5:15 PM - 7:15 PM
[SCG52-P02] Geodetically estimated Nankai asperities below oceanic basins bordered with slow earthquakes
Keywords:Asperities, Seafloor topography, Nankai subduction zone, Geodetic data inversion
A locked zone of a plate boundary fault refers to the interseismically stationary area that eventually incurs frictional failure: an earthquake. The locked zone can be detected from interseismic slip deficit recorded by geodetic data (Savage, 1983). Therefore, the slip deficit normalized by the plate convergence rate, termed plate coupling (Kanamori, 1971), is a proxy of seismic potential (Scholz & Campos, 2012). However, the plate coupling must not be mistaken as the plate locking (Wang and Dixon, 2004). The problem is that the locked zone segment, or an asperity, creates a stress shadow around it, slowing the surrounding creeping zone to fill the stress shadow (Heman et al., 2018; Lindsey et al., 2021). A coupled zone is, in short, an overestimate of a locked zone. Thus, we have developed a method to estimate the locked zone segments, called asperities in fault mechancis, in its original sense of friction (Sato, Hori, & Fukahata, under review; arXiv:2409.14266), where locking (pre-yield) means a stationary phase of zero slip rate while unlocking (post-yield) is interseismically synonymous with a quasi-static phase of zero stressing rate. This presentation reports our preliminary results on the spatial distribution of the locked zone in the Nankai subduction zone and its correlation with seafloor topography and seismic activities.
Our inversion analysis of onshore Global Navigation Satellite System (GNSS) and offshore GNSS-acoustic data by Yokota et al. (2016) has detected five primary asperities (Figure attached). Interestingly, the estimated five asperities are consistent with offshore basins, including those discussed by M. Ando (1975). The asperities correlate with five basins: from the west, Hyuga, Tosa, Muroto, Kumano, and Enshu, the envisioned rupture segments of the Nankai megathrust earthquake (e.g., Hirose et al., 2022). It is intriguing that the “asperity” in fault mechanics, not directly related to the fault surface roughness unlike the original meaning in frictional physics, correlates with seafloor topography, the actual surface roughness, but of the earth.
The estimated asperities constitute the belts of locked zones with a locking gap separating western and eastern segments. The location of the locking gap, the east of the Cape Shionomisaki, is highly consistent with the slip patterns of the 1944 Tonankai and the 1946 Nankai earthquakes estimated by Kikuchi et al. (2003) and Murotani et al. (2015). The rupture initiation point of the 1944 Tonankai earthquake is in the locking gap, and the inverted rupture zone intrudes into the estimated eastern locked segment. The rupture initiation point of the 1946 Nankai earthquake is exactly at the edge of the estimated western locked segment, which includes the slip zone of the 1946 Nankai earthquake. It is physically natural that the earthquake nucleates at the stress concentration zone (Kodaira et al., 2006; Chen & Lapusta, 2009), and we were able to extract the associated interseismic behaviors from the surface deformation. Although rupture initiation points are unclear on or before 1854, the Cape Shionomisaki has been a segmentation boundary of the eastern and western segments, which have hosted megathrust earthquakes separately (Ishibashi, 2004). These facts consistently imply that the locking gap observed from the current geodetic data has been preserved for a geological time scale.
Documented deep low-frequency tremors are all outside the locked zone estimate. The estimated locked zones coincide with the previous focal zones of the same basins (Obara & Kato, 2016) in all basins but the Hyuga. Except for the Hyuga asperity, our results suggest that the seismogenic zones of slow earthquakes are unlocked in long-wavelength and long-time scales. The Hyuga locked zone includes the slip zones of the 1968 Hyuga-nada earthquake (Yagi et al., 1999) and the Bungo-Channel long-term SSEs (Obara & Kato, 2016). Moreover, this zone is supposed to have experienced the fault slip during the 1707 Hoei earthquake (Furumura et al., 2011). These behaviors of the Hyuga (Bungo-Channel) locked zone are highly complex, but here is one simple, consistent interpretation of these behaviors. Namely, the Hyuga locked zone is exceptionally the nucleation zone that often fails to slip faster, as in the Bungo-Channel slow-slip events, but sometimes succeeds, as supposedly in 1707.
Our inversion analysis of onshore Global Navigation Satellite System (GNSS) and offshore GNSS-acoustic data by Yokota et al. (2016) has detected five primary asperities (Figure attached). Interestingly, the estimated five asperities are consistent with offshore basins, including those discussed by M. Ando (1975). The asperities correlate with five basins: from the west, Hyuga, Tosa, Muroto, Kumano, and Enshu, the envisioned rupture segments of the Nankai megathrust earthquake (e.g., Hirose et al., 2022). It is intriguing that the “asperity” in fault mechanics, not directly related to the fault surface roughness unlike the original meaning in frictional physics, correlates with seafloor topography, the actual surface roughness, but of the earth.
The estimated asperities constitute the belts of locked zones with a locking gap separating western and eastern segments. The location of the locking gap, the east of the Cape Shionomisaki, is highly consistent with the slip patterns of the 1944 Tonankai and the 1946 Nankai earthquakes estimated by Kikuchi et al. (2003) and Murotani et al. (2015). The rupture initiation point of the 1944 Tonankai earthquake is in the locking gap, and the inverted rupture zone intrudes into the estimated eastern locked segment. The rupture initiation point of the 1946 Nankai earthquake is exactly at the edge of the estimated western locked segment, which includes the slip zone of the 1946 Nankai earthquake. It is physically natural that the earthquake nucleates at the stress concentration zone (Kodaira et al., 2006; Chen & Lapusta, 2009), and we were able to extract the associated interseismic behaviors from the surface deformation. Although rupture initiation points are unclear on or before 1854, the Cape Shionomisaki has been a segmentation boundary of the eastern and western segments, which have hosted megathrust earthquakes separately (Ishibashi, 2004). These facts consistently imply that the locking gap observed from the current geodetic data has been preserved for a geological time scale.
Documented deep low-frequency tremors are all outside the locked zone estimate. The estimated locked zones coincide with the previous focal zones of the same basins (Obara & Kato, 2016) in all basins but the Hyuga. Except for the Hyuga asperity, our results suggest that the seismogenic zones of slow earthquakes are unlocked in long-wavelength and long-time scales. The Hyuga locked zone includes the slip zones of the 1968 Hyuga-nada earthquake (Yagi et al., 1999) and the Bungo-Channel long-term SSEs (Obara & Kato, 2016). Moreover, this zone is supposed to have experienced the fault slip during the 1707 Hoei earthquake (Furumura et al., 2011). These behaviors of the Hyuga (Bungo-Channel) locked zone are highly complex, but here is one simple, consistent interpretation of these behaviors. Namely, the Hyuga locked zone is exceptionally the nucleation zone that often fails to slip faster, as in the Bungo-Channel slow-slip events, but sometimes succeeds, as supposedly in 1707.