3-D Seismic Image Beneath Hakone Volcano, Japan

Transcription

1 3-D Seismic Image Beneath Hakone Volcano, Japan Yoshiya Oda 1), Shintaro Abe 2), Yasuhira Aoyagi 2) and Takahiro Iwatate 1) 1) Dept. of Civil Engineering, Tokyo Metropolitan University, Tokyo, , Japan. 2) Geosphere Environmental Science Dept., Central Research Institute of Electric Power Industry, Chiba, , Japan. Abstract: Hakone volcano is located in the northern part of Izu peninsula. It is well known as inactive volcano and hot springs resort. But sometimes there are earthquake swarms. There was big earthquake swarm in April Total number of earthquakes was 15,816 during April to October. The cause of this activity is not yet solved. Explosion observation was carried out and 3-D velocity structure was estimated using explosions and micro earthquakes hybrid inversion method to know where magma exists or not. Not only micro earthquakes but also explosions were used, so the high resolution subsurface image and velocity value can be obtained with good accuracy. As results, we can find some faults as lines of hypocenter beneath Hakone volcano, and there is no results that suggests existence of magma. So it is considered that the cause of swarm earthquakes activity that occurred on 2001, was not activity of magma. 1 Introduction Hakone volcano is located in the northern part of Izu penninsula (figure1). It is well known as inactive volcano and hot springs resort. But sometimes there are earthquake swarms. There was big earthquake swarm in April Total number of earthquakes was 15,816 during April to October. The cause of this activity is not yet solved. Explosion observation was carried out and 3-D velocity structure was estimated using explosions and micro earthquakes hybrid inversion method to know where the magma exists or not. Not only micro earthquake but also explosions were used, so high resolution subsurface image and velocity value can be obtain with good accuracy. 2 Method More deeper 3-D image can be estimated by use of earthquakes than explosions. Seismic inversion method using earthquake has following problems. These problems make the solution low resolution and low accuracy. 1. Ill-posed problem: there are many unknown parameters to solve, not only velocity model but also source parameter 2. Data acquisiton problem: We can not control hypocenter location Hybrid inversion method was developed by modifying SIMULPS12 (Evans et al., 1994). SIMULPS12 is one of the most popular inversion program in the world, it can read not only earthquake data but also explosion data and blast data. But in this program, size of horizontal grid can not be changed at each depth, so we can get lower resolution model even if we use many explotions. SIMULPS12 was modified to be able to change horizontal grid size at each depth.

2 Fig. 1 Location of Hakone volcano Comparison of results of checker board test between SIMULPS and modified SIMULPS12 is shown in figure2. In this figure high resolution image is obtained by use of modified SIMULPS12. 3 Data and initial model Following data is used in this study, 1. Explosions: The experiment was carried out on 2/26/2000-2/28/2000 as a cooperative research with Central Research Institute of Electric Power Industry (CRIEPI) and Hot Springs Research Institute of Kanagawa Prefecture (HSRI). 14 explosions was observed by 15 stationary stations and 230 temporary stations. Locations of explosions and seismic stations are shown in figure3. 2. Earthquakes: 6,681 earthquakes was observed by HSRI for the period from1989 to Initial model has two kind of horizontal grid, upper and lower. The grid size of upper model is 2km x 2km x 0.5km and initial velocity was made according to 1-D analysis using explosions. The size of lower model is 4km x 4km x 4km and initial velocity was made according to 1-D analysis using earthquakes. Initial model is shown in table D image Map view of 3-D P wave velocity image is shown figure5.

3 Fig. 2 Comparison SIMULPS12 (left) and modified SIMULPS12 (right) Fig. 3 Locations of explosions and seismic stations

4 35 30' ' 35 15' Depth(km) km Depth(km) Fig. 4 Hypocenter distribution of earthquakes located by HSRI for the period from 1989 to Table 1 Initial 1-D velocity model Depth (km) Vp (km/s) Vp/Vs

5 Depth=-0.5km Depth=4.0km Depth=0.0km Depth=8.0km Depth=0.5km Depth=12.0km Depth=1.0km Vp(km/s) Fig. 5 Map view of 3-D P wave velocity image. Relocated hypocenters are also plotted in each depth.

6 5 Discussion Hybrid inversion program was developed by modifying SIMULPS12. Explosion observation was carried out in Hakone volcano and modified SIMULPS12 was applied to the data. Low velocity anomaly was find out 8km depth beneath Hakone volcano, and the cause of this anomaly is not the magma but hydrothermal because Vp/Vs velocity ratio is not high. Two groups of events are find out beneath Hakone volcano. One group that locate south-west of caldera 4km depth, leads to Northern izu fault system and almost all events of this group are aftershocks of the 1994 outer rim of Hakone volcano earthquake(mj4.8). Events of another group locate from Mt.Komagatake to Mt.Kintoki and to Kojiri (two faults), and depth is around 2km. Almost all earthquake beneath Hakone volcano except aftershock of the 1994 outer rim of Hakone volcano earthquake, occurred in this area. Earthquake swarm on 2001 also took place in this area. It is thought that this faults are with the low degree adherence. 6 Conclusions Hybrid inversion program was developed to be able to change horizontal grid size at each depth and applied to explosions and earthquake data of Hakone volcano. The results of this study is following, 1. There is no results that suggests existence of magma beneath Hakone volcano. 2. We can find some faults as groups of events beneath Hakone volcano. 3. The cause of earthquake swarm on 2001, was not activity of magma (hydrothermal activity?).