ALBERT

Description: We estimate fluid sources around a subducted seamount along the northern Hikurangi subduction margin of New Zealand, using thermomechanical numerical modelling informed by wedge structure and porosities from multichannel seismic data. Calculated fluid sources are input into an independent fluid-flow model to explore the key controls on overpressure generation to depths of 12 km. In the thermomechanical models, sediment transport through and beneath the wedge is calculated assuming a pressure-sensitive frictional rheology. The change in porosity, pressure and temperature with calculated rock advection is used to compute fluid release from compaction and dehydration. Our calculations yield more precise information about source locations in time and space than previous averaged estimates for the Hikurangi margin. The volume of fluid release in the wedge is smaller than previously estimated from margin-averaged calculations (~14 m 3  yr –1  m –1 ), and is exceeded by fluid release from underlying (subducting) sediment (~16 m 3  yr –1  m –1 ). Clay dehydration contributes only a small quantity of fluid by volume (~2 m 3  yr –1  m –1 from subducted sediment), but the integrated effect is still significant landward of the seamount. Fluid source terms are used to estimate fluid pressures around a subducting seamount in the fluid-flow models, using subducted sediment permeability derived from porosity, and testing two end-members for décollement permeability. Models in which the décollement acts as a fluid conduit predict only moderate fluid overpressure in the wedge and subducting sediment. However, if the subduction interface becomes impermeable with depth, significant fluid overpressure develops in subducting sediment landward of the seamount. The location of predicted fluid overpressure and associated dehydration reactions is consistent with the idea that short duration, shallow, slow slip events (SSEs) landward of the seamount are caused by anomalous fluid pressures; alternatively, it may result from frictional effects of changing clay content along the subduction interface.

Description: Earthquakes far from plate boundaries are poorly understood, because of few well-studied examples and uncertainty of what controls the location of these events. In 1969 a damaging local magnitude ( M L ) 6.3 strike-slip earthquake, with no surface expression, occurred in the Ceres–Tulbagh region in the South African stable continental interior. Here, we present a microseismic study of the Ceres–Tulbagh area, conducted over three months in 2012, in which 172 events recorded on at least three stations follow a Gutenberg–Richter relationship for –1.5〈 M L 〈0.5. The events delineate a 5 km wide, subvertical zone that is microseismically active to a depth of 15 km. This fault zone is subparallel to the 1969 aftershock zone and at low angle to the regionally inferred greatest horizontal stress. We argue that the microseismically active zone is guided by inherited structures in the basement geology. This and similar structures may represent significant earthquake hazard in plate interiors. Online Material: Tables of station and event parameters, and P - and S -wave arrival times.

Description: The relationship between folding and faulting in the Cape Fold Belt has been raised as an enigma. The mineral deformation mechanisms accommodating folding are integral to the relation between faults and folds. Here I discuss field and microstructural observations of folded rocks from the Laingsburg region in the Western Cape, and the underlying mineral deformation mechanisms accommodating strain in these rocks. In competent units, deformation was dominantly accomplished by flexural slip faulting and jointing; in relatively incompetent layers macroscale flow was accommodated by dissolution-precipitation creep and distributed cataclasis. Combined with previous studies suggesting deformation occurred at lowermost greenschist facies temperatures, these observations indicate that folding in the Cape Fold Belt occurred at temperatures and pressures within the normally frictional regime. Folding and thrust fault development therefore generally occurred concurrently, and partitioning between localised and distributed deformation was governed by factors such as fluid pressure conditions, strain rate, and relative viscosity.

Description: We document a correlation between along-strike variation in interseismic coupling depth on subduction interfaces, and the upper plate tectonic stress state in New Zealand, Vanuatu, and southwest Japan. Deep interseismic coupling occurs where the upper plate stress regime is contractional to transpressional, whereas a shallowing of interseismic coupling occurs where there is an along-strike shift to back-arc or intra-arc extension. To explain this relationship, we draw on theoretical studies suggesting that the fluid pressure state within the upper plate and on the subduction interface has a strong control on the depth of the transition from frictional to viscous behavior. Lower fluid pressures (e.g., close to hydrostatic) are expected where the over-riding plate is undergoing tectonic extension, whereas higher fluid pressures (e.g., close to lithostatic) are expected where the over-riding plate experiences long-term (e.g., 〉10 5 yr) tectonic shortening. Low fluid pressures within the upper plate may lead to a shallow frictional to viscous transition compared to an upper plate that is highly overpressured. We hypothesize that the state of tectonic stress and structural permeability in the upper plate are yet other variables to consider when evaluating which physical mechanisms control interseismic coupling of subduction megathrusts.

Description: 〈span〉Geophysical observations show spatial and temporal variations in fault slip style on shallow subduction thrust faults, but geological signatures and underlying deformation processes remain poorly understood. International Ocean Discovery Program (IODP) Expeditions 372 and 375 investigated New Zealand's Hikurangi margin in a region that has experienced both tsunami earthquakes and repeated slow-slip events. We report direct observations from cores that sampled the active Pāpaku splay fault at 304 m below the seafloor. This fault roots into the plate interface and comprises an 18-m-thick main fault underlain by ~30 m of less intensely deformed footwall and an ~10-m-thick subsidiary fault above undeformed footwall. Fault zone structures include breccias, folds, and asymmetric clasts within transposed and/or dismembered, relatively homogeneous, silty hemipelagic sediments. The data demonstrate that the fault has experienced both ductile and brittle deformation. This structural variation indicates that a range of local slip speeds can occur along shallow faults, and they are controlled by temporal, potentially far-field, changes in strain rate or effective stress.〈/span〉

Description: We have determined the major element composition, 18 O and D values, and water content of impact-related granophyre, and pseudotachylite, from various Vredefort Dome localities, aiming to constrain the mechanism of melt formation and the relationship between pseudotachylite and granophyre. The granitoid gneisses and the pseudotachylites they host have almost identical average D and 18 O values (–67 and 8.6, and –67 and 8.4, respectively). The water contents of the pseudotachylites are extremely low, consistent with the isolation of the pseudotachylites from free water during and since their formation. There is a bimodal distribution of water content in pseudotachylites, with one group averaging 0.28 ± 0.03 weight % (n = 9) and the other 0.59 ± 0.06 weight % (n = 9). The Vredefort granophyre, which has been interpreted as the pooled product of impact melting, has average D and 18 O values of –69 and 7.6, respectively (n = 2) and also has a very low water content (0.23 weight %). Differences in major element and O-isotope composition between the granophyre and the pseudotachylites are not consistent with a simple relationship, but can be explained by a higher component of greenstone in the granophyre. A strong correlation between host and pseudotachylite 18 O values is consistent with a system where the melt composition is controlled by the immediate surroundings. The pseudotachylites with higher water content have slightly higher 18 O values (9.1 compared to 8.1). This is opposite to the relationship predicted if water content is related to the proportion of biotite entering the melt. It is possible, instead, that this relates to the incorporation of higher proportions of material altered at low temperature in the high-water group.

Description: Subduction of mid-ocean ridges is a common feature in recent convergent margins, but is rarely documented in Proterozoic to Paleozoic orogenic belts. Here we describe evidence for ridge-trench interaction in the deeply eroded late Neoproterozoic Damara orogenic belt, central Namibia. The earliest interaction is indicated by primary intrusive contacts between amphibolite facies mid-ocean ridge metabasalts and trench metasediments. U-Pb zircon ages of 550–540 Ma from syntectonic granites in the forearc indicate the timing of partial melting and mafic underplating of the prism in response to ridge subduction. The thermal peak in the Damara belt, associated widespread granitic and alkalic plutonism, and hydrothermal activity coincide with the waning stages of tectonism at 530–520 Ma and are interpreted to indicate slab window widening and slab delamination. We suggest that the proposed two-stage thermal evolution of the Damara belt, comprising latest Neoproterozoic ridge subduction and early Cambrian slab delamination, represents a fingerprint of ridge subduction in ancient orogens.

Description: The mechanism and evolution of fault linkage is important in the growth and development of large faults. Here we investigate the role of coseismic stress changes in shaping the hard links between parallel normal fault segments (or faults), by comparing numerical models of the Coulomb stress change from simulated earthquakes on two en echelon fault segments to natural observations of hard-linked fault geometry. We consider three simplified linking fault geometries: (1) fault bend, (2) breached relay ramp, and (3) strike-slip transform fault. We consider scenarios where either one or both segments rupture and vary the distance between segment tips. Fault bends and breached relay ramps are favored where segments underlap or when the strike-perpendicular distance between overlapping segments is less than 20% of their total length, matching all 14 documented examples. Transform fault linkage geometries are preferred when overlapping segments are laterally offset at larger distances. Few transform faults exist in continental extensional settings, and our model suggests that propagating faults or fault segments may first link through fault bends or breached ramps before reaching sufficient overlap for a transform fault to develop. Our results suggest that Coulomb stresses arising from multisegment ruptures or repeated earthquakes are consistent with natural observations of the geometry of hard links between parallel normal fault segments. ©2017. American Geophysical Union. All Rights Reserved.