Earth and Life Processes Discovered from Subseafloor Environments

Harold Tobin , ... Elizabeth Screaton , in Developments in Marine Geology, 2014

4.4.1.4 Future Directions

The science plan for the 2013–2023 International Ocean Discovery Program (IODP ) prioritizes several aspects of subduction zone research. Given the past decade's history of destructive megathrust earthquakes and tsunami, understanding the mechanisms that control the occurrence of destructive earthquakes, landslides, and tsunami will certainly continue to drive future subduction zone science. Completion of the work started via NanTroSEIZE and CRISP would help achieve goals of understanding fault zone processes. In particular, realizing the goal of borehole sampling of the deep (>2–3  km) plate boundary fault zones remains high in priority (CHIKYU+10 Steering Committee, 2013; Science Plan Writing Committee, INVEST Steering Committee, & Additional Contributors, 2011). At the time of this writing, plans are active for advancing NanTroSEIZE Site C0002 from 3000 mbsf to ∼5200 mbsf, as a flagship project for Chikyu, with coring and LWD logging of the main fault zone (Figure 4.4.1.8), and casing that main plate boundary fault for installation of a long-term borehole monitoring package (see e.g., Tobin, Hirose, Saffer, & Expedition 348 Shipboard Scientific Party, 2014). CRISP B has also been proposed as one of the flagship projects for the Chikyu platform, but as of 2014, a decision whether or not drilling will go ahead and when has not yet been reached.

Figure 4.4.1.8. Detail of seismic depth section (Figure 4.4.1.2) in the region of NanTroSEIZE Deep riser Site C0002, illustrating casing installed in the borehole on Expeditions 326, 338, and 348, and planned future drilling target across the main high-amplitude plate boundary (so called megasplay) reflector and into its footwall. The NanTroSEIZE science team proposes to drill to ∼5200   m below sea floor, first in a logging-while-drilling hole, then in a second sidetrack hole to collect cores from above, within, and below the main fault zone as interpreted from this reflection.

Because the seismogenic cycle in subduction zones is decades to centuries in duration, a key strategy for future efforts will be to investigate different plate boundaries that are at different points in the megathrust event seismic cycle (see, e.g., Wang, Hu, & He, 2012). Further investigations of the Japan Trench would address, in the near-term postseismic period, questions on the mechanisms of the surprising and devastating 2011 Tohoku earthquake. Another expedition, now scheduled for 2016, would examine the inputs to the Sumatra 2004 M9 nucleation region, also in the early postseismic phase. Both of these complement the Nankai and Costa Rica efforts, which have focused on systems that are at mid- to late stages in the presumed seismic cycle.

Recent advances in broadband seismology and continuous GPS geodetic monitoring have revolutionized our understanding of fault displacement, showing that it can occur across a much wider spectrum of time scales than previously appreciated (e.g., Beroza & Ide, 2011; Obara, 2002; Schwartz & Rokosky, 2007; Wallace & Beavan, 2010; Wallace, Beavan, Bannister, & Williams, 2012, and many others). Slow slip and VLF earthquake events (Ito & Obara, 2006; Sugioka et al., 2012) in particular, those known to occur at shallow depths offshore in several subduction zones including the Hikurangi, Mexico, and Nankai margins, may elucidate fault behavior in the transitional state between genuinely aseismic and fully seismogenic conditions, and are promising targets for future drilling projects, for example, as proposed for the north Hikurangi margin of New Zealand.

Additional high priority scientific goals in the new IODP Science Plan include understanding how subduction zones initiate and how new continental crust is generated. This will be investigated through drilling of regions in incipient or initial stages of subduction, examination of ash as a recorder of arc volcanism, and through drilling transects across regions of ancient and active subduction. Another subduction zone target is the outer trench environment, where normal faulting is suspected to allow fluid influx and mantle serpentinization.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780444626172000207

Orogenic Garnet Peridotites

Cong Zhang , ... Lifei Zhang , in Ultrahigh-Pressure Metamorphism, 2011

16.2.2 Subduction Zone Garnet Peridotite

Subduction zone garnet peridotite acquired their garnet-bearing mineral assemblage as a result of prograde subduction postdating the age of their low-pressure mineral assemblage. The protolith of subduction zone garnet peridotite can be either produced by differentiation from mafic magma of lower crustal affinity or are serpentinite, plagioclase or spinel peridotite formed at shallow depths in the lithosphere. Typically, this kind of garnet peridotite may be interlayered with eclogite of various compositions that share an evolution similar to other components of the subducting crustal slab. Some subduction zone garnet peridotites contain well-preserved prograde, lower-pressure mineral assemblages as inclusions in HP and UHP phases. For example, inclusions of sapphirine, corundum, clinochlore, and amphibole occur in garnet porphyroblasts from the Maowu area of the Dabie orogen in China ( Okay, 1993; Liou & Zhang, 1998). The Fe–Ti type of garnet peridotite in the WGR also belongs to this type (Carswell et al., 1983; Jamtveit, 1987; Vrijmoed et al., 2006). The age of the garnet-bearing assemblage formed in a subduction zone garnet peridotite is always approximately equal to the age of the collision event leading to subduction. The restricted P–T conditions recorded by the mineralogy of subduction zone garnet peridotite: can be used to discriminate the prograde subduction zone type from a syncollisional mantle wedge garnet peridotite type (Figure 16.2B2 and D2 ) which, by definition, must have had a much higher P–T evolution.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780123851444000151

Volume 2

Gautam Sen , Robert J. Stern , in Encyclopedia of Geology (Second Edition), 2021

Abstract

Subduction zones are where Earth's tectonic plates dive back into the mantle, at rates of a few to several centimeters per year. These are key features of Earth's plate tectonic regime. An oceanic trench shows where the plate disappears, and a dipping zone of earthquakes show where the subducting plate is. Subduction zones are where Earth's deepest (~  700   km) and strongest earthquakes (Magnitude ~   9) occur. Plates sinking in subduction zones pull the tectonic plates and provide most of the driving force for plate tectonics. The subducting plate is oceanic lithosphere and carries water and sediments down into the mantle where these interact to cause melting in the overlying mantle "wedge" and a line of volcanoes about 100   km above the downgoing plate; these water-rich magmas produce Earth's most violent eruptions. This "subduction factory" produces continental crust and ore deposits. Subduction can continue indefinitely as long as oceanic plate is involved but collision occurs if continents are subducted. A video about subduction zones can be viewed at https://www.youtube.com/watch?v=6wJBOk9xjto

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780081029084000862

The Crust

G.E. Bebout , in Treatise on Geochemistry (Second Edition), 2014

Abstract

Subduction zones are avenues for the delivery of crustal, atmospheric, and oceanic (including organic) components to the mantle and understanding of subduction zone chemical and isotopic cycling is central to many models of crust–mantle–atmosphere evolution. During subduction, diagenetic and metamorphic reactions and related geochemical effects can profoundly influence the element inventory and isotopic composition of subducting slabs to depths beneath volcanic arcs and as they enter the deeper mantle. Physical juxtaposition of chemically disparate rocks and the generation and mobility of various fluids lead to myriad metasomatic effects in subduction zones. Larger scale manifestations of such processes include the mass transfer that leads to arc magmatism and convergent margin volatile cycling. Subduction zone metasomatism is initiated at very shallow levels, as oceanic slabs entering trenches bend and are infiltrated by seawater, and as sedimentary sections experience physical compaction, fluid expulsion, and diagenetic alteration. Studies of forearc fluid geochemistry track this shallow-level metasomatic alteration, and high- and ultrahigh-pressure metamorphic rock suites provide records of fluid generation and flow, and related metasomatism, to depths exceeding those beneath volcanic fronts. Subduction zone processes act as a geochemical filter, altering the compositions of deeply subducting rocks and generating outputs such as arc magmas (and associated volcanic gases) and chemically and isotopically modified rocks. The latter enter the deeper mantle and influence its long-term geochemical evolution.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780080959757003223

Volume 2

Yong-Fei Zheng , in Encyclopedia of Geology (Second Edition), 2021

Conclusions

Subduction zones may change their P-T conditions in both time and space, so that a change in metamorphic thermal gradients is recorded in mineral assemblages with different peak thermobaric ratios. As a consequence, regional metamorphism of crustal rocks along convergent plate boundaries is generally characterized by two fundamentally different types of metamorphic transformation that operate one after other. An early lower T/P metamorphism is induced by lithospheric subduction to give compressional heating at lower thermal gradients. Its products are not only blueschist to eclogite facies HP to UHP rocks in modern subduction zones but also MP amphibolite to HP granulite in ancient subduction zones. A later Buchan type metamorphism is induced by extensional heating for lithospheric rifting at higher thermal gradients. Its products are LP amphibolite to granulite facies HT to UHT rocks in both modern and ancient subduction zones, together with the abundance of coeval migmatites and granitic plutons.

Although bimodal metamorphism takes place successively, their products show different spatial relationships along converging and converged plate boundaries, respectively. Whereas they are separated from each other along accretionary orogens, they are superimposed with each other along collisional orogens. Therefore, the dualism of regional metamorphism is caused by the transition in stress regime from compression to extension with increasing temperature in the evolution of convergent plate boundaries. As such, the secular change in the thermal state of subduction zones results in their division into subducting and subducted zones. Metamorphic thermal gradients and the products of bimodal metamorphism are the reflection of tectonic regimes during plate tectonic processes. Such a change can be deduced from metamorphic mineral assemblages showing a change in peak thermobaric ratios. High T/P Buchan type metamorphic facies series has been common since the Archean, whereas low T/P Alpine facies series only appeared in the Neoproterozoic. This indicates a secular change in the style of plate tectonics from the Eoarchean to Neoproterozoic.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780081029084000205

Plate Tectonics☆

Fabio Crameri , ... Clinton P. Conrad , in Reference Module in Earth Systems and Environmental Sciences, 2019

Ocean-plate destruction

Subduction zones control plate speed, are single-sided, with only one of the two colliding plates sinking asymmetrically into the mantle, and are intrinsically arcuate from a bird's eye view (see Figs. 2 and 3). The single-sidedness arises due to a strong strength contrast between the sinking plate and the plate interface, and an efficient lubrication effect which decouples the two colliding plates (Crameri et al., 2012). The arcuate shape of sinking plates and their related subduction zones are not a result of what has been called the "Ping-Pong ball effect," but in fact due to the sinking plate's surrounding mantle material and the induced flow of it (Crameri and Tackley, 2014). Sinking plates tend to retreat and thereby move mantle material behind them around the plate edges to their front side (Funiciello et al., 2003, 2004). This induced mantle flow deforms and curves the sinking plate naturally to become arcuate.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780124095489123930

Subduction Zones Parameters☆

Serge Lallemand , Arnauld Heuret , in Reference Module in Earth Systems and Environmental Sciences, 2017

Introduction

Subduction zones cover geodynamic systems where converging tectonic plates interact and evolve at both short-term and geological times scales. Their characteristics including geometry, structure, strain, or kinematics result from the complex interplay between various parameters. To better understand the processes that lead to the ancient or modern subduction settings and behaviors, we explore the spectrum of parameters that are commonly used by scientists. Jarrard in 1986 provided the first global study of 26 parameters for each of 39 modern subduction zones. Twenty years after, Heuret and Lallemand (2005) have updated part of the former compilation, which is available in the submap database (see relevant website at the end of this article). Since the pioneer work of Jarrard, several other authors have contributed in various ways to create and improve datasets related to subduction zones like Müller et al. (1997, 2016) on kinematics, seafloor age and plates reconstructions, Engdhal et al. (1998) on earthquakes catalogs, or Hayes et al. (2012) on slab contours (see relevant websites at the end of this article).

In addition to the classical parameters like "subduction velocity" or "age of plate at trench" (e.g., Otsuki, 1989), we have introduced some new "terms/codes" among the parameters which refer to elements that help to better understand the subduction systems, like "subduction channel," "seismic/aseismic behavior," or "mantle dynamics." These elements are generally not easy to quantify but still play a major role in subduction dynamics. As such, they are studied by many authors. Thanks to these new codes like "sub_chan," "IF_(a)seis," or "Man_dyn," the references to studies based on cross-correlations between parameters in the second section of this review paper is facilitated. In this review, we do not pretend to be exhaustive but we aim to provide both a list of relevant subduction zones parameters, as well as relations or possible interplays that may exist between them.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780124095489094951

Mantle Dynamics

S.D. King , in Treatise on Geophysics, 2007

7.08.7 Summary

Subduction zones are complex regions of the Earth with challenging physical and chemical processes that impact global geodynamics, geochemistry, and thermal evolution of the planet. While we have a fairly good general understanding about the processes occurring in subduction zones, many details (and a few major pieces) remain elusive. It is increasingly evident that future progress requires collaboration and communication among a variety of specialists. So perhaps topping any list of future directions or outstanding questions is the need for better communication between researchers of various backgrounds and expertise. While this has become a common statement in our scientific lexicon, when it comes to understanding subduction zones, it is probably also true. Aside from this, we provide our own list of most interesting outstanding questions:

1.

Why does Earth have plate tectonics while the other terrestrial planets do not? While this has not really been a topic discussed in this review, it is one of the foremost questions related to our understanding of global planetary geodynamics. It is hard to imagine that we really understand the process of subduction if we cannot explain why Earth is the only terrestrial planet with active subduction.

2.

What is the nature of deformation in the seismogenic zone? Are there observations in addition to fore-arc heat flow that can be used to constrain the deformation and shear heating in this region? Our understanding of the deformation and heat generation in the top 50–100   km of a subduction zone is limited. Unfortunately, the thermal structure acquired in the top 100   km is largely translated down slab dip, so our ignorance of the processes in this region is propagated along with the slab.

3.

How strong are slabs and does slab deformation significantly effect the thermal structure of the slab? For many problems of interest, slab strength may not be the major controlling factor, but for understanding the interaction of slabs with phase transformations in the transition zone, understanding the role of slab buoyancy in plate dynamics, and understanding whether slab components are able to separate (which could be significant in the chemical evolution of the mantle) slab rheology is almost certainly critical.

4.

How much water is carried into (and out of) the transition zone by slabs? The recognition that wadsleyite has a large water storage capacity has led to the interesting possibility that a large amount of water could reside in the transition zone. With the evidence for a weak mantle wedge thought to be due to dehydration of the slab and/or serpentinization of the mantle wedge, it is by no means clear that slabs retain any water by the time they reach the transition zone. The fate of water at the base of the transition zone is even more obscure. It is important to remember that this may not be a steady-state process. Water may only be carried into the transition zone during relatively unusual periods of fast subduction.

5.

How does subduction begin? While there have been some interesting developments in this area related to preexisting regions of weakness and/or lithospheric discontinuities, there is a lot more to understand. It seems the easiest way to make a new subduction zone assume preexisting subduction. This begs the question as to how the first subduction zone started. When subduction first began and what subduction looked like in the early Earth are also open questions.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B978044452748600122X

Structural Geomorphology

Jean Pierre Bergoeing , in Geomorphology of Central America, 2015

Subduction Zones

Subduction zones are confrontation areas of two or more tectonic plates colliding on an ongoing basis. This collision makes one plate go under the other, producing a leaning area of seismicity known as a Wadati-Benioff zone, which dives, in some cases, up to 434  mi to the interior of the Earth with an inclination of 40°-60°, while the other plate's ascent gives rise to mountainous reliefs. Subducted plates are generally denser, consist mainly of gabbros and peridotite, and descend gradually toward the interior of the Earth and end up melting in the mantle with the underlying magma. In the subduction process, when a plate reaches depths of 260,000-330,000   ft, hydrated minerals cease to be stable at these depths and temperatures, and they move on to more stable structures and release water they contained (Figure 1.4).

Figure 1.4. Vertical cut of two tectonic plates' collision and subduction area (x). Colliding plates produce the descent of the subducted plate into the asthenosphere, where it dehydrates, melting partially. Water released also partially melts the mantle. The ascending plate is followed by volcanic building, fed by the magmatic chambers' product of magma ascents (www.biologyeducation.net).

That liberated water reduces the melting point of materials of the mantle and melts them partially. From there, the magmatic ascents infiltrating through the collision zone are magmatic reservoirs in the cortex that feed volcanoes that are the external expression of the magmatic surge. These bags or magmatic reservoirs are located approximately 20,000-30000   ft deep in the Earth's crust. It's the difference in density between the oceanic lithosphere and the asthenosphere that increases with time; the latter grows less rapidly, creating the real engine of the subduction zone. The oceanic lithosphere, heavier by increasing density with age, acquires the tendency to immerse itself. Subduction zones, also known as active margins, are where earthquakes and volcanic activity are most frequent (e.g., the Pacific Ocean's Ring of Fire or the Indonesian southern margin).

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128031599000017

Ocean floor tectonics

C.M.R. Fowler , in Regional Geology and Tectonics: Principles of Geologic Analysis, 2012

Seismicity at subduction zones

Subduction zones are characterised by seismicity from the surface down to almost 700 km depth, and are often referred to as Wadati–Benioff zones ( Benioff, 1949; Wadati, 1928, 1935). Seismicity is often classified as shallow (0–70 km), intermediate (70–300 km) and deep (300–700 km). Shallow earthquakes occur in both plates, with most of the low-magnitude events associated with the deformation and volcanism on the overriding plate. The top of the zone of intermediate-depth seismicity beneath volcanic arcs is close to the top of the subducting plates. Intermediate and deep earthquakes occur within the subducting plates rather than being associated with its upper surface. P–T conditions mean that slip on faults should not be a possible mechanism for earthquakes below the seismogenic zone. The maximum depth of seismicity on subduction zones does not vary uniformly with the thermal parameter but deep earthquakes only occur on subduction zones with a thermal parameter over 5000 (Fig. 26.47), which is consistent with an abrupt thermally controlled change controlling deep earthquakes.

Figure 26.47. The maximum depth of subduction zone earthquakes and the corresponding thermal parameter

 (after Kirby et al., 1996b).

Globally, earthquake occurrence is bimodal with depth decreasing exponentially down to 300 km depth with a minimum of about one event per year, below which it remains low until ~550 km where it increases to a maximum at ~600 km and then ceases by 675 km (Frohlich, 1998). On average, deep earthquakes (focal depths 550–675 km) occur at ca. 20 per annum and with a maximum magnitude of ca. 8.2 (Frohlich, 1998). A subduction zone with a young, hot descending plate and fast convergence will generally produce higher magnitude earthquakes than one with an old, cold descending plate and slow convergence (Ruff and Kanamori, 1980).

Using a worldwide catalogue of shallow (<70 km) earthquakes with M S ≥ 7 that occurred between 1900 and 1989, Pacheco and Sykes (1992) show that the seismic moment released at subduction zones is 90% of all the seismic moment released by large shallow earthquakes. Large shallow normal faulting earthquakes occur along the outer arc high and the trench, while large shallow thrust and strike slip-faulting earthquakes occur in the overriding plate. The large shallow thrust-faulting earthquakes at subduction zones can only rupture those parts of the plate interface that are seismically coupled, that is, from the surface down to the onset of ductile deformation (the seismogenic zone). The maximum depth of interplate earthquakes (the brittle-ductile transition) on most subduction zones is about 40 km (Tichelaar and Ruff, 1993). This transition is deeper than 40 km at the intersection of the Japan Trench and the Kurile Trench (52–55 km) and in Central Chile and shallower on the Mexico subduction zone (20–30 km).

Mechanisms proposed for intermediate and deep earthquakes include transformational faulting, deformation embrittlement and ductile shear instability (Frohlich, 1989; Green and Houston, 1995; Kao and Liu, 1995; Kirby et al., 1996a,b; Kirby, 1987, 1991; Peacock, 2001). Figure 26.45 shows the temperature structure and the predicted regions of instability of olivine for a low and a high thermal parameter subduction zone and Figure 26.47 shows the relationship between the thermal parameter and the maximum depth of earthquakes. That transformational faulting of metastable olivine in a wedge-shaped region within the subducting plate can cause deep earthquakes is consistent with the changes in earthquake distribution with depth, mineralogy and high-pressure experimental studies. Transformational earthquake foci should occur along the hotter outer edges of the metastable olivine wedge consistent with the observations that deep earthquakes in both the Tonga and Izu-Bonin subduction zones seem to occur on separated planes.

The very shallow seismicity beneath NE Japan (Fig. 26.44) is associated with shallow deformation, thrusting and volcanism, while low-frequency events beneath the volcanic arc seem to be related to deeper activity of mantle diapirs (Nakajima et al., 2001). The earthquakes in the downgoing Pacific plate are aligned in two almost-parallel planes some 30–40 km apart, a double seismic zone (Hasegawa et al., 1978). The upper plane earthquakes have down-dip compressional stresses, while those lower in the plate have down-dip extensional stresses, mechanisms consistent with the interior of the plate undergoing brittle rupture (compare with Fig. 26.48). Similar double seismic zones have been observed at depths of 50–200 km at several other subduction zones including SW Japan, Tonga, N Chile, E Aleutians and Mariana (e.g., Comte et al., 1999; Engdahl and Scholz, 1977; Kawakatsu, 1986; Samowitz and Forsyth, 1981).

Figure 26.48. Cross-sections through the NE Japan subduction zone illustrating phase transformations, seismicity and isotherms. (A) Temperature, crustal thickness and stress state (offset circles). (B) Seismicity and hydration state (% colour scale) of the subducting oceanic lithosphere with phase transformations

 (from Hacker et al., 2003a).

While aspects of double seismic zones can be explained by the unbending of the subducting plate, other proposals include the transformation of oceanic crust (basalt, gabbro) to eclogite (Kirby et al., 1996a) and the dehydration embrittlement of metamorphosed oceanic crust and mantle of the subducting plate (Meade and Jeanloz, 1991). When the upper seismic plane is taken as marking the dehydration loci of dehydration embrittlement of the subducting oceanic crust, the lower plane is consistent with the lower dehydration loci of serpentine in the mantle (Peacock, 2001; Yamasaki and Seno, 2003). In a study of focal mechanisms for earthquakes beneath NE Japan with magnitude <5, Igarashi et al. (2001) found that at depths of 40–70 km, down-dip compression events underlie low-angle thrust faulting events. The trench axis marks the arc-ward extent of the low-angle thrust faulting events and hence the western extent of active interplate seismicity. At depths greater than 70 km, Igarashi et al. (2001) find the seismicity not to be double but rather triple-planed: (1) the upper plane events are mainly down-dip compression, (2) normal-faulting events also occur in the uppermost part of the downgoing plate very close to the plate boundary and (3) the lower plane events are down-dip extension (Fig. 26.48). Hacker et al. (2003a) demonstrate evidence for a key role for dehydration in causing intermediate earthquakes – intermediate earthquakes (a) occur in downgoing plates that are predicted to be hydrous, and (b) are absent in dehydrated downgoing plates. Figure 26.48B shows the correlation between intermediate seismicity beneath NE Japan and the shallower phase transformations (Fig. 26.46) that occur within the subducting Pacific plate (Hacker et al., 2003a,b).

England et al. (2004) find the range of depths of the intermediate seismicity immediately beneath volcanic arcs to be 65–130 km, with the depth varying little along individual arc segments, but with differences between arc segments (England et al., 2004). The depth to the top of the seismicity beneath volcanic arcs does not correlate with either the thermal parameter of the subduction zone or with the age of the subducting lithosphere, but rather is inversely correlated with the plate descent rate (Fig. 26.49). This is consistent with the primary control on the location of the volcanic arc being the temperature in the overlying mantle wedge (England and Wilkins, 2004).

Figure 26.49. Depth of the top of the intermediate depth seismicity beneath volcanic arcs plotted against the vertical descent rate of the subducting plate (subduction rate multiplied by sine of plate dip angle). Grey filled circles, volcanic arcs having significant back-arc spreading. White circles, volcanic arcs having significant back-arc spreading, with the descent rate recalculated to include the back-arc spreading rate. R, Spearman rank-order correlation coefficient

 (after England et al., 2004, Systematic variation in the depths of slabs beneath arc volcanoes, Geophys. J. Int., 156, 377–408, John Wiley and Sons, with permission.).

Low-frequency tremors have been recorded along the Nankai (SW Japan) subduction zone where the Philippine plate is subducting beneath the Eurasian plate. These tremors, which have periods in the range 0.1–1 s and last for several days, are located within the Philippine plate at an average depth of ~30 km and thus close to the Moho (Obara, 2002). Such low-frequency events are normally associated with active volcanoes, but the epicentres of these tremors are located in a 600 km along-strike, non-volcanic portion of the subduction zone, which has experienced frequent magnitude 8 earthquakes. The Philippine plate is young and has a thick sediment cover. At 40 km depth beneath SW Japan, the temperature is ca. 450 °C, the seismogenic zone is likely to be shallow and narrow and pore fluid-pressure could be very high (Seno and Yamasak, 2003). A detailed inversion of the 3D seismic velocity structure of SW Japan by Wang et al. (2006) shows that the tremors are associated with strong low P-wave and S-wave anomalies at a depth of 32 km. There are two gaps within the 600-km-long tremor belt – within these gaps P-wave velocity and S-wave velocity are high and Poisson's ratio is low. Wang et al. (2006) conclude the tremors are associated with the dehydration reactions occurring in the lowermost crust of the Philippine plate beneath the forearc where there is high pore fluid-pressure and high crack-density, so that the tremors are caused by fluids migrating through faults and the opening and closing of cracks. Similar tremors have also been observed on the Cascadia subduction zone where they are co-located with strong seismic reflectors in the upper plate, and are thought to be associated with trapped fluids from the dehydration of the underlying oceanic crust (Abers et al., 2009; Kao et al., 2005).

Only a fraction of plate motion at subduction zones is accommodated by co-seismic displacement, with the missing displacement being accommodated by aseismic slip. GPS surveys have confirmed that aseismic slip takes several forms: slow slip and tremor occur along with steady state creep (Schwartz and Rokosky, 2007).

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780444530424000261