Subduction Margin Research Articles

Distribution and depth of bottom-simulating reflectors in the Nankai subduction margin

Surface heat flow has been observed to be highly variable in the Nankai subduction margin. This study presents an investigation of local anomalies in surface heat flows on the undulating seafloor in the Nankai subduction margin. We estimate the heat flows from bottom-simulating reflectors (BSRs) marking the lower boundaries of the methane hydrate stability zone and evaluate topographic effects on heat flow via two-dimensional thermal modeling. BSRs have been used to estimate heat flows based on the known stability characteristics of methane hydrates under low-temperature and high-pressure conditions. First, we generate an extensive map of the distribution and subseafloor depths of the BSRs in the Nankai subduction margin. We confirm that BSRs exist at the toe of the accretionary prism and the trough floor of the offshore Tokai region, where BSRs had previously been thought to be absent. Second, we calculate the BSR-derived heat flow and evaluate the associated errors. We conclude that the total uncertainty of the BSR-derived heat flow should be within 25%, considering allowable ranges in the P-wave velocity, which influences the time-to-depth conversion of the BSR position in seismic images, the resultant geothermal gradient, and thermal resistance. Finally, we model a two-dimensional thermal structure by comparing the temperatures at the observed BSR depths with the calculated temperatures at the same depths. The thermal modeling reveals that most local variations in BSR depth over the undulating seafloor can be explained by topographic effects. Those areas that cannot be explained by topographic effects can be mainly attributed to advective fluid flow, regional rapid sedimentation, or erosion. Our spatial distribution of heat flow data provides indispensable basic data for numerical studies of subduction zone modeling to evaluate margin parallel age dependencies of subducting plates.

The origin of strata within the inner accretionary prism of Nankai Trough: Evidence from clay mineral assemblages along the NanTroSEIZE transect

AbstractOne of the more prominent architectural elements of the Nankai subduction margin, offshore southwest Japan, is an out‐of‐sequence thrust fault (megasplay) that separates the inner accretionary prism from the outer prism. The inner prism (hanging wall of the megasplay) is dominated by mudstone, which is enigmatic when the sedimentary facies is compared to coeval deposits in the Shikoku Basin (i.e. inputs from the subducting Philippine Sea plate) and to coarser‐grained turbidite sequences from the Quaternary trench wedge. Clay mineral assemblages amplify the mismatches of sedimentary facies. Mudstones from the inner prism are uniformly depleted in smectite, with average bulk values of 23–24 wt%, whereas the Shikoku Basin deposits show progressive decreases in proportions of smectite over time, from averages of 46–48 wt% at 10 Ma to 17–21 wt% at 1 Ma. Plate‐boundary reconstructions for the Philippine Sea region provide one solution to the conundrum. Between 15 Ma and 10 Ma, the Pacific plate subducted near the NanTroSEIZE transect, and a trench‐trench‐trench triple junction migrated to the northeast. Accretion during that period involved sediments that had been deposited on the Pacific plate. Motion of the Philippine Sea plate changed from 10 Ma to 6 Ma, resulting in sinistral slip along the proto‐Nankai Trough. Sediments accreted during that period probably had been deposited near the triple junction, with a hybrid detrital provenance. Renewed subduction of the Philippine Sea plate at 6 Ma led to reorganization of watersheds near the Izu–Honshu collision zone and gradual incision of large submarine canyons on both sides of the colliding Izu arc. Accreted Pliocene mudstones share more of an affinity to the triple junction paleoenvironment than they do to Shikoku Basin. These differences between subducting Shikoku Basin strata and accreted Pacific plate sediments have important implications for interpretations of frictional properties, structural architecture, and diagenetic fluid production.

Elongate fluid flow structures: Stress control on gas migration at Opouawe Bank, New Zealand

High-resolution 2D and 3D seismic data from Opouawe Bank, an accretionary ridge on the Hikurangi subduction margin off New Zealand, show evidence for exceptional gas migration pathways linked to the stress regime of the ridge. Although the ridge has formed by thrusting and folding in response to a sub-horizontal principal compressive stress (σ1), it is clear that local stress conditions related to uplift and extension around the apex of folding (i.e. sub-vertical σ1) are controlling shallow fluid flow. The most conspicuous structural features are parallel and horizontally-elongated extensional fractures that are perpendicular to the ridge axis. At shallower depth near the seafloor, extensional fractures evolve into more concentric structures which ultimately reach the seafloor where they terminate at gas seeps. In addition to the ridge-perpendicular extensional fractures, we also observe both ridge-perpendicular and ridge-parallel normal faults. This indicates that both longitudinal- and ridge-perpendicular extension have occurred in the past. The deepest stratigraphic unit that we image has undergone significant folding and is affected by both sets of normal faults. Shallower stratigraphic units are less deformed and only host the ridge-parallel normal faults, indicating that longitudinal extension was limited to an older phase of ridge evolution. Present-day gas migration has exploited the fabric from longitudinal extension at depth. As the gas ascends to shallower units it ‘self-generates’ its flow pathways through the more concentric structures near the seafloor. This shows that gas migration can evolve from being dependent on inherited tectonic structures at depth, to becoming self-propagating closer to the seafloor.

Eclogites from the Marun-Keu Complex, Polar Urals, Russia: a record of hot subduction and sub-isothermal exhumation

Abstract Lenses and boudins of eclogitized and non-metamorphic gabbroic and ultramafic rocks occurring in quartzofeldspathic gneisses and migmatites of the Marun-Keu Complex, Polar Urals, are of central interest for the study of fluid-controlled processes in the subducting continental lithosphere. Petrology and mineral chemistry, combined with multi-equilibrium geothermobarometry and phase-equilibrium modelling, applied to mafic eclogitic rocks show a prograde evolution from c. 685°C, 1.89 GPa and a H 2 O = 0.6 to c. 745°C, 2.13 GPa and a H 2 O = 1 followed by near-isothermal decompression down to c. 1.63 GPa. The absence of pre-eclogitic H 2 O-bearing mineral inclusions in garnet and omphacite suggests that the primary magmatic rocks have not been affected by prograde metamorphism during subduction due to the lack of fluid. We suggest that the Marun-Keu Complex represents a coherent crustal block that experienced the same P – T evolution but provides a metamorphic record of different P – T parameters because of an uneven introduction of aqueous-bearing fluids during subduction and exhumation. The tectonometamorphic history of the Marun-Keu Complex is best attributed to the relatively hot regime of the continental margin subduction beneath an island arc, whereas the blueschist-facies part of the complex towards the north could be related to the closure of the Uralian Ocean.

Long-term seismic observations along Myanmar–Sunda subduction margin: insights for 2004 Mw > 9.0 earthquake

The present study investigates the temporal variation of few seismic parameters between the Myanmar (Zone I), Andaman–Nicobar–Northwest Sumatra (Zone II), Southeast Sumatra–West Indonesia (Zone III) and East Indonesia (Zone IV) converging boundaries in reference to the generation of 26 December 2004 Mw > 9.0 off-Sumatra mega-earthquake event. The four segments are distinguished based on tectonics parameters, distinct geological locations, great earthquake occurrences, and the Wadati–Benioff zone characteristics. Two important seismic parameters such as seismic energy and b values are computed over a time-window of 6-month period during the entire 1976–2013 period for these segments. The b values show a constant decrease in Zones II, III, and IV, whereas the Zone I does not show any such pattern prior to the 2004 mega-event. The release of seismic energy was also gradually decreasing in Zones II and III till the 2004 event, and little similar pattern was also noted in Zone IV. This distinct observation might be indicating that the stress accumulation was dominant near the Sumatra–Java area located towards southeast of Zone II and northwest of Zone III. The released strain energy during the 2004 event was subsequently migrated towards north, rupturing ~ 1300 km of the boundary between the Northwest Sumatra and the North Andaman. The occurrence of 2004 mega-event was apparently concealed behind the long-term seismic quiescence existing near the Sumatra and Nicobar margin. A systematic study of the patterns of seismic energy release and b values, and the long-term observation of collective behaviour of the margin tectonics might have had given clues to the possibility of the 2004 mega-event.

Mississippian lamprophyre dikes in western Sierras Pampeanas, Argentina: Evidence of transtensional tectonics along the SW margin of Gondwana

In the Famatina range, Sierras Pampeanas of Argentina (SW Gondwana), subvertical calc-alkaline lamprophyric dike swarms crop out through >300 km. The dikes cut Ordovician units with a prominent NW-SE trending and are covered by continental sedimentary successions of Pennsylvanian to Permian age. The dikes show a strong structural control associated with Riedel fault systems. Detailed field analysis suggested a ∼N-S opening direction oblique to the attitude of dike walls and a left-lateral transtensional tectonics during the emplacement. 40Ar/39Ar geochronology of a lamprophyric sample defined a crystallization age (plateau; whole rock) of 357.1 ± 7.1 Ma (MSWD = 2.3). Coetaneous ductile zones with dominant strike-slip motion, documented along western Argentina for >600 km, suggest a regional event in SW Gondwana during the Mississippian. We propose that this deformation was the result of the counterclockwise fast rotation of Gondwana between 365 and 345 Ma, when the Famatina range and western Argentina occupied a sub-polar position. A transform margin along SW Gondwana better explains our (and others) data rather than a subduction margin. This scenario is also consistent with the occurrence of A-type granites and normal-fault basins within the foreland as well as bimodal volcanics.

The role of protothrusts in frontal accretion and accommodation of plate convergence, Hikurangi subduction margin, New Zealand

The role of protothrusts in frontal accretion and accommodation of plate convergence, Hikurangi subduction margin, New Zealand

Sensitivities of Near‐field Tsunami Forecasts to Megathrust Deformation Predictions

AbstractThis study reveals how modeling configurations of forward and inverse analyses of coseismic deformation data influence the estimations of seismic and tsunami sources. We illuminate how the predictions of near‐field tsunami change when (1) a heterogeneous (HET) distribution of crustal material is introduced to the elastic dislocation model, and (2) the near‐trench rupture is either encouraged or suppressed to invert spontaneous coseismic displacements. Hypothetical scenarios of megathrust earthquakes are studied with synthetic Global Positioning System displacements in Cascadia. Finite‐element models are designed to mimic the subsurface heterogeneity across the curved subduction margin. The HET lithospheric domain modifies the seafloor displacement field and alters tsunami predictions from those of a homogeneous (HOM) crust. Uncertainties persist as the inverse analyses of geodetic data produce nonrealistic slip artifacts over the HOM domain, which propagates into the prediction errors of subsequent tsunami arrival and amplitudes. A stochastic analysis further shows that the uncertainties of seismic tomography models do not degrade the solution accuracy of HET over HOM. Whether the source ruptures near the trench also controls the details of the seafloor disturbance. Deeper subsurface slips induce more seafloor uplift near the coast and cause an earlier arrival of tsunami waves than surface‐slipping events. We suggest using the solutions of zero‐updip‐slip and zero‐updip‐slip‐gradient rupture boundary conditions as end‐members to constrain the tsunami behavior for forecasting purposes. The findings are important for the near‐field tsunami warning that primarily relies on the near‐real‐time geodetic or seismic data for source calibration before megawaves hit the nearest shore upon tsunamigenic events.

The tectonic setting and evolution of the 2.7 Ga Kalgoorlie–Kurnalpi Rift, a world-class Archean gold province

The Yilgarn Craton results from three major mantle input events (at ca 3.0–2.9, 2.8 and 2.7 Ga) that have interacted with > 3.0 Ga continental crust. Zircon geochronology and Sm-Nd isotopic data subdivide the craton into an older Yilgarn proto-craton and the younger, more primitive Eastern Goldfields Superterrane (EGST). Formation of the Kalgoorlie-Kurnalpi Rift (KKR) within the EGST was associated with the 2.7 Ga event, which exploited weakened crust at the eastern margin of the Yilgarn proto-craton where thick sequences of komatiite and basalt were erupted between ca 2710 and 2690 Ma in the Kalgoorlie Terrane. Calc-alkaline volcanism in the Kurnalpi Terrane began at ca 2730 Ma and continued to ca 2690 Ma, overlapping rifting and plume-related volcanism in the Kalgoorlie Terrane. Deposition of siliciclastic sedimentary rocks within basins at ca 2660 resulted from an intra-orogenic extensional event and coincided with the transition from High-Ca to Low-Ca granite magmatism and peak emplacement of intrusions with a metasomatised mantle source component. Most aspects of the KKR are satisfied by broadly coincident plume-related magmatism in the Kalgoorlie Terrane and westward subduction to the east of the Burtville Terrane. Geochemical characteristics of 2730–2700 Ma calc-alkaline volcanism and 2685–2630 Ma low-SiO2 and alkali-rich intrusions support models for a continental margin subduction zone setting. World-class gold deposits formed in reactivated margins of the KKR, which became flux zones for mantle-derived magmas, hydrothermal fluids and heat during 2675–2620 Ma orogenesis. The orogenic gold mineralisation can be subdivided into proximal intrusion-related and distal-source deposits.

A Fluid Pulse on the Hikurangi Subduction Margin: Evidence From a Heat Flux Transect Across the Upper Limit of Gas Hydrate Stability

AbstractA transect of seafloor heat probe measurements on the Hikurangi Margin shows a significant increase of thermal gradients upslope of the updip limit of gas hydrate stability at the seafloor. We interpret these anomalously high thermal gradients as evidence for a fluid pulse leading to advective heat flux, while endothermic cooling from gas hydrate dissociation depresses temperatures in the hydrate stability field. Previous studies predict a seamount on the subducting Pacific Plate to cause significant overpressure beneath our study area, which may be the source of the fluid pulse. Double‐bottom simulating reflections are present in our study area and likely caused by uplift based on gas hydrate phase boundary considerations, although we cannot exclude a thermogenic origin. We suggest that uplift may be associated with the leading edge of the subducting seamount. Our results provide further evidence for the transient nature of fluid expulsion in subduction zones.

Geochemistry and tectonic setting of the Golabad granitoid complex (SW Nain, Iran)

The Oligo-Miocene Golabad granitoid complex intrusive into the Eocene volcanic rocks occurs in the Urumieh-Dokhtar Magmatic Arc (UDMA) in Iran. According to microscopic and chemical studies, the granitoid complex consists of three different rock types: 1) plutonic rocks comprising diorite, quartz diorite, granodiorite and granite; 2) volcanic rocks composed of basalt, andesite basalt, ± pyroxene bearing andesite and rhyolite, and 3) pyroclastic rocks. The main mineral constituents of these rocks are mostly plagioclase (oligoclase and andesine), quartz, K-feldspar, amphibole (magnesio-hornblende and actinolite-hornblende) and Mg-biotite. In addition, apatite, titanite, zircon, and opaque minerals are common accessory minerals. The studied enclaves are classified as mafic micro-granular enclaves (MME) with monzodiorite compositions. Geochemically, the rocks in this study represent medium to high-K calc-alkaline series, metaluminous and I-type nature.Plotting, the chemical composition of plagioclase on the An-Ab-Or ternary diagram, the temperature of crystallization is estimated to range from 700 to 900 °C at a pressure of 4.5 Kbar. High TiO2 values of biotites from the Golabad granitoid complex suggest magmatic origin and the crystallization temperature is estimated to range from 700 to 750 °C. The amphiboles according to their chemical analysis, are classified as igneous amphiboles generated in high oxygen fugacity conditions. The chemical data of the amphiboles and biotites pointed out to the I- type nature of the Golabad granitoid complex emplaced in an active continental margin subduction setting. The amphibole crystallization pressure was estimated by Al in amphibole varies from 1.09 to 2.28 Kbar. Using the calculated pressure the depth of the formation of the Golabad granitoid complex estimated from 4 to 9 Km.

Structure of the (ultra)high-pressure Western Gneiss Region, Norway: Imbrication during Caledonian continental margin subduction

Structure of the (ultra)high-pressure Western Gneiss Region, Norway: Imbrication during Caledonian continental margin subduction

Gas Hydrate Formation Amid Submarine Canyon Incision: Investigations From New Zealand's Hikurangi Subduction Margin

AbstractWe investigate gas hydrate system dynamics beneath a submarine canyon on New Zealand's Hikurangi subduction margin using seismic reflection data and petroleum systems modeling. High seismic velocities just above the base of gas hydrate stability (BGHS) indicate that concentrated gas hydrates exist beneath the canyon. Two‐dimensional gas hydrate formation modeling shows how the process of canyon incision at this location alters the distribution and concentration of gas hydrate. The key modeling result is that free gas is trapped beneath the gas hydrate layer and then “captured” into a concentrated gas hydrate deposit as a result of a downward‐shift in the BGHS driven by canyon incision. Our study thus provides new insight into the functioning of this process. From our data, we also conceptualize two other models to describe how canyons could significantly change gas hydrate distribution and concentration. One scenario is related to deflection of fluid flow pathways from over‐pressured regions at the BGHS toward the canyon, and the other is based on relationships between simultaneous seafloor uplift and canyon incision. The relationships and processes described are of global relevance because of considerations of gas hydrate as an energy resource and the influence of both submarine canyons and gas hydrate systems on seafloor biodiversity.

How Do Tides and Tsunamis Interact in a Highly Energetic Channel? The Case of Canal Chacao, Chile

AbstractThis study aims at understanding the role of tidal level, speed, and direction in tsunami propagation in highly energetic tidal channels. The main goal is to comprehend whether tide‐tsunami interactions enhance/reduce elevation, currents speeds, and arrival times, when compared to pure tsunami models and to simulations in which tides and tsunamis are linearly superimposed. We designed various numerical experiments to compute the tsunami propagation along Canal Chacao, a highly energetic channel in the Chilean Patagonia lying on a subduction margin prone to megathrust earthquakes. Three modeling approaches were implemented under the same seismic scenario: a tsunami model with a constant tide level, a series of six composite models in which independent tide and tsunami simulations are linearly superimposed, and a series of six tide‐tsunami nonlinear interaction models (full models). We found that hydrodynamic patterns differ significantly among approaches, being the composite and full models sensitive to both the tidal phase at which the tsunami is triggered and the local depth of the channel. When compared to full models, composite models adequately predicted the maximum surface elevation, but largely overestimated currents. The amplitude and arrival time of the tsunami‐leading wave computed with the full model was found to be strongly dependent on the direction of the tidal current and less responsive to the tide level and the tidal current speed. These outcomes emphasize the importance of addressing more carefully the interactions of tides and tsunamis on hazard assessment studies.

Foreshocks and delayed triggering of the 2016 MW7.1 Te Araroa earthquake and dynamic reinvigoration of its aftershock sequence by the MW7.8 Kaikōura earthquake, New Zealand

We analyze the preparatory period of the September 2016 MW7.1 Te Araroa foreshock–mainshock sequence in the Northern Hikurangi margin, New Zealand, and subsequent reinvigoration of Te Araroa aftershocks driven by a large distant earthquake (the November 2016 MW7.8 Kaikōura earthquake). By adopting a matched-filter detection workflow using 582 well-defined template events, we generate an improved foreshock and aftershock catalog for the Te Araroa sequence (>8,000 earthquakes over 66 d). Templates characteristic of the MW7.1 sequence (including the mainshock template) detect several highly correlating events (ML2.5–3.5) starting 12 min after a MW5.7 foreshock. These pre-cursory events occurred within ∼1 km of the mainshock and migrate bilaterally, suggesting precursory slip was triggered by the foreshock on the MW7.1 fault patch prior to mainshock failure. We extend our matched-filter routine to examine the interactions between high dynamic stresses resulting from passing surface waves of the November 2016 MW7.8 Kaikōura earthquake, and the evolution of the Te Araroa aftershock sequence. We observe a sudden spike in moment release of the aftershock sequence immediately following peak dynamic Coulomb stresses of 50–150 kPa on the MW7.1 fault plane. The triggered increase in moment release culminated in a MW5.1 event, immediately followed by a ∼3 h temporal stress shadow. Our observations document the preparatory period of a major subduction margin earthquake following a significant foreshock, and quantify dynamic reinvigoration of a distant on-going major aftershock sequence amid a period of temporal clustering of seismic activity in New Zealand.

Recycling of palaeo-Pacific subducted oceanic crust related to a Fe–Cu–Au mineralization in the Xu-Huai region of North Anhui-Jiangsu, East China: Geochronological and geochemical constraints

ABSTRACTIn this study, Early Cretaceous skarn deposits and genesis of their host diorite/monzodiorite porphyry in the Xuzhou-Huaibei (Xu-Huai) region, northern Anhui-Jiangsu have been discussed by detailed geochemical work. In-situ zircon U–Pb dating of the diorites related to Fe–Cu–Au deposits shows that they were formed between 131.4 ± 1.5 Ma and 130.8 ± 1.8 Ma. Geochemical data indicate a depletion of high field strength elements (HFSE) in the diorite porphyry with similarity to that of arc-related igneous rocks. The diorite porphyry was probably derived from typical arc magmas related to continental margin subduction characterized by light rare earth elements (LREEs) enrichment and HFSE depletion. REEs compositions of apatite in the diorite porphyry indicate that the dioritic magma was produced from the metasomatized subcontinental mantle by slab-derived fluids. The magma was proven to be a high oxygen fugacity; thus, it was particularly conducive to the precipitation of Fe, Cu, Au and other ore-forming elements. The δ34S values of pyrite and chalcopyrite of Fe–Cu–Au ores range from −0.2‰ to 2.8‰, indicating that the sulphur in the ore was probably derived from deep-seated magmas. Integrated with geochronological and geochemical analyses, we suggest that the Early Cretaceous igneous suites associated with Fe–Cu–Au deposits in the Xu-Huai region are related to recycling subduction of Pacific oceanic crust.

Paleoarchean bedrock lithologies across the Makhonjwa Mountains of South Africa and Swaziland linked to geochemical, magnetic and tectonic data reveal early plate tectonic genes flanking subduction margins

The Makhonjwa Mountains, traditionally referred to as the Barberton Greenstone Belt, retain an iconic Paleoarchean archive against which numerical models of early earth geodynamics can be tested. We present new geologic and structural maps, geochemical plots, geo- and thermo-chronology, and geophysical data from seven silicic, mafic to ultramafic complexes separated by major shear systems across the southern Makhonjwa Mountains. All reveal signs of modern oceanic back-arc crust and subduction-related processes. We compare the rates of processes determined from this data and balance these against plate tectonic and plume related models. Robust rates of both horizontal and vertical tectonic processes derived from the Makhonjwa Mountain complexes are similar, well within an order of magnitude, to those encountered across modern oceanic and orogenic terrains flanking Western Pacific-like subduction zones. We conclude that plate tectonics and linked plate-boundary processes were well established by 3.2–3.6 Ga. Our work provides new constraints for modellers with rates of a ‘basket’ of processes against which to test Paleoarchean geodynamic models over a time period close to the length of the Phanerozoic.

Sediment underthrusting within a continental magmatic arc: Coast Mountains batholith, British Columbia

AbstractThough continental magmatic arcs are factories for new continental crust, a significant proportion of continental arc magmas are recycled from supracrustal material. To evaluate the relative contributions of retroarc underthrusting and trench side partial sediment subduction for introducing supracrustal rocks to the middle and lower crust of continental magmatic arcs, we present results from the deeply exposed country rocks of the Coast Mountains batholith of western British Columbia. Prior work demonstrates that these rocks underwent widespread partial melting that contributed to the Coast Mountains batholith. We utilize U‐Pb zircon geochronology, Sm‐Nd thermochronology, and field‐based studies to document the protoliths and early burial history of amphibolite and granulite‐facies metasedimentary rocks in the Central Gneiss Complex. U‐Pb detrital zircon data from the structurally highest sample localities yielded ~190 Ma unimodal age peaks and suggest that retroarc rocks of the Stikine terrane constitute a substantial portion of the Central Gneiss Complex. These supracrustal rocks underwent thrust‐related burial and metamorphism at >25 km depths prior to ~80 Ma. These rocks may also be underlain at the deepest exposed structural levels by Upper Cretaceous metasedimentary rocks, which may have been emplaced as a result of trench side underplating or intraarc burial. These results further our understanding of the mechanisms of material transport within the continental lithosphere along Cordilleran subduction margins.

Mapping subduction interface coupling using magnetotellurics: Hikurangi margin, New Zealand

AbstractThe observation of slow‐slip, seismic tremor, and low‐frequency earthquakes at subduction margins has provided new insight into the mechanisms by which stress accumulates between large subduction (megathrust) earthquakes. However, the relationship between the physical properties of the subduction interface and the nature of the controls on interplate seismic coupling is not fully understood. Using magnetotelluric data, we show in situ that an electrically resistive patch on the Hikurangi subduction interface corresponds with an area of increased coupling inferred from geodetic data. This resistive patch must reflect a decrease in the fluid or sediment content of the interface shear zone. Together, the magnetotelluric and geodetic data suggest that the frictional coupling of this part on the Hikurangi margin may be controlled by the interface fluid and sediment content: the resistive patch marking a fluid‐ and sediment‐starved area with an increased density of small, seismogenic‐asperities, and therefore a greater likelihood of subduction earthquake nucleation.

Complex rupture process of the Mw 7.8, 2016, Kaikoura earthquake, New Zealand, and its aftershock sequence

The M7.8 Kaikoura Earthquake that struck the northeastern South Island, New Zealand, on November 14, 2016 (local time), is one of the largest ever instrumentally recorded earthquakes in New Zealand. It occurred at the southern termination of the Hikurangi subduction margin, where the subducting Pacific Plate transitions into the dextral Alpine transform fault. The earthquake produced significant distributed uplift along the north-eastern part of the South Island, reaching a peak amplitude of ∼8 m, which was accompanied by large (≥10 m) horizontal coseismic displacements at the ground surface along discrete active faults. The seismic waveforms' expression of the main shock indicate a complex rupture process. Early automated centroid moment tensor solutions indicated a strong non-double-couple term, which supports a complex rupture involving multiple faults. The hypocentral distribution of aftershocks, which appears diffuse over a broad region, clusters spatially along lineaments with different orientations. A key question of global interest is to shed light on the mechanism with which such a complex rupture occurred, and whether the underlying plate-interface was involved in the rupture. The consequences for seismic hazard of such a distributed, shallow faulting is important to be assessed. We perform a broad seismological analysis, combining regional and teleseismic seismograms, GPS and InSAR, to determine the rupture process of the main shock and moment tensors of 118 aftershocks down to Mw 4.2. The joint interpretation of the main rupture and aftershock sequence allow reconstruction of the geometry, and suggests sequential activation and slip distribution on at least three major active fault domains. We find that the rupture nucleated as a weak strike-slip event along the Humps Fault, which progressively propagated northward onto a shallow reverse fault, where most of the seismic moment was released, before it triggered slip on a second set of strike-slip faults at the northern end of the rupture. The northern and southern strike-slip fault domains have the same orientation but are spatially separated by >15 km. In our model, the low angle splay thrust fault is located above the slab and connects the strike-slip faults kinematically. During the aftershock phase, the entire fault system remained active.