9:45 AM - 10:00 AM
[SCG55-04] Mechanically coupled areas along the Sagami trough and its implications for seismic cycles of thrust-type earthquakes
★Invited Papers
Keywords:Mechanical coupling, Thrust-type earthquakes, Marine terraces, Elastic strain energy
In the Sagami trough subduction zone, great thrust-type earthquakes, such as the 1923 Taisho (Mw8.0) and 1703 Genroku (Mw8.1-8.5) earthquakes, have repeatedly occurred. It has been considered that earthquakes of the same magnitude as the 1923 event (referred to as "Taisho-type earthquakes" hereafter) have occurred at intervals of several hundred years, and that one in several of these earthquakes has developed into an earthquake of the same magnitude as the 1703 event (referred to as "Genroku-type earthquake" hereafter) at intervals of approximately 2000 years. This idea comes from the analogy between the amount of uplift due to the 1923 and 1703 events and the height of marine terraces observed at the coast of the Boso Peninsula. A series of well-developed Holocene terraces (Numa I-IV terraces) were considered to have formed by the Genroku-type earthquakes. These wide terraces can be divided into several narrow steps, which were considered to be due to the Taisho-type earthquakes (Shishikura, 2014 Episodes). According to that idea of earthquake recurrence intervals, the probability of a Taisho-type earthquake is still low because only 100 years have passed since the 1923 event, and the probability of a Genroku-type earthquake is even lower (The Headquarters for Earthquake Research Promotion, 2014).
Saito & Noda (2023 BSSA) estimated the stress accumulation rate at the plate boundary along the Sagami trough from interseismic GNSS displacement rate data. The results revealed the existence of four isolated peaks of stress accumulation (mechanically coupled areas) along the trough, three of which correspond to the source areas of the 1923 and 1703 events. The combination of these coupled areas would create the differences between the 1923 and 1703 earthquakes. Assuming that stress accumulated at a constant rate since the past earthquakes will be completely released during the next earthquake, they also developed several earthquake scenarios in 2023 and examined whether or not each scenario meets the necessary condition for realizing the fault motions in terms of the energy balance. The scenario of a Taisho-type earthquake cannot be realized in 2023, while the scenario of a Genroku-type earthquake meets the necessary condition. This is because the stress has accumulated for 320 years since the 1703 event (up to 3 MPa) in the mechanically coupled area just below the southern tip of the Boso Peninsula. Based on the mechanical coupling, the Genroku-type earthquake is more likely to occur than the Taisho-type earthquake at present, which conflicts with the earthquake recurrence intervals from marine terraces. Furthermore, some earthquake scenarios other than the Taisho-type and Genroku-type earthquakes are also possible. These results indicate that the seismic cycles in the Sagami trough subduction zone are much more complex than the idea of two known types of earthquakes, Taisho- and Genroku-type, are repeating and that an earthquake may occur earlier than the recurrence interval based on marine terraces.
To understand the reason for the conflict, a simulation of Holocene terrace development in the Boso Peninsula by Noda et al. (2018 Tectonophysics) provides an important clue. This study modeled erosion and deposition processes on the coast and simulated the evolution of coastal landforms. The results show that the formation of Numa I-IV terraces requires sea-level fluctuation with a period of 2000-3000 years. Seismic cycles with periods of less than 1200 years had less impact on coastal landforms. It is noteworthy that earthquakes occurring at intervals shorter than 2000 years left few traces and could be overlooked. Therefore, it is not appropriate to discuss the recurrence interval of Genroku-type earthquakes or the rupture of the mechanically coupled area under the Boso Peninsula based only on the marine terraces. Taking into account not only the coastal landforms but also the mechanical coupling at the plate interface, we need to reevaluate the seismic risk under the Boso Peninsula.
Saito & Noda (2023 BSSA) estimated the stress accumulation rate at the plate boundary along the Sagami trough from interseismic GNSS displacement rate data. The results revealed the existence of four isolated peaks of stress accumulation (mechanically coupled areas) along the trough, three of which correspond to the source areas of the 1923 and 1703 events. The combination of these coupled areas would create the differences between the 1923 and 1703 earthquakes. Assuming that stress accumulated at a constant rate since the past earthquakes will be completely released during the next earthquake, they also developed several earthquake scenarios in 2023 and examined whether or not each scenario meets the necessary condition for realizing the fault motions in terms of the energy balance. The scenario of a Taisho-type earthquake cannot be realized in 2023, while the scenario of a Genroku-type earthquake meets the necessary condition. This is because the stress has accumulated for 320 years since the 1703 event (up to 3 MPa) in the mechanically coupled area just below the southern tip of the Boso Peninsula. Based on the mechanical coupling, the Genroku-type earthquake is more likely to occur than the Taisho-type earthquake at present, which conflicts with the earthquake recurrence intervals from marine terraces. Furthermore, some earthquake scenarios other than the Taisho-type and Genroku-type earthquakes are also possible. These results indicate that the seismic cycles in the Sagami trough subduction zone are much more complex than the idea of two known types of earthquakes, Taisho- and Genroku-type, are repeating and that an earthquake may occur earlier than the recurrence interval based on marine terraces.
To understand the reason for the conflict, a simulation of Holocene terrace development in the Boso Peninsula by Noda et al. (2018 Tectonophysics) provides an important clue. This study modeled erosion and deposition processes on the coast and simulated the evolution of coastal landforms. The results show that the formation of Numa I-IV terraces requires sea-level fluctuation with a period of 2000-3000 years. Seismic cycles with periods of less than 1200 years had less impact on coastal landforms. It is noteworthy that earthquakes occurring at intervals shorter than 2000 years left few traces and could be overlooked. Therefore, it is not appropriate to discuss the recurrence interval of Genroku-type earthquakes or the rupture of the mechanically coupled area under the Boso Peninsula based only on the marine terraces. Taking into account not only the coastal landforms but also the mechanical coupling at the plate interface, we need to reevaluate the seismic risk under the Boso Peninsula.