1:45 PM - 3:15 PM
[SSS09-P10] Broadband source characteristics of the 2021 MW 8.2 Chignik, Alaska, interplate earthquake
Keywords:Interplate earthquake, Aleutian trench, Source scaling relationship, Large slip area (Slip asperity area), Strong-motion generation area
On July 29, 2021, an MW 8.2 interplate earthquake (Chignik earthquake) occurred at the interface between the North American and Pacific plates off Alaska Peninsula. Using the waveform data observed around the source region, we investigated the broadband source characteristics of this event, such as the frequency dependence of the seismic radiation. We performed ground motion simulations at two period ranges. We estimated Large Slip Areas (LSAs; slip asperity areas) in the period range of 12.5-33 s (0.03-0.08 Hz) and Strong-Motion Generation Areas (SMGAs) in the period range of 0.1-10 s (0.1-10 Hz). This study will be very useful for updating the method of source modeling in the ground-motion prediction of interplate earthquakes.
First, we extracted an effective rupture area by applying the criterion of Somerville et al. (1999) to the slip distribution inverted by USGS (2021). Based on the extracted rupture area, which roughly covered the entire range of aftershocks, we located a fault plane on the plate boundary. In the long-period range, we constructed a characterized source model consisting of two rectangular LSAs and a background area. We adjusted the location and dimension of the LSAs by trial and error so that the synthetic waveforms can reproduce the observed pulse-like waveforms and amplitude level. The synthetic waveforms were calculated by using the discrete wavenumber method (Bouchon 1981) and the reflection/transmission coefficient matrix method (Kennett and Kerry 1979) and assuming 1D velocity structure models. The 1D velocity structure models were calibrated against the waveforms observed during an MW 5.5 aftershock, where the model of Nayak et al. (2020) was used as an initial model. In the short-period range, we assumed two square SMGAs corresponding to the observed two pulses. We estimated the parameters of these SMGAs using the empirical Green's function method (Irikura, 1986). The above MW 5.5 aftershock was also used as the element event for the empirical Green's function method.
We found that the SMGAs were located at a slightly greater depth than the LSAs. The estimated spatial separation of SMGAs and LSAs is consistent with the depth-varying properties proposed by Lay et al. (2012), where the strong short-period seismic radiation occurs in the downdip portion of the megathrust. We suggest that the LSAs during this event ruptured most of Domain B of Lay et al. (2012) and the SMGAs ruptured the bottom of Domain B or the top of Domain C. We estimated that the LSAs (7840 km2) and SMGAs (3904 km2) occupied 21% and 11% of the rupture area (36652 km2), respectively. The dimensions of the rupture area and LSAs are consistent with those derived from Murotani et al. (2008), a scaling relationship for interplate earthquakes in Japan. However, the dimension of SMGAs is larger than the scaling relationship of Satoh (2010), which suggests that SMGAs occupy 4% of the rupture area. Although the fact that SMGAs have smaller dimension than LSAs is considered to be a source characteristic for interplate events (e.g., Satoh, 2010; Tajima et al., 2013), the ratio of SMGAs to LSAs requires further studies such as explaining broadband ground motions and taking into account the regional differences of oceanic plates.
Acknowledgements: This study was based on the 2022 research project “the study on the characterized source model for interplate earthquakes” by the Secretariat of Nuclear Regulation Authority, Japan. The waveform data were obtained from the Alaska Regional Network and the Alaska Volcano Observatory.
First, we extracted an effective rupture area by applying the criterion of Somerville et al. (1999) to the slip distribution inverted by USGS (2021). Based on the extracted rupture area, which roughly covered the entire range of aftershocks, we located a fault plane on the plate boundary. In the long-period range, we constructed a characterized source model consisting of two rectangular LSAs and a background area. We adjusted the location and dimension of the LSAs by trial and error so that the synthetic waveforms can reproduce the observed pulse-like waveforms and amplitude level. The synthetic waveforms were calculated by using the discrete wavenumber method (Bouchon 1981) and the reflection/transmission coefficient matrix method (Kennett and Kerry 1979) and assuming 1D velocity structure models. The 1D velocity structure models were calibrated against the waveforms observed during an MW 5.5 aftershock, where the model of Nayak et al. (2020) was used as an initial model. In the short-period range, we assumed two square SMGAs corresponding to the observed two pulses. We estimated the parameters of these SMGAs using the empirical Green's function method (Irikura, 1986). The above MW 5.5 aftershock was also used as the element event for the empirical Green's function method.
We found that the SMGAs were located at a slightly greater depth than the LSAs. The estimated spatial separation of SMGAs and LSAs is consistent with the depth-varying properties proposed by Lay et al. (2012), where the strong short-period seismic radiation occurs in the downdip portion of the megathrust. We suggest that the LSAs during this event ruptured most of Domain B of Lay et al. (2012) and the SMGAs ruptured the bottom of Domain B or the top of Domain C. We estimated that the LSAs (7840 km2) and SMGAs (3904 km2) occupied 21% and 11% of the rupture area (36652 km2), respectively. The dimensions of the rupture area and LSAs are consistent with those derived from Murotani et al. (2008), a scaling relationship for interplate earthquakes in Japan. However, the dimension of SMGAs is larger than the scaling relationship of Satoh (2010), which suggests that SMGAs occupy 4% of the rupture area. Although the fact that SMGAs have smaller dimension than LSAs is considered to be a source characteristic for interplate events (e.g., Satoh, 2010; Tajima et al., 2013), the ratio of SMGAs to LSAs requires further studies such as explaining broadband ground motions and taking into account the regional differences of oceanic plates.
Acknowledgements: This study was based on the 2022 research project “the study on the characterized source model for interplate earthquakes” by the Secretariat of Nuclear Regulation Authority, Japan. The waveform data were obtained from the Alaska Regional Network and the Alaska Volcano Observatory.