We used the GSN broadband data downloaded from the IRIS DMC (http://www.iris.washington.edu/). We selected 25 teleseismic P and 21 SH body waveforms, which were first converted to displacement by removing the instrument response and then bandpass filtered from 2 sec to 330 sec. The time window used in this study is about 160 sec. We also included 23 long period Rayleigh waves and 17 long period Love waves into inversions. We defined the fault plane using the hypocenter location of the USGS (Lat.=46.616 deg.; Lon.=153.2240 deg. Depth = 26.7 km) and the moment tensor solution of the GLOABL CMT (http://www.globalcmt.org). The hypocenter depth and strike of fault plane has been slightly modified to match the trench axis. We constrained its rupture process using a finite fault inverse algorithm in wavelet domain (Ji et al, 2002, 2003).
We selected the low angle nodal plane (dip =14.89 deg., strike=220 deg.) as preferred fault plane based upon teleseismic body waves. Its dimension is 400 km (along strike) by 137.5 km, which is further divided into 220 subfaults (20 km by 12.5 km). We found that the along strike variation of fault slip could be well resolved using teleseismic body waves only but the downdip variation is poorly constrained. Fortunately, the resolution could be significantly improved using additional long period surface waves. The seismic moment release of this model is 3.9e+20 N.m using a 1D PREM model, slightly larger than that of GLOBAL CMT (3.3 e+20 N.m).
Caption: A big black arrow indicates the strike of the fault plane. The color shows the amplitude of dislocations and white arrows represent the motion of the hanging wall relative to the footwall. Contours show the rupture initiation time in sec and the red star indicates the hypocenter location.
Caption: Moment rate function.
Caption: The Data are shown in black and the synthetic seismograms are plotted in red. Both of them are aligned on the P or SH arrivals. The number at the end of each trace is the peak amplitude of the observation in micro-meter. The number above the beginning of each trace is the source azimuth and below is the epicentral distance.
Caption: Comparison of long period Rayleigh (UD) and Love (SH) waves (30 mHz to 60 mHz). The number at the end of each trace is the peak amplitude of the observation in millimeter.
Figure: Comparison of long period Rayleigh (UD) or Love (SH) waves (30 mHz to 60 mHz). The number at the end of each trace is the peak amplitude of the observation in millimeter.
Figure: Surface projection of the slip distribution superimposed on topography and bathymetric map ETOPO2. The red contours shows the slip distribution. The black line indicates the plate boundary. The white dots are background seismicity from 1964 to 2004 (Relocated ISC catalog, Engdahl et al, 1998). The red dots are aftershocks (NEIC USGS).
Ji, C., D.J. Wald, and D.V. Helmberger, Source description of the
1999 Hector Mine, California earthquake; Part I: Wavelet domain
inversion theory and resolution analysis, Bull. Seism. Soc. Am., Vol
92, No. 4. pp. 1192-1207, 2002.
Bassin, C., Laske, G. and Masters, G., The Current Limits of Resolution for Surface Wave Tomography in North America, EOS Trans AGU, 81, F897, 2000.
This work is supported by National Earthquake Information Center (NEIC) of United States Geological Survey. This web page is built and maintained by Dr. C. Ji at UCSB.