Search

The geometry of the upper-crustal portion of the Alpine Fault plate boundary on New Zealand’s South Island, is poorly constrained where it cuts through the sedimentary units of the Westland...

Permo-Triassic, Rakaia Terrane, pelitic and psammitic schists at Lake Hāwea, Otago, show progressive metamorphism from TZII, pumpellyite–actinolite facies to TZIV, greenschist facies. The sequence shows evidence of five phases of deformation. D1 and D2 are progressive synmetamorphic phases, forming macroscopic nappe-like folds, interpreted to have formed by internal deformation within the Rakaia–Caples accretionary wedge during terrane collision. Thick S1 quartz veins are the form surface for definition of D2 folding, intensification of which progressively destroys evidence of D1 at higher textural and metamorphic grades. D3–5 are post metamorphic, less pervasive, kink folds (commonly conjugate) and brittle fault zones that predate the intrusion of lamprophyre magma and carbonatitic fluids associated with the Alpine Dike Swarm (c. 25 Ma). They relate to tectonic relaxation and exhumation of the schist in Cretaceous–Paleogene times. Fluid flow has been ubiquitous throughout the deformation, resulti...

We have analysed 11 years of geodetic data from 30 stations distributed over the Otago Fault System in the South Island of New Zealand. Velocities were estimated from time series corrected for coseismic displacements from the 2004 Macquarie Island and 2007 George Sound earthquakes and the coseismic and the short term postseismic deformation associated with the 2009 Dusky Sound earthquake. By dividing the corrected time series in half we were able to demonstrate the existence of a systematic difference between the pre- and post-earthquake velocity fields, associated with a longer term viscoelastic transient related to the 2009 Dusky Sound earthquake. In the northern part of our study area, the geodetic strain rate data are consistent with elastic strain accumulation on the Alpine Fault while in the south and east, the strain rate tensors are consistent with the Otago Fault System. There is a significant change in orientation in the axis of contraction from east to west across the network that corre...

Determining the earthquake segmentation of plate-boundary transform faults remains a scientific challenge because paleoseismic data sets rarely resolve the end points of past ruptures. In this study, we test whether lacustrine paleoseismology can be used to assess rupture end points and the earthquake segmentation of the Alpine fault, one of the longest and fastest-slipping plate-boundary transform faults on Earth. Sediment cores from Lake Ellery record eight episodes of high-intensity shaking (modified Mercalli intensity [MM] IX) from Alpine fault earthquakes as event sequences of a turbidite produced by coseismic subaqueous mass wasting, overlain by deposits representing sediment flux from co- and postseismic landsliding in the fluvial catchment. Age-depth modeling constrains the timing of shaking events at a decadal resolution, facilitating correlation with two previously published lake records to reconstruct the spatial distribution of MM IX shaking along ~150 km of the Alpine fault. When resolved with existing on- and near-fault paleoseismic records, the lake data set demonstrates that independent ruptures of the South Westland and Central segments occurred in A.D. 845–775 and A.D. 739–646, and A.D. 646–592 and A.D. 416–370, respectively. Lakes adjacent to the Alpine fault provide paleoseismic records with sufficient spatial and temporal resolution to define along-strike differences in the pattern of rupture capable of distinguishing rupture termination at a geometric segment boundary. This multilake study suggests that locating the end points of ruptures using lacustrine paleoseismology will be most applicable in midlatitude convergent plate-boundary settings where along-strike topography and Quaternary glaciation have resulted in the widespread distribution of suitable lakes.

The first phase of the Deep Fault Drilling Project (DFDP-1) yielded a continuous lithological transect through fault rock surrounding the Alpine fault (South Island, New Zealand). This allowed micrometer- to decimeter-scale variations in fault rock lithology and structure to be delineated on either side of two principal slip zones intersected by DFDP-1A and DFDP-1B. Here, we provide a comprehensive analysis of fault rock lithologies within 70 m of the Alpine fault based on analysis of hand specimens and detailed petrographic and petrologic analysis. The sequence of fault rock lithologies is consistent with that inferred previously from outcrop observations, but the continuous section afforded by DFDP-1 permits new insight into the spatial and genetic relationships between different lithologies and structures. We identify principal slip zone gouge, and cataclasite-series rocks, formed by multiple increments of shear deformation at up to coseismic slip rates. A 20?30-m-thick package of these rocks (including the principal slip zone) forms the fault core, which has accommodated most of the brittle shear displacement. This deformation has overprinted ultramylonites deformed mostly by grain-size-insensitive dislocation creep. Outside the fault core, ultramylonites contain low-displacement brittle fractures that are part of the fault damage zone. Fault rocks presently found in the hanging wall of the Alpine fault are inferred to have been derived from protoliths on both sides of the present-day principal slip zone, specifically the hanging-wall Alpine Schist and footwall Greenland Group. This implies that, at seismogenic depths, the Alpine fault is either a single zone of focused brittle shear that moves laterally over time, or it consists of multiple strands. Ultramylonites, cataclasites, and fault gouge represent distinct zones into which deformation has localized, but within the brittle regime, particularly, it is not clear whether this localization accompanies reductions in pressure and temperature during exhumation or whether it occurs throughout the seismogenic regime. These two contrasting possibilities should be a focus of future studies of fault zone architecture.

Understanding the seismic hazard posed by large earthquakes requires paleoseismic investigation because most faults have not ruptured repeatedly during the period of historic records. However, determining the location and length of fault ruptures using paleoseismic data remains challenging. Our study demonstrates that lake sediments record the high intensity shaking that occurs proximal to fault rupture, allowing the location and length of ruptures to be reconstructed. In two lakes adjacent to the Alpine Fault, New Zealand, seismic shaking is recorded as subaqueous mass-wasting derived turbidites formed by coseismic subaqueous slope failures, which are overlain by sets of hyperpycnites representing elevated fluvial sediment fluxes from earthquake-induced landslides. Precise radiocarbon age models show that shaking events are synchronous between the two lake sites and correlate with the timing of known Alpine Fault earthquakes. Modelled shaking intensities for the last two Alpine Fault earthquakes show that subaqueous mass-wasting occurs when shaking intensities exceed Modified Mercalli scale (MM) VI–VII, and that fluvial sediment fluxes from earthquake-induced landslides occur when shaking intensities exceed MM IX. The data demonstrate that lake records distinguish between strong (MM VI) and violent (MM IX) shaking at a lake site. The ability to map the spatial extent of MM IX shaking provides new insights into the timing and extent of rupture for the last five earthquakes on the Alpine Fault. The study demonstrates that lake deposits constrain the spatial extent of rupture during large earthquakes and may yield long records of the spatial and temporal patterns of fault rupture.

Continental transform faults are dominantly highly localized strike-slip shear zones hundreds of kilometers long that accumulate tens to hundreds of kilometers of displacement. From work on the Alpine Fault, we pose the questions: what is the deep structure of a continental transform, and how does the displacement become localized? We review research on the Alpine Fault and propose a model in which the fault partitions at depth into a steep zone extending into the mantle with largely fault-parallel motion and a flat ductile decollement in the lower crust. The fault localizes around two-thirds of the plate motion within a 100 km wide zone of distributed deformation. A review of other active continental fault systems suggests that variation between them may reflect their tectonic origins, the nature of the crust in which they develop, the presence of a significant oblique component of motion, and the displacement rate. All however have evidence for the development of a single principal fault zone that carries ≥50% of the total displacement and extends as a localized zone of shear into the upper mantle. We review mechanisms of strain weakening and suggest that localization of a principal fault may be initiated in the seismogenic crust and through a series of positive feedbacks eventually extend through the lower crust into the upper mantle.

Active fault traces are a surface expression of permanent deformation that accommodates the motion within and between adjacent tectonic plates. We present an updated national-scale model for active faulting in New Zealand, summarize the current understanding of fault kinematics in 15 tectonic domains, and undertake some brief kinematic analysis including comparison of fault slip rates with GPS velocities. The model contains 635 simplified faults with tabulated parameters of their attitude (dip and dip-direction) and kinematics (sense of movement and rake of slip vector), net slip rate and a quality code. Fault density and slip rates are, as expected, highest along the central plate boundary zone, but the model is undoubtedly incomplete, particularly in rapidly eroding mountainous areas and submarine areas with limited data. The active fault data presented are of value to a range of kinematic, active fault and seismic hazard studies.

The Eyre Creek Melange, a north–northwest trending 0.5–1 km-wide zone within Caples Terrane rocks in northern Southland, New Zealand, is composed of phyllonite and mylonitic rocks, and includes lenses of spilite, chert and altered microgabbro. This shear-zone melange (SZM) is enclosed within undifferentiated textural zone IIA Caples Terrane sandstones (Q10F14L76) that are metamorphosed to pumpellyite–actinolite grade. The sandstones correlate with Caples Terrane Bold Peak Formation, and display features consistent with deposition as proximal through distal turbidity current deposits, including pelitic and olistostromal material. Altered microgabbros within the SZM contain relict clinopyroxenes that provide an alkali basalt signature, and display within-plate affinities. Banded red and white chert lenses include recrystallized radiolarian tests, and provided a conodont remnant, resembling Carboniferous Idiognathodontinae Taphrognathus varians. The SZM probably developed in multiple stages during subduction...

Lake basins that experience rapid rates of deposition act as high-resolution environmental archives because they produce sedimentary records that have centennial or even decadal resolution. However, identifying target fractions for radiocarbon dating of lake sediments remains problematic because reworked organic material from fluvial catchments can produce anomalously old radiocarbon ages. This study determines the extent to which reworked material from catchment soils impacts radiocarbon dates on pollen and other organic concentrates by comparing radiocarbon dates produced by these techniques against a chronostratigraphic marker in cores from Lake Mapourika, New Zealand. Pollen preferentially preserved and reworked from catchment soils was identified using soil palynology. A technique was then developed to remove reworked pollen types from pollen concentrates extracted from lake sediment. Identification and removal of reworked pollen from pollen concentrates produced ages that were consistently closer to the age of the chronostratigraphic horizon than other organic concentrates. However, these dates were still between 736 and 366 calendar years older than expected. The only organic fractions that reliably reproduced the age of the chronostratigraphic horizon were terrestrial leaf macrofossils, although terrestrial leaf macrofossils isolated from megaturbidite deposits, which are formed by high-energy depositional events, also provided anomalously old ages. The results indicate that leaf material extracted from hemipelagite, which accumulates gradually, is likely to be the only organic fraction to produce reliable chronology in lakes where a component of sedimentation is driven by the fluvial system. The results also demonstrate the importance of conducting a detailed investigation of physical sedimentology before selecting material for radiocarbon dating lake sediments.

Lineations within mylonites exhumed in the hanging wall of New Zealand's active Alpine Fault zone have a complicated relationship to contemporary plate kinematics. The shear zone is triclinic and macroscopic object lineations are not usually parallel to the simple shear direction, despite high total simple shear strains ( γ ≥ 150). This is mostly because the lineations are inherited from pre-mylonitic fabrics, and have not been rotated into parallelism with the mylonitic stretching direction (which pitches c. 44° in the fault plane). Furthermore, some lineations have been variably rotated depending on whether they are present in shear bands or microlithons, which accommodated bulk strains with different vorticities. Total strains required to obtain parallelism between the finite maximum principal stretching direction calculated from transpression models and these mylonitic lineations, are pure shear stretch, S 1 ∼ 3.5; simple shear 11.7 γ

We combine recently acquired airborne light detection and ranging (LiDAR) data along a portion of the Alpine fault with previous work to define the ways in which the plate-boundary structures partition at three different scales from 6 to 10 0 m. At the first order ( 6 –10 4 m), the Alpine fault is a remarkably straight and unpartitioned structure controlled by inherited and active weakening processes at depth. At the second order (10 4 –10 3 m), motion is serially partitioned in the upper ∼1–2 km onto oblique-thrust and strike-slip fault segments that arise at the scale of major river valleys due to stress perturbations from hanging-wall topographic variations and river incision destabilization of the hanging-wall critical wedge, concepts proposed by previous workers. The resolution of the LiDAR data refines second-order mapping and reveals for the first time that at a third order (10 3 –10 0 m), the fault is parallel-partitioned into asymmetric positive flower structures, or fault wedges, in the hanging wall. These fault wedges are bounded by dextral-normal and dextral-thrust faults rooted at shallow depths (

Rock damage during earthquake slip affects fluid migration within the fault core and the surrounding damage zone, and consequently coseismic and postseismic strength evolution. Results from the first two boreholes (Deep Fault Drilling Project DFDP-1) drilled through the Alpine fault, New Zealand, which is late in its 200–400 yr earthquake cycle, reveal a >50-m-thick “alteration zone” formed by fluid-rock interaction and mineralization above background regional levels. The alteration zone comprises cemented low-permeability cataclasite and ultramylonite dissected by clay-filled fractures, and obscures the boundary between the damage zone and fault core. The fault core contains a

Large earthquakes in mountain regions commonly trigger extensive landsliding and are important drivers of erosion, but the contribution of this landsliding to long-term erosion rates and seismic hazard remains poorly understood. Here we show that lake sediments record postseismic landscape response as a sequence of turbidites that can be used to quantify erosion related to large (moment magnitude, M w > 7.6) earthquakes on the Alpine fault, New Zealand. Alpine fault earthquakes caused a threefold increase in sediment flux over the ∼50 yr duration of each postseismic landscape response; this represents considerable delayed hazard following earthquake-induced strong ground motion. Earthquakes were responsible for 27% of the sediment flux from the lake catchment over the past 1100 yr, leading us to conclude that Alpine fault earthquakes are one of the most important drivers of erosion in the range front of the Southern Alps.

In order to assess how different elements of plate boundary-scale shear zone tectonite fabrics reflect the boundary kinematics and strain history, we have examined a range of such structures within the active Alpine Fault mylonite zone, New Zealand. In this oblique slip zone we find that quartz CPO fabric symmetry and individual mica grain lineations reflect incremental stretches but develop stable orientations parallel to the finite stretching direction (S 1 ). This S 1 trends oblique to the plate motion vector because the shear zone experiences some extrusion toward the free surface. Conversely, asymmetric structures (such as shear bands and mica fish) reflect simple shear-dominated deformation late in the shear zone history, parallel to the imposed relative plate motion vector. Finally, linear fabrics are a poor record of the mylonitic deformation in this zone, because (1) they are dominantly inherited and have been only partially re-oriented during mylonitisation, and (2) macroscopically they become less visible at higher strains due to fabric homogenisation. Generally, these results indicate that, particularly in heterogeneous, polyphase shear zones, a full understanding of the movement history can only be obtained by considering kinematic information from as many different sources as possible.

Strata and fault relationships revealed in five trenches excavated across the recent trace of the Alpine fault at the Haast, Okuru, and Turnbull Rivers, South Westland, New Zealand, record the three most recent surface‐faulting events. Using back‐stripping techniques to remove the three faulting events and the sedimentary units associated with the faulting restores the cross‐sections to gravel‐bed floodplains at the Haast and Okuru Rivers, at about A.D. 750. Horizontal and vertical offsets of stream channels and terrace risers reveal characteristic displacements of about 8–9 m dextral and up to 1 m vertical per event. Cumulative dextral displacement is 25±3 m in the past three events. The most recent surface‐rupture event was probably in A.D. 1717, and the next prior events were about A.D. 1230±50 and about A.D.750±50. The timing of these events is consistent with past large‐great earthquakes on the southern section of the Alpine fault inferred from off‐fault data, but there are fewer events identified in trenches. Our three‐event dataset indicates the average surface‐rupture recurrence interval for the South Westland section of the fault is about 480 years, much longer than the current elapsed time of 295 years. Therefore, the Alpine fault in South Westland may not be close to rupture as is often speculated.

Low-grade albite- and peristerite-mylonites occur as a northeast-tapering, 200–300 m wide zone structurally above basal cataclasite on the Alpine fault (New Zealand). They form an inverted metamorphic sequence grading eastward into an ∼1-km-wide, amphibolite-facies−derived mylonite zone in the hanging wall. Their mineralogy matches low-grade protoliths that occur 100–110 km to the southeast. This displacement is compatible with distributed lateral and vertical shear within the mylonite zone since inception of significant convergence on the plate boundary in the late Neogene.

The thermal state within plate boundary fault zones is likely to influence the dimensions of large co-seismic fault ruptures, and strain localisation in underlying mylonitic shear zones. Vein microstructures, biotite chemistry and fluid inclusion data offer new insights into the thermal structure of the mylonitic shear zone down dip of the Alpine Fault, the main Pacific–Australian plate boundary structure through southern New Zealand. Quartz veins within the fault rocks were formed and deformed at temperatures ranging from > 500 °C to immediately above that of the brittle–viscous transition (∼ 325 °C), illustrating that fluids are present throughout the fault zone. Microthermometric data from primary, mixed CO 2 –H 2 O fluid inclusions in boudin-neck quartz–carbonate veins indicate that these inclusions were trapped at temperatures of 325 ± 15 °C and pressures of c. 40 MPa. Vein microstructures indicate that fluid temperatures were equal to rock temperatures. Assuming hydrostatic fluid pressure, the veins were trapped at depths of 2 O–CO 2 fluids in the shallower crust of > 500 kg m − 3 . Biotite grains in the mylonites are a mixture of porphyroclastic grains with high R = Ti/Mg# inherited from the parent Alpine Schist, and smaller, low R grains that recrystallised during mylonitisation, particularly those that are found within syn-mylonitic shear bands. Assuming a simple step-function model approximating diffusive equilibration, the changes in biotite chemistry indicate Δ T during mylonitic deformation of ∼−340 °C according to the Ti-in-biotite thermometer of Henry et al. (2005). A best-fit uplift P – T path to existing data and these new results indicate that the uplift path P – T gradient of fault zone rocks above the brittle–viscous transition is of the order of 40 °C km − 1 . Below the brittle–viscous transition, the uplift path P – T gradient is 10 °C km − 1 . The P – T profile within the fault rocks is comparable to that in the immediate hanging wall Alpine Schist, indicating there is no peak in the crustal isotherms in the fault hanging wall, as was indicated by previous numerical models, and analyses that did not consider data from the fault rocks themselves.

Strong quartz crystallographic preferred orientations (CPOs) were developed during dislocation creep in the mylonite zone exposed in the hanging wall of the dextral reverse Alpine Fault Zone, New Zealand. The CPOs have a consistent asymmetry indicating a high ratio of simple to pure shear strain, with a shear sense of dextral-up to the NW, consistent with the mesoscopic shear sense indicators and with slip on the active fault. There is a transition from Y-maxima and asymmetric single girdles in mylonites and ultramylonites within 300 m of the present fault trace, to cross-girdle fabrics in the protomylonites further from the fault. The strong Y-maxima or single girdle CPOs are ascribed to high ductile shear strains under amphibolite facies conditions while the cross-girdle patterns found in the protomylonites are interpreted to represent deformation under lower temperature conditions. However, the observed fabric transition cannot logically be attributed to variations in temperature during the last increment of deformation. The highly oriented Y-maximum fabrics formed at high temperature contain very few grains suitably oriented for basal slip, so that the slip systems activated during subsequent shear at lower temperatures during exhumation were prism or rhomb . Further from the fault, where shear strains under high temperature conditions were lower, weaker fabrics developed under these conditions were modified at higher levels in the crust into crossed girdle patterns. This interpretation implies that intense localisation of shear strain along the fault zone within the lower crust must have occurred in order for the high-temperature fabrics to become sufficiently intense to be preserved. This in turn implies that deep-seated localised shear was taking place early during the evolution of the current oblique-slip plate boundary fault. The results of this study also indicate that the current model that operation of certain slip systems in quartz is mostly a function of temperature is incorrect. Future work should consider the effects of deformation history, total strain and the presence of other mineral phases.

The Golden Bar gold mine pit near Macraes, east Otago, was developed in Textural Zone (TZ) 3 of the Otago Schist, at least 100 m structurally up‐section from the Hyde‐Macraes Shear Zone. This study tests the hypothesis that the Golden Bar mineralised zone was an extension of the major Hyde‐Macraes Shear Zone mineralisation system. The Golden Bar rocks have been extensively, but subtly, hydro‐thermally altered. Principal hydrothermal alteration effects are replacement of titanite by rutile, replacement of epidote by siderite and kaolinite, and addition of pyrite. Mineralised (gold‐bearing) rocks within this hydrothermal alteration zone are volumetrically minor and include micaceous schist with disseminated pyrite and arsenopyrite, silicified micaceous schist, and quartz veins. The structurally lowest mineralised rocks occur in a well‐defined foliation‐parallel zone of black sheared rocks with abundant hydrothermal graphite and sulfide minerals. The top of the mineralised zone is defined by a duple...

Hydrothermal processes along two regional-scale shear zones in the Otago Schist were dominated by structurally controlled fluid flow and mineralization in the host schist, with relatively minor quartz vein formation, and mineralized rocks are only subtly different from unmineralized rocks. Most Au in the shear zones is associated with sulphide minerals (pyrite and arsenopyrite) disseminated through the host schist or along microshears. Minor enrichment of Sb, Mo and Bi (ppm level) is detectable in the Hyde-Macraes Shear Zone (HMSZ). Hydrothermal muscovite is slightly more aluminous (1–2 wt%) than metamorphic muscovite in both shear zones. HMSZ muscovite averages >900 ppm N, in contrast to metamorphic muscovite that averages c. 200 ppm N. In both shear zones, rutile has replaced metamorphic titanite and epidote has altered to carbonate and phyllosilicates, but these reactions were nearly isochemical. Structurally controlled hydrothermal graphite in the HMSZ occurs in microshears (up to 3 wt%, above background

The Rise & Shine Shear Zone is a late metamorphic deformation zone developed in biotite zone Textural Zone 4 schist in Central Otago. The shear zone has been hydrothermally altered, with addition of gold associated with replacement of schist minerals by pyrite and arsenopyrite. Hydrothermal alteration of schist during mineralisation involved replacement of titanite by rutile, recrystallisation of metamorphic quartz, muscovite and chlorite, and addition of ankerite. Mineralised schist has abundant microshears that have developed parallel and subparallel to the pervasive schist foliation, and these microshears contain much of the hydrothermal sulfides and gold. Microshears have been deformed locally by upright syn‐mineralisation brittle reverse faults and angular folds that have a southerly axial trend. These more brittle deformation zones are confined to the Rise & Shine Shear Zone. Gold‐bearing veins and mineralised breccias, made up of quartz, albite, pyrite, arsenopyrite, calcite, and chlorite,...

Measurements of 10Be concentrations in quartz‐rich boulders exposed by the uplift of anticlinal ranges in Central Otago, New Zealand, are used to investigate the fault propagation styles and rates for the underlying blind reverse faults. 10Be ages along Little Rough Ridge reveal the propagation rate for this fault to be between zero (i.e., not propagating at all, if the ages reflect emergence of a fault of fixed length through a sloping surface) and a maximum of 8.1 mm/yr, if the ages reflect true lateral growth. Regardless of the propagation rate, Little Rough Ridge has established its length rapidly compared with the rate at which it has accumulated displacement. On Raggedy Range, the 10Be ages, together with structural data, imply that the ridge reached its current length by the merging of several small fault segments. This study emphasises the importance of combining 10Be measurements with geomorphic observations. Without understanding the geomorphological context, 10Be exposure ages are diff...

Glacial landforms at 12 localities in 9 river valleys are offset by the southern end of the onshore Alpine fault. Offsets cluster at ∼435, 1240, and 1850 m, consistent with evidence for glacial retreat at 18, 58, and 79 calendar ka. The peak of an offset fluvial aggradation surface is correlated with the Last Glacial Maximum at 22 ka. Displacement rates derived from features aged 18, 22, 58, and 79 cal. ka are 24.2 ± 2.2, 23.2 ± 4.9, 21.4 ± 2.6, and 23.5 ± 2.7 mm/yr, respectively, with uncertainties at the 95% confidence level. The joint probability, weighted mean, and arithmetic mean of all observations pooled by rank are 23.1 ± 1.5, 23.2 ± 1.4, and 23.1 ± 1.7 mm/yr, respectively. We conclude that the mean surface displacement rate for this section of the Alpine fault is 23.1 mm/yr, with standard error in the range of 0.7–0.9 mm/yr. The reduction in estimated long-term slip rate from 26 ± 6 mm/yr to 23 ± 2 mm/yr results in an increase in estimated hazard associated with faulting distributed across the rest of the plate boundary. Model-dependent probabilities of Alpine fault rupture within the next 50 yr are in the range 14%–29%. The 36 ± 3 mm/yr of total plate motion (NUVEL-1A) is partitioned into 23 ± 2 mm/yr of Alpine fault dextral strike slip, 12 ± 4 mm/yr of horizontal motion by clockwise block rotations and oblique dextral-reverse faulting up to 80 km southeast of the Alpine fault, and 5 ± 3 mm/yr of heave on reverse faults at the peripheries of the plate boundary.

The Alpine Fault accommodates around 60–70% of the 37 mm/yr oblique motion between the Australian and Pacific plates in the South Island of New Zealand. Uplift on the fault over the past 5 Ma has led to the exhumation of the deep-seated mylonite zone alongside the present surface trace. Shear strain estimates in the mylonites reach 200–300 in the most highly strained rocks, and provide an integrated displacement across the zone of 60–120 km. This is consistent with the amount of displacement during the last 5 Ma, suggesting that displacement on the fault is localised within a 1–2 km wide ductile shear zone to depths of 25–30 km. Existing geodetic data, together with Late Quaternary slip rate and paleoseismic data, are consistent with the steady build-up and release of elastic strain in the upper crust driven by ductile creep within a narrow mylonite zone at depth. Faults of the Otago Fault System form a parallel array east of the Alpine Fault and accommodate c. 2 mm/yr contraction. Long periods of quiescence on individual structures suggest episodic, or “intermittently characteristic”, behaviour. This is more consistent with failure on faults within an elastico-frictional upper crust above a ductile lower crust. Localisation of crustal deformation may be initiated by inherited weaknesses in the upper crust, with downward propagation of slip causing strain weakening within the ductile zone immediately beneath. Inherited structures of great length focus a greater amount of displacement and hence more rapidly develop underlying zones of ductile shear.

Oblique displacement on the Alpine Fault, which forms the principal structure along the Australian–Pacific plate boundary in South Island, New Zealand, has resulted in exhumation of a kilometre-wide mylonite zone in the hanging wall adjacent to the current brittle fault trace. The mylonites formed under amphibolite facies conditions at depths of ca. 25 km and have been uplifted during the past 5 Ma. A suite of 65–70 Ma pegmatite veins in the hanging wall Alpine schists has been progressively deformed within the mylonite zone and sheared out over a strike length of ca. 100 km. Measurements of the thickness distribution of the pegmatite veins within the non-mylonitised schists and at three localities within the progressively strained mylonites have been used to estimate strain values within the mylonites. The thicknesses approximate a log-normal distribution, with a mean value that is progressively reduced through the protomylonites, mylonites and ultramylonites. By assuming that the thickness distribution currently observed in the schists was the same for the pegmatites within the mylonites before strain, a model of deformation incorporating simple shear and simultaneous pure shear is used to strain the undeformed veins until a fit is obtained with the strained distributions. Shear strains calculated range from 12 to 22 for the protomylonites, 120 to 200 for the mylonites and 180 to 300 for the ultramylonites, corresponding to pure shear values of 1–3 in each case. These values are compatible with the strains predicted if most of the surface displacement on the fault over the past 5 Ma were accommodated within a 1–2-km-wide mylonite zone through the middle and lower crusts. The results suggest that processes such as erosional focussing of deformation and thermal weakening may cause intense strain localisation within the lower crust, with plate boundary deformation restricted to narrow zones rather than becoming increasingly distributed over a widening shear zone with depth.

Published geological data on Late Quaternary offsets on the Alpine Fault, New Zealand, have been assembled into a common format and analysed with respect to uncertainties. Uncertainties arise mainly from measurement of offset features, relating apparent offsets to actual fault slip, and dating the offset features. Despite the considerable uncertainties, the data form a coherent set consistent with a relatively constant rate of strike-slip of 27±5 mm/year between Milford Sound and Hokitika. This rate represents 70–75% of the fault-parallel interplate motion. North of the confluence with the Hope Fault, the rate drops substantially. Dip-slip rates, on the other hand, show considerable variation along strike, rising to a maximum of more than 10 mm/year in the central section and decreasing to zero at the southern end. Partitioning of c. 25% of the interplate slip on to structures east of the Alpine Fault occurs in the central section, consistent with predictions from critical wedge models. The partitioning of all the fault-normal component of displacement on to other structures in the south may be related, in part, to a doubling in width of the deforming wedge to the east. Most probably, however, the fault-normal displacement is mainly accommodated by underthrusting of the Australian plate offshore, due to a change in the nature of the crust from continental to oceanic.

The Akatore Fault comprises the onshore leading edge of the Otago reverse fault‐fold belt, part of the outboard zone of Australia‐Pacific plate continental collision in the South Island. We examine stratigraphic and sedimentologic evidence of late Holocene reverse faulting along the Akatore Fault, and use radiocarbon dating to constrain fault rupture events and associated uplift of the Akatore fault block. Two fault rupture events are recorded in swamps on the downthrown, upstream side of the fault, as forest and/or peat horizons buried by silt, dated at post‐3800 yr BP and 1150–1000 yr BP. The latest event coincides with the timing of uplift of a 3 m marine terrace on the coast, and the uplift of a second (6 m) is tentatively correlated to the penultimate event. The terrace height data, together with the amount of offset of alluvial gravels resting on the fluvial strath surface at the blocked swamp sites, suggests an average of 3 m of throw per uplift event. Distribution of throw along the fault...

The south Otago coast is characterised by a flight of marine terraces which have formed and been uplifted in Quaternary time. Optical luminescence dating along the recently active Akatore Fault has provided burial ages for beach sands resting upon a wave‐cut platform in the up thrown block that are equal to or younger than 71 ± 14 ka, and a burial age of c. 20 ka for the loess cap. The ages on beach sands clearly indicate that motion on this fault has been relatively limited since the last interglacial, while the loess date constrains the age of at least one faulting event on the Akatore Fault at this locality, and two events on the fault farther south, to be younger than 21 ka. The 6 m high (above high sea level) sand units in the terrace shown here to date from the late part of the last interglacial, suggest that sea levels at that time (oxygen isotope substage 5a) may not have been as low as ‐18 m, but rather closer to modern sea level. These applications of optical luminescence dating, using ...

The NNE‐striking faults of the Nokomai valley, Southland, New Zealand, are the southern extension of the Nevis‐Cardrona Fault System, and have similar deformation style. This fault zone defines the boundary between broad, smooth, central Otago ranges and basins dominated by a deformed regional unconformity cut into the basement, and rugged mountains and deeply incised bedrock valleys to the west. Nokomai valley topography is dominated by bounding ranges: Slate and Hector Ranges to the west, and Garvie Range to the east. These ranges are rising on NNE‐striking oblique reverse faults, which dip beneath the ranges in the upper and lower Nokomai valley. The middle reaches of the Nokomai valley are cut by a complex set of NE‐ENE strike‐slip faults with some reverse component. These faults pass through the headwaters of eastern tributaries of the main Nokomai River, which itself passes through a bedrock gorge. Structural depressions have developed in the upper and lower Nokomai valleys, where the obliq...

Blackstone Hill is a northeast‐trending ridge of schist basement rising between the Ida Burn and Manuherikia River valleys. Deformation of young sediments and uplift of strath river terraces cut into both sides of the ridge indicate that its rise is an ongoing, late Cenozoic phenomenon. Deformation of poorly exposed cover sediments at the ridge margins appears to be relatively localised and complex. In the schist basement, a new generation of fractures has developed to accommodate late Cenozoic deformation. The basement also displays a prominent lineation and foliation of Mesozoic age. Lineation defined by quartz rodding trends c. 020°. The orientation of foliation varies and defines the Blackstone Hill Nappe, a large synmetamorphic fold of Mesozoic age with a hinge trending c. 167°. The Otago Peneplain, the regional unconformity between schist basement and overlying sedimentary cover, was a subhorizontal erosion surface before the onset of ridge formation. It is now warped around the late Cenozo...

The Waikukupa thrust is a 4-km long oblique thrust segment of the New Zealand Alpine Fault that has developed over the last 65,000 years, emplacing mylonite and cataclasite over fluvioglacial gravel. During the last 20,000 years or so, a 2-km long ‘out-of-sequence’ imbricate, the Hare Mare thrust, has formed on the east side of the Waikukupa River valley. The average rate of slip on the two thrusts is estimated at 22–30 mm/year. Analysis of the fault structure in terms of critical wedge theory is consistent with imbrication resulting from a reduction in wedge taper below the critical value due to rapid river erosion. The internal structure of the Waikukupa thrust sheet consists of reverse faults and angular folds forming duplex-like systems, associated with minor strike-slip faults. Within the limits of uncertainty, the structures are consistent with strain accumulating during internal transpressional shear of the thrust sheet while the taper of the latter was subcritical prior to the transfer of fault displacement to the Hare Mare imbricate. The thrust complex forms part of a serrated range front fault system consisting of oblique thrust sections linked by strike-slip faults. We introduce the term ‘serial partitioning’ for this type of partitioned transpressional fault system. Its development is linked to the erosional processes at the range front. An alternative system consisting of parallel thrust and strike-slip faults we here term ‘parallel partitioning’. While angle of obliquity is an important parameter in determining which partitioning model develops, we suggest that erosion rate is also an influential factor.

The Otago Fault System in the southern South Island of New Zealand is an area dominated by actively growing asymmetric anticlines above buried reverse faults. Geological studies indicate that the average total shortening rate across the system is 2–3 mm/year. This region was the subject of several early geodetic studies based on terrestrial techniques, which reported high (0.5 ppm/year) strain rates that are difficult to reconcile with geological estimates of fault slip rates.

Central Otago in New Zealand is an area of active continental shortening in which a peneplain surface cut into schist has been deformed by folds, which are developed above buried reverse faults. We use the drainage patterns in this region to demonstrate various processes in fold (and fault) growth and interaction that would be difficult to identify by other means. In particular we show: (1) how simple asymmetric folds can develop into box folds; (2) how apparently continuous ridges were formed by the coalescing of quite separate propagating fold (and fault) segments; (3) evidence for the relative ages (or relative uplift rates) of adjacent structures; and (4) evidence for the propagation direction of folds (or faults) as they grow. The few quantitative estimates we obtain for fault propagation rates suggest an increase in length of 10–50 m per earthquake on faults about 20 km long. These estimates are very uncertain, but are similar in magnitude to an estimate made in Nevada for a normal fault of similar size and are also similar to predicted estimates from theoretical growth models. They raise the question of whether fault growth, earthquake recurrence rates and climate change can interact to produce semi-regular discrete features in an active landscape.

A 400 ± 100 m offset of Lake McKerrow, South Westland, New Zealand, combined with dated (15.6 ka) glacial lake silts, requires an Alpine Fault displacement rate of 26 ± 7 mm/yr. Moraines associated with Hokuri Creek (assumed to be 17 ± 2 ka) are offset by 440 ± 40 m and require a displacement rate on the Alpine Fault of 26 ± 6 mm/yr. Slickensides, fault exposure, and offset topography are consistent with an almost pure dextral sense of movement on a vertical or subvertical fault. Locally, a small vertical component of up‐to‐the‐west movement is observed. Folding in late Quaternary sediments indicates active tilting of sediments at up to 0.4°/ka and variations in local uplift/subsidence rates of up to 4 mm/yr. At one locality c. 1 km northwest of the Alpine Fault and near the core of an anticline, uplifted shells require an uplift rate of 1.4 ± 0.5 mm/yr relative to sea level. Displaced river channels provide estimates of the last two coseismic displacements on the fault of 9 m (penultimate) and 8...

On the north bank of Haast River, the Alpine Fault trace has an 047–057° trend. The trace comprises left and right stepping en echelon, linear segments, each up to 200–300 m long, with a prominent terrace upthrown by between 5 and 10 m on the southeast side. A creek channel, deeply incised into the terrace 1.1 km north of the river, is offset dextrally by 94 m at the fault scarp. Exposed in the creek walls is a sequence of subterrace sediments, up to 9 m thick, that show an upward transition from clean, well‐sorted sands, through fine‐grained sands and carbonaceous clays containing a persistent peat horizon, to an erosive, coarsegrained, wood‐bearing gravel unit. The sequence is interpreted as a regressive transition from beach, through swamp or overbank fluvial, to high‐energy fluvial environments. As in other South Westland localities, the postglacial regression is in marked contrast to the pronounced global transgression predicted from rising sea levels. Radiocarbon dating of the peat horizon ...

Recent mapping of the Alpine Fault trace in the Paringa region has revealed the existence of an extensive Haast Schist‐derived thrust nappe resting on Western Province basement rock and moraine. Erosion of the nappe by the Paringa River and its tributaries, however, has resulted in eastward propagation of the active fault zone, forming southeast‐dipping thrust faults linked by swarms of steeply dipping strike‐slip faults. Late Quaternary sediments of the Paringa Formation have been intensely deformed along a newly developed zone of shortening and uplift on the northeast side of Paringa River. Marine, fluviatile, lacustrine, and terrestrial sediments record progressive uplift east of the Alpine Fault. The occurrence of lake deposits rhythmically interbedded with forest horizons may have resulted from damming of the Paringa River behind the zone of rapid uplift. The uplift rate for this region over the last 16 ka has been calculated at 13.7 ± 1 mm/yr. After removing the effects of tilting due to lo...

Gold‐scheelite (Au‐W) vein mineralisation in the Mesozoic Haast Schist (Otago, Alpine, and Marlborough Schist belts) resulted from metamorphic dewatering and fluid channelling periodically throughout the >100 Ma tectonic history of the schist belt. The earliest metal mobilisation occurred during synmetamorphic vein formation and was both accompanied and succeeded by ductile deformation. Structural style of deformation within vein zones became progressively more brittle as mineralisation accompanied Mesozoic‐Tertiary isostatic and extension‐related uplift of the Otago Schist. Rapid tectonic uplift in the Late Cenozoic resulted in mineralisation in the narrow belt of Alpine Schist along the Alpine Fault. The wide range of structural styles of vein systems seen in the Haast Schist represents different erosion levels exposed in tectonic zones. Geothermal gradients during Otago Schist mineralisation appear to have been close to normal, whereas mineralisation in the Southern Alps took place under anoma...

The Hyde-Macraes shear zone in east Otago, New Zealand, is a low-angle north- to northeast-dipping thrust system in which gold-bearing quartz veins have been deposited. Mineralization accompanied deformation in the shear zone which is preferentially developed in graphitic pelitic schist. The shear zone is a duplex system dominated by steep hinterland-dipping thrust faults between well-defined shallow-dipping upper and lower bounding thrusts. Imbrication and stacking have occurred in the shear zone, especially where intrashear schist is psammitic and relatively brittle. This stacking contributes to lateral thickness changes from 2 up to 125 m. Mineralized veins formed mainly in dilational jogs or localized extensional zones on the relatively shallow-dipping parts of the intrashear thrust faults. Late-stage shallow south-dipping extensional veins crosscut the duplex structures but are related to the same overall deformation. From analysis of the shear zone structure, it is possible to make predictions about the geometry of mineralized schist within unexposed parts of the shear zone. This is a potentially useful tool for further prospecting and mining in hydrothermally mineralized duplex systems.

A study of lineaments traced from aerial photographs shows three prominent orientations in the region immediately east of the Alpine Fault between the Fox and Waiho valleys, Westland, New Zealand. One set trends about 025°, parallel to the foliation in the Alpine Schist. A second, minor set of lineaments trends 130° and may be related to extension fractures. The third lineament set trends east-west. Detailed fieldwork on prominent east-west faults in the Waiho valley shows that they are dextral strike-slip faults, with evidence for additional late-stage normal displacement. All of the faults investigated exhibit only small displacements. Fault orientations are consistent with theoretical and geodetically measured regional strain east of the Alpine Fault.

Stratigraphic sections through sag ponds developed along the Alpine Fault trace between Milford Sound and John O'Groats River record drastic changes in sedimentation. Swamp or forest-floor material has been inundated by up to 0.5 m of cataclasite-derived sand and gravel. Two events, inferred to represent degradation of newly created scarps on the Alpine Fault, occurred at, or later than, 230 ± 50 yr B.P., and 1980 ±60 yr B.P., respectively. At John O'Groats River, large silver beech trees growing on the Alpine Fault scarp have been topped at heights of 8–15 m above ground level by movement on the fault. Unbroken beeches have mean ages calculated at 266 +90, -60 years. On a regional scale, a colluvial fan-forming event at Kaipo River and rotational slumping at Gorge Plateau are broadly synchronous with inferred fault movements near John O'Groats River. It is suggested that a large seismic event, related to the last major movement on the Alpine Fault, occurred in the middle 17th to early 18th century.

The prehnite-pumpellyite to greenschist facies rocks of the classic low-grade Otago schist terrane are subdivided into a number of textural zones based on the development of slaty cleavage and the appearance of segregation lamination in metagreywackes. In this study, conglomerates from fifteen localities have been studied in order to investigate the relationship between ductile strain and the sequence of textural zones. Strain measurements on lithic pebbles show a clear correlation between strain and textural zone, ranging from around 40% shortening in TZ IIA with the development of slaty cleavage to more than 70% in TZ IIIA corresponding to the appearence of segregation lamination. The much lower and more prolate strains of the quartzose pebbles compared to the corresponding lithic clasts may be explained by differential strain partitioning between clast and matrix. Microstructural evidence, including well-developed strain shadows and solution seams, indicates that solution-transfer was the dominant deformation mechanism. The development of quartz segregation laminae in TZ III is due to three related processes: the development of mica-rich solution seams demarcating quartzo-feldspathic microlithons; the development of amalgamated and recrystallised strain shadow overgrowths on clastic grains; and the formation, rotation and disruption of early formed veins. The large increase in strain from TZ II to TZ III is responsible both for increased segregation and for the rotation and flattening of segregation features into the foliation, thus producing a segregation lamination. Large strains developed by solution-transfer during the low-grade metamorphism of greywackes will lead inevitably to segregated schists.

The Alpine Fault of New Zealand forms the western boundary of a zone of distributed deformation formed by the oblique convergence of continental crust belonging to the Pacific and Australian plates. Structural and geodetic data from the Alpine Fault show that a large proportion of the total plate displacement is accommodated by rapid oblique slip on the fault. The remaining displacement is distributed over a 200 km wide zone to the east. The collision may be modelled as a two-sided deforming orogen, with the partitioning of deformation being controlled by erosional differences between the narrow high-strain inboard and broad low-strain outboard sides. In ancient collision zones, little evidence may remain of the nature and amount of displacement on the inboard side. Partitioning of deformation among pre-existing structures complicates interpretation of the outboard zone. Radiometric ages may post-date collision by several tens of millions of years and indicate slow isostatic uplift and unroofing. Collision ages, if preserved, may be recognized by high uplift rates calculated from muscovite-biotite pairs.

Geological observations require that episodic slip on the Alpine fault averages to a long-term displacement rate of 2-3 cm/yr. Patterns of seismicity and geodetic strain suggest the fault is locked above a depth of 6-12 km and will probably fail during an earthquake. High pore-fluid pressures in the deeper fault zone are inferred from low seismic P-wave velocity and high electrical conductivity in central South Island, and may limit the seismogenic zone east of the Alpine fault to depths as shallow as 6 km. A simplified dynamic rupture model suggests an episode of aseismic slip at depth may not inhibit later propagation of a fully developed earthquake rupture. Although it is difficult to resolve surface displacement during an ancient earthquake from displacements that occurred in the months and years that immediately surround the event, sufficient data exist to evaluate the extent of the last three Alpine fault ruptures: the 1717 AD event is inferred to have ruptured a 300-500 km length of fault; the 1620 AD event ruptured 200-300 km; and the 1430 AD event ruptured 350-600 km. The geologically estimated moment magnitudes are 7.9 ± 0.3, 7.6 ± 0.3, and 7.9 ± 0.4, respectively. We conclude that large earthquakes (Mw >7) on the Alpine fault will almost certainly occur in future, and it is realistic to expect some great earthquakes (Mw ≥8).

The Alpine Fault forms part of the on-land Pacific-Australian plate boundary. Generally the fault has a simple straight trace, striking ca. 055° and dipping moderately eastwards, but in central areas it is serially partitioned. Late Quaternary strike-slip rates are 23-25 mm/yr, but dip-slip rates range from ca. 10 mm/yr in the central part, reducing north and south and reaching zero southwest of Jackson Bay. Due to oblique-slip, mylonites are exhumed from ca. 25 km in a ca. 1 km wide fault zone east of the present trace, undergoing intense cataclasis close to the fault at shallow levels. Pseudotachylytes are generated by preferential melting of phyllosilicaterich assemblages. Ductile shear strain is high, reaching 200-300 in ultramylonites. Displacement of metamorphic zones and a pegmatite swarm, with respect to their location in the hanging wall is consistent with ca. 100 km of dextral ductile shear within the mylonites since ca. 5 Ma.

[1] In situ cosmogenic 10Be and 26Al measurements on sedimentary quartzites, together with analysis of abandoned river terraces, can be used to show how drainage responded to the lateral propagation of a late Quaternary anticline forming above a blind reverse fault in Central Otago, New Zealand. A close link between the progression of cosmogenic ages and the tectonic geomorphology allows us to confirm that uplift and propagation rates on the anticline are in the ranges 0.08–0.12 mm yr−1 and 0.8–1.5 mm yr−1, respectively, over the last 550,000 years. The agreement between the isotopic and geomorphological evidence in turn requires that minimum 10Be exposure ages as great as 660 ka are not in steady state with respect to erosion. This is a geochemical result of significance, as it opens the possibility of a more regional analysis of landscape evolution in this region, using the same rocks. On an adjacent anticline, thought to be even older on geomorphological arguments, 10Be concentrations give minimum exposure ages in the range 750–1400 ka. These extremely old minimum ages are rare worldwide, as they are usually limited to smaller values by erosion. They are attributable to the very resistant nature of the quartz-rich boulders in which the measurements were made, which in places are shown to have erosion rates lower than 0.5 mm kyr−1.

$^{10}Be$ concentrations measured in silica-cemented Tertiary sandstones in Central Otago have yielded minimum exposure ages of up to 1400 ka, some of the oldest ever recorded outside the dry valleys of Antarctica and some arid deserts. The silica-cemented sandstones outcrop as boulders in a region where their exposure is caused by the growth of anticlinal ridges above blind reverse faults. Initial studies using a combination of in situ cosmogenic isotope measurements and geomorphological observations on one range have demonstrated the potential of this technique to investigate Late Cenozoic fault growth and erosion rates. The best-exposed and preserved occurrence of the quartz-rich boulders is at German Hill on North Rough Ridge, where their stratigraphic context and their sedimentary and diagenetic origin, together with their method of emplacement and preservation on the modern land surface, can be studied in detail. These are all features that affect their suitability for cosmogenic isotope studies of landscape development and are discussed in this article. $^{10}Be$ concentrations in the German Hill rocks yield minimum exposure ages of 600–1300 ka, and it is possible that these concentrations are not in saturated steady state with respect to erosion. These ages imply that the uplift of North Rough Ridge by its underlying fault is a Quaternary phenomenon. This conclusion has implications for the likely earthquake recurrence on the fault, which with an estimated average long-term slip rate of 0.5 mm yr$^{-1}$, is expected to be infrequent.

We use in situ cosmogenic 10 Be measurements in quartzites to examine the growth and propagation rates of a Late Quaternary anticline forming above a blind reverse fault in Central Otago, New Zealand. We obtain average uplift and propagation rates of 0.10–0.15 mm yr −1 and 1.0–2.0 mm yr −1 respectively over the last 450 000 yr, though it is probable that growth occurred episodically, with more rapid development in the periods 600–400 kyr and 200–100 kyr and relative quiescence between about 400 and 200 kyr ago. These results quantify and confirm a qualitative analysis of the fault evolution based on geomorphology and drainage patterns. We obtain minimum 10 Be exposure ages of up to 750 000 yr for the quartzites, some of the oldest ages obtained by this method outside Antarctica, which are made possible by the extremely low erosion rates of less than 0.6 mm kyr −1 in the resistant quartzites. These rocks consequently provide an opportunity to study geomorphic and tectonic processes in this region over unusually long time periods.

De Pascale et al. (2014) present a new “bimodal” fault rupture behavior model for the dextral-reverse Alpine fault using evidence from several new, small (