subduction - géosciences montpellier · large earthquakes, active volcanoes, and sometimes...

Subduction

Serge Lallemand*Géosciences Montpellier, Université Montpellier 2, Montpellier, France

Synonyms

Convergent plate boundary; Destructive plate boundary; Plate consumption

Definition

At a convergent plate boundary, one of the plates necessarily passes beneath the other. This process,first termed subduction by A. Amstutz in 1951 when discussing the evolution of alpine structures,allows nearly all oceanic lithosphere to be recycled into the mantle. By contrast, only the lowerportion of the continental lithosphere (the lithospheric mantle) may, under certain conditions, betransported into the deep mantle. Since the continental crust resists subduction at depths beyondabout 100 km due to buoyancy forces, this type of convergence is typically called collision andforms an “orogenic” or a “mountain” belt of which the Alps or the Himalayas are classical examples.Subduction zones can generally be traced at the surface of the Earth because they are associated withlarge earthquakes, active volcanoes, and sometimes mountain building.

Classification

There are two main types of subduction (Fig. 1), depending on the nature of the downgoing plate(hereafter called slab):

• Oceanic subduction (often just called subduction) when an oceanic plate subducts• Continental subduction (often just called collision) when a continental plate subducts

Oceanic subduction can last tens or hundreds of millions of years without a significant change inthe topography and tectonics in the upper plate, whereas continental subduction, which typicallyoccurs after a period of oceanic subduction, rapidly causes either a plate reorganization or mountainbuilding. Below, we will focus primarily on oceanic subduction processes (see “▶Orogeny” entryfor details on continental subduction).

Each of these subduction types can be divided in two “subtypes” depending on the nature of theoverriding plate. The most common situation at a convergent plate boundary is the subduction of anoceanic plate beneath a continental plate (�67 %, 45,000 km, see entry “▶Active ContinentalMargin”). Among the remaining 33%, the converging continents contribute for about half, i.e., 17%(11,200 km), of the total, intra-oceanic subduction zones for about 15 % (10,000 km) and continentssubducting beneath oceanic plates for only 1 % (600 km) (Lallemand et al., 2005a).

*Email: [email protected]

Encyclopedia of Marine GeosciencesDOI 10.1007/978-94-007-6644-0_103-1# Springer Science+Business Media Dordrecht 2014

of 16

http://link.springer.com/SpringerLink:ChapterTarget






Many other criteria or combinations of criteria have been proposed to describe the wide variety ofsubduction zones (Jarrard, 1986). The most commonly used, first introduced by Uyeda in 1984, isthe opposition between the Chilean-type and the Mariana-type subductions. The Chilean-type ischaracterized by a young subducting plate with a shallow slab dip, a high degree of interplatecoupling, inducing high stresses and producing large earthquakes, mountains, and thus terrigenoussediment supply feeding an accretionary wedge. On the contrary, the Mariana type is characterizedby an old subducting plate with a steep slab dip; a low degree of interplate coupling, producing lowstresses; and only moderate seismicity, the presence of back-arc extension, and a low topographyunable to feed the trench. Despite the correct description of these two end-member subductionzones, Heuret and Lallemand (2005), based on revised global datasets, have shown that some ofthese associations of parameters are not statistically verified. Those which are statistically verified inmodern subduction zones are the combination of upper plate compressional stress (typically Chile),shallow-dipping slabs (less than 50�), and upper plate seaward absolute motion, or, reversely, upperplate extensional stress (typically Mariana), steeply dipping slabs (more than 50�), and upper platelandward absolute motion (Lallemand et al., 2005b). The other characteristics such as plate age,seismicity, or the presence of an accretionary wedge do not show simple correlations.

Distribution on Earth

Ninety percent of the oceanic subduction zones are concentrated around the Pacific Ocean and inSoutheast Asia (48,700 km). The rest are distributed in the Atlantic Ocean (5,000 km) and theMediterranean Sea (1,350 km). This asymmetric distribution at Earth’s surface results from thebreakup of the Pangea in Late Paleozoic and Early Mesozoic time accompanied by the rifting of theAtlantic and Indian oceans and the closure of the Tethys and Panthalassa oceans. Because the mostrapid subduction rates are presently observed in the Western Pacific and Southeast Asia, Xavier LePichon in 1990 (lessons at Collège de France) compared this relatively small region with a huge pit

Fig. 1 Two types of subduction mainly depending on the nature of the downgoing plate: oceanic subduction, 82 %, andcontinental subduction, 18 %, modified after Lallemand (1999)


of 16

centered on the Philippine Sea where oceanic plates are recycled into the mantle. Subduction ratesgenerally range between 1 cm/year off Sicily, for example, and 10 cm/year off Japan, except inSouthwest Pacific where it exceptionally reaches 24 cm/year offshore the northernmost Tongaarchipelago.

Distinctive Characteristics of Subduction Zones

The plane forming the interface between converging plates has distinctive properties compared witha “classical” active fault. It accommodates the underthrusting of one plate beneath another. Its dipangle may vary between 10� and 35� for the shallow part which is seismogenic (Heuret et al., 2011),and it can reach 60� in its deepest part. The subduction interface outlines the contact surface betweenthe plates, i.e., from the trench down to the limit with the convecting mantle. The main differenceswith a “classical” active fault are:

– A great size in both downdip width (typically between one and two hundreds of km) and lateralextent (length between a few hundreds and a few thousands of km).

– A lower thermal gradient along the interface because of the continuous advection oflow-temperature rocks (top of subducting plate). The thermal state depends on the age of theincoming plate and the rate of subduction (Stein and Stein, 1996).

– Fluid overpressure along the interface caused by the progressive dehydration of the subductingplate. Fluid pressure at a given depth depends on the available input from the subducting slab,temperature, and pressure that control phase changes and permeability of the rocks of the hangingwall (see review by Saffer and Tobin, 2011).

The great size of the fault and the low temperature increases the seismic potential. This is thereason why all giant earthquakes occur at subduction zones.When the subducting plate is old and thesubduction rate is fast (e.g., below Japan), the subducting plate is characterized by a cloud ofintraslab earthquakes termed the “Wadati-Benioff” or “Benioff” zone (Wadati, 1928; Benioff, 1949,see entry “▶Wadati-Benioff Zone”) down to 660 km, which is the depth of the discontinuitybetween the upper and the lower mantle.

The last distinctive characteristic is the volcanic arc that generally outlines the oceanic subductionzones. Fluids stored in the subducting plate are released at various depths from the first kilometers ofsubduction near the trench (pore fluids) down to about 150 km (hydrous minerals) depending on thethermal state of the slab. At depths between 80 and 150 km, slab dehydration is responsible for themetasomatism of the overlying mantle wedge. A small fraction of the asthenospheric wedgeundergoes partial melting and migrates upward. Volcanoes thus develop at a given distancedepending on the slab dip because melts migrate vertically from the slab.

Slab Shape and Dynamics

At a large scale, oceanic plates behave elastically when they are forced to bend in subduction zones.We often observe a small bulge seaward of the trench before the plate sinks into the mantle. Thetrench is the deepest surface expression of the oceanic plate. Its depth depends on the age of thesubducting plate and the amount of sediment infill. The deepest trenches are the Mariana Trough(10,938 m revised depth by Fujioka et al., 2002) and the Philippine trench (10,100 m after


of 16


Lallemand et al., 1998). It is interesting to note that the deepest trenches are located right in the areaof maximum downwelling of oceanic lithosphere on Earth as computed by Lithgow-Bertelloni andRichards (1998). Some authors argue that the elastic properties of the sinking lithosphere help theslab to unbend below the arc so that it recovers a flat profile. In detail, other forces may alsocontribute in the unbending of the plate such as the suction force exerted by the overriding plate andmantle wedge. Slabs may exhibit various shapes such as flat and horizontal in their upper section(Peru or Central Chile), shallow dipping (Japan or Kurile), intermediate dipping (Tonga orKermadec), steeply dipping (Izu-Bonin), vertically dipping (Mariana), or even overturned (Luzonor New Britain).

In modern subduction zones, active compression in the overriding plate is always observed forshallow-dipping slabs (<50�), whereas active back-arc extension is always observed for steeplydipping slabs (>50�). Lallemand et al. (2005b, 2008) have shown that the slab shape is not simplylinked with the age of the subducting plate (the common assumption is that an old plate is thick anddense and thus prone to sink steeply or even vertically in the mantle) but rather with the combinationof absolute velocities between converging plates and the regional mantle dynamics. Moreover, theupper plate strain appears to directly result from this kinematic interaction.

Main Driving and Resistive Forces in Subduction Zones

Today, the main driving mechanism for plate motion in general (see the entry “▶Driving Forces”),and for subduction in particular, is the gravitational sinking of the slab (Turcotte and Schubert, 1982)which is usually slightly denser (about 1 %) than the surrounding mantle because of its lowertemperature (thermal contraction). This force is called slab pull (Fig. 2). The viscous mantle opposes

Fig. 2 Description of driving and resistive forces in a subduction zone modified after Lallemand et al. (2008). Manyforces can point in two opposite directions depending on the geodynamic context. Full lines are the most frequentdirections, and dotted lines are less frequent directions but still observed. Far-field forces commonly originate fromridge-push, slab pull, or collision


of 16


a resistance to plate motion from the ridge to the trench calledmantle drag. The upper portion of theasthenosphere is known to have a low viscosity so that the mantle drag should be moderate. Theviscous shearing increases considerably as the plate sinks into the mantle because it exerts on bothsides – top and bottom – of the descending slab and because the viscosity is pressure dependent(Karato andWu, 1993). It probably increases by at least one order of magnitude between the top andthe bottom of the upper mantle. This resistance to penetration drops again when the slab reaches thediscontinuity between the upper and the lower mantle near 660 km as a result of phase changes inminerals. The mantle opposes another resistance called anchoring, which accounts for mantlereaction to slab facewise translation (Scholtz and Campos, 1995; Lallemand et al., 2008). Since itis a viscous pressure force, it will depend on the velocity of slab rollback or advance associated withtrench migration and increase with depth. The slab also resists to bend and unbend when passing thetrench. This resisting force called bending moment increases with the stiffness of the plate and thusits age. It is inversely proportional to the radius of curvature of the slab (Buffett and Rowley, 2006;Wu et al., 2008).

Among these main internal forces, two are always resisting: the mantle resistance to penetration,by definition, which is effective as soon as a slab dives into the mantle, and the bending moment, alsocalled bending resistance, as the plate bends and unbends. All the other forces can shift from drivingto resisting depending on the geodynamic context. Let us first consider slab pullwhose name comes,by definition, from the capacity of the slab to sink into the mantle because of its negative buoyancy.Most of the time, it is considered as a driving force, but in about 10 % of cases, we observe “flat”slabs, i.e., subducting slabs sliding horizontally beneath the upper plate over hundreds of kilometersbefore sinking. Gutscher et al. (2000) attributed this flatness to a positive buoyancy of the oceanicplate and a delay in the basalt to eclogite transition due to the cool thermal structure of the twooverlapping lithospheres. The best examples are the Nankai or Cascadia subduction zone where thesubducting crust is very young, or the Peru and Central Chile segments along the Andean marginwhere buoyant ridges or plateaus are subducting. Themantle drag generally offers resistance to platemotion near subduction zones except when mantle convective cells, produced by another mecha-nism rather than the subduction, drive the plate down. This could be the case when a spreading ridgesubducts like in South Chile or in Cascadia. The anchoring force always offers resistance to facewisetranslation of the slab but not much to its downward penetration.

There are also second-order external forces originating from the far field (Fig. 2). Their magnitudeand direction depend on the geodynamic context. The so-called ridge-push results from thegravitational collapse or subsidence of the oceanic plates as they drift away from the ridge. Fowler(1990) estimates that, on average, the slab pull force is one order of magnitude greater than the ridge-push force. If the active ridge is located seaward of the trench, the ridge-push force will push theplate toward the trench, but if the spreading ridge is located landward of the trench, it will push theupper plate onto the subducting plate (like in South America with the East Pacific Rise on one sideand the mid-Atlantic Ridge on the other). Far-field slab pull can also provide an additional forceacting in the subduction zone. For example, the Izu-Bonin-Mariana arc overrides the subductingPacific plate but is carried by the Philippine Sea plate which is dragged northwestward asa consequence of the slab pull acting along the Ryukyu subduction zone (Pacanovsky et al.,1999). In this case, the upper plate is pulled away from the trench. Another source of external stressis the effect of a collision like the Ontong Java Plateau with the Solomon trench that impacts thestress field over a huge area from New Guinea to New Hebrides (Mann and Taira, 2004).


of 16

Subduction Interface

The shallowest part of the subduction interface often corresponds to an aseismic decollement zonecharacterized by overpressures resulting from the pore fluids release as lithostatic pressure increases(Saffer and Tobin, 2011). At rather low temperatures between 100–150� and 350–450� (Oleskevichet al., 1999; Hyndman, 2007), the plate interface becomes frictional and thus seismogenic (see entry“▶ Seismogenic Zone” and Dixon and Moore, 2007). All “subduction earthquakes” nucleate in thisregion. The seismogenic zone generally extends from a depth of 11 � 4 to 51 � 9 km. It is about112 � 40 km wide and dips 23 � 8 � (Heuret et al., 2011). At temperatures higher than 350–450�,the slip along the interface is either continuous (creep) or characterized by slow-slip events (SSE)associated with nonvolcanic tremors (NVT) (Rogers and Draggert, 2003). The downdip limit of thesubduction interface is marked by the coupling between the descending plate and the overlyingmantle (Fig. 3). Furukawa (1993) called the depth below which the plates are uncoupled on the longterm Zdec. Downdip of this depth, a significant thickness of mantle is dragged with the slab, called“viscous blanket” by Kincaid and Sacks (1997). The thickness depends on the temperature contrastbetween the slab and the mantle wedge. Beyond Zdec, there is no more plate interface, the slab beingoverlain by the convective mantle wedge.

There is a debate on the nature of the plate interface. Is it an � 1–20 m thin “décollement” zonelike those drilled in Nankai (Bangs et al., 1996) or an � 100–1,000m thick channel across which theshear distributes like in the exhumed Franciscan complex (Cloos and Shreve, 1988)? The discoveryof erosional processes at the hanging wall of the subduction interface in Japan, Middle America, andIzu-Bonin-Mariana trenches since the early 1980s (von Huene et al., 1980; von Huene andLallemand, 1990; von Huene and Scholl, 1991; Lallemand et al., 1992; Meschede et al., 1999)has been decisive in promoting the concept of subduction channel. It has been later confirmed,

Fig. 3 Schematic illustration of the subduction interface in an oceanic subduction zone. Uz is the upper limit and Dz thedowndip limit of the seismogenic zone. Zdec is the decoupling depth. Below that depth, the convective mantle is draggeddown (coupled) with the subducting plate. NVT nonvolcanic tremors and SSE slow-slip events are observed in somesubduction zones at depths between Dz and Zdec


of 16


thanks to the increasing resolution of seismic images as in the Ecuadorian margin (Sage et al., 2006;Collot et al., 2008) and the reinterpretation of some outcrops of exhumed subduction interface like inthe Shimanto belt (Kitamura et al., 2005) or in the Apennines (Vannucchi et al., 2008). Severalprocesses of upward migration of the roof décollement or downward migration of the basaldécollement delimiting the subduction channel have been discussed by Vannucchi et al. (2012)and are not yet completely understood. Hydrofracturing of the hanging wall rocks under increasingfluid pressure coming from the dewatered downgoing plate (von Huene et al., 2004) very probablycontributes to the upward migration of the deformation front as well as the subduction of oceanicreliefs such as seamounts and ridges (Dominguez et al., 1998, 2000).

At the scale of the plate interface, two forces govern local tectonics and hazards: friction andpressure. The net pressure or normal stress exerted along the subduction interface combines thelithostatic pressure, increasing with depth, and the non-isostatic “dynamic” pressure, whichaccounts for the slab pull, the slab-mantle interactions, and the far-field “tectonic” forces. Theidea, developed by Shemenda (1992) in laboratory analog models, where far-field forces and mantledrag and anchoring are neglected, is that a dense (old) and long subducting plate tends to pull theupper plate down producing a subsidence of the forearc, whereas a light (young) or a shortsubducting plate tends to push the upper plate up producing uplift of the forearc. In nature, notonly the age and length of the subducting plate but mantle and far-field forces also contribute to thenormal stress along the interface. The second force is frictional shear, which resists subduction. Thestress is released intermittently as it overcomes the yield strength of the fault, through repeatedearthquakes and subsequent fault ruptures.

Seismicity with Special Emphasis on Mega-thrust Earthquakes

Subduction zones are the locus of most earthquakes occurring on Earth because the plate boundarieshave to accommodate the largest relative motions and subsequent strain, because the plate interfaceis continuously cooled and because the slab still remains cool at large depths. “Subductionearthquakes” are thrust events occurring along the seismogenic plate interface. They representmost of the seismic energy released at subduction zones and in the world (Scholz, 2002). Thereare also intraslab shallow, intermediate, and deep earthquakes that accommodate the high level ofstress during the plate bending and penetration into the viscous mantle and the upper plateearthquakes that accommodate the part of the stress transmitted through the plate interface. Seeentries “▶Earthquake”, “▶Active Continental Margin” and in particular Fig. 2 of “▶ActiveContinental Margin.”

One specificity of the subduction zones is that they can generate giant earthquakes with magni-tudes � 8.5, most of them being tsunamigenic like the 2004 M9.2 event off Sumatra (Indonesia) orthe 2011 M9.0 event off Tohoku (Japan). Back to the 1900–2012 period, Fig. 4 shows the ruptureareas associated with the M � 8.0 earthquakes and the segmentation of the seismogenic zonesmainly based on past earthquake ruptures. It is easy to see that they don’t distribute equally along thesubduction zones. Based on available global datasets, Heuret et al. (2011, 2012) have shown thatgiant earthquakes preferentially occurred near slab edges, where the upper plate was continental, theback-arc strain neutral to compressive, and the trench fill larger than 1 km. A possible explanationwould be that the rupture propagation could be facilitated if the plate interface is smooth (Ruff, 1989)and the normal stress on the interface large enough to accumulate strain but not too large to allowrupture propagation. The continental nature of the upper plate could be preferred because it isassociated with wider (downdip) seismogenic zones. The dependence of earthquake magnitude with


of 16





Fig.4

Map

ofthesubductio

ninterfaceseismicrate,trenchsegm

ents,and

major

ruptureareasof

theM

w�

8.01900

–2012eventsupdatedfrom

Heuretetal.(2011).

Paleblue

arrowsindicatesubductio

nrelativ

evelocitiesafterHeuretand

Lallemand(2005)


of 16

the subduction velocity and plate age, first proposed by Ruff and Kanamori (1980, 1983), is notstatistically verified using revised databases. However, none of the Sumatra and Tohoku earthquakeswere expected to occur based on our knowledge at the time they happened, because the recurrencetime of such huge events is often underestimated and because we generally do not have access topaleoearthquake archives (McCaffrey, 1997, 2008; Lay and Kanamori, 2011).

The degree of coupling between the converging plates may be used to estimate the seismic hazardat a given subduction zone. For many years, authors calculated the seismic coupling coefficient as theratio between the observed seismic moment release rate and the rate calculated from plate tectonicvelocities (Brune, 1968; Davies and Brune, 1971). Scholz and Campos (1995) have shown that thiscoefficient should be used cautiously because the seismic slip was generally computed over a shorttime period (typically a hundred years) compared with the recurrence time for large earthquakes(sometimes up to a thousand years). During the last two decades, the advent of monitoring ofsubduction zones with GPS stations allowed geodesists to measure the interseismic strain accumu-lation rate in the upper plate, which gives a better picture of the interplate coupling than the earliermeasure of energy release rate by summing the earthquake moment over a short period. By doingthis, Scholz and Campos (2012) have revised their estimate of the seismic coupling coefficient inmany subduction settings. They concluded that this new approach agrees well with the revised“classical way” of measuring the coefficient accounting for major earthquakes older than a centurywhen available. It is also more satisfactory for the balance of forces acting along the plate interfacelike the normal stress and the friction making assumptions on their respective origins (see above).They confirm the observation of Chemenda et al. (2000) that high coupling, and thus high friction, isobserved in regions of high normal stress. The limitation of the method, consisting of measuring theinterseismic elastic strain, is that most of the time the coupled area in a subduction zone is locatedoffshore, whereas GPS receivers are deployed onshore. The resolution of the models should beimproved in the future when a significant number of GPS stations can be installed offshore like offJapan. Tremendous developments must be further done to precisely map the complexity of thecoupling distribution along the subduction interface because large lateral variations are predicted byauthors (Ruff, 1992; Bilek, 2007; Scholz and Campos, 2012).

Upper Plate Tectonics

The upper plate is the visible part of a subduction zone, also called active margin. The tectonicregime of the active margin depends on interactions between the converging plates at the level of thetrench, the plate interface, and even the whole overriding plate.

At the trench, sediment carried by the subducting oceanic plate may be offscraped and incorpo-rated at the front of the overriding plate. This process, called frontal accretion, contributes to thegrowth of an accretionary wedge (see entry “▶Accretionary Wedge” for more details or Moore andSilver, 1987). Von Huene and Scholl (1991) have shown that half of the growth of an accretionaryprism can be accounted for by subcrustal or basal accretion, also called underplating. They alsoestimated that only half of the modern convergent margins are accretionary. The other halfundergoes what they called tectonic erosion. This process has been definitely demonstrated in thelate 1970s (e.g., Scholl et al., 1980) after observations from both deep-sea drilling and reflectionseismic imaging, which the Japan margin underwent a Miocene subsidence of at least 2 km. Suchlarge subsidence increasing seaward, together with the absence of an accretionary prism, erosionalfeatures at margin’s front, and simultaneous retreat of the volcanic arc, led the authors to proposea model in which part of the margin was removed both frontally and subcrustally during subduction.


of 16


Since 1980, other margins appeared to also undergo tectonic erosion like in mid-America, Peru,Tonga, Mariana, or Izu-Bonin (von Huene and Lallemand, 1990; Lallemand, 1995; Meschede et al.,1999; Clift and Vannucchi, 2004). See entry “▶ Subduction Erosion” for more details.

The convergent margins deform in response to the stress transmitted across the plate interface butalso in response either to frontal or basal accretion or frontal or basal tectonic erosion. The tectonicregime is often compressive in the frontal part of the overriding plate, but it is also quite common toobserve normal faults dipping either landward or seaward in active convergent margins, especiallywhen they are subject to tectonic erosion. These faults reflect either a trenchward collapse as themargin progressively steepens (Aubouin et al., 1984), or a buckling as underplating occurs(Lallemand et al., 1994). Another mechanism, coseismically induced, has been recently suggested.Indeed, a seaward dipping normal fault has been activated at the front of the Japan margin during the2011 Tohoku mega-earthquake. The explanation provided by McKenzie and Jackson (2012) wouldbe that both the release of gravitational potential energy and the elastic strain account for thiscoseismic behavior.

Trench-parallel strike-slip faults may also be observed in the forearc in case of oblique conver-gence like off Sumatra (Mentawai Fault) or the southern Ryukyus (Yaeyama Fault). Thesetranscurrent subvertical faults accommodate part of the lateral component of the oblique conver-gence, whereas the slip along the frontal part of the subduction thrust is closer to trench normal (e.g.,McCaffrey, 1992). This mechanism is verified by the direction of slip vectors of subductionearthquakes that are often deflected toward trench normal when the convergence is oblique. Thisprocess called slip partitioning has been observed not only at the scale of the forearc but also at thescale of tectonic plates. Indeed, several oblique subduction systems develop transcurrent faults.Chemenda et al. (2000), for example, have investigated using physical models the effect of thepressure and the friction along the plate interface on slip partitioning in a context of obliqueconvergence. They conclude that slip partitioning can occur in the models only when interplatefriction is high and when the overriding plate contains a weak zone. As a matter of fact, lithospherictranscurrent faults develop either at the toe of the rigid backstop (rear of accretionary wedge like offEast Taiwan), along the volcanic arc (Sumatra, the Philippines, South Kuriles), at the back-arc riftaxis, or spreading center (Andaman Sea, Le Havre Trough).

Ultimately, the global strain rate map (Kreemer et al., 2003) indicates that most subductionsystems deform far from the plate boundaries, or one may consider that the plate boundaries arewider than a single fault. They can potentially affect a strip several hundreds of kilometers large atupper plate edge. In addition to interseismic elastic deformation, there are many areas wherepermanent (plastic) deformation localizes. We have seen above that, in some oblique geodynamicsettings, transcurrent faulting may occur, but even more shortening, rifting, or spreading can alsooccur. Significant shortening (>1 cm/year) is observed, for example, along the eastern margin of theJapan Sea, the southern margin of the Okhotsk Sea, or the eastern cordillera of the Andes.Opposingly, rifting or spreading develops in the Sumisu rift and the Mariana Trough (Izu-Bonin-Mariana arc), the Lau Basin and Le Havre Trough (Tonga-Kermadec arc), the Andaman Sea(Andaman arc), the Central Kamchatka graben (North Kurile arc), the Manus Basin (New Britainarc), the North and South Fiji basins (New Hebrides arc), the Scotia Sea (South Sandwich arc), theTyrrhenian Sea (Calabria arc), or the Aegean Sea (Hellenic arc). Several mechanisms have beeninvoked to account for this upper plate deformation playing with interplate stress resulting fromplate kinematics (Dewey, 1980; Lallemand et al., 2008) and/or mantle wedge dynamics (Sleep andToksöz, 1971; Lagabrielle et al., 1997; Faccenna et al., 2010).


of 16


Magmatism at Subduction Zones

Volcanic arcs are the second worldwide magmatic contributor to crustal growth on Earth after theoceanic spreading centers (nearly 1 km3/year according to Crisp, 1984). Most subduction zones arecharacterized by an active volcanic arc. Sometimes, the volcanic activity ceases for a few millionyears and resumes again. The reason for the cessation of volcanism is commonly either the collisionwith a continent or the geometry of the slab which is incompatible with the presence of a mantlewedge. This is the case, for example, with flat slab segments in the Andes (Gutscher et al., 2000). Arcmagmatism results from the metasomatism and partial melting of rocks of the mantle wedge that arehydrated by the water extracted from hydrous minerals in the subducting plate at depths rangingbetween 80 and 150 km. Further details on arc magmatism can be found in Tatsumi (1989) orTatsumi and Kosigo (2003). See also entries “▶Magmatism at Convergent Plate Boundaries”, and“▶ Island Arc Volcanism, Volcanic Arcs”.

Access to a database on kinematic, geometric, seismologic, and structural parameters of oceanicsubduction zones can be found via url: submap.fr.

Summary

Subduction is a lithospheric process which occurs at convergent plate boundaries. One plateunderthrusts another and sinks into the Earth mantle. The subduction mainly involvesa subducting oceanic plate, but sometimes, a continental plate enters into the subduction zone. Inthis last case, a collisional chain develops. This process generates a high level of seismicity as therocks deform in the vicinity of the plate boundary. All mega-earthquakes occur in subduction zones.Oceanic subduction creates favorable conditions for arc magmatism.

Cross-References

▶Accretionary Wedge▶Active Continental Margin▶Crustal Accretion▶Driving Forces▶Earthquake▶ Intraoceanic Subduction Zone▶ Island Arc Volcanism, Volcanic Arcs▶Lithosphere-Composition and Formation▶Magmatism at Convergent Plate Boundaries▶Marine Evaporites▶Morphology Across Convergent Plate Boundries▶Ocean Margin Systems▶Orogeny▶ Seamounts▶ Seismogenic Zone▶ Subduction Erosion▶Wadati-Benioff Zone▶Wilson Cycle-Marine Geosciences


of 16





















Bibliography

Amstutz, A., 1951. Sur l’évolution des structures alpines. Archives Science, 4(5), 323–329.Aubouin, J., Bourgois, J., and Azéma, J., 1984. A new type of active margin: the convergent

extensional margin, as exemplified by the Middle America Trench off Guatemala. Earth PlanetScience Letters, 67, 211–218.

Bangs, N. L., Shipley, T. H., and Moore, G. F., 1996. Elevated fluid pressure and fault zone dilationinferred from seismic models of the northern Barbados Ridge décollement. Journal of Geophys-ical Research, 101(B1), 627–642.

Benioff, H., 1949. Seismic evidence for the fault origin of oceanic deeps. Geological Society ofAmerica Bulletin, 60, 1837–1856.

Bilek, S. L., 2007. Influence of subducting topography on earthquake rupture. In Dixon, T. H., andMoore, J. C. (eds.), The Seismogenic Zone of Subduction Thrust Faults. New York: ColumbiaUniversity Press, pp. 123–146.

Brune, J., 1968. Seismic moment, seismicity, and rate of slip along major fault zones. Journal ofGeophysical Research, 73, 777–784.

Buffett, B. A., and Rowley, D. B., 2006. Plate bending at subduction zones: consequences for thedirection of plate motions. Earth Planet Science Letters, 245, 359–364, doi:10.1016/j.epsl.2006.03.011.

Chemenda, A., Lallemand, S., and Bokun, A., 2000. Strain partitioning and interplate friction inoblique subduction zones: constraints provided by physical modeling. Journal of GeophysicalResearch, 105(3), 5567–5582.

Clift, P., and Vannucchi, P., 2004. Controls on tectonic accretion versus erosion in subduction zones:implications for the origin and recycling of the continental crust. Reviews of Geophysics, 42,RG2001, doi:10.1029/2003RG000127.

Cloos, M., and Shreve, R., 1988. Subduction channel model of prism accretion, melange formation,sediment subduction, and subduction erosion at convergent plate margins: 2. Implications anddiscussion. PAGEOPH, 128(3/4), 501–545.

Collot, J.-Y., Agudelo,W., Ribodetti, A., andMarcaillou, B., 2008. Origin of a crustal splay fault andits relation to the seismogenic zone and underplating at the erosional north Ecuador–southColombia oceanic margin. Journal of Geophysical Research, 113, B12102, doi:10.1029/2008JB005691.

Crisp, J. A., 1984. Rates of magma emplacement and volcanic output. Journal of Volcanology andGeothermal Research, 20, 177–211.

Davies, G., and Brune, J. N., 1971. Global plate motion rates from seismicity data. Nature, 229,101–107.

Dewey, J. F., 1980. Episodicity, sequence, and style at convergent plate boundaries. In Strangway,D. W. (ed.), The continental crust and its mineral deposits, Geological Association of CanadaSpecial paper, Vol. 20, pp. 553–573.

Dixon, T. H., and Moore, J. C. (eds.), 2007. The Seismogenic Zone of Subduction Thrust Faults.New York: Columbia University Press. 680 pp.

Dominguez, S., Lallemand, S. E., Malavieille, J., and von Huene, R., 1998. Upper plate deformationassociated with seamount subduction. Tectonophysics, 293, 207–224.

Dominguez, S., Malavieille, J., and Lallemand, S. E., 2000. Deformation of accretionary wedges inresponse to seamount subduction – insights from sandbox experiments. Tectonics, 19(1),182–196.


of 16

Faccenna, C., Becker, T. W., Lallemand, S., Lagabrielle, Y., Funiciello, F., and Piromallo, C., 2010.Subduction-triggered magmatic pulses : a new class of plumes? Earth Planet Science Letter, 299,54–68, doi:10.1016/j.epsl.2010.08.012.

Fowler, C. M. R., 1990. The Solid Earth: An Introduction to Global Geophysics. New York:Cambridge University Press. 472 pp.

Fujioka, K., Okino, K., Kanamatsu, T., and Ohara, Y., 2002. Morphology and origin of thechallenger deep in the southern Mariana trench. Geophysical Research Letters, 29(10), 1372.

Furukawa, Y., 1993. Depth of the decoupling plate interface and thermal structure under arcs.Journal of Geophysical Research, 98, 20005–20013.

Gutscher, M.-A., Spakman, W., Bijwaard, H., and Engdahl, E. R., 2000. Geodynamics of flatsubduction: seismicity and tomographic constraints from the Andean margin. Tectonics, 19(5),814–833.

Heuret, A., and Lallemand, S., 2005. Plate motions, slab dynamics and back-arc deformation.Physics of the Earth and Planetary Interiors, 149, 31–51.

Heuret, A., Lallemand, S., Funiciello, F., Piromallo, C., and Faccenna, C., 2011. Physical charac-teristics of subduction interface type seismogenic zones revisited. Geochemistry, Geophysics,Geosystems, 12, Q01004, doi:10.1029/2010GC003230.

Heuret, A., Conrad, C. P., Funiciello, F., Lallemand, S., and Sandri, L., 2012. Relation betweensubduction megathrust earthquakes, trench sediment thickness and upper plate strain. Geophys-ical Research Letters, 39, L05304, doi:10.1029/2011GL050712.

Hyndman, R. D., 2007. The seismogenic zone of subduction thrust faults: What we know and don’tknow. In Dixon, T. H., and Moore, J. C. (eds.), The Seismogenic Zone of Subduction ThrustFaults. New York: Columbia University Press, pp. 15–40.

Jarrard, R. D., 1986. Relations among subduction parameters. Reviews of Geophysics, 24, 217–284.Karato, S., and Wu, P., 1993. Rheology of the upper mantle: a synthesis. Science, 260, 771–778.Kincaid, C., and Sacks, I., 1997. Thermal and dynamical evolution of the upper mantle in subduction

zones. Journal of Geophysical Research, 102, 12295–12315.Kitamura, Y., Sato, K., Ikesawa, E., Ikehara-Ohmori, K., Kimura, G., Kondo, H., Ujiie, K., Tiemi

Onishi, C., Kawabata, K., Hashimoto, Y., Mukoyoshi, H., andMasago, H., 2005. Mélange and itsseismogenic roof décollement: a plate boundary fault rock in the subduction zone—an examplefrom the Shimanto Belt, Japan. Tectonics, 24, TC5012, doi:10.1029/2004TC001635.

Kreemer, C., Holt, W. E., and Haines, A. J., 2003. An integrated global model of present-day platemotions and plate boundary deformation. Geophysical Journal International, 154, 8–34.

Lagabrielle, Y., Goslin, J., Martin, H., Thirot, J. L., and Auzende, J. M., 1997. Multiple activespreading centers in the hot North Fiji Basin (Southweast Pacific): a possible model for Archeansea floor dynamics? Earth and Planetary Science Letters, 149, 1–13.

Lallemand, S. E., Schnurle, P., and Manoussis, S., 1992. Reconstruction of subduction zonepaleogeometries and quantification of upper plate material losses caused by tectonic erosion.Journal of Geophysical Research, 97(B1), 217–240.

Lallemand, S. E., Schnurle, P., and Malavieille, J., 1994. Coulomb theory applied to accretionaryand non-accretionary wedges – possible causes for tectonic erosion and/or frontal accretion.Journal of Geophysical Research, 99(B6), 12033–12055.

Lallemand, S. E., 1995. High rates of arc consumption by subduction processes: some conse-quences. Geology, 23(6), 551–554.

Lallemand, S. E., Popoff, M., Cadet, J.-P., Deffontaines, B., Bader, A.-G., Pubellier, M., and Rangin,C., 1998. Genetic relations between the central & southern Philippine Trench and the PhilippineTrench. Journal of Geophysical Research, 103(B1), 933–950.


of 16

Lallemand, S., 1999. La subduction oceanique. Amsterdam: Gordon and Breach Science Pub-lishers. 208 pp.

Lallemand, S., Huchon, P., Jolivet, L., and Prouteau, G., 2005a. Convergence lithospherique. Paris:Vuibert. 182 pp.

Lallemand, S., Heuret, A., and Boutelier, D., 2005b. On the relationships between slab dip, back-arcstress, upper plate absolute motion and crustal nature in subduction zones. Geochemistry,Geophysics, Geosystem, 6, 1, doi:10.1029/2005GC000917.

Lallemand, S., Heuret, A., Faccenna, C., and Funiciello, F., 2008. Subduction dynamics revealed bytrench migration. Tectonics, 27, TC3014, doi:10.1029/2007TC002212.

Lay, T., and Kanamori, H., 2011. Insights from the great 2011 Japan earthquake. Physics Today, 64,33–39.

Lithgow-Bertelloni, C., and Richards, M. A., 1998. The dynamics of Cenozoic and Mesozoic platemotions. Reviews of Geophysics, 36(1), 27–78.

Mann, P., and Taira, P., 2004. Global tectonic significance of the Solomon Islands and Ontong JavaPlateau convergent zone. Tectonophysics, 389, 137–190.

McCaffrey, R., 1992. Oblique plate convergence, slip vectors, and forearc deformation. Journal ofGeophysical Research, 97, 8905–8915.

McCaffrey, R., 1997. Statistical significance of the seismic coupling coefficient. SeismologicalSociety of America Bulletin, 87, 1069–1073.

McCaffrey, R., 2008. Global frequency of magnitude 9 earthquakes. Geology, 36(3), 263–266,doi:10.1130/G24402A.

McKenzie, D., and Jackson, J., 2012. Tsunami earthquake generation by the release of gravitationalpotential energy. Earth and Planetary Science Letters, 345–348, 1–8.

Meschede, M., Zweigel, P., and Kiefer, E., 1999. Subsidence and extension at a convergent platemargin: evidence for subduction erosion off Costa Rica. Terran Nova, 11(2/3), 112–117.

Moore, J. C., and Silver, E. A., 1987. Continental margin tectonics: Submarine accretionary prisms.Reviews of Geophysics, 25(6), 1305–1312.

Oleskevich, D. A., Hyndman, R. D., and Wang, K., 1999. The updip and downdip limits to greatsubduction earthquakes: thermal and structural models of Cascadia, south Alaska, SW Japan, andChile. Journal of Geophysical Research, 104, 14,965–14,991.

Pacanovsky, K. M., Davis, D. M., Richardson, R. M., and Coblentz, D. D., 1999. Intraplate stressesand plate-driving forces in the Philippine sea plate. Journal of Geophysical Research, 104(B1),1095–1110.

Rogers, G. C., and Dragert, H., 2003. Episodic tremor and slip on the Cascadia subduction zone: thechatter of silent slip. Science, 300(5627), 1942–1943.

Ruff, L., and Kanamori, H., 1980. Seismicity and the subduction process. Physics of the Earth andPlanetary Intriors, 23, 240–252.

Ruff, L., and Kanamori, H., 1983. Seismic coupling and uncoupling at subduction zones.Tectonophysics, 99, 99–117.

Ruff, L., 1989. Do trench sediment affect great earthquake occurrence in subduction zones?PAGEOPH, 129, 263–282.

Ruff, L. R., 1992. Asperity distributions and large earthquake occurrence in subduction zones.Tectonophysics, 211, 61–83.

Saffer, D. M., and Tobin, H. J., 2011. Hydrogeology and mechanics of subduction zone forearcs:fluid flow and pore pressure. Annual Review of Earth and Planetary Sciences, 39, 157–186.


of 16

Sage, F., Collot, J.-Y., and Ranero, C. R., 2006. Interplate patchiness and subduction-erosionmechanisms: evidence from depth migrated seismic images at the central Ecuador convergentmargin. Geology, 34(12), 997–1000, doi:10.1130/G22790A.1.

Scholl, D. W., von Huene, R., Vallier, T. L., and Howell, D. G., 1980. Sedimentary masses andconcepts about tectonic processes at underthrust ocean margins. Geology, 8, 564–568.

Scholz, C. H., and Campos, J., 1995. On the mechanism of seismic decoupling and back-arcspreading in subduction zones. Journal of Geophysical Research, 100, 22,103–22,115.

Scholz, C. H., 2002. The Mechanics of Earthquakes and Faulting, 2nd edn. Cambridge: CambridgeUniversity Press. 471 pp.

Scholz, C. H., and Campos, J., 2012. The seismic coupling of subduction zones revisited. Journal ofGeophysical Research, 117, B05310, doi:10.1029/2011JB009003.

Shemenda, A., 1992. Subduction: Insights from Physical Modeling. Netherlands: Kluwer AcademicPublishers. Modern approaches in Geophysics. 215 pp.

Sleep, N., and Toksöz, M. N., 1971. Evolution of marginal basins. Nature, 33, 548–550.Stein, S., and Stein, C. A., 1996. Thermo-mechanical evolution of oceanic lithosphere: Implications

for the subduction process and deep earthquakes. In Bebout, G. E., Scholl, D. W., Kirby, S. H.,and Platt, J. (eds.), Subduction: Top to Bottom. Washington, DC: AGU. Geophysical monograph96.

Tatsumi, Y., 1989. Migration of fluid phases and genesis of basalt magmas in subduction zones.Journal of Geophysical Research, 94(B4), 4697–4707.

Turcotte, D. L., and Schubert, G., 1982. Geodynamics: Applications of Continuum Physics toGeological Problems. New York: Wiley. 450 pp.

Tatsumi, Y., and Kogiso, T., 2003. The subduction factory: Its role in the evolution of the Earth’scrust and mantle. In Larter R. D., and Leat P. T.(eds.), Geological Society of London, SpecialPublication, pp. 55–80.

Uyeda, S., 1984. Subduction zones: their diversity, mechanisms and human impacts. GeoJournal,8(4), 381–406.

Vannucchi, P., Remitti, F. and Bettelli, G., 2008. Geological record of fluid flow and seismogenesisalong an erosive subducting plate boundary. Nature, 451, doi:10.1038/nature06486.

Vannucchi, P., Sage, F., Morgan, J. P., Remitti, F., and Collot, J.-Y., 2012. Toward a dynamic conceptof the subduction channel at erosive convergent margins with impla-ications for interplatematerial transfer. Geochemistry, Geophysics, Geosystems, 13, 1, doi:10.1029/2011GC003846.

von Huene, R., Langseth, M., Nasu, N., 1980. Summary, Japan trench transect. Leg 56–57. InScientific Party, Init. Report DSDP, 56–57, part 1. Washington: U.S. Govt. Printing Office,pp. 473–488.

von Huene, R., and Lallemand, S., 1990. Tectonic erosion along convergent margins. GeologicalSociety of America Bulletin, 102, 704–720.

von Huene, R., and Scholl, D. W., 1991. Observations at convergent margins concerning sedimentsubduction, subduction erosion, and the growth of continental crust. Reviews of Geophysics,29(3), 279–316.

von Huene, R., Ranero, C. R., and Vannucchi, P., 2004. Generic model of subduction erosion.Geology, 32(10), 913–916.


of 16

Wadati, K., 1928. Shallow and deep earthquakes. Geophysical Magazine, 1, 162–202.Wu, B., Conrad, C., Heuret, A., Lithgow-Bertelloni, C., and Lallemand, S., 2008. Reconciling

strong slab pull and weak plate bending: the plate motion constraint on the strength of mantleslabs. Earth and Planetary Science Letters, 272, 412–421, doi:10.1016/j.epsl.2008.05.009.


of 16

subduction - géosciences montpellier · large earthquakes, active volcanoes, and sometimes...

Documents