[S08P-07] ニュージーランド南島北西部、反転テクトニクス地域における中・小規模断層の活動評価
Introduction
In New Zealand, the Pacific plate is subducting beneath the Australian plate beneath North Island and the northern part of the South Island. A compressional stress field with ESE-WNW direction is dominant in the northwestern part of the South Island (e.g., Townend et al., 2012). In the northwestern part of the South Island, many normal faults were formed during the opening of the Tasman Sea (Ghisetti and Sibson, 2006), and these fault planes are presently active as a reverse fault (tectonic inversion fault) (Ghisetti et al., 2014). Earthquakes occur less frequently in the northwestern part of South Island than in many other parts of New Zealand. However, large earthquakes (e.g., the 1929 Buller earthquake, Mw 7.3; the 1968 Inangahua earthquake, Mw 7.2) have occurred in this area. Therefore, assessing the likelihood of the fault slip in this area is important for understanding seismic activity in this area. We evaluated the likelihood of the fault slip for large events (e.g., the Buller earthquake; the Inangahua earthquake) using the Slip Tendency (ST) method (Morris et al., 1996) (e.g., Tagami et al., JpGU 2023). The faults were mostly favorable to slip in the stress field in the northwestern part of South Island. Previous studies had estimated focal mechanisms for most large earthquakes, but the actual fault plane had not been determined meaning that without additional information. Therefore, in this study, we detected moderate to small earthquakes using matched filter detection methods to try to estimate the trend of fault planes. If the aftershocks had a planar distribution, we assumed that the earthquakes occurred on single fault planes. Thus, the earthquake distribution can help discriminate the fault planes from the two nodal planes of the focal mechanism.
Data and Methods
We used matched filter detection methods (EQcorrscan; Chamberlain et al., 2018) to detect events similar to the aftershocks of template events. First, to make the template, we picked P and S arrival times manually using the WIN program (Urabe and Tsukada, 1992); events with quality A or B focal mechanisms estimated by the HASH program (Hardebeck and Shearer, 2002) are employed as template events. Second, we converted WIN wave format data to miniSEED wave format data. Then, we detected the events using the matched filter detection methods; the absolute location of the detected events was calculated using NonLinLoc (Lomax et al., 2000); also, the relative location was determined with GrowClust (Trugman and Shearer, 2017) using differential travel time calculated by cross-correlation of waveforms.
Results
We made 52 templates from May 2013-from May 2013 to December 2013. These 52 earthquakes were also registered in the GeoNet Aotearoa New Zealand Earthquake Catalogue (GNS Science., 1970). Most of the focal mechanism solutions determined in this study were reverse fault type, with some strike-slip mechanisms. Of the 1447 events detected by EQcorrscan, 1383 events, including 1304 new events, were located by NonLinLoc. We selected the fault planes of template focal mechanisms by fitting planes to the hypocenter distributions of our relocated seismicity. Finally, we defined fault planes for 25 minor earthquakes. The ST method (Neves et al., 2009) was also used to evaluate whether the fault is optimal to slip in the stress field. Half of the defined faults had dips of 30-60° and were favorable to slip in the stress field. However, six defined faults had dips of 60-90° and were unfavorable to slip in the stress field. These suboptimal fault clusters were distributed in the northern part of the White Creek fault and around Maruia Springs: an area with clear hot springs activity. In addition, previous studies observed the high Vp/Vs continuously from the top of the subducting Pacific plate to the crust in the northwestern part of the South Island (e.g., Okada et al., 2019). Thus, the possible influence of pore fluid supplied from deep underground may have triggered the local unfavorable fault activities.
In New Zealand, the Pacific plate is subducting beneath the Australian plate beneath North Island and the northern part of the South Island. A compressional stress field with ESE-WNW direction is dominant in the northwestern part of the South Island (e.g., Townend et al., 2012). In the northwestern part of the South Island, many normal faults were formed during the opening of the Tasman Sea (Ghisetti and Sibson, 2006), and these fault planes are presently active as a reverse fault (tectonic inversion fault) (Ghisetti et al., 2014). Earthquakes occur less frequently in the northwestern part of South Island than in many other parts of New Zealand. However, large earthquakes (e.g., the 1929 Buller earthquake, Mw 7.3; the 1968 Inangahua earthquake, Mw 7.2) have occurred in this area. Therefore, assessing the likelihood of the fault slip in this area is important for understanding seismic activity in this area. We evaluated the likelihood of the fault slip for large events (e.g., the Buller earthquake; the Inangahua earthquake) using the Slip Tendency (ST) method (Morris et al., 1996) (e.g., Tagami et al., JpGU 2023). The faults were mostly favorable to slip in the stress field in the northwestern part of South Island. Previous studies had estimated focal mechanisms for most large earthquakes, but the actual fault plane had not been determined meaning that without additional information. Therefore, in this study, we detected moderate to small earthquakes using matched filter detection methods to try to estimate the trend of fault planes. If the aftershocks had a planar distribution, we assumed that the earthquakes occurred on single fault planes. Thus, the earthquake distribution can help discriminate the fault planes from the two nodal planes of the focal mechanism.
Data and Methods
We used matched filter detection methods (EQcorrscan; Chamberlain et al., 2018) to detect events similar to the aftershocks of template events. First, to make the template, we picked P and S arrival times manually using the WIN program (Urabe and Tsukada, 1992); events with quality A or B focal mechanisms estimated by the HASH program (Hardebeck and Shearer, 2002) are employed as template events. Second, we converted WIN wave format data to miniSEED wave format data. Then, we detected the events using the matched filter detection methods; the absolute location of the detected events was calculated using NonLinLoc (Lomax et al., 2000); also, the relative location was determined with GrowClust (Trugman and Shearer, 2017) using differential travel time calculated by cross-correlation of waveforms.
Results
We made 52 templates from May 2013-from May 2013 to December 2013. These 52 earthquakes were also registered in the GeoNet Aotearoa New Zealand Earthquake Catalogue (GNS Science., 1970). Most of the focal mechanism solutions determined in this study were reverse fault type, with some strike-slip mechanisms. Of the 1447 events detected by EQcorrscan, 1383 events, including 1304 new events, were located by NonLinLoc. We selected the fault planes of template focal mechanisms by fitting planes to the hypocenter distributions of our relocated seismicity. Finally, we defined fault planes for 25 minor earthquakes. The ST method (Neves et al., 2009) was also used to evaluate whether the fault is optimal to slip in the stress field. Half of the defined faults had dips of 30-60° and were favorable to slip in the stress field. However, six defined faults had dips of 60-90° and were unfavorable to slip in the stress field. These suboptimal fault clusters were distributed in the northern part of the White Creek fault and around Maruia Springs: an area with clear hot springs activity. In addition, previous studies observed the high Vp/Vs continuously from the top of the subducting Pacific plate to the crust in the northwestern part of the South Island (e.g., Okada et al., 2019). Thus, the possible influence of pore fluid supplied from deep underground may have triggered the local unfavorable fault activities.