JOURNALOF GEOPHYSICALRESEARCH, VOL. 87, NO. Bll, PAGES9389-9407, NOVEMBER 10, 1982

Structure and Origin of the J Anomaly Ridge, Western North

BRIAN E. TUCHOLKE

Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543

WILLIAM J. LUDWIG 1

Lamont-Doherty Geological Observatory of Columbia University, Palisades, New York 10964

The J Anomaly Ridge is a structural ridge or step in oceanic basementthat extends southwestfrom the easternend of the Grand Banks. It lies beneaththe J magneticanomaly at the youngend (M-4 to M- 0) of the M series magnetic anomalies. Its structural counterpart beneath the J anomaly in the eastern Atlantic is the -Tore Rise, but this feature has been overprinted by post-middle Cretaceous deformation and volcanism. In order to study the origin and evolution of the J Anomaly Ridge- Madeira-Tore Rise system, we obtained seismicrefraction and multichannel reflection profiles across the J Anomaly Ridge near 39øN latitude. The western ridge flank consistsof a seriesof crustal blocks downdropped along west-dipping normal faults, but the eastern slope to younger crust is gentle and relatively unfaulted. The western flank also is subparallelto seafloorisochrons, becoming younger to the south. Anomalously smooth basement caps the ridge crest, and it locally exhibits internal, eastward-dippingreflectors similar in configurationto those within subaeriallyemplaced flows on Iceland. When isostaticallycorrected for sedimentload, the northern part of the J Anomaly Ridge has basementdepths about 1400 m shallower than in our study area, and deep sea drilling has shown that the northern ridge was subaeriallyexposed during the middle Cretaceous.We suggestthat most of the system originated under subaerialconditions at the time of late-stagerifting between the adjacent Grand Banks and Iberia. The excessmagma required to form the ridge may have been vented from a mantle plume beneath the Grand Banks-Iberia rift zone and channelledsouthward beneath the rift axis of the abutting Mid-Atlantic Ridge. Resulting edifice-buildingvolcanism constructed the ridge system between anomaliesM-4 and M-0, moving southwardalong the ridge axis at about 50 mm/yr. About M- 0 time, when true drift began between Iberia and the Grand Banks, this southward venting rapidly declined. The resultswere rapid return of the spreadingaxis to normal elevations,division of the ridge systeminto the separateJ Anomaly Ridge and Madeira-Tore Rise, and unusuallyfast subsidenceof at least parts of these ridges to depths that presently are near normal. This proposed origin and evolutionary sequencefor the J Anomaly Ridge-Madeira-Tore Rise systemclosely matchesevents of uplift and unconformity development on the adjacent Grand Banks.

INTRODUCTION in crust dating from slightlyyounger than magneticanomaly Predrift reconstructions of the North Atlantic show the M-1 to slightly younger than M-0, but the amplitude of the southwestern edge of the Grand Banks as a sheared (trans- anomaly also can be accounted for by an increase in the form) margin that developed during the fragmentation of magnetic layer thickness [Sullivan and Keen, 1978]. In the Laurasia and the initial opening of the North Atlantic Ocean western North Atlantic the J anomaly can be traced north- [Pitman and Talwani, 1972; Le Pichon et al., 1977]. The eastward from near the eastern end of the New England SoutheastNewfoundland Ridge has been interpreted as the to the eastern edge of the Grand Banks. The trace of a fracture zone in marking the continu- anomalyhas intermediateamplitudes south of 38øNand high ation of this transform margin. The ridge extends about 900 amplitudesfrom there north to the SoutheastNewfoundland km southeastfrom the southern tip of the Grand Banks with Ridge [Rabinowitz et al., 1979]. Keen et al. [1977] also a subsidiary projection, the J Anomaly Ridge (or Spur suggestedthat the J anomaly continues northward acrossthe Ridge), extending to the southwest (Figures 1 and 2). SoutheastNewfoundland Ridge into the Newfoundland Ba- sin, with a 100-km offset; magnetic profiles in this area The J magnetic anomaly [Pitman and Talwani, 1972] is a suggestthat the amplitude of the anomaly decreasesnorth- linear zone of high-amplitude magnetic anomalies, up to about 1000 gammas, associated with crust formed between ward, especially north of the Newfoundland Seamounts. In magnetic anomalies M-0 and M-1 on both sides of the North the eastern Atlantic, the J anomaly follows the Madeira-Tore Atlantic Ocean (Figure 2) [Rabinowitz et al., 1979]. The Rise (MTR) [Luyendyk and Bunce, 1973; Rabinowitz et al., crust therefore is of Barremian-Aptian age (Early Creta- 1979],and it appearsto continuenorthward possibly as far as ceous, 108-113 Ma), according to the Larson and Hilde the western margin of Galicia Bank [Groupe Galice, 1979]. [1975] time scale used throughout this paper. Rabinowitz et These studies suggest that anomaly amplitudes decrease al. interpreted the J anomalies as isochrons formed at the north and south of the -Gibraltar Ridge in the same fashion and over about the same distances as in the western Mid-Atlantic Ridge axis. They modeled the anomalies by North Atlantic. assumingthat large increasesin magnetizationintensity exist A basement ridge or step generally is coincident with the J anomaly in the western North Atlantic, althoughit does not •Now at Gulf Oil Explorationand ProductionCo., Houston, Texas 77099. strictly parallel isochrons. The ridge is asymmetrical and is defined in seafloor morphology only north of 40øN, where Copyright 1982 by the American GeophysicalUnion. the name 'Spur Ridge' also has been applied [Gradstein et Paper number 2B1215. al., 1977]. To the south between 38ø and 40øN, the J 0148-0227/82/002B- 1215505.00 Anomaly Ridge (JAR) is buried by sedimentsof the Sohm

9389 9390 TUCHOLKEAND LUDWIG: STRUCTUREAND ORIGIN OF J ANOMALY RIDGE

Fig. 1. Locationmap of theJ AnomalyRidge and Madeira-Tore Rise in theNorth Atlantic.

AbyssalPlain, and the ridgecrest commonly exhibits a relief alone are not sufficientto accountfor the amplitudeof smoothsurface (Figures 2 and 3). This part of the ridge the J anomalyif a magneticlayer of constantthickness is slopesgently eastward but sharplywestward in a seriesof assumed [Rabinowitz et al., 1979]. 0.5- to 1.5-kmwest-facing scarps. The changesin basement A crustalridge also occurs beneath the Madeira-ToreRise

65ø 60ø 55ø 50ø 45ø t i i i !•,•"-,L•i I I I f' I '/' I I I I (,.,,• it,• I I I I( i I I/ I"----•

,. , , ,- .•..--:.... ?

ß...:...•.....,•.,, .• -•'•-•/J :i;'' .... ' 45 ø

• c- •.'- • •. /"

., • SON,• ' . - 40 ø

i-'"..•o' %,, 0 I ß;; ß 33 :•,

65 ø 60 ø 55 ø 50 ø 45 ø Fig. 2. Locationmap of Conrad21-11 MCS tracks (heavy lines). Bathymetry inmeters. Black dots are locations of magneticanomalies, asidentified by Rabinowitz etal. [1979].J anomalynorth of SoutheastNewfoundland Ridge from Keenet al. [1977].Inset shows three profiles across the high-amplitudeJ magnetic anomaly, located by dottedlines acrossJ AnomalyRidge. Detail of surveywork (boxed area) is shownin Figure4. Starlocates DSDP site 384. TUCHOLKE AND LUDWIG: STRUCTURE AND ORIGIN OF J ANOMALY RIDGE 9391

M2 WEST • EAST6

_ [-•_•••_, ß •• • -• ...... •• / I•/ ' v•"ø' /

• • •; ...... •' • v2701

I 1

z M4 M2 MO o o ._ :.:,-:;:.•:;,/ ' '- "L'::'-- -%½ • / " Vl713

5 •--• • '• , V2301 • M4I M2I ' MOI _

o U -• ' v2302 • ß M4 M2 : MO

. 58-w 55- 50-

•0'L ,• • : •• • • • V2301

' 5 -

EAST.]4

Fig. 3. Tracingsof single-channelseismic profiles (1-9) acrossJ Anomaly Ridge, alignedalong ridge crest (dotted line), and with M series magnetic anomalies located (modified from Tucholke and Vogt [1979]). Inset shows profile locationson map of depth to basement(in kilometers). Bottom profile showscorrelative basementridge in the eastern North Atlantic; profile is 200 km long, crossingMadeira-Tore Rise 70 km north of Madeira. 9392 TUCHOLKEAND LUDWIG: STRUCTUREAND ORIGIN OF J ANOMALY RIDGE

56 ø 5 5 ø 54 ø 53 ø 52 ø 40 ø

5280

39 ø

- ISOBATH ß•, AIRGUN-sONOBUOY STATION 0 EXPLOSIVES-SONOBUOYSTATION:•i'iI:•...... 5560--

LINE

•SCS LINE

ALPINE MO

38 ø Fig. 4. Locationmap of Conrad21-11 track andMCS, OBH, andsonobuoy measurements over the J AnomalyRidge. Bathymetry in corrected meters. at the position of the conjugateJ anomaly in the eastern netic anomaliesas part of the MesozoicM series[Rabino- North Atlantic (Figure 1). However, the structureof at least witz et al., 1978, 1979] and the occurrenceof the conjugate parts of the Madeira-ToreRise has been modifiedby late Madeira-Tore Rise in the eastern North Atlantic seem to Cretaceous-Tertiarycompressional deformation and Neo- demonstrateconclusively that the J AnomalyRidge is devel- genevolcanism [e.g., Laughton and Whitrnarsh,1974; Wat- opedin oceaniccrust. Vogt andEinwich [ 1979]also assumed kins, 1973]. Although no detailed study and mapping of an oceaniccomposition, and they suggestedthat the magne- seismic reflection data has been made for this region, our tizationchanges and (or) thickenedmagnetic layer associat- examinationof a variety of seismicprofiles suggeststhat at ed with the J anomalyoriginated in responseto the transition leastparts of the Madeira-ToreRise are a mirrorimage of the from a long period of relatively slow seafloorspreading (M J Anomaly Ridge in terms of general configurationand seriesanomalies) to a period of faster spreading(Cretaceous presenceof smoothbasement (Figure 3, bottom). magnetic quiet zone). In 1975, the northern J Anomaly Ridge was drilled on The oceanic basement (seismic layer 2) that forms the J magneticanomaly M-2 in an attemptto resolvethe originof AnomalyRidge and that liesbeneath the SohmAbyssal Plain the ridgeand the J magneticanomaly (DSDP site 384, Figure farther west seldom is well resolved in single-channelseis- 2). Basaltic ocean crust was penetratedat 4234 m below sea mic reflectionrecords made with small-volume(20-60 in3) level, and upper Barremian or lower Aptian shallow-water airguns. This is due to a thick cover (0.7-3.5 km) of carbonates were cored immediately above basement [Tu- interbeddedhemipelagic and coarse turbiditic sediments that cholke et al., 1979]. The basalt was not anomalous in either greatlyattenuate seismic energy. For this reason,it was not magnetizationor composition.However, the later magnetic certainthat the anomalouslysmooth 'acoustic basement' of modeling by Rabinowitz et al. [1979] showed that the J the J Anomaly Ridge observedin single-channelrecords anomaly is best explainedby anomalousmagnetization (or actuallywas basalticcrust. In orderto betterunderstand the anomalous magnetic-layer thickness) in crust between structureand origin of the J Anomaly Ridge, we obtaineda anomalies M-0 and M-l, rather than at M-2. series of refraction and multichannel seismic reflection mea- Gradstein et al. [1977] examinedgeological and geophysi- surementsover the ridge near 39øN(Figure 4). This area was cal data alongthe southeasternedge of the Grand Banksand selectedbecause previous magnetic and single-channelseis- concluded that subsidenceof the northern J Anomaly Ridge mic profilessuggested strong development there of boththe was very similarto that of the adjacentGrand Banks. Thus smooth acoustic basement and the M series anomalies. they inferred that the cored were flows covering METHODS continentalcrust. Suchan explanationseems unlikely in that this part of the ridge extendswell to the southwestof the Approximately3300 km of 24-channelseismic reflection Grand Banks transform margin and has a trend parallel to data, 35 sonobuoyreflection/refraction profiles, and three seafloor isochrons. Furthermore, interpretation of the mag- ocean bottom hydrophone(OBH) refractionprofiles were TUCHOLKEAND LUDWIG: STRUCTUREAND ORIGIN OF J ANOMALYRIDGE 9393

obtained during a 32-day cruise of the R/V Conrad in July graphsand interpretedby the conventionalslope-intercept 1978 (Figures 2 and 4). The objectives of the cruise were methodfor reversedprofiles [e.g., Ludwig et al., 1968]. three-fold. First, we traced key seismicreflecting horizons from south of the New England Seamountsinto the Sohm BASEMENT MORPHOLOGY AND STRUCTURE Abyssal Plain and continental rise off Nova Scotia where Multichannel seismic reflection profiles were obtained reflector identification was poorly controlled. Second, we both perpendicularand parallel to the trend of the J Anomaly investigatedthe structure and seismicvelocity distribution Ridge (Figures 5 and 6). Mapped together with previously of the buried J Anomaly Ridge. Finally, we obtained new obtainedsingle-channel profiles, they show clearly that seismic data bearing on the structure and nature of the regional structurestrend northeast nearly parallel to seafloor 'sedimentary(salt) ridge province' of the central and upper isochrons (Figure 7): Aside from a few small, apparently continental rise off Nova Scotia [e.g., Jansa and Wade, isolated peaks, minimum depths along the ridge crest in the 1975; Uchupi and Austin, 1979]. This report describesthe survey area are slightly less than 6.0 km below sea level. results of our seismic work over the J Anomaly Ridge Acoustic basement depths increase northwestward to about (Figure 4). 7.5 km at a position 80 km from the ridge crest. Toward the southeast, basement dips very gently, reaching depths of Multichannel Seismic Reflection Measurements only 6.2-6.5 km before rising into the flank of the Mid- The multichannel seismic (MCS) reflection equipment Atlantic Ridge [Tucholkeet al., 1982]. Acoustic basement aboard R/V Conrad consistedof a 2400-m-long,48-group observedin MCS profiles is unusually smoothover the crest streamer,four 466 in 3 airguns,and a DFS IV digitalrecord- of the ridge, and along line 151 parallel to the ridge crest it is ing system.The airgunswere fired at 127kg/cm 2 (1800psi) almost perfectly flat for at least 77 km (Figure 6). The every 18-20 s at a ship's speed of 9 km/hr. Shot points were smooth basement is broken by a series of predominantly located by satellite navigation and dead reckoning. The data northwestdipping normal faults that strike northeastparallel were recorded digitally on magnetictape at a 4-ms sampling to the ridge crest. Many of the fault blocks appear to have rate and over a recordingwindow of 8-12 s; the data window rotated, and several faults continue into the overlying sedi- startedjust before the seafloorarrival. Common depth point ment column as growth faults (e.g., line 152, Figure 5). The (CDP) processing of the data tapes was done at Lamont- corners of several fault blocks also appear to be truncated to Doherty Geological Observatory. Standard processing in- form a nearly horizontal surface (see arrows, Figure 5). The faulted smooth basement extends from crust of about anom- cluded demultiplexing and CDP gather, semblancevelocity analysis, normal moveout correction, stack, and deconvolu- aly M-2 age (113 Ma) at least to anomaly M-0 (108 Ma). West tion as described by Talwani etal. [1977]. of anomaly M-2 the basementbecomes an irregular, diffract- ing horizon that is more typical of oceanic crust, althougha Airgun-Sonobuoy Measurements few areas of smooth basement persist (lines 148 and 150, Figure 5). A narrow ridge of irregular crust interrupts the Over the 5000-m-deep Sohm Abyssal Plain, the 2.4-kin smooth basement and bounds a distinct basement step about hydrophone array is too short to resolve sediment interval halfway between anomalies M-2 and M-0 (Figures 5 and 7). velocities accurately, and SSQ-41A sonobuoyswere used to The smooth acoustic basement between the irregular provide velocity/depth information [Houtz et al., 1968; Le basement ridge and anomaly M-0 is distinctly different from Pichon et al., 1968]. These analog recordingswere obtained smooth basement farther west because it locally exhibits at two frequency bands on a shipboard drum recorder. In internal reflectors that dip and diverge toward the east addition, refracted waves generated by explosive charges (Figure 5). Intrabasement reflectors elsewhere often appear were recorded as oscillogramsat several sonobuoylocations to be peg-leg multiples generated by a double travel path and plotted on time-distance diagrams; these data sets were within the sediment column, but the dipping reflectors obtained as a by-product of shooting to ocean bottom cannot be explained in this manner. West of the irregular hydrophones. basementridge the smooth basementalso appearsto contain internal reflectors, but it is difficult to distinguish primary Ocean Bottom Hydrophone Measurements events from multiples because the structure has no signifi- Ocean bottom hydrophone (OBH) measurements were cant dip. Although the basement surface there is much made by using the Lamont-Doherty prototype instrument smoother than normal, it nevertheless has some small-scale [Carmichael et al., 1977]. The OBH is a lightweight, inter- roughness(indicated by hyperbolic diffractions) that may be nally recording, pop-up system that separatesfrom a combi- partially maskedby thin, high-velocitysedimentary beds or nation ballast-weight/power-supplyby a time release mecha- basalt flows. nism. Anchored about 2 m above the seafloor, it allows The overall morphologic development of the J Anomaly increased resolution of crustal velocity structure and re- Ridge system is not strictly parallel to seafloorisochrons. In duced ambiguity in travel time data compared to drifting Figure 8 we plot basement depth, isostatically corrected for sonobuoys. For our project, the OBH was modified to loading by overlying sediments, along several seafloor iso- permit 12 hours of continuousrecording and was fitted with chronsfrom approximately magneticanomaly M-1 l/M-12 to a magnetostrictive pinger for location and navigation pur- anomaly M-0. In crust formed through anomaly M-4 time poses [e.g., Francis et al., 1975]. (117 Ma), there is little indication of ridge development; in The reduction of seismic refraction data recorded at the fact, crustal depths are up to a half kilometer deeper than seaflooris describedby Ewingand Ewing [ 1961],Davis et al. predicted from the North Atlantic empirical age-depth [1976], and Purdy and Detrick [1978], among other. We curve. At anomaly M-2 (113 Ma) the northern part of the recorded refraction profiles in opposite directions at each ridge becomeswell developedadjacent to the Grand Banks; OBH location, using explosives as the sound source. After DSDP site 384 recovered upper Barremian or lower Aptian appropriate filtering, the data were plotted as time-distance shallow-water carbonatesoverlying basementon this part of 9394 TUCHOLKE AND LUDWIG: STRUCTURE AND ORIGIN OF J ANOMALY RIDGE

:..6"Fxt49•-.I x L.INE' 150 x i 48

.9

x i49 x i5o LINE 152 x l,5"! . $E

.

Fig. 5a. MCS stackedrecord sections for lines148, 150, and 152 across J AnomalyRidge. Bandpass filter 10-70 Hz; time-varyinggain and Burg deconvolutionapplied. Scale bars = 20 km. Locationsin Figure4. the ridge. The magnitudeof the positive topographicanoma- Ridge near anomaly M-0. Recognizingthat both sonobuoy ly decreasesalong younger crustal isochrons,but there is a and OBH refraction profiles are localized measurements clear indication that the topographic anomaly migrated which determine average velocities over the line sampled, southwardalong the spreadingaxis at a rate of about 50 mm/ there is good agreement between the reversed and unre- yr (dotted line, Figure 8). By anomaly M-0 time, the crust is versed measurements(Table 1). The airgun-sonobuoytech- within about 200 m of normal elevations, and the zone of nique has greater resolution because of the much higher highestamplitudes of the J magneticanomaly [Rabinowitz et repetition rate of the airgun comparedwith explosives.In al., 1979] is restricted to crust north of this morphologic the reversed refraction lines shot with explosives to the trend (Figure 8, stippled area). OBH (Figures9 and 10) and to sonobuoys,the shot spacing was not close enoughto detect arrivalsfrom acoustic SEISMIC-VELOCITY STRUCTURE basement; hence the velocities were assumed from nearby The velocities and layer thicknesses determined from airgun-sonobuoymeasurements. With one exception(sono- unreversed, airgun-sonobuoyprofiles are based on the as- buoy 221), both smooth and rough acoustic basement (as sumptionof nondippinglayers; thus it is useful to compare observed in the MCS records) have refraction velocities of them with results from reversed sonobuoyand OBH profiles 4.5 km/s or greater. We therefore interpret acousticbase- recorded with explosives to determine true velocity in the ment as basalticocean crust, possiblyinterbedded with high- presence of dip. Such comparison is possible along the velocity elastic sedimentsin the smooth-basementareas. anomalyM-4 isochronand alongthe crest of the J Anomaly The various crustal layers determined by sonobuoy solu- TUCHOLKEAND LUDWIG:STRUCTURE AND ORIGINOF J ANOMALYRIDGE 9395

Fig. 5b. Interpretation of record sectionsin Figure 5a. tions have a fairly wide range of velocities. This is because but because of the small statistical sample it is not known some sonobuoysolutions do not resolve all layers but give that this difference is significant. More importantly, the velocities combining several layers, and becausethe veloci- velocities within the upper 2.5 km of older crust (>M-16) are ties are not corrected for layer dip. For these reasons uniformly 0.4-1.1 km/s higher than at M-0. This differenceis individualsolutions may be misleadingin studyinggeologic exemplified in the shallowest crust where measured veloci- structure. In Figure 11 we have partially circumvented this ties for the older basementrange from 4.95 to 5.75 km/s, but problem by averaging sonobuoy solutionsfor three crustal for M-0 basementall valuesbut one are in the range 4.5-4.7 age groups: M-0, M-4, and a group extending from M-16 (135 km/s (Table 1). Ma) westward into the Jurassic quiet zone (--•155 Ma). In The differencesin crustal velocity structure are illustrated comparingthe averages in this plot, increasesin basement in a cross section extending from the J Anomaly Ridge velocity due to increasesin sedimentary overburden and northwest toward the Nova ScotJanmargin (Figure 12). As hydrostatic pressure are ignored; the changes are small shown by the isovelocity contours, a thick, lower velocity (--•0.2 km/s) and most likely are compensatedby equally crustal section is associated with the J Anomaly Ridge small velocity decreasesrelated to increasingtemperature beginningapproximately at the time of anomaly M-4. The [e.g., Chrostonet al., 1979] with increasingdepth of burial. sametrend exists if the previously developed 'velocity layer' The basementvelocity structureat anomaliesM-0 and M- conventionsof Houtz and Ewing [ 1976] and Houtz [ 1980] are 4 is very nearly the same (Figure 11). There is a suggestion used. These authors designated crustal velocities with extre- that somewhat lower velocities occur 0-2 km subbasement mal rangesof about 4.8-5.6 km/s as 'layer 2B' and velocities beneaththe crest of the J Anomaly Ridge at anomalyM-0, less than 4.1 km/s as 'layer 2A.' With the exception of 9396 TUCHOLKE AND LUDWIG: STRUCTURE AND ORIGIN OF J ANOMALY RIDGE

x x x !50 •5.2 .!48 G

.•':!:1 SW...... I...... :...... ":'"':L/AlEii ii i ,,,,,,, "'"'":',...... :"•'...... "•'::" :[ .....:"1 '•t ..•-."%:::•.• •.•-:'x•-?'• '••>?'-•c--•. ?:-•::?';- •'•:•':',' ,:"•:?::?• ;•::•::. •:g•;•.'..-:.:.-..•>.:'•;.•:•.•-•?.•-'• '::.'•"•...... '-' ':• '• ":'C: ..:...• -:-.'.:: % ':: :?•fi.:. ;.• ": "Z:?7:-::"'% 8 :•.•.-•.•....•...... • '*•...... :•':'• ;,•--:.:'•'::.:: :•:'¾ '•'::••?•.••"•.-:-•-7: -•e•':'•'"'" '• :g•" •'/&•';•-•"•••••• • •-'•- ' .--e•-•:'•:•:'-'•-:z•":•z••' ...... ••••••••••••••••.••••;• -...... '...... •:•;•-..---' .••..•..•.•,• • •-e:::'•-v..- :•-'.-' :•;• ...... :.. 9

10

f! ... X XX 152 1,½8 I50 "t •:•__.•.___sw ...... _...... :...... :.•::,,•LINE ...... 151.:, ...... •...... _• ...... •...... • ...... N•..!

•.•,-'m.-.•- ß•' •-•.] : -'-.-..'- •'t.:.' •'•. '•. •:•¾...•.• "•'.•, ,• --.

Fig. 6. MCS stacked record sectionsfor line 149 along anomaly M-4 isochron(top) and for line 151 near M-0 isochronat crest of J Anomaly Ridge (bottom). Bandpassfilter 10-70 Hz; time-varyinggain and Burg deconvolution applied. Scale bars = 20 km. Locations in Figure 4.

56øW 55 ø 54 ø 53 ø 52 ø 40øN ß /// / /v / // /

/ - / '•••x/•1'.6 ß / '1. •' // // / / • .• •_•, /i/ / fi: / / ß / / / / / / /

(KNOWN, ; NORMALINFERRED)FAULTS BASEMENTSMOOTH 39 ø MAGNETICANOMALIES ß 0 A M-4 M-2 M-O // DEPTHSTOBASEMENT INKM BEL• SEALEVEL ///

--.'-__ // • / //

38 ø Fig. 7. Map of depth to basement and structure of J Anomaly Ridge. Track control shown by light solid lines. TUCHOLKE AND LUDWIG: STRUCTURE AND ORIGIN OF J ANOMALY RIDGE 9397

,j

i

Fig. 8. Perspectivediagram of basementdepths (corrected for isostaticloading by sediments)along isochrons in area of J AnomalyRidge southwestof DSDP site 384. Vertical rulinghighlights depths below empiricalage-depth curve of Tucholke and Vogt [1979]; depths above curve are unpatterned. Stippled area shows zone of anomalous magnetizationbetween anomaliesM-1 and M-0 proposedby Rabinowitz et al. [1979]to accountfor J anomaly.Dotted line showsapproximate trend of southwardridge migrationbetween anomalies M-4 to near M-0; it intersectsM-0 near 38øN, where J anomaly amplitudesdecrease markedly to the south. Axial profile of Iceland and Reykjanes Ridge at upper left [from Vogt, 1974]is shownfor comparisonof vertical and horizontal scales. sonobuoy221, there are no velocities in the layer 2A range in Anomaly Rdge (line 152, Figure 5) are very similar to those our study area. Layer 2B is less than a kilometer thick in observed beneath Iceland [Zverev et al., 1980a, b] in that older crust beneath the Sohm Abyssal Plain, but it is 1.5-2 they dip and diverge toward the accretion axis, and they tend km thick beneath the flank and crest of the J Anomaly Ridge to be rather short (--•10 km) reflective segmentsrather than (Figure 11). Farther east beneath 45-100 Ma old seafloor, the laterally extensive horizons. Geologic and kinematic models shallowest crustal layer also appears to have velocities proposed by Bodvarsson and Walker [1964] and Pallmason characteristic of layer 2B, but the layer thickness there is [1973, 1980], respectively, explain the Icelandic units as only about 1 km [Houtz and Ewing, 1976; Houtz, 1980]. subaerial flows that thicken toward the spreading center; Hence the J Anomaly Ridge appearsto have an anomalously subsidence near the spreading center is due to loading by thick layer 2B. This structure is not reflected in the gravity subsequentflow units. Assuming a relatively fixed distance field, which shows a nearly flat free-air anomaly that indi- over which flow units normally spread, the units ultimately cates isostatic compensationat depth. offlap in the direction of the continually widening rift zone. Similar dipping intrabasement reflectors have been recog- ORIGIN OF TIlE J ANOMALY RIDGE nized at the edge of several passive continental margins, and these also are thought to be characteristic of subaerial To determine the origin of the J Anomaly Ridge (and its seafloor spreading associatedwith initial accretion of ocean eastern Atlantic counterpart, the Madeira-Tore Rise), it is crust [e.g., Mutter et al., 1982]. useful to consider several geologicanalogs in Iceland and the On the J Anomaly Ridge the normal faults that dip away Reykjanes Ridge. The parallels in structure and morphology from the spreading axis also are like those observed in are quite striking, and they are useful in interpreting both the Iceland [Zverev et al., 1980a] and in the Icelandic insular paleodepth and the mechanismof formation of the J Anoma- slope adjacent to the Reykjanes Ridge [Egloff and Johnson, ly Ridge/Madeira-Tore Rise System. 1979]. We presume that they result principally from in- A variety of observations indicates that much of the J creased thermal contraction with distance from the spread- Anomaly Ridge was formed at or even above sea level. The ing axis. Saemundsson[1967] observed that some Icelandic dipping reflectors beneath the smooth basement of the J normal faults have greater displacement in older strata than 9398 TUCHOLKEAND LUDWIGß STRUCTUREAND ORIGIN OF J ANOMALY RIDGE

o

o o••o o o • o••o ø

'D

0 c• 0

0

c• 0 c• ß c• o

o c• o

o

0 0 TUCHOLKE AND LUDWIG: STRUCTURE AND ORIGIN OF J ANOMALY RIDGE 9399

(spuooes) 3lflI.L 93A•'W.L 9400 TUCHOLKE AND LUDWIG: STRUCTURE AND ORIGIN OF J ANOMALY RIDGE

TRAVEL TIME (Seconds)

O

i I

i i TUCHOLKEAND LUDWIG: STRUCTUREAND ORIGIN OF J ANOMALY RIDGE 9401

VELOCITY (l•m/$) level. Crustal accretion and erosion at zero depth seem to 8 9 provide the best explanation of the available geologicaland geophysicalobservations. The analogousdevelopment of the J Anomaly Ridge and the 'Iceland phenomenon' extends beyond basement struc- (6) :-"•'] ] WESTERN ture. The vertical and horizontal scales of the two systems "! 1 • SOHMA.R are very much the same (Figure 8), and there is obvious similarity in the southwarddeepening axes of the J Anomaly M-4 .... and Reykjanes Ridge systems, with resultant southward pointing, V-shaped bathymetric patterns. These larger scale

I trends suggestthat the J Anomaly Ridge system was formed I by a mechanismlike that responsiblefor the Iceland system. Iceland is thought to result from effusionof excessmag- mas above a mantle plume or hot spot [e.g., Morgan, 1981]. In addition, increasingaxial depths and systematicchanges in chemistry of Reykjanes Ridge rift-axis basalts southwest Fig. 11. Plot of velocity versusdepth below basementaveraged from Iceland both have been attributedto the presenceof the for each of three crustal age groups. Number of individual determi- sub-Icelandic mantle plume [e.g., Schilling, 1973; Vogt, nationsused for each averageis given in parentheses(see Table 1). The low average velocity for 0-0.5 km subbasementat M-0 is due 1974, 1976]. These authors suggestthat the ascendingman- largely to inclusion of an anomalous 3.1 km/s refractor observed in tle-plume melts,which normally would migrate radially from sonobuoy221. The same relative pattern of velocity distribution the hot spot, are preferentially channeled into 'pipe flows' versus age group is observedif the total velocity range of each age along and beneath the rift axis of the mid-ocean ridge. Head group is plotted. loss and progressive mixing of primary and mantle-plume melts away from the plume account for the observed topo- in younger strata; thus these faults appear to be comparable graphicand geochemicaleffects. Vogt [1974] calculatedthat to the growth faults observed on the J Anomaly Ridge. the migration rate of the plume melts is of the order of 50- Th e smoothbasement widely developedalong the crestof 200 mm/yr, very similarto the 50 mm/yr southwardmigra- the J Anomaly Ridge also suggestsa subaerialorigin, proba- tion of the topographicexpression of the J Anomaly Ridge. bly as a result of laterally extensive basalt flows coupled All these considerationssuggest that the J Anomaly Ridge with subaerial erosion. In additon, subaerial erosion at sea system originated by a hot spot mechanism or at least with level best explains the apparently truncated tops of tilted the presenceof a hot spot-like magmasource. Discussion of fault blocks near the ridge crest (see arrows, Figure 5). this concept and evolution of the ridge system are outlined The only direct samplingof the J Anomaly Ridge, accom- below. plished on the shallower northern part of the ridge at DSDP EVOLUTION OF J ANOMALY RIDGE site 384 (Figure 2), provides direct evidence of subaerial exposure. The crust there was formed at magnetic anomaly In Early Cretaceous time prior to formation of the J M-2 (-• 113 Ma) and consistsof normal tholeiitic basalts only Anomaly Ridge, the spreadingaxis of the Mid-Atlantic Ridge slightly enriched in large ion lithophile (LIL) elementsand of intersected the southwestern margin of the Grand Banks undistinguishedmagnetization [Tucholke et al., 1979]. Three nearly at right angles. The spreading axis did not extend lava flows were recovered, the lowest of which contains north into Laurasia but was offset along a major transform titanomagnetitesaltered by high-temperaturedeuteric oxida- fault into the Tethys [e.g., Dewey et al., 1973].The adjacent tion that is typical of subaerialextrusion. The upper flows continental crust of Iberia (Iber) and the Grand Banks (GB) are altered by low-temperature, probably submarine, oxida- had experienceda probably Triassic-Liassicphase of rifting tion [Petersen et al., 1979]. Initial subaerial extrusion and/or followed by relative inactivity during the Middle Jurassic later exposure also are indicated by high vesicularity of the [Sibuet and Ryan, 1979]. During the Late Jurassicand Early basalts and by a basaltic 'soil' horizon presumably formed Cretaceous, GB-Iber rifting again commenced, and the first by intense subaerial weathering. true drift and production of oceanic crust between these land The basalt could have been above sea level for as much as massesare inferred to have occurred at the J anomaly about several million years, after which upper Barremian or lower the end of Aptian time (--105 Ma; Sibuet and Ryan [1979] Aptian shallow-water carbonatesaccreted on the ridge. The and Sullivan [1982]). carbonateswere deposited until late Aptian or early Albian We propose that the J Anomaly Ridge and Madeira-Tore time (---105-108 Ma). Petrographicand oxygen isotopic data Rise originated in concert with, and possiblyin responseto, indicate that the carbonates themselves were then subaerial- the late-stage rifting and initial drift between Iberia and the ly exposed [Rothe, 1979; Rothe and Hoers, 1979] before the Grand Banks. Specifically, drawing on the Iceland analog, ridgerapidly subsidedto its presentdepth. we suggestthat partial melts from deep sources within the Given this geologichistory, it is reasonablethat the ridge rift zone were funneled southwardalong subcrustalconduits crest farther south, in our survey area, also formed near sea and enteredthe spreadingaxis of the abuttingMid-Atlantic level. Basement depth at site 384, isostaticallycorrected for Ridge. Extrusion and intrusion of these melts at the spread- sedimentloading, is 4100 m. This is only 1400 m shallower ing axis caused the Mid-Atlantic Ridge crest to accrete to than the comparably corrected basement depth of 5500 m elevations at and above sea level. Southward migration of over the southern part of the ridge. Thus while the M-2 age this 'melt front' (and correspondingtopographic anomaly) crust at site 384 was Subsidingtoward sea level, the M-1 to with time allowed the J Anomaly Ridge to be constructed M-0 age crust farther south probably was accreting at sea acrossseafloor isochronsuntil about 108 Ma (anomaly M-0, 9402 TUCHOLKEAND LUDWIG: STRUCTUREAND ORIGIN OF J ANOMALYRIDGE

MA•'NœTICTOTALIIYTœNSITANOMALY ¾/ /"\ ',.,\ Iø M4 / M•0•,_ 400

-B00

I I -12oo 7' I I 4- SEISMIC STRUCTURE

5 I SOHM ABYSSAL PLA IAI ,I-AAIOMAL Y• RIDGE ]-- 1.60 1._701,78"1.84 j 6 m 1.99 2.35,..,, /-X,l• 6

,

?

e I 8

9 I --9

---IO

--II

12 I

--13

141 8.5 --14

0 50 I00 Km I I I SE _ 15 Fig. 12. Velocity-structuresection and magnetic anomaly profile across the J AnomalyRidge. The configurationof the basementlayer is drawn from MCS line 152(see Figure 5), and isovelocitycontours are basedon averagesin Figure 11' mantle determination at left is from sonobuoy 235.

Figure 8). If we extrapolate the rate of southwardmigration melt to a mantle plume until the hot spot trace is better of the ridge (50 mm/yr) back to the continental margin, the constrained. melt front probably was beneath the edge of the Grand A second possible source for the magma is a large zone of Banks at about 120 Ma. partial melting beneath the GB-Iber rift system. The east- Two possibleorigins of the magma source seemattractive. ward relative movement of the Mid-Atlantic Ridge crest with One is the passageof the Grand Banks over a rising mantle respect to the GB-Iber rift zone may have brought the plume or hot spot. Recent work by Morgan [1981, manu- spreadingaxis to a location where it 'tapped' this melt zone. script in preparation, 1982] suggeststhat the Labrador shelf However, it is not clear that this kind of magma source passedover the Canaries hot spot and that the Grand Banks would be volumetrically important in the absenceof a mantle moved over the Madeira hot spot about 120-110 Ma. The plume. Grand Banks and adjacent rift zone then remained over the Presumingthat the kinematic evolution, if not the dynamic Madeira hot spot continuously up to about 105 Ma. A mechanism of ridge formation proposed above, is valid, we problem with this model is that Morgan's relative positionof can outline the followinggeologic history for the J Anomaly the Madeira hot spot at 120 Ma dependson inception of drift Ridgesystem as it relatedto the adjacentGB-Iber rift zone. between the Grand Banks and Iberia about 80 Ma, whereas evidence presented below and by $ibuet and Ryan [1979] Late-Rift Phase (120-113 Ma) indicatesthat drift began at the end of the Aptian (--105-108 The partial melt migrating southward from the GB-Iber rift Ma). Although this does not invalidate the possiblerole of a zone (Figure 13) probably reached the Newfoundland Frac- hot spot, we cannot clearly attribute the origin of the partial tureZone about 117 Ma (anomalyM-4} andth e latitu•leof TUCHOLKE AND LUDWIG; STRUCTURE AND ORIGIN OF J ANOMALY RIDGE 9403

55øW 50 ø 45 ø 40 ø •ALICIA FLEMISH JEANNE CA• D 'ARC gASIN

WHALE ß

ß IgERIA 45øN ß gAS/N,

S'MI:

TAGU• gA•IN

GOI•lglNGœ 40 ø / IglDGœ

AI•EA

AFI•ICA _.,

.35 ø Fig.13. Reconstructionof the North Atlantic to anomalyM-0 time (108 Ma), based on North America-Africa- Iberia-Europepoles of rotation of K. Klitgordand H. Schouten(manuscript inpreparation, 1982). Contours west of spreading-ridgeaxis(heavy line) are present depths to basement inkilometers below sea level [Tucholke etal., 1982' Jansaand Wade,1975]; east of spreadingcenter, regional depth-to-basement contours are not available,and bathymetriccontours are used to illustrate the general shape of the Madeira-Tore Rise. The overlap (shaded) ofGalicia Bankand Flemish Cap is at the 5-kin ba•m•nt-d•pth contour. The African and Iberian plates are arbitrarily separated alongGorringe Ridge. North of Tore thetriangles locate the J anomalyon the Iberian plate [Groupe Galice, 1979],and the dots locate a possibleJanomaly over Flemish Cap Ridge on the North American plate [Sullivan, 1982]' to thesouth, the rotated Janomalies are coincident. Pico Fracture Zone and Newfoundland Fracture Zone are symbolized by syntheticflow lines (K. Klitgordand H. Schouten,manuscript in preparation, 1982)' these flow lines follow basementstructural trends quite closely. site384 about 114 Ma (anomalyM-3). This is presumedto be basaltflows locally formed the characteristiceastward dip- reflectedin the southward extensionof subaeriallyaccreting pingsubbasement reflectors observed in our multichannel crustalong the Mid-AtlanticRidge. We knowthat by 113Ma seismiclines (Figure5). Subaerialerosion peneplaned this (anomalyM-2), basaltswere beingextruded subaerially at crustto form the unusuallysmooth basement surface. Slight- site 384 on the Mid-Atlantic Ridge crest. Farther south, the ly oldercrust had partially cooled and subsided;this crust spreadingaxis probably was below sea level (Figures 3 and was block-faultedalong west dippingfault planesand some 5). No true oceaniccrust had yet formedin the GB-Iberriff of the blockswere rotated(Figure 5). Erosionproducts from zone. the adjacentsubaerial crust probably helped smooth and mask the basement surface of this older crust. At site 384, Rift-Drift Transition(112-108 Ma) the crust subsided below sea level, and reef carbonates Between anomalies M-1 and M-0 (upper Barremian to accumulated. middleAptian), the melt front and topographic expression of Drift Phase (--•108 Ma) theridge migrated south to a (present)latitude of about38øN (Figure8). The ridgeprobably accreted to heightsnear or We suggestthat the first emplacementof true oceanic slightlyabove sea level along its entire length, and subaerial crust between Iberia and the Grand Banks began at this time 9404 TUCHOLKE AND LUDWIG: STRUCTURE AND ORIGIN OF J ANOMALY RIDGE

(about anomaly M-0). The thermal anomaly and minor We correlate the 'short, sharpbreakup unconformity' with regional uplift that are postulatedto accompanythis conti- subaerialexposure of the reef at site 384 and with truncation nental breakup [Falvey, 1974] seem to explain best two of the corners of tilted fault blocks on the J Anomaly Ridge, unusual facets of the J Anomaly Ridge: (1) the temporary, and we place the time of initial drift between Iberia and the late Aptian or early Albian uplift and subaerialexposure of Grand Banks at anomaly M-0 (--•108 Ma, late Aptian). the reef carbonatesat site 384 and (2) the apparentuplift and Similar manifestationsof brief subaerial exposure probably subaerial truncation of corners of tilted fault blocks near the occur on other basement ridges and seamounts in this ridge crest (arrows, Figure 5). These minor, althoughimpor- region. In deeper basinsthe breakup unconformityprobably tant, features would correlate with Falvey's 'breakup uncon- is not developed, but some shallower basinal sedimentsmay formity.' Furthermore, the entire ridge system apparently have been briefly exposed. A possibleexample of the latter experiencedunusually rapid subsidenceshortly following occurs in the South Whale Basin (Amoco and Imperial anomalyM-0. We attribute this to rapid decay of the initial [1974], their Figure 9) where the Puffin B-90 well encoun- thermal regime that accompanied the breakup, and we tered nonmarine facies and an overlying unconformity in the suggestthat the mid-oceanic ridge axis rapidly evolved to Aptian-Turonian section. In most other areas the uplift near-normal depths greater than 2000 m. either causedreexcavation of the earlier synrift unconformi- ty or maintained the subaerial exposure and excavation DISCUSSION inherited from the previous uplift. Regional subsidence The suggested timing of the above events correlates shortly following the breakup event allowed Upper Creta- closely both with Falvey's [1974] model of continental ceous marine facies to become widely distributed across the breakup and with observationsof regionalgeology. Falvey Grand Banks [Gradstein et al., 1975]. predictedtwo phasesof uplift duringthe rifting history of a Our placement of the ocean-continentboundary between passivecontinental margin. The first could occurwell before Iberia and the Grand Banks at anomaly M-0 (J anomaly) has (>50 Ma) the initiation of true drift and consistsof regional several interesting consequences. First is the implication uplift over the rift zone in responseto a thermal anomaly that there is deeply subsided continental crust (8+ km) impinging on the base of the lithosphere. The second is beneath the Newfoundland, Tagus, and Iberia basins(Figure predicted to be a 'short, sharp' period of uplift associated 13). The depth of these basins perhapsis not too surprising with convective heat transport at the time of breakup. A because the Jeanne D'Arc Basin beneath the Grand Banks period of relative subsidenceis thought to occur between has more than 12 km of Mesozoic/Cenozoic fill (Figure 13), thesephases, possibly as a result of thermal metamorphism but the size of the basins is significantly larger and their [Falvey, 1974]. sedimentaryfill is much smaller (--•5 km). These basinsmay The early-rift uplift correlates with widespread develop- contain continental crust that was extensively thinned (and ment of an erosional unconformity over the Grand Banks, intruded) during the rifting phase. Margin basins of similar Orphan Knoll, Galicia Bank, and in the Lusitanian and size and depth occur farther north in the Flemish Pass and Aquitanian basins of Europe [Amoco and Imperial, 1974; the Orphan Basin. Refraction work there by Keen and Gradstein et al., 1975; Jansa and Wade, 1975; Ruffman and Barrett [1981] suggests that extensively thinned (--•50%) van Hinte, 1973; Laughton et al., 1972; Sibuet and Ryan, continental crust underlies the basins and that total crustal 1979;Dardel and Rosset, 1971; Winnock, 1971]. The uncon- thicknessis 15-17 km, includingup to --•5km of Cretaceous/ formity is typified in Grand Banks wells, where the hiatal Tertiary sediment. gap usually spans the Late Jurassicthrough middle Creta- A secondconsequence of postulatingthat breakup did not ceous (Aptian). Thus, following initial uplift, subsidence precede anomaly M-0 is that Flemish Cap and Galicia Bank over most of the Grand Banks apparently was so slight that exhibit minimal overlap (Figure 13). The 70-km overlap no Lower Cretaceous marine sequencesaccumulated. We north of Tore Seamount in the M-0 reconstructionmight be have suggestedthat the adjacent oceanic crust of the J explained by subsequent extension (spreading?) east of Anomaly Ridge systemalso was subaeriallyexposed; hence Galicia Bank (extensional basins of Rdhault and Mauffret the Lower Cretaceous unconformity may be continuous [1979]) and/or possibly in the deep graben west of Flemish across both continental crust and the oceanic crust of the Cap [Haworth and Keen, 1979]. ridge. Finally, the westernmost Newfoundland Seamounts and There are two kinds of locations where this widespread Tore Seamount are formed very near the ocean/continent hiatal development may have been restricted to a shorter boundary but on presumed continental crust. This raises the time interval. One is in basins which initially were less interestingpossibility that the unusualring structureof Tore elevated or which subsided so rapidly that Lower Creta- Seamount actually is a collapsed caldera formed under ceous marine sedimentary sequences could accumulate. subaerialconditions. Laughton et al. [1975] consideredthis Examples occur along the southwestmargin of the Grand possibility but dismissed it because of the unusually large Banks in the South Whale Basin (Puffin B-90 well, Gradstein magma chamber that would be necessary. However, if the et al. [1975]; Amoco and Imperial [1974]) and over Galicia seamountformed near a newly developed spreadingcenter Bank [Sibuet and Ryan, 1979]. The Newfoundland, Iberia, subject to mantle-plume influence, the explanation may not and Tagus basins probably also accumulatedLower Creta- be so unrealistic as was once thought. ceous marine sequences,and they may have been so deep The positionof the ocean/continentboundary south of the that no hiatus developed (Figure 13). A second, very local- Grand Banks is equally interesting.The data and model that ized, restriction of unconformity developmentprobably oc- we have presentedfor the J Anomaly Ridge systemindicate curred at loci of subaerial volcanism. Volcaniclastics and that only oceanic crust occurs south of the Newfoundland lavas of mixed Late Jurassic-Aptian agescan be expected to Fracture Zone (Figure 13).We also suggestthat oceaniccrust interruptand overlie the synrift unconformitythroughout the extends northward to the Pico Fracture Zone near the edge region. of the Grand Banks. Seismic refraction profilesjust south- TUCHOLKE AND LUDWIG: STRUCTURE AND ORIGIN OF J ANOMALY RIDGE 9405

west of the Grand Banks appear to confirm the occurrenceof ling, 1973], radial flow from the plume probably would be oceanic crust seaward of the Pico Fracture Zone [Jackson et concentrated along rift-axis conduits toward the south (and al., 1974]. In contrast, Grant [1977, 1979] inferred from the toward the north unless attenuated in the continental rift observed distribution of the Lower Cretaceous unconformi- zone). Progressivemixing of anomalousmantle-plume melts ty in seismic reflection records that continental crust was with primary rift-axis magmas along the conduits could present in this region beneath the Southeast Newfoundland produce a trend of decreasingly magnetized crustal rocks Ridge. However, as noted earlier, the unconformitycould be with distance from the plume. Or emplacement of excess continuous over both continental and oceanic crust. Further- magma nearer the plume may simply have produced a more, if continental crust were present between these frac- thicker magnetic layer. The presenceof a mantle plume also ture zones off the Grand Banks, we would also expect could explain several other observations: continental crust to occur in conjugate areas of the eastern 1. The steplike decay in anomaly amplitude north of the Atlantic, for example within the stronglydeformed Gorringe Newfoundland Seamounts, south of 38øN, and again at the Ridge. Neither geophysical or dredge results indicate the New England Seamounts: these locations appear to be presence of continental crust there [e.g., Purdy, 1975]. If significant structural boundaries (fracture zones?) in the oceaniccrust in fact is presentimmediately southwest of the crust, where 'pipe flow' from a mantle plume would be Grand Banks, the two southwest trending basement highs attenuated or blocked [e.g., Vogt, 1974]. there require explanation(Figure 13). It is possiblethat they 2. The timing of actual occurrenceof breakup at anoma- are earlier formed versions of the J Anomaly Ridge, em- ly M-0: arrival of an especially weakened zone of rifted placed at the crest of the Mid-Atlantic Ridge as it migrated continental crust over a mantle plume could explain the eastward along the Grand Banks transform margin. initiation of true drift. While the structural origin of the J Anomaly Ridge/ 3. The anomalously rapid subsidenceof the J Anomaly Madeira-Tore Rise system can be correlated rather well with Ridge, SoutheastNewfoundland Ridge, and probably parts the history of rifting between Iberia and the Grand Banks, of the adjacent continental crust following anomaly M-0: the origin of the J magneticanomaly itself remainsuncertain. subsidenceat rates significantlyfaster than predicted by the The restriction of anomalous magnetization to the zone conventionalt•/2 relation might be explainedby a lossof between anomalies M-1 and M-0 [Rabinowitz et al., 1979] dynamic uplift if a mantle plume either became inactive or suggestsa direct relation to the final rifting and initial drift migrated away from the region following GB-Iber breakup. between Iberia and the Grand Banks. Any satisfactory CONCLUSIONS model, however, must also explain the apparentlydecaying amplitude of the J anomaly both northward and southward The J Anomaly Ridge and Madeira-Tore Rise are paired from the SoutheastNewfoundland Ridge [Keen et al., 1977; aseismic rises formed by excess volcanism at the crest of the Groupe Galice, 1979; Rabinowitz et al., 1979]. Mid-Atlantic Ridge between about anomaly M-4 and M-0. Before considering the mechanism by which the anoma- The edifice-building volcanism that formed these ridges lous magnetization was produced, we should discuss its migrated southward at about 50 mm/yr, constructingmuch source. The absenceof any significantgravity anomaly over of the ridge system to and above sea level. This construction the J Anomaly Ridge and the normal velocity structure and migration appear to have been triggered when the (excepting the thickened layer 2B) eliminate the possibility eastward moving Mid-Atlantic Ridge axis arrived at and that the anomaly is produced by intrusionsof dense materi- 'tapped' a large magma source (possibly a mantle plume) als such as intensely magnetized gabbros or chromium- beneath the GB-Iber rift zone. The mantle plume concept spinel-bearingultrabasic rocks [Tucholke and Vogt, 1979]. most readily explains both the large magma source neces- However, induced magnetizationcaused by the presenceof sary to form the ridge system and the eventual formation of a large body of serpentinizedperidotitc cannot be discount- the J magnetic anomaly. ed. Similarly, emplacementof unusually magnetizedrocks The evolution of the J Anomaly Ridge/Madeira-Tore Rise becauseof late-stage magma differentiation, thermal bound- system is intimately related to the GB-Iber rift-drift se- ary effects, or passage over a mantle plume are viable quence. The widespread 'synrift' unconformity over the alternatives. Finally, an unusually thick magnetic layer may adjacentcontinental margins appears to be laterally continu- be present. This possibility may be supportedby the pres- ous with a subaerially eroded unconformity developed on ence of a layer 2B that is up to 2 km thick beneath the J oceanic crust of the J Anomaly Ridge and probably the Anomaly Ridge, compared with about 1-km thicknessesin SoutheastNewfoundland Ridge. The short, sharp 'breakup both younger crust [Houtz and Ewing, 1976] and older crust unconformity' [Falvey, 1974] at or shortly following M-0 is (pre-M-4, Figure 12). The magnitudeof the velocities almost manifested on the J Anomaly Ridge by brief subaerial certainly has changed with time (i.e., the layer 2A-2B exposure of Aptian reefal carbonates and by truncation of conversion of Houtz and Ewing [1976]), but the relative the corners of tilted fault blocks. On the adjacent continental lateral differencesin velocity probably reflect initial differ- margins the breakup unconformity mostly is superimposed encesin crustal structure. Thus the original layer 2A (and the on the 'synrift' unconformity, but in deeper basinsit may not magnetic layer) at the ridge crest may have been twice as be present, dependingupon local uplift/subsidencehistories. thick as normal. Following the initiation of true drift between Iberia and the The mechanismby which the anomalousmagnetic body Grand Banks, the J Anomaly Ridge, the Madeira-Tore Rise, was emplaced perhaps is easier to explain. Although there and probably other adjacent oceanic and continental crustal are uncertaintiesabout the past location of hot spotsin this segments subsided at rates significantly greater than those region, the presenceof a mantle plume beneath the adjacent expressedby the normalt •/2relation. The northernpart of SoutheastNewfoundland Ridge at anomaly M-1 to M-0 most the oceanic J Anomaly Ridge and the continental crust conveniently explains the north-southdistribution of anoma- beneath the adjacent Grand Banks subsidedat very similar ly amplitudes. Again referring to the Iceland analog [Schil- rates [Gradstein et al., 1977]. However, other crustal seg- 9406 TUCHOLKE AND LUDWIG: STRUCTUREAND ORIGIN OF J ANOMALY RIDGE

ments appear to have had significantly different subsidence Gradstein, F. M., A. C. Grant, and L. F. Jansa, Grand Banks and rates (even within the J Anomaly Ridge), and a single the J-Anomaly Ridge: A geologiccomparison, Science, 197, 1074- 1076, 1977. regional subsidencecurve may not be applicable. Grant, A. C., Multichannel seismic reflection profiles of the conti- The model developed here agreeswith available geological nental crust beneath the Newfoundland Ridge, Nature, 270, 22- and geophysicaldata but needs further testing. Probably the 25, 1977. best tests will be found in carefully planned seismicrefrac- Grant, A. C., Geophysical observationsbearing upon the origin of tion and multichannel seismic reflection surveys acrossthe the Newfoundland Ridge, Tectonophysics,59, 71-81, 1979. Groupe Galice, The continental margin off Galicia and : proposed ocean-continent boundaries around the Grand Acoustical stratigraphy,dredge stratigraphy, and structuralevo- Banks. Although similar work in the eastern North Atlantic lution, Initial Rep. Deep Sea Drill. Proj., 47(2), 633-662, 1979. will be useful, it must contend with Late Cretaceous and Haworth, R. T., and C. E. Keen, The Canadian Atlantic margin: A Tertiary tectonic and volcanic modification of crust caused passive margin encompassingan active past, Tectonophysics,59, 83-126, 1979. by relative plate motions within the 'Eurasian plate' and Houtz, R. E., Crustal structure of the North Atlantic on the basis of between Eurasia and Africa. large-airgun-sonobuoydata, Geol. Soc. Am. Bull., 91,406-413, 1980. Acknowledgments. This researchwas supportedby the Office of Houtz, R. E., and J. I. Ewing, Upper crustal structureas a function Naval Research, contracts N00014-75-C-0210 and N00014-80-C- of plate age, J. Geophys. Res., 81, 2490-2498, 1976. 0098 to Lamont-Doherty Geological Observatory and contract Houtz, R., J. Ewing, and X. Le Pichon, Velocity of deep-sea N00014-79-C-0071to Woods Hole OceanographicInstitution. C. C. sedimentsfrom sonobuoydata, J. Geophys. Res., 73, 2615-2641, Windischserved as Chief Scientiston R/V Conrad cruise21, leg 11. 1968. We thank the officers and crew of R/V Conrad for their assistance. Jackson, R., C. E. Keen, and M. J. Keen, Seismic structure of the The OBH measurements were made under the direction of J. continental margins and ocean basins of southeasternCanada, Diebold. We also thank H. Schoutenand K. Klitgord for providing Geol. Surv. Pap. Geol. Surv. Can., 74-51, 1-13, 1974. unpublishedpoles of rotation for the reconstructionin Figure 13, W. Jansa, L. F., and J. A. Wade, Geology of the continentalmargin off J. Morgan for providing a preprint of his paper on North Atlantic hot Nova Scotia and Newfoundland, Offshore Geology of Eastern spot traces, and C. E. Keen for making us aware of K. D. Sullivan's Canada, edited by W. J. M. Van der Linden and J. A. Wade, work in the Newfoundland Basin. M. Sundvik and S. Rappelyea Geol. Surv. Pap. Geol. Surv. Can. 74-30, 51-106, 1975. provided technical assistance.The manuscriptwas reviewed by J. I. Keen, C. E., and D. L. Barrett, Thinned and subsided continental Ewing, G. M. Purdy, and E. Uchupi, who made many helpful crust on the rifted margin of eastern Canada: Crustal structure, comments. Contribution 5117 of Woods Hole OceanographicInsti- thermal evolution and subsidencehistory, Geophys.J. R. Astron. tution and contribution 3382 of Lamont-Doherty GeologicalObser- Soc., 65, 443-465, 1981. vatory. Keen, C. E., B. R. Hall, and K. D. Sullivan, Mesozoic evolution of the Newfoundland Basin, Earth Planet. $ci. Lett., 37, 307-320, REFERENCES 1977. Amoco Canada Petroleum Co. Ltd. and Imperial Oil Limited, Larson, R. L., and T. W. C. Hilde, A revised time scaleof magnetic Offshore Exploration Staffs, Regional geology of Grand Banks, reversalsfor the Early Cretaceousand Late Jurassic,J. Geophys. Am. Assoc. Pet. Geol. Bull., 58, 1109-1123, 1974. Res., 80, 2586-2594, 1975. Bodvarsson, G., and G. P. L. Walker, Crustal drift in Iceland, Laughton, A. S., and R. B. Whitmarsh, The Azores-Gibraltar plate Geophys. J. R. Astron. Soc., 8, 285-300, 1964. boundary, in Geodynamics of Iceland and the North Atlantic Carmichael, D., W. J. Ludwig, J. Ewing, G. Carpenter, and A. C. Area, edited by L. Kristjansson, pp. 63-81, D. Reidel, Hingham, Hubbard, A pop-up ocean bottom hydrophone, Mar. Geophys. Mass., 1974. Res., 3, 225-228, 1977. Laughton, A. S., et al., Site 111, Initial Rep. Deep Sea Drill. Proj., Chroston, P. N., C. J. Evans, and C. Lee, Laboratory measure- 12, 33-159, 1972. ments of compressionalwave velocities and electrical resistivity Laughton, A. S., D. G. Roberts, and R. Graves, Bathymetry of the of basaltsfrom D.S.D.P. leg 49, Initial Rep. Deep Sea Drill. Proj., northeast Atlantic: Mid-Atlantic Ridge to southwest Europe, 49, 761-763, 1979. Deep Sea Res., 22, 791-810, 1975. Dardel, R.A., and R. Rosset, Histoire g6ologiqueet structuraledu LePichon, X., J. Ewing, and R. Houtz, Deep-sea sedimentvelocity bassin de Parentis et de son prolongementen met, in Histoire determination made while reflection profiling, J. Geophys. Res., Structurale du Golfe de Gascogne, vol. 2, pp. 1-28, Technip, 73, 2597-2614, 1968. Paris, 1971. LePichon, X., J.-C. Sibuet, and J. Francheteau, The fit of the Davis, E. E., C. R. B. Lister, and B. T. R. Lewis, Seismic structure continents around the North Atlantic Ocean, Tectonophysics,28, of the Juan de Fuca Ridge: Ocean bottom seismometer results 169-209, 1977. from the median valley, J. Geophys. Res., 81, 3541-3555, 1976. Ludwig, W. J., J. I. Ewing, an.dM. Ewing, Structure of Argentine Dewey, J. F., W. C. Pitman, III, W. B. F. Ryan, and J. Bonnin, continental margin, Am. Assoc. Pet. Geol. Bull., 52, 2337-2368, Plate tectonicsand the evolution of the Alpine System, Geol. Soc. 1968. Am. Bull., 84, 3137-3180, 1973. Luyendyk, B. P., and E. T. Bunce, Geophysical study of the Egloff, J., and G. L. Johnson,Erosional and depositionalstructures northwest African margin off Morocco, Deep Sea Res., 20, 537- of the southwest Iceland insular margin: thirteen geophysical 549, 1973. profiles, Geologicaland GeophysicalInvestigations of Continen- Morgan, W. J., Hotspot tracks and the opening of the Atlantic and tal Margins, edited by J. S. Watkins, L. Montadert, and P. W. Indian Oceans, in The Sea, vol. 7, edited by C. Emiliani, pp. 443- Dickerson, Mere. Am. Assoc. Pet. Geol., 29, 43-63, 1979. 487, John Wiley, New York, 1981. Ewing, J., and M. Ewing, A telemetering ocean bottom seismo- Mutter, J. C., M. Talwani, and P. L. Stoffa, Origin of seaward- graph, J. Geophys. Res., 66, 3863-3878, 1961. dipping reflectors in oceanic crust off the Norwegian margin by Falvey, D. A., The development of continental margins in plate "sub-aerial seafloor spreading," Geology, 10, 353-357, 1982. tectonic theory, APEA J., 14, 95-106, 1974. Pillmason, G., Kinematics and heat flow in a volcanic rift zone, with Francis, T. J. E., I. T. Porter, R. D. Lane, P. J. Osborne, J. E. application to Iceland, Geophys. J. R. Astron. $oc., 33, 451-481, Pooley, and P. K. Tomkins, Ocean bottom seismograph,Mar. 1973. Geophys. Res., 2, 195-213, 1975. Pillmason, G., A continuum model of crustal generationin Iceland; Gradstein, F. M., G. C. Williams, W. A.M. Jenkins, and P. Ascoli, kinematic aspects,J. Geophys., 47, 7-18, 1980. Mesozoic and Cenozoic stratigraphy of the Atlantic continental Peterson, N., V. Bleil, and P. Eisenack, Rock- and paleomagnetism margin, eastern Canada, Canada's Continental Margins and Off- of leg 43 basalts, Initial Rep. Deep Sea Drill Proj., 43, 773-780, shore Petroleum Exploration, edited by C. S. Yorath, E. R. 1979. Parker, and D. J. Glass, Mere. Can. Soc. Pet. Geol., 4, 103-131, Pitman, W. C., III, and M. Talwani, Sea-floor spreading in the 1975. North Atlantic, Geol. $oc. Am. Bull., 83, 619-646, 1972. TUCHOLKEAND LUDWIG: STRUCTUREAND ORIGIN OF J ANOMALY RIDGE 9407

Purdy, G. M., The eastern end of the Azores-Gibraltar plate Tucholke, B. E., and P. R. Vogt, Western North Atlantic: Sedimen- boundary, Tectonophysics, 43, 973-1000, 1975. tary history and aspects of tectonic evolution, Initial Rep. Deep Purdy, G. M., and R. S. Detrick, A seismicrefraction experiment in Sea Drill. Proj., 43, 791-825, 1979. the central Banda Sea, J. Geophys. Res., 83, 2247-2257, 1978. Tucholke, B. E., et al., Site 384: The Cretaceous/Tertiaryboundary, Rabinowitz, P. D., S. Cande, and D. E. Hayes, Grand Banks and J- Aptian reefs, and the J-Anomaly Ridge, Initial Rep. Deep Sea Anomaly Ridge, Science, 202, 71-73, 1978. Drill. Proj., 43, 107-165, 1979. Rabinowitz, P. D., S. Cande, and D. E. Hayes, The J Anomaly in Tucholke, B. E., R. E. Houtz, and W. J. Ludwig, Sediment the central North Atlantic Ocean, Initial Rep. Deep Sea Drill. thickness and depth to basement in the western North Atlantic Proj., 43, 879-885, 1979. Ocean basin, Am. Assoc. Pet. Geol. Bull., 66, in press, 1982. R6hault, J.-P., and A. Mauffret, Relationships between tectonics Uchupi, E., and J. A. Austin, Jr., The geologichistory of the passive and sedimentationaround the northwesternIberian margin, Initial margin off New England and the Canadian Maritime Provinces, Rep. Deep Sea Drill. Proj., 47(2), 663-681, 1979. Tectonophysics,59, 53-69, 1979. Rothe, P., Shallow-water carbonates from site 384, Initial Rep. Vogt, P. R., The Iceland phenomenon: Imprints of a hot spot on the Deep Sea Drill. Proj., 43, 421-435, 1979. ocean crust, and implications for flow below the plates, in Rothe, P., and J. Hoefs, Isotopic composition of Cretaceous shal- Geodynarnicsof Iceland and the North Atlantic Area, edited by low-water carbonates from site 384, Initial Rep. Deep Sea Drill. L. Kristjansson, pp. 105-125, D. Reidel, Hingham, Mass., 1974. Proj., 43, 719-720, 1979. Vogt, P. R., Plumes, subaxial pipe flow, and topography along the Ruffman, A., and J. E. Van Hinte, Orphan Knoll--A "chip" off the mid-oceanic ridge, Earth Planet. Sci. Lett., 29, 309-325, 1976. North American "Plate," Earth Science Symposiumon Offshore Vogt, P. R., and A.M. Einwich, Magnetic anomalies and seafloor- Eastern Canada, edited by P. J. Hood, Geol. Surv. Can. Geol. spreadingin the western North Atlantic, and a revised calibration Surv. Pap., 71-23, 407-449, 1973. of the Keathley (M) geomagneticreversal chronology,Initial Rep. Saemundsson,K., Vulkanismus und Tecktonik des Hengill-Ge- Deep Sea Drill. Proj., 43, 857-876, 1979. bietes in Sudwest Island, Acta Nat. Isl., 2(7), 195 pp., 1967. Watkins, N. D., Paleomagnetismof the Canary Islands and Madei- Schilling, J. G., Iceland mantle plume: Geochemical study of ra, Geophys. J. R. Astron. Soc., 32, 249-267, 1973. Reykjanes Ridge, Nature, 242, 565-571, 1973. Winnock, E., Geologie succinte du Bassin d'Aquitaine (contribution Sibuet, J.-C., and W. B. F. Ryan, Site 398: Evolution of the west a l'histoire du Golfe de Gascogne), in Histoire Structurale du Iberian passive continental margin in the framework of the early Golfe de Gascogne, vol. 1, pp. 1-30, Technip, Paris, 1971. evolution of the North Atlantic Ocean, Initial Rep. Deep Sea Zverev, S. M., I. V. Litvinenko, G. Palmason, G. A. Yaroshevs- Drill. Proj., 47(2), 761-775, 1979. kaya, N. N. Osokin, and M. A. Akhmetjev, A seismicstudy of the Sullivan, K. D., The Newfoundland Basin: Ocean-continent bound- rift zone in northern Iceland, J. Geophys., 47, 191-201, 1980a. ary and Mesozoic seafloor spreadinghistory, Earth Planet Sci. Zverev, S. M., I. V. Litvinenko, G. Palmason, G. A. Yaroshevs- Lett., in press, 1982. kaya, and N. N. Osokin, A seismic crustal study of the axial rift Sullivan, K. D., and C. E. Keen, On the nature of the crust in the zone in southwest Iceland, J. Geophys., 47, 202-210, 1980b. vicinity of the SoutheastNewfoundland Ridge, Can. J. Earth Sci., 15, 1462-1471, 1978. Talwani, M., C. C. Windisch, P. L. Stoffa, P. Buhl, and R. E. Houtz, Multichannel seismic study in the Venezuela Basin and the Curacao Ridge, in Island Arcs, Deep Sea Trenches, and Back- (Received March 31, 1982; Arc Basins, Maurice Ewing Ser., vol. 1, edited by M. Talwani and revised August 3, 1982; W. C. Pitman, III, pp. 83-98, AGU, Washington, D.C., 1977. accepted August 12, 1982.)