Journal of Asian Earth Sciences 98 (2015) 26–49

Contents lists available at ScienceDirect

Journal of Asian Earth Sciences

journal homepage: www.elsevier.com/locate/jseaes

Review article The 2011 Tohoku earthquake (Mw 9.0) sequence and subduction dynamics in Western Pacific and

Dapeng Zhao

Department of Geophysics, Tohoku University, Sendai 980-8578, article info abstract

Article history: We review recent findings on the causal mechanism of the great 2011 Tohoku earthquake (Mw 9.0) Received 10 September 2014 sequence and related issues on seismic structure and subduction dynamics in Western Pacific and East Received in revised form 20 October 2014 Asia. High-resolution tomography revealed significant lateral heterogeneities in the interplate mega- Accepted 27 October 2014 thrust zone beneath the Tohoku, South Kuril and Southwest Japan forearc regions. Large megathrust Available online 7 November 2014 earthquakes since 1900 generally occurred in or around high-velocity (high-V) patches in the megathrust zone, which may reflect asperities resulting from subducted seamounts, oceanic ridges and other topo- Keywords: graphic highs on the Pacific seafloor. In contrast, low-velocity (low-V) patches in the megathrust zone Tohoku earthquakes may contain more sediments and fluids, where the subducting oceanic plate and the overlying continen- Subduction dynamics Western Pacific tal plate are less coupled or even decoupled. The nucleation of large crustal earthquakes in the Japan East Asia Islands, including the 11 April 2011 Iwaki earthquake (M 7.0) in SE Tohoku, is affected by arc magma Megathrust zone and fluids resulting from slab dehydration. The plate has subducted aseismically down Pacific plate to 430–460 km depth under East China Sea, Tsushima Strait and Japan Sea. A window in the aseismic Philippine Sea slab is detected, which may be caused by splitting of weak parts of the slab at the subduct- Intraplate volcanism ed ridges (e.g., Kyushu-Paula ridge) and hot upwelling in the mantle wedge above the Pacific slab. The Deep earthquakes intraplate volcanism in Northeast Asia is caused by hot and wet upwelling flows in the big mantle wedge above the stagnant Pacific slab in the mantle transition zone. Frequent generation of large deep earth- quakes (>500 km depth) in the Pacific slab may supply additional fluids preserved in the slab to the man- tle wedge under the Changbai volcano, making Changbai the largest and most active intraplate volcano in Northeast Asia. Fluids may be involved in nucleation and rupture processes of all types of earthquakes. Ó 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Contents

1. Introduction ...... 27 2. Tohoku arc and 2011 megathrust earthquakes ...... 28 3. Forearc crustal earthquakes and water wall...... 33 4. South Kuril megathrust zone...... 34 5. Southwest Japan arc...... 35 6. Aseismic deep subduction of Philippine Sea slab...... 37 7. Slab deep subduction and intraplate magmatism ...... 39 8. Discussion and summary...... 40 8.1. Outer-rise earthquakes and slab hydration ...... 40 8.2. Megathrust-zone heterogeneity and megathrust earthquakes ...... 40 8.3. Intraslab earthquakes ...... 42 8.4. Arc magmatism and mantle wedge dynamics...... 42 8.5. Slab dehydration and crustal earthquakes...... 43 8.6. Deep earthquakes, BMW and intraplate volcanism...... 45 8.7. Fluids and earthquakes ...... 45 8.8. Future perspectives ...... 45

E-mail address: [email protected] http://dx.doi.org/10.1016/j.jseaes.2014.10.022 1367-9120/Ó 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49 27

Acknowledgements ...... 45 Appendix A. Supplementary material...... 46 References ...... 46

1. Introduction and Zhao, 2013a). The overall level of seismicity in the Japan Islands has increased significantly soon after the Mw 9.0 main The great 2011 Tohoku-oki earthquake (Mw 9.0) sequence shock (e.g., Okada et al., 2011). wrote a new chapter of active seismotectonics and subduction The Japanese Islands are part of the Western Pacific trench – dynamics in the Western Pacific and East Asian region. This earth- arc–back-arc system, forming a typical subduction zone (Fig. 1). quake sequence, taking place in the Northeast Japan (Tohoku) fore- At least four lithospheric plates exist in and around this region arc area (Figs. 1 and 2), started with a foreshock (M 7.3) at local and they are strongly interacting with each other. The Pacific plate time 11:45 on 9 March 2011 (Shao et al., 2011; Huang and Zhao, is subducting beneath Hokkaido and eastern Honshu at the Kuril 2013a). Its main shock (Mw 9.0) occurred at 14:46 on 11 March, and Japan trenches at a rate of 7–10 cm/year (Bird, 2003). The Phil- which was followed by two big aftershocks at 15:08 off Iwate (M ippine Sea plate is descending beneath Central and SW Japan at the 7.4) and at 15:15 off Ibaraki (M 7.7) on the same day (Fig. 2). All Sagami and Nankai troughs at a rate of 4–5 cm/year. Hokkaido and of the four big events had megathrust mechanisms, which were Tohoku belong to the , whereas SW Japan belongs to caused by sudden ruptures of the boundary between the subduct- the Eurasian (or Amur) plate (e.g., Seno et al., 1996; Bird, 2003). ing Pacific plate and the overlying Okhotsk plate beneath the Because of the strong interactions of the four lithospheric plates, Tohoku forearc. About 40 min after the main shock, a big outer-rise seismicity is very active in this region (Fig. 1b). Approximately event (M 7.5) took place at 15:25 with a normal-faulting mecha- 10,000 earthquakes (M P 1.5) take place every month in and nism (Fig. 2). On April 7, a big intraslab earthquake (M 7.1) around Japan, which have been recorded by the dense and high- occurred within the subducting Pacific plate with a focal depth of sensitivity seismic networks on the Japan Islands (Fig. 1a). Because 70 km and a normal-faulting mechanism. On April 11, one month of the high level of seismicity and availability of the dense and after the main shock, the Iwaki earthquake (M 7.0) took place at high-sensitivity seismic networks (Fig. 1a) operated by the Japan 8 km depth in the upper crust of the Okhotsk plate beneath SE Meteorological Agency (JMA), the Japanese national universities, Tohoku, which was caused by rupture of the Idosawa normal . and the National Research Institute for Earth Science and Disaster Four months after the main shock, on 10 July another megathrust Prevention (Okada et al., 2004), a great amount of high-quality aftershock (M 7.3) occurred in the middle between the Japan seismic data have been accumulated for a long time and they have Trench and the main shock hypocenter (Fig. 2). To date, more than been used to study seismotectonics and the three-dimensional (3- 100 aftershocks with M P 6.0 and tens of thousands of smaller D) seismic structure of the crust and upper mantle of the Japan events have occurred in the Tohoku forearc region (e.g., Huang subduction zone. These studies have been made continuously in

Fig. 1. (a) Distribution of seismic stations on the Japan Islands. (b) Distribution of earthquakes in and around the Japan Islands during June 2006 to September 2010. The color shows the focal depth; its scale is shown at the bottom. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 28 D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49

by these tomographic studies and discuss their seismotectonic and geodynamic implications.

2. Tohoku arc and 2011 megathrust earthquakes

Earthquakes in the crust and in the subducting Pacific and Phil- ippine Sea slabs under the Japan land areas can be located precisely because they occurred directly beneath the dense seismic net- works (Fig. 1a). In contrast, earthquakes under the Pacific Ocean and Japan Sea (Fig. 1b) take place outside the seismic network, hence they have poor hypocenter locations because only first P- and S-wave arrival times are used in the routine earthquake loca- tion by the seismic network (e.g., Okada et al., 2004; Zhao, 2012). Precise hypocenter locations are important in all branches of seis- mology for estimating focal mechanism solutions, rupture process on a fault plane, strong ground motions in the source area, crustal and upper-mantle structure, and so on. Among the four hypocentral parameters (i.e., latitude, longitude, focal depth, and origin time), the focal depth is usually harder to determine for the events outside a seismic network. To locate earthquakes accurately, seismologists have used the so-called depth phases (e.g., pP, sP, etc.) in teleseis- mic distances of thousands of kilometers for large or moderate-size (M > 5.0) suboceanic events (e.g., Herrmann, 1976; Forsyth, 1982; Engdahl and Billington, 1986). These depth phases are reflected waves from the surface (or seafloor) with bouncing points close to the epicenter (Fig. 3d), hence their travel times are very sensitive to the focal depth. Therefore, hypocentral locations of the suboce- anic events, in particular, focal depths, can be constrained well by using the depth-phase data. The use of a depth-phase datum in earthquake location is just like installing a new seismic station close to the epicenter (at the bouncing point on the Earth’s surface or seafloor) (Zhao et al., 2009a; Zhao, 2012). In order to locate precisely the suboceanic events surrounding NE Japan, Umino and Hasegawa (1994) and Umino et al. (1995) detected sP depth phases in a local distance (<300 km) on short- Fig. 2. Map showing epicentral locations of 101 large Tohoku-oki earthquakes (JMA period seismograms of shallow earthquakes that occurred under magnitude P 6.0) during 9 March to 31 December 2011 (Huang and Zhao, 2013a). the Japan Sea and Pacific Ocean recorded by the seismic network The red stars denote the events with M P 7.0: (A) local time 11:45, 9 March (M in Tohoku. Figs. 3 and 4 show a few examples of seismograms in 7.3); (B) 14:46, 11 March (M 9.0); (C) 15:08, 11 March (M 7.4); (D) 15:15, 11 March which sP depth phases are identified. Many studies have shown (M 7.7); (E) 15:25, 11 March (M 7.5); (F) 23:32, 7 April (M 7.1); (G) 17:16, 11 April (M 7.0); (H) 9:57, 10 July (M 7.3). The open circles show the events with that the suboceanic events off the Japan Islands could be relocated 7.0 > M P 6.0. The circle-cross symbols denote seven events which occurred after 9 precisely using the sP depth-phase data, and the location accuracy March but before the M 9.0 main shock. The blue squares show the seismic stations (<3 km) is comparable to that of the earthquakes beneath the seis- used to relocate these events. The red triangles show active arc volcanoes. The solid mic network (Umino and Hasegawa, 1994; Umino et al., 1995; and dotted contour lines show the depths to the upper boundary of the subducting Wang and Zhao, 2005; Gamage et al., 2009; Huang et al., 2011a; Pacific slab. The thick solid curve denotes the . (For interpretation of the references to colour in this figure legend, the reader is referred to the web Huang and Zhao, 2013a,b; Liu et al., 2013a,b). version of this article.) Zhao et al. (2002) determined the first P-wave tomography beneath the Tohoku forearc from the Japan Trench to the Pacific coast using P-wave arrival times of suboceanic events that were relocated precisely with sP depth-phase data, and suggested that the past decades using various seismological methods, making the their approach is a reliable way of tomographic imaging outside Japan Islands the best-studied subduction zone in the world (for a seismic network. Later, both P- and S-wave arrival times from details, see the reviews by Hasegawa et al., 2009; Zhao, 2012). the relocated suboceanic events as well as earthquakes under the During the past three years, many researchers have studied land area were used together to determine 3-D P- and S-wave focal mechanisms and rupture processes of the large Tohoku-oki velocity (Vp, Vs) and Poisson’s ratio images under Hokkaido and earthquakes (see reviews by Koketsu et al., 2011; Lay et al., Tohoku forearc areas (Mishra et al., 2003; Wang and Zhao, 2005; 2012; Tajima et al., 2013, and many related publications such as Zhao et al., 2007a, 2009a) and the back-arc area under the eastern those in two special issues of Earth Planets Space, http:// margin of the Japan Sea (Zhao et al., 2011a). www.earth-planets-space.com/). At the same time, many high- Huang et al. (2011a) determined detailed 3-D Vp and Vs tomog- resolution tomographic studies have been made to investigate the raphy of the entire Tohoku arc using a large number of P- and S- detailed 3-D structure of the crust and upper mantle in the source wave arrival times from local earthquakes that occurred in the zone of the Tohoku-oki earthquakes and surrounding regions, crust and the subducting Pacific slab from the Japan Trench to which have provided important new insights into the nucleation the Japan Sea (Fig. 5). They relocated the suboceanic events pre- mechanism of the interplate and intraplate earthquakes, arc and cisely using P- and S-wave arrival times as well as sP depth-phase back-arc magmatism, and subduction dynamics in Western Pacific data which were measured from the seismograms recorded by the and East Asia. In this article, we review the new findings obtained seismic stations on the Tohoku land area. Their 3-D velocity model D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49 29

Fig. 3. (a and b) Examples of vertical-component seismograms with sP depth phases from two suboceanic earthquakes recorded by the Hi-net seismic stations (open symbols) on Honshu Island as shown in (c). The star in (c) shows the epicenter of the 2011 Tohoku-oki earthquake (Mw 9.0). The origin time and magnitude of each earthquake are shown above the seismograms (a and b). The station code and epicentral distance are shown on the left of each seismogram. (d) A cartoon showing the ray paths of the sP depth phase and direct P-wave from a suboceanic event. After Huang and Zhao (2013b).

Fig. 4. An example of vertical-component seismograms of a crustal earthquake beneath the Japan Sea (Zhao et al., 2011a). Hypocenter parameters of the earthquake are shown in (a). The station codes and epicenter distances are shown on the left (b). Clear sP depth phases are visible at 5–6 s after the first P-wave arrivals. The epicenter (star) and seismic stations (solid squares) are shown in (a). Solid triangles in (a) denote active arc volcanoes.

(Figs. 6 and S1) was used to relocate the 2011 Tohoku-oki earth- earthquakes (Zhao et al., 2011b; Huang and Zhao, 2013a)(Fig. 7). quakes and examine the influence of structural heterogeneity in The tomographic image of the Tohoku forearc is updated by adding the megathrust zone on the generation of the 2011 Tohoku-oki P- and S-wave arrival-time data of the 2011 Tohoku-oki 30 D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49

aftershocks which were well located with sP depth phases (Huang and Zhao, 2013b; Fig. S2), though main features of the tomography remain the same (Fig. S3). Liu et al. (2014) determined detailed 3-D P- and S-wave atten- uation (Qp and Qs) tomography of the crust and upper mantle beneath the entire Tohoku arc from the Japan Trench to the Japan Sea coast using a large number of t⁄ data measured precisely from P- and S-wave spectra of local shallow and intermediate-depth earthquakes. For the suboceanic events used for the Q tomography, they adopted the precise hypocenter parameters determined by Huang and Zhao (2013a,b) with sP depth phases. The pattern of the velocity tomography (Fig. 6) is generally con- sistent with that of the attenuation tomography (Fig. 8), though the Q model has, inevitably, a lower spatial tomography than the velocity model (Liu et al., 2014). The subducting Pacific slab is imaged clearly as a dipping high-V and high-Q zone from the Japan Trench to a depth of 200 km (Figs. 6 and 8). Intermediate-depth earthquakes occur actively in the subducting Pacific slab and form a clear double seismic zone. Prominent low-V and low-Q anoma- lies are visible in the crust and uppermost mantle beneath the active arc volcanoes, and they extend down to 150 km depth in the central part of the mantle wedge. The low-V and low-Q anom- alies reflect the source zone of arc magmatism and volcanism, which are produced by the joint effect of convective circulation process in the mantle wedge and fluids from the dehydration pro- cess of the subducting Pacific slab (e.g., Tatsumi, 1989; Zhao et al., 1992; Hasegawa and Zhao, 1994; van Keken et al., 2011). Low- frequency microearthquakes occur in the lower crust and uppermost Fig. 5. (a) Map view and (b) east–west vertical cross-section of 3204 events under mantle and they are located in or around the low-V/low-Q zones the land area (gray crosses) and 1451 events under the Pacific Ocean and the Japan Sea (open circles) which were used for tomographic imaging by Huang et al. beneath the active arc volcanoes (Figs. 6 and 8 and S1), indicating (2011a). The suboceanic events were relocated precisely using sP depth phases. The that they are associated with the fluids and magmatic activity solid and dashed lines in (a) show the plate boundaries. The red star denotes the beneath the volcanic front and back-arc area (Hasegawa and 2011 Tohoku-oki earthquake (Mw 9.0). (For interpretation of the references to Zhao, 1994; Nakamichi et al., 2003). Fig. S1 shows vertical colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 6. Vertical cross-sections of P-wave tomography along eight profiles shown on the inset map (Huang et al., 2011a). The red and blue colors denote slow and fast velocities, respectively. The velocity perturbation scale is shown on the right. The horizontal bar atop each cross-section shows the land area where seismic stations are deployed. The solid triangles denote active volcanoes within a 30-km width of each profile. The white and red circles denote the relocated suboceanic events with sP depth phase and low-frequency microearthquakes within a 30-km width of each profile, whereas the white dots denote the background seismicity within a 7-km width of each profile. The three curved lines denote the Conrad and Moho discontinuities and the upper boundary of the subducting Pacific slab, whereas the dashed lines denote the inferred lower boundary of the Pacific slab. The red triangles and curved lines on the inset map denote active volcanoes and plate boundaries, respectively. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49 31 cross-sections of Vp and Vs images along the volcanic front and the high-V/high-Q patches or at the boundary between the low-V/ second volcanic front along the Japan Sea coast. Continuous low-V low-Q and high-V/high-Q zones. Only a few megathrust events zones are visible in the central part of the mantle wedge under the occurred in the low-V/low-Q patches (Figs. 7 and 9). second volcanic front, whereas intermittent low-V zones exist The 2011 Tohoku-oki main shock and its foreshock (M 7.3) are beneath each group of active volcanoes along the volcanic front located in a significant high-V/high-Q zone off Miyagi. The off- (Fig. S1), which support the hot finger model proposed by Iwate aftershock (M 7.4) is located at the boundary between the Tamura et al. (2002). off-Sanriku low-V zone and a high-V zone in the north (Fig. 7). Significant lateral variations in seismic velocity and attenuation The largest aftershock (M 7.7) is located at the northern edge of are revealed in the interplate megathrust zone under the Tohoku the off-Ibaraki low-V zone (Fig. 7). Such a pattern of hypocenter forearc (Figs. 7 and 9 and S3). Low-V/low-Q anomalies exist off distribution for the 2011 Tohoku-oki earthquakes is quite consis- Sanriku, Fukushima and Ibaraki. Detailed resolution analyses indi- tent with that of the large earthquakes from 1900 to 2008 (Fig. 7). cate that the main features of the tomographic images in the mega- The low-V/low-Q patches in the megathrust zone (Figs. 7 and 9) thrust zone are reliable (Huang and Zhao, 2013b; Liu et al., 2014). may contain more subducted sediments and fluids associated with There is a correlation between the tomography and the distribu- slab dehydration than the surrounding areas (Mishra et al., 2003; tion of large earthquakes (MJMA P 6.0) that occurred after 1900, Zhao et al., 2011b; Huang and Zhao, 2013a,b; Liu et al., 2014). Thus most of which were megathrust earthquakes (Umino et al., 1990; the subducting Pacific plate and the overriding Okhotsk plate may Usami, 2003; Yamanaka and Kikuchi, 2004; Zhao et al., 2009a, become weakly coupled or even decoupled in the low-V/low-Q 2011b). Most of the large megathrust events are located in the areas. Large-amplitude reflected waves from the slab boundary were detected in a low-seismicity area under the Sanriku forearc (Fujie et al., 2002; Ye et al., 2012), as were some slow and ultra- slow megathrust earthquakes (Heki et al., 1997; Kawasaki et al., 2001). Both the seismic reflectors and the slow megathrust events were associated with fluids at the slab boundary (Fujie et al., 2002; Kawasaki et al., 2001), and they are all located in the off-Sanriku low-V/low-Q zone. In contrast, the high-V/high-Q patches in the megathrust zone (Figs. 7 and 9) may result from subducted oceanic ridges, sea- mounts and other topographic highs, as well as compositional vari- ations, in the seafloor of the Pacific plate, and they have become asperities where the subducting Pacific plate and the overriding Okhotsk plate are strongly coupled (Zhao et al., 2011b). Thus, tec- tonic stress tends to accumulate at those asperities for a relatively long time during subduction, leading to the nucleation of large and great megathrust earthquakes there. The off-Miyagi high-V zone, where the Tohoku-oki main shock and its largest foreshock occurred (Fig. 7), corresponds to the area with large coseismic slips (>30 m) during the Tohoku-oki main shock (e.g., Lay et al., 2011; Koper et al., 2011; Shao et al., 2011; Iinuma et al., 2012) (Fig. 10). This result indicates that the off-Miyagi high-V zone rep- resents a large asperity or a cluster of asperities in the megathrust zone which ruptured during the M 9.0 main shock (Zhao et al., 2011b). The seismic velocity and Q tomographic images of the mega- thrust zone and their correlation with the distribution of mega- thrust earthquakes (Figs. 7 and 9) suggest varying degrees of interplate seismic coupling from north to south in the Tohoku fore- arc, which may control the nucleation of megathrust earthquakes. The great 2011 Tohoku-oki earthquake sequence may be related to such a process. Differences in the interplate seismic coupling could result from variations in the frictional behavior of materials (e.g., Pacheco et al., 1993; Heki et al., 1997; Kato and Hirasawa, 1997; Miura et al., 2003; Lay et al., 2012; Romano et al., 2014). The veloc- ity and Q variations in the megathrust zone may be a manifestation of such variations in the frictional behavior. The pattern of velocity tomography (Fig. 7) is also consistent with the results of multi- Fig. 7. Epicentral distribution of large earthquakes (M P 6.0) and P-wave tomog- channel seismic surveys which revealed along-arc variations of int- raphy in the megathrust zone directly above the upper boundary of the subducting erplate sediments at the Japan Trench (Tsuru et al., 2002; Fujie Pacific slab from the Japan Trench to the 60-km depth contour of the Pacific slab et al., 2013a). The sediments near the trench axis are very thick (Zhao et al., 2011b; Huang and Zhao, 2013a). The red and blue colors denote low off Sanriku, very thin or none off Miyagi, and moderately thick and high velocities, respectively. The velocity perturbation scale is shown at the bottom. A–C denote three low-velocity patches in the megathrust zone. The stars but extend for a wide range in the forearc off Fukushima and denote megathrust earthquakes which occurred during 9 March to 31 December Ibaraki (Fig. S4). 2011. The open circles denote large earthquakes (M P 6.0) from 1900 to 2008, most These results suggest that the structural heterogeneity in the of which were megathrust earthquakes. The solid and dotted contour lines show megathrust zone may control the interplate seismic coupling the depths to the upper boundary of the Pacific slab. The thick solid curve denotes and the nucleation of megathrust earthquakes. If this is true and the Japan Trench. The black triangles show active arc volcanoes. (For interpretation of the references to colour in this figure legend, the reader is referred to the web if the structural heterogeneity mainly results from the structure version of this article.) of the upper boundary of the subducting Pacific plate, then it has 32 D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49

Fig. 8. Vertical cross-sections of S-wave attenuation tomography along the profiles as shown on the inset map (Liu et al., 2014). The surface topography and the land area (the black horizontal bar) along each profile are shown atop each cross-section. The red triangles denote active arc volcanoes within a 20-km width of each profile. The background seismicity and low-frequency microearthquakes that occurred within a 20-km width of each profile are shown in white crosses and red dots, respectively. The white dots denote the earthquakes within a 20-km width of each profile, which were used in the tomographic inversion. The three curved lines show the Conrad and Moho discontinuities and the upper boundary of the Pacific slab, whereas the dotted lines denote the inferred lower boundary of the Pacific slab. The red star denotes the 2011 Tohoku-oki earthquake (Mw 9.0). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 9. (a) P and (b) S wave attenuation (Q) tomography of the Tohoku megathrust zone (Liu et al., 2014). The black stars denote megathrust earthquakes (M P 7.0) during January 1900 to February 2011, whereas the red stars denote megathrust events (M P 7.0) during March 2011 to December 2013. The earthquake magnitude and Q perturbation scales are shown at the bottom. The black triangles denote active arc volcanoes. The sawtooth line indicates the Japan Trench. The dashed lines show the depth contours to the upper boundary of the subducting Pacific slab. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49 33

and the subsequent , which is located only 60 km north- east of Iwaki (Fig. 11). Tong et al. (2012) determined high-resolution Vp and Vs tomography of the crust and upper mantle beneath the Iwaki and FNPP area using a large number of P- and S-wave arrival-time data from local shallow and intermediate-depth earthquakes which occurred from June 2002 to October 2011. They revealed significant low-V and high Poisson’s ratio (high-r) anomalies in the crust and upper mantle beneath the Iwaki source zone. The Pacific slab is imaged as a high-V zone, and prominent low-V anomalies are revealed in the crust and mantle wedge under the active Nasu and Azuma volcanoes (Fig. 11), similar to the whole- Tohoku model (Fig. 6). Under the Iwaki hypocenter, a low-V zone is visible in the lower crust and mantle wedge and it extends down to the subducting Pacific slab (Fig. 11a). A thin, vertical low-V anomaly is revealed in the lower crust and upper mantle beneath the Futaba fault which is located 6 km west of FNPP, and the anomaly is connected with the Pacific slab (Fig. 11b and d). These features have been confirmed by detailed resolution analyses and synthetic tests (Tong et al., 2012). The temperature is low under the forearc area as compared with that under the volcanic front and the back-arc area, hence magma cannot be produced and so no volcano exists in the forearc (see Fig. 2 on the distribution of active arc volcanoes in Tohoku). Because the Iwaki earthquake occurred in the Tohoku forearc and 70 km away from the volcanic front (Figs. 11 and S5), the low- V and high-r anomaly beneath the Iwaki hypocenter may reflect fluids that affected the rupture nucleation, similar to the seismo- genic process in the source zone of the 1995 Kobe earthquake which occurred in the SW Japan forearc (Zhao et al., 1996, 2010; Salah and Zhao, 2003; Tong et al., 2011). As the Pacific plate sub- Fig. 10. Comparison of P-wave attenuation (Qp) tomography of the Tohoku ducts, the temperature and pressure in the subducting slab gradu- megathrust zone with coseismic slip distributions of the M 7.3 foreshock (black ally increase, causing hydrated minerals within the slab to undergo contours with interval of 0.5 m) (Ohta et al., 2012), the Mw 9.0 main shock (blue contours with interval of 10 m) (Iinuma et al., 2012), and two large aftershocks (M dehydration decomposition. This process generates aqueous fluids 7.4 and 7.7) on 11 March 2011 (pink contours with interval of 0.2 m and 1.26 m, which are less dense than the surrounding rock and so can migrate respectively) (Munekane, 2012; Kubo et al., 2013). The red contours (with interval upward to the overlying crust. When the fluids enter an active fault of 1 m) denote the afterslip distribution from 2 March to 12 October 2011 (Ozawa (such as the Idosawa fault) in the crust, fault-zone frictions will be et al., 2012). The other labeling is the same as that in Fig. 9. After Liu et al. (2014). (For interpretation of the references to colour in this figure legend, the reader is reduced. This process, together with the exertion of horizontally referred to the web version of this article.) extensional stress regime induced by the Tohoku-oki main shock, resulted in reactivation of the Idosawa normal fault, leading to the 2011 Iwaki earthquake and its aftershocks (Tong et al., 2012). an important implication that the distribution pattern of the mega- Previous studies have already found that crustal fluids were thrust earthquakes (Fig. 7) will remain the same in the coming involved in several large crustal earthquakes in the Japan Islands 10,000 years. The subduction rate of the Pacific plate is 7–10 cm/ (e.g., Wang and Zhao, 2006a,b; Xia et al., 2008a; Zhao et al., year at the Japan Trench (Bird, 2003), hence the Pacific slab 2010; Cheng et al., 2011; Padhy et al., 2011; Wei and Zhao, 2013). beneath the Tohoku arc would move less than 1 km in the Similar to the Idosawa fault, the active Futaba fault near FNPP is 10,000 years from now. also underlain by a significant low-V and high-r anomaly (Fig. 11b and d). Based on the above discussion, the anomaly under Futaba 3. Forearc crustal earthquakes and water wall may be also associated with the ascending fluids from the subduct- ing Pacific slab. The ascending fluids may be able to enter the Futa- The Iwaki earthquake (M 7.0) on 11 April 2011 was one of the ba fault and trigger a large crustal earthquake in the future, just major aftershocks following the Tohoku-oki main shock and the like the 2011 Iwaki earthquake. Therefore much attention should strongest one hit the Japan land area (Figs. 11 and S5). This large be paid to the seismic safety in and around the FNPP site from crustal earthquake occurred at a depth of 8.0 km and was located now. Tong et al. (2012) also conducted separate tomographic 200 km southwest of the Tohoku-oki main shock. It was caused inversions using data from events which occurred before 11 March by normal faulting with some strike-slip component along the 2011 and those after that date. Their results (Fig. S6) show that the Idosawa fault (Fig. S5). About 11-km long coseismic surface rup- tomographic images are generally similar to each other, indicating ture was recognized along the Idosawa fault, which was a surface that the low-V and high-r anomalies under the Idosawa and Futa- manifestation of the fault reactivation associated with the event ba faults had already existed before the great 2011 Tohoku-oki (Kato et al., 2013). This normal-faulting crustal earthquake is in earthquake. contrast to the compressional stress regime in Tohoku and may Zhao et al. (2015) presented detailed Vp, Vs and Poisson’s ratio reflect enhanced extensional stress in the overriding plate induced (r) images in source areas of 26 large crustal earthquakes (M 6.0– by the Tohoku-oki main shock (Asano et al., 2011). The variations 7.2) which occurred in Tohoku during 120 years from 1894 to in stress field give rise to high seismicity in the Iwaki source area 2014. Prominent low-V and high-r anomalies are revealed in the (Tong et al., 2012). The disabled Fukushima nuclear power plant crust and mantle wedge under the source areas. Beneath the volca- (FNPP) suffered major damage from the Tohoku-oki earthquake nic front and back-arc areas, the low-V and high-r zones reflect 34 D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49

Fig. 11. Vertical cross-sections of (a and b) P-wave and (c and d) S-wave velocity images along the profiles AB and CD as shown on the inset map (modified from Tong et al., 2012). The small white dots denote seismicity during 3 June 2002 to 27 October 2011, which are located within a 20-km width of each profile. The purple star and square symbols denote the 2011 Iwaki earthquake (M 7.0) and the Fukushima nuclear power plant (FNPP), respectively. The red triangles represent active arc volcanoes. The three dashed lines denote the Conrad and Moho discontinuities and the upper boundary of the Pacific slab. The red and blue colors denote low and high velocities, respectively. The velocity perturbation (in %) scale is shown at the bottom. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) arc-magma related high-temperature anomalies which cause the Eurasian (or Amur) plate (e.g., Zhao et al., 2011a). Earthquakes locally thinning and weakening of the brittle seismogenic layer occur actively in the crust and the subducting Pacific slab down to above them. Low-frequency micro-earthquakes are observed in a depth of >300 km. In this region, large triggered by the lower crust and uppermost mantle in or around the low-V megathrust earthquakes have recurred at 500 year interval with zones, which reflect ascent of arc magma and fluids from the man- the most recent event in the 17th century (Nanayama et al., tle wedge to the crust (e.g., Hasegawa and Zhao, 1994; Xia et al., 2003; Satake et al., 2008). 2007; Zhao et al., 2011c). As mentioned above, because no volcano So far many studies have been made on the crust and upper and magma exist in the forearc area due to low temperature there, mantle structure beneath the South Kuril arc (e.g., Katsumata the low-V zones in the forearc reflect fluids from the Pacific slab et al., 2006; Miller et al., 2006; Wang and Zhao, 2009; Kita et al., dehydration. They proposed that the ascending fluids have pro- 2012), which have improved our understanding of seismotectonics duced a water wall in the mantle wedge and lower crust beneath and subduction dynamics of the region. However, these previous the forearc, which affects the generation of large crustal earth- studies mainly focused on the region right beneath Hokkaido quakes in the Tohoku forearc region (Fig. 21b). Island where a dense seismic network exists (Fig. 1a). A few seis- mic surveys were conducted for the forearc region east off Hokka- ido, but they were limited in both space and time (e.g., Murai et al., 4. South Kuril megathrust zone 2003; Nakanishi et al., 2004; Machida et al., 2009; Azuma et al., 2012). The detailed 3-D velocity structure and dynamics of the In the South Kuril arc, the Pacific plate is subducting beneath entire South Kuril arc are still not very clear, because few seismic the Okhotsk plate at the Kuril–Japan Trench (Figs. 1 and 12). In stations exist in the Pacific Ocean and the Japan Sea (Fig. 1a). the Japan Sea west off Hokkaido and Tohoku, an active nascent Liu et al. (2013a) determined detailed 3-D Vp and Vs tomogra- plate boundary has formed that separates the Okhotsk plate from phy of the crust and upper mantle under the entire South Kuril arc D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49 35

Fig. 12. (a) Map view and (b) east–west and (c) north–south vertical cross-sections showing the distribution of earthquakes used by Liu et al. (2013a) for tomographic inversion. The black crosses denote events that occurred beneath the seismic network and were located using P and S wave arrival times. The red and green dots denote suboceanic events that occurred outside the seismic network and were relocated using sP depth phase data. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) from the Kuril–Japan Trench to the back-arc area under the Japan megathrust zone may contain more subducted sediments and flu- Sea and the Sea of Okhotsk using a great number of high-quality ids released from the slab dehydration. arrival-time data from many local suboceanic earthquakes which The Hidaka collision zone is situated at the southern Hidaka are relocated precisely with sP depth phases (Figs. 12 and S7). In Mountains, central Hokkaido Island, where the Kuril forearc sliver addition, they also determined the first 3-D P-wave anisotropy collides with the Tohoku arc, and the lower crust of the Kuril arc tomography under the entire South Kuril arc. They collected and thrusts upon the Tohoku arc (Kimura, 1996). Liu et al. (2013a) used much more data than all of the previous tomographic studies revealed a clear correlation between the seismic heterogeneity and of this region. the Cenozoic thrust faults that may have contributed to this arc– Strong lateral heterogeneities exist in the megathrust zone arc collision. In the Hidaka collision zone, a prominent high-V zone under the Kuril forearc, and the Vp and Vs images are generally thrusts upon a low-V zone, and their boundary is the Hidaka main similar to each other (Fig. 13). Three prominent high-V zones sep- thrust corresponding to the fault plane of the 1970 Hidaka earth- arated by low-V anomalies are visible in the megathrust zone. quake (MJMA 6.7) (Kita et al., 2012; Liu et al., 2013a). The high-V zone Large earthquakes (MJMA P 6.0) in the forearc region during probably represents the westward thrust of the lower crustal rocks 1900–2011 are mainly located in or around high-V zones of the Kuril arc, and the low-V zone may reflect the structure of the (Fig. 13a and b), similar to those in the Tohoku forearc (Fig. 7). north–south fold-and-thrust belt with thick sediments on the west The three high-V zones generally coincide with the areas with of Hidaka Mountains (e.g., Arita et al., 1998; Iwasaki et al., 2004; large coseismic slips of the megathrust earthquakes (Fig. 13c and Itoh et al., 2005). The high-V zone extends southward to the source d) (Nagai et al., 2001; Yamanaka and Kikuchi, 2003, 2004; areas of the 1952 and 2003 Tokachi-oki earthquakes, which may Katsumata and Yamanaka, 2006; Yamanaka, 2006). Similar to the have contributed to the formation of the asperity in the megathrust Tohoku megathrust zone, the high-V zones may reflect the sub- zone (Liu et al., 2013a). The boundary of the high-V and low-V zones ducted seamounts or other topographic highs on the Pacific sea- in Hokkaido coincides with a north–south geological boundary sep- floor, and they may have formed asperities in the megathrust arating the Tohoku arc from the Kuril arc (Kimura, 1986, 1996). zone where the subducting Pacific plate is strongly coupled with These results suggest that high-resolution seismic tomography can the overriding Okhotsk plate (Liu et al., 2013a). Stress concentra- well explain local geological features. tion occurs at the asperities with the subduction of the Pacific plate, which breeds megathrust earthquakes (Bowman and King, 5. Southwest Japan arc 2001). Thus, large slips occurred in and around the high-V zones (asperities) when the megathrust earthquakes take place. Along the off SW Japan, the relatively young However, the ruptures of megathrust earthquakes seem to start (15–50 Ma) Philippine Sea (PHS) plate has been subducting at the margins of the high-V zones, adjacent to which significant beneath the Eurasian (or Amur) plate (e.g., Deschamps and low-V anomalies exist (Fig. 13c and d). The low-V patches in the Lallemand, 2002; Hall, 2002; Sdrolias et al., 2004)(Fig. 14). Many 36 D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49

Fig. 13. (a and c) P-wave and (b and d) S-wave velocity tomography of the megathrust zone directly above the upper boundary of the subducting Pacific slab. The black stars in (a and b) denote megathrust earthquakes (M P 6.0) during 1901–2011. The earthquake magnitude and velocity perturbation scales are shown at the bottom. The red contour lines in (c and d) denote coseismic slip distributions of the megathrust earthquakes in 1931, 1968 and 1994 (Nagai et al., 2001; Yamanaka and Kikuchi, 2004), in 2003 (Yamanaka and Kikuchi, 2003), and in 1973 and 2004 (Katsumata and Yamanaka, 2006; Yamanaka, 2006). The inner contour lines denote larger slips. The black triangles denote active arc volcanoes. The sawtooth line indicates the Kuril–Japan Trench. The black dashed lines show depth contours to the upper boundary of the subducting Pacific slab. After Liu et al. (2013a). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

active arc volcanoes exist and they form a clear volcanic front S8). They collected and used much more data than all the previous (Fig. 14d). Earthquakes occur to a depth of 200 km under Kyushu tomographic studies of this region, hence their tomography shows and 80 km beneath Shikoku and SW Honshu. Deep earthquakes clearly the high-V subducting PHS slab and low-V anomalies in the under this region take place within the subducting Pacific slab crust and mantle wedge under the active arc and back-arc volca- which is located below the and the subducting noes. They also revealed low-V zones in the Nankai forearc region, PHS slab (Figs. 16 and 17). especially off Kyushu and Shikoku Islands (Fig. 15), which are gen- Many local tomographic studies have been made for SW Japan erally consistent with the tomographic results obtained with the (see recent reviews by Hasegawa et al., 2009; Zhao et al., 2011c; OBS data (Tahara et al., 2008). The forearc low-V zones may reflect Zhao, 2012). P-wave anisotropy tomography revealed trench- a highly serpentinized and hydrated forearc mantle wedge result- normal fast-velocity directions in the mantle wedge under Kyushu, ing from the dehydration of the young and warm PHS slab which may reflect corner flow driven by the active subduction of (Hyndman and Peacock, 2003; Xia et al., 2008b). The serpentinized the PHS plate (Ishise and Oda, 2008; Wang and Zhao, 2012, area has a low strength and so controls the down-dip limit of 2013). These previous tomographic studies mainly focused on megathrust earthquakes. Repeating earthquake activity seems to the SW Japan land area where the dense seismic network share the same down-dip limit under the Hyuganada area in the (Fig. 1a) allows for high-resolution seismic imaging. A few active- Kyushu forearc (Yamashita et al., 2012). This feature is different source seismic surveys were conducted in the Nankai forearc from that in Tohoku and South Kuril which are underlain by the region (e.g., Kodaira et al., 2000, 2002; Bangs et al., 2009). old and cold Pacific slab. However, a similar feature was revealed Liu et al. (2013b) determined detailed 3-D Vp and Vs tomogra- in central Cascadia where the young Juan de Fuca slab is subduct- phy of the entire Nankai subduction zone from the Nankai Trough ing beneath the North American plate (e.g., Bostock et al., 2002). to the Japan Sea using a large number of high-quality arrival-time Small earthquakes occur actively within the subducting PHS slab data from many local inland earthquakes as well as suboceanic under the Nankai forearc, which may be related to the slab dehy- events which are relocated with sP depth phases (Figs. 14 and dration (Xia et al., 2008b; Zhao et al., 2011c; Liu et al., 2013b). D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49 37

Fig. 14. (a) Map view and (b) east–west and (c) north–south vertical cross-sections showing the distribution of earthquakes used by Liu et al. (2013b) for tomographic inversion. The green and yellow dots denote events that occurred beneath the seismic network and were located using P and S wave arrival times. The red and purple dots denote suboceanic events that occurred outside the seismic network and were relocated using sP depth phase data. (d) Distribution of active arc volcanoes on the Japan Islands and intraplate volcanoes in Northeast Asia (red triangles). The green hatched part denotes the SW Japan arc. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Vp and Vs images in the Nankai megathrust zone are generally 6. Aseismic deep subduction of Philippine Sea slab similar to each other, which exhibit strong lateral heterogeneities (Fig. 15). Megathrust earthquakes (M P 6.0) during 1901–2011 The seismicity within the subducting PHS slab ends at a depth are mainly located in high-V patches, or at the boundary between of 60–80 km under western Honshu and at a depth of 180 km the high-V and low-V zones (Fig. 15), similar to those in the under Kyushu (Figs. 1 and S9b). Does the tip of the PHS slab reach Tohoku and Kuril arcs (Figs. 7 and 13). those depths or has the PHS slab subducted aseismically to a Three prominent seamount chains exist on the PHS plate close greater depth? P-wave tomography of the Japan subduction zone to the Nankai Trough. From the west to east, they are the Kyu- down to a depth of 700 km was determined by joint inversions shu-Palau ridge, the Kinan seamount chain, and the Izu-Bonin arc of local and teleseismic data recorded by the dense Japanese seis- (Fig. 15). The Kyushu-Palau ridge separates the younger Shikoku mic networks, revealing that the PHS slab under SW Japan has sub- Basin from the older West Philippine Basin (Okino et al., 1999; ducted aseismically down to a depth of 430 km under East China Deschamps and Lallemand, 2002). Active-source seismic surveys Sea off Kyushu, Tsushima Strait, and the Japan Sea off western revealed an irregular Vp structure in the northwestern extension Honshu (Abdelwahed and Zhao, 2007; Zhao et al., 2012a,b). The of the Kyushu-Palau ridge beneath Hyuganada, which may be asso- tomographic images of the PHS slab are further improved by ciated with the subducted part of this ridge (Nishizawa et al., Huang et al. (2013) who collected and combined a large number 2009). They also found high-V anomalies north of this irregular of high-quality arrival-time data of local and teleseismic events Vp structure, where the 1968 Hyuganada earthquake (M 7.5) recoded by the dense seismic networks in both SW Japan and occurred (Fig. 15c and d). A seismic reflection survey also revealed South Korea (Fig. 16). This new result shows that, beneath the seis- a subducting seamount under the forearc off Shikoku Island mic PHS slab, a high-V zone is visible down to a depth of 430 km (Kodaira et al., 2002; Fig. 15c and d). This seamount probably under the Japan Sea off western Honshu and down to a depth of belongs to the Kinan seamount chain which originated from the 460 km under western Kyushu, indicating that the aseismic igneous activity after back-arc spreading of the Shikoku Basin PHS slab has reached the mantle transition zone (Fig. 16). (Kobayashi et al., 1995; Okino et al., 1999). The Izu-Bonin arc Along the Nankai Trough off SW Japan, the PHS plate is com- marks the boundary between the Shikoku Basin and the Pacific posed of several blocks with ages increasing from the east to west, plate. The Zenisu ridge is a sub-branch of this arc and subparallel including the Izu-Bonin arc and back-arc (0–2 Ma), the Shikoku with the Nankai Trough (Fig. 15). Seismic reflection and refraction Basin (15–30 Ma), the Kyushu-Palau Ridge, and the Amami Plateau surveys revealed a subducting ridge subparallel with the Zenisu (40–49 Ma) (Hilde and Lee, 1984; Hall et al., 1995; Huang et al., ridge beneath the Nankai accretionary wedge (Park et al., 2004). 2013). The subduction rate of the PHS plate also increases from The estimated locations of these subducted ridges and seamounts 28 mm/year at the to 50 mm/year off southern are generally consistent with those of the high-V patches in the Kyushu (Seno et al., 1993; Bird, 2003). The subduction of the PHS megathrust zone revealed by Liu et al. (2013b) (Fig. 15). plate may have occurred intermittently through the late Cenozoic, 38 D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49

Fig. 15. (a and c) P-wave and (b and d) S-wave velocity tomography of the megathrust zone along the upper boundary of the subducting Philippine Sea slab beneath the Nankai forearc (Liu et al., 2013b). The red and blue colors denote low and high velocities, respectively. The red stars denote large earthquakes (M P 6.0) during 1900–2011. The earthquake magnitude scale is shown in (a). The velocity perturbation scales is shown in (b). The areas shown in yellow dashed contour lines in (c and d) denote the inferred subducted oceanic ridges and seamounts. From the west to east, they are the subducted part of the Kyushu-Palau ridge (Nishizawa et al., 2009), the Kinan seamount chain (Kodaira et al., 2002) and the Izu-Bonin arc (Park et al., 2004). The unsubducted parts of these topographic highs on the ocean floor are also shown in (c and d). Coseismic slips of the large earthquakes in 1944 (blue lines, Kikuchi et al., 2003), 1946 (white lines, Sagiya and Thatcher, 1999), 1968 (red lines, Yagi et al., 1998) and 1996 (red lines, Yagi et al., 1999) are also shown in (c and d). The inner contour lines denote larger coseismic slips (see the above references for details). The solid triangles denote active arc volcanoes. The sawtooth line denotes the Nankai Trough. The dashed lines show depth contours of the upper boundary of the subducting Philippine Sea plate. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) depending on the rates and modes of destruction and formation of dipping angle of the PHS slab changes from Shikoku-Chugoku to plates in the Philippine Sea (Uyeda and Miyashiro, 1974; Hirahara, Kyushu (Fig. 17), which may also facilitate the segmentation of 1981). The tomographic result (Fig. 16) shows that the total length the aseismic PHS slab (Zhao et al., 2012a). In addition, the PHS slab of the PHS slab under SW Japan is 800–900 km from the Nankai window is located directly above the hinge portion of the Pacific Trough to the tip of the aseismic slab. Suppose this part of the PHS slab which becomes flat at 575 km depth (Fig. 17). Recent com- slab started its subduction 40–15 Ma ago, then its average subduc- puter simulations (Kameyama and Nishioka, 2012; He, 2014) show tion rate is 20–60 mm/year from the tomographic results (Zhao that ascending hot flows can occur easily in the mantle wedge et al., 2012a; Huang et al., 2013). This value is roughly consistent above the hinge portion of the Pacific slab, which may have caused with the estimates (28–50 mm/year) by the above-mentioned the Changbai intraplate volcanism, well explaining a low-V anom- studies using other approaches. Note that the convergence rate aly in the upper mantle under the Changbai volcano as revealed by may have varied in different periods of the PHS plate subduction seismic tomography (Zhao et al., 2004, 2009b; Lei and Zhao, 2005; (Uyeda and Miyashiro, 1974; Hirahara, 1981). Duan et al., 2009). In and around the PHS slab window, low-V The morphology of the aseismic PHS slab under SW Japan has anomalies exist (A2–A3 in Fig. 17), which may reflect ascending been estimated from the obtained images of teleseismic tomogra- hot flow in the mantle wedge above the Pacific slab. The slab win- phy (Zhao et al., 2012a; Huang et al., 2013). The aseismic slab is dow may be caused the joint effects of these factors. The opening of visible in two areas: one is under the Japan Sea off western Hon- a slab window can cause various geological processes including shu, the other is under the East China Sea off western Kyushu magmatism, changes in plate kinematics and deformation, (Fig. 17). However, the aseismic slab disappears between the two mineralization, high-temperature metamorphism, and changes in areas, where a slab window may have formed (Fig. 17). This feature sedimentary basin evolution (see Eyuboglu, 2013 and many has been confirmed by detailed resolution analyses and synthetic references cited therein). tests (Zhao et al., 2012a; Huang et al., 2013). The PHS slab has a relatively simple geometry down to a depth The cause of the slab window is still not very clear. One possi- of 250 km beneath Kyushu, whereas the aseismic PHS slab at bility is that the PHS slab is segmented and broken because of depths of 280–460 km exhibits a change in its morphology off wes- the subduction of the Kyushu-Palau Ridge and the Kinan Seamount tern Kyushu (Fig. 17). It is still unclear what caused the slab geom- Chain (Figs. 15 and 17) where the PHS slab may be much weaker etry change. One possibility is that the hot mantle upwelling above than the surrounding parts of the slab, and thermal and mechanical the Pacific slab pushed the PHS slab from below. This together with erosions may occur there, resulting in the slab window. The the compression from the Pacific plate may have changed the PHS D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49 39

Fig. 16. Vertical cross-sections of P-wave tomography along the profiles shown on the inset map (Huang et al., 2013). The red and blue colors denote low and high velocities, respectively. The velocity perturbation scale is shown on the right. The red and black triangles denote active volcanoes. The surface topography is shown atop each cross- section. The three dashed lines in each cross-section represent the Moho, 410- and 660-km discontinuities. The white dots denote seismicity that occurred within a 20-km width of each profile. The estimated upper boundary of the subducting Philippine Sea slab is shown in a dashed line. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) slab geometry, which may be also associated with the formation of local and regional earthquakes recorded by over 2000 seismic sta- the slab window (Zhao et al., 2012a,b; Huang et al., 2013). To clar- tions in East Asia (Fig. S9), which has improved the earlier regional ify these issues, it is necessary to conduct geophysical, geochemical tomography by Huang and Zhao (2006). The new tomography and computer-simulation studies considering the subduction model has confirmed the existence of the stagnant slab under history of the PHS plate. Korea and East China and low-V anomalies in the BMW under the active intraplate volcanoes, such as Changbai, Wudalianchi, 7. Slab deep subduction and intraplate magmatism Jeju, and Ulleung. (Fig. 18). The subducting PHS slab is also imaged clearly, and it has subducted aseismically down to the MTZ According to the studies of seismic tomography (e.g., Zhao, (Fig. 18k–r), being consistent with the local-scale tomography 2001, 2004; Fukao et al., 2001; Huang and Zhao, 2006; Wei et al., under SW Japan (Fig. 16). 2012; Zhao et al., 2004, 2013) and waveform modeling (e.g., Among the active intraplate volcanoes in NE Asia, Changbai is Tajima et al., 2009; Ye et al., 2011; Li et al., 2013), the Pacific plate the largest and most active one and it has erupted several times is subducting beneath the Japan Islands and the PHS plate, and in the past 2000 years (e.g., Liu, 1999; Xu et al., 2012; Wei et al., then it becomes stagnant in the mantle transition zone (MTZ) 2013a). Then a question arises: why is Changbai so active and dis- under the Korean Peninsula and East China. A big mantle wedge tinct from the other intraplate volcanoes in NE Asia? Is there any (BMW) has formed in the upper mantle above the dipping Pacific local cause of the Changbai volcanism in addition to the hot slab and the stagnant slab (Fig. 21a), and hot and wet upwelling upwelling in the BMW? Zhao and Tian (2013) investigated this flows in the BMW and perhaps deep slab dehydration as well have issue and proposed a model which shows a link between deep caused the intraplate volcanoes in East Asia (Zhao et al., 2004, earthquakes in the Pacific slab with the Changbai volcanism 2007b; Lei and Zhao, 2005). This BMW model for the intraplate vol- (Fig. 21a). canism and mantle dynamics in East Asia has been supported by Very deep earthquakes (500–600 km depths) in the Pacific slab petrological and geochemical studies (e.g., Zou et al., 2008; under East Asia occur 200–300 km to the east of the Changbai vol- Ohtani and Zhao, 2009; Kuritani et al., 2009, 2011, 2013; cano (Fig. 19). Some of the deep events are quite large (M > 7.0) Sakuyama et al., 2014) and various geophysical observations (see and they take place frequently in the region. A high-resolution Zhao and Tian, 2013 for a recent review). local tomography down to a depth of 650 km under Changbai Wei et al. (2012) determined an updated P-wave mantle tomog- was obtained by using a large number of high-quality local and raphy under East Asia using a better data set of arrival times from teleseismic data recorded by 645 seismic stations of the local 40 D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49

deep portion of the oceanic lithosphere along the active normal faults which generated the large outer-rise earthquakes (e.g., Peacock, 2001; Ranero et al., 2003; Ranero and Sallares, 2004; Zhao et al., 2007b; Faccenda et al., 2008, 2009; Fujie et al., 2013b). The seawater or fluids may be preserved in the active faults even after the Pacific plate subducts into the mantle. Among the large deep events which occur frequently in the subducting Pacific slab under the Japan Sea and the NE Asian margin, at least some of them may be caused by reactivation of the faults preserved in the subducting slab (e.g., Silver et al., 1995; Zhao et al., 2007b), and the fluids preserved in the faults within the slab may cause the observed non-double-couple components in the deep earthquake faulting (e.g., Frohlich, 1994, 2006; Julian et al., 1998; Miller et al., 1998). The fluids preserved in the slab may be released to the overlying mantle wedge through the large deep earthquakes. Because the deep events occur frequently in the vicinity of the Changbai volcano, much more fluids could be supplied to this vol- cano than other areas in NE Asia, making Changbai the largest and most active intraplate volcano in the region. Zhao and Tian (2013) also showed some examples indicating that not only (some) deep earthquakes but also (some) intermediate-depth earthquakes beneath the Tohoku and Kuril arcs were caused by reactivation of preexisting faults preserved in the subducting Pacific slab.

8. Discussion and summary Fig. 17. Depth contours to the upper boundary of the subducting Pacific slab (the blue dashed lines) and that of the subducting Philippine Sea (PHS) slab. The bold The active northwestward subductions of the Pacific and Philip- blue line denotes the 575-km depth contour of the Pacific slab which becomes pine Sea plates beneath the Okhotsk and Eurasian plates have stagnant in the mantle transition zone to the west of this contour line, according to played a key role in the seismic and volcanic activities, and causing Zhao (2004). The gray lines denote the seismic parts of the PHS slab estimated from the seismicity in the PHS slab and local-earthquake tomography (Nakajima et al., significant structural heterogeneities and dynamic evolution in the 2009). The red lines show the upper boundary of the aseismic PHS slab estimated crust and mantle beneath Western Pacific and East Asia. In the fol- from teleseismic tomography (Huang et al., 2013). The red triangles denote active lowing, we describe and discuss the major subduction processes volcanoes. (For interpretation of the references to colour in this figure legend, the and consequences, with an emphasis on seismotectonics and mag- reader is referred to the web version of this article.) matism, from the toward the arc, back-arc and the East Asian continent, which form a broad deformation zone of 2000 km wide due to the strong interactions among the four seismic networks in NE Asia and 19 portable seismic stations lithospheric plates (Fig. 21). deployed in the Changbai area (Zhao et al., 2009b). Fig. 20 shows five vertical cross-sections of the local tomography passing through the Changbai volcano. A low-V anomaly of 100 km wide 8.1. Outer-rise earthquakes and slab hydration is clearly visible down to a depth of 300 km under the volcano, and the low-V zone extends toward the east in the depth range of 300– Near the oceanic trench, normal-faulting earthquakes occur fre- 480 km. In the lower portion of the MTZ, a continuous high-V quently in the outer-rise portion because of the upward bending of anomaly is clearly visible, which represents the stagnant Pacific the oceanic lithosphere before the plate subduction (Fig. 21a). slab (Fig. 20). The local tomographic images are compared with Large outer-rise earthquakes are associated with deep normal the distribution of earthquakes (M P 4.0) that occurred after faults in the oceanic lithosphere, such as the 3 March 1933 San- 1964 in NE Asia (Fig. 20). It is clear that the prominent low-V riku-oki earthquake (Mw 8.4) (Kanamori, 1971) and the 11 March anomaly under Changbai extends down to a depth of 410 km and 2011 Miyagi-oki earthquake (Mw 7.5) soon after the Mw 9.0 mega- it spreads toward the east to the upper portion of the MTZ above thrust earthquake (Obana et al., 2012). A large amount of seawater or close to swarms of deep earthquakes in the Pacific slab may permeate through the deep portion of the Pacific plate along (Fig. 20). Very deep earthquakes also occurred close to the active those normal faults at least soon after every large outer-rise earth- Ulleung volcano in the Japan Sea (Fig. 20e). However, the local quake, resulting in slab hydration (e.g., Peacock, 2001; Ranero tomography has a poor resolution under the Ulleung volcano, et al., 2003; Ranero and Sallares, 2004; Zhao et al., 2007b; because there is no seismic station in the Japan Sea (Zhao and Faccenda et al., 2009; Fujie et al., 2013b). When the Pacific plate Tian, 2013). subducts to a certain depth (e.g., 50 km under Tohoku), the stress Integrating all the geophysical, geochemical and petrological regime in the upper portion of the slab is changed from extension findings obtained so far, Zhao and Tian (2013) proposed that the to compression (e.g., Gamage et al., 2009) and so the normal faults occurrence of deep earthquakes in the Pacific slab may have con- are closed within the slab. Thus a large amount of water can be tributed to the Changbai volcanism (Fig. 21a). To date, many litho- preserved within the slab and brought down to the MTZ depth spheric normal-faulting earthquakes have been recorded that beneath the Korean Peninsula and East China. occurred frequently (perhaps repeatedly) at the outer rise right before the Pacific plate enters the trench axis (e.g. Kanamori, 8.2. Megathrust-zone heterogeneity and megathrust earthquakes 1971, 2004; Hino et al., 2009). One recent example of the outer-rise earthquakes is that (M 7.5) which occurred within one hour fol- Because of the seismic and mechanical coupling between the lowing the great Tohoku-oki earthquake on 11 March 2011 subducting oceanic plate and the overlying continental plate, large (Obana et al., 2012; Fig. 2). Seawater may permeate through the and small thrust earthquakes occur actively in the megathrust D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49 41

Fig. 18. Vertical cross-sections of P-wave tomography along the profiles shown on the inset map (Wei et al., 2012). The red and blue colors represent low and high velocities, respectively. The velocity perturbation scale is shown at the bottom. The white dots indicate the background seismicity that occurred within a 30-km width of each profile. The red triangles denote active intraplate volcanoes. The dashed lines denote the 410- and 660-km discontinuities. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) zone beneath the forearc region down to 60 km depth, such as the areas where megathrust events reoccurred are not always con- the great 2011 Tohoku-oki earthquakes, although events with sistent with their high coseismic slips. For example, the coseismic other types of mechanisms (strike-slip and normal faulting) also slip distributions of the 1944 Tonankai earthquake (M 7.9) (Kikuchi occur in and around the megathrust zone (e.g., Asano et al., et al., 2003) and the 1946 Nankai earthquake (M 8.0) (Sagiya and 2011; Huang and Zhao, 2013a; Yoshida et al., 2012). The degree Thatcher, 1999) are not limited in the high-V patches (asperities) of interplate seismic coupling is not uniform because of the exis- where the initial ruptures occurred (Fig. 15). In contrast to the tence of strong structural heterogeneities in the megathrust zone high-V patches (strong coupled areas) in the megathrust zone, as revealed by seismic tomography. Most of the large megathrust the low-V patches probably represent weakly coupled or even earthquakes after 1900 are found to occur in or around high-V decoupled areas. Active-source seismic surveys also revealed a patches in the megathrust zones under the Tohoku, Kuril and Nan- low seismic-impedance layer (i.e., a low-V layer) in the megathrust kai forearc regions (Zhao et al., 2011b; Huang and Zhao, 2013b; Liu zone southeast off the Kii Peninsula, which represents a weakly et al., 2013a,b). The high-V patches may reflect subducted sea- coupled area (Bangs et al., 2009). Due to the weak interplate cou- mounts, oceanic ridges and other topographic highs on the sea- pling in the low-V patches, ruptures of some megathrust earth- floor, and they can form asperities in the megathrust zone. In quakes could unimpededly pass through the low-V areas contrast, the low-V patches in the megathrust zone may contain adjacent to the high-V patches (Liu et al., 2013a,b). The asperities more subducted sediments and fluids released from slab dehydra- may also act as barriers hindering or even ceasing the rupture pro- tion. The fluids control the pore fluid pressure and so play an cesses of some megathrust earthquakes triggered by other asperi- important role in the nucleation of megathrust earthquakes. ties (e.g., Nakanishi et al., 2002; Park et al., 2004). This scenario The asperities resulting from the subducted oceanic ridges and may explain a feature that many interplate earthquakes (M 6–7) seamounts may increase the shear stress/strain and promote the occurred beneath the Kyushu forearc where many small high-V nucleation of megathrust earthquakes (e.g., Cloos, 1992; Scholz and low-V anomalies coexist and so limit the coseismic rupture and Small, 1997; Baba et al., 2001; Lay et al., 2012). However, extension (Nishizawa et al., 2009; Liu et al., 2013b). In contrast, 42 D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49

Fig. 19. Tectonic background of Northeast Asia (Zhao and Tian, 2013). The dots show epicenters of large earthquakes (M P 6.0) that occurred from January 1900 to September 2009 compiled by the International Seismological Center. The earthquake magnitude scale is shown at the bottom. The color of the dots denotes the focal depth; its scale is shown at the bottom. The black stars represent the Tohoku-oki main shock (Mw 9.0) and an aftershock (Mw 7.5) in the outer-rise which occurred on 11 March 2011. The contour lines show the upper boundary of the subducting Pacific slab down to 575 km depth, while toward the west the Pacific slab becomes flat in the mantle transition zone. The blue sawtooth line denotes the Japan Trench. The red triangles denote the active Changbai and Ulleung intraplate volcanoes. The black triangles denote active arc volcanoes on the Japan Islands. The black lines in China denote the major active faults. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) interconnected low-V anomalies are revealed south off Shikoku 2003; Wiens et al., 2008; Gerya, 2011). Much water is expelled and Honshu, which may lead to unimpeded rupture processes of from the subducting slab and contributes to melt generation in great megathrust events, such as the 1946 M 8.0 Nankai earth- the hot portion of the mantle wedge under the volcanic front and quake (Fig. 15). The same scenario may have happened in the back-arc. The fluids reduce the melting temperature of rocks, 1968 M 8.3 and the 2003 M 8.0 Tokachi-oki earthquakes which can produce significant amounts of melt. van Keken et al. (Fig. 13), as well as the 2011 Mw 9.0 Tohoku-oki earthquake (Figs. 7 (2011) used a global compilation of subduction-zone thermal mod- and 10). els to predict the metamorphic facies and H2O content of subduct- ing slabs. Their results show that mineralogically bound water can 8.3. Intraslab earthquakes pass efficiently through old and fast subduction zones such as in the Western Pacific, whereas nearly complete slab dehydration Ruptures within the subducting slab cause large intraslab earth- occurs in hot subduction zones such as SW Japan and Cascadia. quakes at intermediate-depths, such as the 2001 Geiyo earthquake They found that the top of the slab is sufficiently hot in all subduc- (M 6.7) which occurred in the PHS slab under SW Japan (Zhao et al., tion zones, and so the subducted upper crust including sediments 2002), the 2003 Miyagi-oki earthquake (M 7.0; Mishra and Zhao, can dehydrate significantly. The degree and depth of dehydration 2004), and the 7 April 2011 Miyagi-oki earthquake (M 7.1; in the deeper crust and uppermost mantle of the slab are highly Nakajima et al., 2011) which occurred within the Pacific slab under diverse and depend strongly on composition and local pressure the Tohoku forearc. It is found that the large intraslab earthquakes and temperature conditions. The mantle part of the slab dehy- generally occurred in lower-velocity zones within the subducting drates at intermediate depths in all but the coldest subduction slab, which are attributed to the process of dehydration embrittle- zones. On average, about 1/3 of the bound H2O subducted globally ment resulting from the dehydration of hydrous minerals within in slabs reaches a depth of 240 km, carried principally and roughly the slab, causing the intraslab events by enhancing pore pressures equally in the gabbro and peridotite sections. The predicted global along the pre-existing faults/fractures within the slab (e.g., Mishra flux of H2O to the deep mantle is small but still amounts to about and Zhao, 2004). one ocean mass over the Earth’s history (van Keken et al., 2011). On the other hand, the solid-state mantle wedge flow in sub- 8.4. Arc magmatism and mantle wedge dynamics duction zones plays an important role in controlling the thermal structure and dynamics of subduction zones (e.g., Wada et al., To date, many multidisciplinary studies have shown that slab 2011). As the cold oceanic lithosphere subducts, it cools the over- dehydration and convective circulation (corner flow) in the mantle riding mantle wedge. The corner flow, however, replenishes the wedge are two most important processes to produce arc and back- wedge with hot mantle material, providing a thermal condition arc magmatism and volcanism (e.g., Tatsumi, 1989; Peacock, 1990; necessary for melt generation and arc magmatism. The hot mantle Zhao et al., 1992; Hasegawa and Zhao, 1994; Iwamori and Zhao, also heats up the top of the subducting slab and promotes the slab 2000; Scambelluri and Philippot, 2001; Stern, 2002; van Keken, dehydration, which is important to such processes as intraslab D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49 43

Fig. 20. (a–e) Vertical cross-sections of P-wave local tomography along the profiles shown on the inset map (f). The blue and red colors denote high and low velocities, respectively. The cross symbols denote crustal earthquakes. Deep earthquakes are shown in open circles (M < 7.0) and red stars (M P 7.0). The velocity perturbation and earthquake magnitude scales are shown below (c). The two dashed lines denote the Moho and 410-km discontinuities. The other labeling is the same as thatinFig. 19. After Zhao and Tian (2013). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) seismicity and volatile recycling (e.g., Wada et al., 2011; van Keken modeling studies is generally consistent with that revealed by seis- et al., 2011). mic tomography (e.g., Zhao et al., 1992, 2015; Huang et al., 2011a). The mantle-wedge corner flow and aqueous fluids from slab The across-arc and along-arc variations of the corner flow in the dehydration are imaged as low-V zones by seismic tomography mantle wedge have been also revealed by seismic anisotropy stud- (Figs. 6 and S1). Low-V zones are also visible in the crust and ies using both shear-wave splitting measurements (e.g., Huang uppermost mantle under the forearc (Fig. 11), which reflects shal- et al., 2011b,c; Long, 2013) and P-wave anisotropic tomography low slab dehydration. Low-frequency microearthquakes occur in or (e.g., Wang and Zhao, 2008, 2010, 2013; Huang et al., 2011a; around low-V zones in the lower crust and uppermost mantle Tian and Zhao, 2012; Liu et al., 2013a). under the volcanic front and back-arc areas, which are good indica- The low-V zones in the mantle wedge under Tohoku extend tors of magmatic activity (Figs. 6 and 8 and S1). Beneath the volca- westward under the eastern margin of the Japan Sea (e.g., nic front, the low-V zones become more prominent beneath each Abdelwahed and Zhao, 2007; Zhao et al., 2011a, 2012a), but it is group of active volcanoes (Fig. S1), which indicate along-arc varia- still unclear exactly where the upwelling flow originates under tions in the fluid and magmatic activity in the mantle wedge, sup- the back-arc region, because the tomographic image has a lower porting the hot finger model (Tamura et al., 2002). Small-scale resolution under the Japan Sea due to the lack of seismic stations convection may occur in the mantle wedge, because the viscosity there (Figs. 1a and 21a). there can be reduced significantly by fluids released from the slab (Honda and Saito, 2003). Another possibility for the along-arc var- 8.5. Slab dehydration and crustal earthquakes iation in seismic property is diapiric ascent of metasomatized man- tle (Hall and Kincaid, 2001; Gerya and Yuen, 2003). Numerical Large crustal earthquakes occur actively in the overlying conti- modeling studies have dealt explicitly with the seismic signature nental plate from the forearc, arc to back-arc areas. Fig. 21b shows of diapirs in the mantle wedge (e.g., Gerya et al., 2006; Gorczyk a qualitative model to explain the seismogenesis of large crustal et al., 2006; Gerya, 2011). The seismic structure predicted by these earthquakes in Japan, based on the studies until now (e.g., 44 D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49

Fig. 21. (a) A cartoon showing main features of the upper-mantle structure and dynamics under Western Pacific and East Asia, emphasizing the possible relationship between the Changbai intraplate volcanism and deep earthquakes in the Pacific slab (Zhao and Tian, 2013). DXAL, Daxing-Anling. (b) Schematic illustration of across-arc vertical cross- section of the crust and upper mantle under the Japan subduction zone. The generation of large crustal earthquakes is controlled by the tectonic background and structural heterogeneities, in particular, arc magma and fluids from slab dehydration (Zhao et al., 2015).

Hasegawa and Zhao, 1994; Zhao et al., 1996, 2010, 2015; Hasegawa 2007 Niigata (M 6.8), 2005 west off Fukuoka (M 7.0), and the et al., 2005, 2009; Xia et al., 2008a; Tong et al., 2011, 2012; Wei and 2008 Iwate-Miyagi (M 7.2) earthquakes (e.g., Zhao et al., 2010), Zhao, 2013). as well as the large historic earthquakes (Zhao et al., 2015). The In the forearc area, the temperature is lower, hence magma can- large crustal earthquakes are not apt to occur within the very weak not be produced and no volcano exists there, thus the fluids from low-V zones but in their edge portions where the mechanical slab dehydration migrate up to the crust and form a water wall strength of rocks is stronger than that of the low-V zones but still in the mantle wedge and the lower crust (Zhao et al., 2015). If weaker than the normal sections of the seismogenic layer. Thus the the fluids enter an active fault in the crust, pore pressure will edge portions of the low-V areas become the ideal locations to gen- increase and fault-zone friction will decrease, which may trigger erate large crustal earthquakes which produce active faults reach- large crustal earthquakes in the forearc, such as the 1995 Kobe ing the Earth’s surface or blind faults within the brittle upper crust earthquake (M 7.2) in SW Japan (Zhao et al., 1996; Salah and (Fig. 21b; Zhao et al., 2002, 2010). Although the tomographic Zhao, 2003; Tong et al., 2011) and the 2011 Iwaki earthquake (M images are complex and their details vary in different earthquake 7.0) in Tohoku (Tong et al., 2012). areas, low-V and/or high-r anomalies are generally visible below Under the volcanic front and back-arc areas, arc magma is pro- or around the main shock hypocenters. This suggests one common duced because of the high temperature in the mantle wedge and feature, i.e., significant influence of fluids and magma on the rup- fluids from slab dehydration. Migration of magma up to the crust ture nucleation of large crustal earthquakes (Zhao et al., 2010, results in active arc volcanoes and causes lateral heterogeneity 2015). and weakening of the seismogenic upper crust. The ascending arc Structural heterogeneities (in particular, crustal fluids) and magmas raise the temperature and reduce the seismic velocity of their effects on seismogenesis are also revealed in source areas of crustal materials around them, causing the brittle seismogenic many large crustal earthquakes in the continental regions (e.g., layer above them to become locally thinner and weaker Kayal et al., 2002; Huang et al., 2002; Mishra and Zhao, 2003; (Hasegawa et al., 2005). Compressive stress regime due to the con- Huang and Zhao, 2004, 2009; Mukhopadhyay et al., 2006; Qi tinuous plate subduction causes stress concentration in those weak et al., 2006; Tian et al., 2007; Lei and Zhao, 2009; Wang et al., sections of the seismogenic layer, leading to large crustal earth- 2010, 2013; Wei et al., 2013a,b; Chen et al., 2014; Cheng et al., quakes, such as the 2000 Western-Tottori (M 7.3), 2004 and 2014; Xia and Zhao, 2014). Recent lunar studies show that even D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49 45 the distribution and generation of moonquakes are affected by areas which can be detected with geophysical imaging (Zhao structural heterogeneities in the lunar interior (Zhao et al., 2008, et al., 2002, 2015). 2012b), suggesting similar seismogenic processes in both the Earth and Moon. 8.8. Future perspectives

8.6. Deep earthquakes, BMW and intraplate volcanism The future advances of seismic imaging of the Western Pacific and East Asian region depend on the progress in both seismic Focal-mechanism solutions of deep earthquakes show that instrumentation and methodology. The data coverage determines compressive stress regime occurs in the Pacific slab at depths of the first-order features of a tomographic result. At present, most 200–600 km under the Japan Sea (Zhao et al., 2009b; Li et al., of the seismic stations are installed in the land areas, whereas 2013), because the slab meets strong resistance at the 670-km dis- there are few stations in the Pacific Ocean and the marginal seas. continuity due to the sudden increase in viscosity from the MTZ to Gradual deployment of seismometers in seafloor and in those less the lower mantle (e.g., Fukao et al., 2001; Zhao, 2001, 2004). Hence instrumented land areas will be the most important task for seis- the Pacific slab becomes stagnant in the lower part of the MTZ mologists from now (Zhao, 2012). Now a permanent OBS network beneath the Korean Peninsula and East China, and a big mantle is in construction in the Tohoku and Hokkaido forearc region wedge (BMW) has formed in the upper mantle and the upper part (Uehira et al., 2013). Once it is available, it will provide crucial data of the MTZ above the subducting Pacific slab and the stagnant slab for clarifying the mechanism of megathrust earthquakes and (Zhao et al., 2004, 2007b). Mineral physics studies show that dehy- dynamic processes in the forearc region of eastern Japan. Installing dration reactions still occur in the Pacific slab in the MTZ, because a dense permanent or portable OBS network in the marginal seas the Pacific slab is very old and cold and so some hydrous phases in (e.g., the East China Sea, the Japan Sea, and the Sea of Okhotsk) will the slab keep stable down to the MTZ depths (e.g., Ohtani and be necessary to resolve important geodynamic issues in subduc- Zhao, 2009; Kuritani et al., 2011, 2013). Fluids from the slab dehy- tion zones, such as the fate of subducting slabs, the structure and dration and corner flow in the BMW may cause upwelling of hot dynamics in the BMW, the mechanism of deep earthquakes, and and wet asthenospheric materials, resulting in intraplate volca- the back-arc and intraplate magmatism. A joint use of the OBS data nism in NE Asia. More magmas may be supplied to the Changbai and the land-based network data will result in much better tomo- volcano, because the frequent occurrence of large deep earth- graphic images, which will greatly improve our understanding of quakes may release additional fluids to the BMW from the pre- the structure and dynamics in Western Pacific and East Asia. served faults in the Pacific slab (Fig. 21a). As a result, Changbai The tomographic images shown above were mainly obtained has become much more active and prominent than the other intra- using body-wave travel-time tomography which is the most plate volcanoes in NE Asia (Zhao and Tian, 2013). straightforward, robust and mature tool that has produced many more credible and geologically reasonable results than any other 8.7. Fluids and earthquakes tomographic methods. So far, however, most travel-time tomogra- phy studies have used only first P- and S-wave arrivals. From now, A great number of studies using various approaches have shown a large number of later-phase data of reflected and converted that fluids exist widely in subduction zones (see Hasegawa et al., waves should be collected from three-component seismograms 2009; Zhao, 2012 for recent reviews). Fluids are found to play an by taking advantage of progresses in modern seismology, such as important role in the generation of outer-rise earthquakes, mega- waveform modeling (e.g., Helmberger et al., 2001; Abdelwahed thrust earthquakes, intraslab earthquakes, crustal earthquakes, and Zhao, 2005, 2014). The later-phase data are very important and intermediate-depth and deep earthquakes (e.g., Zhao et al., for improving the resolution of tomographic images (e.g., Zhao 1996, 2002; Mishra and Zhao, 2003, 2004; Jiang et al., 2008; et al., 2005, 2013; Salah et al., 2005; Lei and Zhao, 2006; Xia Hasegawa et al., 2009; Jiang and Zhao, 2011; Faccenda, 2014). It et al., 2007; Sun et al., 2008; Gupta et al., 2009). seems reasonable to believe that fluids are involved in the genera- Most of the tomographic images obtained so far are P-wave tion of all types of earthquakes and volcanism. The existence of flu- velocity tomography. In future studies, high-quality S-wave arriv- ids in the crust and upper mantle affects the long-term structural als, surface-wave and waveform data from local and teleseismic and compositional evolution of the fault zones, change the strength events recorded by the dense seismic networks in Japan, Korea, of the fault zone, and alter the local stress regime (e.g., Sibson, China and other countries should be collected and used for seismic 1992; Hickman et al., 1995; Zhao et al., 1997, 2010; Huang et al., imaging, thus high-resolution tomographic images of P- and S- 2011d; Yoshida et al., 2012). These factors can enhance stress con- wave velocity (Vp, Vs), Poisson’s ratio, seismic attenuation (Qp, centration in the seismogenic layer leading to mechanical failure. Qs), and seismic anisotropy can be determined (e.g., Liu et al., The above-mentioned many pieces of evidence indicate that the 2013a,b, 2014; Tian and Zhao, 2013; Wang et al., 2014; Zhao generation of a large earthquake is not entirely a mechanical pro- et al., 2015), which will better characterize the 3-D structure and cess, but is closely related to the physical and chemical properties subduction dynamics, in particular, the fluid and magmatic pro- of materials in the crust and upper mantle, such as magma and cesses. In addition, findings of other branches of Earth sciences, fluids (Zhao et al., 2002, 2010). The rupture nucleation zone is such as mineral physics and computer simulations, should be also not limited in a 2-D fault plane but has a 3-D spatial extent, as sug- exploited for better interpreting the seismological results. gested earlier by Tsuboi (1956) with the concept of earthquake vol- ume. Complex physical and chemical reactions take place in the Acknowledgements earthquake volume, causing heterogeneities in the material prop- erty and stress field, which can be detected with seismic tomogra- This work was partially supported by Grant-in-aid for Scientific phy and other geophysical methods. The source zones of M 6–9 Research (Kiban-A 17204037, Kiban-S 11050123) from Japan Soci- earthquakes extend from 10 to 500 km. The spatial resolution ety for the Promotion of Science as well as the Global-COE program of current tomography is close to or even shorter than the scale of Earth and Planetary Sciences, Tohoku University. The author is of the large-earthquake source zones, enabling us to image the grateful to the effective collaborations and helpful discussions with earthquake-related heterogeneities (i.e., earthquake volume) in many colleagues and coworkers, including Zhouchuan Huang, Xin the crust and upper mantle. These results indicate that large earth- Liu, Jian Wang, You Tian, Wei Wei, Ping Tong, Dayong Yu, A. quakes do not strike anywhere randomly, but only in anomalous Hasegawa, H. Kanamori, N. Umino, Y. Yamamoto, T. Yanada, G. 46 D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49

Toyokuni, and H. Kitagawa. Prof. Bor-ming Jahn (the Editor) and Fukao, Y., Widiyantoro, S., Obayashi, M., 2001. Stagnant slabs in the upper and lower two anonymous referees provided thoughtful review comments mantle transition region. Rev. Geophys. 39, 291–323. Gamage, S., Umino, N., Hasegawa, A., Kirby, S., 2009. Offshore double-planed which have improved this paper. shallow seismic zone in the NE Japan forearc region revealed by sP depth phases recorded by regional networks. Geophys. J. Int. 178, 195–214. Gerya, T., 2011. Future directions in subduction modeling. J. Geodyn. 52, 344–378. Appendix A. Supplementary material Gerya, T., Yuen, D., 2003. Rayleigh–Taylor instabilities from hydration and melting propel ‘‘cold plumes’’ at subduction zones. Earth Planet. Sci. Lett. 212, 47–62. Supplementary data associated with this article can be found, in Gerya, T., Connolly, J., Yuen, D., Gorczyk, W., Capel, A., 2006. Seismic implications of the online version, at http://dx.doi.org/10.1016/j.jseaes.2014.10. mantle wedge plumes. Phys. Earth Planet. Inter. 156, 59–74. Gorczyk, W., Gerya, T., Connolly, J., Yuen, D., Rudolph, M., 2006. Large-scale rigid- 022. body rotation in the mantle wedge and its implications for seismic tomography. Geochem. Geophys. Geosyst. 7, Q05018. Gupta, S., Zhao, D., Ikeda, M., Ueki, S., Rai, S., 2009. Crustal tomography under the Median Tectonic Line in Southwest Japan using P and PmP data. J. Asian Earth Sci. 35, 377–390. References Hall, P., Kincaid, C., 2001. Diapiric flow at subduction zones: a recipe for rapid transport. Science 292, 2472–2475. Hall, R., 2002. Cenozoic geological and plate tectonic evolution of SE Asia and the Abdelwahed, M.F., Zhao, D., 2005. Waveform modelling of local earthquakes in SW Pacific: computer-based reconstructions, model and animations. J. Asian southwest Japan. Earth Planets Space 57, 1039–1054. Earth Sci. 20, 353–431. Abdelwahed, M., Zhao, D., 2007. Deep structure of the Japan subduction zone. Phys. Hall, R., Ali, J., Anderson, C., Baker, S., 1995. Origin and motion history of the Earth Planet. Inter. 162, 32–52. Philippine Sea plate. Tectonophysics 251, 229–250. Abdelwahed, M.F., Zhao, D., 2014. Genetic waveform modeling for the crustal Hasegawa, A., Zhao, D., 1994. Deep structure of magmatic regions as structure in Northeast Japan. J. Asian Earth Sci. 89, 66–75. inferred from seismic observations. In: Ryan, M.P. (Ed.), Magmatic Systems. Arita, K., Ikawa, T., Ito, T., Yamamoto, A., et al., 1998. Crustal structure and tectonics Academic Press, pp. 179–195. of the Hidaka Collision Zone, Hokkaido (Japan), revealed by vibroseis seismic Hasegawa, A., Nakajima, J., Umino, N., Miura, S., 2005. Deep structure of the reflection and gravity surveys. Tectonophysics 290, 197–210. and its implications for crustal deformation and shallow Asano, Y., Saito, T., Ito, Y., Shiomi, K., Hirose, H., Matsumoto, T., Aoi, S., Hori, S., seismicity. Tectonophysics 403, 59–75. Sekiguchi, S., 2011. Spatial distribution and focal mechanisms of aftershocks of Hasegawa, A., Nakajima, J., Uchida, N., Okada, T., Zhao, D., Matsuzawa, T., Umino, N., the 2011 off the Pacific coast Tohoku Earthquake. Earth Planets Space 63, 669– 2009. Plate subduction, and generation of earthquakes and magmas in Japan as 673. inferred from seismic observation: an overview. Gondwana Res. 16, 370–400. Azuma, R., Murai, Y., Katsumata, K., et al., 2012. Was the 1952 Tokachi-oki He, L., 2014. Numerical modeling of convective erosion and peridotite–melt earthquake (Mw = 8.1) a typical underthrust earthquake?: Plate interface interaction in big mantle wedge: implications for the destruction of the North reflectivity measurement by an air gun-ocean bottom seismometer China Craton. J. Geophys. Res. 119, 3662–3677. experiment in the Kuril Trench. Geochem. Geophys. Geosyst. 13, Q08015. Heki, K., Miyazaki, S., Tsuji, H., 1997. Silent fault slip following an interplate thrust Baba, T., Hori, T., Hirano, S., Cummins, P., Park, J., Kameyama, M., Kaneda, Y., 2001. earthquake at the Japan trench. Nature 386, 592–597. Deformation of a seamount subducting beneath an accretionary prism: Helmberger, D., Song, X.J., Zhu, L., 2001. Crustal complexity from regional waveform constraints from numerical simulation. Geophys. Res. Lett. 28, 1827–1830. tomography: aftershocks of the 1992 Landers earthquake. J. Geophys. Res. 106, Bangs, N., Moore, G., Gulick, S., Pangborn, E., Tobin, H., Kuramoto, S., Taira, A., 2009. 609–620. Broad, weak regions of the Nankai megathrust and implications for shallow Herrmann, R., 1976. Focal depth determination from the signal character of long- coseismic slip. Earth Planet. Sci. Lett. 284, 44–49. period P waves. Bull. Seismol. Soc. Am. 66, 1221–1232. Bird, P., 2003. An updated digital model of plate boundaries. Geochem. Geophys. Hickman, S., Sibson, R., Bruhn, R., 1995. Introduction to special section: mechanical Geosyst. 4, 1027. involvement of fluids in faulting. J. Geophys. Res. 100, 12831–12840. Bostock, M., Hyndman, R., Rondenay, S., Peacock, S., 2002. An inverted continental Hilde, T., Lee, C., 1984. Origin and evolution of the west Philippine basin: a new Moho and serpentinization of the forearc mantle. Nature 417, 536–538. interpretation. Tectonophysics 102, 85–104. Bowman, D., King, G., 2001. Accelerating seismicity and stress accumulation before Hino, R., Azuma, R., Ito, Y., Yamamoto, Y., Suzuki, K., Tsushima, H., 2009. Insight into large earthquakes. Geophys. Res. Lett. 28, 4039–4042. complex rupturing of the immature bending normal fault in the outer slope of Chen, C., Zhao, D., Wu, S., 2014. Crust and upper mantle structure of New Madrid the Japan Trench from aftershocks of the 2005 Sanriku earthquake (Mw = 7.0) Seismic Zone: insight into intraplate earthquakes. Phys. Earth Planet. Inter. 230, located by ocean bottom seismometry. Geochem. Geophys. Geosyst. 10, 1–14. Q07O18. Cheng, B., Zhao, D., Zhang, G., 2011. Seismic tomography and anisotropy in the Hirahara, K., 1981. Three-dimensional seismic structure beneath southwest Japan: source area of the 2008 Iwate-Miyagi earthquake (M 7.2). Phys. Earth Planet. the subducting Philippine Sea plate. Tectonophysics 79, 1–44. Inter. 184, 172–185. Honda, S., Saito, M., 2003. Small-scale convection under the back-arc occurring in Cheng, B., Cheng, S., Zhang, G., Zhao, D., 2014. Seismic structure of the Helan– the low viscosity wedge. Earth Planet. Sci. Lett. 216, 703–715. Liupan–Ordos western margin tectonic belt in North-Central China and its Huang, J., Zhao, D., 2004. Crustal heterogeneity and seismotectonics of the region geodynamic implications. J. Asian Earth Sci. 87, 141–156. around Beijing, China. Tectonophysics 385, 159–180. Cloos, M., 1992. Thrust-type subduction-zone earthquakes and seamount Huang, J., Zhao, D., 2006. High-resolution mantle tomography of China and asperities: a physical model for seismic rupture. Geology 20, 601–604. surrounding regions. J. Geophys. Res. 111, B09305. Deschamps, A., Lallemand, S., 2002. The West Philippine Basin: an Eocene to early Huang, J., Zhao, D., 2009. Seismic imaging of the crust and upper mantle under Oligocene back arc basin opened between two opposed subduction zones. J. Beijing and surrounding regions. Phys. Earth Planet. Inter. 173, 330–348. Geophys. Res. 107, 2322. Huang, J., Zhao, D., Zheng, S., 2002. Lithospheric structure and its relationship to Duan, Y., Zhao, D., Zhang, X., et al., 2009. Seismic structure and origin of active seismic and volcanic activity in southwest China. J. Geophys. Res. 107, intraplate volcanoes in Northeast Asia. Tectonophysics 470, 257–266. JB000137. Engdahl, E., Billington, S., 1986. Focal depth determination of central Aleutian Huang, Z., Zhao, D., 2013a. Relocating the 2011 Tohoku-oki earthquakes (M 6.0– earthquakes. Bull. Seismol. Soc. Am. 76, 77–93. 9.0). Tectonophysics 586, 35–45. Eyuboglu, Y., 2013. Slab window magmatism and convergent margin tectonics. Huang, Z., Zhao, D., 2013b. Mechanism of the 2011 Tohoku-oki earthquake (Mw 9.0) Geosci. Front. 4, 349–351. and tsunami: insight from seismic tomography. J. Asian Earth Sci. 70, 160–168. Faccenda, M., 2014. Water in the slab: a trilogy. Tectonophysics 614, 1–30. Huang, Z., Zhao, D., Wang, L., 2011a. Seismic heterogeneity and anisotropy of the Faccenda, M., Burlini, L., Gerya, T., Mainprice, D., 2008. Fault-induced seismic Honshu arc from the Japan Trench to the Japan Sea. Geophys. J. Int. 184, 1428– anisotropy by hydration in subducting oceanic plates. Nature 455, 1097–1100. 1444. Faccenda, M., Gerya, T., Burlini, L., 2009. Deep slab hydration induced by bending- Huang, Z., Zhao, D., Wang, L., 2011b. Shear-wave anisotropy in the crust, mantle related variations in tectonic pressure. Nat. Geosci. 2, 790–793. wedge and the subducting Pacific slab under Northeast Japan. Geochem. Forsyth, D., 1982. Determination of focal depths of earthquakes associated with the Geophys. Geosyst. 12, Q01002. bending of oceanic plates at trenches. Phys. Earth Planet. Inter. 28, 141–160. Huang, Z., Zhao, D., Wang, L., 2011c. Frequency-dependent shear-wave splitting and Frohlich, C., 1994. Earthquakes with non-double-couple mechanisms. Science 264, multilayer anisotropy in Northeast Japan. Geophys. Res. Lett. 38, L08302. 804–809. Huang, Z., Zhao, D., Wang, L., 2011d. Stress field in the 2008 Iwate-Miyagi Frohlich, C., 2006. Deep Earthquakes. Cambridge University Press, p. 573. earthquake (M 7.2) area. Geochem. Geophys. Geosyst. 12, Q06006. Fujie, G., Kasahara, J., Hino, R., Sato, T., Shinohara, M., Suyehiro, K., 2002. A Huang, Z., Zhao, D., Hasegawa, A., Umino, N., Park, J., Kang, I., 2013. Aseismic deep significant relation between seismic activities and reflection intensities in the subduction of the Philippine Sea plate and slab window. J. Asian Earth Sci. 75, Japan Trench region. Geophys. Res. Lett. 29, GL013764. 82–94. Fujie, G., Miura, S., Kodaira, S., Kaneda, Y., et al., 2013a. Along-trench structural Hyndman, R., Peacock, S., 2003. Serpentinization of the forearc mantle. Earth Planet. variation and seismic coupling in the northern Japan subduction zone. Earth Sci. Lett. 876, 417–432. Planets Space 65, 75–83. Iinuma, T. et al., 2012. Coseismic slip distribution of the 2011 off the Pacific Coast of Fujie, G., Kodaira, S., Yamashita, M., Sato, T., Takahashi, T., Takahashi, N., 2013b. Tohoku earthquake (M 9.0) refined by means of seafloor geodetic data. J. Systematic changes in the incoming plate structure at the Kuril trench. Geophys. Res. 117, B07409. Geophys. Res. Lett. 40, 88–93. D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49 47

Ishise, M., Oda, H., 2008. Subduction of the Philippine Sea slab in view of P-wave Lei, J., Zhao, D., 2005. P-wave tomography and origin of the Changbai intraplate anisotropy. Phys. Earth Planet. Inter. 166, 83–96. volcano in Northeast Asia. Tectonophysics 397, 281–295. Itoh, Y., Ishiyama, T., Nagasaki, Y., 2005. Deformation mode in the frontal edge of an Lei, J., Zhao, D., 2006. A new insight into the Hawaiian plume. Earth Planet. Sci. Lett. arc–arc collision zone: subsurface geology, active faults and paleomagnetism in 241, 438–453. southern central Hokkaido, Japan. Tectonophysics 395, 81–97. Lei, J., Zhao, D., 2009. Structural heterogeneity of the zone and Iwamori, H., Zhao, D., 2000. Melting and seismic structure beneath the northeast the mechanism of the 2008 Wenchuan earthquake (Ms 8.0). Geochem. Japan arc. Geophys. Res. Lett. 27, 425–428. Geophys. Geosyst. 10, Q10010. Iwasaki, T., Adachi, K., Moriya, T., et al., 2004. Upper and middle crustal deformation Li, J., Wang, X., Wang, X., Yuen, D., 2013. P and SH velocity structure in the upper of an arc–arc collision across Hokkaido, Japan, inferred from seismic refraction/ mantle beneath Northeast China: evidence for a stagnant slab in hydrous wide-angle reflection experiments. Tectonophysics 388, 59–73. mantle transition zone. Earth Planet. Sci. Lett. 367, 71–81. Jiang, G., Zhao, D., 2011. Metastable olivine wedge in the subducting Pacific slab and Liu, J., 1999. Volcanoes in China. Science Press, Beijing, 219 pp. its relation to deep earthquakes. J. Asian Earth Sci. 42, 1411–1423. Liu, X., Zhao, D., Li, S., 2013a. Seismic heterogeneity and anisotropy of the southern Jiang, G., Zhao, D., Zhang, G., 2008. Seismic evidence for a metastable olivine wedge Kuril arc: insight into megathrust earthquakes. Geophys. J. Int. 194, 1069–1090. in the subducting Pacific slab under Japan Sea. Earth Planet. Sci. Lett. 270, 300– Liu, X., Zhao, D., Li, S., 2013b. Seismic imaging of the Southwest Japan arc from the 307. Nankai trough to the Japan Sea. Phys. Earth Planet. Inter. 216, 59–73. Julian, B., Miller, A., Foulger, G., 1998. Non-double-couple earthquakes 1. Theory. Liu, X., Zhao, D., Li, S., 2014. Seismic attenuation tomography of the Northeast Japan Rev. Geophys. 36, 525–549. arc: insight into the 2011 Tohoku earthquake (Mw 9.0) and subduction Kameyama, M., Nishioka, R., 2012. Generation of ascending flows in the Big Mantle dynamics. J. Geophys. Res. 119, 1094–1118. Wedge (BMW) beneath northeast Asia induced by retreat and stagnation of Long, M., 2013. Constraints on subduction geodynamics from seismic anisotropy. subducted slab. Geophys. Res. Lett. 39, L10309. Rev. Geophys. 51, 76–112. Kanamori, H., 1971. Great earthquakes at island arcs and the lithosphere. Machida, Y., Shinohara, M., Takanami, T., Murai, et al., 2009. Heterogeneous Tectonophysics 12, 187–198. structure around the rupture area of the 2003 Tokachi-oki earthquake Kanamori, H., 2004. The diversity of the physics of earthquakes. Proc. Jpn. Acad. Ser. (Mw = 8.0), Japan, as revealed by aftershock observations using Ocean Bottom B 80, 297–316. Seismometers. Tectonophysics 465, 164–176. Kato, A. et al., 2013. Imaging the source regions of normal faulting sequences Miller, A., Foulger, G., Julian, B., 1998. Non-double-couple earthquakes 2. induced by the 2011 M 9.0 Tohoku-Oki earthquake. Geophys. Res. Lett. 40, 273– Observations. Rev. Geophys. 36, 551–568. 278. Miller, M., Kennett, B., Gorbatov, A., 2006. Morphology of the distorted subducted Kato, N., Hirasawa, T., 1997. A numerical study on seismic coupling along Pacific slab beneath the Hokkaido corner, Japan. Phys. Earth Planet. Inter. 156, subduction zones using a laboratory-derived friction law. Phys. Earth Planet. 1–11. Inter. 102, 51–68. Mishra, O.P., Zhao, D., 2003. Crack density, saturation rate and porosity at the 2001 Katsumata, K., Yamanaka, Y., 2006. The 29 November 2004 M7.1 Kushiro-oki Bhuj, India, earthquake hypocenter: a fluid driven earthquake? Earth Planet. Sci. earthquake: an event between the on-going seismic quiescence area and the Lett. 212, 393–405. asperity ruptured by the 1973 Nemuro-oki earthquake. Geophys. Bull. Mishra, O.P., Zhao, D., 2004. Seismic evidence for dehydration embrittlement of the Hokkaido Univ. 69, 23–39. subducting Pacific slab. Geophys. Res. Lett. 31, L09610. Katsumata, K., Wada, N., Kasahara, M., 2006. Three-dimensional P and S wave Mishra, O.P., Zhao, D., Umino, N., Hasegawa, A., 2003. Tomography of Northeast velocity structures beneath the Hokkaido corner, Japan–Kurile arc–arc junction. Japan forearc and its implications for interplate seismic coupling. Geophys. Res. Earth Planets Space 58, e37–e40. Lett. 30, GL017736. Kawasaki, I., Asai, Y., Tamura, Y., 2001. Space–time distribution of interplate Miura, S., Kodaira, S., Nakanishi, A., Tsuru, T., 2003. Structural characteristics moment release including slow earthquakes and the seismo-geodetic coupling controlling the seismicity of southern Japan Trench fore-arc region, revealed by in the Sanriku-oki region along the Japan trench. Tectonophysics 330, 267–283. ocean bottom seismographic data. Tectonophysics 363, 79–102. Kayal, J.R., Zhao, D., Mishra, O.P., Reena, D., Singh, O., 2002. The 2001 Bhuj Mukhopadhyay, S., Mishra, O.P., Zhao, D., Kayal, J.R., 2006. 3-D seismic structure of earthquake: tomographic evidence for fluids at the hypocenter and its the source area of the 1993 Latur, India, earthquake and its implications for implications for rupture nucleation. Geophys. Res. Lett. 29, GL015177. rupture nucleations. Tectonophysics 415, 1–16. Kikuchi, M., Nakamura, M., Yoshikawa, K., 2003. Source rupture processes of the Munekane, H., 2012. Coseismic and early postseismic slips associated with the 2011 1944 Tonankai earthquake and the 1945 Mikawa earthquake derived from low- off the Pacific coast of Tohoku earthquake sequence. EOF analysis of GPS gain seismograms. Earth Planets Space 55, 159–172. kinematic time series. Earth Planets Space 64, 1077–1091. Kimura, G., 1986. Oblique subduction and collision: Forearc tectonics of the Kuril Murai, Y., Akiyama, S., Katsumata, K., et al., 2003. Delamination structure imaged in arc. Geology 14, 404–407. the source area of the 1982 Urakawa-oki earthquake. Geophys. Res. Lett. 30, 1490. Kimura, G., 1996. Collision orogeny at arc–arc junctions in the Japanese Islands. Nagai, R., Kikuchi, M., Yamanaka, Y., 2001. Comparative study on the source process Island Arc 5, 262–275. of recurrent large earthquakes in Sanriku-Oki region: the 1968 Tokachi-Oki Kita, S. et al., 2012. High-resolution seismic velocity structure beneath the Hokkaido earthquake and the 1994 Sanriku-oki earthquake. J. Seismol. Soc. Jpn. 52, 267– corner, northern Japan: arc–arc collision and origins of the 1970 M 6.7 Hidaka 289. and 1982 M 7.1 Urakawa-oki earthquakes. J. Geophys. Res. 117, B12301. Nakajima, J., Hirose, F., Hasegawa, A., 2009. Seismotectonics beneath the Tokyo Kobayashi, K., Kasuga, S., Okino, K., 1995. Shikoku Basin and its margins. In: Taylor, metropolitan area, Japan: effect of slab–slab contact and overlap on seismicity. B. (Ed.), Back-arc Basins: Tectonics and Magmatism. Plenum Press, New York, J. Geophys. Res. 114, B08309. pp. 381–405. Nakajima, J., Hasegawa, A., Kita, S., 2011. Seismic evidence for reactivation of a Kodaira, S., Takahashi, N., Park, J., et al., 2000. Western Nankai Trough seismogenic buried hydrated fault in the Pacific slab by the 2011 M9.0 Tohoku earthquake. zone: result from a wide-angle ocean bottom seismic survey. J. Geophys. Res. Geophys. Res. Lett. 38, L00G06. 105, 5887–5905. Nakamichi, H., Hamaguchi, H., Tanaka, S., Ueki, S., Nishimura, T., Hasegawa, A., 2003. Kodaira, S., Kurashimo, E., Park, J., Takahashi, N., et al., 2002. Structural factors Source mechanisms of deep and intermediate-depth low-frequency controlling the rupture process of a megathrust earthquake at the Nankai earthquakes beneath Iwate volcano, northeastern Japan. Geophys. J. Int. 154, trough seismogenic zone. Geophys. J. Int. 149, 815–835. 811–828. Koketsu, K. et al., 2011. A unified source model for the 2011 Tohoku earthquake. Nakanishi, A., Kodaira, S., Park, J., Kaneda, Y., 2002. Deformable backstop as seaward Earth Planet. Sci. Lett. 310, 480–487. end of coseismic slip in the Nankai Trough seismogenic zone. Earth Planet. Sci. Koper, K., Hutko, A., Lay, T., Ammon, C., Kanamori, H., 2011. Frequency-dependent Lett. 203, 255–263. rupture process of the 11 March 2011 Mw 9.0 Tohoku-oki earthquake: Nakanishi, A. et al., 2004. Structural factors controlling the coseismic rupture zone Comparison of short-period P wave back-projection images and broadband of the 1973 Nemuro-Oki earthquake, the southern Kuril Trench seismogenic seismic rupture models. Earth Planets Space 63, 599–602. zone. J. Geophys. Res. 109, B05305. Kubo, H., Asano, K., Iwata, T., 2013. Source-rupture process of the 2011 Ibaraki-oki, Nanayama, F., Satake, K., Furukawa, R., Shimokawa, K., Atwater, B., Shigeno, K., Japan, earthquake (Mw 7.9) estimated from the joint inversion of strong-motion Yamaki, S., 2003. Unusually large earthquakes inferred from tsunami deposits and GPS data: relationship with seamount and Philippine Sea plate. Geophys. along the Kuril trench. Nature 424, 660–663. Res. Lett. 40, 3003–3007. Nishizawa, A., Kaneda, K., Oikawa, M., 2009. Seismic structure of the northern end of Kuritani, T., Kimura, J., Miyamoto, T., Wei, H., et al., 2009. Intraplate magmatism the Ryukyu Trench subduction zone, southeast of Kyushu, Japan. Earth Planets related to deceleration of upwelling asthenospheric mantle: implications from Space 61, e37–e40. the Changbaishan shield basalts, northeast China. Lithos 112, 247–258. Obana, K., Fujie, G., Takahashi, T., et al., 2012. Normal-faulting earthquakes beneath Kuritani, T., Ohtani, E., Kimura, J., 2011. Intensive hydration of the mantle transition the outer slope of the Japan Trench after the 2011 Tohoku earthquake: zone beneath China caused by ancient slab stagnation. Nat. Geosci. 4, 713–716. implications for the stress regime in the incoming Pacific plate. Geophys. Res. Kuritani, T., Kimura, J., Ohtani, E., Miyamoto, H., Furuyama, K., 2013. Transition zone Lett. 39, L00G24. origin of potassic basalts from Wudalianchi volcano, northeast China. Lithos Ohta, Y. et al., 2012. Geodetic constraints on afterslip characteristics following the 156–159, 1–12. March 9, 2011, Sanriku-oki earthquake, Japan. Geophys. Res. Lett. 39, L16304. Lay, T., Ammon, C., Kanamori, H., Kim, M., Xue, L., 2011. Outer trench-slope faulting Ohtani, E., Zhao, D., 2009. The role of water in the deep upper mantle and transition and the 2011 Mw 9.0 off the Pacific coast of Tohoku Earthquake. Earth Planets zone: dehydration of stagnant slabs and its effects on the big mantle wedge. Space 63, 713–718. Russ. Geol. Geophys. 50, 1073–1078. Lay, T., Kanamori, H., Ammon, C., Koper, K., et al., 2012. Depth-varying rupture Okada, T., Yoshida, K., Ueki, S., Nakajima, J., Uchida, N., et al., 2011. Shallow inland properties of subduction zone megathrust faults. J. Geophys. Res. 117. http:// earthquakes in NE Japan possibly triggered by the 2011 off the Pacific coast of dx.doi.org/10.1029/2011JB009133. Tohoku earthquake. Earth Planets Space 63, 749–754. 48 D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49

Okada, Y., Kasahara, K., Hori, S., Obara, K., 2004. Recent progress of seismic Tian, Y., Zhao, D., 2012. Seismic anisotropy and heterogeneity in the Alaska observation networks in Japan – Hi-net, F-net, K-NET and KiK-net. Earth Planets subduction zone. Geophys. J. Int. 190, 629–649. Space 56, xv–xxviii. Tian, Y., Zhao, D., 2013. Reactivation and mantle dynamics of North China Craton: Okino, K., Ohara, Y., Kasuga, S., Kato, Y., 1999. The Philippine Sea: new survey results insight from P-wave anisotropy tomography. Geophys. J. Int. 195, 1796–1810. reveal the structure and the history of the marginal basins. Geophys. Res. Lett. Tian, Y., Zhao, D., Sun, R., Teng, J., 2007. The 1992 Landers earthquake: effect of crustal 26, 2287–2290. heterogeneity on earthquake generation. Chin. J. Geophys. 50, 1488–1496. Ozawa, S., Nishimura, T., Munekane, H., Suito, H., et al., 2012. Preceding, coseismic, Tong, P., Zhao, D., Yang, D., 2011. Tomography of the 1995 Kobe earthquake area: and postseismic slips of the 2011 Tohoku earthquake, Japan. J. Geophys. Res. comparison of finite-frequency and ray approaches. Geophys. J. Int. 187, 278–302. 117, B07404. Tong, P., Zhao, D., Yang, D., 2012. Tomography of the 2011 Iwaki earthquake (M 7.0) Pacheco, J., Sykes, L., Scholz, C., 1993. Nature of seismic coupling along simple plate and Fukushima nuclear power plant area. Solid Earth 3, 43–51. boundaries of the subduction type. J. Geophys. Res. 98, 14133–14159. Tsuboi, C., 1956. Earthquake energy, earthquake volume, aftershock area, and Padhy, S., Mishra, O.P., Zhao, D., Wei, W., 2011. Crustal heterogeneity in the 2007 strength of the Earth’s crust. J. Phys. Earth 4, 63–66. Noto-Hanto earthquake area and its geodynamical implications. Tsuru, T., Park, J., Miura, S., Kodaira, S., Kido, Y., Hayashi, T., 2002. Along-arc Tectonophysics 509, 55–68. structural variation of the plate boundary at the Japan Trench margin: Park, J., Moore, G., Tsuru, T., Kodaira, S., Kaneda, Y., 2004. A subducted oceanic ridge implications of interplate coupling. J. Geophys. Res. 107, 2357. influencing the Nankai megathrust earthquake rupture. Earth Planet. Sci. Lett. Uehira, K., Kanazawa, T., Noguchi, S., et al., 2013. Installation of ocean bottom 217, 77–84. seismic and tsunami network along the Japan Trench – Off Boso system. Peacock, S., 1990. Fluid processes in subduction zones. Science 248, 329–345. Programme and Abstracts of Seismological Society of Japan 2013 Fall Meeting, Peacock, S., 2001. Are the lower planes of double seismic zones caused by D11-12. serpentine dehydration in subducting oceanic mantle? Geology 29, 299–302. Umino, N., Hasegawa, A., 1994. Aftershock focal depths of the 1993 Hokkaido- Qi, C., Zhao, D., Chen, Y., 2006. 3-D P and S wave velocity structures and their Nansei-Oki earthquake estimated from sP depth phase at small epicentral relationship to strong earthquakes in the Chinese capital region. Chin. J. distances. J. Phys. Earth 42, 321–329. Geophys. 49, 805–815. Umino, N., Hasegawa, A., Takagi, A., 1990. The relationship between seismicity Ranero, C., Sallares, V., 2004. Geophysical evidence for hydration of the crust and patterns and fracture zones beneath northeastern Japan. Tohoku Geophys. J. 33, mantle of the Nazca plate during bending at the north Chile trench. Geology 32, 149–162. 549–552. Umino, N., Hasegawa, A., Matsuzawa, T., 1995. SP depth phase at small epicentral Ranero, C., Phipps Morgan, J., McIntosh, K., Reichert, C., 2003. Bending-related distances and estimated subducting plate boundary. Geophys. J. Int. 120, 356– faulting and mantle serpentinization at the Middle America trench. Nature 425, 366. 367–373. Usami, T., 2003. Materials for Comprehensive List of Destructive Earthquakes in Romano, F., Trasatti, E., Lorito, S., Piromallo, C., Piatanesi, A., Ito, Y., Zhao, D., Hirata, Japan. University of Tokyo Press, Tokyo. K., Lanucara, P., Cocco, M., 2014. Structural control on the Tohoku earthquake Uyeda, S., Miyashiro, A., 1974. and the Japanese Islands: a synthesis. rupture process investigated by 3D FEM, tsunami and geodetic data. Sci. Rep. 4, Geol. Soc. Am. Bull. 85, 1159–1170. 5631. http://dx.doi.org/10.1038/srep05631. van Keken, P., 2003. The structure and dynamics of the mantle wedge. Earth Planet. Sagiya, T., Thatcher, W., 1999. Coseismic slip resolution along a plate boundary Sci. Lett. 215, 323–338. megathrust: the Nankai Trough, southwest Japan. J. Geophys. Res. 104, 1111– van Keken, P., Hacker, B., Syracuse, E., Abers, G., 2011. Subduction factory: 4. Depth-

1129. dependent flux H2O from subducting slabs worldwide. J. Geophys. Res. 116, Sakuyama, T., Nagaoka, S., Miyazaki, T., et al., 2014. Melting of the uppermost B01401. metasomatized asthenosphere triggered by fluid fluxing from ancient Wada, I., Rychert, C., Wang, K., 2011. Sharp thermal transition in the forearc mantle subducted sediment: constraints from the Quaternary basalt lavas at wedge as a consequence of nonlinear mantle wedge flow. Geophys. Res. Lett. 38, Chugaryeong volcano, Korea. J. Petrol. 55, 499–528. L13308. Salah, M., Zhao, D., 2003. 3-D seismic structure of Kii Peninsula in Southwest Japan: Wang, J., Zhao, D., 2008. P-wave anisotropic tomography beneath Northeast Japan. evidence for slab dehydration in the forearc. Tectonophysics 364, 191–213. Phys. Earth Planet. Inter. 170, 115–133. Salah, M., Zhao, D., Lei, J., Farouk, M., 2005. Crustal heterogeneity beneath Wang, J., Zhao, D., 2009. P-wave anisotropic tomography of the crust and upper southwest Japan estimated from direct and Moho-reflected waves. mantle under Hokkaido, Japan. Tectonophysics 469, 137–149. Tectonophysics 395, 1–17. Wang, J., Zhao, D., 2010. Mapping P-wave anisotropy of the Honshu arc from Japan Satake, K., Nanayama, F., Yamaki, S., 2008. Fault models of unusual tsunami in the Trench to the back-arc. J. Asian Earth Sci. 39, 396–407. 17th century along the Kuril trench. Earth Planets Space 60, 925–935. Wang, J., Zhao, D., 2012. P wave anisotropic tomography of the Nankai subduction Scambelluri, M., Philippot, P., 2001. Deep fluids in subduction zones. Lithos 55, 213– zone in Southwest Japan. Geochem. Geophys. Geosyst. 13, Q05017. 227. Wang, J., Zhao, D., 2013. P-wave tomography for 3-D radial and azimuthal Scholz, C., Small, C., 1997. The effect of seamount subduction on seismic coupling. anisotropy of Tohoku and Kyushu subduction zones. Geophys. J. Int. 193, Geology 25, 487–490. 1166–1181. Sdrolias, M., Roest, W., Muller, R., 2004. An expression of Philippine Sea plate Wang, J., Zhao, D., Yao, Z., 2013. Crustal and uppermost mantle structure and rotation: the Parece Vela and Shikoku Basins. Tectonophysics 394, 69–86. seismotectonics of North China Craton. Tectonophysics 582, 177–187. Seno, T., Stein, S., Gripp, A., 1993. A model for the motion of the Philippine Sea plate Wang, J., Wu, H., Zhao, D., 2014. P wave radial anisotropy tomography of the upper consistent with NUVEL-1 and geological data. J. Geophys. Res. 98, 17941– mantle beneath the North China Craton. Geochem. Geophys. Geosyst. 15, 2195– 17948. 2210. Seno, T., Sakurai, T., Stein, S., 1996. Can the Okhotsk plate be discriminated from the Wang, Z., Zhao, D., 2005. Seismic imaging of the entire arc of Tohoku and Hokkaido North American plate? J. Geophys. Res. 101, 11305–11315. in Japan using P-wave, S-wave and sP depth-phase data. Phys. Earth Planet. Shao, G., Ji, C., Zhao, D., 2011. Rupture process of the 9 March, 2011 Mw 7.4 Sanriku- Inter. 152, 144–162. Oki, Japan earthquake constrained by jointly inverting teleseismic waveforms, Wang, Z., Zhao, D., 2006a. Seismic evidence for the influence of fluids on the 2005 strong motion data and GPS observations. Geophys. Res. Lett. 38, L00G20. west off Fukuoka prefecture earthquake in southwest Japan. Phys. Earth Planet. Sibson, R., 1992. Implications of fault-valve behavior for rupture nucleation and Inter. 155, 313–324. recurrence. Tectonophysics 211, 283–293. Wang, Z., Zhao, D., 2006b. Seismic images of the source area of the 2004 Mid-Niigata Silver, P., Beck, S., Wallace, T., Meade, C., Myers, S., James, D., Kuehnel, R., 1995. prefecture earthquake in Northeast Japan. Earth Planet. Sci. Lett. 244, 16–31. Rupture characteristics of the deep Bolivian earthquake of 9 June 1994 and the Wang, Z., Zhao, D., Wang, J., 2010. Deep structure and seismogenesis of the north– mechanism of deep-focus earthquakes. Science 268, 69–73. south seismic zone in Southwest China. J. Geophys. Res. 115, B12334. Stern, R., 2002. Subduction zones. Rev. Geophys. 40, RG000108. Wei, H., Liu, G., Gill, J., 2013a. Review of eruptive activity at Tianchi volcano, Sun, A., Zhao, D., Ikeda, M., Chen, Y., Chen, Q., 2008. Seismic imaging of southwest Changbaishan, northeast China: implications for possible future eruptions. Bull. Japan using P and PmP data: implications for arc magmatism and Volcanol. 75, 706–722. seismotectonics. Gondwana Res. 14, 535–542. Wei, W., Zhao, D., 2013. The 2008 Iwate-Miyagi earthquake (M 7.2) and arc Tahara, M., Uehira, K., Shimizu, H., Nakada, M., et al., 2008. Seismic velocity volcanism: insight from irregular-grid tomography. Earth Sci. Front. 20, 155–171. structure around the Hyuganada region, Southwest Japan, derived from Wei, W., Xu, J., Zhao, D., Shi, Y., 2012. East Asia mantle tomography: new insight into seismic tomography using land and OBS data and its implications for plate subduction and intraplate volcanism. J. Asian Earth Sci. 60, 88–103. interplate coupling and vertical crustal uplift. Phys. Earth Planet. Inter. 167, Wei, W., Zhao, D., Xu, J., 2013b. P-wave anisotropic tomography in Southeast Tibet: 19–33. new insight into the lower crustal flow and seismotectonics. Phys. Earth Planet. Tajima, F., Katayama, I., Nakagawa, T., 2009. Variable seismic structure near the Inter. 222, 47–57. 660 km discontinuity associated with stagnant slabs and geochemical Wiens, D., Conder, J., Faul, U., 2008. The seismic structure and dynamics of the implications. Phys. Earth Planet. Inter. 172, 183–198. mantle wedge. Annu. Rev. Earth Planet. Sci. 36, 421–455. Tajima, F., Mori, J., Kennett, B., 2013. A review of the 2011 Tohoku-oki earthquake Xia, S., Zhao, D., 2014. Late Mesozoic magmatic plumbing system in the onshore– (Mw 9.0): large-scale rupture across heterogeneous plate coupling. offshore area of Hong Kang: Insight from 3-D active-source seismic Tectonophysics 586, 15–34. tomography. J. Asian Earth Sci. 96, 46–58. Tamura, Y., Tatsumi, Y., Zhao, D., Kido, Y., Shukuno, H., 2002. Hot fingers in the Xia, S., Zhao, D., Qiu, X., 2008a. The 2007 Niigata earthquake: effect of arc magma mantle wedge: new insight into magma genesis in subduction zones. Earth and fluids. Phys. Earth Planet. Inter. 166, 153–166. Planet. Sci. Lett. 197, 105–116. Xia, S., Zhao, D., Qiu, X., 2008b. Tomographic evidence for the subducting oceanic Tatsumi, Y., 1989. Migration of fluid phases and genesis of basalt magmas in crust and forearc mantle serpentinization under Kyushu, Japan. Tectonophysics subduction zones. J. Geophys. Res. 94, 4697–4707. 449, 85–96. D. Zhao / Journal of Asian Earth Sciences 98 (2015) 26–49 49

Xia, S., Zhao, D., Qiu, X., Nakajima, J., Matsuzawa, T., Hasegawa, A., 2007. Mapping Zhao, D., Kanamori, H., Wiens, D., 1997. State of stress before and after the 1994 the crustal structure under active volcanoes in central Tohoku, Japan using P Northridge earthquake. Geophys. Res. Lett. 24, 519–522. and PmP data. Geophys. Res. Lett. 34, L10309. Zhao, D., Mishra, O.P., Sanda, R., 2002. Influence of fluids and magma on Xu, J., Liu, G., Wu, J., Ming, Y., et al., 2012. Recent unrest of Changbaishan volcano, earthquakes: seismological evidence. Phys. Earth Planet. Inter. 132, 249–267. northeast China: a precursor of a future eruption? Geophys. Res. Lett. 39, Zhao, D., Lei, J., Tang, R., 2004. Origin of the Changbai intraplate volcanism in L16305. Northeast China: evidence from seismic tomography. Chin. Sci. Bull. 49, 1401– Yagi, Y., Kikuchi, M., Yoshida, S., Yamanaka, Y., 1998. Source process of the Hyuga- 1408. nada earthquake of April 1, 1968 (Mjma 7.5), and relationship to the subsequent Zhao, D., Todo, S., Lei, J., 2005. Local earthquake reflection tomography of the seismicity. J. Seismol. Soc. Jpn. 51, 139–148. Landers aftershock area. Earth Planet. Sci. Lett. 235, 623–631. Yagi, Y., Kikuchi, M., Yoshida, S., Sagiya, T., 1999. Comparison of the coseismic Zhao, D., Wang, Z., Umino, N., Hasegawa, A., 2007a. Tomographic imaging outside a rupture with the aftershock distribution in the Hyuga-nada earthquakes of seismic network: application to the northeast Japan arc. Bull. Seismol. Soc. Am. 1996. Geophys. Res. Lett. 26, 3161–3164. 97, 1121–1132. Yamanaka, Y., 2006. Asperity map in Hokkaido and southern Kurile arc. Earth Mon. Zhao, D., Maruyama, S., Omori, S., 2007b. Mantle dynamics of western Pacific to East 28, 427–431. Asia: new insight from seismic tomography and mineral physics. Gondwana Yamanaka, Y., Kikuchi, M., 2003. Source process of the recurrent Tokachi-oki Res. 11, 120–131. earthquake on September 26, 2003, inferred from teleseismic body waves. Earth Zhao, D., Lei, J., Liu, L., 2008. Seismic tomography of the Moon. Chin. Sci. Bull. 53, Planets Space 55, e21–e24. 3897–3907. Yamanaka, Y., Kikuchi, M., 2004. Asperity map along the subduction zone in Zhao, D., Wang, Z., Umino, N., Hasegawa, A., 2009a. Mapping the mantle wedge and northeastern Japan inferred from regional seismic data. J. Geophys. Res. 109, interplate thrust zone of the northeast Japan arc. Tectonophysics 467, 89–106. B07307. Zhao, D., Tian, Y., Lei, J., Liu, L., Zheng, S., 2009b. Seismic image and origin of the Yamashita, Y., Shimizu, H., Goto, K., 2012. Small repeating earthquake activity, Changbai intraplate volcano in East Asia: role of big mantle wedge above the interplate quasi-static slip, and interplate coupling in the Hyuga-nada, stagnant Pacific slab. Phys. Earth Planet. Inter. 173, 197–206. southwestern Japan subduction zone. Geophys. Res. Lett. 39, L08304. Zhao, D., Santosh, M., Yamada, A., 2010. Dissecting large earthquakes in Japan: role Ye, L., Li, J., Tseng, T., Yao, Z., 2011. A stagnant slab in a water bearing mantle of arc magma and fluids. Island Arc 19, 4–16. transition zone beneath northeast China: implications from regional SH Zhao, D., Huang, Z., Umino, N., Hasegawa, A., Yoshida, T., 2011a. Seismic imaging of waveform modeling. Geophys. J. Int. 186, 706–710. the Amur-Okhotsk plate boundary zone in the Japan Sea. Phys. Earth Planet. Ye, L., Lay, T., Kanamori, H., 2012. The Sanriku-Oki low-seismicity region on the Inter. 188, 82–95. northern margin of the great 2011 Tohoku-Oki earthquake rupture. J. Geophys. Zhao, D., Huang, Z., Umino, N., Hasegawa, A., Kanamori, H., 2011b. Structural Res. 117. http://dx.doi.org/10.1029/2011JB008847. heterogeneity in the megathrust zone and mechanism of the 2011 Tohoku-oki Yoshida, K., Hasegawa, A., Okada, T., Iinuma, T., Ito, Y., Asano, Y., 2012. Stress before earthquake (Mw 9.0). Geophys. Res. Lett. 38, L17308. and after the 2011 great Tohoku-oki earthquake and induced earthquakes in Zhao, D., Wei, W., Nishizono, Y., Inakura, H., 2011c. Low-frequency earthquakes and inland areas of eastern Japan. Geophys. Res. Lett. 39, L03302. tomography in western Japan: insight into fluid and magmatic activity. J. Asian Zhao, D., 2001. Seismic structure and origin of hotspots and mantle plumes. Earth Earth Sci. 42, 1381–1393. Planet. Sci. Lett. 192, 251–265. Zhao, D., Yanada, T., Hasegawa, A., Umino, N., Wei, W., 2012a. Imaging the Zhao, D., 2004. Global tomographic images of mantle plumes and subducting slabs: subducting slabs and mantle upwelling under the Japan Islands. Geophys. J. Int. insight into deep Earth dynamics. Phys. Earth Planet. Inter. 146, 3–34. 190, 816–828. Zhao, D., 2012. Tomography and dynamics of Western-Pacific subduction zones. Zhao, D., Arai, T., Liu, L., Ohtani, E., 2012b. Seismic tomography and geochemical Monogr. Environ., Earth Planets 1, 1–70. evidence for lunar mantle heterogeneity: comparing with Earth. Global Planet. Zhao, D., Tian, Y., 2013. Changbai intraplate volcanism and deep earthquakes in East Change 90, 29–36. Asia: a possible link? Geophys. J. Int. 195, 706–724. Zhao, D., Yamamoto, Y., Yanada, T., 2013. Global mantle heterogeneity and its Zhao, D., Hasegawa, A., Horiuchi, S., 1992. Tomographic imaging of P and S wave influence on teleseismic regional tomography. Gondwana Res. 23, 595–616. velocity structure beneath northeastern Japan. J. Geophys. Res. 97, 19909– Zhao, D., Kitagawa, H., Toyokuni, G., 2015. A water wall in the Tohoku fore-arc 19928. causing large crustal earthquakes. Geophys. J. Int. 200, 149–172. Zhao, D., Kanamori, H., Negishi, H., Wiens, D., 1996. Tomography of the source area Zou, H., Fan, Q., Yao, Y., 2008. U–Th systematics of dispersed young volcanoes in NE of the 1995 Kobe earthquake: evidence for fluids at the hypocenter? Science China: asthenosphere upwelling caused by piling up and upward thickening of 274, 1891–1894. stagnant Pacific slab. Chem. Geol. 255, 134–142.