Bulletin of the Seismological Society of America, Vol. 97, No. 6, pp. 2190–2197, December 2007, doi: 10.1785/0120070066
Ⓔ
Short Note A Study of the Regional Variation of Low-Frequency Q1 Lg around the Korean Peninsula by Tae Woong Chung*, Myung-Hyun Noh, Junk-Kyong Kim, Yun-Kyong Park, Hyun-Jae Yoo, and Jonathan M. Lees 1 We studied regional Q1 Lg at 1 Hz (Q0 ) around the Korean Peninsula based on broadband vertical component seismic records of six Incorporated Research Institutions for Seismology (IRIS) Global Seismographic Network stations and 19 Korean stations of the Korea Institute of Geoscience and Mineral Resources (KIGAM). Using 177 seismic events with magnitudes between 5.3 and 5.7 and depths less than 50 km, the reversed two station method was applied, and 94 high quality interstation paths were selected from 869 possible pairs. These results show high Q1 0 paths around the Sea of Japan (East Sea) reflecting the typical oceanic structure and low Q1 0 paths around northeastern China related to inactive seismicity. Assigning these path values into 193 cells around South Korea with a size of 1° by 1°, we observed that the 3 regional Q1 0 decreased gradually from east to west between 2 and 1 × 10 .
Abstract
Online Material: Table of earthquake-station pairs analyzed.
Introduction The M 7.0 Japanese coastal earthquake that occurred on 20 March 2005 (Fig. 1) severely shook the Korean Peninsula (hereafter, KP) and illustrated the potential hazard of large seismic events outside the KP. Because Lg waves are the most prominent phase observed for Korean events, the amplitude decrease of Lg waves with distance is very important from the viewpoint of seismic hazard. Lg can be thought of either as a sum of higher-mode surface waves in a group velocity window between 3.5 and 3:0 km s1 (e.g., Wang and Herrmann, 1988) or multiple supercritically reflected S waves in the crust (e.g., Kennett, 1986). Aside from geometrical spreading effects and reflection and transmission coefficients at discontinuities, Lg amplitude decrease is caused by internal friction and scattering, commonly described as the inverse of the quality factor, 1 Q1 Lg . QLg is sensitive to the shape and lateral variations of the crustal wave guide (e.g., Bostock and Kennett, 1990). Significantly high Q1 Lg is observed for rapid changes in crustal thickness (e.g., Fan and Lay, 1998) and for the thinned crust typical for oceanic crust (e.g., Zhang and Lay, 1995; Calvert et al., 2000), which has a characteristic thickness range of 5–8.5 km (White et al., 1992). Tectoni*
Present address: Lab of Seismology, Department of Geological Sciences, 209 Mitchell Hall, CB#3315, University of North Carolina, Chapel Hill, North Carolina 27599-3315.
cally active regions also show high Q1 Lg caused by the scattering of Lg energy from fractures within the crust (Aki, 1980). In addition, high temperature and fluid content in the crust may also account for higher Q1 Lg (e.g., Xie, 2002). Regional Q1 Lg around the KP received attention related to the issue of monitoring North Korea’s recent nuclear test because it is important to know where phases are attenuated when discriminating between explosions and earthquakes (e.g., Cong et al., 1996). Laterally heterogeneous Q1 Lg reflecting crustal variations and tectonic activity was clearly shown by an extensive tomographic study of the Eurasia continent (Xie et al., 2006). Q1 Lg at 1 Hz around the KP appeared to be relatively high due to tectonically active regions such as Japan and eastern China. The value for the KP was, however, less constrained because Xie et al. (2006) used only one South Korean station and there were no Lg data crossing the KP. Because digital network South Korean stations were not deployed nationwide until 1999, our initial efforts to study regional Q1 Lg (e.g., Chung and Lee, 2003) were inconclusive due to insufficient data. The reversed two station method (RTSM) (Chun et al., 1987) gave anomalously high Q1 Lg (Chung, 2002; Chung and Lee, 2002) due to short interstation paths (< 130 km) for South Korean stations. Relaxing the collinear restriction on the sources and receivers (Shih et al., 1994), the extended interstation paths (< 505 km)
2190
Short Note
2191
Figure 1.
Map of 19 KIGAM stations and six IRIS stations (the inset in the lower left-hand corner) connected by the RTSM paths used in this study with colors indicating the Q1 0 values estimated for the paths. A star represents the epicenter of the earthquake of 20 March 2005 (M 7.0).
gave a lower Q1 Lg (< 0:002) for 1–3 Hz in South Korea (Chung et al., 2005), typical low values for tectonically inactive regions. This earlier study, however, neglected regional variations such as potential high Q1 Lg in the coastal region of the Sea of Japan (East Sea). In the present study, we obtained more than 800 interstation paths greater than 150 km around the KP by combining data of South Korean stations and foreign stations from the Data Management Center of the Incorporated Research Institutions for Seismology (IRIS). Regional variations were resolved by the RTSM, which can obtain reliable results from a small data set compared to the previous study (Chung et al., 2005) where the colinearity requirement was relaxed.
Data and Q Measurements This study is based on vertical component broadband data sampled at 20 samples from the following sources (Fig. 1): six Global Seismographic Network (GSN) stations of the IRIS and 19 South Korean stations of the Korea Institute of Geoscience and Mineral Resources (KIGAM). One hundred seventy-four seismic event data from 1999–2006
with magnitudes between 5.3 and 5.7 and depths less than 50 km were retrieved from the Data Management Centers of IRIS and KIGAM (Table 1, and Ⓔ supplemental material in the electronic edition of BSSA). To exclude the influence of previous phases such as Sn, 700 Lg-signal seismograms were required to meet the condition that the root-meansquare (rms) amplitude for the velocity window 3:5–3:0 km= sec was larger than two times that of the 4:5–4:0 km=sec window (Fig. 2); few of these traverse long paths over the typical oceanic crust (Figs. 2 and 3). For suppression of the source and receiver effects, we used the RTSM for stations and events lying within 30° of azimuth from the great-circle path connecting the two station-event pairs. Dense coverage of station-event pairs is illustrated in Figure 3. Following Chun et al. (1987), the RTSM calculates the attenuation coefficient γ as Fd1;2 Fd2;1 d1;2 d2;1 0:5 2γΔ e ; d1;1 d2;2 Fd1;1 Fd2;2
(1)
where F denotes the spectral amplitude of the Lg wave, di;j is the epicentral distance from source i to station j, and Δ is
2192
Short Note
Table 1 Interstation Path Values (Fig. 1) Showing Relatively Small Errors (Standard Deviations) from 869 RTSM Pairs Q1 0 Station Pairs
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
MDJ HIA MAJ MAJ INC MAJ SSE SSE INC SND HDB HKU SND ULJ KSA HSB ULL SES HSB KWJ ULL SSE NPR NPR SES ULL SND SND GSU HDB JRB JRB JRB KWJ SEO KWJ KWJ SND NPR CHN KSA HDB HSB KWJ SES GKP ULJ NPR TJN SND ULJ GKP ULJ ULJ HDB HDB GKP GKP HDB JRB
BJT MDJ MDJ INC BJT SSE BJT HIA HIA SNU INC INC INC MDJ MDJ BRD INC INC INC HKU CHN MDJ GKP INC HIA KSA SEO MDJ HSB SEO HSB SEO SES SES BRD SEO HDB BRD HIA BJT BJT MAJ HIA INC MDJ INC INC BRD BRD CHN CHN SNU SEO SNU MDJ CHN HKU SEO SNU NPR
Number of Event Pairs
315 177 54 46 32 17 14 11 10 7 6 6 6 5 5 5 5 4 4 4 4 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
δQ1 0 (×103 )
0.78 1.23 2.98 2.50 1.38 2.08 2.45 1.79 1.44 1.27 0.81 1.83 0.14 2.32 3.27 0.91 1.88 0.42 0.54 2.38 2.98 2.13 1.15 1.11 1.76 3.56 3.84 2.49 1.30 2.97 1.80 1.77 1.24 2.44 1.41 3.05 0.58 2.09 1.49 0.56 1.46 2.88 1.79 2.23 2.00 1.19 0.93 2.72 2.28 0.73 0.49 1.42 2.40 3.43 2.43 3.58 0.57 1.31 2.23 2.27
0.03 0.05 0.07 0.08 0.09 0.13 0.15 0.17 0.19 0.47 0.21 0.32 0.12 0.26 0.26 0.24 0.36 0.12 0.15 0.31 0.33 0.34 0.67 0.36 0.36 0.29 0.42 0.30 0.51 0.39 0.20 0.20 0.19 0.36 1.00 0.34 0.91 0.50 0.44 0.39 0.47 0.37 0.44 0.96 0.41 0.36 0.42 0.67 0.54 0.33 0.33 0.40 0.61 0.94 0.41 0.49 0.26 0.35 0.49 0.22
η
δη
Δ (km)
0.48 1.01 0.50 0.47 0.50 0.51 0.67 0.73 0.95 4.50 0.48 1.60 0:41 0.87 0.81 0.27 2.11 0:92 0:62 1.00 1.25 0.84 4.94 1.07 1.05 0.41 1.67 0.95 2.64 1.31 0:36 0:36 0:47 1.07 10.15 0.91 6.17 1.56 1.05 0.72 1.30 0.52 1.07 7.01 0.87 0.49 0.87 1.34 1.87 0.24 0.28 0.76 2.58 6.60 0.82 1.44 0:29 0.39 1.45 0:58
0.06 0.08 0.04 0.05 0.10 0.09 0.10 0.17 0.27 2.62 0.40 0.52 0.66 0.22 0.15 0.33 0.70 0.18 0.19 0.28 0.27 0.31 4.45 0.73 0.45 0.12 0.34 0.25 1.74 0.34 0.09 0.09 0.12 0.33 10.94 0.23 14.60 0.70 0.66 1.26 0.82 0.20 0.55 4.57 0.41 0.47 0.88 0.64 0.79 0.57 0.88 0.52 1.11 2.74 0.32 0.38 0.39 0.38 0.61 0.07
1215 916 1153 1035 949 1683 1086 2025 1427 167 315 116 195 877 675 232 378 79 103 164 344 1669 158 162 1493 256 169 830 204 298 174 268 205 187 203 267 228 232 1584 963 1059 797 1524 260 907 250 262 288 203 193 267 228 237 233 985 347 138 237 290 116
(continued)
Short Note
2193 Table 1 (Continued) Q1 0 Station Pairs
61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94
JRB GKP GSU BRD INC KSA HKU NPR GSU KWJ GSU HDB JRB GKP HDB SNU ULJ HDB ULL BRD GSU JRB JRB KWJ GSU JRB KWJ JRB BGD ULL BGD JRB MAJ ULL
SNU MDJ MDJ INC SND MAJ SSE SSE MAJ MAJ BJT BJT MAJ BRD BRD BRD BRD GKP BJT MDJ INC INC HIA HIA BRD BRD BRD HKU HIA SNU INC MDJ SND SEO
Number of Event Pairs
2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
δQ1 0 (×103 )
1.67 2.04 1.75 2.41 0.14 2.99 0.88 0.81 1.90 2.08 1.45 1.29 2.54 1.98 1.89 2.05 2.55 2.14 2.07 1.95 1.13 1.85 1.49 1.39 1.40 1.69 1.33 1.96 1.56 3.34 1.67 1.56 6.60 2.35
0.28 0.44 0.46 1.31 0.12 0.52 0.50 0.51 0.49 0.53 0.58 0.58 0.45 0.73 0.61 1.21 0.71 0.47 0.61 0.83 0.50 0.31 0.48 0.56 0.45 0.34 0.36 0.26 0.63 0.79 0.83 0.47 0.55 0.64
η
0:16 1.04 1.18 6.65 0:41 0.54 0.40 0.48 0.38 0.56 0.81 0.83 0.18 1.69 0.99 5.58 3.66 0.24 1.01 2.28 0.39 0:60 0.32 0.71 1.14 0.32 0.46 0:83 1.08 2.01 2.30 0.25 0.67 1.17
δη
Δ (km)
0.15 0.47 0.63 5.52 0.66 0.28 0.81 0.96 0.36 0.40 0.76 0.87 0.22 1.14 0.69 5.08 2.25 0.27 0.64 1.66 0.62 0.12 0.43 0.71 0.76 0.27 0.40 0.09 0.91 0.84 1.96 0.38 0.14 0.65
256 971 1057 177 195 899 835 759 926 1024 1179 1253 965 417 486 288 439 74 1307 842 290 268 1695 1680 436 403 373 157 1774 349 369 1056 842 348
Figure 2. Typical seismograms of IRIS (a) and KIGAM (b) meeting the condition that the rms amplitude for the velocity window of 3:5–3:0 km=sec (black bar) was larger than two times that of the 4:5–4:0 km=sec window (gray bar). All seismograms were band-pass filtered between 0.05 and 5 Hz. (Continued)
2194
Short Note
Figure 2.
Figure 3.
Continued.
Map showing RTSM paths connecting stations and events used in this study.
Short Note
2195 We assumed a power-law frequency dependent model of η 1 1 at 1 Hz and η is its Q1 f Q1 Lg 0 f , where Q0 is Q power-law frequency dependence. After smoothing the results of (1) with a simple running average over a frequency η interval δf 0:234 Hz, we fit Q1 0 f models by leastsquares regression for 0.1–1.0 Hz. The fit analysis is illustrated in Figure 4 for three station pairs; multiple event pairs were used for each station pair. We selected 94 interstation path values (Fig. 1) showing relatively small errors (Table 1) from the total 869 RTSM pairs.
Result
Figure 4. Examples of observed γf values (gray dots) for each interstation path with least-squares regression (solid line) using η the model Q1 0 f . Multiple event pairs are used for each station pair (Table 1).
the interstation path length Δ d1;2 d1;1 or d2;1 d2;2. γ is related to the quality factor Q as γ
πf ; QU
where f is the frequency and U is the group velocity, assuming 3:5 km=sec.
The distinctive features in Figure 1 are high Q1 0 paths 1 around the East Sea and low Q0 paths around northeastern China. The continental crust of the KP, with a ∼30-km thick crust, varies significantly near the coastal margins of the East Sea and thins to a 8.5-km thick crust at deeper than 3000-m bathymetry (Hirata et al., 1992). The high Q1 0 values reflect thinned crust and active tectonics of the Japan Arc while the low Q1 0 values correlate with the passive tectonics of northeastern China. Estimates of frequency dependence (η) were, it should be noted, generally unstable and poorly constrained as compared to Q1 0 . This is inevitably due to the narrow frequency range used in this analysis, between 0.1 and 1 Hz (Xie et al., 2004). Values with small error were derived from event pair numbers larger than 10 (Table 1); all of these are greater than 0.5. To visualize the spatial variation of Q1 0 , the study area was divided into 193 1° × 1° cells of constant Q1 0 . The value in each cell was determined by averaging the value of all traversing paths within the cell; each path value was weighted by its inverse variance and proportional path length. The resulting model (Fig. 5) agrees generally with ray paths presented in Figure 1, showing high Q1 0 around the East Sea and low Q1 0 around northeastern China. In addition, Q1 0 values higher than 0.002 along the eastern KP gradually decrease to the west and form a low attenuation belt (∼0:001) trending north–south. The range of the Q1 0 value is supported by the previous study of Chung et al. (2005). Where tectonics are inactive, high attenuation estimates on the eastern KP are related to the rapid change of crustal thickness. While the coastal region is explained by transition from continental to oceanic crust, variations associated with the inner continental area of the eastern and northeastern KP appear to be caused by KP orogenies as indicated by the inset topographic map (Fig. 5). In southwestern China Q1 0 estimates (∼0:002) most likely reflect moderate tectonism as suggested by Xie et al. (2006).
Acknowledgments We would like to thank the IRIS-DMC for providing data used in this research. The research was supported by funds of CATER 2006-5104 provided by the Center for Atmosphere Sciences and Earthquake Research (CATER) of Korea.
2196
Short Note
1 Figure 5. LgQ1 0 map around the KP. Values for each 1° × 1° cell are the Q0 × 10; 000. Topographic map of the KP (the inset) represents elevations from 3969 to 2362 m.
References Aki, K. (1980). Scattering and attenuation of shear waves in the lithosphere, J. Geophys. Res. 85, 6496–6504. Bostock, M., and B. L. N. Kennett (1990). The effect of 3-D structure on Lg propagation patterns, Geophys. J. Int. 101, 355–365. Calvert, A., E. Sandvol, D. Seber, M. Barazangi, S. Roecker, T. Mourabit, F. Vidal, G. Alguacil, and N. Jabour (2000). Propagation of regional
seismic phases (Lg and Sn) and Pn velocity structure along the AfricaIberia plate boundary zone, Geophys. J. Int. 142, 384–408. Chun, K.-Y., G. F. West, R. J. Kokoski, and C. Samson (1987). A novel technique for measuring Lg attenuation—results from eastern Canada between 1 to 10 Hz, Bull. Seismol. Soc. Am. 77, 398–419. Chung, T. W. (2002). A study of Q1 Lg by the reversed two station method in the crust of central South Korea, J. Korean Geophys. Soc. 5, 211–218 (in Korean).
Short Note Chung, T. W., and K. Lee (2002). Attenuation of high-frequency Lg waves around the Yangsan Fault area, the southeast Korea, J. Korean Geophys. Soc. 5, 1–8. Chung, T. W., and K. Lee (2003). A study of high-frequency Q1 Lg in the crust of South Korea, Bull. Seismol. Soc. Am. 93, 1401–1406. Chung, T. W., Y. K. Park, I. B. Kang, and K. Lee (2005). Crustal Q1 Lg in South Korea using the source pair/receiver pair method, Bull. Seismol. Soc. Am. 95, 512–520. Cong, L., J. Xie, and B. J. Mitchell (1996). Excitation and propagation of Lg from earthquakes in central Asia with implications for explosion/earthquake discrimination, J. Geophys. Res. 101, 27,779–27,789. Fan, G.-W., and T. Lay (1998). Statistical analysis of irregular wave-guide influences on regional seismic discriminants in China, Bull. Seismol. Soc. Am. 76, 74–88. Hirata, N., B. Y. Karp, T. Yamaguchi, T. Kanazawa, K. Suyehiro, J. Kasahara, H. Shiobara, M. Shinohara, and H. Kinoshita (1992). Oceanic crust in the Japan Basin of the Japan Sea by the 1990 Japan-USSR expedition, Geophys. Res. Lett. 19, 2027–2030. Kennett, B. L. N. (1986). Lg waves and structural boundaries, Bull. Seismol. Soc. Am. 76, 1133–1141. Shih, X. R., K.-Y. Chun, and T. Zhu (1994). Attenuation of 1–6 s Lg waves in Eurasia, J. Geophys. Res. 99, 23,859–23,874. Wang, C.-Y., and R. B. Herrmann (1988). Synthesis of coda waves in layered medium, Pure. Appl. Geophys. 128, 7–42. White, R. S., D. McKenzie, and R. K. O’Nions (1992). Oceanic crustal thickness from seismic measurements and rare earth element inversions, J. Geophys. Res. 97, 19,683–19,715. Xie, J. (2002). LgQ in the eastern Tibetian Plateau, Bull. Seismol. Soc. Am. 92, 871–876. Xie, J., R. Gok, J. Ni, and Y. Aoki (2004). Lateral variations of crustal seismic attenuation along the INDEPTH profiles in Tibet from LgQ inversion, J. Geophys. Res. 109, B10308, doi 10.1029/ 2004JB002988. Xie, J., A. Wu, R. Liu, D. Schaff, Y. Liu, and J. Liang (2006). Tomographic regionalization of crustal Lg Q in eastern Eurasia, Geophys. Res. Lett. 33, L03315, doi 10.1029/2005GL024410.
2197 Zhang, T. R., and T. Lay (1995). Why the Lg phase does not traverse oceanic crust, Bull. Seismol. Soc. Am. 85, 1665–1678.
Sejong University Seoul 143-747, Korea
[email protected] (T.W.C.)
Korea Institute of Nuclear Safety Daejon 305-338, Korea (M.-H.N.)
Semyung University Chungbuk, 390-711, Korea (J.-K.K.)
Korea Institute of Geoscience and Mineral Resources Daejon 305-350, Korea (Y.-K.P.)
Korea Polar Research Institute KORDI, Incheon 406-840, Korea (H.-J.Y.)
University of North Carolina Chapel Hill, North Carolina 27599-3315 (J.M.L.) Manuscript received 15 March 2007
