SR

Da

b

c

d

a

ARRA

KCSBSI

1

nacTt1122aa

aouuncs

0d

Physics of the Earth and Planetary Interiors 173 (2009) 197–206

Contents lists available at ScienceDirect

Physics of the Earth and Planetary Interiors

journa l homepage: www.e lsev ier .com/ locate /pepi

eismic image and origin of the Changbai intraplate volcano in East Asia:ole of big mantle wedge above the stagnant Pacific slab

apeng Zhaoa,∗, You Tiana,b, Jianshe Lei c, Lucy Liua, Sihua Zhengd

Department of Geophysics, Tohoku University, Sendai 980-8578, JapanCollege of Geoexploration Science and Technology, Jilin University, Changchun 130026, ChinaInstitute of Crustal Dynamics, China Earthquake Administration, Beijing 100085, ChinaInstitute of Earthquake Science, China Earthquake Administration, Beijing 100036, China

r t i c l e i n f o

rticle history:eceived 23 June 2008eceived in revised form 17 November 2008ccepted 23 November 2008

a b s t r a c t

We determined P-wave tomography of the crust and upper mantle under the Changbai volcanic areain Northeast Asia by using 289,318 arrival-time data from 4802 local earthquakes and 9599 teleseismicevents. A prominent low-velocity anomaly is clearly imaged in the crust and upper mantle down to about

eywords:hangbai volcanotagnant Pacific slabig mantle wedgeeismic tomography

410 km depth beneath the Changbai volcano. A broad high-velocity anomaly is revealed in the mantletransition zone and deep earthquakes occurred actively there, suggesting that the subducting Pacific slabis stagnating in the mantle transition zone under East Asia. Our results suggest that the formation ofthe Changbai and other intraplate volcanoes in Northeast Asia is related to the upwelling of hot andwet asthenospheric materials in the big mantle wedge above the stagnant Pacific slab. Hence the activeintraplate volcanoes in Northeast Asia are not hotspots related to deep mantle plumes but are caused by

n the

ntraplate volcanism plate tectonic processes i

. Introduction

In the East Asia continent there are many Cenozoic volca-oes, but only a few are still active now, such as the Changbaind Wudalianchi volcanoes in Northeast (NE) China and Teng-hong volcano in Southwest China (Liu, 1999; Liu, 2000) (Fig. 1).he three active volcanoes have erupted many times in the his-ory. For example, the Changbai volcano erupted six times in BC120, AD 1050, 1413, 1597, 1668, and 1702 (Simkin and Siebert,994), the Wudalianchi volcano erupted during 1719–1721 (Liu,000), and the Tengchong volcano erupted during 1465–1620 (Liu,000). Although many studies have been made by using variouspproaches, the origins of these intraplate volcanoes in East Asiare still not very clear.

The Changbai intraplate volcano is located close to the bound-ry between China and Korea (Fig. 1). It is also called Tianchir Baiton-shan volcano. In recent years many researchers havesed multidisciplinary approaches to investigate the crust and

pper-mantle structure of the Changbai volcano. For example, mag-etotelluric soundings detected a high-conductivity layer in therust under the volcano (Tang et al., 2001); seismic explosiontudies revealed low-velocity (low-V) anomalies in the crust and

∗ Corresponding author.E-mail address: zhao@aob.geophys.tohoku.ac.jp (D. Zhao).

031-9201/$ – see front matter © 2008 Elsevier B.V. All rights reserved.oi:10.1016/j.pepi.2008.11.009

upper mantle.© 2008 Elsevier B.V. All rights reserved.

uppermost mantle under the volcano, which are considered toreflect magma chambers (Zhang et al., 2002). Some researchers con-sidered the Changbai volcano as a hotspot like Hawaii (e.g., Turcotteand Schubert, 1982), while others suggested that its formation wasassociated with the subduction of the Pacific plate, upwelling ofthe hot asthenospheric materials, and lithospheric fractures (e.g.,Tatsumi et al., 1990).

Because of the lack of digital seismic stations in NE Asia, theseismic structure of the crust and mantle under this region hadnot been determined till the very recent time. A portable seismicnetwork consisting of 19 stations was installed in the Changbaivolcanic area from June 1998 to April 1999, which recorded seis-mic waveforms from some distant earthquakes (Wu and Hetland,1999). The waveform data were used to estimate the depth dis-tribution of the Moho, 410 and 660 km discontinuities under theChangbai area with teleseismic receiver-function methods (Ai etal., 2003; Li and Yuan, 2003; Hetland et al., 2004). Zhao et al.(2004) and Lei and Zhao (2005) collected 548 high-quality P-wave arrival times from 68 teleseismic events recorded by the19 portable stations and three permanent stations (Mudanjiang,Beijing, and Hailaer) which belong to the Chinese Digital SeismicNetwork (CDSN), and used the data to determine the first 3-D P-

wave velocity structure of the upper mantle under the Changbaivolcano. Their results show a prominent low-V zone down to about400-km depth under the Changbai volcano and a high-velocity(high-V) anomaly in the mantle transition zone. However, the pre-vious tomographic images (Zhao et al., 2004; Lei and Zhao, 2005)

198 D. Zhao et al. / Physics of the Earth and Planetary Interiors 173 (2009) 197–206

F les sht ows th

hs

Pbdra

F(

ig. 1. Map showing surface topography of Western Pacific to East Asia. Red trianghe Wadati-Benioff deep seismic zone in the subducting Pacific slab. The red box sh

ave a low resolution because of the small data set used in thosetudies.

In the present work we have determined a high-resolution 3-D

-wave velocity structure down to 650 km depth under the Chang-ai volcano by using simultaneously a great number of arrival-timeata from local and regional earthquakes and teleseismic eventsecorded by many permanent and portable seismic stations in andround Changbai. Our present results shed new light on the deep

ig. 2. (a) Distribution of 645 seismic stations (black triangles) used in this study. Solid sqb) Epicenter distribution of 4802 local and regional earthquakes used in this study. The r

ow the three active intraplate volcanoes. Dashed lines show the depth contours ofe present study area.

structure and origin of the Changbai intraplate volcano and mantledynamics under East Asia.

2. Data and method

In this work we used 645 analog and digital seismic stationsof the Chinese Seismic Network and 19 portable seismic stationsdeployed by Wu and Hetland (1999) (Figs. 2a and 3). The 19 portable

uares show the 19 portable seismic stations deployed around the Changbai volcano.ed triangles show the location of the Changbai volcano.

D. Zhao et al. / Physics of the Earth and Planetary Interiors 173 (2009) 197–206 199

Fig. 3. (a) Map showing the surface topography of East Asia. The box shows the location of the present study area enlarged in (b). The three triangles show the intraplatevolcanoes in China (from north to south: Wudalianchi, Changbai, and Tengchong). (b) Tectonic background around the Changbai volcano (the red triangle). Solid lines showthe major active faults. Black and yellow triangles denote the permanent and portable seismic stations, respectively.

Fig. 4. Epicenter locations of the 9599 teleseismic events (star symbols) used in thisstudy. The concentric circles show the distances (30, 60 and 90◦) from the center ofthe study area.

Fig. 5. One-dimensional P-wave velocity model used in this study (grey lines) whichis modified slightly from the iasp91 Earth model (dashed lines).

2 d Plan

smbHesu

ea1esL

matrt

FaT(

00 D. Zhao et al. / Physics of the Earth an

tations were located around the Changbai volcano (Fig. 3), whileost of the permanent seismic stations are located west of Chang-

ai and there are no stations on the Korea Peninsula and Japan Sea.owever, the use of data from many teleseismic events in the east-rn hemisphere (Fig. 4) recorded by the portable and permanenttations has provided crisscrossing seismic rays in the upper mantlender the Changbai volcano.

We collected 49,799 arrival times from 4802 local and regionalarthquakes occurred during 1980–2005 (Fig. 2b), and 239,519rrival times from 9599 teleseismic events occurred during980–2002 (Fig. 4). The picking accuracy of the arrival times isstimated to be 0.1–0.2 s. The 548 teleseismic data recorded by 22tations in the Changbai area measured by Zhao et al. (2004) andei and Zhao (2005) are also used in this study.

To analyze the arrival-time data, we used the tomographic

ethod of Zhao et al. (1994, 2004) to combine the teleseismic rel-

tive residuals with local and regional earthquake arrival times inhe tomographic inversion. The local and regional earthquakes areelocated in the inversion process. Hypocenter parameters of theeleseismic events determined by Dr. E.R. Engdahl with a procedure

ig. 6. Plan views of P-wave tomography at different depth slices. The depth is shown at thnomalies (in %) from the 1-D velocity model shown in Fig. 5. The velocity perturbation she bold black lines in (a–d) show the active faults. The cross symbols show the earthquag–i).

etary Interiors 173 (2009) 197–206

described by Engdahl et al. (1998) are used in this study and theyare fixed in the inversion process because relative teleseismic resid-uals are used (see Zhao et al., 1994, 2006). The 1-D starting model(Fig. 5) for the 3-D tomographic inversion is derived from the iasp91Earth model (Kennett and Engdahl, 1991) with slight changes in thecrust and uppermost mantle according to a previous tomographicstudy for East Asia (Huang and Zhao, 2006). A 3-D grid is arrangedin the modeling space. P-wave velocity perturbations at the gridnodes are taken as unknown parameters. The velocity perturba-tion at any point in the model is calculated by linearly interpolatingthe velocity perturbations at the eight grid nodes surrounding thatpoint. A 3-D ray tracing method is used to calculate travel timesand ray paths (Zhao et al., 1992; Zhao, 2004). The large and sparsesystem of observation equations is resolved by using a conjugate-gradient algorithm with damping and smoothing regulations (Zhao,

2004).

We conducted many inversions and synthetic tests with dif-ferent grid intervals, and found that the optimal grid spacing is0.8 × 0.8◦ in the horizontal direction and 20–80 km in depth. Thetotal number of grid nodes in the model is 7488, but only the

e lower-right corner of each map. Red and blue colors denote slow and fast velocitycale is shown at the bottom. The red triangles denote the active Changbai volcano.kes within 20 km depth of each layer in (a and b) and within 50 km of each layer in

D. Zhao et al. / Physics of the Earth and Planetary Interiors 173 (2009) 197–206 201

F omogra ion scd cross sa

vw

3

vbracitiattaHsdtgandt

dt

ig. 7. North-south (a) and east-west (b) vertical cross sections of P-wave velocity tnomalies (in %) from the 1-D velocity model shown in Fig. 5. The velocity perturbatashed lines show the 410-km discontinuity. (c) Map showing locations of the twond white dots show the shallow (0–30 km) and deep (450–600 km) earthquakes.

elocity perturbations at 5987 grid nodes with hit count over 30ere resolved by inverting the 289,318 travel-time data.

. Results and resolution analyses

Figs. 6 and 7 show the final P-wave tomographic images in planiews and along two vertical cross sections beneath the Chang-ai volcano. The images are generally consistent with the previousesults (Zhao et al., 2004; Lei and Zhao, 2005). A clear low-Vnomaly is visible down to 410 km depth under the Changbai vol-ano. In the N–S vertical cross section (Fig. 7b) the low-V zones confined right beneath the Changbai volcano, while it spreadsoward the west and east in the asthenosphere (250–400 km depth)n the E–W cross section (Figs. 6 and 7a). A continuous high-Vnomaly is clearly visible in the lower portion of the mantle transi-ion zone, and deep earthquakes occurred within the high-V zone inhe depth range of 480–600 km. Referring to the previous regionalnd global tomography results (Zhao, 2004; Lei and Zhao, 2005;uang and Zhao, 2006), we believe that the high-V zone repre-

ents the stagnant Pacific slab in the transition zone. A significantifference between the present result and the previous ones ishat the stagnant Pacific slab is about 100 km thick in our tomo-raphic images (Fig. 7), while it was imaged as a high-V zone ofbout 200 km thick by the previous tomographic studies. This sig-ificant improvement in the imaging of the stagnant Pacific slab is

ue to the much better data set used in this work, which leads tohe higher-resolution of our present tomography.

To confirm the main features of the tomographic result, we con-ucted extensive resolution analyses. Here we show the results ofwo synthetic tests. One is checkerboard resolution test (CRT), the

aphy along profiles shown in (c). Red and blue colors denote slow and fast velocityale is shown beside (a). The black triangles denote the active Changbai volcano. Theections in (a) and (b). The red triangle denotes the Changbai volcano. Black crosses

other is a synthetic resolution test. In CRT, we assigned positive andnegative velocity perturbations of ±3% alternately to the grid nodesand calculated travel times for this model to make synthetic data,and then we inverted the synthetic data. Fig. 8 shows the resultsof CRT along two profiles as shown in Fig. 7c. As a whole, the CRTresults are very good and the checkerboard pattern is well recov-ered under the Changbai volcano down to 660 km depth, suggestingthat the main features of the tomographic images under Changbaiare reliable.

Fig. 9 shows the results of a synthetic test for assessing the reli-ability of the high and low velocity zones beneath the Changbaivolcano along the two profiles shown in Fig. 7c. Unlike the CRT,the pattern in the synthetic test is not the alternative positive andnegative velocity perturbations, but instead some main featuresappeared in the obtained tomographic result (Fig. 7). It is clear thatthe input velocity anomalies are generally well recovered, thoughthere are some differences in the amplitude of velocity anomalies(Fig. 9). This proves the reliability of the low-V anomaly in the depthrange of 0-410 km and the high-V zone in the mantle transition zonebeneath the Changbai volcano.

4. Discussion

4.1. Other seismic evidence

Fig. 10 shows two vertical cross sections of whole-mantle P-wave tomography under the Changbai and Wudalianchi volcanoesfrom the global tomography model of Zhao (2004). About one mil-lion arrival times of first P-wave and later phases (pP, PP, PcP, andPdiff waves) were used in the inversion, and depth variations of the

202 D. Zhao et al. / Physics of the Earth and Planetary Interiors 173 (2009) 197–206

F ownv f the C

M(PaPVafB

sttncttsiw

utdfdl1

ig. 8. Results of a checkerboard resolution test in two vertical cross sections as shelocity perturbation scale is shown on the right. The triangles show the location o

oho, 410 and 660 km discontinuities were also taken into accountfor details, see Zhao, 2004). In the upper mantle, the subductingacific slab is imaged clearly and earthquakes occurred down tobout 600 km depth within the slab (Fig. 10). Under East Asia, theacific slab becomes stagnant in the transition zone. Pieces of high-anomalies are visible in the lower mantle under the stagnant slab,

nd they have piled up in the lowermost mantle (Fig. 10). Similareatures were also found in other global tomographic models (e.g.,ijwaard et al., 1998).

We also estimated the stress regime in the subducting Pacificlab by using the Harvard CMT solutions of deep earthquakes underhe Japan Sea and East Asia (Ekstrom et al., 2005). The result showshat the compressive stress axes of almost all deep earthquakes areearly parallel with the down-dip direction of the slab (Fig. 11), indi-ating that the Pacific slab is under the compressive stress regime inhe depth range of 200–600 km. This result is well consistent withhe earlier findings by Isacks and Molnar (1971). We consider thatuch a stress regime in the slab is caused by the stagnant slab meet-ng strong resistance at the 660-km discontinuity, being consistent

ith the seismic tomography results (Figs. 7 and 10).All these seismological results suggest that the subducting slab

nder NE Asia does not easily penetrate down to the lower man-le, but meets strong resistance when it encounters the 660-km

iscontinuity. The slab bends horizontally, and accumulates thereor a long time (ca. 100–140 m.y.), and then finally collapses to fallown as blobs onto the core-mantle boundary as a result of very

arge gravitational instability from phase transitions (Maruyama,994; Mitrovica et al., 2000; Zhao, 2004). The receiver-function

in Fig. 7c. Black and white circles denote low and high velocities, respectively. Thehangbai volcano. The dashed lines show the 410-km discontinuity.

studies using the waveform data from the portable seismic networkobtained similar results (Ai et al., 2003; Li and Yuan, 2003). Fig. 10also shows that very slow anomalies exist in the upper mantle rightbeneath the Wudalianchi and Changbai volcanoes, right above thestagnant Pacific slab in the mantle transition zone. The global tomo-graphic images (Fig. 10) are generally consistent with the regionaltomography results (Fig. 7; Huang and Zhao, 2006). These resultsare also quite similar to the images under the Fiji-Tonga regionwhere the back-arc volcanoes in Fiji and the Lau spreading cen-ter are located above very slow anomalies in the mantle wedgeright above the subducting Tonga slab (Zhao et al., 1997; Conderand Wiens, 2006).

4.2. Big mantle wedge model

There are several types of volcanoes on Earth, such as the mid-ocean ridge volcanoes, subduction zone volcanoes, hotspots causedby deep mantle plumes, and intraplate volcanoes associated withupwelling of hot asthenospheric materials and lithospheric frac-tures (e.g., Tatsumi et al., 1990). Apparently, the Changbai andWudalianchi volcanoes are not mid-ocean ridge volcanoes. Becausethe subducting Pacific slab is stagnant in the mantle transition zoneunder NE Asia, the volcanoes are not hotspot like Hawaii, Iceland,

and Eifel (Ritter et al., 2001; Waite et al., 2006).

The regional and global tomographic images (Figs. 7 and 10) sug-gest that the Changbai and Wudalianchi volcanoes are a kind ofback-arc intraplate volcanoes whose formations are closely relatedto the deep subduction of the Pacific slab and its stagnancy in the

D. Zhao et al. / Physics of the Earth and Planetary Interiors 173 (2009) 197–206 203

F olcanom phicv of the

msul

FfTs

ig. 9. Results of a synthetic test for the velocity anomalies under the Changbai vodels, while the right panels show the output models resulting from the tomogra

elocity perturbation scale is shown at the bottom. The triangles show the location

antle transition zone. The broad upper-mantle region above thetagnant Pacific slab may have formed a big mantle wedge (BMW)nder the Japan Sea and NE China (Fig. 10). Slow velocity anoma-

ies in the back-arc region are generally associated with the back-arc

ig. 10. (a and b) Vertical cross sections of whole-mantle P-wave tomography (Zhao, 2004)rom the 1-D iasp91 Earth model. The velocity perturbation scale is shown below the crhe two lines show the 410 and 660 km discontinuities. Solid triangles show the intraplahowing locations of the two cross sections in (a) and (b). Triangles denote hotspots or in

along AB (a) and CD (b) profiles shown in Fig. 7c. The left panels show the inputinversion. Black and white circles denote low and high velocities, respectively. TheChangbai volcano. The dashed lines show the 410-km discontinuity.

magmatism and volcanism caused by the deep dehydration processof the subducting slab and the convective circulation process of themantle wedge (Zhao et al., 1994, 1997; Stern, 2002). These processesmay have led to the large-scale upwelling of the hot asthenospheric

. Red and blue colors denote slow and fast velocity perturbations (in %), respectively,oss sections. White dots denote earthquakes within 150-km width of the profiles.te volcanoes. The reverse triangles show the location of the Japan Trench. (c) Maptraplate volcanoes on Earth.

204 D. Zhao et al. / Physics of the Earth and Planetary Interiors 173 (2009) 197–206

Fig. 11. (a) Plan view and (b) east-west vertical cross section showing the distribution of compressive axis (P-axis) of deep earthquakes under the Japan Sea and East Asia.The red triangles denote active and Quaternary intraplate volcanoes. The vertical cross section is along line A–B shown in (a). The two dashed lines show the 410 and 660 kmdiscontinuities.

Fig. 12. (a) Surface tectonic features in Northwest Pacific and Northeast Asia. Black patches denote the Cenozoic basalts. (A) Baikal rift; (B) Shanxi graben; (C) Tancheng-Lujiangfault zone; (D) Okinawa trough. (b) A schematic east-west vertical cross section showing the upper-mantle structure beneath Northeast Asia. The subducting Pacific slabbecomes stagnant in the mantle transition zone. The convective circulation process in the big mantle wedge (BMW) and deep dehydration reactions of the subducting slabcause upwelling of hot and wet asthenospheric materials, leading to the formation of the continental rift systems as well as intraplate volcanoes in Northeast Asia (modifiedfrom Tatsumi et al., 1990 and Zhao et al., 2004).

d Planetary Interiors 173 (2009) 197–206 205

mic(amizDh1

ebbmsvfmos2ctesspi

isWlltjodt

Nwidtdwlad1tmtecs(dtza

v

D. Zhao et al. / Physics of the Earth an

aterials under NE Asia and caused the continental rift systems andntraplate volcanism in the region. Because the very old (hence veryold) Pacific plate is subducting beneath East Asia at a rapid rate7–10 cm/year), the dehydration reactions may not fully completet the shallow depth (100–200 km) of the mantle. Hydrous Mg–Siinerals in the subducting Pacific slab may continue to release flu-

ds through dehydration reactions at the depths of mantle transitionone (Inoue, 1994; Ohtani et al., 2004; Komabayashi et al., 2004).eep dehydration reactions of the rapidly subducting Pacific slabave also been found in the Tonga subduction zone (Zhao et al.,997; Conder and Wiens, 2006).

Tatsumi et al. (1990) first invoked the asthenospheric injection toxplain the formation of the Wudalianchi and Changbai volcanoes,ut they did not consider the stagnant Pacific slab under the regionecause such a slab structure was unknown at that time. Here weodify their model to emphasize the roles of the stagnant Pacific

lab and the big mantle wedge in the formation of the intraplateolcanism in East Asia (Fig. 12b). The extensional rift systems andaults widely existing in East Eurasia (Fig. 12a) may be the surface

anifestation of a deep dynamic process such as deep subductionf the Pacific slab, upwelling of hot asthenospheric materials, litho-pheric fractures and intracontinental rifting (e.g., Zhao et al., 2004,006). Recent results of S-wave splitting (Liu et al., 2008) and geo-hemical analysis (Chen et al., 2007; Zou et al., 2008) also supporthis BMW model. Zou et al. (2008) presented U-series data and tracelement data which argue against the deep dehydration process anduggested that the piling up and thickening of the stagnant Pacificlab in the transition zone, together with the convective circulationrocess in the BMW, help drive the asthenosphere upwelling and

nduce decompression melting.In this work we determined high-resolution tomographic

mages of the upper mantle under the active Changbai volcano, butuch a detailed mantle tomography is still not yet available for the

udalianchi volcano. Global tomographic images show that thearge-scale structure under the Wudalianchi volcano is quite simi-ar to that under the Changbai volcano (Fig. 10). Hence we considerhat the two volcanoes have the same origin. To confirm this con-ecture, it is necessary to install a dense seismic network of portabler permanent stations in the Wudalianchi area in the near future toetermine a detailed 3-D structure of the crust and mantle underhe volcano to clarify its origin.

Recently water content and geotherm of the upper mantle underE China were estimated from the electrical conductivity and P-ave velocity under the region (Ichiki et al., 2006) (Fig. 13). It

s found that in the deep part of the upper mantle (250–400 kmepth), neither the dry pyrolite nor the dry harzburgite condi-ion provide consistent electrical and seismic geotherms, while theiscrepancy can be explained by allowing for a small amount ofater (500–1000 ppm H/Si) with the seismic geotherm. In the shal-

ow mantle (<250 km depth), the electrical and seismic geothermsre consistent with each other within 1500–1700 ◦C under thery harzburgite condition, but they are inconsistent by more than00 ◦C under the dry pyrolite condition (Fig. 13b). Alternatively,he wet pyrolite condition applied to the deeper part of the upper

antle also satisfies the electrical conductivity and seismic struc-ure in the shallow mantle (Fig. 13a). These results suggest that thentire upper mantle or at least the asthenosphere under NE Asiaontains fluids, which is consistent with the tomographic resultshowing prominent low-velocity anomalies down to 410 km depthFigs. 7 and 10). The wet upper mantle is possibly caused by theeep dehydration reactions of the stagnant Pacific slab that con-

ains sufficient amount of hydrous minerals down to the transitionone depths (Shieh et al., 1998; Ohtani et al., 2004; Komabayashi etl., 2004).

Recent studies using GPS (global positioning system) obser-ations indicate that not all the plate boundaries are narrow

Fig. 13. Two plausible models on the wet upper mantle beneath Northeast Asia esti-mated from electrical and seismic observations (modified from Ichiki et al., 2006).

deformation zones as assumed by the theory of plate tectonicswhen it was first proposed (Stein and Freymueller, 2002). Someplate boundaries can be very broad deformation zones, for example,the continental collision between the Indian and Eurasian plates hasresulted in the Tibetan plateau of a few thousands kilometers wide.The Changbai and Wudalianchi volcanoes are located in the interiorof the Eastern Asia continent, leaving the Japan Trench farther than1000 km (Fig. 1). The present and previous tomographic imagesshow that the subducting Pacific slab is stagnant in the mantle tran-sition zone under NE Asia, and it has greatly affected the intraplateseismic and volcanic activities in NE China (Zhao, 2004; Huang andZhao, 2006). These results suggest that the western Pacific plateboundary is a deformation zone as broad as over 1000 km wide,and the subduction of the Pacific plate has far-reaching effects onthe volcanism and seismotectonics of the interior of the Eurasiancontinent and mantle dynamics under this wide region.

Our present results on the origin of the active intraplate vol-canoes in NE Asia suggest that not all the intraplate volcanoes arecaused by deep mantle plumes (Wilson, 1963; Morgan, 1971). Someof the intraplate volcanism as that in NE Asia can be explained by theplate tectonic processes in the upper mantle and the mantle tran-sition zone, as pointed out earlier by Anderson (1995) and Foulger(2003). However, the nature is complex, and various processes maycoexist in the mantle. Some of the strong hotspots, such as those inHawaii, Iceland, South Pacific and Africa, are associated with deepmantle plumes (Zhao, 2004, 2007).

5. Conclusions

We used a large number of arrival-time data from local, regionaland teleseismic events to determine a 3-D P-wave structure ofthe crust and upper mantle under the Changbai volcanic area inNortheast Asia. Our results show a prominent low-velocity anomalyfrom the surface down to 410 km depth beneath the Changbai vol-cano and a broad high-velocity anomaly in the mantle transition

zone, which are generally consistent with the previous global andregional tomographic studies for East Asia. Focal-mechanism solu-tions of deep earthquakes indicate that the subducting Pacific slabunder the Japan Sea and the East Asia margin is subject to compres-sive stress regime. These results suggest that the Pacific slab meets

2 d Plan

ssmteovtnftmuishcr

A

1ZSpTKc

R

A

A

B

C

C

E

E

FH

H

I

I

06 D. Zhao et al. / Physics of the Earth an

trong resistance at the 660-km discontinuity and so it becomestagnant in the mantle transition zone under NE Asia. The upperantle under NE Asia has formed a big mantle wedge (BMW) above

he stagnant slab. The BMW exhibits low seismic-velocity and highlectrical-conductivity, which is hot and perhaps also wet becausef the deep dehydration reactions of the stagnant slab and the con-ective circulation process in the BMW. These processes lead tohe upwelling of hot and wet asthenospheric materials and thin-ing and fracturing of the continental lithosphere as well as the

ormation of the active intraplate volcanoes in NE Asia. Thereforehe active intraplate volcanism in NE Asia is not related to deep

antle plumes but is caused by the plate tectonic processes in thepper mantle and the mantle transition zone. A better understand-

ng of these processes can be achieved by deploying a network ofeismic stations on the Korea Peninsula and Japan Sea to determineigher-resolution mantle tomography under NE Asia in addition toonducting other geophysical and geochemical investigations of theegion.

cknowledgements

This work was partially supported by a grant (Kiban-A7204037) from Japan Society for the Promotion of Science to D.hao. Some of the figures are made by using GMT (Wessel andmith, 1998). We are grateful to Prof. Ruomei Sun for providingart of the data used in this work and to Profs. E. Ohtani, H. Zou, J.eng, R. Sun, Y. Taniguti and S. Ueki for thoughtful discussions. Prof.. Zhang and an anonymous referee provided constructive reviewomments which improved the manuscript.

eferences

i, Y., Zheng, T., Xu, W., He, Y., Dong, D., 2003. A complex 660 km discontinuitybeneath northeast China. Earth Planet. Sci. Lett. 212, 63–71.

nderson, D., 1995. Lithosphere, asthenosphere and perisphere. Rev. Geophys. 33,125–149.

ijwaard, H., Spakman, W., Engdahl, E., 1998. Closing the gap between regional andglobal travel time tomography. J. Geophys. Res. 103, 30055–30078.

hen, Y., Zhang, Y., Graham, D., Su, S., Deng, J., 2007. Geochemistry of Cenozoic basaltsand mantle xenoliths in Northeast China. Lithos 96, 108–126.

onder, J., Wiens, D., 2006. Seismic structure beneath the Tonga arc and Lau back-arcbasin determined from joint Vp, Vp/Vs tomography. Geochem. Geophys. Geosyst.7, Q03018.

kstrom, G., Dziewonski, A., Maternovskaya, N., Nettles, M., 2005. Global seismicityof 2003: centroid-moment-tensor solutions for 1087 earthquakes. Phys. EarthPlanet. Inter. 148, 327–351.

ngdahl, R., van der Hilst, R., Buland, R., 1998. Global teleseismic earthquake reloca-tion with improved travel times and procedures for depth determination. Bull.Seismol. Soc. Am. 88, 722–743.

oulger, G., 2003. Plumes, or plate tectonic processes? Astron. Geophys 43, 6.19-6.23.etland, E., Wu, F., Song, J., 2004. Crustal structure in the Changbaishan volcanic

area, China, determined by modeling receiver functions. Tectonophysics 386,157–175.

uang, J., Zhao, D., 2006. High-resolution mantle tomography of China and surround-ing regions. J. Geophys. Res. 111, B09305.

chiki, M., Baba, K., Obayashi, M., Utada, H., 2006. Water content and geotherm in theupper mantle above the stagnant slab: interpretation of electrical conductivityand seismic P-wave velocity models. Phys. Earth Planet. Inter. 155, 1–15.

noue, T., 1994. Effect of water on melting phase relations and melt compositionin the system Mg2SiO4–MgSiO3–H2O up to 15 GPa. Phys. Earth Planet. Inter 85,237–263.

etary Interiors 173 (2009) 197–206

Isacks, B., Molnar, P., 1971. Distribution of stresses in the descending lithospherefrom a global survey of focal-mechanism solutions of mantle earthquakes. Rev.Geophys. Space Phys. 9, 103–174.

Kennett, B., Engdahl, E., 1991. Traveltimes for global earthquake location and phaseidentification. Geophys. J. Int. 105, 429–465.

Komabayashi, T., Omori, S., Maruyama, S., 2004. Petrogenetic grid in the systemMgO–SiO2–H2O up to 30 GPa, 1600 degree C: applications to hydrous peridotitesubducting into the Earth’s deep interior. J. Geophys. Res 109, B03206.

Lei, J., Zhao, D., 2005. P-wave tomography and origin of the Changbai intraplatevolcano in Northeast Asia. Tectonophysics 397, 281–295.

Li, X., Yuan, X., 2003. Receiver functions in northeast China—implications for slabpenetration into the lower mantle in Northwest Pacific subduction zone. EarthPlanet. Sci. Lett. 216, 679–691.

Liu, J., 1999. Volcanoes in China. Science Press, Beijing, p. 219.Liu, K., Gao, S., Gao, Y., Wu, J., 2008. Shear wave splitting and mantle flow associ-

ated with the deflected Pacific slab beneath northeast Asia. J. Geophys. Res. 113,B01305.

Liu, R., 2000. Active Volcanoes in China. Seismological Press, Beijing, p. 114.Maruyama, S., 1994. Plume tectonics. J. Geol. Soc. Jpn. 100, 24–49.Mitrovica, J., Mound, J., Pysklywec, R., Milne, G., 2000. Sea-level change on a dynamic

Earth. In: Boschi, E., Esktrom, G., Morelli, A. (Eds.), Problems in Geophysics Forthe New Millennium. Editrice Compositori, pp. 499–529.

Ohtani, E., Litasov, K., Hosoya, T., Kubo, T., Kondo, T., 2004. Water transport into thedeep mantle and formation of a hydrous transition zone. Phys. Earth Planet. Inter.143, 255–269.

Morgan, W., 1971. Convective plumes in the lower mantle. Nature 230, 42–43.Ritter, J., Jordan, M., Christensen, U., 2001. A mantle plume below the Eifel volcanic

fields, Germany. Earth Planet. Sci. Lett. 186, 7–14.Shieh, S., Mao, H., Hemley, R., Ming, L., 1998. Decomposition of phase D in the lower

mantle and the fate of dense hydrous silicates in subducting slabs. Earth Planet.Sci. Lett. 159, 13–23.

Simkin, T., Siebert, L., 1994. Volcanoes of the World. Geoscience Press, WashingtonDC, p. 349.

Stein, S., Freymueller, J., 2002. Plate Boundary Zones. AGU, Washington DC, p. 425.Stern, R., 2002. Subduction zones. Rev. Geophy. 40 (4), RG000108.Tang, J., Deng, Q., Zhao, G., 2001. Electric conductivity and magma chamber at the

Tianchi volcano area in Changbaishan. Seismo. Geol. 23, 191–200.Tatsumi, Y., Maruyama, S., Nohda, S., 1990. Mechanism of backarc opening in the

Japan Sea: role of asthenospheric injection. Tectonophysics 181, 299–306.Turcotte, D., Schubert, G., 1982. Geodynamics. John Wiley and Sons Press, p. 450.Waite, G., Smith, R., Allen, R., 2006. Vp and Vs structure of the Yellowstone hot spot

from teleseismic tomography: evidence for an upper mantle plume. J. Geophys.Res. 111, B04303.

Wessel, P., Smith, W., 1998. New, improved version of the Generic Mapping Toolsreleased. Eos. Trans. AGU 79, 579.

Wilson, J., 1963. A possible origin of the Hawaiian Islands. Can. J. Phys. 41, 863–870.Wu, F., Hetland, E., 1999. The Changbaishan, China, PASSCAL experiment. In: 21st

Seismic Research Symposium: Technologies for Monitoring the ComprehensiveNuclear-Test-Ban Treaty, Las Vegas, NV, USA.

Zhang, C., Zhang, X., Zhao, J., Liu, B., 2002. Crust-mantle structure of the ChangbaishanTianchi volcanic area and its vicinity: an exploratory study and its implications.Chin. J. Geophys. 45, 862–871.

Zhao, D., 2004. Global tomographic images of mantle plumes and subducting slabs:insight into deep Earth dynamics. Phys. Earth Planet. Inter. 146, 3–34.

Zhao, D., 2007. Seismic images under 60 hotspots: search for mantle plumes. Gond-wana Res. 12, 335–355.

Zhao, D., Hasegawa, A., Horiuch, S., 1992. Tomographic imaging of P and S wavevelocity structure beneath northeastern Japan. J. Geophys. Res. 97, 19909–19928.

Zhao, D., Hasegawa, A., Kanamori, H., 1994. Deep structure of Japan subductionzones as derived from local, regional, and teleseismic events. J. Geophys. Res.99, 22313–22329.

Zhao, D., Xu, Y., Wiens, D., Dorman, L., Hildebrand, J., Webb, S., 1997. Depth extentof the Lau back-arc spreading center and its relation to subduction processes.Science 278, 254–257.

Zhao, D., Lei, J., Tang, R., 2004. Origin of the Changbai intraplate volcanism in North-

east China: evidence from seismic tomography. Chin. Sci. Bull. 49, 1401–1408.

Zhao, D., Lei, J., Inoue, T., Yamada, A., Gao, S., 2006. Deep structure and origin of theBaikal rife zone. Earth Planet. Sci. Lett. 243, 681–691.

Zou, H., Fan, Q., Yao, Y., 2008. U-Th systematics of dispersed young volcanoes in NEChina: asthenosphere upwelling caused by piling up and upward thickening ofstagnant Pacific slab. Chem. Geol. 255, 134–142.