Journal of Volcanology and Geothermal Research

Transcription

1 Journal of Volcanology and Geothermal Research 187 (2009) Contents lists available at ScienceDirect Journal of Volcanology and Geothermal Research journal homepage: Short Communication P-wave velocity structure beneath Asama Volcano, Japan, inferred from active source seismic experiment Yosuke Aoki a,, Minoru Takeo a, Hiroshi Aoyama b, Jun Fujimatsu c, Satoshi Matsumoto d, Hiroki Miyamachi e, Haruhisa Nakamichi f, Takahiro Ohkura g, Takao Ohminato a, Jun Oikawa a, Rie Tanada c, Tomoki Tsutsui h, Keigo Yamamoto i, Mare Yamamoto j, Hitoshi Yamasato k, Teruo Yamawaki l a Earthquake Research Institute, University of Tokyo, 1-1 Yayoi 1, Bunkyo-ku, Tokyo , Japan b Institute of Seismology and Volcanology, Hokkaido University, N10W8, Kita-ku, Sapporo, Hokkaido , Japan c Seismological and Volcanological Department, Japan Meteorological Agency, 3-4 Otemachi 1, Chiyoda-ku, Tokyo , Japan d Seismological and Volcanological Department, Kyushu University, Shinyama 2, Shimabara, Nagasaki , Japan e Department of Earth and Environmental Sciences, Kagoshima University, Koorimoto 1, Kagoshima, Kagoshima , Japan f Research Center for Seismology, Volcanology, and Disaster Mitigation, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, Aichi , Japan g Aso Volcanological Laboratory, Kyoto University, Kawayo, Minami-Aso, Kumamoto , Japan h Faculty of Engineering and Resource Science, Akita University, 1 Tegatagakuen-machi 1, Akita, Akita , Japan i Sakurajima Volcano Research Center, Disaster Prevention Research Institute, Kyoto University, Sakurajima-Yokoyama-cho, Kagoshima, Kagoshima , Japan j Department of Geophysics, Tohoku University, 3 Aoba 6, Aoba-ku, Sendai, Miyagi , Japan k Seismology and Volcanology Research Department, Meteorological Research Institute, 1 Nagamine 1, Tsukuba, Ibaraki , Japan l Volcanic Fluid Research Center, Tokyo Institute of Technology, Kusatsu, Kusatsu, Gumma , Japan article info abstract Article history: Received 1 October 2008 Accepted 1 September 2009 Available online 17 September 2009 Keywords: active source seismic experiment P-wave velocity structure dike intrusion magma pathway An active seismic survey with unprecedented density of seismometers found a high velocity zone to the west of the summit of Asama Volcano, Japan. The high velocity zone coincides with an area of magma-filled crack (dike) intrusion associated with the 2004 eruptions inferred from precise earthquake relocations and ground deformation modeling. It also coincides with an area of high resistivity surrounded by low resistivity, indicating that the solidification of magma due to repeating intrusions is responsible for the high velocity. This thus endorses the magma pathway previously speculated by seismic and geodetic observations. These findings demonstrate that dense seismic exploration combined with geophysical monitoring is an effective way to understand the dynamics of volcanic eruptions Elsevier B.V. All rights reserved. 1. Introduction Corresponding author. Tel.: ; fax: addresses: yaoki@eri.u-tokyo.ac.jp (Y. Aoki), takeo@eri.u-tokyo.ac.jp (M. Takeo), aoyama@uvo.sci.hokudai.ac.jp (H. Aoyama), j-fujimatsu@met.kishou.go.jp (J. Fujimatsu), matumoto@sevo.kyushu-u.ac.jp (S. Matsumoto), miya@sci.kagoshima-u.ac.jp (H. Miyamachi), nakamiti@nagoya-u.jp (H. Nakamichi), bonkura@aso.vgs.kyoto-u.ac.jp (T. Ohkura), takao@eri.u-tokyo.ac.jp (T. Ohminato), oikawa@eri.u-tokyo.ac.jp (J. Oikawa), r.tanada@met.kishou.go.jp (R. Tanada), tom@geophys.mine.akita-u.ac.jp (T. Tsutsui), yamamoto@svo.dpri.kyoto-u.ac.jp (K. Yamamoto), mare@zisin.geophys.tohoku.ac.jp (M. Yamamoto), yamasato@mri-jma.go.jp (H. Yamasato), yamawaki@ksvo.titech.ac.jp (T. Yamawaki). A volcanic eruption is an ejection of magma transported from depth. How magma is transported from depth to the surface is thus one of the fundamental questions to understand how a volcano works. One of the effective ways to address this question is to image the magma pathway through geophysical monitoring including 1) precise earthquake relocations, 2) modeling geodetic data, and 3) exploring subsurface structure because earthquakes and ground deformation are representations of magma intrusion (e.g., Morita et al., 2006). Furthermore, the subsurface structure, density in particular, of the host rock plays an important role in dynamics of magma transport (e.g., Rubin, 1995). Here we investigate shallow subsurface structure of Asama volcano, Japan, where there already exists a dense seismic and geodetic monitoring network (Aoki et al., 2005; Takeo et al., 2006), with active seismic sources and densely deployed temporal seismic network to gain more insights into magma transport beneath the volcano. We employed active sources to take advantage of known locations and origin time. Asama Volcano is one of the arc volcanoes associated with westward subduction of the Pacific plate (Fig. 1a). It is an andesitic volcano with explosive eruptions such as the ones with a Volcano Explosivity Index (Newhall and Self, 1982) of 5 in 1108 and 1783 (Simkin and Siebert, 1994). The volcano has been relatively active in the 20th century until 1960s with frequent eruptions with VEIs up to 3. The eruptions became infrequent since then except for moderate-sized (VEI=2) ones in 1973, 1982, 1983, and 2004 (Simkin and Siebert, 1994; Nakada et al., 2005) and minor ones in Even during its dormancy, various types of volcanic earthquakes have been observed just beneath the summit of the volcano (e.g., Aoyama and Takeo, 2001) which may reflect some magmatic activity there /$ see front matter 2009 Elsevier B.V. All rights reserved. doi: /j.jvolgeores

2 Y. Aoki et al. / Journal of Volcanology and Geothermal Research 187 (2009) related to subsurface structure. To meet this goal, we conducted a seismic exploration in October, 2006, with artificial sources and densely deployed seismometers by taking advantage of known locations and origin time of artificial sources. 2. Active seismic experiment The seismic exploration is conducted by five dynamite sources of kg and 464 temporary 2-Hz seismometers (Mark Products L22D)withanaveragespacingofapproximately150m(Fig. 1b). Table 1 describes information on the active sources. To delineate seismic velocity structure around an area of dike intrusion during the 2004 eruptions, we constructed two lines of seismometers striking roughly north south and east west, respectively (Fig. 1b), crossing the area of dike intrusion. In addition, several arrays are constructed around the edifice (Fig. 1b) to observe scattered waves likely from the edifice. We do not discuss the data from these arrays here. It will be discussed elsewhere. Each site recorded seismic waves by a 250 Hz sampling interval. The observed waveforms indicate that seismic waves not traversing the area of dike intrusion have the maximum energy at first arrivals, while those traversing the area of dike intrusion have different waveforms with more energy in later arrivals. For example, waveforms exerted from S1 with epicentral distances greater than 12 km (Fig. 2a) and those from S2 with epicentral distances greater than 8 km(fig. 2d) have more energy in later arrivals than first arrivals. This feature is probably due to the scattering of seismic waves at the area of diking where the seismic structure is likely to be quite inhomogeneous. We also note that seismic waves traversing the summit are subject to substantial scattering. For example, waveforms exerted eastward from S3 with epicentral distances greater than 5 km (Fig. 2e) and those from S4 with epicentral distances greater than 10 km (Fig. 2f) are subject to scattering, resulting in more energy in later arrivals than first arrivals. 3. P-wave velocity structure Fig. 1. (a) Tectonic setting around Asama Volcano. Plate boundary and active volcanoes are represented by solid lines and gray triangles, respectively. Asama Volcano is shown by a black triangle. PAC stands for the Pacific plate which subducts beneath Asama Volcano. Direction of plate convergence is also shown. (b) Spatial distribution of active sources (red stars) and seismometers (black dots). The summit of Asama Volcano is shown by a green triangle. Bouguer gravity anomaly with a reduction density of 2670 kg/m 3 (Geological Survey of Japan, 2000). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) Precisely relocated earthquakes and geodetic modeling from dense monitoring network in Asama enabled us to gain some insights into the indirect evidence for the magma pathway. Aoki et al. (2005) and Takeo et al. (2006) demonstrated that dike intrusion associated with the 2004 eruptions occurred ~4 km to the west of the summit at ~1.5 km below sea level approximately 6 weeks before the first eruption. They also suggested that the intruded magma may have propagated horizontally to right beneath the summit and then propagated vertically to make the surface. The motivation of this study is to gain more insights into the dynamics of magma transport beneath Asama Volcano by understanding how the magma pathway derived from geophysical monitoring is As a first step to delineate the shallow subsurface P-wave velocity structure of the volcano, we manually picked the first arrivals of each trace. First, each author independently inspected all the traces to pick the arrival times manually, then metto classify the first arrivals into five classes, A, B, C, D, and F, respectively, according to the accuracy of picking, where A, B, C, and D refer to the ones with picking errors of less than 0.01, , , and s, respectively, and F refers to the ones unable to pick first arrivals within 0.2 s. Of the five classes, we left the picks of classes A, B, and C for the further analysis, resulting in a total of 346 picks along the north south profile and 299 picks along the east west profile, respectively. Then, we constructed P-wave velocity models to be consistent with first arrivals in each trace first by a forward modeling (Zelt and Smith, 1992) and then applying a regularized inversion of the travel times with the initial model obtained through the previous forward modeling (Zelt and Barton, 1998). The forward modeling before the inversion is required in this case because a small number of shots (only three shots per profile; see Fig. 1b) does not allow convergence to the optimum Table 1 Locations, times, and sizes of explosions. Latitude and longitude are based on the WGS- 84 reference frame. Shot times are based on Japanese local time (GMT+9 hours). Shot Latitude ( N) Longitude ( E) Altitude (m) Depth (m) Date Time Charge (kg) S Oct. 13, :01: S Oct. 13, :07: S Oct. 13, :12: S Oct. 13, :17: S Oct. 13, :22:

3 274 Y. Aoki et al. / Journal of Volcanology and Geothermal Research 187 (2009) Fig. 2. Record sections of vertical seismograms along the north south profile ((a) for S1, (b) for S3, and (c) for S5, respectively) and east west profile ((d) for S2, (e) for S3, and (f) for S4, respectively). Each trace is normalized by its maximum amplitude and reduced by 6 km/s. The horizontal and vertical axes represent the reduced time from the shot time and the horizontal distance from each shot point, respectively. solution in the inversion from a bad initial model. Fig. 3 shows the initial P-wave velocity and the comparison between the observed and calculated travel times. It shows that the initial velocity model is already good in terms of the fit between observed and calculated travel times notwithstanding the starting model is spatially smooth enough not to produce an arbitrary result through an inversion. Fig. 4 shows the P-wave velocity structure for both profiles, in which Fig. 4a and b depict the comparison between observed and calculated travel times and P-wave velocity structure for the north south profile, respectively, and Fig. 4d and e depict those for the east west profile, respectively. Fig. 4a and d indicate that the calculated travel times fit well with the observed travel times for both profiles. Also Fig. 5 indicates that the root mean squares of the travel time residuals for north south and east west profiles are 0.07 and 0.10s, respectively. Fig. 4b clearly shows a high velocity zone around S3 (distance=0 km) at a depth of 2 km below sea level and shallower, the area of dike intrusion during the 2004 eruptions inferred from seismic and geodetic measurements (Aoki et al., 2005; Takeo et al., 2006). Fig. 4e shows a high velocity zone N3 km to the west of the summit; combining this with the velocity along the north south profile suggests that the high velocity zone strikes east west with its top deepening to the east. The north south extent of the high velocity zone is as narrow as approximately 5 km. We conclude that this is due to solidified magma resulting from repeating dike intrusions because the location of the high velocity zone roughly coincides with the area of diking associated with the 2004 eruptions. Note that the

4 Y. Aoki et al. / Journal of Volcanology and Geothermal Research 187 (2009) Fig. 3. (a) Comparison between the observed (black) and calculated (red) first arrival times from the initial velocity model for the north south profile. Error bars denote the 1σ confidence interval. (b) Initial P-wave velocity structure along the north south profile. (c)(d) Same as (a) and (b), respectively, except that these are for the east west profile. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) high velocity zone is not formed by a single diking, for example, in 2004; geodetic data shows that the dike intruded in 2004 is at less than 1 m thick (Aoki et al., 2005; Takeo et al., 2006), too thin to be imaged in this analysis if the high velocity zone was formed by a single diking in Also note that we did not put aprioriinformation that the velocity structure is the same at the intersection of north south and east west profiles, that is, horizontal distances of 0 km in both profiles (Fig. 4). We conclude that the velocity structure at the intersection has a reasonable agreement in the absence of aprioriinformation (Fig. 4b, d). Fig. 4c and f display the ray paths calculated by the bending method of Um and Thurber (1987) over the inferred P-wave velocity structure for north south and east west profiles, respectively. They Fig. 4. (a) Comparison between the observed (black) and calculated (red) first arrival times for the north south profile. Error bars denote the 1σ confidence interval. (b) P-wave velocity structure along the north south profile. (c) Ray paths along the north south profile calculated by the method of Um and Thurber (1987). (d)(e)(f) Same as (a), (b), and (c), respectively, except that these are for the east west profile. The approximate position of the intruded dike during the 2004 eruptions (Aoki et al., 2005; Takeo et al., 2006) is projected in (b) and (e). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

5 276 Y. Aoki et al. / Journal of Volcanology and Geothermal Research 187 (2009) Fig. 5. Histograms of the residual between observed and calculated travel times from the obtained P-wave velocity model for (a) north south and (b) east west profiles. Number of travel times (N) and root mean squares (rms) of both profiles are also shown. show that the ray coverage is good to a depth of at least 1 2 km below sea level for the north south profile (Fig. 4c) and 1 km below sea level for the east west profile, endorsing that a high seismic velocity at the area of inferred diking during the 2004 eruptions is a robust feature. 4. Discussion Active and passive seismic surveys found high velocity zones in many active volcanoes (e.g., Zollo et al., 1996, 1998; Okubo et al., 1997; Tanaka et al., 2002; Yamawaki et al., 2004). We interpret the high velocity zone as solidified magma due to past intrusions, as previous studies did. Here we would like to note that the high velocity zone we found is consistent with the regional geology (Aramaki, 1963) that Quaternary volcanoes distribute roughly east west, the eastern end and the youngest of which is Asama. Our result is also consistent with insights gained from other geophysical observations, i.e., precise earthquake locations, ground deformation, gravity anomaly, and electromagnetic structure. Fig. 1b indicates that the spatial distribution of the Bouguer gravity in the area has east west trending local maxima to the west of the summit, the area of high velocity, and the eastern end of which is at Asama. This saddle-shaped gravity high suggests that the high velocity zone is imaged by a high density zone, consistent with our hypothesis that the high velocity zone is due to the solidification of repeatedly intruded magma; note that when magma is cooled slowly, it will become fractureless, thus dense, rock with high seismic velocity. Electromagnetic data depicts that the area of the high velocity zone is less conductive than surrounding areas (Aizawa et al., 2008), also consistent with our hypothesis because slowly solidified magma has low permeability, resulting in high resistivity. Aizawa et al. (2008) attributed to this high resistivity, extending up to about 1 km above sea level, as solidified magma associated with eruptions of Kurofu (Fig. 6) which has been active until 24,000 years ago (Aramaki, 1963). Fig. 6 gives a schematic overview of magma plumbing system of Asama Volcano. The intruded dike to the west of the summit cannot migrate upward shallower than a depth of 1 km or so below sea level due to negative buoyancy; the host rock there is dense due to repeating dike intrusions and solidification associated with Kurofu eruptions. Magma then found an easier way to make the surface to the east, which formed the present magma pathway of Asama Volcano. Our hypothesis that the high velocity zone is due to solidified magma implies that some amount of magma should fail to make the surface during eruptions. In fact, geodetic data, field observations, and Airborne Synthetic Aperture Radar observations all suggest that in the 2004 eruptions, only a small fraction of intruded magma was ejected (Nakada et al., 2005; Oki et al., 2005). We speculate that in Asama Volcano, much of the intruded magma did not make the surface but was arrested and solidified in past eruptions as well as the 2004 eruptions. 5. Conclusion We conducted an active seismic experiment with five dynamites and more than 400 temporary seismometers to delineate the seismic structure of shallow subsurface of Asama Volcano, Japan. The analysis of first arrival times indicate that there is an area of high P-wave velocity to the west of the summit, the area of the dike intrusion during the 2004 eruptions delineated by earthquake relocations and modeling of geodetic data. We interpret that this high velocity zone is due to the solidification of repeatedly intruded magma. In other words, our interpretation is endorsed by the low conductivity in the area inferred from electromagnetic data. Although the observed waveforms suggest that the seismic structure below the summit may be quite inhomogeneous, our analysis using travel times only is not capable of delineating the fine structure of the magma conduit. A more sophisticated analysis using waveforms, for example, will be needed to gain more insights into the finer structure of Asama Volcano. Acknowledgments Fig. 6. Schematic overview of magma plumbing system of Asama Volcano. Black dots represent the schematics of earthquake locations during the 2004 eruptions (Takeo et al., 2006). The inferred high velocity zone, high resistivity area (Aizawa et al., 2008), and location of diking during the 2004 eruptions (Aoki et al., 2005; Takeo et al., 2006) are also shown. No vertical exaggerations. We are grateful to the participants of the active seismic experiment. The active seismic experiment is supported by the National Project for the Prediction of Volcanic Eruption and Japan Meteorological Agency. Comments by an anonymous reviewer improved the manuscript. Some figures are created with the Generic Mapping Tools (Wessel and Smith, 1998).

6 Y. Aoki et al. / Journal of Volcanology and Geothermal Research 187 (2009) References Aizawa, K., Ogawa, Y., Hashimoto, T., Koyama, T., Kanda, W., Yamaya, Y., Mishina, M., Kagiyama, T., Shallow resistivity structure of Asama Volcano and its implications for magma ascent process in the 2004 eruption. J. Volcanol. Geotherm. Res. 173, Aoki, Y., Watanabe, H., Koyama, E., Oikawa, J., Morita, Y., Ground deformation associated with the unrest of Asama Volcano, Japan. Bull. Volcanol. Soc. Jpn. 50, (in Japanese with English abstract). Aoyama, H., Takeo, M., Wave properties and focal mechanisms of N-type earthquakes at Asama volcano. J. Volcanol. Geotherm. Res. 105, Aramaki, S., Geology of Asama Volcano. J. Fac. Sci. Univ. Tokyo, Sec. II 14, Geological Survey of Japan, Gravity of Japan. Gelogical Survey of Japan (CD-ROM). Morita, Y., Nakao, S., Hayashi, Y., A quantitative approach to the dike intrusion process inferred from a joint analysis of geodetic and seismological data for the 1998 earthquake swarm off the east coast of Izu Peninsula, central Japan. J. Geophys. Res. 111, B doi: /2005jb Nakada, S., Yoshimoto, M., Koyama, E., Tsuji, H., Urabe, T., Comparative study of the 2004 eruption with old eruptions at Asama Volcano and activity evaluation. Bull. Volcanol. Soc. Jpn. 50, (in Japanese with English abstract). Newhall, C.G., Self, S., The volcanic explosivity index (VEI) an estimate of explosive magnitude for historical volcanism. J. Geophys. Res. 87, Oki, S., Murakami, M., Watanabe, N., Urabe, B., Miyawaki, M., Topographic change of the summit crater of the Asama Volcano during 2004 eruption derived from repeated Airborne Synthetic Aperture Radar (SAR) measurements. Bull. Volcanol. Soc. Jpn. 50, Okubo, P.G., Benz, H.M., Chouet, B.A., Imaging the crustal magma sources beneath Mauna Loa and Kilauea volcanoes, Hawaii. Geology 25, Rubin, A.M., Propagation of magma-filled cracks. Ann. Rev. Earth Planet. Sci. 23, Simkin, T., Siebert, L., Volcanoes of the World: A Regional Directory, Gazetteer, and Chronology of Volcanism During the Last 10,000 Years, 2nd ed. Geoscience Press, Tucson, AZ, USA, p Takeo, M., Aoki, Y., Ohminato, T., Yamamoto, M., Magma supply path beneath Mt. Asama volcano, Japan. Geophys. Res. Lett. 33, L doi: /2006gl Tanaka, S., Hamaguchi, H., Nishimura, T., Yamawaki, T., Ueki, S., Nakamichi, H., Tsutsui, T., Miyamachi, H., Matsuwo, N., Oikawa, J., Ohminato, T., Miyaoka, K., Onizawa, S., Mori, T., Aizawa, K., Three-dimensional P-wave velocity structure of Iwate volcano, Japan from active seismic survey. Geophys. Res. Lett. 29, doi: /2002gl Um, J., Thurber, C., A fast algorithm for two-point seismic ray tracing. Bull. Seismolol. Soc. Am. 77, Wessel, P., Smith, W.H.F., New, improved version of the Generic Mapping Tools released. EOS Trans. AGU 79, 579. Yamawaki, T., Tanaka, S., Ueki, S., Hamaguchi, H., Nakamichi, H., Nishimura, T., Oikawa, J., Tsutsui, T., Nishi, K., Shimizu, H., Yamaguchi, S., Miyamachi, H., Yamasato, H., Hayashi, Y., Three-dimensional P-wave velocity structure of Bandai volcano in northeastern Japan inferred from active seismic survey. J. Volcanol. Geotherm. Res. 138, Zelt, C.A., Barton, P.J., Three-dimensional seismic refraction tomography: a comparison of two methods applied to data from the Faeroe Basin. J. Geophys. Res. 103, Zelt, C.A., Smith, R.B., Seismic traveltime inversion for 2-D crustal velocity structure. Geophys. J. Int. 108, Zollo, A., Gasparini, P., Virieux, J., Le Meur, H., de Natale, G., Biella, G., Boschi, E., Capuano, P., de Franco, R., dell'aversana, P., de Matteis, R., Guerra, I., Iannaccone, G., Mirabile, L., Vilardo, G., Seismic evidence for a low-velocity zone in the upper crust beneath Mount Vesuvius. Science 274, Zollo, A., Gasparini, P., Viriuex, J., Biella, G., Boschi, E., Capuano, P., de Franco, R., Dell Aversana,P.,deMatteis,R.,DeNatale,G.,Iannaccone,G.,Guerra,I.,LeMeur, H., Mirabile, L., An image of Mt. Vesuvius obtained by 2D seismic tomography.j.volcanol.geotherm.res.82,