Crustal structure in an onshore–offshore transitional zone near Hong Kong, northern South China Sea

Crustal structure in an onshore–offshore transitional zone near Hong Kong, northern South China Sea

Journal of Asian Earth Sciences 37 (2010) 460–472 Contents lists available at ScienceDirect Journal of Asian Earth Sciences journal homepage: www.el...

4MB Sizes 3 Downloads 50 Views

Journal of Asian Earth Sciences 37 (2010) 460–472

Contents lists available at ScienceDirect

Journal of Asian Earth Sciences journal homepage: www.elsevier.com/locate/jseaes

Crustal structure in an onshore–offshore transitional zone near Hong Kong, northern South China Sea Shaohong Xia a,b,1, Minghui Zhao a, Xuelin Qiu a,*, Huilong Xu a, Xiaobin Shi a a b

CAS Key Laboratory of Marginal Sea Geology, South China Sea Institute of Oceanology, Chinese Academy of Sciences, Guangzhou 510301, China Graduate University of the Chinese Academy of Sciences, Beijing 100039, China

a r t i c l e

i n f o

Article history: Received 22 March 2009 Received in revised form 8 September 2009 Accepted 20 November 2009

Keywords: Crustal structure Littoral Fault Zone Onshore–offshore transitional zone Onshore–offshore seismic experiment The South China Sea Continental South China

a b s t r a c t To study the crustal structure beneath the onshore–offshore transitional zone, a wide-angle onshore–offshore seismic experiment was carried out in northern South China Sea near Hong Kong, using large volume airgun sources at sea and seismic stations on land. The crustal velocity model constructed from traveltime fitting shows that the sedimentary thickness abruptly increases seaward of the Dangan Islands based on the characteristics of Pg and Multiple Pg, and the crustal structure beneath the sedimentary layer is relatively simple. The Moho depth is about 25–28 km along the profile and the P-wave velocity increases gradually with depth. The velocities in the upper crust range from 5.5 to 6.4 km/s, while that in the lower crust is 6.4–6.9 km/s. It also reveals a low velocity zone with a width of more than 10 km crossing the crust at about 75–90 km distance, which suggests that the Littoral Fault Zone (LFZ) exists beneath the onshore–offshore transitional zone. The magnetism anomalies, bouguer gravity anomalies and active seismic zone along the coastline imply the LFZ is a main tectonic fault in the onshore–offshore area. Combined with two previously published profiles in the continental South China (L–G profile) and in the northern margin of South China Sea (OBS1993) respectively, we constructed a land-sea super cross-section about 1000 km long. The results show the onshore–offshore transitional zone is a border separating the unstretched and the stretched continental crust. The low velocity layer (LVL) in the middle crust was imaged along L–G profile. However, the high velocity layer (HVL) in the lower crust was detected along OBS1993. By analyzing the mechanisms of the LVL in the middle crust and HVL in the base of crust, we believe the crustal structures had distinctly different attributes in the continental South China and in the northern SCS, which indicates that the LFZ could be the boundary fault between them. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction Rifted margins are created by extension and breakup of continental crust prior to the formation of ocean basins. They are often classified as volcanic or nonvolcanic based on the amount of extrusion and intrusive magmatic activity during rifting (Louden and Lau, 2001). The mechanical behavior of the lithosphere under extension is well studied on nonvolcanic margins where the extensional fabric has not been modified by large volumes of synrift or postrift volcanism (Funck et al., 2003). Yan et al. (2001) proposed a nonvolcanic nature of the northern margin of South China Sea (SCS) according to the geological and geophysical evidences. However, the mechanism responsible for the rifting of the China continental margin and the subsequent opening of the SCS is still widely discussed due to the complicated interaction among the Pacific Plate in the east, the Eurasian Plate in the north and the Indian* Corresponding author. Tel.: +86 20 89023157. E-mail addresses: [email protected] (S. Xia), [email protected] (X. Qiu). 1 Tel.: +86 20 89023192. 1367-9120/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jseaes.2009.11.004

Australian Plate in the west and south (Qiu et al., 2001). Some contradicting hypotheses have been proposed (Taylor and Hayes, 1983; Ru and Pigott, 1986; Tapponnier et al., 1986; Briais et al., 1993; Nissen et al., 1995; Lee and Lawer, 1995; Zhou et al., 1995). To understand the rifting process, several geological and geophysical investigations have been carried out across the northern margin of SCS and the continental South China in the last two decades. In order to exploring oil and gas-bearing sediments and gas hydrates, a large number of multi-channel seismic, magnetic, drilling and geochemical surveys have been completed across the Pearl River Mouth Basin, Qiongdongnan Basin, Yinggehai Basin and Southwest Taiwan Basin in the northern margin of SCS (Chen et al., 1987; Wu, 1994). Investigations on the crustal structure in the northern margin of SCS also have been carried out since 1980s. During the Sino-US joint study in 1985, three transects of expanding spread profiles were measured (ESP-W, ESP-C, ESP-E in Fig. 1), and the crustal structures were imaged along the northern margin of SCS (Yao et al., 1994; Nissen et al., 1995; Hayes et al., 1995). Yan et al. (2001) obtained a velocity profile based on the

S. Xia et al. / Journal of Asian Earth Sciences 37 (2010) 460–472

461

Fig. 1. Bathymetry and deep seismic profiles in the continental margin of the northern SCS. The blue circles denote the earthquakes with ML more than 3.5 since 1970. The red triangles indicate the historical earthquakes with magnitude higher than 5.5. Deep seismic profiles in the northern margin of SCS are labeled by OBH, OBS, ESP and OOS. The black solid line called L–G denotes the Lianxian–Boluo–Gankou exploration profile in the South China. F–S denotes the Fuzhou–Quanzhou–Shantou profile. B–J denotes the Baiyan–Jianghong profile. The black broken line indicates the Littoral Fault Zone. The red solid lines denote the two onshore–offshore seismic profiles (OOS2004) analyzed in this paper. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

ocean-bottom seismometers deployed by the joint Sino-Japanese team in 1993 (OBS1993 in Fig. 1), which indicated that the high velocity layer (HVL) in the base of the thinned continental crust was related to magmatism occurred after the cessation of seafloor spreading of SCS. The velocity structure along OBS1995 profile (Fig. 1) across the Hengchun Peninsula off Southern Taiwan shows that 11-km thick, transitional crust of the Eurasian Plate subducts beneath a large, rapidly growing accretionary prism (Nakamura et al., 1998; McIntosh et al., 2005). Qiu et al. (2001) carefully studied the crustal structure beneath the Xisha Trough by using oceanbottom hydrophones in 1996 (OBH-IV in Fig. 1). Their results show a similar velocity structure of the continental crust on both sides of the Xisha Trough and crustal thinning centered at the Xisha Trough. The lack of high velocity body in the lower crust implies no magmatic underplating beneath the Xisha Trough. Imaging along a seismic profile in the northeastern margin of SCS, using ocean-bottom seismometers in 2001 (OBS2001 in Fig. 1) shows a HVL with a thickness of 2–3 km in the base of crust. Wang et al. (2006) compared with the differences of HVL distribution and thickness along profiles OBS2001, OBS1993 and OBH-IV and suggested that the magmatic underplating in the northeastern margin of SCS was less active than that in the middle northern margin of SCS, and no magmatic underplating occurred in the northwestern margin of SCS. On land, Liao et al. (1988) obtained a velocity model of the crust and upper mantle, based on the exploration seismic profiles across the Fuzhou–Quanzhou–Shantou region (F–S in Fig. 1), and found a horizontal Low Velocity Layer (LVL) in the middle crust closely relating with the hot spring. The profile across the Lianxian–Boluo–Gankou was acquired in 1988–1989 (L–G in Fig. 1), and a LVL in the middle crust was also identified according to the travel time fitting (Yin et al., 1999). Recently, Jia et al. (2006) imaged the crustal structure along Baiyan–Jiangdong (B–J in Fig. 1) and found that there was a LVL at about 20–23 km depths and no obvious crust-mantle transitional zone existed.

However, all previous studies on the crustal structure are mostly located either onshore or at sea only, so that an imaging gap exists in the onshore–offshore transitional zone in shallow water (Fig. 1), where plate boundaries may exist (Okaya et al., 2002; Qiu et al., 2004). According to the previous geophysical data and large historical earthquakes (Liu, 1981; Wei et al., 2001; Xu et al., 2007), the Littoral Fault Zone (LFZ) exists in the onshore–offshore transitional zone, which might separate the oceanic lithosphere of SCS from the continental lithosphere of South China. Zhao et al. (2004) detected a low velocity zone in the northeastern onshore–offshore transitional zone based on the wide-angle seismic data (OOS2001 in Fig. 1), which constrained the position of the LFZ. To better understand the crustal structure of the onshore–offshore transitional zone in the northern SCS, we carried out a wide-angle onshore–offshore seismic experiment in the northern margin of SCS near Hong Kong in 2004 (OOS2004 in Fig. 1). The main objective of this work is to obtain a 2-D velocity model using these data. First, the velocity model was determined using a joint refraction/reflection travel time inversion method (Zelt and Smith, 1992). Second, the resolution and robustness of velocity model were estimated by performing the checkerboard tests and calculating the ray densities. Finally, combined with the results of OBS1993 and L–G profile, a velocity structure was constructed crossing from land to sea, and the geodynamic implications are also discussed.

2. Tectonic setting The SCS is one of the largest marginal seas in the western Pacific. It is located at the junction of the Eurasian, Indian-Australian and Pacific Plates. The SCS mainly consists of three parts: the northern continental margin, the oceanic basin and the southern continental margin, and they are aligned in a NE-SW trend. It is

462

S. Xia et al. / Journal of Asian Earth Sciences 37 (2010) 460–472

generally agreed that the SCS originated from passive rifting, pullapart and breakup of the South China continent (Taylor and Hayes, 1980, 1983; Tapponnier et al., 1986). According to magnetic lineation analysis, the South China continent started to stretch in about S–N direction from Late Cretaceous to Middle Miocene (Nissen et al., 1995; Zhou et al., 1995). Because of the continental breakup and the seafloor spreading, the pull-apart oceanic basin of the SCS was formed during the Late Oligocene–Middle Miocene. Since Middle Miocene, the SCS was in compression due to the eastward subduction of the SCS beneath the Manila Trench and the collision between the Nansha (Spratly) Islands and Liyue Bank with Kalimantan (Taylor and Hayes, 1980, 1983; Wang et al., 2006). The northern margin of SCS is an extensional continental margin. The amount of northern continental crustal extension is much less along the east and central segments of the margin than along the western segment, which results from the east–west variations in the rheology of the pre-rift crust and associated east–west variations in the thermal structure of the pre-rift lithosphere (Hayes and Nissen, 2005). Intrusion of magma and igneous rocks existed in the northern margin of SCS (Yan et al., 2001; Wang et al., 2006). Structurally, the northern margin is characterized by normal faults, which trend mainly in NNE, NW, and EW direction (Pigott and Ru, 1994; Liu et al., 1997; Ludmann and Wong, 1999). The LFZ exists in the onshore–offshore transitional zone in shallow water, along which many earthquakes occurred and the anomalous characteristics of the gravity and geomagnetism are very obvious (Hao et al., 2002). Its strike is NEE–SWW direction and the dip is SE direction. A large number of cracked metamorphic granites could be observed in the Dangan Islands and the Nanpeng Islands, which may be caused by the LFZ. Most of the faults mainly trend NE to NNE in the coastal land areas of South China, although some trend ENE and NW.

3. Onshore–offshore wide-angle seismic experiment 3.1. Data acquisition and processing In July 2004, two onshore–offshore wide-angle seismic profiles were shot across the onshore–offshore transitional zone offshore Hong Kong (Fig. 2). Both were NNW–SSE in the direction and more than 50 km in length. The survey was carried out by SCSIO’s R/V Shiyan 2. The seismic source consisted of 4 new Bolt model 1500LL airguns. The chamber volume of the individual airgun was 1500 in.3 and the total volume of the airgun array was 6000 in.3. The working pressure was 2000 psi. The 26 receiving stations comprised three parts: a portable seismic station on the Dangan Islands (DGN), the Hong Kong short-period Seismographic Network (nine stations), and the Guangdong Seismographic Network (16 stations) (Fig. 2a). The seismometer at DGN is a 3-component short-period instrument with 100 sps sampling rate. The Hong Kong short-period Seismographic Network consists of THK, HKSL, HKLK, HKLM, YHK, HKCD, CCHK, HKLI and HKCS. These seismometers are 3-component instruments with 100 sps sampling rate and piezoelectric sensors having a dynamic range of about 100 dB and a frequency response of 1–30 Hz. The Guangdong Seismographic Network includes six broadband stations (GZH, SHZ, SHT, ZHJ, HUD and SHG) and 10 short-period stations (LIN, XNY, TIS, LCH, SHW, MEZ, ZHQ, YGJ, ZHH and XFJ). These 16 stations are 3-component seismometers with 50 sps sampling rate. Fig. 2b is the enlargement of the profile 1 and the near stations along the profile extension. For navigation and shot timing, the Global Positioning System (GPS) was used. The ship speed was between 3.7 knots and 4.3 knots and the source array was fired at a time interval of 110 s, resulting in shot distance intervals ranged from 210 m to 240 m.

11 4 . E

11 0 E 25N

111 E

11 2 E

11 4 E

11 3 E

11 5 E

11 6 E

11 4 . 5 E

11 7 E SHZ

22.5N

SHG

LIN

MEZ

HKLM

LCH 24N

HKCS

GZH XFJ ZHQ

SHT

HKCD

HUD

23N

SHW DGN

SHZ XNY

TIS

HongKong seismic network

ZHH

YGJ

Da

nga

n

n Isla

ds

DGN

e

le 1

lt Z o n

Profi

al Fau

le 1

L it t o r

Pr of il e 2

ZHJ

Profi

22N

22.0N

(a)

21N

21.5N

(b)

Fig. 2. (a) Locations of the two profiles of the onshore–offshore seismic experiment and the 26 seismic stations. Solid bold lines show the shot lines. Circles indicate the locations of the Hongkong seismic network (Red circles mean detectable signals and black circles mean no detectable signals). Triangles denote the stations of the Guangdong seismic network (Red triangles mean detectable signals and black triangles mean no detectable signals). Red inverted triangle denotes the location of the portable seismic station on the Dangan Islands. (b) The enlargement of the profile 1 and the near stations along the profile extension. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

463

S. Xia et al. / Journal of Asian Earth Sciences 37 (2010) 460–472

distance (km) 60

70

110

100

90

80

t-x/6 (s)

6 4 2 0 10

20

30

40

(a)

50

offset km distance (km) 0

40

20

Dangan

0

80

60

100

depth (k)

10 20 30

(b)

t x 6 s

40

6 4

PmP

2

Multiple

Pg

Pg 0

50

60

(c) 90

80

70

100

110

distance (km) Fig. 3. (a) Record section for the portable Dangan station. (b) Rays path diagram. (c) Observed (color vertical bars) and calculated (thin solid lines) travel times of identified seismic phases. Horizontal scale in the record section is shot-receiver distance (offset), and the vertical scale is the travel time using a reduction velocity of 6.0 km/s. Horizontal axes in the ray path diagram and the travel time window are the distance relative to the SHZ station. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

A total of 657 shots were fired along these two lines. Water depths along the two lines were obtained by converting measurements from the onboard echo sounder. The range of the water depth in this region is about 40–70 km. During the field operation, we encountered good weather conditions. After a series of processing (Xia et al., 2007), the continuously recorded data were converted into standard SEG-Y format and stored in common receiver gathers. Amplitude spectra shown the main seismic energy was in the frequency range from 3 to 15 Hz. Thus, the record sections were band-pass filtered from 3 to 15 Hz to enhance the S/N ratio of some weak seismic phases. Finally, single receiver record sections were plotted with a reduction velocity of 6.0 km/s (Figs. 3–5). On these record sections, deep seismic phases like Pg, Multiple Pg and PmP could be recognized. However, the dataset seems to have a bit strong ringing as shown in Figs. 3 and 4, which could be related with water-bottom multiples at the range of 5–70 km offset. The objective of the present work is mainly to obtain the cross-section velocity structure along the onshore–offshore transitional zone. There are only two stations with detectable airgun signals along the profile 2. Therefore we only

modeled the two-dimensional crustal structure based on the data of five seismic stations along the profile 1 extension (Fig. 2). 3.2. Methodology To apply a two-dimensional modeling approach, we took the straight extension of the shot line as the baseline. The Dangan and HKCD stations are located on the baseline. The other receiver positions were projected onto the baseline while maintaining the true source-receiver offsets for subsequent modeling. The deviations of receiver sites from the baseline are about 1 km for SHZ station, 2 km for HKLM station, and 5 km for HKCS, which is a relatively small amount for crustal-scale experiments and could be neglected. The travel time forward and inverse program RAYINVR of Zelt and Smith (1992) is used to construct the crustal velocity model. This method is particularly useful for crustal seismic data with a relatively low shot-receiver density with respect to the requirements of conventional tomographic or full wavefield techniques, and it determines the optimum number and location of velocity

464

S. Xia et al. / Journal of Asian Earth Sciences 37 (2010) 460–472

distance (km) 60

70

110

100

90

80

8 6 4 2 0 50

70

60

80

90

100

(a) distance (km) 0

distance (km)

0

HKLM

80

60

40

20

100

10 20 30

(b) 40 8 6 4 Multiple

2

Pg PmP

Pg

(c)

0 50

60

90

80

70

100

110

distance (km) Fig. 4. The same as Fig. 3 but for the HKLM station. This station belongs to the Hong Kong seismic network. The signals of other stations in the seismic network is similar with the HKLM.

and boundary nodes by providing the desired trade-off between traveltime fit and parameter resolution, as well as the ability to trace rays to all observations (Zelt and Smith, 1992). First, the observed travel times were roughly fitted by forward modeling to obtain a starting model which was developed from top to bottom. In a second step, the forward model was refined to obtain a match between the observed and calculated data by using inversion algorithm in RAYINVR. 3.3. Seismic data Most of the wide-angle data along the profile are generally of good quality. Some key record sections are shown in Figs. 3–5 together with the calculated travel times through the final velocity model to allow the reader to access how well the model fits the data. These figures stand for the portable seismic station, the Hong Kong seismic network and the Guangdong seismic network respectively. Wide-angle data are modeled in terms of phases that are identifiable and laterally coherent over many traces. Names of individual seismic phases are based on the later interpretation of the velocity model. Refraction within the crust is labeled Pg, while the refraction with the multiple reflection from the base of sediment is named Multiple Pg. PmP is the Moho reflection. For the

remainder of this paper, ‘‘travel time” will refer to reduced time, ‘‘offset” will refer to the distance between shot and receiver, and ‘‘distance” will refer to a location along the profile 1 relative to SHZ at 0 km. The clear intra crustal refraction phase, Pg, is observed at offsets less than 80 km on almost all record sections. The pattern of the Pg arrivals varies significantly from land to sea. The Pg travel time is abruptly delayed near 12 km to the south of Dangan Islands, and the varied characteristics are similar on Dangan and HKLM record sections (Figs. 3 and 4). The apparent velocity of Pg is about 6.0 km/ s near the Dangan Islands, and decreases to 5.5–5.7 km/s seaward. This may indicate the lateral heterogeneity of the sedimentary cover in the transitional zone. The delay of Pg along the profile clearly demonstrates the abrupt change of the crystalline basement, which suggests the existence of faults. Part of this variation might be due to the presence of the sediments in the sedimentary basin (Pearl River Mouth Basin). Prominent phases after the first Pg arrivals were observed in the record sections of Dangan and HKLM (Figs. 3 and 4). These phases are interpreted as Multiple Pg, which reflect within the sediments and then refract in the crust. Fig. 6a shows a schematic illustration of ray path of Multiple Pg. These phases have the following features. (1) They are extremely similar to the Pg (Figs. 3 and 4). (2)

465

S. Xia et al. / Journal of Asian Earth Sciences 37 (2010) 460–472

distance (km) 60

70

80

60

70

80

90

100

110

90

100

110

8 6 4 2 0

(a)

offset (km) distance (km) 0SHZ

distance (km)

0

80

60

40

20

100

10 20 30

(b) 40 8 6 4

PmP

2

(b) 0 60

50

110

100

90

80

70

distance (km) Fig. 5. The same as Fig. 3 but for the SHZ station. This station belongs to the Guangdong seismic network.

distance (km) 0

distance (km)

0

20 

40

60

80

100



10 20 30

Moho 28 km

(a)

40

shot number 1.5 1.0 0.5 0.0 - 0.5

(b)

Fig. 6. (a) The ray path diagrams of Pg (broken line) and Multiple Pg (solid line) in the forward model. (b) First-arrival Pg picks from off-line stations of profile 1 with corresponding shot number.

466

S. Xia et al. / Journal of Asian Earth Sciences 37 (2010) 460–472

The time delay relative to Pg is about 1–1.5 s. (3) Amplitudes of these phases are large and sometimes larger than those of Pg phases, which imply that the phases are generated at a very sharp velocity boundary. (4) Apparent velocities of Pg and Multiple Pg phases are almost consistent. In addition, the amplitudes of the Multiple Pg phases are a little weaker on Dangan record section than that on HKLM record section, which may result from the almost same arrival time between the PmP and the Multiple Pg on HKLM record section. The time delay of Multiple Pg relative to Pg could be used to control the thickness of the sediments. The reflection phase from the crust-mantle boundary, PmP, is observed on most of the instruments with variable quality. The amplitude and energy of PmP are weak at the 40–50 km offset on the record section of Dangan station (Fig. 3), but increase rapidly at more than 70 km offset on the record sections of SHZ station (Fig. 5), which suggests the reflected energy at the pre-critical offset is relatively weak, but increases rapidly near the critical point (Cerveny, 1966; Zeng, 1984; Qiu et al., 2001). PmP phases are much

more energetic and easier to identify beneath Dangan islands and landward (Figs. 3–5), indicating the presence of a better developed crust-mantle boundary. Due to the interference of the Multiple Pg, the PmP phase is a little difficult to be picked on HKLM record sections. The absence of Pn and clear PmP may imply the complicated crust-mantle boundary with an abrupt velocity discontinuity (Qiu et al., 2001).

4. Modeling and results 4.1. Velocity model The first-arrival Pg phases from other off-line stations showed similar shot number-time trends without strong variations correlated with receiver positions (Fig. 6b), which implied the variations of Pg were mainly caused by the variations of the sedimentary thickness beneath the shot lines. Therefore, we constrained the

Input model distance (km)

depth (km)

(a)

Output model

depth (km)

(b)

-0.2

-0.1

0.0

0.1

0.2 km/s

distance (km)

depth (km)

(c) 0.78

0.74

0.75

0.88

0.84

0.23

0.49

0.45

0.63

0.53

0.23

0.29 0.92

0.94

0.90 0.11

0.95 0.12

0.53

0.16

0.89

0.85

0.74

0.69

0.59

0.41

0.41 0.98

0.99

0.90

0.13

Fig. 7. (a) Synthetic velocity perturbation for Checkerboard test. (b) Calculated velocity perturbation for Checkerboard test. (c) Final model parameters with the value of resolution. Circles denote the velocity nodes, while squares are the Moho boundary nodes.

467

S. Xia et al. / Journal of Asian Earth Sciences 37 (2010) 460–472

thickness of the sediments from the intercept times of the first observed refractions Pg by assuming a constant velocity of 3.0 km/s within the sediments (Qiu et al., 2001; Yan et al., 2001; Zhang and Wang, 2007; Li et al., 2008). Then the Multiple Pg arrivals were used to optimize the thickness of sediments by a simple velocity model (Fig. 6a). Because the ray-paths of Pg and Multiple Pg are almost the same beneath the basement as shown in Fig. 6a, and the time delays between Pg and Multiple Pg could better constrain the variations of sediments. According to the dataset as shown in Figs. 3 and 4, we can know that the average time-lag of the Pg multiples is about 1.5–1.8 s and the average thickness of the sedimentary layer is about 1.5–2.0 km.

In order to better parameterize the model and refine the velocity structure in the inversion, we performed a checkerboard test (Zhao et al., 1992) to constrain the resolution of the model (Fig. 7). A synthetic data set was calculated for the synthetic velocity model with the same source-receiver geometry as in the real data. The synthetic model was obtained by adding the velocity anomalies shown in Fig. 7a. The size of each anomalous block is about 10  8 km, and the maximum amplitude is 0.2 km/s. Random noise with 0.1 s has been added to the synthetic data. The result of the inversion is shown in Fig. 7b. The pattern of the velocity anomalies is well recovered within the upper crust, while it is a little poor in the lower crust. Therefore, we set some velocity nodes

distance (km) 0

60

40

20

80

100

distance (km)

0 10 20 30

( a)

40 8 6 4 2

( b)

0 50

60

70

110

100

90

80

distance (km) 0

20

60

40

0

80

Litto

5. 8 6. 0

10

aul ral F

6. 2 6. 4

ne t Zo

depth (km)

100

5. 6

6. 6

20

6. 8

30

( c)

0

20

40

60

7.0

6.0

5.0

80

100

aul ral F

10

Litto ne t Zo

depth (km)

0

20

(( dd))

30

- 3%

0

3%

Fig. 8. (a) Ray paths through the final model for all seismic phases. (b) Comparison of observed and calculated travel times of all phases through the final model. (c) Velocity model in the onshore–offshore transitional zone. Two dashed lines indicate the location of the Littoral Fault Zone. (d) The velocity perturbation.

468

S. Xia et al. / Journal of Asian Earth Sciences 37 (2010) 460–472

with a grid spacing of about 10 km in the upper crust while 20 km in the lower crust for dense ray areas, and some Moho boundary nodes with a spacing of 10–20 km in the forward model (Fig. 7c) to better refine the velocity model in the inversion. The final P-wave velocity model for the profile 1 is shown in Fig. 8, together with the ray coverage and comparison of observed and calculated arrivals. Two main interfaces, the basement interface and the Moho interface, divide the model section into three parts of the sedimentary layer, the crust and the uppermost mantle. No major sediment cover is detected beneath the Dangan Islands and landward according to the characteristics of Pg and Multiple Pg phases. However, a significant change of sedimentary layer exists at about 70 km distance, and thickness of the sediments increases abruptly to about 1–2 km by assuming the average velocity of 3.0 km/s within the sediment. Good correlation exists between the final velocity model and the geological topography. The cracked and weathered granites mainly distribute on the surface (Vp = 5.5 km/s, 0–70 km distance along the profile) in the region with absence of sediment cover, while the present sedimentary layer is considered to be associated with the sedimentary basin. The large-scale fault may be closely related to the abruptly lateral variation of the sedimentary layer. Two additional boundaries were initially used during the modeling to constrain the crustal structure without first-order velocity discontinuity (Fig. 8a). However, the results of the travel time modeling suggest that these are no clear reflections from inside the crust, and these two boundaries were turned into isovelocity lines. The P-wave velocity increases gently from 5.5 to 5.6 km/s on the top to 6.8–6.9 km/s on the bottom of the crust. Taking the 6.4 km/s isoline as the boundary between the upper and lower crust, the upper crust has a thickness of about 13–14 km with the velocities from 5.5 to 6.4 km/s, while the velocities are from 6.4 to 6.9 km/s in the lower crust with the thickness of about 11–13 km landward of 70 km distance. Seaward of 70 km distance, the upper crust has a thickness of about 11–12 km with the velocities from 5.5 to 6.4 km/s, while the velocities of lower crust change from 6.4 to 6.8, and the thickness is about 10–12 km. The velocity isolines show a weak sign of ‘‘U” character at 70–90 km distance (Fig. 8c), indicating a low velocity zone that may correspond to the LFZ. The Moho depth gradually becomes shallow seaward, although with some variations along the profile. The velocity perturbation was imaged to better show the crustal lateral heterogeneities (Fig. 8d). A clear low velocity zone crossing the crust locates at about 80 km distance. The velocity perturbation of the low velocity zone is about 2%. Its width is about 20 km in shallow depth and narrows to about 10 km in deep crust. 4.2. Model resolution and uncertainty Travel time residuals, number of observations, and normalized

v2 for individual phases are summarized in Table 1. The estimated pick uncertainty changed from 50 ms to 150 ms depending on the quality of each individual travel time pick. The total RMS misfit is 106 ms for the final model. The normalized v2 is 1.439, which is close to the optimum value of 1.

Table 1 Number of observations, RMS misfit between calculated and picked travel times, and Normalized v2 for individual seismic phases. Phase

Number of observations

RMS Misfit, ms

Normalized v2

Pg Multiple Pg PmP Total

787 341 466 1594

90 87 90 106

1.308 0.753 2.252 1.439

One advantage of the applied inversion method is that it could provide a measure of resolution and uncertainties of the estimated model parameters according to a least-squares norm. Resolution values generally range from between 0 and 1, which depend on the relative number of rays sampling each model parameter. The inversion parameters are 1.0 for the overall damping factor, 1.0 km/s for a priori uncertainty in velocity parameter and 1.0 km for a priori uncertainty in depth parameter. Fig. 7c shows the obtained values of resolution for velocity and Moho boundary nodes. Resolution values of greater than 0.5 indicate reasonably well resolved model parameters (Lutter and Nowack, 1990). Resolution for velocity nodes is good within the upper crust with values from 0.5 to 0.9, while that is weak in the lower crust. Resolution for Moho boundary nodes is excellent with values of greater than 0.9 in the central part. They are fairly consistent with the ray coverage (Fig. 8a). The dense rays passed through the upper crust due to large amount of Pg and Multiple Pg phases. But only PmP data controlled the velocity structure of lower crust. The parameter uncertainty test (Zelt and Smith, 1992; Zelt, 1999) suggests that velocity uncertainties associated with the upper and lower crust are 0.05 and 0.13 km/s, respectively, and the average depth uncertainty of Moho boundary is of the order of 1.5 km. As the above mentioned results, we think that the interpretation of this wide-angle seismic experiment can only resolve a first-order variation in crustal properties. The main deficiency of the model is embedded in the geometry of the data acquisition. The use of land-stations together with marine shots does not produce any reversed ray-paths which are important to constrain any wide-angle seismic model. However,

Table 2 The RMS values and chi-square v2 from the different laterally homogenous models in the crystalline crust. Velocity (km/s)

Moho (km)

RMS/ms

Normalized v2

5.5–6.9 5.6–6.9 5.7–6.9 5.8–6.9 5.9–6.9 6.0–6.9 5.5–6.9 5.6–6.9 5.7–6.9 5.8–6.9 5.9–6.9 6.0–6.9 5.5–6.9 5.6–6.9 5.7–6.9 5.8–6.9 5.9–6.9 6.0–6.9 5.5–6.8 5.6–6.8 5.7–6.8 5.8–6.8 5.9–6.8 6.0–6.8 5.5–6.8 5.6–6.8 5.7–6.8 5.8–6.8 5.9–6.8 6.0–6.8 5.5–6.8 5.6–6.8 5.7–6.8 5.8–6.8 5.9–6.8 6.0–6.8

30 30 30 30 30 30 28 28 28 28 28 28 26 26 26 26 26 26 30 30 30 30 30 30 28 28 28 28 28 28 26 26 26 26 26 26

539 458 412 413 454 530 336 257 233 283 373 479 173 141 151 266 388 514 599 512 469 450 489 567 399 310 277 303 378 478 227 138 148 245 365 487

56.870 41.126 33.227 32.143 38.532 52.351 23.388 13.312 10.866 15.773 26.831 44.817 6.137 2.764 4.581 14.401 29.864 52.084 72.795 54.727 43.930 40.167 45.766 56.388 33.350 20.915 16.043 18.227 27.779 43.723 11.150 4.199 4.711 12.042 26.104 46.857

469

S. Xia et al. / Journal of Asian Earth Sciences 37 (2010) 460–472

like many other wide-angle experiments, reverse profiles are too expensive to conduct. In order to demonstrate clearly that the resulting velocity model is unique, we maintain the same sediments as that in the resulting model, while use different models with no lateral velocity change in the crystalline crust to trace the data. The calculated RMS values and the chi-square errors (Table 2) are much higher than those from final model. The results show that a laterally homogenous model in the crystalline crust does not allow to sufficiently fit the data.

is consistent with the characteristics of the Pg phases. In the crust, there is a low velocity zone crossing the entire crust at about 76 km distance. Combined with the gravity and magnetism anomaly (Fig. 10), the characteristics of cracked granites in the field works and the distribution of earthquakes in the onshore–offshore transitional zone, we suggest that the low velocity zone is related with the LFZ existed beneath the onshore–offshore transitional zone (Fig. 8d). Fig. 8 shows the sedimentary layer increases abruptly seaward of the low velocity zone, which implies that the LFZ could control the abrupt variations of the sediment. The Moho depth is about 25–28 km beneath the onshore–offshore transitional zone, but it is not sure whether there is abrupt change in Moho thickness in this transect due to less control on the Moho depth south of LFZ. Compared with the results beneath the northeastern onshore–offshore transitional zone of SCS (Zhao et al., 2004), we found that the detected low velocity zone also denoted the LFZ. The variation of

5. Discussion 5.1. Crustal structure beneath the onshore–offshore transitional zone The basement depth and the thickness of the sediments are strongly variable in the onshore–offshore transitional zone, which

0

2

4

t-x/6 (s)

6

8

10

distance (km)

70

80

90

100

110

( a)

60

70

80

90

100

110

(b)

6 0

2

4

t-x/6

8

10

60

distance ( km) 0 0

20

60

40

6. 2

6. 6 6. 8

30

Zone ault ral F

depth (km)

6. 0

6. 4

20

100

Litto

5. 8

10

80

5. 6

((c) c)

Fig. 9. (a) Record section of Dangan station. (b) Record section of HKLM station. (c) Velocity model with the location of Littoral Fault Zone. Red squares indicate the weak amplitudes as shown in (a) and (b). According to the same positions of weak amplitudes and the low velocity zone, it suggests the Littoral Fault Zone has weakened the shot energy and resulted in the weak amplitudes. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

470

S. Xia et al. / Journal of Asian Earth Sciences 37 (2010) 460–472

sedimentary layer was influenced by the LFZ, and Moho depth was consistent in the transitional zone, which further testified the LFZ is the main tectonic fault in the onshore–offshore transitional zone. However, the LVL in the middle crust is absent in present study (Fig. 8c), which is different from the results in the northeastern onshore–offshore transitional zone (Zhao et al., 2004, 2007) and in the continental South China (Yin et al., 1999; Zhang and Wang, 2007). It might be difficult to use the OOS2004 dataset to argue whether there is a LVL in the middle crust offshore Hong Kong due to the very simple layered structure model in this region.

ble penetration of the LFZ into the uppermost mantle (Xu et al., 2007). According to the results of magnetism anomaly and bouguer gravity anomaly (Figs. 10d and e), the LFZ is located the gradient belt of magnetism and gravity anomalies. We collected historical large earthquake data with the magnitudes larger than 5.5 along the transitional zone from 1000 to 1953 AD and the earthquakes with a magnitude larger than 3.5 that occurred along the coastline since 1970 (Fig. 1), and found that a seismic active zone existed in the onshore–offshore transitional zone. Liu (1981) carefully studied the LFZ and deduced its strike was NNE–NE according to historical earthquake epicenters and water depth data. Through detailed study of the Nan’ao earthquakes (M7.5) in 1918, which occurred in the northeast coastal areas of Guangdong Province, Xu et al. (2006) suggested that the earthquake epicenter was located in the intersection part of the LFZ and the Huanggangshui Fault with NW direction strike. In 2006, an earthquake with ML 4.0 occurred over the sea area near the Dangan Islands. Wong et al. (2007) determined its focal mechanism and suggested that its epicenter was located on the ENE–WSW trending Dangan Islands Fault, which could be related to the LFZ. Therefore, we infer that the occurrence of earthquakes along the onshore–offshore transitional zone may result from combined effect between the LFZ and some other faults. However, we could not accurately determine the width of the LFZ due to the facts that this profile is un-reversed and the LFZ is on the limit of resolution in this study. The presence of the LFZ in our model could be tested by the use of Ocean-Bottom Seismometers with dense spacing and shooting on land. We intend to follow this up in further work.

5.2. The Littoral Fault Zone (LFZ) The obtained crustal structure indicated the LFZ developed in the northern onshore–offshore transitional zone with the characteristics of the low velocity zone (Figs. 8c–d). We also observed that the amplitudes of Pg are strongly attenuated on the Dangan and HKLM record sections in the squares as shown in Figs. 9a–b. The locations of weak amplitudes are almost same at the 80– 90 km distance, which is consistent with the position of low velocity zone (Fig. 9c). It suggests that the LFZ has weakened the shot energy and resulted in the weak amplitudes. Based on the onshore–offshore seismic experiment in the northeastern margin of SCS, Zhao et al. (2004) also obtained a low velocity zone crossing the entire crust, and suggested that it is the LFZ. The seismic tomography in the northeastern SCS and adjacent region showed the slow velocities in the uppermost mantle appeared near the LFZ along the coast of southeastern China, which indicated a possi-

L- G p r o f i l e

NW

OOS2004

South China continental crust 0

depth ( km)

OBS1993

o n s h o r e - o ff s h o r e transitional crust

extended continental crust 5.4

6.20

6.0

6.0

5.8 6.0

6.4

6.4

6.5

6. 03 5.80

7.0

20

7.2

Moho

6.80

Moho 100

300

5.5 5.8 6.3 6.8

6.4 6.9

6.8

6.8

Moho

7.4

low e r

s lo p e

(b)

400

100

0

(c) 200

100

0

distance ( km)

distance ( km)

6.8

7.3

upper slope

Littoral Fault Zone

(a) 200

5.2 5.8

5.2 5.8 6.1 6.4

7.4

6.60

30

oceanic crust

volcanism

6.00

10

SE

ocean- continent crust ( OCT)

300

400

distance ( km) A'

A Magnetism anomaly

200

Δ m / nT

Sediments Upper crust Low velocity Lower crust High velocity layer layer

100

L

0

- 100

A G

(d)

OOS

- 200

200 4

Bouguer gravity anomaly

270

1993

Δ g / mGal

OBS

220

170

A'

120 70 20

(e) 0

100

200

300

400

500

distance ( km) Fig. 10. (a–c) Velocity model of super cross-section from land to sea; (d) the magnetism anomaly along AA’ profile; (e) the bouguer gravity anomaly along AA’ profile. The inserted map shows the locations of three deep seismic profiles from land to sea. The magnetism and bouguer gravity anomalies are cited from Hao et al. (2008).

S. Xia et al. / Journal of Asian Earth Sciences 37 (2010) 460–472

5.3. Land-sea super cross-section We constructed a super cross-section from land to sea between the continental South China and the SCS in order to understand the lateral variation of the crustal structure beneath the northern margin of SCS. The cross-section is about 1000 km long and is composed of our profile 1, the L–G profile, and the OBS1993 profile (Fig. 10). Magnetism anomaly and bouguer gravity anomaly along AA’ profile were shown in Fig. 10d and e. The Fig. 10 shows that the thickness of the crust gradually thins from land to sea when combined data from L–G and OBS1993 profiles. The results of the L–G profile showed that the crustal thickness changed from 34 to 29 km over the 400 km NW–SE profile. The velocity of the upper crust was about 5.4–6.2 km/s, and that of the lower crust was about 6.4–6.9 km/s. A LVL in the middle crust was imaged (Yin et al., 1999; Zhang and Wang, 2007). The result of the OBS1993 profile showed the crustal thickness was 26 km in the northernmost part of OBS1993, and gradually decreased to 8 km in deep sea. A HVL with velocities of 7.1–7.4 km/s was detected in the base of crust in the northern SCS (Yan et al., 2001). Over the onshore– offshore transitional zone of about 120 km between L–G profile and OBS1993 profile, the crust thickness thins quickly from 29 to 26 km. If we consider the crust with average 30 km thickness onshore South China as unstretched continental crust, the onshore– offshore transitional zone should be a transit zone between the unstretched South China and the stretched continental margin. The L–G profile shows that the LVL in the middle crust widely develops in the continental South China (Yin et al., 1999), and the distribution of LVL has a close relationship with the hot springs (Liao et al., 1988). The previous studies suggested the LVL in the middle crust might result from the crustal partial melting (Nelson et al., 1996; Yang et al., 2003; Zhao et al., 2007), and the LVL could enhance the decoupling of stronger layers and played an important role in the generation of large earthquake in the crust (Zhou and He, 2002). These previously published results (Liao et al., 1988; Yin et al., 1999; Jia et al., 2006) announced that the LVL in the middle crust was a typical character in the continental South China. However, the LVL in the middle crust is not detected so far in the SCS, according to already published results (Nissen et al., 1995; Qiu et al., 2001; Yan et al., 2001; Wang et al., 2006). Whereas the HVL in the base of crust were imaged in the northern SCS, and gradually appeared and thickened to about 5–7 km in the shelf and the slope, then pinched out in the ocean basin (Yan et al., 2001; Wang et al., 2006). The HVL might originate from magmatic underplating or serpentinized peridotites. The serpentinized peridotite HVL in the base of crust mainly distributed in the western Iberia margin and the Newfoundland margin with the 7.6 km/s velocity in the base of HVL (Funck et al., 2003; Whitmarsh et al., 1993). The tomographic results shown no distinct velocity anomaly within hundreds of kilometers depth under the SCS (Fukao et al., 1994). The observed heat flow values (40–90 mW/m2) were not generally very high except for the rare spots (post-rifting volcanoes), and rifting-related eruptive rocks were also sparsely found in the northern margin of SCS (Yan et al., 2001), and the OBH-IV transect shown the lack of HVL in the northwestern SCS (Qiu et al., 2001). As compared with above investigated results, the HVL in the northern margin of SCS is less likely explained as the serpentinized peridotites or the hot mantle plume, but might be most likely related to partial melting magmatic underplating (Yan et al., 2001). From above analysis, we know that the crustal structure in the continental South China is distinctly different from that in the SCS, which further proves the onshore–offshore transitional zone is the transit zone separating the continental South China from the northern SCS. As mentioned above, the LFZ is the main tectonic fault crossing the crust in the transit zone, corresponds to magne-

471

tism and bouguer gravity anomalies, and has closely relations with active seismic zone along the coastline. These evidences suggest that the LFZ locating in the onshore–offshore transitional zone might be the boundary fault between the continental South China and the northern SCS. Combined with the distribution of LVL and HVL, volcanism, the magnetism and bouguer gravity anomalies, we divided the crustal structure into five parts from land to sea: the continental crust; the onshore–offshore transitional crust; the extended continental crust; the ocean-continent transition (OCT) crust; and the oceanic crust (Fig. 10). The crust below the lower slope is seen as the ocean-continent transition (OCT) with the Moho depth of 12–22 km, the crustal thickness of 8–18 km, the thick HVL and volcanoes (Fig. 10; Yan et al., 2001; Wang et al., 2006). 6. Conclusions The crustal structure beneath the onshore–offshore transitional zone of the northern SCS was imaged, based on the onshore–offshore seismic experiment near Hong Kong. The results suggest that the thickness of the sedimentary layer is strongly variable along the profile. The velocity is 5.5–5.6 km/s at the top to 6.8–6.9 km/ s at the bottom in the crust, the Moho depth is about 25–28 km and the LFZ exists in the transitional zone with the characteristic of low velocity crossing the entire crust. Together with the L–G profile and the OBS1993 profile, a super cross-section of the crustal structure was constructed from land to sea between the continental South China and the northern SCS. The characteristics of crustal structure in the continental South China are obviously different from that in the northern SCS, which might imply that the LFZ is the boundary fault between them. Acknowledgements The field work of this study was assisted by the Engineering Center of SCSIO, the Seismic Monitoring Center of EAGP, and the captain and crew of R/V Shiyan 2. This work was jointly supported by the Chinese Ministry of Science and Technology (2007CB411701), the National Natural Science Foundation of China (40674051, 40776025, U0933006), and the Key Laboratory of Marginal Sea Geology, Chinese Academy of Sciences (MSGL0704). We thank two anonymous reviewers provided thoughtful review comments, which improved the manuscript. References Briais, A., Patriat, P., Tapponnier, P., 1993. Updated interpretation of magnetic anomalies and seafloor spreading stages in the South China Sea: implications for the tertiary tectonics of Southeast Asia. Journal of Geophysical Research 98, 6299–6328. Cerveny, V., 1966. On dynamic properties of reflected and head waves in the n-layered Earth’s crust. Geophysical Journal of the Royal Astronomical Society 11, 139–147. Chen, B., Lei, Z., Zhou, Y., 1987. Major oil accumulation characteristics and exploration direction in the Pear River Mouth Basin. China Oil 17, 23. Fukao, Y., Maruyama, S., Kobayashi, M., Inoue, H., 1994. Geologic implication of the whole mantle P-wave tomography. Journal of Geological Society Japan 100, 4–23. Funck, T., Hopper, J., Larsen, H., Louden, K., Tucholke, B., Holbrook, W., 2003. Crustal structure of the ocean-continent transition at Flemish Cap: Seismic refraction results. Journal of Geophysical Research 108 (B11), 2531. doi:10.1029/ 2003JB002434. Hao, T.Y., Liu, J.H., Song, H.B., Xu, Y., 2002. Geophysical evidences of some important faults in South China and adjacent marginal seas region. Progress of Geophysics 17 (1), 13–23 (in Chinese). Hao, T.Y., Huang, S., Xu, Y., et al., 2008. Comprehensive geophysical research on the deep structure of Northeastern South China Sea. Chinese Journal of Geophysics 51 (6), 1785–1796 (in Chinese). Hayes, D.E., Nissen, S.S., Buhl, P., Diebold, J., Yao, B., Zeng, W., Chen, Y., 1995. Through-going crustal faults along the northern margin of the South China Sea and their role in crustal extension. Journal of Geophysical Research 100 (B11), 22435–22446. Hayes, D., Nissen, S., 2005. The South China Sea margins: implications for rifting contrasts. Earth and Planetary Science Letters 237, 601–616.

472

S. Xia et al. / Journal of Asian Earth Sciences 37 (2010) 460–472

Jia, S., Xiong, L., Xu, Z., et al., 2006. Crustal structure features of the Leiqiong depression in Hainan province. Chinese Journal of Geophysics 49, 1385–1394 (in Chinese). Lee, T.Y., Lawer, L.A., 1995. Cenozoic plate reconstruction of Southeast Asia. Tectonophysics 251, 85–138. Li, H., Morozov, I.B., Smithson, S.B., 2008. 3D seismic analysis of the Coast Shear Zone in SE Alaska and Western British Columbia: broadside analysis of ACCRETE wide-angle data. Tectonophysics 448, 20–32. Liao, Q., Wang, Z., Wang, P., Yu, Z., 1988. Explosion seismic study of the crustal structure in Fuzhou–Quanzhou–Shantou region. Chinese Journal of Geophysics 31, 270–280 (in Chinese). Liu, Y., 1981. Analysis of Regional Faults Along the Coast of South China. Seismological Press, Beijing (in Chinese). Liu, Z., Zhao, Y., Zhang, Y., Zhou, X., He, S., Xie, Y., Jiang, S., Wang, Q., Huang, Z., Lin, J., 1997. A comprehensive study on Zhujiang River Mouth-Liyue Bank-Nansha Trough transect in the South China Sea. Nanhai Studia Marina Sinica 12, 1–24 (in Chinese). Louden, K.E., Lau, H., 2001. Insights from scientific drilling on rifted continental margins. Geoscience Canada 28, 187–195. Ludmann, T., Wong, H.K., 1999. Neotectonic regime on the passive continental margin of the northern South China Sea. Tectonophysics 311, 113–138. Lutter, W.J., Nowack, R.L., 1990. Inversion for crustal structure using reflections from the PASSCAL Ouachita experiment. Journal of Geophysical Research 95, 4633–4646. McIntosh, K.D., Nakamura, Y., Wang, T.K., Shih, R.C., Chen, A.T., Liu, C.S., 2005. Crustal-scale seismic profiles across Taiwan and the western Philippine Sea. Tectonophysics 401, 23–54. Nakamura, Y., McIntosh, K., Chen, A.T., 1998. Preliminary results of a large offset seismic survey west of Hengchun Peninsula, Southern Taiwan. Terrestrial, Atmospheric and Oceanic Sciences 9, 395–408. Nelson, K.D., Zhao, W., Brown, L.D., et al., 1996. Partially molten middle crust beneath southern Tibet: synthesis of project INDEPTH results. Science 274, 1684–1688. Nissen, S.S., Hayes, D.E., Buhl, P., Diebold, J., Yao, B.C., Zeng, W.J., Chen, Y.Q., 1995. Deep-penetrating seismic sounding across the northern margin of the South China Sea. Journal of Geophysical Research 100 (B11), 22407–22433. Okaya, D., Henrys, S., Stem, T., 2002. Double-side onshore–offshore seismic imaging of a plate boundary: ‘‘super-gathers” across South Island, New Zealand. Tectonophysics 255, 247–263. Pigott, J.D., Ru, K., 1994. Basin superposition on the northern margin of the South China Sea. Tectonophysics 235, 27–50. Qiu, X., Ye, S., Wu, S., Shi, X., Zhou, D., Xia, K., Flueh, E., 2001. Crustal structure across the Xisha Trough, northwestern South China Sea. Tectonophysics 341, 179–193. Qiu, X., Zhao, M., Ye, C., et al., 2004. Ocean Bottom Seismometer and onshore– offshore seismic experiment in northeastern South China Sea. Geotectonica et Metallogenia 28 (1), 28–35. Ru, K., Pigott, J.D., 1986. Episodic rifting and subsidence in the South China Sea. AAPG Bulletin 70, 1136–1155. Tapponnier, P., Peltzer, G., Armijo, R., 1986. On the mechanics of the collision between India and Asia. In: Coward, M.P., Ries, A.C. (Eds.), Collision Tectonics, vol. 19. Geological Society London Special Publication, Black-well Scientific, Oxford, pp. 115–157. Taylor, B., Hayes, D., 1980. The tectonic evolution of the South China Sea Basin. In: Hayes, D. (Ed.), Tectonic and Geologic Evolution of Southeast Asian Seas and Islands, Geophysical Monograph Series, vol. 23. AGU, Washington, DC. Taylor, B., Hayes, D., 1983. Origin and history of the South China Sea Basin. In: Hayes, D. (Ed.), Tectonic and Geologic Evolution of Southeast Asian Seas and Islands, Geophysical Monograph Serial, vol. 27. AGU, Washington, DC.

Wang, T.K., Chen, M., Lee, C., Xia, K., 2006. Seismic imaging of the transitional crust across the northeastern margin of the South China Sea. Tectonophysics 412, 237–254. Wei, B., Feng, X., Cheng, D., et al., 2001. Characteristics of the seismic activity in the southeast coastal region. Seismological Press, Beijing (in Chinese). Whitmarch, R.B., Pinheiro, L.M., Miles, P.R., et al., 1993. Thin crust at the Western Iberia ocean-continent transition and ophiolites. Tectonics 12, 1230–1239. Wong, W.T., Chan, Y.W., Kang, Y., Yang, X., 2007. Source parameters of an ML 4.0 earthquake near Dangan Islands on 14.9.2006. Asian-Pacific Network of Centers for Earthquake Engineering Research 2007 Meeting, Hong Kong, China, 29–30 May 2007. Wu, J., 1994. Evaluation and models of Cenozoic sedimentation in the South China Sea. Tectonophysics 235, 77–98. Xia, S., Qiu, X., Zhao, M., Ye, C., Chan, Y., Xu, H., Wang, P., 2007. Data processing of onshore–offshore seismic experiment in Hongkong and Zhujiang River Delta region. Journal of Tropical Oceanography 26 (1), 35–38 (in Chinese). Xu, H.L., Qiu, X.L., Zhao, M.H., Sun, J.L., Zhu, J.J., 2006. Characteristics of the crustal structure and hypocentral tectonics in the epicentral area of Nan’ao earthquake (M7.5), the northeastern South China Sea. Chinese Science Bulletin 51 (Suppl.), 95–106. Xu, Y., Li, Z., Hao, T.Y., et al., 2007. Pn wave velocity and anisotropy in the northeastern South China Sea and adjacent region. Chinese Journal of Geophysics 50 (5), 1473–1479 (in Chinese). Yan, P., Zhou, D., Liu, Z., 2001. A crustal structure profile across the northern continental margin of the South China Sea. Tectonophysics 338, 1–21. Yang, X., Ma, J., Zhang, X., 2003. Summarization of genesis of low-velocity layer in continental crust. Geological Science and Technology Information 22, 35–41. Yao, B., Zeng, W., Hayes, D.E., Spangler, S., 1994. The Geological Memoir of South China Sea Surveyed Jointly by China and the USA. The Press of China University of Geoscience, Wuhan. Yin, Z., Lai, M., Xiong, S., et al., 1999. Crustal structure and velocity distribution based on deep seismic sounding along the Lianxian–Boluo–Gankou profile in southern China. Chinese Journal of Geophysics 42 (3), 383–392 (in Chinese). Zelt, C.A., Smith, R.B., 1992. Seismic travel time inversion for 2-D crustal velocity structure. Geophysical Journal International 108, 16–34. Zelt, C.A., 1999. Modeling strategies and model assessment for wide-angle seismic traveltime data. Geophysical Journal International 139, 183–204. Zeng, R.S., 1984. Introduction of Solid Geophysics. Science Press, Beijing. 446 pp (in Chinese). Zhang, Z., Wang, Y., 2007. Crustal structure and contact relationship revealed from deep seismic sounding data in South China. Physics of the Earth and Planetary Interiors 165, 114–126. Zhao, D., Hasegawa, A., Horiuchi, S., 1992. Tomographic imaging of P and S wave velocity structure beneath northeastern Japan. Journal of Geophysical Research 97, 19909–19928. Zhao, M., Qiu, X., Ye, C., Xia, K., Huang, C., Xie, J., Wang, P., 2004. Analysis on deep crustal structure along the onshore–offshore seismic profile across the Binhai (littoral) fault zone in northeastern South China Sea. Chinese Journal of Geophysics 47 (5), 845–852 (in Chinese). Zhao, M., Qiu, X., XU, H., et al., 2007. Distribution and identification of the lowvelocity layer in the northern South China Sea. Progress in Natural Science 17, 591–600. Zhou, D., Ru, K., Chen, H.Z., 1995. Kinematics of Cenozoic extension on the South China Sea continental margin and implications for the tectonic evolution of the region. Tectonophysics 251, 153–160. Zhou, Y., He, C., 2002. The relationship between low velocity layers and rheology of the crust in North China and its effect on strong earthquake. Seismology and Geology 24, 124–132.