Amos Ostler Fault summary Tectonics 2010

33 Pages·2010·11.86 MB·English

TECTONICS,VOL.29, TC4021,doi:10.1029/2009TC002613,2010
‐
Along strike growth of the Ostler fault, New Zealand:
Consequences for drainage deflection above active thrusts
Colin B. Amos,1,2 Douglas W. Burbank,1,3 and Stuart A. L. Read4
Received14September 2009; revised 2 March2010; accepted 9March 2010;published 6 August 2010.
[1] Rarely are geologic records available to constrain lateral propagation. Oscillations of climate at ∼104‐yr
thespatialandtemporalevolutionofthrust‐faultgrowth timescalesmodulatetheformationandincisionofgeo-
asslipaccumulatesduring repeated earthquakeevents. morphicsurfacesduringsuccessiveglacialstages.Super-
Here, we utilize multiple generations of dated and de- imposed on apparently steadier rates of fault slip, such
formed fluvial terraces to explore two key aspects of climate‐dependent surfaces contribute to a pattern of
the along‐strike kinematic development of the Ostler progressive drainage deflection along the central Ostler
fault zone in southern New Zealand over the past fault zone that is largely independent of fault propaga-
∼100 k.y.: accumulation of fault slip through space tion. Citation: Amos,C.B.,D.W.Burbank,andS.A.L.Read
and time and fixed‐length thrust growth that results (2010), Along‐strike growth of the Ostler fault, New Zealand:
in patterns of drainage diversion suggestive of later- Consequences for drainage deflection above active thrusts,
ally propagating faults. Along the Ostler fault, sur- Tectonics,29,TC4021,doi:10.1029/2009TC002613.
face deformation patterns revealed by topographic
surveying of terrace profiles in nine transverse drai- 1. Introduction
nages define systematic variations in fault geometry
[2] A quantitative understanding of thrust fault growth
and suggest deformation over both listric and planar
remains integral to our knowledge of how active compres-
thrust ramps. Kinematic modeling of folded terrace
sional mountain belts evolve during ongoing deformation
profilesand>100fault‐scarpsurveysalongmajorfault and crustal shortening. Actual kinematic constraints on the
sections reveals remarkably similar slip distributions three‐dimensional, spatiotemporal development of active
for multiple successions of geomorphic surfaces span- thrust systems, however, are surprisingly scarce. Both geo-
ning ∼100 k.y. Spatially abrupt and temporally sus- morphic surfaces and syntectonic sediments have emerged
tained displacement gradients across zones of fault as useful archives of strain by linking the kinematic pre-
section overlap suggest that either persistent barriers dictionsofthrustfault‐relatedfolding(i.e.,fault‐bend,fault‐
propagation, and trishear models) [Suppe, 1983; Suppe and
to fault propagation or interference between overlap-
Medwedeff,1990;Erslev,1991]toactivedeformationofthe
ping faults dominate the interactions of fault tips from
ground surface [e.g., Lavé and Avouac, 2000; Thompson
the scale of individual scarps to the entire fault zone.
et al., 2002; Hardy and Poblet, 2005; Dolan and Avouac,
Deformedterrace surfacesdated usingoptically stimu-
2007, and references therein]. Despite their role in provid-
latedluminescenceandcosmogenicradionuclidesindi-
ing constraints on the style, rate, and magnitude of active
catesteady,maximumratesoffaultslipof∼1.9mm/yr thrusting from topographic and seismic data, such studies
during the Late Quaternary. Slip data synthesized typically present atwo‐dimensional, transport‐parallel view
along the central Ostler fault zone imply that displace- into the growth and architecture of active thrust systems.
ment accumulated at approximately constant fault The tendency for thrusts to lengthen with increasing dis-
lengths over the past ∼100 k.y. A northward temporal placement [e.g., Elliott, 1976; Walsh and Watterson, 1988;
progressionofabandonedwindgapsalongthissection Cowie and Scholz, 1992], however, underscores the neces-
sity of both transport‐parallel and ‐perpendicular perspec-
thusreflectslateraltiltinginresponsetoamplificationof
tives in providing a comprehensive view of thrust fault
displacement, rather than simple fault lengthening or
evolution [Cooperet al., 2003;Bernal etal., 2004].
[3] By exploiting linkages between thrust faulting and
folding of the Earth’s crust [e.g., Boyer and Elliot, 1982],
1DepartmentofEarthScience,UniversityofCalifornia,SantaBarbara, numerous studies have utilized along‐strike analysis of
California,USA. active fault‐related folds as a proxy for delineating thrust
2NowattheDepartmentofEarthandPlanetaryScience,Universityof
growth in three dimensions [e.g., Jackson et al., 1996;
California,Berkeley,California,USA.
3InstituteforCrustalStudies,UniversityofCalifornia,SantaBarbara, Mueller and Talling, 1997; Delcaillau et al., 1998; Bull,
California,USA. 2009]. Such studies typically rely on combined structural
4InstituteofGeologicalandNuclearSciencesLtd.,LowerHutt,New andgeomorphicanalysesofemergentfoldsandchronologic
Zealand. constraints on deformed geomorphic surfaces [Keller et al.,
1998; Jackson et al., 2002; Hetzel et al., 2004; Bennett
Copyright2010bytheAmericanGeophysicalUnion.
0278‐7407/10/2009TC002613 et al., 2005, 2006], magnetostratigraphy of growth strata
TC4021 1 of33
TC4021 AMOS ET AL.:OSTLER FAULTGROWTH TC4021
[Chen et al., 2002, 2007], or ergodic, space‐for‐time sub- 2. Study Area
stitutions [Burbank et al., 1999; Keller et al., 1999] to
2.1. Regional Setting and Background
document growth histories along the underlying active
thrusts. Taken together, these studies suggest that thrust [6] The Ostler fault represents a zone of east‐directed
faultgrowthisoftennonlinear,andthatratesoflateralthrust thrusting within the intermontane Mackenzie Basin in the
propagation can be rapid (10 s of mm/yr) and variable with Southern Alps of New Zealand (Figure 1). The fault con-
respecttothetotalfaultdisplacement[Bennettetal.,2006]. stitutes one of several structures east of the plate‐bounding
Such data fill a crucial gap in our understanding of the Alpinefaultthataccommodateapproximatelyone‐quarterto
evolution of fault length and displacement relationships, one‐third of the ∼37 mm/yr of relative motion between the
given that kinematic constraints on competing models for obliquelycollidingPacificandAustralianplates[DeMetset
fault growth are generally absent [Walsh et al., 2002]. Slip al., 1994; Pearson et al., 1995; Tippett and Hovius, 2000;
histories constructed from both active and inactive fault Norris and Cooper, 2001]. Resolved onto the Alpine fault,
systems, however, are generally ambiguous as to whether this motion corresponds to velocities parallel and per-
they reflect lateral fault growth or sustained displacement pendicular to the plate boundary of ∼35 and ∼10 mm/yr,
gradientsonthecontrolling thrust [Kelleretal., 1999].The respectively[NorrisandCooper,2001].Theroughlynorth–
actual kinematics of displacement transfer and segment northeast‐striking Ostler fault is oriented oblique to the
linkageforactivethrustsarealsoconfoundedbyinsufficient direction of plate convergence (Figure 1) in a structural
temporalcontrol[Davisetal.,2005],againhighlightingthe transition zone that interrupts the predominately plate‐
need for quantitative measures of along‐strike thrust fault boundary‐parallel orientation of folds and thrusts in the
growth. centralSouthIsland[Uptonetal.,2009].Thiszoneconsists
[4] Here, we define and analyze patterns of thrust devel- of several north‐ to northwest‐striking reverse faults and
opment for the Ostler fault zone, a well‐expressed Quater- isolated sinistral strike‐slip faults localized along an inher-
nary thrust fault in southern New Zealand (Figure 1). By ited, pre‐collisional transition in crustal thickness and rheo-
integrating kinematic analysis of displaced geomorphic logical properties between Otago to the southwest and
surfaces from nine transverse drainages along the ∼60‐km Canterbury tothenortheast [Uptonetal.,2009].
surface trace, we delineate spatial patterns of deformation [7] The crustal composition and structure of the Mack-
parallelandperpendiculartothefaulttransportdirection.As enzieBasinprimarilyreflectstheassemblageandsubsequent
tools, we combine the results of detailed geomorphic map- tectonic and erosional modification of the Permo‐Triassic
ping with high‐resolution topographic surveying of both Torlesse terrane. Ranges surrounding the Mackenzie Basin
deformed geomorphic surfaces and active fault scarps, and comprise deformed Torlesse greywacke, foliated meta-
with geochronologic constraints on terrace formation. Our greywacke, and its low‐grade schist equivalent [Sporli and
synthesis indicates that, over the past ∼70–100 k.y., growth Lillie, 1974]. The Torlesse terrane originated within an
of the Ostler fault zone occurred through gradual amplifi- accretionary prism bounding the convergent margin of
cation of the displacement profile without resolvable fault Gondwana [MacKinnon, 1983]. Later rifting of this margin
lengthening. Displacement and slip‐rate estimates for mul- in the Cretaceous resulted in pervasive extensional defor-
tiple generations of deformed surfaces along each fault mation of Torlesse rocks and is recorded in a series of Ter-
section reveal a range of profile shapes, with either asym- tiary basins localized along the crustal boundary separating
metric or centrally localized maxima and smoothly varying Otago and Canterbury [Deckart et al., 2002; Upton et al.,
gradientsgenerallysteepeningtowardthefaulttips.Steeper 2009]. Within the Mackenzie Basin, Cretaceous‐Miocene
and persistent displacement‐rate gradients occur at segment sedimentary rocks are nowhere exposed but are interpreted
boundaries where overlapping fault strands suggest signifi- fromseismicimagingtorestunconformablyaboveTorlesse
cant slip transfer or incipient linkage. basement[Longetal.,2003].Basedontheinferredpresence
[5] Maintenance of these displacement‐rate gradients ofthesedepositsatdepth,Ghisettietal.[2007]suggestthat
through time also contributes to the general northward pat- the Mackenzie Basin may represent an inverted Tertiary
tern of drainage deflection and the younging of paleo‐out- extensional basin, controlled by the Ostler fault zone as
wash valleys preserved along the Ostler fault zone [Amos reactivated normal fault.
and Burbank, 2007]. In contrast to conventional inter- [8] Upper portions of the Mackenzie Basin stratigraphy
pretations of progressive wind‐gap emplacement as consist of Plio‐Pleistocene molasse of the Kurow Group
reflecting lateral fault propagation [Jackson et al., 1996; [Mildenhall, 2001], recording earlier periods of uplift and
Delcaillau et al., 1998; Keller et al., 1998; Burbank et al., unroofing of the Southern Alps [Chamberlain et al., 1999].
1999; Keller et al., 1999], drainage diversion along the Kurow Group sediments are primarily exposed in uplifted
Ostler fault instead reflects lateral tilting in response to and backtilted dip panels above active thrusts in the
amplification of the displacement profile in the absence of Mackenzie Basin [Gair, 1967]. The modern geomorphic
significantfaultlengthening.Thisresulthighlightspotential surfaceoftheMackenzieBasinisdominatedbyPleistocene
discrepancies between rates of drainage diversion and outwash and fluvial gravels, and estimates of basin depth
lengthwise fault growth and emphasizes the importance of limit the total thickness of Cenozoic deposits to less than
geochronologic control on the creation of wind gaps in 2 km [Long et al., 2003].
effortstoreconstructthegrowthhistoryofactivethrustsand [9] Theactive surface trace oftheOstler faultzonespans
related folds. ∼60 km, from the Ahuriri River in the south (Figure 1) to
2 of 33
TC4021 AMOS ET AL.:OSTLER FAULTGROWTH TC4021
Figure1. TheOstlerfaultzone,SouthIsland,NewZealand.Overviewmapshowsthelocationoffault
scarps, major fault section boundaries, and moraine complexes in the Ohau and Pukaki Valleys. Paleo‐
outwashvalleysatWillowbankSaddle(WS)andClearburn(CB)recordprogressivenorthwarddrainage
deflection toward the modern position of the Ohau River (OR). Inset depicts active fault traces in the
MackenzieBasin(MB),Otago(OT),andCanterbury(CB),aswellasthetectonicsettingoftheobliquely
colliding Australian (AUS) and Pacific (PAC) plates. The location ofthe Ostler faultzone (OFZ) within
the small bounding box and the Great Groove fault (GGF) are also shown for reference.
Whale Stream at the northern end of Lake Pukaki. To the From north to south, theNorthern, Haybarn, North Central,
north, the Ostler fault merges with the Great Groove fault, and South Central sections define an average fault strike of
which extends an additional ∼30 km northward along the 015°. Despite the fault zone’s relatively discontinuous sur-
foot of the Ben Ohau Range [Ward and Sporli, 1979]. face trace and the presence of several km‐scale step‐overs
Complexandhighlysegmentedsurfacerupturesdefinefour (Figure 1), individual sections along the Ostler fault are
main fault sections along the Ostler fault zone (Figure 1). interpreted as kinematically linked at depth [Davis et al.,
3 of 33
TC4021 AMOS ET AL.:OSTLER FAULTGROWTH TC4021
2005].Offsetgeomorphicsurfacespreservedacrossthefault 2.2. Geomorphology of the Ostler Fault Zone
show primarily west‐side‐up displacements without a sig-
[13] Several generations of Late Pleistocene glacial mor-
nificantcomponentofobliqueslip[Read,1984;Davisetal., aines, outwash surfaces, and fluvial terraces dominate the
2005]. Along the Great Groove fault, however, displace-
surficial geology of the Mackenzie Basin [Read, 1984].
ment of Kurow Group sediments includes both reverse and
Terrace and outwash surfaces are graded to moraine com-
left‐lateral strike slip motion [Templeton et al., 1999]. plexes reflecting intermittent glacial occupation ofthe Lake
[10] Fault planes imaged using ground‐penetrating radar OhauandLakePukakivalleys[Blicketal.,1989;Readand
the along the North Central section immediately west of
Blick, 1991]. Quaternary landforms and deposits in the
Twizel (Figure 1) have average dips of 56 ± 9° (2s) in the Mackenzie Basin were originally categorized through rela-
shallowsubsurface[Amosetal.,2007].Thesedipsgenerally
tive dating and field relationships into four major age divi-
agree with estimates from offsets of multiple terrace levels
sions[Gair,1967].Webuilduponthisclassificationscheme
(50 ± 18° (2s)) [Davis et al., 2005] and deeper seismic
and subdivide geomorphic surfaces into Wolds, Balmoral,
reflection and refraction data (∼50–60°) [Ghisetti et al.,
Mount John, and Tekapo age associations (T1 to T4), from
2007; Campbell et al., 2010]. Ground‐penetrating radar
oldesttoyoungest(Figure2).Eachageassociationcontains
studiesfarthersouthalongtheNorthCentralsectionreveala multiple terrace levels, noted using apostrophes, e.g., T1′.
range in fault dips between ∼30° [McClymont et al., 2008] Dating of terrace surfaces spanning the length of the Ostler
and ∼50° [Wallace et al., 2010], indicating some degree of
fault (discussed below) confirms synchronous times of ter-
dipvariabilityintheshallowsubsurface.Foldedgeomorphic
race formation between individual drainages.
surfaces in the hanging wall of the Ostler fault zone com- [14] Flights of terraces preserved across the Ostler fault
monly define highly asymmetric anticlines with broad,
represent strath terraces eroded into bedrock, aggradational
west‐tiltedbacklimbsandhalf‐wavelengthsofuptoseveral
Quaternary outwash surfaces, or isolated cut terraces
km [Davis et al., 2005]. In an earlier study [Amos et al., beveled into Pleistocene outwash. Exposures into strath
2007], we interpreted this pattern of surface deformation
terrace surfaces reveal up to several meters of fluvial sand
as representing slip along listric, or curviplanar thrusts
and gravel resting above the bedrock strath. In contrast,
rootedatshallowdepthsintoplanarfaultramps,potentially terrace surfaces with thicker gravel cover (i.e., ≥∼3 m) are
controlled by the mechanical stratigraphy and depth of classified here as outwash terraces, which are aggradational
Cenozoic basin fill.
features. Basal exposures of outwash surfaces are uncom-
[11] Estimates of the total offset on the Ostler fault stem mon, although where visible, they indicate on the order of
from seismic imaging along the North Central fault section several decimeters or less of overlying gravel cover. Intri-
[Ghisetti et al., 2007; Campbell et al., 2010]. Several Plio‐
cate texturing of strath and outwash terrace surfaces with
Pleistocene sedimentary sequences that are identified and
abundant paleochannels attests to their relatively pristine
correlated across the fault by Ghisetti et al. [2007] yield a and uneroded nature, despite a locally irregular loess and
total fault throw of ∼800 m and a corresponding shortening
soil cover typically less than a meter thick [Maizels, 1989].
of ∼30%. Constraints on the timing of fault onset are less
[15] Geomorphic surfaces on the Ostler fault also include
clear, although Davis et al. [2005] suggest a mid‐Pleisto- threebroad,originallyflat‐bottomedoutwashvalleyseroded
cene origin for the Ostler fault based on extrapolation of through Plio‐Pleistocene bedrock across the North Central
fault length‐displacement scaling relationships. Exposed
section[AmosandBurbank,2007](Figure1).Thesouthern
growthstrataintheOstlerfaulthangingwall(seediscussion
twooutwashvalleys(WillowbankandClearburn)nolonger
belowandalsoAmosetal.[2007]andGhisettietal.[2007]) contain major, active drainages and are preserved as wind
also demonstrate that deformation was clearly ongoing
gapsupliftedabovethebasinfloor. Westward tilting dueto
during deposition of Plio‐Pleistocene molasse.
foldingovertheOstlerfaulthaslocallyreversedtheoriginal
[12] Previous work on deformed terraces crossing the eastward slope of the outwash surfaces in each wind gap.
Ostler fault zone suggests slip rates in excess of ∼1 mm/yr
Drainage from the Lake Ohau valley now follows the
[Blicketal.,1989;ReadandBlick,1991;Davisetal.,2005;
northernmostvalleyandcrossestheNorthCentralsectionof
Amos et al., 2007]. This rate corresponds to ∼10% of the the Ostler fault zone immediately south of the town of
distributed Pacific plate motion southeast of the Alpine
Twizel (Figure 1).
fault. Geodetically observed vertical deformation of the
[16] If we assume that deformed strath terraces formed
Ostlerfaulthangingwalloccursatasimilarrate,suggestive
with a long profile geometry similar to modern streams
of either interseismic storage of elastic strain or aseismic [Pazzaglia and Brandon, 2001], widespread and laterally
folding [Blick et al., 1989]. Paleoseismic trenching on the
continuousterracesurfacespreservedacrosstheOstlerfault
northern Ostler fault [van Dissen et al., 1994], however,
zone represent ideal, dateable strain markers for recording
impliesthatlarge,surface‐rupturingearthquakesoccuralong fault growth. This assumption is most likely to be valid
the Ostler fault zone. The timing of these paleoearthquakes
where longitudinal profiles along a strath surface can be
between∼3–4ka,∼6–8ka,and∼10kaindicatesarecurrence
compared to the modern bedrock channel, because straths
intervalof∼2–5kyforthenorthernOstlerfault[vanDissen representformerequilibriumpositionsoftheriverbedprior
etal.,1994]. Away fromtransverse drainages, a ∼1m‐high
to incision [Bull, 1991]. For outwash or fill terraces, po-
scarp cuts the foot of young hillslopes and recent landslide
tential dissimilarities between the initial surface geometry
depositsalongmuchofthefault’slengthandmayrepresenta and modern stream profiles are more likely, given that
single‐eventscarpformedduringthemostrecentearthquake.
bedrock geometry beneath outwash surfaces is commonly
4 of 33
TC4021 AMOS ET AL.:OSTLER FAULTGROWTH TC4021
Figure2. Terracestratigraphy/QuaternarychronologyforsurfacespreservedalongtheOstlerfaultzone.
OSL dating of terrace fills from this study and Amos et al. [2007] bracket correlation of the Balmoral
surfacewithcoldperiodsduringeitherMarineIsotopeStage4or5b.Meancosmogenicexposureagesona
Tekapomoraine match well with latest‐LGM morainesdated by Schaeferetal.[2001, 2006].
unknown, the thickness of outwash gravels may approach strath terrace surfaces, we also surveyed isolated points
tensofmeters,andincreasedsedimentloadduringoutwash along the strath‐gravel contact in addition to points along
events probably required a steeper transport gradient. This the terrace tread. Additionally, topographic profiles mea-
discrepancy is exemplified by the modern Ohau River, sured from a TOPSAR 10‐m digital elevation model were
where average long profile slopes on the Mount John out- used to supplement our survey coverage. DEM elevations
wash surface (∼0.9°) exceed modern local river gradients were tied to our survey measurements through repeated
(∼0.4°) upstream of the area affected by folding. As such, surveying of known benchmarks throughout the field area.
the actual magnitude of post‐abandonment westward
(upstream)tiltingofthissurfaceisprobablyunderestimated
2.3. Surface Chronology
through comparison with the modern channel long profile.
Assessmentofpost‐abandonmenttiltingofoutwashsurfaces [18] Existing radiocarbon, luminescence, and cosmogenic
thusrequirescomparisontosurfaceprofilesoutsidethezone exposure ages in the Mackenzie Basin [Maizels, 1989;
oftilting and deformation. Schaeferetal.,2001,2006;Amosetal.,2007]lendsupport
[17] We surveyed terrace treads, outwash surfaces, and to correlation of the Tekapo and Mount John age divisions
activechannelsineachtransversedrainage alongtheOstler with the Last Glacial Maximum (LGM), or Otiran Glacial
faultusingaTrimble4700differentialGPSwithcentimeter‐ period in New Zealand [Suggate, 1990]. The association of
level vertical and horizontal precision. This degree of pre- older features with Late Quaternary stratigraphic subdivi-
cisionislessthanthe“geomorphicnoise,”orgroundsurface sionsinNewZealand,however,typicallyreliesontentative
irregularities caused by variations in soil or loess cover, correlations to climatic proxies such as the marine isotopic
generallylessthan∼10cm.Scatterinoursurveydatainstead record [e.g., Martinson et al., 1987] or paleoclimatic data
reflectsubiquitouspaleochannelrillstypicallywith<1mof from ice cores [Petit et al., 1999]. Actual connections be-
relief on surface treads. Where exposed and accessible for tween these records and Quaternary landscape features re-
5 of 33
TC4021 AMOS ET AL.:OSTLER FAULTGROWTH TC4021
Figure 3. (a) Photograph and (b) interpretation of OSL sample locality WLL494 within terrace fill
deposits on a prominent Balmoral outwash terrace. (c) Photo and (d) stratigraphic interpretation of
OSL sample locality WLL495, where Mount John capping gravels cover older Balmoral fill below a
prominent unconformity. Hammer for scale in each photo.
quireadditionalgeochronologiccontrolbeyondtherangeof (Tables 1a and 1b) represent interbedded sands and silts
radiocarbon dating [Sutherland et al., 2007]. We supple- collected from an exposure into the same Balmoral surface
mentexisting ageconstraints ongeomorphic features inthe dated in our previous study where it is cut by a prominent
Mackenzie Basin using optically stimulated luminescence scarpalongtheHaybarnfaultsection(Figures1,3a,and3b).
(OSL) dating of terrace sediments and cosmogenic radio- Although these samples originate from approximately the
nuclide dating (CRN) of glacial moraines and outwash same stratigraphic interval, their calculated OSL ages differ
surfaces to bracket the timing of terrace formation. by ∼70 kyr. This age discrepancy may result from the fact
[19] OSL dating of Quaternary sediments can provide thattheolderofthesetwosamples(WLL493)islikelytobe
accurate constraints on depositional age provided that ma- close to saturation, resulting in large data scatter and high
terial was sufficiently exposed to sunlight prior to burial so errormargins(∼20%)inthecalculatedOSLage[U.Rieser,
thatitsOSLageisresettozero[Duller,1996;Aitken,1998]. personalcommunication,2006].Incompleteexposuretolight
To maximize the likelihood of resetting prior to deposition, priortoburialwillalsobiassamplestoolderages.Giventhat
we sampled terrace silts and fine sands indicative of over- the OSL date for sample WLL493 (143 ± 23 ka, 1s) is
bankdepositioninarelativelyshallowandquiescentfluvial significantly older than the maximum age of two samples
environment. Our results (Tables 1a and 1b) complement from beneath the terrace gravels (∼105 ka) [Amos et al.,
previous luminescence dating of silts beneath Balmoral 2007], we use only the younger sample age of 75.1 ±
outwash gravels just north of the Ohau River where it 8.6 ka (1s, WLL494) as a minimum constraint on the
crosses the North Central section of the Ostler fault zone Balmoral T2 terrace (Figure 2).
(Figure 1) [Amos et al., 2007]. Those dates indicated a [20] Samples WLL495 and WLL496 also represent inter-
maximumageof∼105kaandcorrelationwitheithermarine bedded sand and silt collected from Balmoral outwash
isotopestage(MIS)5bor4.SamplesWLL493andWLL494 gravels(Figures3cand3d).Theseages,69.0±6.8and72.9±
Table1a. Luminescence DatingResults: Water andRadionuclide Contents
U(mg/g) Th(mg/g)
Sample Location Altitude SampleDepth Water U(mg/g) From226Ra, U(mg/g) From208Tl,
Numbera (Latitude/Longitude) (m) (m) Contentdb From234Th 214Pb,214Bic From210Pb 212Pb,228Acc
WLL493 −44.23887°/170.02568° 561 3.4 1.192 2.66±0.26 2.62±0.17 2.70±0.24 10.5±0.13
WLL494 −44.23881°/170.02582° 556 3.15 1.144 3.08±0.30 2.84±0.19 2.94±0.26 10.88±0.15
WLL495 −44.26513°/170.04730° 526 1.3 1.078 2.03±0.25 2.21±0.16 2.45±0.23 9.72±0.13
WLL496 −44.26527°/170.04723° 526 0.85 1.092 2.07±0.26 2.12±0.17 2.36±0.23 9.76±0.13
aSamplepreparationandmeasurementsattheLuminescenceDatingLaboratory,SchoolofEarthSciences,VictoriaUniversity,Wellington,NZ.
bRatiowetsampletodrysampleweight.Errorsassumed50%(d‐1).
cUandTh‐contentiscalculatedfromtheerrorweightedmeanoftheisotopeequivalentcontents.
6 of 33
TC4021 AMOS ET AL.:OSTLER FAULTGROWTH TC4021
Table1b.LuminescenceDatingResults:MeasuredKContent,a‐ValueandEquivalentDose,CosmicDoseRate,TotalDoserate,and
OSLAge
Alpha
Sample Effectiveness EquivalentDose, DoseRatedD/dt OSL‐age(ka)and Age
Numbera K(%) a‐Valueb D (Gy) dD/dt(Gy/ka)c (Gy/ka) 1sUncertaintyd Associatione FieldCode
e c
WLL493 2.15±0.05 0.05±0.02 535.7±76.2 0.1448±0.0072 3.75±0.30 142.7±23.3 Balmoral HBOSL‐05‐02
WLL494 2.03±0.05 0.07±0.03 312±21.5 0.1495±0.0075 4.16±0.38 75.1±8.6 Balmoral HBOSL‐05‐03
WLL495 1.99±0.04 0.117±0.015 314.3±27.6 0.1911±0.0096 4.56±0.20 69.0±6.8 Balmoral EROSL‐05‐01
WLL496 2.02±0.04 0.073±0.009 296.8±10.1 0.2033±0.0102 4.07±0.16 72.9±3.8 Balmoral EROSL‐05‐02
aSamplepreparationandmeasurementsattheLuminescenceDatingLaboratory,SchoolofEarthSciences,VictoriaUniversity,Welllington,NZ.
bItalicsindicateanestimateda–valueduetosaturationofthea–irradiatedsubsample.
cContributionofcosmicradiationtothetotaldoserate,calculatedasproposedbyPrescottandHutton[1994].
dSampleWLL493wasclosetosaturationandthelargeerrormarginsreflectdifficultiesinfittingdatapoints(U.Rieser,personalcommunication,2006).
eAgedesignationbasedonpreviouslydefinedQuaternarychronologies.Seetextfordiscussion.
3.8 ka (1s), respectively, are indistinguishable from sample further supports correlation of Balmoral terraces with MIS
WLL494 at 95% confidence, and they further support a 5borearlystage4(Figure2).Thisperiodalsooverlapswith
minimum Balmoral age of 70–75 ka (Tables 1a and 1b). the timing of widespread glacial advance recognized
Samples WLL495 and WLL496, however, originate from throughout New Zealand, culminating at ∼80 ka [Preusser
an outwash exposure beneath a prominent Mount John out- et al., 2005; Sutherland et al., 2007].
wash terrace (Figures 3c and 3d). Whereas the minimum [22] Geochronologic constraints on both the Wolds and
Balmoral age from the Haybarn scarp to the northwest Tekapo surfaces come from CRN exposure‐age dating of
(WLL494)comesfromaboveabedrockstrathwith<10mof boulders and cobbles preserved on these landforms. Expo-
gravel cover, the presence of older fill beneath Mount John sure‐age dates were calculated by measuring cosmogenic
terrace gravels suggests a significant southward thickening 10Be present in quartz extracted from Torlesse metasedi-
of Balmoral outwash in the Ohau River valley (Figure 4). mentary rocks deposited on end moraines and outwash
Overlap among these three minimum ages also suggests surfaces. For glacial moraines, we sampled the uppermost
synchrony between the deposition of terrace gravels and surface of the largest boulders, typically >1 m in diameter
outwash accumulation during creation of the Balmoral T2 fortheyoungestmoraines,presentneartheterminalmoraine
terrace. crest. Samples collected from outwash surfaces include
[21] Weighted averages of OSL dates presented here and both the largest individual cobbles present and an amalgam
in our previous study [Amos et al., 2007, Table 1b] bracket of surface chips from ∼10 additional cobbles of meta-
formation of the Balmoral strath between 72.4 ± 3.1 and greywacke. The resulting exposure ages (Table 2) were
104.9 ± 6.3 ka (1s), respectively. Comparison with paleo- computed using the CRONUS‐Earth online exposure age
climatic proxy records reveals atmospheric cooling in the calculator, Version 2.2, as described by Balco et al. [2008]
southernhemisphereduringthistime[Petitetal.,1999]and using the scaling scheme or spallation of Lal [1991] and
Figure4. TopographicprofileacrosstheOhauRiverValleyfromtheTOPSAR10‐mDEMshowingthe
generalpatternofoldersurfacessteppingdowninelevationtothemodernchannel.Theprofilelocationis
shownonFigure1.
7 of 33
TC4021 AMOS ET AL.:OSTLER FAULTGROWTH TC4021
Agefsociation WoldsWoldsWoldsWoldsWoldsTekapoTekapoTekapoTekapoWoldsWoldsWoldsWolds fromLal Sant[od2n3]ecl[aE2sx0tp0eo0ros].usiAroelnl.amgeosdeclaelcxuploasteudrefaogresfiavsesubmoeulzdeerrosbaomupldleesr
As on fromaprominentWoldsendmoraineatWillowbanksaddle
10BeModele(ky)Age 84.8±8.644.9±4.668.6±10.885.1±8.858.5±7.314.4±1.912.4±1.715.9±2.218.2±1.776.9±8.3100.0±9.7122.9±12.1109.5±12.2 hemeforspallati rpls(iaOartntomSlgbepTianloCbegvisRleibNrtryleea‐tfpd3wlee0aenc4mets‐ni1otdyn∼igmft4hu5iirnnnocuduattiingivoodhinnduo85af5o)lfthakeegcaxsoehess(iTmabagainotbedgsleess(nuFu2ibgic)sgg.tuae1rAns0etBtsi5aectaloh)mtahsrtecprvtoaohteusteaegistlrhees,
c
gs post‐depositional erosion. The smooth, relatively subdued
10dMeasuredBe6(10atoms/gSiO)2 0.680±0.0320.362±0.0190.552±0.0710.683±0.0360.469±0.0410.144±0.0140.125±0.0130.158±0.0170.180±0.0070.716±0.0430.927±0.0361.133±0.0461.012±0.066 08],usingthescalin gscmsdcuoeiosarovcfrrmeacaorciiuotneaynrerpteeohctrchfiorcoeeslnssiaostesrunigsaaretfngeoadenbfcsteostwuihonaleimdptohepecureossraucnropbasasmtrsniosttcaebutrensieunstnidmaoatitlfneibodngtonithfutfiohlosudfrfsoeimtrvuheseogxe.rhheatAuirimotmnshseieainsotugianrocnenohgodf,ofaltttwinhhhthdeeee
0
2 Wolds glaciation and surface formation.
e/TotalBe−12)(10 53±0.03959±0.01782±0.09876±0.04097±0.02417±0.01971±0.02797±0.02932±0.02093±0.06545±0.06643±0.11967±0.075 alcoetal., beno[du24lm]deoIrnrasiacnmoenpptlrreaessstef,rrovCmeRdNaonreextlhpaetoivsnueorlryethafergrenesshsecacantlidcounrloaotcefkdtyhfeToOreksfaotlpueorr
10B 0.80.30.70.70.20.20.20.20.51.01.72.91.1 u/)[B cfaluusltteartbeTtowpeeMnc∼M12illaanndS1t8reakma a(nTdMoMveMrla‐p05w‐1ellthwroituhginh24s)
d
e
BecCarrier(mg) 0.30170.29780.29860.29900.30130.29830.29230.29480.29940.29700.29810.30090.2988 hington. uf(uFnnicgceutrirtoaenin5otbife)ts,hse(usTegagabeglseetsin2eg)x.hmiAbiniticsmoamaslptrpooosnisgtte‐pdepearpokobsaaittbi1oil5ni.ta2yl±edr2eo.n4ssiioktyna
s
a
w of these boulders and the moraine crest. The calculated
TopographicShieldingFactor 0.999780.999780.999780.999780.999780.997600.997600.997600.996900.998720.998720.998720.99872 p://hess.ess. cyreoopmuonprgtoeesdrittfehoarpnleaatthkeesti‐msLeGiannMgCoTRoedNkapaegoxrpemoesomurreaenintaegsaenadt(1a7od.nj4alyc±ens1tl.i0gLhaktklaye)
htt Pukaki [Schaefer et al., 2006]. Schaefer et al. [2001] also
s 2; reportaCRNexposureagefortheoldestLGMMountJohn
Table2.AnalyticalResultsof10BeGeochronology SamplebSampleLocationAltitudeQuartzThicknesaSample(g)Type(Latitude/Longitude)(m)(cm) ‐‐−OSTCRN3041moraine44.38601°/169.94792°58524.73035‐‐−OSTCRN3042moraine44.38706°/169.94756°58218.62885‐‐−OSTCRN3043moraine44.38857°/169.94798°58527.53895‐‐−OSTCRN3044moraine44.38765°/169.94887°58822.09065‐‐−OSTCRN3045moraine44.38634°/169.94893°57811.88465‐‐−TMMM0501moraine44.10312°/170.05282°83827.34945‐‐−TMMM0502moraine44.10242°/170.05339°84239.44355‐‐−TMMM0503moraine44.10273°/170.05340°83934.51185‐‐−TMMM0504moraine44.10446°/170.05153°82657.03485‐‐−OSTCRN2241clast44.15680°/170.02949°75029.75113‐‐−OSTCRN2242clast44.15755°/170.03074°75037.09463‐‐−OSTCRN2243clast44.15588°/170.03062°75051.89163‐‐−OSTCRN2246clastamalgam44.157°/170.030°75022.62973 aSamplespreparedandmeasuredatthePRIMELab,PurdueUniversity,WestLafayette,IN.b3Exposureagecalculationsemployasampledensityof2.7g/cmforallsamples.cBecarrierconcentrationis1.45mg/g.dPropagateduncertaintiesincludeerrorsintheblank,carrier,andcountingstatistics.‐eModelagesanduncertaintiescalculatedusingtheCHRONUSEarthonlinecalculator(Version[1991]andStone[2000].fAgedesignationbasedonpreviouslydefinedQuaternarychronologies.Seetextfordiscussion. mgJssPitNCaossiTrsotmatlatudnenhhiieoeuapenolfnekouhRro[mpotddeeimcmrhgedpocrasenk2ivsartrreicNasng5oecldiotc.auioapenTri]t1cohnlnetnakhirbbmliTee2nn1esiuedmotsts(eaioismcSexotrh>3eslttmiexgftoonirrVsni.paasosn[1zasupraattdsbkeAnmuioeuuhanca0timoaknaaBbnaahnterstmrylgeo0nstipegifltfltume∼aiohuefaedcwa(loatLbdrltp1eTyfhkncDsrocaetesesroeawsheH0evraegnamarsrok[brepai)rabceythOaeeStasaeafkekipraorklmtiaygcaorecieySwtslsa(waeoasaSPhtbiseiuec1tpracTctnnlteoi2satDafuared1iktlalrodgrCdroorD)rheen.kaee0ttvnry,fndruesunhsfenlaaaRyosw.mereouoed8rerOmtky2rdfrtwoNacirnitiyeoais0oSnhfttoStfgtt±eeruhaoal‐io0ooheotuleS(oLxtya2mcrltutyt7enefdthlenhaant1epot2tand.sr]ihpagienlg‐e2tds1oudn4e.elgmahmfelcee..6.aasna‐heoe,it,ai5ultcTuinmgh1s.tcgsaeroprptn52tniSarOaheik,srine(trkpga,ce0domtiaotkopoera2g±httSkaasl0elneuefgamo,poe)prhasnpDrT1asdntr.ee4aoesalf3geiherdt∼i]CesnLrcsgs.mnef(c.,eTsd0yW1st1stoFtesrRetsGsooaa0Cipr–eaerei.ntacSkrBrnugkNegoMofilr2sveaytgooFeatldvmseepulpfeinrne‐evsd,ntdnieeWerr2mtcthge,f6lesweadhoweracoo(otc2leeraltnFm)emtdowleroir,tor41hrlutseti,lpsubrbwhCg‐),iiegstdosrtsiacn.ormO1eeiurhegtwutsibRhihnaeetshnt,CtpbgeuirwsswighscrnhnNeeTscplgttefsatiTereatiilaseeehbohntireta1lcler5enatanhevcaieeiresi,hrfeMzlbncwkncgnugenatoaises)tdtlldeLtt‐leffrrtfoohha,eyiaro∼tmsrassaaoaasuioueetattttifrtunn7ehhhhgktnzsoa2oestlimmnasldhdgnd7eeeeeeeeesss-r)ftt.
8 of 33
TC4021 AMOS ET AL.:OSTLER FAULTGROWTH TC4021
Figure5. IndividualandcompositeprobabilitydensityfunctionsforCRNexposureagesfor(a)Wolds
moraineatWillowbankSaddle,(b)TekapomoraineatTopMcMillanStream,and(c)aprominentWolds
outwash surface at Dry Stream. Peak ages for the composite curves represent mean ages (vertical white
line)andstandarderrors(graybars)forallagesatTopMcMillanStreamandthethreeoldestagesatDry
Stream.
exposure ages constrain the timing of the Wolds glaciation active surface trace is buried beneath scree slopes immedi-
to lie within the penultimate glacial maximum, or the ately north of Whale Stream. At the southern tip of the
WaimeaglacialperiodinNewZealand[Suggate,1990],and Northern section, the surface trace dies out as subtle anti-
suggest potential correlation with MIS 6 (Figure 2). clinalwarpingofterracesnorthofFraserStream(Figure6a).
[26] The general correspondence among terrace ages [28] StreamsdrainingtheBenOhauRangeareflankedby
bracketedherewithOSLandCRNexposuredating,climate flights of Late Quaternary strath terraces that cross the
proxy records, and documented Late‐Pleistocene glacial Northern section of the Ostler fault zone. Gravels capping
intervals elsewhere in New Zealand [Suggate, 1990; Bull, these surfaces are typically less than 1–3 m in thickness.
1991; Preusser et al., 2005; Sutherland et al., 2007] lends Terrace treads, moraines, fault scarps, and fold axes were
support to the increasingly widespread recognition that mappedusingacombinationof1:40,000and1:10,000scale
climateexertsfundamentalcontroloverfluvialandoutwash aerial photography (Figure 6). Adjacent terrace treads are
terrace genesis [e.g., Bull, 1991; Molnar et al., 1994; Pratt separated by risers and display the general pattern of older
et al., 2002; Formento‐Trigilio et al., 2003; Litchfield and surfacessteppingdowninelevationtoyoungerterracesand
Berryman, 2005; Bookhagen et al., 2006]. Climate reg- themodernriver.Strathexposuresarerelativelycontinuous
ulateslocalvariationsinsedimentsupplyanddischargethat along these surfaces and reveal both Kurow Group and
determine boththetiming andpattern oflateralerosionand Torlesse bedrock (Figure 7). Last‐Glacial age terraces
incision during terrace formation [Hancock and Anderson, (Mount John and Tekapo) predominate toward the northern
2002]. Subsequent deformation of terrace and outwash end of the fault and were identified based on their pristine
surfaces at the basin scale serves to amplify the effects of surface texture and height above the active channel. CRN
regionalincisionthatisultimatelydriventectonicbase‐level exposure dating of a prominent terminal moraine at Top
lowering[NicolandCampbell,2001;Panetal.,2003;Bull, McMillanStream(Figure5b)aidsindifferentiatingbetween
2007]. these two LGM age divisions and in correlating surfaces
within adjacent drainages. Widespread aggradation within
thefootwalloftheNorthernOstlerfaultandfromnumerous
3. Deformed Geomorphic Surfaces on the smaller tributary drainages, however, diminishes the cer-
Ostler Fault tainty of this distinction elsewhere along the Northern fault
section (Figure 6a). Nested within the LGM terraces, mul-
3.1. Northern Section tiplelevelsofrelativelydiscontinuoussurfacesflankeachof
[27] TheNorthern section of theOstlerfault zonebounds themodernstreams.Wegroupedthesesurfacesasundiffer-
thesoutheastern margin of theBen Ohau Range(Figure 1). entiatedHoloceneterracesbecausethescaleofourmapping
Along the southern half of this strand, the fault raises Tor- generally exceeds the size of individual treads in this age
lesse metasediments in its hanging wall over Quaternary range (Figure 6a). Additionally, isolated Balmoral terrace
gravels and glacial deposits on the basin floor (Figure 6a). remnantsoccupythehighestrelativepositionsaboveDryand
The fault consists of numerous overlapping splays and sec- Fraser Stream. Paleochannel rills are generally absent from
tions forming a complex, arcuate surface trace with an av- theseBalmoraltreads,thereforesubduingtheoveralltextural
eragetrendof∼025°.Northernportionsofthisfaultsection appearance of these relatively older surfaces (Figure 6a).
cut into the range near the headwaters of the Twizel River With the exception of active alluvial surfaces flanking the
and form a prominent bedrock scarp with a quasi‐linear modernchannels,theNorthernfaultsectiondisplaceseachof
surfacetrace.Farthernorth,thenorthernOstlerfaultsection thethreeterraceagegroupsalongitslength.
displaces lateral moraines at Boundary Stream before the
9 of 33
TC4021 AMOS ET AL.:OSTLER FAULTGROWTH TC4021
Figure 6. Geomorphic andneotectonic map displayingterrace treads,faultscarps,andfoldaxesalong
the (a) Northern and Haybarn and (b) Haybarn and North Central fault sections of the Ostler fault zone.
10of 33