Bull. Earthq. Res. Inst. Univ. Tokyo Vol. 2.,**3 pp. ++/ +,/ MeSO-net +, - - + - + * -. +, -. Estimation of the Azimuth of MeSO-net Borehole Seismometers Based on Long-period Seismic Ground Motion +, - - + Katsuhiko Shiomi *, Shunji Sasaki, Shin ichi Sakai, Keiji Kasahara, Syutaro Sekine, Shigeki Nakagawa, Kazushige Obara, Naoshi Hirata +, -. - + -. National Research Institute for Earth Science and Disaster Prevention Association for the Development of Earthquake Prediction Earthquake Research Institute, University of Tokyo Hot Springs Research Institute of Kanagawa Prefecture and Toshikazu Tanada Abstract A remarkable high-density seismograph network, called MeSO-net, is under construction in the Tokyo metropolitan region, central Japan. At the MeSO-net station, a three-component accelerograph is installed at the bottom of a,*-m deep borehole. To analyze horizontal component waveforms observed at a borehole-type seismic station, we first have to know the azimuth of the seismometer. In this study, based on cross-correlation analyses of long wavelength seismograms with a high S/N, we estimate sensor azimuths at MeSO-net stations. As reference stations, we use seismograms observed by NIED F-net broad-band seismographs and Hi-net tiltmeters. At more than 2* of stations, N-components of seismographs are within +* from magnetic north. We confirm that the sensors at three stations are rotated more than 3*. Although our results agree well with results estimated using P-wave first motion of a local earthquake, there are systematical azimuthal di# erences between two results at stations distributed in the northern Kanto region. Because the subducting Pacific slab lies beneath the Kanto region, this feature may indicate that the di# racted wave passing through the high-velocity slab was observed as a first pulse at stations located in the northeastern Kanto region. Key words : MeSO-net, NIED F-net/Hi-net, sensor azimuth, long wavelength seismogram, crosscorrelation + Ishida +33, Nakajima et al.,**3,**. * e-mail: shiomi@bosai.go.jp -*/ ***0 - + 115
MeSO-net,**1 MeSO-net,**3,**3,**3 +** km +32, Okada et al.,**.,* km,**3 MeSO-net / km,**3,**3,* m - F-net,**3.** Hi-net MeSO-net,**- Hi-net F-net, radial S MeSOnet N,**- Hi-net MeSO-net MeSO-net MeSO-net,**3 SN +32-,**+, MeSO-net MeSO-net M 1.. MeSO-net,**3,**2 1,**2 3 ++ Mw 0. 2 Table + P MeSO-net,**- Hi-net +** s,** s N MeSO-net,* 3 SN,**3,**3 F-net Hi-net 116
Table +. Earthquakes used in this study. Fig. +. Distribution of stations used in this study. Gray and black triangles show the locations of MeSO-net stations installed in FY,**1 and FY,**2, respectively. Open diamonds represent the locations of NIED Hi-net/F-net stations. Region where the distance from MeSO-net stations is,, km or more is shaded.,**- S Hi-net,/ s -./ kms 22 km,**- /* s, +. +** s,, km Fig. + MeSO-net Hi-net,,/ s F-net Table, 117
Table,. Information on reference stations. F-net Table + /* s +** s SN Hi-net,**2 /, Table + U+ N KOTH E YYIM F-net Hi-net Fig., NS EW MeSO-net Hi-net - N q, /* s +** s,**- MeSO- MeSO-net net, E N 118
Fig.,. Examples of observed waveforms of,**2 Eastern Sichuan earthquake (U+ in Table + ). (a) Accelerograms recorded by a tiltmeter at the N.KOTH station. At this station, N- and E-components of the tiltmeter are oriented to N.3 E, N+-3E, respectively. (b) Displacement waveforms integrated from waveforms shown in (a). Band-pass filter from /* s to +** s is applied. (c) NS and EW component displacement waveforms. These waveforms are calculated from a.3 anticlockwise rotation of waveforms shown in (b). (d) Accelerograms recorded at the E.YYIM station. (e) Displacement waveforms calculated using waveforms shown in (d). 3* + q * d i + E. YYIM Fig. -a N. TKRH * +13 * +2* N. KOTH +. j i Fig. -b Fig. - b, E. YYIM * N N-, E N-/2E i, j C * Fig. -a MeSO-, net E YYIM Hi-net N. TKRH C * C MeSO-net,**2 A C Table + N+ * 1** km,, km, q i, j C * w i, j i, j i, j i, j i, j max *,. cos djp.* *2. *2., wij Aij *0/. *-/. Aij *0/. * Aij *0/. i, j j.,**2 MeSO-net +12 N Table - Fig.. +* SN 119
Fig. -. (a) Maximum cross-correlation coe$ cients C ij (*) between displacement waveforms observed at E. YYIM and N. TKRH. Line colors and types are related to earthquakes listed in Table +. (b) Averaged cross-correlation coe$ cients for E. YYIM. Fig... Estimated azimuths of the MeSO-net borehole sensors. Each arrow indicates the direction of the N-component of a sensor. Diamonds and circles represent locations of reference stations (NIED Hi-net/F-net) and MeSO-net stations, respectively. 120
Table -. Estimated azimuths of MeSO-net borehole sensors. 121
Fig. 0. Distribution of N-component sensor azimuthal di# erences estimated by this study and Sasaki et al. (,**3). P / s,**3., Table - Fig. / Fig. / Fig. /. N-component sensor azimuths of the MeSO-net. Solid and open arrows represent azimuths estimated Fig. / by this study and by Sasaki et al. (,**3), respectively.,**3,**3 MeSO-net N., Fig. /, N-1E 0 31 Fig. 0 Fig. 0 N / N-+, EN-, E 2 +*,**3 1 N-+1EN-E,* 1. +/ - 3* / / Fig. 0 Hi-net,**3 F-net 1 Fig. /,**3,**2 1,. MeSO-net, Mj 0. 2 ++* km i k MeSO-net P x i, 122
Fig. 1. Relative sensor azimuths di# erences x( i, k) z( i, k) between MeSO-net stations within a +* -km interval. Relative sensor azimuths x ( i, k) are estimated by a cross-correlation analysis between MeSO-net stations i and k. z ( i, k) are derived using absolute sensor azimuth at stations i and k estimated in this study (Table -). Circles on the map show the locations of stations we evaluated. k Table - MeSO-net, zi, k xi, kzi, k Fig. 1 Fig. 1 P MeSO-net, +* km Ishida, +33, Fig. 1 x i, kz i, k - MeSO-net,**3 x i, k z i, k Fig. 2,**3 N a,**- Hi-net b P Fig. 2a ab 123
P 0 MeSO-net Hi-net F-net MeSO-net N 2 N +* 3* - P,**3 MeSO-net Fig. 2. Schematic illustration of seismic wave di# raction e# ect on sensor azimuth estimation. (a) Star, dotted arrow, and thin arrows indicate epitively. Parameters a, b refer to back azimuth and center, seismic ray pass, and sensor azimuths, respec- azimuth of principal axis of P-wave first motion, respectively. Gray ellipse denotes particle motion of P-wave. (b) After rotating the sensor with ( ba ) anticlockwise (thin arrows) to put the azimuth of the principal axis of P-wave on the backazimuth (thick arrow). Fig. 2b N abe N ab P Nakajima et al.,**3 Uchida et al.,**3 MeSO-net Wessel and Smith +332 GMT USGS PDE +32,, -/.*+.,0.,**3 2..+ /0. Ishida, M., +33,, Geometry and relative motion of the Philippine Sea plate and Pacific plate beneath the Kanto- Tokai district, Japan, J. Geophys. Res., 31,.23 /+-, doi: +*. +*,3/ 3+ JB *,/01.,**. http ://www.jishin. go.jp/main/chousa/ *. aug_ sagami/index.htm,**3 +*,3,**3 124
MeSO-net 2. 1+ 22.,**3 MeSO-net,**+ CD- KiK-net ROM S,,+ -P**-., +32- +1 /..,+.,3. ++/ ++2.,**3,**2 1,.,**-Hi-netKiK- M 02. 2+ net, +20 +3/. /0 33 ++*.,**3,**3 Hi- F-net +*, 0+ S3 -S +2. net, 0+ S+ -S 2. Nakajima, J., F. Hirose, and A. Hasegawa (,**3), Seismotec- Uchida, N., A. Hasegawa, J. Nakajima, and T. Matsuzawa tonics beneath the Tokyo metropolitan area, Japan: (,**3), What controls interplate coupling?: Evidence for E# ect of slab-slab contact and overlap on seismicity, J. abrupt change in coupling across a border between two Geophys. Res., ++., B *2-*3, doi: +*. +*,3 /,**2 JB **0+*+. overlying plates in the NE Japan subduction zone, Okada, Y., K. Kasahara, S. Hori, K. Obara, S. Sekiguchi, H. Earth Planet. Sci. Lett.,,2-, +++ +,+, doi: +*. +*+0/j.epsl. Fujiwara, and A. Yamamoto,,**., Recent progress of,**3.*..**-. seismic observation networks in Japan Hi-net, F-net, Wessel, P. and W.H.F. Smith, +332, New improved version of K-NET and KiK-net, Earth Planets Space, /0, xv- the Generic Mapping Tools released, EOS Trans. Am. xxviii. Geophys. Union, 13, /13. (Received October -*,,**3) (Accepted November -*,,**3) 125