著者

浜田 信生 野坂 大輔 小林 正志 吉川 一光 石垣 祐三 田利 信二朗

出版者

公益社団法人 日本地震学会

雑誌

地震 第2輯 (ISSN:00371114)

巻号頁・発行日

vol.64, no.4, pp.197-209, 2012-06-25 (Released:2012-08-08)

参考文献数

28

---

Matsushiro earthquake swarm is a most well known and well studied earthquake swarm in the history of seismology. From viewpoints of geophysics, geology and geochemistry, various observations, field surveys and analyses had been made to reveal characteristics of the swarm. On the basis of these studies, several models had been proposed to explain why and how the swarm originated and developed. However, majority of studies had focused on the activity around the climatic stage of the swarm in 1966 and so far, studies on the initial stage of the swarm in mid 1965 had remained few due to lack of sufficient observation data. The exact area where the swarm was born had not been known and the swarm was vaguely believed to originate from the area around Mt. Minakami in former Matsushiro town (now belongs to Nagano city). Considering that the manner of initial development of swarm activity represents an important characteristic of the swarm earthquake, we tried to get more clear view about initial stage of the swarm activity in this study. We re-investigated seismograms obtained by routine observation and studied seismograms of temporal stations for the first time which had not been processed yet. By scanning analogue seismograms, we made a complete data set of S-P times with high precision for each station. Although the data set of S-P times from two stations are not sufficient for conventional hypocenter location, we were able to narrow down a possible source area of the swarm activity under the reasonable assumptions. By considering direction of initial motion of P waves and assuming local velocity model of the upper crust in the region and plausible focal depth of 4.5km of the swarm earthquake, we found that the area of swarm in the very beginning in August, 1965 is located about 4km north-east of the Matsushiro earthquake fault (MEF) in former Wakaho town near its border with former Matsushiro town. Size of the initial swarm area was 3-4km in diameter. While swarm activity in the initial region was gradually decaying in September new swarm activity appeared separately from the initial swarm area around southern and south-west part of the MEF. Activity in the new swarm area had been increasing and it was developed to more intense swarm after October, 1965 when establishment of temporal seismic station network of the Earthquake Research Institute, University of Tokyo enabled detailed hypocenter location. The new swarm area coincides with the area where large amount of ground water moved upward and was released on the ground surface in the climatic stage of the swarm. It was well known that source area of the swarm was split and expanded toward north-east and south-west after March, 1967 in its climatic stage. Present study on the initial development of swarm area suggested that characteristics of the Matsushiro earthquake swarm such as splitting and expansion of its source area toward northeast and southwest were inherent in their early stage of the activity in August and September, 1965.

著者

溜渕 功史 山田 安之 石垣 祐三 高木 康伸 中村 雅基 前田 憲二 岡田 正実

出版者

SEISMOLOGICAL SOCIETY OF JAPAN

雑誌

地震 第1輯 (ISSN:00371114)

巻号頁・発行日

vol.62, no.4, pp.193-207, 2010

---

We found eight <I>M</I> 5.1 characteristic earthquakes regularly occurring since 1966 on the plate boundary between the Eurasian plate and the Philippine Sea plate near Miyakojima Island, the Ryukyu Arc, Japan. The quake recurrence interval was 5.89 years in average, and the standard deviation was only 0.73 years. The accumulating stress presumably ruptured the same asperity enclosed by the creeping zone repeatedly. Also, we found three other groups of small repeating earthquakes of <I>M</I> 4, which occurred close to the hypocenters of the <I>M</I> 5 events. Those groups also occurred regularly and we can consider them to be 'characteristic' earthquake sequences. Now, we called those groups A, B, and C. It is not clear whether groups A and B had an intrinsic recurrence interval or if they influenced each other. However, two events of group C occurred within one week after the <I>M</I> 5 quakes, indicating that the <I>M</I> 5 events triggered the group C events whose asperity had suffcient strain energy. No earthquake exceeding <I>M</I> 7, which could change the recurrence intervals, has been observed on the subduction zone around the Ryukyu Islands. Therefore, there should be numerous characteristic earthquake sequences in other areas of the Ryukyu district. We expect that the next <I>M</I> 5 earthquake at 50 km depth on the plate boundary near Miyakojima Island will occur between September 2012 and July 2014 with 70% probability, using the small-sample theory with a log-normal distribution model. Moreover, the <I>M</I> 5 event may be accompanied by an <I>M</I> 4 quake that could rupture the asperity of group C within one week.

著者

溜渕 功史 山田 安之 石垣 祐三 高木 康伸 中村 雅基 前田 憲二 岡田 正実

出版者

公益社団法人 日本地震学会

雑誌

地震 第2輯 (ISSN:00371114)

巻号頁・発行日

vol.62, no.4, pp.193-207, 2010-03-15 (Released:2012-03-26)

参考文献数

26

---

We found eight M 5.1 characteristic earthquakes regularly occurring since 1966 on the plate boundary between the Eurasian plate and the Philippine Sea plate near Miyakojima Island, the Ryukyu Arc, Japan. The quake recurrence interval was 5.89 years in average, and the standard deviation was only 0.73 years. The accumulating stress presumably ruptured the same asperity enclosed by the creeping zone repeatedly. Also, we found three other groups of small repeating earthquakes of M 4, which occurred close to the hypocenters of the M 5 events. Those groups also occurred regularly and we can consider them to be ‘characteristic’ earthquake sequences. Now, we called those groups A, B, and C. It is not clear whether groups A and B had an intrinsic recurrence interval or if they influenced each other. However, two events of group C occurred within one week after the M 5 quakes, indicating that the M 5 events triggered the group C events whose asperity had suffcient strain energy. No earthquake exceeding M 7, which could change the recurrence intervals, has been observed on the subduction zone around the Ryukyu Islands. Therefore, there should be numerous characteristic earthquake sequences in other areas of the Ryukyu district. We expect that the next M 5 earthquake at 50 km depth on the plate boundary near Miyakojima Island will occur between September 2012 and July 2014 with 70% probability, using the small-sample theory with a log-normal distribution model. Moreover, the M 5 event may be accompanied by an M 4 quake that could rupture the asperity of group C within one week.