5:15 PM - 6:45 PM
[U15-P14] Hypocenter distribution of the Noto Peninsular Earthquake by a hypocenter location method using multiple template earthquakes (hypoTD)
Keywords:hypoTD, hypoDD, relative hypocenter location, Noto Peninsula Earthquake
1. Introduction It is well established that there are major differences in the underground seismic velocity structure between ocean and land. Since the Noto Peninsula earthquake occurred at boundary region between ocean and land areas, this heterogeneity may be a major source of error in hypocenter location. Horiuchi et al.(2022) developed a method to locate hypocenter using a large number of template earthquakes. Horiuchi and Sato (present meeting) show the method to re-located hypocenters of earthquakes in and around Japan Islands by using arrival time data of . Japan Meteorological Agency's(JMA’s) unified catalogue. In the following, this hypocenter location method will be referred to as hypoTD. In this report, we introduce hypoTD and show the relocated hypocenter distribution of the Noto Peninsula earthquake.
2. HypoTD In the conventional hypocenter location method using template earthquakes, one template earthquake is used to locate multiple earthquakes. HypoTD uses a large number of template earthquakes, 50 in this study, to locate each earthquake. JMA unified hypocenter catalogue includes hypocenters determined through automatic processing, and may contain data with low accuracy in arrival time pickings. HypoTD has a function to detect incorrect pickings, as shown in https://www.homeseismo.com/html/sotaijido.html. We didn’t use arrival times which are considered to be low accuracy. In the case of a large earthquake occurrence, the S wave initial motion is buried in the P wave coda wave and is difficult to pick correct arrival in principle. Because hypoTD can provide a stable solution even using only P-wave arrival time data, we use only P-wave data for earthquakes with a magnitude greater than 5.0.
3. Results We created a three-dimensional display software that shows the hypocenter distribution. It can compare hypocenter distribution by hypoTD and that by JMA catalogue for earthquakes occurring throughout Japan Islands. We got a conclusion that hypoTD can locate accurate hypocenters. Additionally, we compare hypocenter distributions by hypoDD (Waldhauser & Ellsworth (2002) with hypoTD for earthquakes beneath the Tokyo metropolitan area, aftershocks of the 2004 Chuetsu Earthquake, and 2004 Off Tokaido Earthquake. These comparisons show that hypoTD can locate more accurate and stable hypocenters than those by hypoDD. The characteristics of the hypocenter distribution of the Noto Peninsula earthquake are as follows.1) An accurate hypocenter distribution was obtained by hypoTD. 2) Southwest side of the aftershock area has aftershocks that occur along a southeast-dipping plane, but the area of the northeastern part, aftershocks occur along a southeast-dipping plane that is perpendicular to the main fault. 3) We randomly created 100 sets of 50 template earthquakes and calculated 100 sets of the main shock hypocenters using only P waves so as to check its accuracy. As a result, the variations in the hypocenter locations of the 100 sets were very small (Fig.1). In a case when we use S wave arrivals, located depths become deep and the travel time residual becomes significantly large, suggesting the difficulty of S wave arrival time picking for large earthquakes.
4. Release of the software and computed results The hypoDD proposed by Waldhauser & Ellsworth (2000) is an excellent method for locating hypocenters that can eliminate the effects of heterogeneity, but it has limitations such as not being applicable to earthquakes that occur over a wide area. HypoTD does not have such restrictions and is expected to have the potential to be used in research on seismic activity in various regions in the future. We will publish the software of hypoTD, from the web sites of Home Seismometer Co., Ltd (https://www.homeseismo.com). In addition, we plan to make the website at Hakusan Kogyo Co., Ltd. (https://www.hakusan.co.jp), where one can see the distribution of hypocenters throughout the Japan Islans based on relative hypocenter determination over 22 years with using 3D display software.
2. HypoTD In the conventional hypocenter location method using template earthquakes, one template earthquake is used to locate multiple earthquakes. HypoTD uses a large number of template earthquakes, 50 in this study, to locate each earthquake. JMA unified hypocenter catalogue includes hypocenters determined through automatic processing, and may contain data with low accuracy in arrival time pickings. HypoTD has a function to detect incorrect pickings, as shown in https://www.homeseismo.com/html/sotaijido.html. We didn’t use arrival times which are considered to be low accuracy. In the case of a large earthquake occurrence, the S wave initial motion is buried in the P wave coda wave and is difficult to pick correct arrival in principle. Because hypoTD can provide a stable solution even using only P-wave arrival time data, we use only P-wave data for earthquakes with a magnitude greater than 5.0.
3. Results We created a three-dimensional display software that shows the hypocenter distribution. It can compare hypocenter distribution by hypoTD and that by JMA catalogue for earthquakes occurring throughout Japan Islands. We got a conclusion that hypoTD can locate accurate hypocenters. Additionally, we compare hypocenter distributions by hypoDD (Waldhauser & Ellsworth (2002) with hypoTD for earthquakes beneath the Tokyo metropolitan area, aftershocks of the 2004 Chuetsu Earthquake, and 2004 Off Tokaido Earthquake. These comparisons show that hypoTD can locate more accurate and stable hypocenters than those by hypoDD. The characteristics of the hypocenter distribution of the Noto Peninsula earthquake are as follows.1) An accurate hypocenter distribution was obtained by hypoTD. 2) Southwest side of the aftershock area has aftershocks that occur along a southeast-dipping plane, but the area of the northeastern part, aftershocks occur along a southeast-dipping plane that is perpendicular to the main fault. 3) We randomly created 100 sets of 50 template earthquakes and calculated 100 sets of the main shock hypocenters using only P waves so as to check its accuracy. As a result, the variations in the hypocenter locations of the 100 sets were very small (Fig.1). In a case when we use S wave arrivals, located depths become deep and the travel time residual becomes significantly large, suggesting the difficulty of S wave arrival time picking for large earthquakes.
4. Release of the software and computed results The hypoDD proposed by Waldhauser & Ellsworth (2000) is an excellent method for locating hypocenters that can eliminate the effects of heterogeneity, but it has limitations such as not being applicable to earthquakes that occur over a wide area. HypoTD does not have such restrictions and is expected to have the potential to be used in research on seismic activity in various regions in the future. We will publish the software of hypoTD, from the web sites of Home Seismometer Co., Ltd (https://www.homeseismo.com). In addition, we plan to make the website at Hakusan Kogyo Co., Ltd. (https://www.hakusan.co.jp), where one can see the distribution of hypocenters throughout the Japan Islans based on relative hypocenter determination over 22 years with using 3D display software.