Seismicity associated with the 1991-1995 dome growth at Unzen Volcano, Japan

K. Umakoshi, N. Takamura, N. Shinzato, K. Uchida, N. Matsuwo, Hiroshi Shimizu

研究成果: ジャーナルへの寄稿記事

22 引用 (Scopus)

抄録

From May 1991 until February 1995, seismicity in the crater area of Unzen Volcano, southwest Japan, intensified in conjunction with the growth of a dacite lava dome. We used data from seismic stations located near the crater to identify approximately 580,000 summit earthquakes with maximum amplitudes equal to or greater than 1 × 10- 3 cm/s. The temporal characteristics of the seismicity level were different for exogenous and endogenous periods of dome growth. Periods of solely exogenous growth were accompanied by several days or weeks of increased seismicity, and levels of seismicity were notably reduced between successive seismically active periods. In contrast, levels of seismicity were generally high during periods when the dome grew endogenously, with repeated cycles of increasing and decreasing seismicity of one to two months duration. We classified the waveforms of summit earthquakes into high-frequency (HF), medium-frequency (MF), and low-frequency (LF) types on the basis of spectral analysis. Dominant waveform types varied significantly over time: HF was dominant in May 1991, LF from June 1991 until August 1993, MF during September and October 1993, and HF and MF from November 1993. HF and MF events are mainly distributed at depths of 500-1100 m above sea level (ASL), just below the lava dome, while LF events are widely distributed at depths from 500 m ASL to the interior of the dome itself. To efficiently detect earthquake families, we conducted a cross-correlation analysis of waveforms. Using the events one-by-one as reference events, we calculated peak correlation coefficients between each reference event and events that occurred within 24 h either side of the reference event. The results show that many earthquake families of all waveform types occurred throughout the growth period of the dome. The durations of most families were less than two weeks and were not related to the intensity of seismic activity. The incidence rate of events within each family reached a peak in the middle of the respective activity period. The results also suggest that the waveforms in several analyzed families gradually evolved over time. During periods when HF events intensified, several families appeared contemporaneously. In contrast, during periods when LF events intensified, the occurrence pattern of earthquake families was relatively simple: a new family became active only once the activity of the previous family had declined or ceased completely. A possible source mechanism for families of HF events is stick-slip within the stiff rocks surrounding the conduit; however, several different types of source mechanisms should be considered for families of LF events.

元の言語英語
ページ(範囲)91-99
ページ数9
ジャーナルJournal of Volcanology and Geothermal Research
175
発行部数1-2
DOI
出版物ステータス出版済み - 7 30 2008

Fingerprint

Volcanoes
Domes
domes
volcanoes
dome
seismicity
Japan
volcano
Earthquakes
Sea level
waveforms
earthquakes
earthquake
low frequencies
lava dome
Stick-slip
lava
sea level
crater
craters

All Science Journal Classification (ASJC) codes

  • Geophysics
  • Geochemistry and Petrology

これを引用

Seismicity associated with the 1991-1995 dome growth at Unzen Volcano, Japan. / Umakoshi, K.; Takamura, N.; Shinzato, N.; Uchida, K.; Matsuwo, N.; Shimizu, Hiroshi.

:: Journal of Volcanology and Geothermal Research, 巻 175, 番号 1-2, 30.07.2008, p. 91-99.

研究成果: ジャーナルへの寄稿記事

Umakoshi, K. ; Takamura, N. ; Shinzato, N. ; Uchida, K. ; Matsuwo, N. ; Shimizu, Hiroshi. / Seismicity associated with the 1991-1995 dome growth at Unzen Volcano, Japan. :: Journal of Volcanology and Geothermal Research. 2008 ; 巻 175, 番号 1-2. pp. 91-99.
@article{1f43f709576d4d94b02aee6c52a8732d,
title = "Seismicity associated with the 1991-1995 dome growth at Unzen Volcano, Japan",
abstract = "From May 1991 until February 1995, seismicity in the crater area of Unzen Volcano, southwest Japan, intensified in conjunction with the growth of a dacite lava dome. We used data from seismic stations located near the crater to identify approximately 580,000 summit earthquakes with maximum amplitudes equal to or greater than 1 × 10- 3 cm/s. The temporal characteristics of the seismicity level were different for exogenous and endogenous periods of dome growth. Periods of solely exogenous growth were accompanied by several days or weeks of increased seismicity, and levels of seismicity were notably reduced between successive seismically active periods. In contrast, levels of seismicity were generally high during periods when the dome grew endogenously, with repeated cycles of increasing and decreasing seismicity of one to two months duration. We classified the waveforms of summit earthquakes into high-frequency (HF), medium-frequency (MF), and low-frequency (LF) types on the basis of spectral analysis. Dominant waveform types varied significantly over time: HF was dominant in May 1991, LF from June 1991 until August 1993, MF during September and October 1993, and HF and MF from November 1993. HF and MF events are mainly distributed at depths of 500-1100 m above sea level (ASL), just below the lava dome, while LF events are widely distributed at depths from 500 m ASL to the interior of the dome itself. To efficiently detect earthquake families, we conducted a cross-correlation analysis of waveforms. Using the events one-by-one as reference events, we calculated peak correlation coefficients between each reference event and events that occurred within 24 h either side of the reference event. The results show that many earthquake families of all waveform types occurred throughout the growth period of the dome. The durations of most families were less than two weeks and were not related to the intensity of seismic activity. The incidence rate of events within each family reached a peak in the middle of the respective activity period. The results also suggest that the waveforms in several analyzed families gradually evolved over time. During periods when HF events intensified, several families appeared contemporaneously. In contrast, during periods when LF events intensified, the occurrence pattern of earthquake families was relatively simple: a new family became active only once the activity of the previous family had declined or ceased completely. A possible source mechanism for families of HF events is stick-slip within the stiff rocks surrounding the conduit; however, several different types of source mechanisms should be considered for families of LF events.",
author = "K. Umakoshi and N. Takamura and N. Shinzato and K. Uchida and N. Matsuwo and Hiroshi Shimizu",
year = "2008",
month = "7",
day = "30",
doi = "10.1016/j.jvolgeores.2008.03.030",
language = "English",
volume = "175",
pages = "91--99",
journal = "Journal of Volcanology and Geothermal Research",
issn = "0377-0273",
publisher = "Elsevier",
number = "1-2",

}

TY - JOUR

T1 - Seismicity associated with the 1991-1995 dome growth at Unzen Volcano, Japan

AU - Umakoshi, K.

AU - Takamura, N.

AU - Shinzato, N.

AU - Uchida, K.

AU - Matsuwo, N.

AU - Shimizu, Hiroshi

PY - 2008/7/30

Y1 - 2008/7/30

N2 - From May 1991 until February 1995, seismicity in the crater area of Unzen Volcano, southwest Japan, intensified in conjunction with the growth of a dacite lava dome. We used data from seismic stations located near the crater to identify approximately 580,000 summit earthquakes with maximum amplitudes equal to or greater than 1 × 10- 3 cm/s. The temporal characteristics of the seismicity level were different for exogenous and endogenous periods of dome growth. Periods of solely exogenous growth were accompanied by several days or weeks of increased seismicity, and levels of seismicity were notably reduced between successive seismically active periods. In contrast, levels of seismicity were generally high during periods when the dome grew endogenously, with repeated cycles of increasing and decreasing seismicity of one to two months duration. We classified the waveforms of summit earthquakes into high-frequency (HF), medium-frequency (MF), and low-frequency (LF) types on the basis of spectral analysis. Dominant waveform types varied significantly over time: HF was dominant in May 1991, LF from June 1991 until August 1993, MF during September and October 1993, and HF and MF from November 1993. HF and MF events are mainly distributed at depths of 500-1100 m above sea level (ASL), just below the lava dome, while LF events are widely distributed at depths from 500 m ASL to the interior of the dome itself. To efficiently detect earthquake families, we conducted a cross-correlation analysis of waveforms. Using the events one-by-one as reference events, we calculated peak correlation coefficients between each reference event and events that occurred within 24 h either side of the reference event. The results show that many earthquake families of all waveform types occurred throughout the growth period of the dome. The durations of most families were less than two weeks and were not related to the intensity of seismic activity. The incidence rate of events within each family reached a peak in the middle of the respective activity period. The results also suggest that the waveforms in several analyzed families gradually evolved over time. During periods when HF events intensified, several families appeared contemporaneously. In contrast, during periods when LF events intensified, the occurrence pattern of earthquake families was relatively simple: a new family became active only once the activity of the previous family had declined or ceased completely. A possible source mechanism for families of HF events is stick-slip within the stiff rocks surrounding the conduit; however, several different types of source mechanisms should be considered for families of LF events.

AB - From May 1991 until February 1995, seismicity in the crater area of Unzen Volcano, southwest Japan, intensified in conjunction with the growth of a dacite lava dome. We used data from seismic stations located near the crater to identify approximately 580,000 summit earthquakes with maximum amplitudes equal to or greater than 1 × 10- 3 cm/s. The temporal characteristics of the seismicity level were different for exogenous and endogenous periods of dome growth. Periods of solely exogenous growth were accompanied by several days or weeks of increased seismicity, and levels of seismicity were notably reduced between successive seismically active periods. In contrast, levels of seismicity were generally high during periods when the dome grew endogenously, with repeated cycles of increasing and decreasing seismicity of one to two months duration. We classified the waveforms of summit earthquakes into high-frequency (HF), medium-frequency (MF), and low-frequency (LF) types on the basis of spectral analysis. Dominant waveform types varied significantly over time: HF was dominant in May 1991, LF from June 1991 until August 1993, MF during September and October 1993, and HF and MF from November 1993. HF and MF events are mainly distributed at depths of 500-1100 m above sea level (ASL), just below the lava dome, while LF events are widely distributed at depths from 500 m ASL to the interior of the dome itself. To efficiently detect earthquake families, we conducted a cross-correlation analysis of waveforms. Using the events one-by-one as reference events, we calculated peak correlation coefficients between each reference event and events that occurred within 24 h either side of the reference event. The results show that many earthquake families of all waveform types occurred throughout the growth period of the dome. The durations of most families were less than two weeks and were not related to the intensity of seismic activity. The incidence rate of events within each family reached a peak in the middle of the respective activity period. The results also suggest that the waveforms in several analyzed families gradually evolved over time. During periods when HF events intensified, several families appeared contemporaneously. In contrast, during periods when LF events intensified, the occurrence pattern of earthquake families was relatively simple: a new family became active only once the activity of the previous family had declined or ceased completely. A possible source mechanism for families of HF events is stick-slip within the stiff rocks surrounding the conduit; however, several different types of source mechanisms should be considered for families of LF events.

UR - http://www.scopus.com/inward/record.url?scp=50049124158&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=50049124158&partnerID=8YFLogxK

U2 - 10.1016/j.jvolgeores.2008.03.030

DO - 10.1016/j.jvolgeores.2008.03.030

M3 - Article

AN - SCOPUS:50049124158

VL - 175

SP - 91

EP - 99

JO - Journal of Volcanology and Geothermal Research

JF - Journal of Volcanology and Geothermal Research

SN - 0377-0273

IS - 1-2

ER -