5:15 PM - 6:45 PM
[U15-P45] Early forecast of ground shaking caused by aftershocks of the 2024 Noto Peninsula Earthquake
Keywords:The 2024 Noto Peninsula Earthquake, Aftershocks, Early forecast, Extreme value statistics
The early forecast method for ground shaking by aftershocks (Sawazaki, 2021), which utilizes extreme value statistics, is applied to the Hi-net continuous seismograms that recorded the aftershock sequence of the 2024 Noto Peninsula Earthquake (MJ7.6).
The 2024 Noto Peninsula Earthquake triggered aftershock activity spanning up to 150 km in the NE-SW direction. According to the Headquarters for Earthquake Research Promotion, 1398 felt earthquakes were observed between January 1, 2024, at 16:00 and January 15, 2024, at 08:00. In this study, continuous seismograms from the Hi-net stations N.SUZH and N.TGIH are used; each is located directly above the mainshock hypocenter and close to southwestern edge of the mainshock rupture zone, respectively. The maximum amplitudes at every constant time interval (interval maximum amplitude, IMA) are calculated for the three-component vector sum of each Hi-net seismogram. The waveforms showing mechanical saturations are replaced by waveforms of the concurrent KiK-net subsurface records after correcting for the response function as much as possible. Extreme value analysis is applied to these IMAs, and exceedance probabilities are calculated for the maximum amplitude and the number of felt earthquakes (PGV>0.065cm/s, following Ooi et al. (2002)) from 1, 6, 24, and 72 hours of the mainshock. The forecasted period covers up to seven days from the time of the forecast.
The maximum amplitude caused by the mainshock is 23 cm/s for both N.SUZH and N.TGIH stations. This value is lower than actual maximum amplitude since the Hi-net system is insensitive at frequencies lower than 1 Hz. Furthermore, the shakings observed on the ground surface are larger than these values recorded at depths of 100 meters or more. The probabilities of observing maximum amplitudes exceeding that of the mainshock within seven days at the issued times of 1, 6, 24, and 72 hours of the mainshock are as follows:
N.SUZH: 55% > 40% > 18% > 13%
N.TGIH: 49% > 37% > 29% > 27%
Actually, shakings that exceed the mainshock maximum amplitude have not been observed. However, the probabilities of exceeding the mainshock maximum amplitude are considerably high at early lapse times after the mainshock, reflecting very intense aftershock activity immediately after the mainshock. This probability decreases as time elapses, reflecting the decay of aftershock rate following the Omori-Utsu law. This temporal decay is weaker for N.TGIH than N.SUZH, probably due to the aftershocks activated in shallow part of the southwestern area of the mainshock rupture zone.
The forecast range and observed cumulative number of felt earthquakes within seven days at the issued times of 1, 6, 24, and 72 hours of the mainshock are as follows:
N.SUZH: 103-293(136) > 86-154(87) > 39-70(49) > 23-42(27)
N.TGIH: 293-848(508) > 195-337(388) > 161-260(247) > 87-135(124)
Here, the forecast range is defined by the 10th to 90th percentile values, and the numbers in parentheses indicate the actual number of felt earthquakes observed within seven days. The observed number of felt earthquakes falls within the forecast range except for the underestimation for N.TGIH at the issued time of 6 hours, which is reasonable considering that such probability would be 80 % in theory. Similarly to the exceedance probability for the mainshock maximum amplitude, the forecasted number of felt earthquakes is also large immediately after the mainshock, and then shows temporal decays. Both the forecasted and observed numbers at N.TGIH are more than double that at N.SUZH. The median estimates of the m-value of the Ishimoto-Iida (1939) formula are 1.86-1.99 and 2.00-2.15 for N.SUZH and N.TGIH, respectively. Namely, the relative frequency of smaller earthquakes is higher for N.TGIH, which may be the reason for observing a significant discrepancy in the number of felt earthquakes between the two stations.
Acknowledgements: This study is funded by Grants-in-Aid for Scientific Research C (Grant Number 21K03686) and Seismology TowArd Research innovation with data of Earthquake (STAR-E).
The 2024 Noto Peninsula Earthquake triggered aftershock activity spanning up to 150 km in the NE-SW direction. According to the Headquarters for Earthquake Research Promotion, 1398 felt earthquakes were observed between January 1, 2024, at 16:00 and January 15, 2024, at 08:00. In this study, continuous seismograms from the Hi-net stations N.SUZH and N.TGIH are used; each is located directly above the mainshock hypocenter and close to southwestern edge of the mainshock rupture zone, respectively. The maximum amplitudes at every constant time interval (interval maximum amplitude, IMA) are calculated for the three-component vector sum of each Hi-net seismogram. The waveforms showing mechanical saturations are replaced by waveforms of the concurrent KiK-net subsurface records after correcting for the response function as much as possible. Extreme value analysis is applied to these IMAs, and exceedance probabilities are calculated for the maximum amplitude and the number of felt earthquakes (PGV>0.065cm/s, following Ooi et al. (2002)) from 1, 6, 24, and 72 hours of the mainshock. The forecasted period covers up to seven days from the time of the forecast.
The maximum amplitude caused by the mainshock is 23 cm/s for both N.SUZH and N.TGIH stations. This value is lower than actual maximum amplitude since the Hi-net system is insensitive at frequencies lower than 1 Hz. Furthermore, the shakings observed on the ground surface are larger than these values recorded at depths of 100 meters or more. The probabilities of observing maximum amplitudes exceeding that of the mainshock within seven days at the issued times of 1, 6, 24, and 72 hours of the mainshock are as follows:
N.SUZH: 55% > 40% > 18% > 13%
N.TGIH: 49% > 37% > 29% > 27%
Actually, shakings that exceed the mainshock maximum amplitude have not been observed. However, the probabilities of exceeding the mainshock maximum amplitude are considerably high at early lapse times after the mainshock, reflecting very intense aftershock activity immediately after the mainshock. This probability decreases as time elapses, reflecting the decay of aftershock rate following the Omori-Utsu law. This temporal decay is weaker for N.TGIH than N.SUZH, probably due to the aftershocks activated in shallow part of the southwestern area of the mainshock rupture zone.
The forecast range and observed cumulative number of felt earthquakes within seven days at the issued times of 1, 6, 24, and 72 hours of the mainshock are as follows:
N.SUZH: 103-293(136) > 86-154(87) > 39-70(49) > 23-42(27)
N.TGIH: 293-848(508) > 195-337(388) > 161-260(247) > 87-135(124)
Here, the forecast range is defined by the 10th to 90th percentile values, and the numbers in parentheses indicate the actual number of felt earthquakes observed within seven days. The observed number of felt earthquakes falls within the forecast range except for the underestimation for N.TGIH at the issued time of 6 hours, which is reasonable considering that such probability would be 80 % in theory. Similarly to the exceedance probability for the mainshock maximum amplitude, the forecasted number of felt earthquakes is also large immediately after the mainshock, and then shows temporal decays. Both the forecasted and observed numbers at N.TGIH are more than double that at N.SUZH. The median estimates of the m-value of the Ishimoto-Iida (1939) formula are 1.86-1.99 and 2.00-2.15 for N.SUZH and N.TGIH, respectively. Namely, the relative frequency of smaller earthquakes is higher for N.TGIH, which may be the reason for observing a significant discrepancy in the number of felt earthquakes between the two stations.
Acknowledgements: This study is funded by Grants-in-Aid for Scientific Research C (Grant Number 21K03686) and Seismology TowArd Research innovation with data of Earthquake (STAR-E).