Ground Motion and Rupture Process of the 2004 Mid Niigata Prefecture Earthquake Obtained from Strong Motion Data of K-NET and KiK-net

　On 23 October 2004 at 17:56 (JST), the 2004 Mid Niigata Prefecture earthquake (MJMA 6.8; Japan Meteorological Agency) struck mid Niigata prefecture, central Japan. The mechanism of the earthquake is estimated as a reverse fault dipping to northwest by Hi-net (High-sensitivity seismograph network; Obara, 2002) and F-net (Full range seismograph network of Japan; Fukuyama et al., 1996) as shown in Fig. 1. We estimated the rupture process by the waveform inversion analysis. Three different velocity structure models are constructed to get appropriate Green's functions in the waveform inversion.

Fault model

We construct a fault model for our waveform inversion analysis in order to obtain the rupture process. The fault size is assumed to be 42 x 24 km to cover aftershock distribution within 24 hours after the main shock (Fig. 1). We adopt N211E and 52 degrees as the strike and dip angles, respectively, of the moment tensor solution of F-net (Fig. 1). Referring to the hypocenter 37.289N, 138.870E, 13.1 km determined by JMA unified catalog, we assume a rupture stating point of our rupture process model at 13.4 km below the epicenter.

Data

For the analysis, we used strong motion data from 9 stations of K-NET and KiK-net. Their locations are shown in Fig. 1. The observed acceleration records were integrated into velocity and bandpass filtered between 0.1 and 0.67 Hz. We inverted 10s of the S-wave portion from 1 s before the S-wave arrival. Set up for the inversion analysis is shown in a Table.

Velocity structures

In order to calculate suitable Green's functions, we construct velocity models replacing shallow part of a crust and upper mantle model proposed by Ukawa et al. (1984) according to geological information and borehole logging data. Underground structure of the source region is divided into two parts by Shibata-Koide Tectonic Line. The causative fault is estimated to occur along this tectonic line. In the southeast of the tectonic line, Pre-Neogene hard rocks outcrop and in the northwest side, there are deep sedimentary basins.

　For the hanging wall stations, we construct a velocity model with 6 km sedimentary basin based on the geological information and borehole logging data. For the stations southeast of the tectonic line, the foot wall stations in other words, we introduce sedimentary layers of several hundred meters based on geological information (e.g., Yanagisawa et al., 1986). Based on borehole logging data for a few dozens to hundred meters deep for each KiK-net and K-NET station, we modify the shallowest part of our velocity structure models.

　We construct another velocity structure model for NIG019 referring to Yamanaka et al. (2005). They modeled velocity structure beneath NIG019 from array observations of microtremors. The model for NIG019 has sedimentary basin with slower P- and S-wave velocities than those for the other hanging wall stations.

　Using them as initial models, we modified the structures by trial and error of forward modeling to fit the synthetic seismograms to observed ones for several aftershocks. The final models that we use for calculation of Green's functions are shown in Fig. 2. Classification of stations by the velocity model is shown in Fig. 1.

Results

　Figure 3 shows the total slip distribution and moment rate functions at each subfault obtained by the waveform inversion analysis. Figure 4 compares observed seismograms and synthetics for the estimated rupture process. The overall waveform matching is good at all the stations. We can see two major asperities; (a) around the hypocenter, and (b) the upper-middle of the fault plane. The largest slip of 3.8 m occurs at the asperity (a). Asperity (a) contributes to the waveforms observed at the north stations (e.g., NIG019, NIGH01). Especially, main pulses observed at NIG019 are generated by this largest slip. Waveforms observed at the foot wall stations, NIGH12 and NIGH15, are mainly generated from the asperity (b). An additional small asperity (c) at southwest of the hypocenter plays an important role to explain characteristic waveforms which are widely observed to the southwest of the source region. Solid lines under the seismograms observed at NIG012, NIG022, and NIGH11 indicate pulse from the small asperity.

　The time progression of the rupture is given in Fig. 5. Slip at the hypocenter is dominant for 3 seconds after rupture starting. A small asperity to the southwest of the hypocenter ruptures at 3.0 - 6.0 seconds after the rupture starting. The major rupture propagates toward east with the first-time-window propagation velocity of 2200 m/s. This is about 67 % of S-wave velocity at the hypocenter. The total seismic moment is 1.2×10**19 Nm which corresponds to Mw 6.7.

　Moment rate functions are different from each other according to depth of the asperity (Fig. 3). The moment rate functions of deep two asperities, (a) and (c), have short rise time and a sharp peak. Seismic waves radiated from these asperities largely contribute to the generation of the large, brief pulses at stations on the hanging wall. In contrast, the source time functions of the shallow asperity (b) have longer rise time, and caused ground motion dominant in low frequency. This rise time variation suggest that stress drop at the deeper asperity is larger than that of the shallower one. Our inversion result shows that the rupture propagates from the hypocenter mainly to the up-dip direction. This corresponds well with the fact that the centroid depth of moment tensor solution, 5 km deep determined by F-net, is shallower than the depth of hypocenter, 13 km, determined by Hi-net data.