1:45 PM - 2:00 PM
[SSS07-25] Earthquake source parameters and their implications for fluid movement in the northern Hikurangi subduction zone, New Zealand
Introduction
Beneath the North Island of New Zealand, the Pacific plate is subducting under the Australian plate along the Hikurangi Trough, forming the Hikurangi subduction zone. In the Hikurangi subduction zone, some recent studies suggest that fluid movement from the oceanic crust to the plate boundary or into the Australian plate impacts various seismic activities such as SSEs (slow slip events), tremors, and earthquake swarms. A fundamental limitation in understanding how fluid movement affects slow earthquakes and regular earthquakes is the variable quality of earthquake hypocenter locations on and around the Hikurangi subduction zone. The main objective of this study is to clarify how regular earthquakes are related to fluid movement and slow earthquakes in the northern Hikurangi subduction zone. As a start, we classified the earthquakes to be analyzed into those occurring within the Australian plate (AUS), within the Pacific plate (PAC), or at the interplate (INT), and examined the spatio-temporal distribution of the source parameters of these earthquakes.
Data and Method
We selected all earthquakes in the GeoNet catalog of M > 3 occurring between 2003 and 2020 within a region of 37-39.5°S and 177.2-179.5°E and shallower than 150 km for analysis. Data from GeoNet broadband (HH) and short-period (EH) seismometers were used for the study.
The classification into AUS, INT, and PAC mainly consists of the four steps: (1) classification based on the hypocenters of GeoNet catalog, (2) classification based on the re-determined hypocenters by this study, (3) classification based on the focal mechanisms from the P-wave polarity, and (4) classification based on the cross-correlation of the waveforms. The number of earthquakes classified in each step was 2,671 in the first step, 507 in the second step, 11 in the third step, and 22 in the fourth step.
Result and Discussion
PAC earthquakes had a wide distribution, while AUS and INT earthquakes were distributed locally. The AUS and INT earthquakes in Tolaga Bay are located near tremors, and the timing of earthquake and tremor occurrences is synchronous. Furthermore, at the GNSS stations near Tolaga Bay, the timing of the eastward transient displacements, which are caused by the SSE, is sometimes synchronous with the timing of the AUS and INT earthquakes. This synchronization can be explained by fluid movement. That is, fluid movement from the oceanic crust to the plate boundary or upper plate before and after the SSE, as proposed by Nishikawa et al. (2021), may have triggered SSEs, tremors, and AUS or INT earthquakes.
Finally, we estimated the corner frequency (fc) for all the earthquakes of 3 < M < 4, and examined the spatial distribution of M0fc3, which is proportional to the stress drop, and found that the closer the PAC hypocenter is to the upper surface, the smaller M0fc3 becomes, or the lower the stress drop becomes. This result suggests that fluid pressure is higher in the shallow part of the subducting crust, which can be explained by considering that fluid movement to be shallow, as proposed by Warren-Smith et al. (2019).
Beneath the North Island of New Zealand, the Pacific plate is subducting under the Australian plate along the Hikurangi Trough, forming the Hikurangi subduction zone. In the Hikurangi subduction zone, some recent studies suggest that fluid movement from the oceanic crust to the plate boundary or into the Australian plate impacts various seismic activities such as SSEs (slow slip events), tremors, and earthquake swarms. A fundamental limitation in understanding how fluid movement affects slow earthquakes and regular earthquakes is the variable quality of earthquake hypocenter locations on and around the Hikurangi subduction zone. The main objective of this study is to clarify how regular earthquakes are related to fluid movement and slow earthquakes in the northern Hikurangi subduction zone. As a start, we classified the earthquakes to be analyzed into those occurring within the Australian plate (AUS), within the Pacific plate (PAC), or at the interplate (INT), and examined the spatio-temporal distribution of the source parameters of these earthquakes.
Data and Method
We selected all earthquakes in the GeoNet catalog of M > 3 occurring between 2003 and 2020 within a region of 37-39.5°S and 177.2-179.5°E and shallower than 150 km for analysis. Data from GeoNet broadband (HH) and short-period (EH) seismometers were used for the study.
The classification into AUS, INT, and PAC mainly consists of the four steps: (1) classification based on the hypocenters of GeoNet catalog, (2) classification based on the re-determined hypocenters by this study, (3) classification based on the focal mechanisms from the P-wave polarity, and (4) classification based on the cross-correlation of the waveforms. The number of earthquakes classified in each step was 2,671 in the first step, 507 in the second step, 11 in the third step, and 22 in the fourth step.
Result and Discussion
PAC earthquakes had a wide distribution, while AUS and INT earthquakes were distributed locally. The AUS and INT earthquakes in Tolaga Bay are located near tremors, and the timing of earthquake and tremor occurrences is synchronous. Furthermore, at the GNSS stations near Tolaga Bay, the timing of the eastward transient displacements, which are caused by the SSE, is sometimes synchronous with the timing of the AUS and INT earthquakes. This synchronization can be explained by fluid movement. That is, fluid movement from the oceanic crust to the plate boundary or upper plate before and after the SSE, as proposed by Nishikawa et al. (2021), may have triggered SSEs, tremors, and AUS or INT earthquakes.
Finally, we estimated the corner frequency (fc) for all the earthquakes of 3 < M < 4, and examined the spatial distribution of M0fc3, which is proportional to the stress drop, and found that the closer the PAC hypocenter is to the upper surface, the smaller M0fc3 becomes, or the lower the stress drop becomes. This result suggests that fluid pressure is higher in the shallow part of the subducting crust, which can be explained by considering that fluid movement to be shallow, as proposed by Warren-Smith et al. (2019).