17:15 〜 18:45
[SCG40-P45] Preliminary numerical modeling of slow and fast earthquakes on the subducting Philippine Sea Plate in the Kanto region
キーワード：スロー地震、関東、数値シミュレーション、プレート間地震
Recent studies revealed stress accumulation rate on a subducting boundary based on geodetic data (e.g., Saito and Noda, 2022, 2023). The stress accumulation rate is likely to be more relevant to the frictional property on the plate interface than slip deficit, as the stress change is given by the temporal integration of stress accumulation rate. In the Kanto region, the Philippine Sea Plate is subducting below the inland plate. Devastating megathrust earthquakes repeatedly occur in the Kanto region and, slow slip events (SSEs) are frequently found off the Boso Peninsula. Saito and Noda (2023) estimated stress accumulation rate in this region and suggested three major regions with high stress accumulation rate. In this study, we tried to reproduce the megathrust earthquakes and SSEs based on their result.
To model megathrust earthquakes and SSEs, we adopt a similar approach as in our previous studies (Matsuzawa et al., 2010, 2013), in which rate- and state-dependent friction law (RS-law) with cutoff velocities is used. Based on Saito and Noda (2023), we assume negative (a-b) value in the RS-law within the high stress accumulation region, while positive (a-b) value (i.e., stable sliding) is assumed in the other region. The negative (a-b) region is defined by the region where stress accumulation rate exceeds a threshold value. In the following results, we show the cases with the threshold value of 0.05 MPa/year (Model 1) and 0.02 MPa/year (Model 2). In the off-Boso SSE region, effective normal stress is set to lower value than that at the same depth to reproduce SSEs (e.g., Matsuzawa et al., 2010). The subducting Philippine Sea plate is modeled by about 73,000 triangular elements. Temporal evolution of slip velocity is numerically simulated, introducing elastic response of semi-infinite medium and realistic configuration of the plate interface.
In the numerical result, we successfully reproduced repeating megathrust earthquakes and off-Boso SSEs. As shown in Saito and Noda (2023), four characteristic regions (Odawara, Miura, Awa, and Boso) are reproduced in our simulated result. In Model 1, SSEs repeatedly occur in the Boso region at the interval of about 10 years. On the other hand, the typical interval of the Boso SSEs becomes longer (about 20 years) in Model 2. This is due to the difference of the patch size of the SSE, as Model 2 has wider negative (a-b) region.
In the Odawara (Region O), Miura (Region M), and Awa (Region A) regions, megathrust earthquakes repeatedly occur in the both models. Earthquakes in Region M and A tend to occur more simultaneously in Model 2 than the case of Model 1. This may be caused by the negative (a-b) region which connects the Region A and M in Model 2, while two regions are separated in Model 1. However, occurrences of earthquakes in Region O show no clear relevance to earthquakes in Region M and A in the both models. This may be due to the isolated negative (a-b) region in Region O in the both cases. As the Region O and M are thought to be the slip region at the 1923 Kanto earthquake, our result suggests that further tuning of frictional parameters seems to be necessary. Perhaps, extra negative (a-b) region between Region O and M, or hierarchical frictional property might enable us to reproduce the observations.
To model megathrust earthquakes and SSEs, we adopt a similar approach as in our previous studies (Matsuzawa et al., 2010, 2013), in which rate- and state-dependent friction law (RS-law) with cutoff velocities is used. Based on Saito and Noda (2023), we assume negative (a-b) value in the RS-law within the high stress accumulation region, while positive (a-b) value (i.e., stable sliding) is assumed in the other region. The negative (a-b) region is defined by the region where stress accumulation rate exceeds a threshold value. In the following results, we show the cases with the threshold value of 0.05 MPa/year (Model 1) and 0.02 MPa/year (Model 2). In the off-Boso SSE region, effective normal stress is set to lower value than that at the same depth to reproduce SSEs (e.g., Matsuzawa et al., 2010). The subducting Philippine Sea plate is modeled by about 73,000 triangular elements. Temporal evolution of slip velocity is numerically simulated, introducing elastic response of semi-infinite medium and realistic configuration of the plate interface.
In the numerical result, we successfully reproduced repeating megathrust earthquakes and off-Boso SSEs. As shown in Saito and Noda (2023), four characteristic regions (Odawara, Miura, Awa, and Boso) are reproduced in our simulated result. In Model 1, SSEs repeatedly occur in the Boso region at the interval of about 10 years. On the other hand, the typical interval of the Boso SSEs becomes longer (about 20 years) in Model 2. This is due to the difference of the patch size of the SSE, as Model 2 has wider negative (a-b) region.
In the Odawara (Region O), Miura (Region M), and Awa (Region A) regions, megathrust earthquakes repeatedly occur in the both models. Earthquakes in Region M and A tend to occur more simultaneously in Model 2 than the case of Model 1. This may be caused by the negative (a-b) region which connects the Region A and M in Model 2, while two regions are separated in Model 1. However, occurrences of earthquakes in Region O show no clear relevance to earthquakes in Region M and A in the both models. This may be due to the isolated negative (a-b) region in Region O in the both cases. As the Region O and M are thought to be the slip region at the 1923 Kanto earthquake, our result suggests that further tuning of frictional parameters seems to be necessary. Perhaps, extra negative (a-b) region between Region O and M, or hierarchical frictional property might enable us to reproduce the observations.