- Email: [email protected]
1D velocity structure beneath broadband seismic stations in the Cretaceous Gyeongsang Basin of Korea by receiver function analyses
Tectonophysics 472 (2009) 158–168
Contents lists available at ScienceDirect
Tectonophysics j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / t e c t o
1D velocity structure beneath broadband seismic stations in the Cretaceous Gyeongsang Basin of Korea by receiver function analyses Seon Jeong Park 1, Jung Mo Lee ⁎, In-Chang Ryu 2 Department of Geology, Kyungpook National University, Daegu 702-701, Republic of Korea
A R T I C L E
I N F O
Article history: Received 20 March 2007 Received in revised form 4 May 2008 Accepted 26 May 2008 Available online 5 June 2008 Keywords: Crustal velocity structure Receiver function Moho Cretaceous Gyeongsang Basin
A B S T R A C T The crustal velocity structures beneath four broadband seismic stations (GKP, GSU, HDB, and BUS) in the Cretaceous Gyeongsang Basin, southeastern Korea, are estimated by using receiver function analyses employing teleseismic waveforms. The genetic algorithm is adopted to avoid the inherent non-uniqueness problem of the inversion. The inversion results are constrained by surface-wave dispersions to complement the shortcoming of the receiver function. The selected records of earthquakes distributed in three quadrants from each seismic station in the years from 2002 to 2005 are analyzed. Although limited quantity of data is used due to short operation periods, the resolution and confidence level are expected to be better than those using less data. The Moho depth under GKP is estimated to be 30 km showing no distinctive indication of dipping. This depth is well coincident with those in nearby velocity cross-sections obtained from crustal-scale seismic profiles compared to the 36-km deep Moho previously reported by other independent work. At GSU the Moho appears at the 32-km depth and a distinct low-velocity anomaly is detected at the depth of about 10 km. They agree with those in the velocity tomogram of the nearby survey line. An adakitic intrusion which results from the partial melting of a young and hot subducted oceanic crust in the basin during the Cretaceous is suggested as a possible geologic interpretation of the low-velocity zone. The Moho beneath HDB is 28-km deep, and agrees with those in nearby velocity cross-sections obtained from crustal-scale seismic profiles. This Moho depth at HDB is rather shallower than those of other stations. The significant PS phase amplitude and arrival time differences in the radial receiver functions conclude the Moho to dip southwestwardly. This is supported by polarities of direct P and PS in transverse receiver functions. Beneath BUS, P wave velocity increases gradually from the 32-km depth and it reaches 7.6 km/s at the 36-km depth. The Moho discontinuity beneath BUS is thought to be at an around 35-km depth. © 2008 Elsevier B.V. All rights reserved.
1. Introduction The crustal velocity structure is one of the most important subjects in seismology because it controls seismic wave propagation as well as provides fundamental geological and geophysical information. Traveltimes of seismic phases are most widely used for the velocity structure studies since it is easy to obtain from earthquake records. The data, however, lack the continuity of phases in Korea since seismic stations are sparsely distributed and earthquakes of magnitude larger than 5 have not frequently occurred. Subsequently, many uncertain crustal velocity models came out and they have been under controversy (e.g., Lee, 1979; Kim and Kim, 1983; Kim, 1995). Korean Crust ⁎ Corresponding author. Tel.: +82 53 950 6347; fax: +82 53 950 5362. E-mail addresses: [email protected] (S.J. Park), [email protected] (J.M. Lee), [email protected] (I.-C. Ryu). 1 Now with Civil Engineering Department, Korea Power Engineering Company, Inc., 351-1 Gugal-dong, Giheung-gu, Yongin-si, Gyeonggi-do 446-713, Republic of Korea. Tel.: +82 31 899 2227; fax: +82 31 899 2219. 2 Tel.: +82 53 950 5359; fax: +82 53 950 5362. 0040-1951/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.tecto.2008.05.032
Research Team (KCRT), has systematically approached to obtain reliable high resolution 3D crustal velocity structures of the Korean Peninsula area by earthquake data inversions and deep seismic profiling (e.g., Chang et al., 2004; Chang and Baag, 2005; Cho et al., 2006; Kim et al., 2007). The map showing the seismic profiles with 4 broadband seismic stations analyzed in this work is presented in Fig. 1. Two velocity cross-sections, the one for the 2002 WNW-ESE line and the other for the 2004 NWN-SSE line are presented in Figs. 2 and 3, respectively (Cho et al., 2006; Kim et al., 2007). Meanwhile 1D velocity structures beneath all 22 broadband seismic stations in the southern Korea were estimated by receiver function analyses, and the Moho depth distribution and the average crustal P wave velocities were reported (Chang et al., 2004; Chang and Baag, 2005). The reported Moho depth distribution is presented in Fig. 4 for the comparison purpose. At a glance, some discrepancy in the Moho depth around GKP can be found. The reported Moho depth beneath GKP by receiver function analyses is 36 km (Fig. 4) while those around GKP in the 2002 WNW-SES velocity cross-section are around 30 km (Fig. 2). Modern broadband seismic network has been installed in South Korea
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
159
Fig. 1. Index map showing locations of the KCRT2002 (black dotted line) and KCRT2004 (red dotted line) crustal-scale seismic survey lines, and 4 broadband seismic stations in the Gyeongsang Basin. The topography is presented in color. Shot locations in the survey lines are presented by oplus symbols with abbreviated names and coordinates in the same index color. The seismic stations are presented by red inverted triangles with station codes and coordinates in blue. Abbreviated shot location names and seismic station codes are: SS; Seosan, YD; Yeongdong, GJ; Gyeongjoo, YC; Yeoncheon, JP; Jeungpyeong, GS; Goseong, GKP; the Kyungpook National University Station, HDB; the Hyodongri Station, BUS; the Busan Station, and GSU; the Gyeongsang National University Station.
since 1996, and Chang and Baag (2005) should use limited data in their receiver function analyses. Especially four stations – GKP, GSU, HDB, and BUS – in the Gyeongsang Basin have been operated since 1998, 2000, 2000, and 2001, respectively. They could use only three earthquake data for GKP, GSU, HDB, and PUS each. It is worthwhile to note that PUS (35.1010°N, 129.0338°E) has moved to BUS (35.2487°N, 129.1125°E) in 2001. The coordinates of four stations are presented in Fig. 1. The Gyeongsang Basin is furthermore important in geology and tectonic setting of the Korean Peninsula. The purpose of the present work is to estimate the crustal velocity structures beneath the broadband seismic stations in the Gyeongsang Basin where GKP, GSU, HDB and BUS are located using the joint inversion of receiver functions and surface-wave dispersions. Since recent large earthquake data such as the Sumatra–Andaman Earthquake (Mw = 9.0, 26 Dec. 2004) are to be included in the analyses, not only the signal-to-noise ratio in the data but also the back azimuth coverage should be improved. More reliable velocity structures which may resolve the abovementioned discrepancy between the GKP result
of Chang and Baag (2005) and the velocity profiles of Cho et al. (2006) as well as reconcile themselves to some geological features in the Gyeongsang Basin are expected. 2. Geology of the Cretaceous Gyeongsang Basin Geologic terrenes in Korea include the Nangrim Massif (NM), Pyeongnam Basin (PB), Imjingang Belt (IB), Gyeonggi Massif (GM), Okchon and Taebacksan basins (OB-TB), Yeongnam Massif (YM), and Gyeongsang Basin (GB), from northwest to southeast (see index map in Fig. 5). The Cretaceous Gyeongsang Basin is situated in the southeastern part of the Korean Peninsula (Fig. 5). The basin is bounded on the west and north by the Precambrian metamorphic rocks (i.e., the Yeongnam Massif) as well as the Jurassic granitoids, and is overlapped on the east by the Early Tertiary calc-alkaline volcanic succession intercalated with minor amounts of sedimentary rocks (Fig. 5). An aggregate thickness of more than 9 km of siliciclastic and volcanic strata is preserved in the basin. These strata form the
160
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
Fig. 2. Velocity cross-section of the KCRT2002 survey line (Fig. 1) obtained by iterative forward ray tracing following least-square traveltime inversion. In addition to first breaks, traveltimes of late arrivals are included in the inversion. Major tectonic boundaries are indicated on the top of the cross-section. The Moho depths increase from 30 km at the WNW end to 34 km at the center of the transect, and then decrease to 28 km at the ENE end. A mid-crustal velocity discontinuity at depths of 15 ± 1 km is recognized in the WNW part of the transect, and extends up to a 110-km horizontal distance before fading out (after Cho et al., 2006).
Cretaceous Gyeongsang Supergroup that is subdivided into three lithostratigraphic units: the lower Sindong Group, the middle Hayang Group, and the upper Yucheon Group (Fig. 5; Chang 1975). Major sedimentation in the basin began in the pre-Barremian (?Hauterivian) and continued through the Aptian into the early Albian, which resulted in a thick siliciclastic succession of the Sindong and Hayang groups (Fig. 5; Chang, 1975). The initiation of major sedimentation in the basin was mainly attributed to an E–W crustal extension that resulted from the sinistral, brittle shearing in the southeastern part of the Korean Peninsula during the Early Cretaceous (Chough et al. 2000). This sinistral, brittle shearing might be due to the northward (oblique) subduction of the Izanagi Plate underneath the eastern margin of the Eurasian Plate during the Early Cretaceous. Although intermittent alkaline volcanisms occurred during the late Early Cretaceous (i.e., Hayang time), substantial volumes of volcanic rocks were erupted from the Late Cretaceous. As a result, overlying the Hayang Group is the thick volcanic sequence of the Yucheon Group that mainly comprises andesite and rhyolitic flows as well as their associated tuffs (Chang, 1975). The Cretaceous Gyeongsang Supergroup is extensively intruded by granitoids that are collectively termed as the Bulgugsa Granite (Fig. 5). They are in general fine- to medium-grained and are com-
posed of granite, granodiorite, and tonalite (Jin, 1985). Many of them are closely associated with volcanic rocks of the Late Cretaceous Yuchon Group in time and space, although some granitoids appear to be associated with the Early Cretaceous Hayang Group. The Late Cretaceous granitoids in the Gyeongsang Basin are correlated with the granitoids in the southwestern Japan that have similar intrusion ages (Maruyama et al., 1997). The intrusive ages of granitoids in both areas cluster between 100 Ma and 50 Ma (Maruyama et al., 1997). It is generally agreed that the Late Cretaceous to Early Tertiary magmatism in the Gyeongsang Basin and the southwestern Japan took place by the subduction of the proto-Pacific (Izanagi) Plate (Uyeda and Miyashiro, 1974; Jin, 1985). 3. Receiver function, genetic algorithm, and joint inversion The theory and methods for estimation of crustal velocity structures using receiver functions have been studied and developed by many seismologists since 1970s (e.g., Helmberger and Wiggins, 1971; DeySarkar and Wiggins, 1976; Langston, 1979; Owens et al., 1984; Ammon et al., 1990; Chang et al., 2004). Teleseismic P waveforms recorded at a three-component seismic station contain a wealth of information on the earthquake source, the Earth structure in the vicinity of both the source,
Fig. 3. Contoured velocity tomogram of KCRT2004 survey line (Fig. 1) obtained by first break inversion. Major tectonic boundaries and shot locations are indicated on the top of the tomogram. The Moho depths shallow to approximately 31-km deep about 60 km SSE of the NNW, and deepen and fluctuate to SSE. Abbreviated major tectonic unit names are: GM; Gyeonggi Massif, OFB; Okchon fold belt, YB; Yeongdong Basin, RM; Yeongnam Massif, and GB; Gyeongsang Basin. The arrow indicates the low-velocity zone which is found in the result of the GSU receiver function analysis of the present work (see Discussion). Abbreviated shot location names are presented in Fig. 1 (after Kim et al., 2007).
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
Fig. 4. Distribution of the Moho depths in southern Korea. Contours in kilometers were constructed by the interpolation of estimated values from velocity structures beneath 22 broadband seismic stations by receiver function analyses. Dashed lines imply uncertain contours owing to the scarce distribution of stations (after Chang and Baag, 2005).
161
Earthquake data from south-west, south-east, and north-east to the stations were collected to construct receiver functions with different back azimuths. The earthquakes of magnitude larger than 6.0 occurred from 2002 to 2005 were used in this study. We selected and analyzed 8 earthquake data recorded at GKP, 7 at GSU, 10 at HDB, and 6 at BUS, respectively. Earthquake data used in this work for each station are listed in Table 1 with magnitudes, epicenters, and back azimuths. The waterlevel parameter is set to 0.01. The parameter of the Gaussian filter is set to 2.5, which gives an effective high frequency limit of 0.5Hz. Not normalized amplitudes of the receiver function but true amplitudes are used to obtain the information about the shallow velocity structure. Calculated receiver functions are generally stacked to enhance their signal-to-noise ratio (Cassidy, 1992). The stacking bounds are 11° in back azimuth and 13° in epicentral distance, and they are tighter than those suggested by Owens (1984). Rayleigh wave phase velocities for the period range of 10–25 s with a 5-second interval which are estimated by Chang and Baag (2005) are used. The velocity models from which the phase velocities computed differ from the observed phase velocities by 0.1–km/s in even a considered period are rejected during inversion. The density is assumed as ρ = 0.32Vp + 0.77 (Berteussen, 1977). Although the receiver function modeling is in terms of S wave velocities, P wave pffiffiffi velocities converted from S wave velocities assuming VP = VS = 3 are consistently used in this paper to facilitate the comparison with other velocity cross-sections. 5. Results
the receiver and mantle propagation effects (Cassidy,1992). If it is able to remove the propagation effect and influence of the source, the waveforms have only the information about the Earth structure in the vicinity of the receiver. The crustal velocity structures under the seismic stations could be estimated by analyzing those waveforms. This is accomplished by deconvolving the vertical component of ground motion from the horizontal components (Langston, 1979). The details of the method are outlined in Langston (1979) and Ammon (1991). The genetic algorithm (GA) is an optimization technique to find the global solution. It searches the solution by reproduction and widens the range of searching by crossover and mutation. Because the GA is performed with a lot of initial models randomly generated, it does not strongly depend on the initial model. The GA is appropriate for the velocity inversion in the Korean Peninsula which has little a priori information about crustal structures (Chang and Baag, 2005). Uniform crossover micro-GA suggested by Syswerda (1989) is adopted in this work. The non-uniqueness problem may also occur by the velocity-depth trade-off since the relative traveltime is represented in receiver functions. Receiver functions are sensitive to shear wave velocity contrasts while surface-wave dispersions are sensitive to the average shear wave velocity. It is well known the joint inversion of these two datasets mutually complements (e.g., Last et al., 1997; Julià et al., 2000; Chang et al., 2004). The joint inversion is carried out by employing the surface surface-wave phase velocity data as constraints in the receiver function inversion. 4. Data processing There are several short period seismic stations and four broadband seismic stations which are operated by KIGAM (Korea Institute of Geoscience And Mineral Resources) and KMA (Korea Meteorological Administration) in the Gyeongsang Basin. The earthquake data recorded at the short period seismic stations are not adequate to be used in the receiver function analysis due to lack of the low frequency information. The earthquake data recorded at the broadband seismic stations – GKP, GSU, HDB, and BUS – are analyzed. The locations of the seismic stations with their coordinates are presented in Fig. 1.
The records of earthquakes occurred in the areas of Sumatra, the Solomon Islands, and Alaska were used to construct receiver functions for GKP. Their back azimuths are 228°, 143°, and 48°, respectively. The estimated velocity structures and the receiver functions are presented in Fig. 6. The records of three earthquakes (12/26/2004, 04/28/2005, 05/14/2005) occurred in the Sumatra area are stacked. In the result of inversion, a low-velocity zone exists at the depth of 8 km, and a sharp velocity jump from 6.6 to 7.4 km/s appears at the 30-km depth. This is expected to be the Moho. In the velocity profile estimated by using the records of earthquakes (12/12/2002, 06/07/2003, 10/08/2004) occurred in the Solomon Islands area, a velocity discontinuity considered as the Moho appears as a velocity jump from 7.0 to 7.4 km/s at the depth of 30 km, however, the velocity decreases and increases again just below it. A low-velocity zone is also shown at the 8-km depth. The records of earthquakes (05/25/2002, 06/23/2003) occurred in the Alaska area are stacked. A low-velocity zone appears at the depth of 8 km. P wave velocity increases gradually from the depth of 30 km to 34 km and it reaches 7.8 km/s. This is thought to be the Moho. Polarities of the transverse receiver function can indicate the orientation of a dipping interface (Langston, 1979). Transverse receiver functions are examined to identify the possible dipping Moho under GKP. Amplitudes of transverse receiver functions are too small to indicate the dipping Moho. The records of earthquakes occurred in the areas of Sumatra, the Solomon Islands, and the Aleutian Islands areas are analyzed to estimate the velocity structure under GSU. Their back azimuths are 229°, 144°, 50°, respectively. The records of two earthquakes (12/26/2004, 05/14/2005) occurred in Sumatra area, three (03/11/2003, 06/12/ 2003, 10/17/2003) in the Solomon Islands area, and two (06/23/ 2003, 06/14/2005) in Aleutian Islands area are selected to construct the receiver functions. The receiver functions and the estimated crustal velocity structures are presented in Fig. 7. A distinctive lowvelocity zone at the 10-km depth appears commonly in all velocity profiles. The Moho is expected to be at the depth of 32 km, however, the velocity contrast is not so distinctive. Examination of the transverse receiver functions indicates the flat Moho under GSU. Records of three earthquakes (12/26/2004, 05/14/2005, 05/19/ 2005) occurred in the Sumatra area, three (03/25/2003, 04/17/2004, 11/11/2004) in the Timor region, two (10/17/2003, 04/11/2005) in
162
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
Fig. 5. Geologic map of the Cretaceous Gyeongsang Basin, southeastern Korea. Abbreviations in index map: NM = Nangrim Massif, PB = Pyeongnam Basin, IB = Imjingang Belt, GM = Gyeonggi Massif, OB-TB = Okchon and Taebacksan basins (or Okchon fold belt in Fig. 2 and OFB in Fig. 3), YM = Yeongnam Massif, and GB = Gyeongsang Basin.
the Solomon Islands area, and two (05/25/2002, 06/14/2005) in the Alaska area are used to estimate the velocity structures beneath HDB. Their back azimuths are 228°, 186°, 150°, and 48°, respectively. Estimated velocity structures and receiver functions are presented in Fig. 8. In two velocity profiles estimated by using earthquake data from the Sumatra and Alaska areas, the velocity increases abruptly from 7.2 to 7.7 km/s at the depth of 30 km. The estimated velocity
structure from receiver function constructed by earthquakes occurred in the Timor region, however, shows a velocity jump from 7.3 to 7.8 km/s at the 28-km depth. In the velocity structures estimated by using earthquakes in the Solomon Islands area, the velocity starts to increase gradually from the 28-km depth. The individual receiver functions before stacking are computed and presented in Fig. 9 in order to examine the dipping Moho under
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
163
Table 1 List of earthquakes for each station used in this work. Station code
Time (yyyy/mm/dd hh:mm:ss.ssss)
Latitude (deg.)
Longitude (deg.)
Magnitude (Mw)
Back azimuth (deg.)
Angular distance (deg.)
GKP
2002/05/25 05:36:32.5300 2002/12/12 08:30:44.9200 2003/06/07 00:32:45.5700 2003/06/23 12:12:31.5100 2004/10/08 08:27:53.5400 2004/12/26 00:58:53.4500 2005/04/28 14:07:33.7000 2005/05/14 05:05:18.4800 2003/03/11 07:27:32.0100 2003/06/12 08:59:20.2300 2003/06/23 12:12:31.5100 2003/10/17 10:19:09.1200 2004/12/26 00:58:53.4500 2005/05/14 05:05:18.4800 2005/06/14 17:10:17.0000 2002/05/25 05:36:32.5300 2003/03/25 02:53:26.6800 2003/10/17 10:19:09.1200 2004/04/17 15:58:24.6100 2004/11/11 21:26:41.1500 2004/12/26 00:58:53.4500 2005/04/11 12:20:05.9600 2005/05/14 05:05:18.4800 2005/05/19 01:54:52.8500 2005/06/14 17:10:17.0000 2002/02/05 13:27:26.0000 2003/01/20 08:43:53.0000 2003/11/17 06:43:00.0000 2004/12/26 00:58:53.4500 2005/06/14 17:10:17.0000 2005/07/05 01:52:04.0000
53.903 − 4.828 − 5.095 51.445 − 10.951 3.308 2.132 0.587 − 4.702 − 6.001 51.445 − 5.503 3.308 0.587 51.309 53.903 − 8.328 − 5.503 − 7.352 − 8.152 3.308 − 3.484 0.587 1.989 51.309 − 5.370 − 10.420 51.300 3.308 51.309 1.852
− 161.081 153.302 152.502 176.708 162.161 95.874 96.799 98.459 153.219 154.798 176.708 152.178 95.874 98.459 179.413 − 161.081 120.671 152.178 128.373 124.868 95.874 145.909 98.459 97.041 179.413 151.240 160.700 178.600 95.874 179.413 97.047
6.1 6.6 6.0 6.4 6.1 9.0 6.2 6.8 6.7 6.1 6.4 6.2 9.0 6.8 6.8 6.1 6.2 6.2 6.1 6.5 9.0 6.5 6.8 6.7 6.8 6.6 7.2 7.8 9.0 6.8 6.7
45.98 145.07 146.21 50.10 139.29 230.46 228.52 225.33 144.07 143.06 49.43 145.85 230.47 228.11 49.99 45.99 192.35 147.76 181.50 186.48 231.46 154.90 226.36 229.14 49.89 148.24 140.65 49.84 231.65 49.89 229.05
50.76 46.65 46.52 37.39 56.33 44.27 44.67 44.92 46.15 48.04 38.75 46.36 43.48 43.84 39.83 50.42 44.59 46.26 42.86 43.85 44.67 41.96 45.27 45.02 39.16 45.45 54.36 38.65 44.24 39.16 44.63
GSU
HDB
BUS
HDB which has been suggested by Chang and Baag (2005). The receiver functions of the Sumatra Andaman Earthquake (the bottom one in Fig. 9) show the large sharp direct P phase and the relatively smaller PS phase compared to others in the radial component, and the little transverse component. This may come from lack of high frequency energy of very large earthquakes. Polarity difference is not
clear in the transverse components, either. In the radial components, the amplitudes and arrival times of PS phases significantly vary, which has been suggested by Cassidy (1992) as an indication of dipping layers. The PS phases of the Alaska area earthquakes (top two in Fig. 9) arrive earlier than those of the Sumatra area earthquakes (bottom two excluding the Sumatra Andaman event). The former shows PS
Fig. 6. Estimated velocity structures beneath GKP and corresponding stacked receiver functions from earthquakes occurred in the Sumatra area (left), the Solomon Islands area (middle), and the Alaska area (right). The receiver functions synthesized from the estimated velocity structures (red dash-dot line) are compared with the observed ones (black solid line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
164
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
Fig. 7. Estimated velocity structures beneath GSU and corresponding stacked receiver functions from earthquakes occurred in the Sumatra area (left), the Solomon Islands area (middle), and the Aleutian Islands area (right). The receiver functions synthesized from the estimated velocity structures (red dash-dot line) are compared with the observed ones (black solid line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
amplitudes much smaller than those of the latter. They are located almost symmetrically around HDB. This suggests the southwestward dipping Moho interface. If the polarity difference in the transverses components is investigated with this information, the polarities of direct P and PS agree reasonably with the synthetic results of dipping interfaces by Cassidy (1992) — negative direct P and positive PS polarities of the Solomon Islands area (third and forth in Fig. 9) and the Timor region (fifth to seventh) earthquakes, and large direct P
amplitudes of the Solomon Islands area events which is located almost in the strike direction. Records of two (12/26/2004, 07/05/2005) earthquakes occurred in the Sumatra area, two (01/20/2003, 02/05/2003) in the Solomon Islands area, and two (11/17/2003, 06/14/2005) in the Aleutian Islands area are used to construct receiver functions for BUS. Their back azimuths are 230°, 144°, and 50°, respectively. The estimated velocity models and receiver functions are presented in Fig. 10. A sharp
Fig. 8. Estimated velocity structures beneath HDB and corresponding stacked receiver functions from earthquakes occurred in the Sumatra area, the Timor region, the Solomon Islands area, and the Alaska area (left to right). The receiver functions synthesized from the estimated velocity structures (red dash-dot line) are compared with the observed ones (black solid line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
165
Fig. 9. Individual radial (left) and transverse (right) receiver functions of HDB with event dates and back azimuths. From top to bottom, uppermost two are constructed from the earthquakes occurred in the Alaska area, upper two are from those in the Solomon Islands area, middle three are from those in the Timor region, and lower three are from those in the Sumatra area. All the receiver functions are plotted at the same amplitude scale. Although polarity difference is not clear in the transverse components, the amplitudes and arrival times of PS phases in the radial components significantly vary, which indicate the dipping Moho.
velocity increase from 6.1 to 6.6 km/s appears at the depth of 10 km in two velocity profiles using data from the areas of Sumatra and Solomon Islands. The velocity increases sharply from 7.0 to 7.6 km/s at the 34-km depth in the velocity profile using data from the Aleutian Islands area, and the velocity increases from 7.2 to 7.6 km/s at the 36km depth in the velocity structure from data of the Sumatra area. In the velocity profile using data from the Solomon Islands, the velocity increases gradually from 6.6 to 7.5 km/s at the 32–36-km depth interval. The Moho is expected to be at the depths between 34 km and 36 km. It is the deepest Moho in the Gyeongsang Basin analyzed in this work.
6. Discussion The crustal velocity structures of the Gyeongsang Basin have been estimated by analyzing receiver functions. To facilitate discussion, the composite velocity structures were constructed by averaging the velocity structures estimated from individual stacked receiver functions under each station. They are presented in Fig. 11. The 1D velocity structures of the locations nearest to GKP and HDB are extracted from the KCRT2002 velocity cross-section (Fig. 2), and are also presented in Fig. 11. Those for GSU are extracted from the KCRT2004 velocity cross-section (Fig. 3), and are also presented in Fig. 11. The distinctive
166
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
Fig. 10. Estimated velocity structures beneath BUS and corresponding stacked receiver functions from earthquakes occurred in the Sumatra area (left), the Solomon Islands area (middle), and the Aleutian Islands area (right). The receiver functions synthesized from the estimated velocity structures (red dash-dot line) are compared with the observed ones (black solid line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
features of Fig. 11 are the Moho depths under GKP and HDB, and the shallow low-velocity discontinuity under GSU. They agree with results of KCRT2002 pffiffiffi and KCRT2004. It is worthwhile to note that the assumed VP = VS = 3 is probably too low for the mantle. This partly explains the low mantle velocity of about 7.5 km/s in the present results. The followings come out as a result in the course of this work. 6.1. Velocity discontinuity and low-velocity anomaly All of the velocity profiles estimated in this work show some kind of weak or strong velocity discontinuities at about 10-km depths. Especially distinctive low-velocity anomalies appear at about 10-km depths in the estimated velocity structures beneath GKP and GSU (Fig. 11). The low-velocity zone under GSU is also shown in the velocity tomogram of KCRT2004 survey line (Figs. 3 and 11). The low-velocity
zone the velocity of which is about 5.8 km/s appears at around 10-km depths near the SSE end, and extends to a 35-km horizontal distance. GSU stands apart from the KCRT2004 survey line by less than 15 km (Fig. 1). Although further study related to these features is needed, a possible explanation for these shallow low-velocity anomalies may be continental margin volcanism (e.g., adakitic intrusion) in the basin, which is partly attributed to the upwelling flow in the upper mantle resulted from ridge subduction and subsequent slab window formation during the Late Cretaceous. The ridge subduction and subsequent slab window formation have been considered as principal causes of adakitic volcanism and low-velocity anomaly in active continental margin (Thorkelson and Breitsprecher, 2005). Thorkelson and Breitsprecher (2005) demonstrated a connection between subduction of young oceanic crusts and production of intermediate to felsic igneous rocks which bear the signature of a garnetifeours residuum. Such
Fig. 11. Composite velocity structures constructed by averaging the velocity structures estimated from analyses of stacked receiver functions for GKP, GSU, HDB, and BUS. The 1D velocity structures of the locations nearest to GKP and HDB are extracted from the KCRT2002 velocity cross-section (Fig. 2), and are presented (red dotted lines). Those for GSU are extracted from the KCRT2004 cross-section (Fig. 3), and are also presented (blue dotted line). Arrows indicate the Moho depths estimated in this work.
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
magmatic rocks are compositionally similar to Tertiary lavas on Adak Island in the Aleutian arc, which were identified as products of slab melting by Kay (1978). This petrologic family, termed adakites, was described by Drummond and Defant (1990) as high-alumina, intermediate to felsic volcanic rocks typically hosting phenocrysts of plagioclase, amphibole, mica, and orthophyroxene, and lacking phenocrysts of clinophyroxene. Recent petrochemical study on the granitoids within the Cretaceous Gyeongsang Basin indicates that adakitic volcanism locally has occurred in the basin. Petrogenesis of these granitoids is consistent with a slab window model in which parental magmas were derived by decompression melting of the suboceanic mantle (i.e., subducted paleo-Pacific Plate or Izanagi Plate) that upwelled into the opening that formed beneath continental margin following ridge subduction since Late Cretaceous (Wee et al., 2006). Adakitic rocks of Late Cretaceous ages have also been reported in the southwestern Japan, which are usually considered to be generated by the melting of a young and hot subducted oceanic crust, i.e. paleo-Pacific or Izanagi Plate (Takahasi et al., 2005). As well recent geophysical investigation along the northeastern Pacific margin indicates that a large low-velocity anomaly is found, which may reflect the upwelling mantle in the Pacific-Juan de Fuca slab window near the subducted edge of the Pacific Plate (Qi et al., 2007). 6.2. Moho depth and characteristics The Moho under GKP appears as a sharp velocity discontinuity from 6.8 to 7.4 km/s at the 30-km depth in velocity structures estimated by analyses of receiver functions constructed from events in the Sumatra and the Solomon areas. In contrast with this, it appears a gradual velocity increase from 6.7 to 7.9 km/s in the 30–36-km depth range in the velocity structure by the receiver function from events in the Alaska region (Fig. 6). No dipping evidence such as polarity reversals in the transverse receiver functions has been found. It appears at the 30-km depth as a sharp velocity discontinuity in the composite velocity structure, and the Moho depth around GKP in the KCRT2002 profile coincides with it (Fig. 11). In contrast with the 36-km deep Moho reported without definite conclusion by Chang and Baag (2005), a 30-km deep Moho appears to be reasonable under GKP. The Moho under GSU appears to be at the 32-km depth, which agrees with that in the velocity tomogram of the KCRT2004 survey line. All velocity structures under GSU from three back azimuths show a similar trend. No polarity reversal has been found in the transverse receiver functions. It is reasonable to conclude that the Moho under GSU is flat. The Moho depths under HDB from individual receiver functions are 28– 30 km and it appears at the 28-km depth in the composite velocity structure. It agrees with that in the KCRT2002 velocity cross-section (Fig. 11). It is the shallowest one among those analyzed in this work. As discussed in the previous section, the southwestward dipping Moho is suggested in contrast with westward dipping one by Chang and Baag (2005). The Moho depth under BUS is estimated to be 34–36 km, and is the deepest among those analyzed in this work. The Moho depths vary as less than 2 km in the velocity profiles from receiver functions with different back azimuths, and polarity reversals have not been found in the transverse receiver functions. It is reasonable to conclude that the Moho under BUS is flat. All of these results will contribute to mapping the Moho depths in the southeast Korean Peninsula, and it is expected that more realistic Moho depth distribution which reconciles to the results of KCRT seismic profiles will come out in the near future. 7. Conclusions Crustal velocity structures beneath the broadband seismic stations in the Gyeongsang Basin, GKP, GSU, HDB, and BUS, are estimated by using the joint inversion of receiver functions and surface-wave dispersions. Recent earthquake data with good signal-to-noise ratio are included and back azimuth coverage is extended. The former
167
improves resolution in the estimated velocity structures, and the latter does the confidence level in the interpretation. The estimated Moho depths are 30 km, 32 km, 28 km and 34–36 km under GKP, GSU, HDB, and BUS, respectively. Moho depths under GKP, GSU, and HDB harmonize with those from nearby velocity cross-sections obtained from crustal-scale seismic profiles. The estimated Moho depth under BUS appears to be reasonable although no other data to verify it are available at present. Some kind of mid-crustal velocity discontinuity at the depth of around 10 km is commonly found. The distinctive 8–10-km deep low-velocity zone under GSU is also shown in the velocity tomogram of the KCRT2004 seismic profile, and the partial melting of a young and hot subducted oceanic crust in the Cretaceous is a possible geologic interpretation of this low-velocity zone. The southwestward dipping Moho under HDB is identified from the variation of amplitudes, arrival times, and polarities of PS in the individual receiver functions. Acknowledgements The authors thank the anonymous reviewers and the guest editors, Professors. T. Ito, H. Thybo, and T. Iwasaki, for their efforts and constructive comments. This work was funded by the Korea Meteorological Administration Research and Development Program under Grant CATER 2006-5206. References Ammon, C.J., 1991. The isolation of receiver effects from teleseismic P waveforms. Bull. Seism. Soc. Am. 81, 2504–2510. Ammon, C.J., Randall, G., Zandt, G., 1990. On the nonuniqueness of receiver function inversions. J. Geophys. Res. 95, 15303–15318. Berteussen, K.A., 1977. Moho depth determinations based on spectral ratio analysis of NORSAR long-period P waves. Phys. Earth Planet. Inter. 15, 13–27. Cassidy, J.F., 1992. Numerical experiments in broadband receiver function analysis. Bull. Seism. Soc. Am. 82, 1453–1474. Chang, K.H., 1975. Cretaceous stratigraphy of southeast Korea. J. Geol. Soc. Korea 11, 1–23. Chang, S.J., Baag, C.E., 2005. Crustal structure in Southern Korea from joint analysis of teleseismic receiver functions and surface-wave dispersion. Bull. Seism. Soc. Am. 95, 1516–1534. Chang, S.J., Baag, C.E., Langston, C.A., 2004. Joint analysis of teleseismic receiver functions and surface wave dispersion using genetic algorithm. Bull. Seism. Soc. Am. 94, 691–704. Cho, H.M., Baag, C.E., Lee, J.M., Moon, W.M., Jung, H., Kim, K.Y., Asudeh, I., 2006. Crustal velocity structure across the southern Korean Peninsula from seismic refraction survey. Geophys. Res. Lett. 33, L06307. doi:10.1029/2005GL025145. Chough, S.K., Kwon, S.T., Ree, J.H., Choi, D.K., 2000. Tectonic and sedimentary evolution of the Korean peninsula: a review and new view. Earth Sci. Rev. 52, 175–235. Dey-Sarkar, S.K., Wiggins, R., 1976. Upper mantle structure in western Canada. J. Geophys. Res. 81. doi:10.1029/OJGREA000081000320003619000001. Drummond, M.S., Defant, M.J., 1990. A model for trondhjemite–tonalite–dacite genesis and crustal growth via slab melting; archean to modern comparisons. J. Geophys. Res. 95, 21503–21521. Helmberger, D., Wiggins, R.A., 1971. Upper mantle structure of Midwestern United States. J. Geophys. Res. 76, 3229–3245. Jin, M.S., 1985. Geochemistry of the Cretaceous to early Tertiary granitic rocks in southern Korea: part I, major elements geochemistry. J. Geol. Soc. Korea 21, 297–316. Julià, J., Ammon, C.J., Herrmann, R.B., Correig, A.M., 2000. Joint inversion of receiver function and surface wave dispersion observations. Geophys. J. Int. 143, 99–112. Kay, R.W., 1978. Aleutian magnesian andesites; melts from subducted Pacific Ocean crust. J. Volcanol. Geotherm. Res. 4, 117–132. Kim, S.K., 1995. A study on the crustal structure of the Korean Peninsula. J. Geol. Soc. Korea 31, 393–403. Kim, S.J., Kim, S.G., 1983. A study on the crustal structure of South Korea by using seismic waves. J. Korean Inst. Mining Geol. 16, 51–61. Kim, K.Y., Lee, J.M., Moon, W., Baag, C.E., Jung, H., Hong, M.H., 2007. Crustal structure of the Korean Peninsula from seismic waves generated by large explosions in 2002 and 2004. Pure Appl. Geophys. 164, 97–113. Langston, C.A., 1979. Structure under Mount Rainier, Washington, inferred from teleseismic body waves. J. Geophys. Res. 84, 4749–4762. Last, R.J., Nyblade, A.A., Langston, C.A., 1997. Crustal structure of the east African plateau from receiver functions and Rayleigh wave phase velocities. J. Geophys. Res. 102, 24469–24483. Lee, K.,1979. On crustal structure of the Korean Peninsula. J. Geol. Soc. Korea 15, 253–258. Maruyama, S., Isozaki, Y., Kimura, G., Terabayashi, M., 1997. Paleo-geographic maps of the Japanese islands. plate tectonic synthesis from 750 Ma to the present. Island Arc. 6, 121–142.
168
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
Owens, T. J., 1984. Determination of crustal and upper mantle structure from analysis of broadband teleseismic P-waveforms, Ph.D. Thesis, University of Utah, Salt Lake City, 146. Owens, T.J., Zandt, G., Taylor, S.R., 1984. Seismic evidence for an ancient rift beneath the Cumberland Plateau, Tennessee: a detailed analysis of broadband teleseismic P waveforms. J. Geophys. Res. 89, 7783–7795. Qi, C., Zhao, D., Chen, Y., 2007. Search for deep slab segments under Alaska. Phys. Earth Planet. Inter. 165, 68–82. Syswerda, G., 1989. Uniform crossover in genetic algorithms. In: Schaffer, J (Ed.), Proceedings of the Third International Conference on Genetic Algorithms. Morgan Kaufmann Publishers, Los Altos, CA, pp. 2–9.
Takahasi, Y., Kagashima, S.-I., Mikoshiba, M., 2005. Geochemistry of adakitic quartz diorite in the Yamizo Mountains, central Japan: implications for Early Cretaceous adakitic magmatism in the inner zone of southwest Japan. Island Arc 14, 150–164. Thorkelson, D.J., Katrin Breitsprecher, K., 2005. Partial melting of slab window margins: genesis of adakitic and non-adakitic magmas. Lithos 79, 25–41. Uyeda, S., Miyashiro, A., 1974. Plate tectonics and the Japanese Islands: a synthesis. Bull. Geol. Soc. Am. 85, 1159–1170. Wee, S.M., Choi, S.G., Ryu, I.C., Shin, H.J., 2006. Geochemical characteristics of the Cretaceous Jindong granites in the southeastern part of the Gyeongsang Basin, Korea: focused on adakitic signatures. J. Korea Soc. Econ. Environ. Geol. 39, 555–566.
Contents lists available at ScienceDirect
Tectonophysics j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / t e c t o
1D velocity structure beneath broadband seismic stations in the Cretaceous Gyeongsang Basin of Korea by receiver function analyses Seon Jeong Park 1, Jung Mo Lee ⁎, In-Chang Ryu 2 Department of Geology, Kyungpook National University, Daegu 702-701, Republic of Korea
A R T I C L E
I N F O
Article history: Received 20 March 2007 Received in revised form 4 May 2008 Accepted 26 May 2008 Available online 5 June 2008 Keywords: Crustal velocity structure Receiver function Moho Cretaceous Gyeongsang Basin
A B S T R A C T The crustal velocity structures beneath four broadband seismic stations (GKP, GSU, HDB, and BUS) in the Cretaceous Gyeongsang Basin, southeastern Korea, are estimated by using receiver function analyses employing teleseismic waveforms. The genetic algorithm is adopted to avoid the inherent non-uniqueness problem of the inversion. The inversion results are constrained by surface-wave dispersions to complement the shortcoming of the receiver function. The selected records of earthquakes distributed in three quadrants from each seismic station in the years from 2002 to 2005 are analyzed. Although limited quantity of data is used due to short operation periods, the resolution and confidence level are expected to be better than those using less data. The Moho depth under GKP is estimated to be 30 km showing no distinctive indication of dipping. This depth is well coincident with those in nearby velocity cross-sections obtained from crustal-scale seismic profiles compared to the 36-km deep Moho previously reported by other independent work. At GSU the Moho appears at the 32-km depth and a distinct low-velocity anomaly is detected at the depth of about 10 km. They agree with those in the velocity tomogram of the nearby survey line. An adakitic intrusion which results from the partial melting of a young and hot subducted oceanic crust in the basin during the Cretaceous is suggested as a possible geologic interpretation of the low-velocity zone. The Moho beneath HDB is 28-km deep, and agrees with those in nearby velocity cross-sections obtained from crustal-scale seismic profiles. This Moho depth at HDB is rather shallower than those of other stations. The significant PS phase amplitude and arrival time differences in the radial receiver functions conclude the Moho to dip southwestwardly. This is supported by polarities of direct P and PS in transverse receiver functions. Beneath BUS, P wave velocity increases gradually from the 32-km depth and it reaches 7.6 km/s at the 36-km depth. The Moho discontinuity beneath BUS is thought to be at an around 35-km depth. © 2008 Elsevier B.V. All rights reserved.
1. Introduction The crustal velocity structure is one of the most important subjects in seismology because it controls seismic wave propagation as well as provides fundamental geological and geophysical information. Traveltimes of seismic phases are most widely used for the velocity structure studies since it is easy to obtain from earthquake records. The data, however, lack the continuity of phases in Korea since seismic stations are sparsely distributed and earthquakes of magnitude larger than 5 have not frequently occurred. Subsequently, many uncertain crustal velocity models came out and they have been under controversy (e.g., Lee, 1979; Kim and Kim, 1983; Kim, 1995). Korean Crust ⁎ Corresponding author. Tel.: +82 53 950 6347; fax: +82 53 950 5362. E-mail addresses: [email protected] (S.J. Park), [email protected] (J.M. Lee), [email protected] (I.-C. Ryu). 1 Now with Civil Engineering Department, Korea Power Engineering Company, Inc., 351-1 Gugal-dong, Giheung-gu, Yongin-si, Gyeonggi-do 446-713, Republic of Korea. Tel.: +82 31 899 2227; fax: +82 31 899 2219. 2 Tel.: +82 53 950 5359; fax: +82 53 950 5362. 0040-1951/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.tecto.2008.05.032
Research Team (KCRT), has systematically approached to obtain reliable high resolution 3D crustal velocity structures of the Korean Peninsula area by earthquake data inversions and deep seismic profiling (e.g., Chang et al., 2004; Chang and Baag, 2005; Cho et al., 2006; Kim et al., 2007). The map showing the seismic profiles with 4 broadband seismic stations analyzed in this work is presented in Fig. 1. Two velocity cross-sections, the one for the 2002 WNW-ESE line and the other for the 2004 NWN-SSE line are presented in Figs. 2 and 3, respectively (Cho et al., 2006; Kim et al., 2007). Meanwhile 1D velocity structures beneath all 22 broadband seismic stations in the southern Korea were estimated by receiver function analyses, and the Moho depth distribution and the average crustal P wave velocities were reported (Chang et al., 2004; Chang and Baag, 2005). The reported Moho depth distribution is presented in Fig. 4 for the comparison purpose. At a glance, some discrepancy in the Moho depth around GKP can be found. The reported Moho depth beneath GKP by receiver function analyses is 36 km (Fig. 4) while those around GKP in the 2002 WNW-SES velocity cross-section are around 30 km (Fig. 2). Modern broadband seismic network has been installed in South Korea
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
159
Fig. 1. Index map showing locations of the KCRT2002 (black dotted line) and KCRT2004 (red dotted line) crustal-scale seismic survey lines, and 4 broadband seismic stations in the Gyeongsang Basin. The topography is presented in color. Shot locations in the survey lines are presented by oplus symbols with abbreviated names and coordinates in the same index color. The seismic stations are presented by red inverted triangles with station codes and coordinates in blue. Abbreviated shot location names and seismic station codes are: SS; Seosan, YD; Yeongdong, GJ; Gyeongjoo, YC; Yeoncheon, JP; Jeungpyeong, GS; Goseong, GKP; the Kyungpook National University Station, HDB; the Hyodongri Station, BUS; the Busan Station, and GSU; the Gyeongsang National University Station.
since 1996, and Chang and Baag (2005) should use limited data in their receiver function analyses. Especially four stations – GKP, GSU, HDB, and BUS – in the Gyeongsang Basin have been operated since 1998, 2000, 2000, and 2001, respectively. They could use only three earthquake data for GKP, GSU, HDB, and PUS each. It is worthwhile to note that PUS (35.1010°N, 129.0338°E) has moved to BUS (35.2487°N, 129.1125°E) in 2001. The coordinates of four stations are presented in Fig. 1. The Gyeongsang Basin is furthermore important in geology and tectonic setting of the Korean Peninsula. The purpose of the present work is to estimate the crustal velocity structures beneath the broadband seismic stations in the Gyeongsang Basin where GKP, GSU, HDB and BUS are located using the joint inversion of receiver functions and surface-wave dispersions. Since recent large earthquake data such as the Sumatra–Andaman Earthquake (Mw = 9.0, 26 Dec. 2004) are to be included in the analyses, not only the signal-to-noise ratio in the data but also the back azimuth coverage should be improved. More reliable velocity structures which may resolve the abovementioned discrepancy between the GKP result
of Chang and Baag (2005) and the velocity profiles of Cho et al. (2006) as well as reconcile themselves to some geological features in the Gyeongsang Basin are expected. 2. Geology of the Cretaceous Gyeongsang Basin Geologic terrenes in Korea include the Nangrim Massif (NM), Pyeongnam Basin (PB), Imjingang Belt (IB), Gyeonggi Massif (GM), Okchon and Taebacksan basins (OB-TB), Yeongnam Massif (YM), and Gyeongsang Basin (GB), from northwest to southeast (see index map in Fig. 5). The Cretaceous Gyeongsang Basin is situated in the southeastern part of the Korean Peninsula (Fig. 5). The basin is bounded on the west and north by the Precambrian metamorphic rocks (i.e., the Yeongnam Massif) as well as the Jurassic granitoids, and is overlapped on the east by the Early Tertiary calc-alkaline volcanic succession intercalated with minor amounts of sedimentary rocks (Fig. 5). An aggregate thickness of more than 9 km of siliciclastic and volcanic strata is preserved in the basin. These strata form the
160
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
Fig. 2. Velocity cross-section of the KCRT2002 survey line (Fig. 1) obtained by iterative forward ray tracing following least-square traveltime inversion. In addition to first breaks, traveltimes of late arrivals are included in the inversion. Major tectonic boundaries are indicated on the top of the cross-section. The Moho depths increase from 30 km at the WNW end to 34 km at the center of the transect, and then decrease to 28 km at the ENE end. A mid-crustal velocity discontinuity at depths of 15 ± 1 km is recognized in the WNW part of the transect, and extends up to a 110-km horizontal distance before fading out (after Cho et al., 2006).
Cretaceous Gyeongsang Supergroup that is subdivided into three lithostratigraphic units: the lower Sindong Group, the middle Hayang Group, and the upper Yucheon Group (Fig. 5; Chang 1975). Major sedimentation in the basin began in the pre-Barremian (?Hauterivian) and continued through the Aptian into the early Albian, which resulted in a thick siliciclastic succession of the Sindong and Hayang groups (Fig. 5; Chang, 1975). The initiation of major sedimentation in the basin was mainly attributed to an E–W crustal extension that resulted from the sinistral, brittle shearing in the southeastern part of the Korean Peninsula during the Early Cretaceous (Chough et al. 2000). This sinistral, brittle shearing might be due to the northward (oblique) subduction of the Izanagi Plate underneath the eastern margin of the Eurasian Plate during the Early Cretaceous. Although intermittent alkaline volcanisms occurred during the late Early Cretaceous (i.e., Hayang time), substantial volumes of volcanic rocks were erupted from the Late Cretaceous. As a result, overlying the Hayang Group is the thick volcanic sequence of the Yucheon Group that mainly comprises andesite and rhyolitic flows as well as their associated tuffs (Chang, 1975). The Cretaceous Gyeongsang Supergroup is extensively intruded by granitoids that are collectively termed as the Bulgugsa Granite (Fig. 5). They are in general fine- to medium-grained and are com-
posed of granite, granodiorite, and tonalite (Jin, 1985). Many of them are closely associated with volcanic rocks of the Late Cretaceous Yuchon Group in time and space, although some granitoids appear to be associated with the Early Cretaceous Hayang Group. The Late Cretaceous granitoids in the Gyeongsang Basin are correlated with the granitoids in the southwestern Japan that have similar intrusion ages (Maruyama et al., 1997). The intrusive ages of granitoids in both areas cluster between 100 Ma and 50 Ma (Maruyama et al., 1997). It is generally agreed that the Late Cretaceous to Early Tertiary magmatism in the Gyeongsang Basin and the southwestern Japan took place by the subduction of the proto-Pacific (Izanagi) Plate (Uyeda and Miyashiro, 1974; Jin, 1985). 3. Receiver function, genetic algorithm, and joint inversion The theory and methods for estimation of crustal velocity structures using receiver functions have been studied and developed by many seismologists since 1970s (e.g., Helmberger and Wiggins, 1971; DeySarkar and Wiggins, 1976; Langston, 1979; Owens et al., 1984; Ammon et al., 1990; Chang et al., 2004). Teleseismic P waveforms recorded at a three-component seismic station contain a wealth of information on the earthquake source, the Earth structure in the vicinity of both the source,
Fig. 3. Contoured velocity tomogram of KCRT2004 survey line (Fig. 1) obtained by first break inversion. Major tectonic boundaries and shot locations are indicated on the top of the tomogram. The Moho depths shallow to approximately 31-km deep about 60 km SSE of the NNW, and deepen and fluctuate to SSE. Abbreviated major tectonic unit names are: GM; Gyeonggi Massif, OFB; Okchon fold belt, YB; Yeongdong Basin, RM; Yeongnam Massif, and GB; Gyeongsang Basin. The arrow indicates the low-velocity zone which is found in the result of the GSU receiver function analysis of the present work (see Discussion). Abbreviated shot location names are presented in Fig. 1 (after Kim et al., 2007).
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
Fig. 4. Distribution of the Moho depths in southern Korea. Contours in kilometers were constructed by the interpolation of estimated values from velocity structures beneath 22 broadband seismic stations by receiver function analyses. Dashed lines imply uncertain contours owing to the scarce distribution of stations (after Chang and Baag, 2005).
161
Earthquake data from south-west, south-east, and north-east to the stations were collected to construct receiver functions with different back azimuths. The earthquakes of magnitude larger than 6.0 occurred from 2002 to 2005 were used in this study. We selected and analyzed 8 earthquake data recorded at GKP, 7 at GSU, 10 at HDB, and 6 at BUS, respectively. Earthquake data used in this work for each station are listed in Table 1 with magnitudes, epicenters, and back azimuths. The waterlevel parameter is set to 0.01. The parameter of the Gaussian filter is set to 2.5, which gives an effective high frequency limit of 0.5Hz. Not normalized amplitudes of the receiver function but true amplitudes are used to obtain the information about the shallow velocity structure. Calculated receiver functions are generally stacked to enhance their signal-to-noise ratio (Cassidy, 1992). The stacking bounds are 11° in back azimuth and 13° in epicentral distance, and they are tighter than those suggested by Owens (1984). Rayleigh wave phase velocities for the period range of 10–25 s with a 5-second interval which are estimated by Chang and Baag (2005) are used. The velocity models from which the phase velocities computed differ from the observed phase velocities by 0.1–km/s in even a considered period are rejected during inversion. The density is assumed as ρ = 0.32Vp + 0.77 (Berteussen, 1977). Although the receiver function modeling is in terms of S wave velocities, P wave pffiffiffi velocities converted from S wave velocities assuming VP = VS = 3 are consistently used in this paper to facilitate the comparison with other velocity cross-sections. 5. Results
the receiver and mantle propagation effects (Cassidy,1992). If it is able to remove the propagation effect and influence of the source, the waveforms have only the information about the Earth structure in the vicinity of the receiver. The crustal velocity structures under the seismic stations could be estimated by analyzing those waveforms. This is accomplished by deconvolving the vertical component of ground motion from the horizontal components (Langston, 1979). The details of the method are outlined in Langston (1979) and Ammon (1991). The genetic algorithm (GA) is an optimization technique to find the global solution. It searches the solution by reproduction and widens the range of searching by crossover and mutation. Because the GA is performed with a lot of initial models randomly generated, it does not strongly depend on the initial model. The GA is appropriate for the velocity inversion in the Korean Peninsula which has little a priori information about crustal structures (Chang and Baag, 2005). Uniform crossover micro-GA suggested by Syswerda (1989) is adopted in this work. The non-uniqueness problem may also occur by the velocity-depth trade-off since the relative traveltime is represented in receiver functions. Receiver functions are sensitive to shear wave velocity contrasts while surface-wave dispersions are sensitive to the average shear wave velocity. It is well known the joint inversion of these two datasets mutually complements (e.g., Last et al., 1997; Julià et al., 2000; Chang et al., 2004). The joint inversion is carried out by employing the surface surface-wave phase velocity data as constraints in the receiver function inversion. 4. Data processing There are several short period seismic stations and four broadband seismic stations which are operated by KIGAM (Korea Institute of Geoscience And Mineral Resources) and KMA (Korea Meteorological Administration) in the Gyeongsang Basin. The earthquake data recorded at the short period seismic stations are not adequate to be used in the receiver function analysis due to lack of the low frequency information. The earthquake data recorded at the broadband seismic stations – GKP, GSU, HDB, and BUS – are analyzed. The locations of the seismic stations with their coordinates are presented in Fig. 1.
The records of earthquakes occurred in the areas of Sumatra, the Solomon Islands, and Alaska were used to construct receiver functions for GKP. Their back azimuths are 228°, 143°, and 48°, respectively. The estimated velocity structures and the receiver functions are presented in Fig. 6. The records of three earthquakes (12/26/2004, 04/28/2005, 05/14/2005) occurred in the Sumatra area are stacked. In the result of inversion, a low-velocity zone exists at the depth of 8 km, and a sharp velocity jump from 6.6 to 7.4 km/s appears at the 30-km depth. This is expected to be the Moho. In the velocity profile estimated by using the records of earthquakes (12/12/2002, 06/07/2003, 10/08/2004) occurred in the Solomon Islands area, a velocity discontinuity considered as the Moho appears as a velocity jump from 7.0 to 7.4 km/s at the depth of 30 km, however, the velocity decreases and increases again just below it. A low-velocity zone is also shown at the 8-km depth. The records of earthquakes (05/25/2002, 06/23/2003) occurred in the Alaska area are stacked. A low-velocity zone appears at the depth of 8 km. P wave velocity increases gradually from the depth of 30 km to 34 km and it reaches 7.8 km/s. This is thought to be the Moho. Polarities of the transverse receiver function can indicate the orientation of a dipping interface (Langston, 1979). Transverse receiver functions are examined to identify the possible dipping Moho under GKP. Amplitudes of transverse receiver functions are too small to indicate the dipping Moho. The records of earthquakes occurred in the areas of Sumatra, the Solomon Islands, and the Aleutian Islands areas are analyzed to estimate the velocity structure under GSU. Their back azimuths are 229°, 144°, 50°, respectively. The records of two earthquakes (12/26/2004, 05/14/2005) occurred in Sumatra area, three (03/11/2003, 06/12/ 2003, 10/17/2003) in the Solomon Islands area, and two (06/23/ 2003, 06/14/2005) in Aleutian Islands area are selected to construct the receiver functions. The receiver functions and the estimated crustal velocity structures are presented in Fig. 7. A distinctive lowvelocity zone at the 10-km depth appears commonly in all velocity profiles. The Moho is expected to be at the depth of 32 km, however, the velocity contrast is not so distinctive. Examination of the transverse receiver functions indicates the flat Moho under GSU. Records of three earthquakes (12/26/2004, 05/14/2005, 05/19/ 2005) occurred in the Sumatra area, three (03/25/2003, 04/17/2004, 11/11/2004) in the Timor region, two (10/17/2003, 04/11/2005) in
162
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
Fig. 5. Geologic map of the Cretaceous Gyeongsang Basin, southeastern Korea. Abbreviations in index map: NM = Nangrim Massif, PB = Pyeongnam Basin, IB = Imjingang Belt, GM = Gyeonggi Massif, OB-TB = Okchon and Taebacksan basins (or Okchon fold belt in Fig. 2 and OFB in Fig. 3), YM = Yeongnam Massif, and GB = Gyeongsang Basin.
the Solomon Islands area, and two (05/25/2002, 06/14/2005) in the Alaska area are used to estimate the velocity structures beneath HDB. Their back azimuths are 228°, 186°, 150°, and 48°, respectively. Estimated velocity structures and receiver functions are presented in Fig. 8. In two velocity profiles estimated by using earthquake data from the Sumatra and Alaska areas, the velocity increases abruptly from 7.2 to 7.7 km/s at the depth of 30 km. The estimated velocity
structure from receiver function constructed by earthquakes occurred in the Timor region, however, shows a velocity jump from 7.3 to 7.8 km/s at the 28-km depth. In the velocity structures estimated by using earthquakes in the Solomon Islands area, the velocity starts to increase gradually from the 28-km depth. The individual receiver functions before stacking are computed and presented in Fig. 9 in order to examine the dipping Moho under
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
163
Table 1 List of earthquakes for each station used in this work. Station code
Time (yyyy/mm/dd hh:mm:ss.ssss)
Latitude (deg.)
Longitude (deg.)
Magnitude (Mw)
Back azimuth (deg.)
Angular distance (deg.)
GKP
2002/05/25 05:36:32.5300 2002/12/12 08:30:44.9200 2003/06/07 00:32:45.5700 2003/06/23 12:12:31.5100 2004/10/08 08:27:53.5400 2004/12/26 00:58:53.4500 2005/04/28 14:07:33.7000 2005/05/14 05:05:18.4800 2003/03/11 07:27:32.0100 2003/06/12 08:59:20.2300 2003/06/23 12:12:31.5100 2003/10/17 10:19:09.1200 2004/12/26 00:58:53.4500 2005/05/14 05:05:18.4800 2005/06/14 17:10:17.0000 2002/05/25 05:36:32.5300 2003/03/25 02:53:26.6800 2003/10/17 10:19:09.1200 2004/04/17 15:58:24.6100 2004/11/11 21:26:41.1500 2004/12/26 00:58:53.4500 2005/04/11 12:20:05.9600 2005/05/14 05:05:18.4800 2005/05/19 01:54:52.8500 2005/06/14 17:10:17.0000 2002/02/05 13:27:26.0000 2003/01/20 08:43:53.0000 2003/11/17 06:43:00.0000 2004/12/26 00:58:53.4500 2005/06/14 17:10:17.0000 2005/07/05 01:52:04.0000
53.903 − 4.828 − 5.095 51.445 − 10.951 3.308 2.132 0.587 − 4.702 − 6.001 51.445 − 5.503 3.308 0.587 51.309 53.903 − 8.328 − 5.503 − 7.352 − 8.152 3.308 − 3.484 0.587 1.989 51.309 − 5.370 − 10.420 51.300 3.308 51.309 1.852
− 161.081 153.302 152.502 176.708 162.161 95.874 96.799 98.459 153.219 154.798 176.708 152.178 95.874 98.459 179.413 − 161.081 120.671 152.178 128.373 124.868 95.874 145.909 98.459 97.041 179.413 151.240 160.700 178.600 95.874 179.413 97.047
6.1 6.6 6.0 6.4 6.1 9.0 6.2 6.8 6.7 6.1 6.4 6.2 9.0 6.8 6.8 6.1 6.2 6.2 6.1 6.5 9.0 6.5 6.8 6.7 6.8 6.6 7.2 7.8 9.0 6.8 6.7
45.98 145.07 146.21 50.10 139.29 230.46 228.52 225.33 144.07 143.06 49.43 145.85 230.47 228.11 49.99 45.99 192.35 147.76 181.50 186.48 231.46 154.90 226.36 229.14 49.89 148.24 140.65 49.84 231.65 49.89 229.05
50.76 46.65 46.52 37.39 56.33 44.27 44.67 44.92 46.15 48.04 38.75 46.36 43.48 43.84 39.83 50.42 44.59 46.26 42.86 43.85 44.67 41.96 45.27 45.02 39.16 45.45 54.36 38.65 44.24 39.16 44.63
GSU
HDB
BUS
HDB which has been suggested by Chang and Baag (2005). The receiver functions of the Sumatra Andaman Earthquake (the bottom one in Fig. 9) show the large sharp direct P phase and the relatively smaller PS phase compared to others in the radial component, and the little transverse component. This may come from lack of high frequency energy of very large earthquakes. Polarity difference is not
clear in the transverse components, either. In the radial components, the amplitudes and arrival times of PS phases significantly vary, which has been suggested by Cassidy (1992) as an indication of dipping layers. The PS phases of the Alaska area earthquakes (top two in Fig. 9) arrive earlier than those of the Sumatra area earthquakes (bottom two excluding the Sumatra Andaman event). The former shows PS
Fig. 6. Estimated velocity structures beneath GKP and corresponding stacked receiver functions from earthquakes occurred in the Sumatra area (left), the Solomon Islands area (middle), and the Alaska area (right). The receiver functions synthesized from the estimated velocity structures (red dash-dot line) are compared with the observed ones (black solid line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
164
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
Fig. 7. Estimated velocity structures beneath GSU and corresponding stacked receiver functions from earthquakes occurred in the Sumatra area (left), the Solomon Islands area (middle), and the Aleutian Islands area (right). The receiver functions synthesized from the estimated velocity structures (red dash-dot line) are compared with the observed ones (black solid line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
amplitudes much smaller than those of the latter. They are located almost symmetrically around HDB. This suggests the southwestward dipping Moho interface. If the polarity difference in the transverses components is investigated with this information, the polarities of direct P and PS agree reasonably with the synthetic results of dipping interfaces by Cassidy (1992) — negative direct P and positive PS polarities of the Solomon Islands area (third and forth in Fig. 9) and the Timor region (fifth to seventh) earthquakes, and large direct P
amplitudes of the Solomon Islands area events which is located almost in the strike direction. Records of two (12/26/2004, 07/05/2005) earthquakes occurred in the Sumatra area, two (01/20/2003, 02/05/2003) in the Solomon Islands area, and two (11/17/2003, 06/14/2005) in the Aleutian Islands area are used to construct receiver functions for BUS. Their back azimuths are 230°, 144°, and 50°, respectively. The estimated velocity models and receiver functions are presented in Fig. 10. A sharp
Fig. 8. Estimated velocity structures beneath HDB and corresponding stacked receiver functions from earthquakes occurred in the Sumatra area, the Timor region, the Solomon Islands area, and the Alaska area (left to right). The receiver functions synthesized from the estimated velocity structures (red dash-dot line) are compared with the observed ones (black solid line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
165
Fig. 9. Individual radial (left) and transverse (right) receiver functions of HDB with event dates and back azimuths. From top to bottom, uppermost two are constructed from the earthquakes occurred in the Alaska area, upper two are from those in the Solomon Islands area, middle three are from those in the Timor region, and lower three are from those in the Sumatra area. All the receiver functions are plotted at the same amplitude scale. Although polarity difference is not clear in the transverse components, the amplitudes and arrival times of PS phases in the radial components significantly vary, which indicate the dipping Moho.
velocity increase from 6.1 to 6.6 km/s appears at the depth of 10 km in two velocity profiles using data from the areas of Sumatra and Solomon Islands. The velocity increases sharply from 7.0 to 7.6 km/s at the 34-km depth in the velocity profile using data from the Aleutian Islands area, and the velocity increases from 7.2 to 7.6 km/s at the 36km depth in the velocity structure from data of the Sumatra area. In the velocity profile using data from the Solomon Islands, the velocity increases gradually from 6.6 to 7.5 km/s at the 32–36-km depth interval. The Moho is expected to be at the depths between 34 km and 36 km. It is the deepest Moho in the Gyeongsang Basin analyzed in this work.
6. Discussion The crustal velocity structures of the Gyeongsang Basin have been estimated by analyzing receiver functions. To facilitate discussion, the composite velocity structures were constructed by averaging the velocity structures estimated from individual stacked receiver functions under each station. They are presented in Fig. 11. The 1D velocity structures of the locations nearest to GKP and HDB are extracted from the KCRT2002 velocity cross-section (Fig. 2), and are also presented in Fig. 11. Those for GSU are extracted from the KCRT2004 velocity cross-section (Fig. 3), and are also presented in Fig. 11. The distinctive
166
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
Fig. 10. Estimated velocity structures beneath BUS and corresponding stacked receiver functions from earthquakes occurred in the Sumatra area (left), the Solomon Islands area (middle), and the Aleutian Islands area (right). The receiver functions synthesized from the estimated velocity structures (red dash-dot line) are compared with the observed ones (black solid line). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
features of Fig. 11 are the Moho depths under GKP and HDB, and the shallow low-velocity discontinuity under GSU. They agree with results of KCRT2002 pffiffiffi and KCRT2004. It is worthwhile to note that the assumed VP = VS = 3 is probably too low for the mantle. This partly explains the low mantle velocity of about 7.5 km/s in the present results. The followings come out as a result in the course of this work. 6.1. Velocity discontinuity and low-velocity anomaly All of the velocity profiles estimated in this work show some kind of weak or strong velocity discontinuities at about 10-km depths. Especially distinctive low-velocity anomalies appear at about 10-km depths in the estimated velocity structures beneath GKP and GSU (Fig. 11). The low-velocity zone under GSU is also shown in the velocity tomogram of KCRT2004 survey line (Figs. 3 and 11). The low-velocity
zone the velocity of which is about 5.8 km/s appears at around 10-km depths near the SSE end, and extends to a 35-km horizontal distance. GSU stands apart from the KCRT2004 survey line by less than 15 km (Fig. 1). Although further study related to these features is needed, a possible explanation for these shallow low-velocity anomalies may be continental margin volcanism (e.g., adakitic intrusion) in the basin, which is partly attributed to the upwelling flow in the upper mantle resulted from ridge subduction and subsequent slab window formation during the Late Cretaceous. The ridge subduction and subsequent slab window formation have been considered as principal causes of adakitic volcanism and low-velocity anomaly in active continental margin (Thorkelson and Breitsprecher, 2005). Thorkelson and Breitsprecher (2005) demonstrated a connection between subduction of young oceanic crusts and production of intermediate to felsic igneous rocks which bear the signature of a garnetifeours residuum. Such
Fig. 11. Composite velocity structures constructed by averaging the velocity structures estimated from analyses of stacked receiver functions for GKP, GSU, HDB, and BUS. The 1D velocity structures of the locations nearest to GKP and HDB are extracted from the KCRT2002 velocity cross-section (Fig. 2), and are presented (red dotted lines). Those for GSU are extracted from the KCRT2004 cross-section (Fig. 3), and are also presented (blue dotted line). Arrows indicate the Moho depths estimated in this work.
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
magmatic rocks are compositionally similar to Tertiary lavas on Adak Island in the Aleutian arc, which were identified as products of slab melting by Kay (1978). This petrologic family, termed adakites, was described by Drummond and Defant (1990) as high-alumina, intermediate to felsic volcanic rocks typically hosting phenocrysts of plagioclase, amphibole, mica, and orthophyroxene, and lacking phenocrysts of clinophyroxene. Recent petrochemical study on the granitoids within the Cretaceous Gyeongsang Basin indicates that adakitic volcanism locally has occurred in the basin. Petrogenesis of these granitoids is consistent with a slab window model in which parental magmas were derived by decompression melting of the suboceanic mantle (i.e., subducted paleo-Pacific Plate or Izanagi Plate) that upwelled into the opening that formed beneath continental margin following ridge subduction since Late Cretaceous (Wee et al., 2006). Adakitic rocks of Late Cretaceous ages have also been reported in the southwestern Japan, which are usually considered to be generated by the melting of a young and hot subducted oceanic crust, i.e. paleo-Pacific or Izanagi Plate (Takahasi et al., 2005). As well recent geophysical investigation along the northeastern Pacific margin indicates that a large low-velocity anomaly is found, which may reflect the upwelling mantle in the Pacific-Juan de Fuca slab window near the subducted edge of the Pacific Plate (Qi et al., 2007). 6.2. Moho depth and characteristics The Moho under GKP appears as a sharp velocity discontinuity from 6.8 to 7.4 km/s at the 30-km depth in velocity structures estimated by analyses of receiver functions constructed from events in the Sumatra and the Solomon areas. In contrast with this, it appears a gradual velocity increase from 6.7 to 7.9 km/s in the 30–36-km depth range in the velocity structure by the receiver function from events in the Alaska region (Fig. 6). No dipping evidence such as polarity reversals in the transverse receiver functions has been found. It appears at the 30-km depth as a sharp velocity discontinuity in the composite velocity structure, and the Moho depth around GKP in the KCRT2002 profile coincides with it (Fig. 11). In contrast with the 36-km deep Moho reported without definite conclusion by Chang and Baag (2005), a 30-km deep Moho appears to be reasonable under GKP. The Moho under GSU appears to be at the 32-km depth, which agrees with that in the velocity tomogram of the KCRT2004 survey line. All velocity structures under GSU from three back azimuths show a similar trend. No polarity reversal has been found in the transverse receiver functions. It is reasonable to conclude that the Moho under GSU is flat. The Moho depths under HDB from individual receiver functions are 28– 30 km and it appears at the 28-km depth in the composite velocity structure. It agrees with that in the KCRT2002 velocity cross-section (Fig. 11). It is the shallowest one among those analyzed in this work. As discussed in the previous section, the southwestward dipping Moho is suggested in contrast with westward dipping one by Chang and Baag (2005). The Moho depth under BUS is estimated to be 34–36 km, and is the deepest among those analyzed in this work. The Moho depths vary as less than 2 km in the velocity profiles from receiver functions with different back azimuths, and polarity reversals have not been found in the transverse receiver functions. It is reasonable to conclude that the Moho under BUS is flat. All of these results will contribute to mapping the Moho depths in the southeast Korean Peninsula, and it is expected that more realistic Moho depth distribution which reconciles to the results of KCRT seismic profiles will come out in the near future. 7. Conclusions Crustal velocity structures beneath the broadband seismic stations in the Gyeongsang Basin, GKP, GSU, HDB, and BUS, are estimated by using the joint inversion of receiver functions and surface-wave dispersions. Recent earthquake data with good signal-to-noise ratio are included and back azimuth coverage is extended. The former
167
improves resolution in the estimated velocity structures, and the latter does the confidence level in the interpretation. The estimated Moho depths are 30 km, 32 km, 28 km and 34–36 km under GKP, GSU, HDB, and BUS, respectively. Moho depths under GKP, GSU, and HDB harmonize with those from nearby velocity cross-sections obtained from crustal-scale seismic profiles. The estimated Moho depth under BUS appears to be reasonable although no other data to verify it are available at present. Some kind of mid-crustal velocity discontinuity at the depth of around 10 km is commonly found. The distinctive 8–10-km deep low-velocity zone under GSU is also shown in the velocity tomogram of the KCRT2004 seismic profile, and the partial melting of a young and hot subducted oceanic crust in the Cretaceous is a possible geologic interpretation of this low-velocity zone. The southwestward dipping Moho under HDB is identified from the variation of amplitudes, arrival times, and polarities of PS in the individual receiver functions. Acknowledgements The authors thank the anonymous reviewers and the guest editors, Professors. T. Ito, H. Thybo, and T. Iwasaki, for their efforts and constructive comments. This work was funded by the Korea Meteorological Administration Research and Development Program under Grant CATER 2006-5206. References Ammon, C.J., 1991. The isolation of receiver effects from teleseismic P waveforms. Bull. Seism. Soc. Am. 81, 2504–2510. Ammon, C.J., Randall, G., Zandt, G., 1990. On the nonuniqueness of receiver function inversions. J. Geophys. Res. 95, 15303–15318. Berteussen, K.A., 1977. Moho depth determinations based on spectral ratio analysis of NORSAR long-period P waves. Phys. Earth Planet. Inter. 15, 13–27. Cassidy, J.F., 1992. Numerical experiments in broadband receiver function analysis. Bull. Seism. Soc. Am. 82, 1453–1474. Chang, K.H., 1975. Cretaceous stratigraphy of southeast Korea. J. Geol. Soc. Korea 11, 1–23. Chang, S.J., Baag, C.E., 2005. Crustal structure in Southern Korea from joint analysis of teleseismic receiver functions and surface-wave dispersion. Bull. Seism. Soc. Am. 95, 1516–1534. Chang, S.J., Baag, C.E., Langston, C.A., 2004. Joint analysis of teleseismic receiver functions and surface wave dispersion using genetic algorithm. Bull. Seism. Soc. Am. 94, 691–704. Cho, H.M., Baag, C.E., Lee, J.M., Moon, W.M., Jung, H., Kim, K.Y., Asudeh, I., 2006. Crustal velocity structure across the southern Korean Peninsula from seismic refraction survey. Geophys. Res. Lett. 33, L06307. doi:10.1029/2005GL025145. Chough, S.K., Kwon, S.T., Ree, J.H., Choi, D.K., 2000. Tectonic and sedimentary evolution of the Korean peninsula: a review and new view. Earth Sci. Rev. 52, 175–235. Dey-Sarkar, S.K., Wiggins, R., 1976. Upper mantle structure in western Canada. J. Geophys. Res. 81. doi:10.1029/OJGREA000081000320003619000001. Drummond, M.S., Defant, M.J., 1990. A model for trondhjemite–tonalite–dacite genesis and crustal growth via slab melting; archean to modern comparisons. J. Geophys. Res. 95, 21503–21521. Helmberger, D., Wiggins, R.A., 1971. Upper mantle structure of Midwestern United States. J. Geophys. Res. 76, 3229–3245. Jin, M.S., 1985. Geochemistry of the Cretaceous to early Tertiary granitic rocks in southern Korea: part I, major elements geochemistry. J. Geol. Soc. Korea 21, 297–316. Julià, J., Ammon, C.J., Herrmann, R.B., Correig, A.M., 2000. Joint inversion of receiver function and surface wave dispersion observations. Geophys. J. Int. 143, 99–112. Kay, R.W., 1978. Aleutian magnesian andesites; melts from subducted Pacific Ocean crust. J. Volcanol. Geotherm. Res. 4, 117–132. Kim, S.K., 1995. A study on the crustal structure of the Korean Peninsula. J. Geol. Soc. Korea 31, 393–403. Kim, S.J., Kim, S.G., 1983. A study on the crustal structure of South Korea by using seismic waves. J. Korean Inst. Mining Geol. 16, 51–61. Kim, K.Y., Lee, J.M., Moon, W., Baag, C.E., Jung, H., Hong, M.H., 2007. Crustal structure of the Korean Peninsula from seismic waves generated by large explosions in 2002 and 2004. Pure Appl. Geophys. 164, 97–113. Langston, C.A., 1979. Structure under Mount Rainier, Washington, inferred from teleseismic body waves. J. Geophys. Res. 84, 4749–4762. Last, R.J., Nyblade, A.A., Langston, C.A., 1997. Crustal structure of the east African plateau from receiver functions and Rayleigh wave phase velocities. J. Geophys. Res. 102, 24469–24483. Lee, K.,1979. On crustal structure of the Korean Peninsula. J. Geol. Soc. Korea 15, 253–258. Maruyama, S., Isozaki, Y., Kimura, G., Terabayashi, M., 1997. Paleo-geographic maps of the Japanese islands. plate tectonic synthesis from 750 Ma to the present. Island Arc. 6, 121–142.
168
S.J. Park et al. / Tectonophysics 472 (2009) 158–168
Owens, T. J., 1984. Determination of crustal and upper mantle structure from analysis of broadband teleseismic P-waveforms, Ph.D. Thesis, University of Utah, Salt Lake City, 146. Owens, T.J., Zandt, G., Taylor, S.R., 1984. Seismic evidence for an ancient rift beneath the Cumberland Plateau, Tennessee: a detailed analysis of broadband teleseismic P waveforms. J. Geophys. Res. 89, 7783–7795. Qi, C., Zhao, D., Chen, Y., 2007. Search for deep slab segments under Alaska. Phys. Earth Planet. Inter. 165, 68–82. Syswerda, G., 1989. Uniform crossover in genetic algorithms. In: Schaffer, J (Ed.), Proceedings of the Third International Conference on Genetic Algorithms. Morgan Kaufmann Publishers, Los Altos, CA, pp. 2–9.
Takahasi, Y., Kagashima, S.-I., Mikoshiba, M., 2005. Geochemistry of adakitic quartz diorite in the Yamizo Mountains, central Japan: implications for Early Cretaceous adakitic magmatism in the inner zone of southwest Japan. Island Arc 14, 150–164. Thorkelson, D.J., Katrin Breitsprecher, K., 2005. Partial melting of slab window margins: genesis of adakitic and non-adakitic magmas. Lithos 79, 25–41. Uyeda, S., Miyashiro, A., 1974. Plate tectonics and the Japanese Islands: a synthesis. Bull. Geol. Soc. Am. 85, 1159–1170. Wee, S.M., Choi, S.G., Ryu, I.C., Shin, H.J., 2006. Geochemical characteristics of the Cretaceous Jindong granites in the southeastern part of the Gyeongsang Basin, Korea: focused on adakitic signatures. J. Korea Soc. Econ. Environ. Geol. 39, 555–566.