5:15 PM - 6:45 PM
[U15-P12] Relocations and uncertainty quantifications of earthquake hypocenters in and around the source area of the 2024 Mj 7.6 Noto Peninsula earthquake, Japan
Keywords:2024 Noto Peninsula Earthquake, Hypocenter determination, MCMC
A large crustal earthquake (Mj 7.6) struck the Noto Peninsula, central Japan, at 16:10 (JST = UT + 9 hours) on New Year's Day, 2024. The mainshock rupture extended ~150 km in length and generated tsunami. The rupture area covered the source regions of intense swarm activity in the northeastern tip of the peninsula [Amezawa et al., 2023] and the previous large crustal earthquakes, such as the 2007 (Mj 6.9) event. Therefore, revealing the detailed hypocenter distribution in and around the source area of the 2024 event will provide fundamental information for understanding the rupture process of the mainshock and seismotectonics in the Noto Peninsula.
In this study, we tried to precisely determine locations and quantify the uncertainty of the earthquake hypocenters that occurred around the Noto Peninsula. Specifically, we used the three-dimensional velocity structure model [Nakajima, 2022] for relocation, instead of a one-dimensional velocity model, such as the JMA2001 velocity model [Ueno et al., 2002]. Following the methodology of Shiina and Kano [2022], we applied the Markov Chain Monte Carlo (MCMC) technique and jointly determined the hypocenter locations and station correction terms. We collected earthquake data during from January 2005 to January 2024 from the Japan Meteorological Agency unified earthquake catalog. The lower limit of magnitude is 1.0. 42,583 earthquakes were collected in the analysis period, which included 14,299 aftershocks. An advantage of the MCMC-based relocation method is that we can directly evaluate the uncertainty of estimated parameters. For example, a certain number of aftershocks occurred offshore, indicating that these offshore events may get worse estimation accuracy due to limited station coverage compared with the inland area. As a result of the relocation considering the three-dimensional heterogeneity, the depth of these offshore events became shallower by about 5 km, which was larger than the uncertainty in observations, and most of them were relocated at a depth of less than 15 km. The results suggest that the mainshock rupture in the offshore area also occurred near the ground surface, as discussed based on the observations of the tsunami [e.g., Fujii and Satake, 2024].
In this study, we tried to precisely determine locations and quantify the uncertainty of the earthquake hypocenters that occurred around the Noto Peninsula. Specifically, we used the three-dimensional velocity structure model [Nakajima, 2022] for relocation, instead of a one-dimensional velocity model, such as the JMA2001 velocity model [Ueno et al., 2002]. Following the methodology of Shiina and Kano [2022], we applied the Markov Chain Monte Carlo (MCMC) technique and jointly determined the hypocenter locations and station correction terms. We collected earthquake data during from January 2005 to January 2024 from the Japan Meteorological Agency unified earthquake catalog. The lower limit of magnitude is 1.0. 42,583 earthquakes were collected in the analysis period, which included 14,299 aftershocks. An advantage of the MCMC-based relocation method is that we can directly evaluate the uncertainty of estimated parameters. For example, a certain number of aftershocks occurred offshore, indicating that these offshore events may get worse estimation accuracy due to limited station coverage compared with the inland area. As a result of the relocation considering the three-dimensional heterogeneity, the depth of these offshore events became shallower by about 5 km, which was larger than the uncertainty in observations, and most of them were relocated at a depth of less than 15 km. The results suggest that the mainshock rupture in the offshore area also occurred near the ground surface, as discussed based on the observations of the tsunami [e.g., Fujii and Satake, 2024].