V23I-01 INVITED
Reconciling Process Variability and Configuration Uniformity of Subduction Zones
Subduction zone processes vary dramatically with the age of the subducting slab and hence the thermal state. In subduction zones of young and warm slabs, such as Cascadia and Nankai, intraslab earthquakes terminate at rather shallow depths, and arc volcanism is feeble. In contrast, subduction zones of old and cold slabs, such at NE Japan and Kermadec, intraslab earthquakes extend to large depths and arc volcanism is strong. However, despite these large variations, the configuration of subduction zones is surprisingly uniform. Regardless of the slab age and thermal state, the depth of the dipping slab is 100±20 km beneath the volcanic arc. Towards reconciling the striking variability and uniformity, we have developed numerical thermal models including viscous mantle wedge flow for many subduction zones having different slab ages, convergence rates, and slab geometry. Where there are adequate surface heat flow constraints, such at Cascadia, NE Japan, and Kamchatka, we find that the heat flow can be explained only if the interface between the slab and the mantle wedge is decoupled to a depth of 70-80 km. The deeper part of the interface is coupled, resulting in a slab-driven "corner flow" similar to what is seen in all thermal models of this spatial scale. A common depth of decoupling of 70-80 km not only explains the heat flow data but also reconciles the variability and uniformity of subduction zones. The abrupt decoupling-coupling transition due to the temperature and stress-dependent non-linear mantle rheology gives rise to a sharp transition in the flow and thermal regimes of the mantle wedge. Overlying the decoupled part of the interface, the wedge is stagnant and cold, with the exact temperature depending on the age of the slab. Overlying the coupled part of the interface, the mantle wedge flows at full speed, and the circulation brings in hot material from greater depths, resulting in temperatures of 1220-1340°C regardless of the age of the slab. These temperatures are considered adequate for magma generation, provided that there is sufficient fluid supply from the dehydrating slab to trigger melting. The models predict that slab dehydration peaks in the 80-140 km depth range in most subduction zones. The requirement of both high temperature and ample fluids explains the spatial limitation of volcanic arcs and thus the similar configuration of different subduction zones. The models also predict that very warm slabs dehydrate at shallow depths and mostly in the forearc region, explaining the feeble arc volcanism, shallow intraslab earthquakes, strong evidence for serpentinization of the forearc mantle wedge, and perhaps the occurrence of non-volcanic tremor and slow slip events in these subduction zones.
V23I-02
Sub Moho boundary beneath island arc, Japan
Moho is a seismic boundary discovered by Mohorovicic in 1909. It is generally considered as the boundary between Earthfs crust and mantle, although the origin of the Moho and the definition of the crust-mantle boundary are still debated. Regardless whether it represents a phase or chemical boundary, it has been investigated world wide by seismologists since its discovery. Reflection/refraction data acquired from active sources were the main sources of the investigation. Recently, receiver function analysis provided an emerging tool for mapping this boundary below a seismic station. Since reflection/refraction data sample P- wave structure while receiver-function data are basically sensitive to S-wave velocity, a comparison of the two can provide better constraints on the composition of the lowermost crust and the uppermost mantle. In this study, we used receiver function data to determine subsurface boundaries. Receiver functions are calculated from teleseismic records of a borehole seismic network in Japan, the Hi-net, which covers the entire Japan with a 20-km station spacing. We used the multi-taper deconvolution method to generate individual receiver functions and a common-conversion-point gathering method for stacking the receiver functions. Results along several profiles where seismic refraction surveys have been conducted in the past were selected for comparison. We found remarkable difference between southwestern and central/northeastern Japan. In southwestern Japan both receiver function and refraction data yield a consistent and distinct Moho at around 35 km. In central and northeastern Japan, however, receiver function images reveal two P-to-S conversion events at approximately 30 km and 40 km, respectively. The shallow 30-km deep boundary is consistent with the refraction Moho. The deep boundary (sub-Moho) is about 10 km deeper than the refraction Moho boundary. The uppermost mantle seismic structure at central and northeastern Japan was known to be reflective to seismic waves as compared to the continental ones which are relatively transparent. Locations of the observed sub-Moho appear to correlate well with the active volcanic front, suggesting that the sub-Moho might be formed as the result of arc magmatism.
V23I-03
Seismological constraints on crustal formation processes in the Izu-Bonin intra-oceanic subduction zone
In order to investigate formation, deformation and alternation processes of arc crust in the Izu-Bonin intra- oceanic subduction zone where the old Pacific plate is subducting beneath the young Philippine Sea plate, JAMSTEC has acquired large amount of active-source seismic data to cover the entire arc system. We found several new seismological constraints on formation processes of arc crust from the seismic data along the arc. For examples, along the current volcanic front which is believed to preserve an entire crustal formation process since a steady-state plate subduction has been occurred in the Oligocene age, crust of continental composition having Vp of 6 - 6.8 km/s has been predominantly generated beneath basaltic volcanic center. In the rear-arc region which is proposed to be formed by rifting of the arc crust, we discovered a seismological evidence indicating a paleo-arc structure. A variation pattern of the volume of continental component of crust along the rear arc is similar to that we found in the present-day volcanic front. It is also deduced, from a comparison between the structural variation pattern along the current arc and the rear-arc, that the rifting direction (NNE-SSW) was highly oblique to the current plate convergent direction. Petrological studies in the forearc region demonstrated that magmatism of an early stage of an arc generation is characterized by the boninitic volcanism rather than the basaltic volcanism observed in the current volcanic front. Data from a recent active-source seismic survey shows a unique seismic character along the forarc; i.e., layers having crustal seismic velocity is remarkably thin (less than 10 km) even under the large forearc ridge (the Bonin ridge) and the thin crustal-velocity layers are observed only in the center of the forearc ridge. The seismic image is well corresponds to the gravity data showing a north-south trend of a high gravity anomaly along the forearc ridge. This seismological observation suggests that the crust at the center of the forearc ridge formed by the boninitic volcanism in the early stage of subduction have generated the thin crustal-velocity layers, and this structure only extends in the narrow region along the forearc.
V23I-04
Seismogenic Coupling at Convergent Margins Geophysical Observations from the South American Subduction Zone and the Alpine Rock Record
Convergent continental margins are the Earth's principal locus of important earthquake hazards with nearly all interplate megathrust earthquakes (M>8) in the seismogenic coupling zone between the converging plates. Despite the key importance of this zone, the processes that shape it are poorly understood. This is underscored by a number of novel observations attributed to processes in the interface zone that are attracting increasing attention: silent slip events, non-volcanic tremors, afterslip, locked patches embedded in a creeping environment, etc. We here compare the rock record from a field study with recent results from two major geophysical experiments (ANCORP and TIPTEQ) that have imaged the South Chilean subduction zone at the site of the largest historically recorded earthquake (Valdivia, 1969; Mw = 9.5) and the plate boundary in Northern Chile, where a major seismic event is expected in the near future (Iquique segment). The reflection seismic data exhibit well defined changes of reflectivity and Vp/Vs ratio along the plate interface that can be correlated with different parts of the coupling zone as well as with changes during the seismic cycle. Observations suggest an important role of the hydraulic system. The rock record from the exhumed Early Tertiary seismogenic coupling zone of the European Alps provides indications for the mechanisms and processes responsible for the geophysical images. Fabric formation and metamorphism in a largely preserved subduction channel chiefly record the deformation conditions of the pre-collisional setting along the plate interface. We identify an unstable slip domain from pseudotachylytes occurring in the temperature range between 200-300°C. This zone coincides with a domain of intense veining in the subduction mélange with mineral growth into open cavities, indicating fast, possibly seismic, rupture. Evidence for transient near-lithostatic fluid pressure as well as brittle fractures competing with mylonitic shear zones continues into the region below the occurrence of pseudotachylytes, possibly reflecting a zone of conditionally stable slip. This latter zone is characterized by a pervasive fabric in the subduction mélange with elongated clasts that may well contribute to a layered reflection image. The zone above the unstable slip area has a chaotic appearance and is devoid of veins, but displays ample evidence of fluid-assisted processes similar to the deeper zone: solution-precipitation creep and dehydration reactions in the mélange matrix, hydration and sealing of the base of the upper plate. Seismic rupture here is possibly expressed by ubiquitous localized deformation zones. We hypothesize that trenchward sealing of parts of the plate interface as well as reaction-enhanced destruction of upper plate permeability is an important component localizing the unstable slip zone. Temporal variations in the hydraulic system are probably related to the seismic cycle and may be candidates for the geophysical observations identified.
V23I-05
Magnetotelluric Imaging of Fluids and Melts: Examples from the Chilean and the Costa Rican Margins
Long-period magnetotellurics is one of the methods of choice if large volumes of fluids and melts shall be detected in the deep crust and upper mantle in a subduction setting. However, upper-crustal conductivity anomalies frequently obscure resolution at depth. Nevertheless, under favorable conditions, even the asthenospheric wedge may resolved as a highly-conductive feature. In this contribution I'll present case studies from two key tectonic regions, the Andean plateau of North Chile and Bolivia, and the Central American subduction zone in Costa Rica. Both subduction settings are very different considering crustal thickness, age of oceanic plate, etc., but several common electrical features could be derived. The main finding in the southern Altiplano plateau is a huge crustal conductor, interpreted as an image of large-scale silicic melting of the crust. Where this conductor is missing, as in the northern Altiplano, the asthenospheric wedge is resolved as a highly-conductive structure, but surprisingly not beneath the arc volcanoes, but rather beneath the backarc, i.e, the center of the plateau. This may hint as an emerging shift in the position of the volcanic arc as has occurred frequently in Andean history. In Costa Rica, too, the backarc is generally more conductive along a recently conducted traverse, but the asthenospheric wedge is not resolved due to a number of crustal anomalies, e.g., the southern boundary of the Chortis block, which underlies much of the Central American isthmus. Further anomalies hint at a large fluid input into the continental crust in the forearc. A common feature of both study areas is a missing conductor beneath the volcanic arc itself.
V23I-06
Where Does Subducted Water Go? A Summary
Large amounts of water are carried down in subduction zones, about one world ocean per billion years. Most appears to be driven upward into the overlying upper mantle and crust. A major consequence is the volatile-rich volcanic arc magmas, but this may represent only ~10 percent of the total water. The downdip rates of fluid expulsion can be estimated from numerical thermal models and dehydration P-T phase diagrams. For subduction of young hot oceanic lithosphere, most expulsion is beneath the forearc and little reaches the arc. For old cold subduction most is deeper beneath the backarc. From seismic velocity data landward of the arc, most of the fluid is concluded to be trapped in hydrating forearc mantle serpentinite. The ~30 percent serpentinization in the Cascadia forearc mantle approximately represents the total amount of water released from the downgoing plate under this region in the most recent ~43 Ma subduction phase. At shallower depths, from seismic data and exhumed sections, there also is evidence in the forearc crust for silica deposited from the rising fluids. This is the region of ETS seismic tremor which could result from quartz vein formation by mobile hydrofracture emplacement. Further landward, most backarcs are uniformly very hot, 800-900C at the Moho and thin ~60 km lithospheres, not just the arc (500-1000 km wide hot 'backarc' for Cordillera). The high temperatures are interpreted to be due to the upward input of water that reduces the viscosity of the upper mantle and facilitates vigorous shallow convection.
V23I-07
Seismic anisotropy of subduction zone minerals - contribution of hydrous phases
The seismology is the most effective method to explore the structure of subduction zones to great depth. The distinguishing feature of the mantle in the subduction regions is the presence of hydrated phases, which transport water into the Earth's interior and release it with dramatic local consequences, triggering earthquakes and melting. The seismological detection of these hydrous phases and geodynamic interpretation of flow in the hydrous mantle depend on knowledge of the anisotropic elastic properties and the characteristics of the wave propagation in anisotropic media. We present the current knowledge of the anisotropic seismic properties of hydrous minerals in the upper mantle, transition zone, and lower mantle that are stable along low temperature geotherms associated with subduction, and identify which minerals are likely to influence seismological observations because they have very high volume fractions, or very high anisotropies, or both of these. In the upper mantle antigorite and talc are exceptionally anisotropic and chlorite is also very anisotropic for S-waves. In the transition zone the major phases hydrous wadsleyite and ringwoodite have moderate to weak anisotropies. The DHMS phase D at 24 GPa is the only hydrous phase that can transport hydrogen from the transition zone into the lower mantle. The phase D is more anisotropic for S than P waves like many of the hydrous phases. From these data it is clear that hydrous phases are in general very anisotropic. However, pressure can play a strong role in reducing anisotropy. It is the case for talc, in which increasing pressure from ambient to 4 GPa reduces the anisotropy by about 50% for both P and S waves. In contrast, the anisotropy of the DHMS phase A does not change significantly with pressure. Our picture of the seismic anisotropy of hydrated minerals remains incomplete; the elastic properties of many have not been measured even at ambient conditions (e.g. phase E) or not measured in their true elastic symmetry (e.g. clinochlore). The majority of hydrous minerals have not been measured at high pressure, and none related to hydrated mantle have been measured at elevated temperature. We have shown in the few cases where hydrous minerals have been measured as a function of pressure, that this variable has an important effect on the velocity distribution and in most cases reduces the degree of anisotropy, hence we would expect seismic anisotropy to play a key role in the determination of the shallow structure of subduction zones in the upper mantle.
V23I-08
Channelised melt flow in downwelling mantle: implications for 226Ra-210Pb disequilibria in arc magmas
We present the results of an analytical model of porous flow of viscous melt into a steadily dilating 'channel' (defined as a cluster of dilating veins) in downwelling subarc mantle. The model predicts the pressure drop in the mantle wedge matrix surrounding the channel needed to drive melt flow as a function of position and time. Melt is sucked towards the dilatant region at a near-constant velocity (10-5 s-1) until veins comprising the channel stop opening. Our results make it possible to calculate the region of influence sampled by melt that surrounds the channel. This region is large compared to the model size of the channelized region driving flow. For a baseline dilation time of 1 yr, and channel half-width of 2 m, melt can be extracted from an 80 m radius and has the opportunity to sample matrix material with potentially contrasting chemistry on geologically short timescales. Our mechanical results are consistent with a downgoing arc mantle wedge source region where melting and melt extraction by porous flow to a channel network are sufficiently rapid to preserve source-derived 238U-230Th-226Ra, and potentially also 226Ra-210Pb, disequilibria, prior to magma ascent to the surface. Since this is the rate- determining step in the overall process, it allows the possibility that such short-lived disequilibria measured in arc rocks at the surface are derived from deep in the mantle wedge.