean: Ocean Crust, Maurice EwingSer.,edited byM. Talwani, C.G.A.Hamson, D.E.Hayes, 2, pp.151-159, AGU,Washington, DC, 1979. Schouten, H., andC. Denham,Virtualoceancrust,EOSTrans.AGU, 76(17), S48, 1995. Talwani, M., C.G.A.Harrison, andD.E.Hayes,DSDPResults in theAt- patible withthelocation ofJaramillo normal polarity chron. Po- lanticOcean:OceanCrust,MauriceEwingSer.,2, AGU, Washinglarityboundaries haveashallow dipwithin theupper undeformed ton, DC, 1979. pillowed lavas anda steeper dipintheunderlying massive and Tivey,M.A.,Thevertical magnetic structure ofocean crust determined tectonized extrusive lavasequence. Fromforward models, using fromnear-bottom magnetic fieldmeasurements, J. Geophys. Res., themapped geometry andmeasured magnetization intensity, we 101, 20,275-20,296,1996. neticanomaly amplitude to account forthemajority (50-75%) of crustal structure, J. Geophys. Res.,99,4833-4855,1994. M.A.,Thefine-scale magnetic anomaly fieldovertheSouthern find thatthe extrusive lava sequence generates sufficient mag- Tivey, JuandeFucaRidge:Theaxialmagnetic lowandimplications for theobserved seasurface magnetic anomaly signal.Intriguingly, Vine,F.J.,andD.H.Matthews, Magnetic anomalies overoceanic ridges, theremaining source (25%)mustbe accounted for by deeper Nature,199, 947-949, 1963. D.S.,andR.N.Hey,TheGalapagos axialmagnetic anomaly: crustalunits,eitherwithinthedikeor gabbrosections.We con- Wilson, Evidence fortheEmperor EventwithintheBrunhes andfora two- cludethatforyoung crust(- 1 Ma),theextrusive lavasformthe layer magnetic source, Geophys. Res. Lett.,8, 1051-1054, 1981. Won,I.J.,andM.G.Bevis,Computing thegravitational andmagnetic primary source ofthemarine magnetic anomaly signal. anomalies dueto a polygon: Algorithms andFortran subroutines, Acknowledgments. We thankthecrewof Atlantis II andWHOIDSOGfor theireffortsin makingtheAlvin magneticsurveys a success. We thankD. Van PattenandD. Naidoofor theirhelpon thecruise.We Geophysics, 52, 232-238,1987. M.A.Tivey, Woods HoleOceanographic Institution, Woods Hole,MA alsothank3 anonymous reviewers fortheircomments whichhelped im- 02543-1542(email:[email protected]) prove themanuscript. M.A.Tiveywassupported byNSFgrant OCE- 9400623;H.P. Johnson wassupported by NSFgrantOCE-9405078.(ReceivedFebruary20, 1998;revisedJuly13, 1998; WHOI Contribution #9778. accepted August14, 1998.)
© Copyright 2025 Paperzz