3, PAGES 325-342, JUNE 2001. Cenozoic tectonics of the Cape Roberts Riff Basin and Transantarctic Mountains Front,. Southwestern Ross Sea, Antarctica.
TECTONICS, VOL. 20, NO. 3, PAGES 325-342, JUNE 2001
Cenozoic tectonics of the Cape Roberts Riff Basin and Transantarctic Mountains Front, Southwestern Ross Sea, Antarctica
RheaJ. Hamilton, • BruceP. Luyendyk, andChristopher C. Sotlien Institutetbr CrustalStucl•cs,Universityof Cal•tbrn•a,SantaBarbara,Calttbrn•a
Louis R. Bartek2 Departmentof Geology, Universityof Alabama, Tuscaloosa,Alabama Abstract.
We conducted
a !nultichanncl
SOlSlille
reflection
survey offshore Gape Roberts, Antarctica, and conlbined our f•!ld!llgS with the results of the Cape Roberts International Drilling Project (CRP). This allows us to interpret Cenozoic tectonics in the southwest sector of the Ross Sea including tile history of tiplift of the Transantarctic Mountains (TAM) and subsidence of the Victoria La!ld Basin (VLB). Seisln•c stratigraphic sequencesnlapped offshore Cape Roberts are tilted eastwardand t11•ckeninto tile VLB where they conlprise nlore than half the fill seen on selsnlic
records. Normal
faults a
during the Cenozoic [Me/nat c! a/. 1975; Stock and Me/nat, 1982, 1987]. Related to this by ilnplication are the details of tile uplift history of the TransantarcticMountains (TAM)and subsidenceof tile major basilis in the Ross Sea, especially the westernmost Victoria Land Basin (VLB), and how these events
may be explained witlain a plate tectonic setting. The tiplift history of the TAM was one focus Of Cape Roberts InternationalDrilling Project (CRP) recently completed at the !nargin of the Ross Sea adjacent to the TAM [Barrett and Dr•vc')', 1992; International Steering Committee, 1994]. We conducted a marine geophysical survey there in February 1996 using the RVIB Nathaniel B. Palmer (NBP9601) as part of the preparation for the CRP (Figtire 1). We acquired SeaBean1 2112 swath bathylnetry, chirp sonar, gravity, n•agncticfield, and over 250 kin of single-cha!lnel (SCS; two channel and single fold) and 22-fold !l•ulticha!lnel (MCS)
few kilonletcrs offshore cut these sequences and define a north trending rift graben. Drilling resultsfi'oln the CRP sho•v that thesestrataare latest Eocene(?), Oligocene, and younger Ill age' nluch younger than previously inferred. Wc lllterpret this pattern to be due to an episode of E-W extension and related subsidence that occurred across the major basins ill tile western Ross Sea durillg the early Ce!lozoic. The rift grabcnofigho!'cand adjacent to Cape Roberts is bounded on the west b• a m•kjornorth trendingfittilt/one. At Cape Roberts rials fault systClll !nay 111xeI•'o!116 to 9 km of vertical separation.Tills timIt system is part of a larger zone along tile coastlille in southernVictoria Land that aCCOlnmodated tiplift of the TAM in Oligoce!le t•me. We name !t here the McMurdo Sound Fault Zone. A late Oligocelle angular unconfornlity that is seen in s½lsnlicdata and sampled by CRP drillillg marks tile end of east tilting of the stratigraphic sequences. We interpret this as the end of tile main uplift of the TAM coinciding with a c]lallge fi'O111E-W extension to NW-SE oblique rifting at that time. Upl•t't of tile TAM and subsidence 111the VI,B 11135, be I•nkcd with seafloor sp!'eact•l•gon the Adarc Trough to the northwestof the Ross Sea between43 and 26 Ma. Th•s woulct •n•ply a plate boul•dal'y between East and West Antarctica crossing tllrou-h the western Ross Sea in Eocene and Oligoc½llCt•me.
Cml•llasize results from the inte!'l•retation of our seismic reflection data that place the drilling results of tile CRP irl a strat•grapllic and st!-uctt•ral context. We also relate our tectonic interpretations for the southxvestc!'nRoss Sea to larger-scaleplate tectonic episodes that formed and shaped tiaoIlla.jor N-S trending basins ill the Ross Sea.
1.
I.I.
Introduction
seismic
data.
Cape Roberts is found in the southwestern Ross Sea, Antarctica,
125 kill north of tile United
States McMurdo
Research Station on Ross Island (Figtire 1). The CRP employed drilling from annual sea ice a few kilometers offsllorc into a suspectedsequenceof Cenozoic and Mesozoic strata. The CRP was ailned at recovering strata up to 100 million years ill age adjacent to the range fi-ont of tiao TAM. Tile project goals are to a!lSWCFquestions conocr!ling the glacial history of Antarctica, tile history of the Ross Sea rift, and the uplift of the TAM. In our study we address specific issues
about
Tectonic
this
sector
Evolution
of the southwestern
of the Western
Ross
Sea.
We
Ross Sea
The in]portant exents of the Ross Sea tectonic l•istor5 are bel•cvectto be of Mesozoic and Cenozoic age. Tile oilset of the Gonctwanabreakupin the Jurassicleft its imprint in the Ross Sea and TAM region. This was markoctby a major •gncous •NowatIESE,Un•xers•dad deNaxarra, Barcelona, Spa•n 2Nowat [)cpart•ncnt of Geological Sciences, Unix crs•tyof North event at •18() Ma that •nvolvcd the eruption and intrusion of Carolina, Chapel 11•11, North Ca•oi•na the Jurassic tholci•tic Kirkpatrick Basttits and the Fe!TalDoloritc that are widespread in the TAM [K)'/e ct rz/., 1981; Copyright2001 by the American(•cophys•calUnion E//ioi, 1992]. l)rzs'c)'rtnd Hrrt•ir'o/i•ii [1995] discussed four tectonic CVClltsthat were significant in the exolution of tile Papc• nitrebel 2{)()0TC001218 0278-74(}7/01/20(10TC00121 Ross Sea in Early Cretaceousand later t•nle. These •nclude (I)
Along-standing problem in Antarctic tectonic history is xvhctherthe Antarctic plate was separated into two plates
325
326
}iAIVIilZi'ON E'I' AL.: CENOZOIC TECT(}NICS, SOUTtiWESTERN ROSS SL!A
AT.• 1•. ø
-•-
•--
Pacific Ocean
'••P2-73 "/'T•'
•Ooo ••
..Ros'Seai. --.
Cape '••--•
....
• '..•-':•
.Marie•
EastAntarctica
Figure I. (a) RossSearegionmapshowingthe locationof the studyarea,outcroponshore(dark shading),DeepSea Drilling sites,the Transantarctic Mountains,and CapeRoberts.AT, Adal-eTrough(dashedarrowsindicateformer spreading directions [Candceta/., 2000]); NB, NorthernBasin;VLB, VictoriaLandBasin;CT, CentralTrough;EB, Eastern Basin; NVL, Northern Victoria Land; SVL, Southern Victoria Land. The West Antarctic Rift System is
outlinedby the light shading[Le/Plaxurier,1990].(b) Map of regionincludingCapeRobertsand McMurdo Sound, showingthe surveytrack of N. B Palmercruise96-01, locationsof CapeRobertsProjectdrill holes,and the McMurdo Sound Fault Zone (dashed l•nes; see text). Drill holes CIROS-I and MSSTS-! are described by Barrett
[ 1986, 1989],respectively. The Dry Valley Drilling Pro, jcct(DVDP) is described by/?lcGinnix[ ! 98! ].
a in[[jor extension episode that began •105 Ma that correlated xvith tile end of PJloelllX plate subduction tllldel' Antarctica and Gondxvana[[.•l'•ir/.¾h(iil', 1989; Lt(l'ct•r/t'k, 1995], (2) seafloor spreading between West Antarctica aild Nc\v Zcaland/Calnpbell Plateau that began •83 Ma [Stor'k (•'an(h', 1999, 2001], (3) the main uplift of the TAM starting at 55 Ma [Fitzgcru/r/, 1992] possibly in responseto the end of spl'cadil•g in the Tasman Sea and acceleration of Sj)l'eadillg bctxvccnAustralia and Antarctica [ll,'c,i.v.¾cl ut a/., 1977; anr/ :¾htllcr, 1982], and (4) volcanism and tl-anstfnslona] faulting beginning in Oligocene time. Item 4 may be related to m•ddle Cenozoic seafloor spreading in the Adare Trough northwestof the Ross Sea (Figure 1) [Stock ant/ Cant&, ! 999, 2001; Crmdc eta/., 2000].
1.2. Stratigraphy of the Victoria Land Basin
it has beenproposedthat rifting and subsequentextension of tile Ross Sea in the iniddle and Late Cretaceous time led to
the initiation of the Northern Basin (NB), the Central Trough (CT), and the VLB (Figure 1) [e.g., Drtt,'cst', 1981]. These basins were actively subsiding during Mesozoic time and dul-ing later phasesof rifting in Cenozoic time. The illot-e rcccnt (Late Cenozoic) lootis Of activity in the Ross Sea has been in tile Terror Rill of the VLB that is immediately east o• Cape Roberts (Figure !) [Dave)' m•d BFam'o/ini, 1995]. The VLB was described by Cooper (tnd Dczv•Lv [1985] a•d Cooper c! rJl. [1987, 1991] as a rift depression over 150 kin wide that extends fi-om Ross Island to Cape Washington (Figure I). Greater than average heat flow has bcell nlfasured ill the VLB [[•/(IC/(i'l'l[iF1 CI (l/. 1987] that can bc cxplained by a shallow mantic and igneous activity in an area of extended crust. An estimated 14 km of strata fill the basin [COOl)C,F el a/., 1987]. A thick sectionof Oligoccnesedimentsin tile Cape Roberts drill holes can be con-elatedto the VLB, suggesting that much of this ellOl'lllO[iS thickness
of' sediment
is of latest
Eocene and younger agc[C•q•e RobefLy Scicm'c Tca,•, 2000; Da•'c•' ct •t/.. 2000].
HAMILTON ET AL.' CENOZOIC TECTONICS, SOUT!4WESTERN ROSS SEA
32'7
-76 ø 15' WESTERN
ROSS
SEA
-76 ø 30'
12'"•::"•,:;:•.':•,CRP-I' 2/2',•' 3
-77 ø 00'
74 ......:%:!
.......',.
...... '.....
'......................
;.'.:.'.½3•. 4. ':' ': ':.......... :"..7'-'.:................................
..........
'"':'"' ............ :
CapeRoberts\
•, \ •
'" ': NBP9601Shiptrack
",.
.........
MCMU' .RDO•.....
SOUVD,,, ..... .,,'
'•' ROSSISLAND
-77 ø 30'
....
r• CIRO
Mt. Erebus
61.2MCMURDO
STATION/
SCOTT BASE
-78 ø 00'
km
•0
-78 ø 15'
164
1
168 ø
Figure 1. (continued)
Cooper et al. [1987] established a nomenclature for seismic
units
in the VLB
and western
Ross
Sea based on six
seismic stratigraphic units (V1-V$ sedimentary and V6 volcanic) above an acousticbasement(V?) (Figure 2). These V seque•ces have been correlated to seisrhic sequences in the Northern, Central, and Eastern Basins (Ross Sea Seismic (RSS) units and U unconformities of Hinz and B/ock [1984]) by Branco/ini eta/. [1995]. However, seismic units in the west rind easternRoss Sea do not carry across the highs bounding the basins, so that seismic lines do not tic them together. Bortek et •/. [1996] later developed a more detailed letter nomenclaturefor the V sequencesbased on interpretations of coherentVLB seismicsequencesseenon RV Po/ar Duke highresolution seismic data. The V sequenceshave been correlated to Roberts Ridge offshore Cape Roberts by Da•'ey eta/. [2oo]. Age control on V seismicsequenceshas been uncertain and has relied on data fi'om a'f•w drill sites prior to the CRP [Dave)' e! a/.. 2001]. Prior drill sites sampled mid-Cenozoic sectionsand includethe Deep Sea Drilling Project(DSDP) Leg 28, Site 270 in the central Ross Sea [H,::o'e,v and F,,'akes,1975] and the Cenozoic Investigation of the Ross Sea core at the southern enct of McMurdo Sound (CIROS-1 [e.g., Ha•n•d•, 1994; I4/i/.vone/a/., 1998]) (Figure I). The oldest sediments
Ill CIROS-1areglacialandglacial-marine stratathat areearlylateEocenein age[Barn'ell,1989; Hannah, 1997; Hannahct a/., 1997; Wi/.vonel a/., 1998]. Unfortunately, faulting, unconformitics, and shalloxvwater bottom multiples have madetile correlation of Eoceneand Oligocenc strata sampled at CIROS-I to the VLB and to Cape Roberts uncertain [Bartek eta/., 1996; Henrys eta/., 1998]. Despite the previous drilling efforts, little is known about the time period between 160 and 45 Ma [Barrett eta/., 1995; Bartek eta/., 1996]. No rocksor sedimentsof this age have been sampled in the TAM or in other parts of East Antarctica [Davey, 1987; Barrett et a/., 1995].
The lack of stratigraphiccontrol in the westernRoss Sea has recentlybeen addressed by the drilling effortsof the CRP. Over 1500 rn of stratarepresentingthe period from 16 to •34 rnillion years ago have been recoveredin three drill holes. More than 750 m of core were obtained fi'om CRP-1 (oldest material Miocene) and CRP-2/2A (oldest material Oligocene) drill sites, positioned on the thick sea ice offshore Cape Roberts [Barrett eta/., 1998, Fie/cling and Thomson, 1999] (Figure 1). The third site, CRP-3, drilled in November 1999, penetratedrocks interpretedas Paleozoic Beacon sandstone below Oligocene and latest Eocene (?) sedimentary rocks [Cape RobertsScienceTeam, 2000].
328
HAMII.TON ET AL.' CENOZOIC TF.CI'ONI('S, SOUT!tWESrI'ERN ROSS SEA
UNIT
AGE
BAR•:.
BARTEK
..... TEcToNICS COOPER ETAL.(fromCRP-I,'2J2A ETAL.-igT. AL. ........:;.:•:.• •:.......... ......
.:
...... •::•. -.,,..=..:.., :. .... •:• .....................
FACIES/COMMENTS
[1987]; thispaper • .............. :..............................................................................................
:.........
Strike-slip faulting
•
Decreased uplift
=
inTAM;
V3
•
•t Volcanism
r.•
,-4 ........................................ 90 mbsfC•P-2./2
•, l D.ecreased uplift :ii'i .......:......... •_ •_.
p/Q
< 21.2 Ma Miocene
.m.XAM; I;' V4a ¾•!canism ii:•:• ...... ,•::•
'• •t
.23..7 •24..1Ma Lateofigocene
•...... •'-:,•'•"•.:i•11 ':' ..... •' •..'"•p'T' •:'•"•i!e•
•':":•'•:':'•'• .....• ......': '•'":"• .......... •"'•::•:'•:•"•'""'• :'•........ •............ •.•-•-:•:•½;::½•:•.•=:•...•½•G•:; •...................................'"'•
Tfa•tblock tilting :•: "V4b •::'
•
androtation. =:'•
•
•-:
••
......................
-• •••
...... '•
:•:• • Late
Glacial marine sediments.
Evidence ofglacial scouring inseismics.
Glacial marine sediments.
R
Voicanics and pumice tuff and detritus ß ß
•
demvedfrom gramte basement
............................................................
•
:•
Glacial marine sediments.
S
De•itus from J•assic dolerites andlavas
?:=OHgocene ?
Angular unconformity attop and onlap surface atbase.
3•-robs. C•P?•.&;;......::....:-.:•:: ......... ..-.:.•. :..•.:•:.::•..;-":'•: •::'• "•'•'"':•'.............. ?":(?ab0•'•2•'-•bel• •.... ............ ..• :..•:.•;-. L.: =-•• ->": ..... .... ' U'nc6 "'.... "•6'F•it:y""RSU-6? :"'"" .......... :-• •'z'•'•' "';'""''• •" •"
=• • _• •. •= E-W rifting } ;•.•:•.-• :• •,v; •. •via Increased uplift •-• ..:..•":':-•a' .• •ariy ongocene• _ _ •:• • ........... :•'2• •
] •9 '1'ß ]••
Glacial marine sediments. Oldest s•ata sampled byCRP-•2A. -' •o seismic evidence for angularunconformities
rifting • •:•:• •:•'":"'•'•: ?? Lat•tEocene • E-W lnc•ased •lift ?•---/•'•5:b:.+V' -'.:• -. V5
-
Sampled inCR-
V"6 '?•Late O!igocene V6
--
[
....... "
Pull-aparts, •=-> >...........:'•• .•[•• toRecent
Sampi. CmOS-l Volcani•
..............................• ,••-•••,•••• ......................................................................................................................................................
•'¾' '7• Gondwana •:'=•Z7 bre•up •t.......
i Ill
II
I I I nilI1
II
I
•urassicV7 I
I Illllll IllIll
Sedimen•ry (Beacon), volcanic &igneous & metamorphic basement rocks
• IlllllllllIlll
II
Ill
II
IllllII I
I
Figure 2. The stratigraphicuntt.sintcrprctcdoffshoreCape Robcrts,Antarctica, based on our intcrprctatton of Cape Roberts ctriil•ng results and the NBP96()I reflection seismic data [modtficd fi-om Hamilton ct o/., 1998]. A comparisonis made t'othe nomcnclatt•rcs developedby Cr,o!)crcl a/ [1987], Burrell c[ ri/ [1995], and Btirlc/( el rt/. [1996] basedon •ntcrprctat•onsin southernV•ctoria Land and the westernRossSea. Agcs and geologic descriptions l•r the V units are fi'om ('oo/w•' cl o/. [ 1987] and the Cape Robertsreports(see text). V3, V4, and V5 are sedimentary units, V6 is x-olcantc, and V7 is acoustic basement (undiffCrcntiated),
1.3.
Transantarctic
[Mountains
and
West
Antarctic
Rift System
The TAM are a 4000 km long and 4 km high mountain range that was created in a nonconvcrgtng plate tectonic setting during Cenoxo•c time (Figure 1) [e.g., Dav•3,' and Brrmco/ini. 1995]. The TAM const'•tutethe wcst shouldcr of a major continental rift system, the West Antarctic Rift System ('WARS). The WARS also includesthe Ross Embaymcnt (Ross Sea rift [Tcsscnso/•nand 14'oi'ncr.1991]) and western Marie Byrd Land that comprise an extended region of crust and the easternriff shoulder,respectively(F•gure i) [c.g., LcMa.s'itriei'. 1990; Bc/trcndl ½!a/.. 1991]. The WARS is one of the largest extensionalregimeson Earth and is comparable in size to the Basin and Range province of the western United States [LcMa.s'ui-ici'.1990]. Much of the WARS lies under an ice sheet. maki•g it dtfficull to dclcrrn•nc lhc history and current actixit5 in the rift. The WARS •s considered active today by many researchersbecauseext'cns•vclate Cc•o7.oic volcanism is tbund here along with a riff shoulder w•th 4-7 km of relief at it's western border along the TAM [Bc/v'cndt c! o/.. 1991: Bchrc•tdt •t•d ('oot)cr, 1991]. Ex,•denccfi'om apatitc fission
track dating [Filz,,4crald,1992, 1994, 1999] suggests that significantexhumationof the TAM •n southernVictoria Land beganat •55 Ma. Ftlz•crrt/d cl tt[ [20()(}1find evidence Igr exhumation
of the western TAM
•n southern
begtnning tn m•d-Cretaceous ttmc
Victoria
Land
t:itzgcrold [1992]
estimated the total amount of Cenozoic rock uplift •n the TAMtobc6kma!lowinglbr 4.5-5.(} km ofcrosion along the east stde of the TAM
1.4. Cal)e Roberts
CapeRobertsis locatedin southernV•ctoria Land at the foo! of the TAM Front, describedby Bt,'rc!! c! el. [1995] as a 30 km w•dc zone of faults bordcrin,, the cdf•c of the uplifted crustal block of the TAM. The bathymctry off'shore Cape Robcrtsconsistsof the V-shaped(in map x tcxv)trough that we intcrprct as a grabcn, the Cape Roberts Rift Basin (CRRB; Figure 3). This 400-500 m deep trough is bounded by the foothills of the TAM to the west and by Roberts Ridge, a 100 m deep bathymetric high •15 km off,shore [Barre!! c! 1995.' Bar!el,'el a/., 1996]. The three drill holes of the Cape RobertsProject,CRP-I, CPR-2/2A, and CRP-3, are located on
HAMILTON
162ø50'
163"00'
ET AL.' CENOZOIC TECTONICS, SOUTHWES•I'ERN ROSS SEA
163"10'
163"20'
163"30'
163ø40'
329
163ø50'
-76" 55'
-77 00' ..
i;:.i:"
_ .•.... .....•BP89
CapeRObe rts.:"
•• 162"50'
NBP87
• 163'00'
163' 10'
--,•
' i ".:.- •
:;:•.
-':L'•"...... ::•:'•" :'::•::"•;::•:'
.•i:':,:-•
Cape Robe•s Rift Basin 0 5 10 163' 20'
1•' 30'
163'40'
'•'•'"•'•--...:.• .
163' 50'
Figure3. A high-resolution, shaded reliefbathymetric mapfromSeaBeam 2112datashowingthe N. B. Palmership trackduringthe NBP9601RossSeaexpedition.RobertsRidge,a bathymetric high on the easternmarginof the surveymea,theMacKaySeaValley,a deepglaciallyscoured troughto thenorth,andthe CapeRobertsRift Basin,a V-shapedbathymetric low,areprominentfeaturesof the studyareaoffshoreCapeRoberts.The CRP-1,CRP-2/2A, andCRP-3drillingsitesareshownalongwith locations of seismiclines87, 89, and97. (Thecolorversionsof this and the seismic reflection figtires are available on the web at the UCSB Cape Roberts researchwebsite, http://www. crustal.ucsb.edu/caperoberts.)
330
HAMILTONETAL.:CENOZOIC TECTONICS, SOUTHWESTERN ROSSSEA
RobertsRidge (Figures 2 and 3). The Pleistocene age MacKay Sea Valley truncates the CRRB to the north in Granite Harbour. Bathymetric depths range fi'om 100 m along the easternand western margins of the CRRB to depths of over 1000 m in the MacKay Sea Valley. Along the coastat Cape Roberts, Late Precambrianfoliated and nonfoliated gneissic granites form the basement rocks of TAM. These were deformed during the early Paleozoic Ross orogeny [see Stump. 1992]. The basement is separated from overlying Devonian to Triassic terrestrial sedimentary rocks of the Beacon Supergroup [see Woo/re and Barrett, 1995] by the Kukri Peneplain, a regional erosion surface found in the TAM except for northern Victoria Land [Barrett e/a/., 1972; Gunn and Warren, 1962]. Basement rocks are intruded by Ordovician Granite Harbour lntrusives (granites). The Beacon Supergroup is intruded by distinctive tholeiitic Jurassic Ferrar Dolerite sills and overlain by the equivalent Kirkpatrick Basalt [Gunn and I4/rarren, 1962; Fitzgerald, 1992; Barrett et al., 1995]. The sills and basalts mark a major igneous event at-177_+2 Ma [Heimann et al., 1994], corresponding to the initial stages of Gondwana supercontinent breakup in Jurassictime. The (late) Cenozoic McMurdo alkali volcanics [Kyle, 1990] are found 80-150 km to the east and south of Cape Roberts at the south end of the Terror Rift. Igneous intrusions are interpreted along the west side of the CRRB that may be part of this series [Hamilton et al.. 1998; Bozzo et ai. 1997a, 1997b]. The NBP9601 geophysicaldata, combined with the results fi'om the CRP holes, allow interpretations of the stratigraphic and structural relationships offshore of the TAM at Cape Roberts. We propose that a series of tectonic events, commencing with increased uplift in the TAM and E-W extension in the western Ross Sea, influenced the evolution
of the CRRB and RobertsRidge.
could not handle reflections from the complex structures within the CRRB. Typically, the split-step migrationresulted in the overmigration of reflections. Because higher frequenciesare spatially aliasedduring F-K filtering, a 6-50 Hz or 6-40 Hz band-passfilter was applied to all of the final displays. In general, there are significant problems with the data,
especiallyreflection data fi'om sinallowwaterdepths. This is due in part to a noisy near trace and dead third trace or to a problem in synchronizing gun firing. The fact that only near traces can be used at small travel time due to NMO
it is difficult
Seismic
Reflection
Data
Two hundred-fifty kilometers of MCS and SCS data were recordedsimultaneouslyby firing alternatingsourcesevery 5
s. The MCS sourceconsistedof five 3.44 L generator/injector
to choose accurate velocities
for local moderate
or steepdips within the CRRB.
3. Interpretations of Stratigraphy for Offshore Cape Roberts All available MCS and SCS data (NBP9601, RV Polar Duke
(1990), OsservatorioGeofisico Sperimentale(OGS) (1990) and U.S. Geological Survey (USGS)(1984)) were used to correlatethe seismic V sequencesfi'om the VLB to the CRP drill sites on RobertsRidge [Dave)'eta/., 2001]. This effort resolved tiaoambiguities in seismicstratigraphy on Roberts Ridge pointedout earlier by Heno,seta/. [1998]. The seismic
reflections were correlated to the drilling results by conqputingsyntheticseismogramsusing velocity data from
holeCRP-2[Henoweta/., 2001] andverticalseismicprofiles in CRP-3 [Cal)e RobertxScienceTeam,2000]. In our study we identified the V3-V5 and V7 sequence boundaries of Coot)er et al. [1987] according to new interpretations of seismicdataon RobertsR•dge[Drive)'eta/. 2001]. We defined a seriesof subsequences(V4a, V4b, V5a, V5b, V5c, V7a, and V7b)on the basis of high-amplitude reflections
2.
stretch also
made the removalof multiplesin shallow water very difficult. Becausestacking velocities increasefor a particular interval velocity structureby 1/cosineof the dip angle [Yi/maz, 1987],
that show distinct
characteristics
and/or bound
seismicsequencestratigraphicunits (Figure 2). 3.1. V3/V4aBoundary Our V3/V4a boundarywas sampledby CRP-2/2A and dated
(Gi)air guns. Data were recorded on 44 live channels (out of 48), xvithhydroplaoncgroup spacing at 25 m and a maximum
at •21 Ma IF/c/clingand Thomxon,1999].In the northernpart of the study,V3 onlapsV4a, and bothunitsshowevidencefor
shotgrouprangeof 1420 m. Shots wereequally spacedat 25 m to resultin 22-fold stacks.One 3.44 L G! air gun was used as the sourcefor the SCS data. GPS was used for navigation throughout the survey. The suppression of strong waterbottom multiples was the main goal in processingthe SOlSlilledata. Velocity analyses were done every k•lometer. A fi'equency-wavenumber (F-K) filter was applied after a partial normal move out (NMO)
glacial erosion.
correctionusingvelocitiesintermediate betweenthe primary and multiple arrivals. After predictive and spiking deconvolution,both ordinary and median amplitude stacks were created.The median amplitude stacks were effective at removing multiple energybut degradedthe data above 0.5 s
two-waytravel time (TWTT).In somecasesan F-K migration at 1460 m/s was performedpost-stack,and a split-step migration [Stoffa eta/., 1990] was performed(done on line 97). Bothof thesemigrations wereusedto aid in interpreting the dataon a line by line basis.However,the F-K migration
3.2. V4a/V4b Boundary
We interpret the V4a/V4b boundary as an angular unconformity that coincides witIn an abrupt change in "clast and grain parameters"at 307 m below seafloor (mbsf) in the CRP-2/2A core hole [Fielding and Thomson, 1999; Sine/lie et al, 1999]. Henrys et al. [2001] present a synthetic seismogramwhere top V4b is interpretedat 270 mbsf in CRP2/2A. This conflict is explained if their seismic correlation is one cycle too shallow. Drilling results date the oldest sediments above the 307 mbsf unconformity separating V4a and V4b in CRP-2/2A as 24.1 Ma [Cape Roberts Science Team, 2000]. The youngest dated sediments below the V4a/V4b boundary are in V5a and are •29 Ma (Figure 2) [Cape Roberts Science Team, 2000]. On Roberts Ridge the V4a/V4b unconformity reflection onlaps the V4b/V5a boundary to the west (Figure 4).
HAMILTON ET AL.: CENOZOIC TECTONICS, SOUTHWESTERNROSSSEA
Two-Way Travel Time (s)
ß•>
(s) eWLt P^•J.L/•M-OMI
•
•i >
331
332
HAMILTON ETAL.:CENOZOICTECTONICS,SOUTHWESTERN ROSSSEA
3.3. V4b/V5a Boundary Unit
V4b
thickens
toward
3.6. Acoustic Basement, V7a and V7b the
east
into
the
VLB.
It is
separatedfrom VSa by a sequenceboundary that is an onlap surfacecorrespondingto 443 mbsf in CRP-2/2A and to the late-early Oiigocene boundary [Fielding and Thomson, 1999]. Davey et al. [2001] interpret the V4b/V5a boundary as RSU6, a regional sequenceboundarydefined in the Eastern Basin by Hinz and Block [1984]. RSU6 is older than 26 Ma, where it was first defined in the Eastern Basin near DSDP 270 [H•ows and Frakes, 1975]. Unit VSa crops out at the seafloor on the
western side of Roberts Ridge and gradually thickens eastward, similar to V4b. The oldest strata sampled by CRP2/2A
were within
the VSa unit
and were dated
at > 29
Ma
(early O!igocene [Cape Robertx Sewnee Team, 2000]). Hole CRP-2/2A 3.4.
reached 624 m below seafloor.
Subdivisions
of V5
We subdivide
V5
into subunits
V5a, V5b, and
V5c,
separatedby strongreflections.Unit V5 is several kilometers thick in theVLB,
and these subdivisions allow us to discuss
the t•ming of initial erosionin the TAM and deposition•n the VLB. The velocity and two-way travel times derived by the Cape RobertsScienceTeam[2000] indicatethat the reflection at the top of V5b may be an increase in velocity at a conglomerate bed near 100 mbsf in CRP-3 (their reflector "p"). it is difficult to trace the top V5c reflection to CRP-3, but it may be the top of high seismic velocity conglomerates near 770 mbsf in CRP-3 (Cape RobectxScience Team [2000] their reflector "w"). The V5b and V5c strata (Figure 5) thicken east toxvard the western VLB, where deformation of V5b is l•mited to gentle tilting and minor (300 m or less) separations acrosshigh-anglefaults [e.g., Cooper eta/., 1987].
ntunerous Ulqconforlnities, fattits,
fi'om acoustic basement are seen under Roberts
between 2.5 and 2.8 s TWTI'
beneath CRRB
and beneath
the
crest of Roberts Ridge at 1.4 s TWTT. This V7 reflection, which
is seen at the intersection
of seismic
lines
USGS-401
and 403, is on strike with the location of CRP-3 (location in Figure 6). V7 pickedon the USGS lines ties to a reflectionjust above 1.5 TWTTon NBP9601 line 97 at shot point 2320. We label
this
reflection
as V7a.
CRP-3
bottomed
in
sandstone
interpreted as Devonian Arena Sandstone of the Beacon Supergroup [Cape Roberts Science Team, 2000]; this is V7a. The contactof the Beacon (V7a) with overlying Oligocene (?) conglomerateunits at 823 mbsf is an erosional unconformity [Cape Roberts Science Team, 2000]. We identify another basement reflection, V7b, under the east flank of Roberts
Ridge at 2.4-2.5 s TWTT (Figure 4). The reflection characterof V7a and V7b are distinctly different. Although both are acoustic basement reflections, they are not necessarily the same geologic unit. V7b on line 97 ties to a reflection on cross line NBP9601-94 (Figure 6)that in turn ties to a reflection interpretedas acousticbasementon 1%69 by Dave)' c/ u/. [2001]. Therefore we believe Otll' interpretation of this reflection is robust. A reflection at 1.8 s TWTT in CRRB may be Beacon(V7a, Figure 4), but this is much shallowerthan the •nterpretation of V7 by Cooper e! a/. [1987]. Th•s reflection could be from volcanic units (V6) instead.
4. Interpretations Cape Roberts
of Structure Offshore
Our interpretations fi-om the NBP9601 seismic reflection
3.5. Correlation to Cape Roberts Rift Basin The
Reflections
Ridge and in the CRRB (Figures 4 and 5). Cooper et al. [1987] interpretedan acousticbasementreflection (V7) as the Beacon Supergroupoffshore Cape Roberts. They interpret V7
and
igneous
IntrusionsoffshoreCape Robertsmade it difficult to correlate seismic reflections continuously across the CRRB. We interpret the reflective sequencesin the upper 150 ms TWTT in the southernpart of the CRRB to be a veneerof Quaternary (?) glacial sediments.A thin remnantof more reflective V4b strata is bounded above and below by unconformitiesin the CRRB. We correlatehigh-amplitude reflections seen below thesesequencesto be the erodedtop of the V5a unit (Figure 5). The V5b unit can be traced into the basin acrossthe west dippingfaultson RobertsRidge;however,this unit is poorly imaged. Cooper eta/. [1987] have also interpreted V5 (undifferentiated)in the CRRB. We are unsurewheretop V5c is in CRRB, so we do not attemptto differentiateit there.
datasuggest thepresence of threedist,inct faulttrends:north to NNW striking and NNE str•kin& normal-separationfaults borderingthe CRRB, and ENE striking faults, crosscuttingthe graben and Roberts Ridge. A fault map on the top of V4b is shownin Figure6. Someof the faults shown in Figures4 and 5 are not mapped in Figure 6 becausethese Faults do not offset
V4b.
4.1. North to NNW Striking Faults
The major north and NNW striking (labeled A, B, and C) and NNE striking (D and E)normal faults dip toward the CRRB and define its V shapein map vlew (Figure 3). These faults werenot mappedin the glacially erodedtrough of the MacKay SeaValley due to the lack of' E-W seismic profiles
Figure 5. ParallelE-W seismicprofilesacrossthe escarpmentseparatingRobertsRidge (RR) and the Cape Roberts rift basin. Line 87 is the northernmost,and line 89 is the southernmost(Figure 6). NNE striking, west dipping normal-separation faults including "D" and "E," and NNW striking flower structurespresumedto be right-lateral faults (e.g., "G")are shown.BeneathRobertsRidge the V3/V4a boundaryis locally scottred,and there is a local angularunconformitywithin V4a; V4a/V4b is an angularunconformity,and V4b/V5a is a subtleonlap surfaceon theseprofiles.For line 97, "TB" pointsto a tilted block, and "FPR"points to a non migrated,westdipping fault planereflection.Projectedlocationof drill holes CRP-I and CRP-2/2A and CRP-3 are shownon line 89 (bottom). The westdippingfault F is from Hamiltoneta/., [1998]. This fattit is not mappedin Figure6 becauseit does not cut the top of V4b, and its vertical separationdeeper in the sectionis uncertain.
HAMILTON ET AL.: CENOZOIC TECTONICS, SOUTHWESTERN ROSS SEA
.333
West
1•0õ
115•
1•05
t-91 [
•05•
•0o5
East
t-92Line 87t-401t.9.••'•1 t-403
lk
-*, '
_ .
•
,,-%•V•.an• younger
..... ..-
.•....
.. ':'
o
... '• - '"""•::••-,•4"'•'•"• .......... :L• ....",...... '-• ..... •-•'•:...::,•..:.•:..-.•.• .-•-• .-:•.' ............ • .....- '.•.-............. ,...... -'-•-.• .... •'r-'••V5•, •-,. ::..•:......... "•': ':.:-:5'•,-' .,.:-• :'•'•'.-:.,: •..',:....•,-:-{?-.'•;..:::::•u '?:"'b.•: ':'"'•',.::;•, :'•:''' .....::": ' t •'"'"• "*-• ..... •''""" .....• : '- '......... •
