5:15 PM - 6:45 PM
[U15-P10] Fault configuration of the 2024 Noto earthquake (Mw = 7.5) as deduced from hypocenters and focal mechanism solutions with finite mixture modeling
Keywords:2024 Noto earthquake, fault configuration, hypocentral distribution, focal mechanismsolution
The 2024 Noto earthquake (Mw = 7.5), an intra-plate crustal earthquake, struck the northern part of the Noto Peninsula, central Japan, after active seismic swarm (e.g., Amezawa et al., 2023 GRL) and a moderate-size crustal earthquake (Mw = 6.3) that occurred on May 5, 2023. The relationship between the mainshock and the preceding seismic activities is quite interesting and the physical process behind the activity is worth investigating to deepen our understanding on earthquake generation process.
Fault configuration of a large earthquake play an important role in the physical process. However, the investigation of the configuration is often conducted rather subjectively with generating many cross sections along various directions, which requires a lot of time and effort. A huge amount of seismic data being available in recent years may contribute to clarify the complexity of fault configuration, but it is sometimes difficult to reveal such complexity only with visual inspections. Therefore, we are developing a new method that helps such analysis with providing possible fault configurations subjectively (Horikawa et al., 2023, JpGU). In this study, we apply the method to relocated hypocenters and focal mechanism solutions to infer the fault configuration of the 2024 Noto earthquake.
From the relocated catalog by Shiina et al. (2024, this meeting) with using the three-dimensionally heterogeneous velocity model inferred by Nakajima (2022, EPS), we used the 13,935 hypocenters of the mainshock and aftershocks that occurred until Jan. 31, 2024. Focal mechanism solutions used in this study were the best double couples of the moment tensor solutions of the F-net. The number of the solutions were 194.
On inferring the fault configuration of rectangular faults, we used a finite mixture model in which each mixture component corresponds to a fault. Assuming the number of the rectangular faults in advance, we inferred the following parameters of each fault plane: location of the center of the fault, length and width, strike and dip angles. As for the hypocentral distribution around each fault, we assumed the Gaussian distribution along strike and dip directions whose center is the same as that of the fault, and the standard deviations correspond to length and width. As for the across-fault direction, we assumed the power-law distribution (Powers and Jordan, 2010 JGR). In addition, we took account of the difference in the angle between the normal vector of the fault plane and that of nodal planes for the measure of the agreement between the fault plane and the focal mechanism solutions, assuming that the variation of focal mechanism solutions obeyed the double-couple rotational Cauchy distribution (Kagan, 1992 GJI).
We identified a total of seven fault planes, among which four faults with dip angles of ~30 degrees were along the northern and western coasts of the Noto Peninsula although none of the faults reaches the ground surface. Three faults overlap each other at the northwestern corner of the peninsula. These intersecting faults look a single J-shaped fault. One more fault with northeast-southwest strike was found at the western offshore of the peninsula, which is located at the western end of the series of the faults. The other two faults were identified to the northeastern offshore of the peninsula, but the dip angles and dipping directions of these two faults are poorly constrained because the hypocenters are not well determined due to poor coverage of seismic stations.
Acknowledgments. This study was supported by MEXT Project for Seismology toward Research Innovation with Data of Earthquake (STAR-E) Grant Number JPJ010217. We used the supercomputer of ACCMS, Kyoto University.
Fault configuration of a large earthquake play an important role in the physical process. However, the investigation of the configuration is often conducted rather subjectively with generating many cross sections along various directions, which requires a lot of time and effort. A huge amount of seismic data being available in recent years may contribute to clarify the complexity of fault configuration, but it is sometimes difficult to reveal such complexity only with visual inspections. Therefore, we are developing a new method that helps such analysis with providing possible fault configurations subjectively (Horikawa et al., 2023, JpGU). In this study, we apply the method to relocated hypocenters and focal mechanism solutions to infer the fault configuration of the 2024 Noto earthquake.
From the relocated catalog by Shiina et al. (2024, this meeting) with using the three-dimensionally heterogeneous velocity model inferred by Nakajima (2022, EPS), we used the 13,935 hypocenters of the mainshock and aftershocks that occurred until Jan. 31, 2024. Focal mechanism solutions used in this study were the best double couples of the moment tensor solutions of the F-net. The number of the solutions were 194.
On inferring the fault configuration of rectangular faults, we used a finite mixture model in which each mixture component corresponds to a fault. Assuming the number of the rectangular faults in advance, we inferred the following parameters of each fault plane: location of the center of the fault, length and width, strike and dip angles. As for the hypocentral distribution around each fault, we assumed the Gaussian distribution along strike and dip directions whose center is the same as that of the fault, and the standard deviations correspond to length and width. As for the across-fault direction, we assumed the power-law distribution (Powers and Jordan, 2010 JGR). In addition, we took account of the difference in the angle between the normal vector of the fault plane and that of nodal planes for the measure of the agreement between the fault plane and the focal mechanism solutions, assuming that the variation of focal mechanism solutions obeyed the double-couple rotational Cauchy distribution (Kagan, 1992 GJI).
We identified a total of seven fault planes, among which four faults with dip angles of ~30 degrees were along the northern and western coasts of the Noto Peninsula although none of the faults reaches the ground surface. Three faults overlap each other at the northwestern corner of the peninsula. These intersecting faults look a single J-shaped fault. One more fault with northeast-southwest strike was found at the western offshore of the peninsula, which is located at the western end of the series of the faults. The other two faults were identified to the northeastern offshore of the peninsula, but the dip angles and dipping directions of these two faults are poorly constrained because the hypocenters are not well determined due to poor coverage of seismic stations.
Acknowledgments. This study was supported by MEXT Project for Seismology toward Research Innovation with Data of Earthquake (STAR-E) Grant Number JPJ010217. We used the supercomputer of ACCMS, Kyoto University.