10:15 AM - 10:30 AM
[SSS12-18] Intraslab reverse faulting adjacent to the hypocenter of the 1923 Kanto earthquake: The Mw 5.0 western Kanagawa earthquake on 9 August 2024
Keywords:Fault geometry, Hypocenter clustering, Intraslab reverse fault, Machine learning, Kanto Earthquake
An Mj 5.3 earthquake occurred in western Kanagawa Prefecture on 9 August 2024 at 7:57:38 pm JST (UTC+9), resulting in a maximum Japan Meteorological Agency (JMA) seismic intensity scale of 5 Lower in this region. The F-net moment tensor solution indicated a reverse-fault-type focal mechanism of Mw 5.0 at a depth of 11 km, characterized by a north–south compressional axis. The Tokyo Metropolitan area, including Kanagawa Prefecture, has historically experienced huge earthquakes, known as the “Great Kanto earthquakes”. The most recent earthquake on 1 September 1923 was a megathrust-type event on the Philippine Sea plate (PHS). The rupture of the 1923 Kanto earthquake is considered to have initiated in western Kanagawa (Kanamori and Miyamura 1970), near the epicenter of this Mw 5.0 earthquake. Therefore, a geophysical interpretation of this Mw 5.0 earthquake is crucial for understanding of the potential for future Kanto earthquakes.
Interestingly, the epicenter of this earthquake was located on the Matsuda-kita active fault. Additionally, the northward collision of the Izu Peninsula has further complicated the seismotectonic of this area. This raises the scientific question: “What type of faulting occurred—interplate, active-fault, or intraslab earthquake?” To tackle this question, we examined the fault geometry of this earthquake from the hypocenter distribution using machine learning techniques.
The general flow of this study consists of three steps: (1) picking of P- and S-wave arrivals for selected events using a deep-neural-network-based picker (Zhu and Beroza 2019) which was re-trained on the JMA catalog (Naoi et al. 2024); (2) hypocenter relocation with picked onsets and waveform correlation values; and (3) extraction of fault planes through hypocenter clustering analysis (Sawaki et al., 2025). We selected 867 events with Mj 0.5 or greater from January 2004 to October 2024 for relocation and used 364 events after 9 August 2024 for hypocenter clustering.
We extracted four fault planes through the hypocenter clustering. The upper edge depth of the four planes was nearly 12 km. In the northeastern area with the mainshock, a south-dipping plane was identified, with a dip angle of 67°. A cross-sectional view showed that this cluster consists of aligned hypocenters and exhibits a steeply dipping structure. The mainshock focal mechanism in the same view showed that the other steep nodal plane, different from the one parallel to the PHS surface, is on the extracted plane. Considering the PHS surface depth of 10 km, our results demonstrate that the mainshock occurred as an intraslab earthquake on a steep reverse fault, as opposed to thrusting on the PHS surface. We will discuss the geophysical implications of intraslab steep reverse earthquakes on earthquake generation and seismotectonics.
[Acknowledgments]
We used JMA Unified Catalog and phase picks. We analyzed seismic waveforms recorded by NIED Hi-net, JMA, ERI, and Hot Springs Research Institute of Kanagawa Prefecture. This study was supported by MEXT Project for Seismology toward Research Innovation with Data of Earthquake (STAR-E) Grant Number JPJ010217.
Interestingly, the epicenter of this earthquake was located on the Matsuda-kita active fault. Additionally, the northward collision of the Izu Peninsula has further complicated the seismotectonic of this area. This raises the scientific question: “What type of faulting occurred—interplate, active-fault, or intraslab earthquake?” To tackle this question, we examined the fault geometry of this earthquake from the hypocenter distribution using machine learning techniques.
The general flow of this study consists of three steps: (1) picking of P- and S-wave arrivals for selected events using a deep-neural-network-based picker (Zhu and Beroza 2019) which was re-trained on the JMA catalog (Naoi et al. 2024); (2) hypocenter relocation with picked onsets and waveform correlation values; and (3) extraction of fault planes through hypocenter clustering analysis (Sawaki et al., 2025). We selected 867 events with Mj 0.5 or greater from January 2004 to October 2024 for relocation and used 364 events after 9 August 2024 for hypocenter clustering.
We extracted four fault planes through the hypocenter clustering. The upper edge depth of the four planes was nearly 12 km. In the northeastern area with the mainshock, a south-dipping plane was identified, with a dip angle of 67°. A cross-sectional view showed that this cluster consists of aligned hypocenters and exhibits a steeply dipping structure. The mainshock focal mechanism in the same view showed that the other steep nodal plane, different from the one parallel to the PHS surface, is on the extracted plane. Considering the PHS surface depth of 10 km, our results demonstrate that the mainshock occurred as an intraslab earthquake on a steep reverse fault, as opposed to thrusting on the PHS surface. We will discuss the geophysical implications of intraslab steep reverse earthquakes on earthquake generation and seismotectonics.
[Acknowledgments]
We used JMA Unified Catalog and phase picks. We analyzed seismic waveforms recorded by NIED Hi-net, JMA, ERI, and Hot Springs Research Institute of Kanagawa Prefecture. This study was supported by MEXT Project for Seismology toward Research Innovation with Data of Earthquake (STAR-E) Grant Number JPJ010217.