Consistent kinematic architecture in the damage zones of intraplate strike-slip fault systems in North Victoria Land, Antarctica and implications for fault zone evolution

 Sheet Music

 3 views
of 14
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Description
Consistent kinematic architecture in the damage zones of intraplate strike-slip fault systems in North Victoria Land, Antarctica and implications for fault zone evolution
Share
Tags
Transcript
  Consistent kinematic architecture in the damage zones of intraplatestrike-slip fault systems in North Victoria Land, Antarctica andimplications for fault zone evolution Fabrizio Storti  a, * , Federico Rossetti  a , Andreas L. La¨ufer  b , Francesco Salvini  a a  Dipartimento di Scienze Geologiche, Universita` “Roma Tre”, Largo S. L. Murialdo 1, I-00146 Rome, Italy b  Bundesanstalt fu¨ r Geowissenschaften und Rohstoffe, Stilleweg 2, 30655 Hannover, Germany Received 27 January 2005; received in revised form 8 August 2005; accepted 16 September 2005Available online 28 October 2005 Abstract Cumulative, polymodal normal (Gaussian distribution) statistics was applied to subsidiary fault data collected from damage zonesassociated with the Cenozoic Lanterman and Priestley intraplate right-lateral strike-slip fault systems in North Victoria Land, Antarctica.Results show that five Gaussian peaks out of seven in the Lanterman Fault and five out of nine in the Priestley Fault have almost coincidentazimuthal values. We named these Gaussian peak pairs as  consistent fault sets , arranged in a  consistent kinematic architecture  that iscompatible with the Cenozoic regional strike-slip environment. Angular and kinematic relationships among subsidiary fault sets within theconsistent kinematic architecture provide constraints for the inference of the state of stress along the Lanterman and Priestley fault systems.We interpret the fault pattern of the consistent kinematic architecture to be produced by early localisation of the principal displacement zonealong pre-existing mechanical discontinuities inherited from the Early Paleozoic Ross Orogeny. Shear localisation was followed bysubsidiary faulting at an angle to the principal displacement zone according to the Mohr–Coulomb–Byerlee failure criterion. q 2005 Elsevier Ltd. All rights reserved. Keywords:  Subsidiary fault; Consistent fault set; Consistent kinematic architecture; Cumulative statistics; Paleostress; Antarctica 1. Introduction Intraplate strike-slip fault systems can accommodate tensto hundreds of kilometres of horizontal displacementbetween adjacent lithospheric blocks (Woodcock andSchubert, 1994; Storti et al., 2003a). They typically includealmost straight belts alternated with bends and offsets inindividual segments, each of them consisting of complexarrays of anastomosing fault strands (Deng et al., 1986;Vauchez et al., 1995; Ludman, 1998; Holdsworth andPinheiro, 2000; Li et al., 2001; Faulkner et al., 2003). Thekinematics of the fault strands depends on their orientationwith respect to the total stress field along the principaldisplacement zone (PDZ in Tchalenko (1970)). The totalstress field results from the interplay between the stress fieldproduced by fault motion (i.e. the kinematically-inducedstress) and the regional stress field (e.g. Davis, 1984; Mandl,2000). Strike-slip, transpressional and transtensional faultsegments typically coexist along intraplate strike-slip faultsystems (Woodcock and Schubert, 1994). This may occur atdifferent scales and may result in the development of verycomplex fault patterns (in space, time and kinematics;Sylvester, 1988).In many cases, large-offset faults are not preservedwithin strike-slip fault systems at the continental scale, dueto the progression of the tectonic activity and to preferentialerosion and sedimentation in highly fractured fault zones(Price and Carmichael, 1986; Umhoefer, 2000). Never-theless, small-scale subsidiary faults are widespread in thedamage zones of major fault strands (Keller et al., 1995;McGrath and Davison, 1995; Little, 1996; Kim et al., 2004).Detailed fieldwork on the three-dimensional architecture of subsidiary faults along intraplate fault systems providesimportant information on the modalities offault propagationand displacement accommodation (Deng and Zhang, 1984;Pachell and Evans, 2002), the role offault re-activation (e.g. Journal of Structural Geology 28 (2006) 50–63www.elsevier.com/locate/jsg0191-8141/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.doi:10.1016/j.jsg.2005.09.004 *  Corresponding author. E-mail address:  storti@uniroma3.it (F. Storti).  Fig. 1. TectonicsketchmapofNorthVictoriaLandandthewesternRossSeashowingtheCenozoictectonicarchitecturedominatedbyNW–SEstrikingintraplateright-lateralstrike-slipfaultsystems(afterSalvinietal.,1997).Thetwofaultsystemsegmentsanalysedinthisstudyareindicatedbythebrokenrectangles.Contours(Schmidt net,lower hemisphere) of polesto faultsandslickenlinesare providedfor each faultsegment. Contours (Schmidt net,lower hemisphere) of polestothe mainregionalfoliation occurring in the Early Paleozoic metamorphic rock relicts exposed along the Lanterman and Priestley fault systems are also shown. Contouring interval 4%. F. Storti et al. / Journal of Structural Geology 28 (2006) 50–63  51  Holdsworth et al., 2001), and on stress conditions alongspecificfaultstrands(Rispoli,1981;AydinandSchultz,1990).However, the possible variability of dynamically- andkinematically-induced stress conditions along fault segmentsand their along- and across-strike geometric irregularity,imply that the generalization of the state of stress obtained atthe local scale may lead to biased inferences whenextrapolated at the scale of the entire strike-slip fault system.Cumulativestatistics ofsubsidiaryfault data indamagezonesmay provide a useful tool for removing the stress, geometric,and structural variability at the local scale along intraplatestrike-slip fault systems, thus contributing to constrain thestress regime under which they developed.In this paper, we apply cumulative, polymodal normal(Gaussian distribution) statistics to subsidiary fault datacollected along the damage zones ofthe CenozoicLanterman(Rossetti et al., 2002) and Priestley (Storti et al., 2001) right- lateral strike-slip fault systems in North Victoria Land,Antarctica (Fig. 1). These continentally sized strike-slip faultsystems have similar overall orientations, affect similar rock types, and developed in the same geodynamic scenario. Ourresults encourage the use of cumulative statistical analysis of large data sets as an additional tool for the completecharacterisation of intraplate strike-slip fault systems. 2. Geological framework of the Lanterman and Priestleyfaults The tectonic architecture of North Victoria Land hasbeen classically interpreted by the assembly of three majorterranes (Fig. 1). These are from west to east the Wilson,Bowers and Robertson Bay terranes (Kleinschmidt andTessensohn, 1987). The Wilson Terrane generally consistsof granitic rocks that are Cambrian to Ordovician in age(Granite Harbour Intrusive Complex; Gunn and Warren,1962; Rocchi et al., 1998), intercalated with low- to high-grade metamorphic rocks (Gair, 1967; Skinner and Richer,1968; GANOVEX Team, 1987; Lombardo et al., 1989). TheBowers Terrane generally comprises very low- to low-grademostly clastic and volcanic Cambrian rocks (Weaver et al.,1984; GANOVEX Team, 1987). Very low- to low-gradeCambrian to Ordovician terrigenous metasedimentarysequences are the dominant rock type exposed in theRobertson Bay Terrane (GANOVEX Team, 1987). Theassembly of these terranes occurred during the EarlyPaleozoic Ross-Delamerian Orogeny (Kleinschmidt andTessensohn, 1987; Flo¨ttmann et al., 1993). Recent revisitingof the Paleozoic orogenic evolution of Victoria Landquestioned the classical terrane model, emphasising asubduction–accretion setting (Finn et al., 1999; Rolandet al., 2004).The Mesozoic break-up of the Antarctic, Australian andPacific plates resulted in the opening of the Southern Oceanand the Ross Sea, at the northern and eastern boundaries of Victoria Land, respectively (Lawver and Gahagan, 1994;Sutherland, 1995; Mukasa and Dalzier, 2000). In Eocenetimes (from 50 to 40 Ma), activation of intraplate right-lateral strike-slip tectonics occurred in North Victoria Landand overprinted the still ongoing extension in the Ross Sea(Salvini et al., 1997; Salvini and Storti, 1999; Rossetti et al.,2003). The interplay between right-lateral strike-slipfaulting in North Victoria Land and extensional faulting inthe Ross Sea resulted in partitioned transtension along thewestern shoulder of the basin (Wilson, 1995; Salvini et al.,1997; Hamilton et al., 2001; Rossetti et al., 2000, 2003). Atthis time, emplacement of extrusive and intrusive rocksbelonging to the McMurdo Volcanic Group (Kyle and Cole,1974; Rocchi et al., 2002) also started along the westernshoulder of the Ross Sea.The Lanterman Fault and Priestley Fault are part of theCenozoic array of NW–SE striking right-lateral strike-slipfault systems that cut through the continental crust of NorthVictoria Land and continue into the western Ross Sea(Fig. 1) (Salvini et al., 1997). The Lanterman Fault follows the boundary of the Wilson and Bowers terranes andconsists of two major segments: a NNW–SSE strikingsegment in the northern sector and a considerably longer,NW–SE striking segment to the SE (Fig. 1). The faultsystem records a complex deformation history with bothductile and brittle features preserved (Capponi et al., 1999;Rossetti et al., 2002). Late Cenozoic activity in its offshoresegment is indicated by dextral shearing of sedimentsyounger than 32 Ma (Salvini et al., 1997). Apatitethermochronology supports the Late Cenozoic activity of the onshore segments (Rossetti et al., 2003). The onshorelength of the Lanterman Fault is approximately 430 km.However, when the offshore segment located in the RossSea is included, the total length of this intraplate faultsystem exceeds 750 km.The Priestley Fault cuts across the Wilson Terrane, re-activating ductile shear zones that developed during theRoss Orogeny (Storti et al., 2001). Its Cenozoic activity issupported by pseudotachylyte-bearing fault splays exposedat its southern tip, which were dated at ca. 34 Ma (DiVincenzo et al., 2004). This Cenozoic activity is alsosupported by the syntectonic injection of McMurdo dykeswithin the fault zone (Storti et al., 2001). The Priestley Faulthas an onshore length of approximately 220 km from theRoss Sea coast to the polar plateau, where the ice capprevents its further tracing. The Priestley Fault terminatesoffshore in the Terror Rift transtensional basin, along thewestern shoulder of the Ross Sea (Storti et al., 2001)(Fig. 1).Displacement computation is difficult for both faultsystems, since the Priestley Fault occurs entirely withingranitic rocks, and no significant markers can be correlatedacross the Lanterman Fault. A cumulative displacement of approximately 270 km has been proposed for the array of Cenozoic right-lateral strike-slip fault systems (Fig. 1)based on restoration of the continental shelf boundary in theSouthern Ocean (Storti et al., 2003b). F. Storti et al. / Journal of Structural Geology 28 (2006) 50–63 52  3. Cumulative data analysis The study areas include the NW–SE striking segment of the Lanterman Fault and the exposed onshore segment of the Priestley Fault (Fig. 1). Both fault systems crosscut theCambrian–Ordovician Granite Harbour Intrusive Complex.These rocks are well exposed and are relatively homo-geneous in texture. In a few cases, fault data were alsocollected in scattered metamorphic rock remnantsembedded within the granitic bodies. The structuralarchitecture and kinematics of the Lanterman Fault andPriestley Fault are described in Rossetti et al. (2002) andStorti et al. (2001), respectively. For the purposes of thiswork, the same datasets were statistically analysed.Subsidiary faults in the damage zones of both faultsystems have an overall NW–SE strike and are predomi-nantly steeply dipping (Fig. 1). This attitude is similar tothat of the regional foliation recognized in Early Paleozoicmetamorphic rocks exposed along both the Priestley andLanterman fault systems (Fig. 1). Subsidiary faults mostlyinclude low displacement (some centimetres) single shearsurfaces and larger displacement (ten of metres) compositefault zones forming decimetre- to metre-thick cataclasticcores. Kinematic indicators include lunate fractures (Petit,1987) and quartz slickenfibres (Fig. 2). Slickenlines along the Lanterman Fault are well clustered into major strike-slipand dip-slip populations. Minor oblique-slip populationswere also observed. Slickenlines along the Priestley Faulthave a similar distribution (Fig. 1).No systematic overprinting relationship was recognisedin the field among the different subsidiary fault sets. Theevidence that both the attitude and kinematics of thedifferent subsidiary fault sets are compatible with theregional Cenozoic dextral shearing (Storti et al., 2001;Rossetti et al., 2002), and radiometric (Di Vincenzo et al.,2004) and apatite fission track dating (Rossetti et al., 2003), support the overall contemporaneous activity of the wholefault pattern in post Eocene times (Rossetti et al., 2005). 3.1. Polymodal Gaussian distribution statistics The normal or Gaussian distribution is by far the mostused probability distribution in statistical analysis and isbased on the reasonable assumption that the data have anormal distribution around their mean value (Swan andSandilands, 1995). Applications of Gaussian distributionstatistics to structural geology include lineament swarmanalysis (Wise et al., 1985) and cleavage populationanalysis (Salvini et al., 1999; Tavani et al., 2004). In thiswork we limited Gaussian distribution statistics to faultazimuth data because the PDZ in both the Lanterman andPriestley fault systems is near vertical (Salvini et al., 1997),and because most subsidiary faults in their damage zoneshave a very steep attitude. The average dip of subsidiaryfaults in the damage zones of the Lanterman Fault is in fact77.4 G 14.0 8 . In the Priestley Fault it is 74.4 G 14.7 8 , and themean dip value obtained from merging the two datasets is77.3 G 9.3 8  (Fig. 3a).In our statistical analysis, all fault data collected along agiven fault system were merged into a single dataset from Fig. 2. Kinematic indicators used for shear sense determination of subsidiary faults in granitic rocks. (a) Map view of synthetic shearsemanating from the through-going slip surface and forming an acute anglewith the fault itself, opening in the same direction as the slip sense of theopposite side (to the right). The resulting slip sense of the fault in thisexample is left-lateral. (b) Cross-sectional view of a fault surface showingthe occurrence of lunate fractures produced by the intersection of thesynthetic shears emanating from the through-going slip surface (Petit,1987). A lunate fracture opens in the same direction as the slip sense of theopposite side (to the left) and the resulting slip sense of the fault in thisexample is right-lateral. (c) Quartz shear fibres on a fault surface. Thegrowing direction is the same as the slip sense of the opposite side (to theleft) and the resulting slip sense of the fault in this example is right-lateral. F. Storti et al. / Journal of Structural Geology 28 (2006) 50–63  53  which data belonging to the four major kinematic types(extensional, reverse, right-lateral strike-slip, and left-lateral strike-slip) were extracted. A frequency analysisshowed that right-lateral strike-slip and reverse faults are thedominant kinematic types in both cumulative datasets (i.e.more than 50 fault data in each set) (Fig. 3b). Accordingly,only the strike of right-lateral strike-slip and reversefaults were initially analysed using azimuth-frequencyhistograms. Original data histograms were smoothed toreduce noise component of the raw data (Fig. 3a). Thesmoothing procedure consists of a selected number of moving weighted averages (e.g. Wise et al., 1985). Thesmoothed value  Y  0 i  at each  i  interval was computed by(Salvini et al., 1999): Y  0 i Z P i C k   =  2  j Z i K k   =  2 Y   j k  (1)where  k   is the width of the smoothing interval. Thesmoothed histogram peaks were automatically fitted withGaussian curves using the approximation of the polymodalprobability function in Fraser and Suzuki (1966) andimplemented as described in Wise et al. (1985):  f  ð  x Þ Z X  N i Z 1 h i  exp K 4 ln 2  x K m i D i    2 (2)In this function,  x  is the variable (fault strike in thisstudy),  h  is the peak height,  m  is the mean value of   x ,  N   is thenumber of distributions and  D i  is the curve width measuredat its half height, which relates to the standard deviation of the  i  Gaussian  s i  according to the following equation (seealso Salvini et al., 1999): D i y s i  ffiffiffiffi 2 i p   (3)In order to reduce the ambiguity between adjacent,independent Gaussian curves and to limit the range of thisfunctionwithinthe0–360 8 interval,eachGaussiancurvewasconsidered only in its G 1.5  s  interval (Wise et al., 1985).Polymodal Gaussian fits were used to identify differentfault azimuthal sets within the two analysed kinematictypes. The fitting procedure increases the number of Gaussian curves until the residual is below a given thresholdvalue (10%) of the maximum height of the curves. To have areproducible criterion for selecting the proper number of Gaussian peaks in each analysis, we varied the parameter  D i by trial-and-error until the azimuthal difference betweenadjacent peaks was greater than 15 8 . This angular differencewas derived from the prediction of possible fault types andtheir expected orientation along a strike-slip fault zone bypartitioning the stress field induced by the fault motion intofault-parallel simple shear, fault-parallel compression, andfault-parallel extension (Swanson, 1988) (Fig. 4). Gaussian peaks in the same or different datasets may have verysimilar azimuthal values thus raising the problem of faultparallelism. We considered Gaussian peaks as azimuthallyequivalent when the difference between their azimuthalmean values was lower than the corresponding standarddeviations. 3.2. Right-lateral strike-slip faults Application of the angular limiting criterion describedabove to the right-lateral strike-slip subsidiary faults of the Fig. 3. (a) Cumulative, polymodal Gaussian distribution statistics of faultdip values obtained from the merged datasets of the Lanterman andPriestley right-lateral strike-slip fault systems. The histogram of thesrcinal data distribution, the corresponding smoothed histogram, and theGaussian best-fit curves are shown. In the table, % Z percent of data fittedby each Gaussian curve; Nor  H  . Z normalised height of the Gaussian peak withrespectto thehighestpeak; Max  H  . Z maximumheightofthe Gaussianpeak expressed as smoothed number of data by 10 8  of interval; s.d. Z standard deviation. It is worth noting that the sum of the two % valuesexceeds 100% because of the overlapping tails of the Gaussian curves. (b)Frequency diagrams of fault kinematic types in the Lanterman Fault andPriestley Fault. In both cases, right-lateral strike-slip and reverse faults areby far more abundant than left-lateral strike-slip and extensional ones. F. Storti et al. / Journal of Structural Geology 28 (2006) 50–63 54
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks