Shallow S-wave Seismic Reflection Survey around the K
Transcription
Shallow S-wave Seismic Reflection Survey around the K
Shallow S-wave Seismic Reflection Survey around the K-NET Anamizu Site A. Nozomi KOBAYASHI Engineering Department, Kyoto University B. Hiroyuki GOTO & D. Sumio SAWADA Disaster Prevention Research Institution, Kyoto University C. Koji YAMADA HANSHIN CONSULTANTS CO., LTD. ABSTRACT: Around the K-NET Anamizu site, some residences are damaged during the 2007 Noto Hanto Earthquake, Japan. Hayashi et al. [2007] reported relations between the residences damages and the subsurface structure estimated from surface-wave survey and so on. The subsoil boundary of estimated velocity structure is expected to be more distinct by applying shallow S-wave seismic reflection survey. We develop the seismic reflection survey system to reduce the survey costs by devising equipment and materials. The shallow S-wave seismic reflection survey delineated detailed structures of the surface layers shallower than 20m. In addition, we verified the structures by comparing to the theoretical refracted travel time. 1 INTRODUCTION Noto Hanto Earthquake (Mj = 6.9) occurred on March 25, 2007 at Ishikawa prefecture in Japan,.The hypocenter is located at 37.22° N, 136.69° N and 11 km depth. One person was killed and more than 300 people were injured. There were numerous slope failures and liquefaction. About 700 residences in Wazima city, Nanao city, Anamizu town were seriously damaged. Strong ground motion over 100 cm/s of peak ground velocity was observed at KNET Anamizu site, which is organized by NIED. However the damage distribution in Anamizu town was not homogeneous. It implies that the ground motion during the main shock was distributed inhomogeneously. Iwata et al. [2007] observed the aftershocks in Anamizu town on one day after the main shock. They reported that the amplification of S-wave at the downtown site of Anamizu town was remarkable respect to an outcrop site. Hayashi et al. [2007] carried out shallow subsurface structure surveys that are a surface-wave survey and a sounding test in Anamizu downtown. The results of survey suggested that the shallow subsurface structure with approximately 0-20 m depth varied significantly in a horizontal direction around K-NET Anamizu site (Fig. 1) and these complex structures are also distributed in the Anamizu downtown. Then, it is important to know the detail subsurface structure in order to discuss a relationship between the damages and the subsurface structure in Anamizu town. Figure 1. Subsurface structure around K-NET Anamizu site estimated from a surface-wave survey andsounding test (Hayashi et al. 2007) We perform shallow S-wave seismic reflection survey in order to estimate the detail boundary of velocity structure. A simple survey system is developed in order to reduce a survey cost. 2 SYSTEM OF SHALLOW S-WAVE SEISMIC REFLECTION SURVEY 2.1 Survey system The ground exploration using a seismic reflection survey has been developed for the sake of a discovery of oil well. In engineering seismology, the technique is applied to investigate subsurface structures, shapes of seismic and engineering basement. Table 1. Parameter sets of survey system S-wave source Wooden hammer Receivers 28 Hz horizontal geophones, 24 sets Data logger ES-3000(GEOMETRICS); 24-bit, 8 channels Stack 10 Shot interval (Line L) 1.0 m (Line S) 0.5 m Receiver interval (Line L) 1.0 m (Line S) 0.5 m Total profile length (Line L) 40 m (Line S) 15 m Figure 2. Schematic figure of the survey system Figure 3. Schematic figure of the First observation Photo 1. Survey scene Figure 4. Schematic figure of the second observation fallowing the first (Fig. 3) Photo 2. Mortar board to set a vertical-type sensor into horizontal direction Seismic body wave consists of P-wave and Swave. Conventional P-wave reflection surveys are powerful to detect a deep structure because the phases are not disturbed by a pure S-wave and translated S-waves. However, for relatively shallow structure, P-waves with a long wavelength are not well scattered and trapped. S-wave with a short wave length increases a resolution of the survey results. S-wave seismic sources had been used Dynamites and Hammers. Now the formers are applied to Omnipuise and Hydraulic Impacter. The latter are applied to Marthor by IFP and S-wave vibrator by Conoco and Large-scale S-wave reflection surveys are conducted. We performed S-wave reflection survey in order to identify a shallow complex structure as expected by Hayashi et al. [2007]. For the sake of reducing the survey casts, we develop a small-scale system, which can be performed only by two persons, as follows. S-waves are generated by a wooden hammer stroke on one side of a wooden plank placed perpendicular to the survey line (Photo 1). 10 times stack on each side enhances a reflected signal. 24 geophones with a natural frequency of 28 Hz observe a horizontal ground motion. Since the geophones con- Table 2. Process of the survey data 1. Setting Line Offset-distance: 1m 2. Pre-Filter Line L: 6 ~ 90 [Hz] Line S: 6 ~ 50 [Hz] 3. Amplitude Recovery 4. AGC Gate length: 180 [msec] 5. Deconvolution 6. Filter Same as “Pre-Filter” 7. AGC Gate length: 180 [msec] 8. Velocity Analysis 9. Residual Statics 10. Time Valiant Filter Figure 5. Survey layout and a location of K-NET site in Omachi West Children’s Park Figure 6. Soil condition gif at K-NET Anamizu [NIED KNET ISK005] tain only one component of vertical-type sensor. They are fixed on mortar boards to be placed parallel to the direction of the source plank (Photo 2). Observed analog data is converted to digital data on ES-3000 (Geometrics), which supports only 8 channel data (Table 1, Fig. 2). To use 24 receivers, three sets of observations with a common source point are conducted. To select 8 receivers belonging to one set, we use a switch box selecting the occupied receivers (Fig. 3). When one observation finished, the closest receiver is removed, and the source moves to the point that the removed receiver placed (Fig. 4). The observation is regarded as full 24 channels observation when we analyze a survey data. 2.2 Field survey The survey is performed in Omachi West Children’s Park on 22-23, November 2007. K-NET Anamziu site is located at the north-western corner in the park (Fig. 5). According to the boring data of K-NET Anamizu site, N-values of the surface layer (0m~2m) are about 7. The deeper layer (3m~15m) is Figure 7. Time-distance profile of Line S very soft because the most of N-values are 0. The bottom layer (15m~) becomes stiff (Fig. 6). Two survey lines, Line S and Line L, are placed. Line L is 40 m length from almost south to north. Line S is 15m length from almost west to east. The starting points of two lines are overlapped. Figure 8. Velocity Analysis of Line L in 0-20m Figure 9. Velocity Analysis of Line L in 20-40m Figure 10. Time-distance profile of Line L 3 DATA ANALYZING OF THE SURVEY 3.1 Data processing The common-mid-point (CMP) stack is applied to a data processing of the survey. The data were processed on PC through the steps listed in Table 2. Initial processing steps include line setting, pre-filter, amplitude recovery, automatic gain control (AGC), deconvolution and filter to enhance the signal and suppress coherent and random noise. Pre-filter trims a target frequency component by band pass filters, 690 Hz for Line L and 6-50 Hz for Line S. Amplitude recovery process recovers signal amplitudes absorbed in the structure by assuming geometrical attenuation and inelasticity. The gate length of AGC is set to be 180 msec. The data are analyzed in velocity analysis, NMO and residual statics. After CMP stack, the data sets are filtered by time valiant filter and migration in order to estimate a final profile. Time valiant filter cuts a high frequency component in the deep region. Figure 11. Depth-distance profile of Line L 3.2 The result of processing Hayashi et al. [2007] reported the survey result of surface waves only along Line L as shown in Fig. 1. First, we show a time-distance profile of Line S in Fig. 7. The horizontal axis is distance from the crossing point with Line L and the vertical axis is two-way time. Sources and receivers spaces are set to be 0.5 m for Line S. A river is located on an extension of the line. Emphasized phases are given between 0.0 s and 0.05 s. The image implies the existence of the clear structure boundary in a shallow region. Emphasized images in deeper region are not clear rather than the first boundary. For Line L, sources and receivers spaces were set to be 1.0 m and the source energy was the same as Line S. K-NET Anamizu site is located on the extension of the line. It has been already reported that the subsurface structure of Line L may vary in the horizontal direction. The characteristics are shown in velocity analysis (Fig. 8 and 9). Figure 8 shows the phase peaks for CMP of 0-20m and Figure 9 shows them for CMP of 20-40m. Layer velocities for 020m increases with depth, while the second layer velocity for 20-40m is slower than the first. Figure 10 shows a time-distance profile of Line L. The subsoil boundary between 0.0 s to 0.05 s is almost flat from 0-9 m. The boundary from 9 m to 40 m gradually decline. The declination of two-way time range is about 0.25 s from 0.15 s to 0.4 s. Figure 11 shows depth-distance profile of Line L. The depth-distance profile also shows a boundary change of the layer. The difference is almost 14 m. Comparing between images of Fig. 1 and Fig. 11, boundary shapes are almost identical, in 0-10 m and 30-40m. The shapes in 10-30 m are different. It may be related to the difference of the survey methods because the surface wave survey assumes moderate change of layer boundary. On the other hand, we apply migration process to estimate the boundary shape in analyzing the data of reflection survey in order to understand the detail slopes. 3.3 Comparison of first travel time The survey also observes waveforms in order to discuss a first travel time of Line L in 0-24 m. The observation was a set of line surveys including two sources at the ends of the target line that are 0 m and 24 m of Line L. To calculate a theoretical first travel time of S-waves, we construct a model with two layers based on the reflection survey as shown in Figure 11. The theoretical first travel times are compared to the observed waveforms. Figure 12 shows a comparison between the theoretical first travel time and the observed waveforms for the case of 0 m source, and Figure 13 shows one for the case of 24 m source. First travel time of theoretical calculation almost corresponds to the observed time that is recognized as the first phases. Figure 12. Comparison between theoretical first travel time and the observed waveforms for 0m case Figure 13. Comparison with theoretical first travel time and the observed waveforms for 24m case 4 CONCLUSION AND DISCUSSION We estimate a subsurface velocity structure by applying a shallow S-wave reflection survey around the K-NET Anamizu site. Distinct image with the complex ground structure is investigated. A simple survey system of two participants is applied by developing the equipments in order to save steps moving sources and receivers. Comparing with the results of the previous survey, the result of this survey was almost identical to that in the flat region of boundaries. The depth of the boundary rapidly changes. First arrival time of theoretical calculation related to the estimated structure corresponded to the observed travel time. We verified the steep structure change around the K-NET site. The similar structures are distributed in Anamizu town. The relation- ship between the complex structures and damages will be discussed in the future works. 5 REFERENCES Hayashi, K., Tamura, M., Hirade, T., Yu, S., Muraoka, M., Kikuti, Y. 2007. Site Investigation by Geophysical Exploration and Sounding Around the K-NET Anamizu Observation Station. PROGRAMME AND ABSTRACTS THE SEISMOLOGICAL SOCIETY OF JAPAN 2007, FALL MEETING A12-07: 11. Iwata, T., Asano, K., Kuriyama, M., Iwaki, A. 2008. NONLINEAR SITE RESPONSE CHARACTERISITCS OF KNET ISK005 STATION AND RELATION TO THE EARTHQUAKE DISASTER DUARING THE 2007 NOTO-HANTO EARTHQUAKE, CENTRAL JAPAN. The 14th World Conference on Earthquake Engineering October 12-17, 2008, Beijing, China The Social of Exploration Geophysicists of Japan. 1998. The Geophysical Exploration Hand Book: 93 National Research Institute for Earth Science and Disaster Prevention K-NET ISK005 Soil condition gif Vidale, J. 1988. Finite-Difference calculation of travel times, Bull. Seism. Soc. Am. 78: 2062-2076