Dr. Chun-Hsiang Kuo’s (NCREE) email (2018-08-03):
The Mw bias of BATS in the period 2008 to 2014 is believed caused by the technical problem of Antelope.
Ms. 李文蕙’s (IES) email (2018-08-17):
附件為BATS CMT解的規模異常期間(2007/01/01~2014/03/01)
From SSC’s earthquake catalog report:
本目錄考慮通用性與時間性等因素,採用全球地震觀 測網的全球地震張量解(Global Centroid Moment Tensor, GCMT)觀測結果做 為震矩規模的依據,經由回歸式將各種規模轉換成全球地震觀測網的震矩 規模\(M_W\)(Harvard),各種規模所對應的回歸式如表 4.2 所示。 本目錄蒐集之規模最多可能有三大類,分別為全球地震張量解(GCMT)的震矩規模 \(M_W\)(Harvard)、經回歸式轉換過的台灣寬頻地震網(Broadband Array in Taiwan for Seismology, BATS) BATS \(M_W\)(Harvard-BATS)與原始目錄規模轉 換的\(M_W\)(原始規模來源可能為CWB、ISC或CSN ),若同一筆地震擁有多 種規模的時,規模選用規則參照轉換公式的複雜程度與標準差大小,公式 越簡易、標準差越小或越有足夠樣本可以分析轉換結果者越優先使用,選 用順序依照下列(1)到(5)排序,(1)全球地震觀測\(M_W\)(Harvard)(不須轉換)、 (2)由BATS規模回歸的\(M_W\)(Harvard-BATS)、(3)由CWB 規模回歸的 \(M_W\)(Harvard-CWB)、(4)由ISC規模回歸的\(M_W\)(Harvard-ISC)、(5)由CSN 規模回歸的\(M_W\)(Harvard-CSN) 。
Table4.2 in SSC EQ report
鄭世楠等人(2010)
圖8顯示1996-2002年中央研究院地球科學研究所寬頻地震網 (IES_BAT)使用CMT方法定出的震矩規模(MW-IES)與全球網CMT方法定出震矩規 模(MW-HRV)關係圖。
Cheng et al.(2010) BATS-GCMT MW
The convertion equation for BATS MW to GCMT MW is not affected by the technical problem of Antelope since Cheng et al.(2010) only use BATS MW from 1996 to 20026. There’re total of 98 events in GMC’s GM database using the MW converted from BATS’s MW catalog. During the affected period (2007-01-01 ~ 2014-03-01), there’re 36 events in GMC’s GM database, these events have the problem.
The following table shows the event number using in each GMPE, the number of BATS(to GCMT)Affted is the event number that have affated by the technical problem of Antelope in BATS. That about ** 6%, 8% and 4% ** of events respectively used in ajdustment and development GMPEs in GMC model.
| Events | Adjusted GMPE | Chao18 | Phung18 |
|---|---|---|---|
| Total | 104 | 198 | 158 |
| BATS(to GCMT)Affted | 6 | 15 | 6 |
| BATS(to GCMT)OK | 19 | 35 | 29 |
| Convet | 59 | 44 | 34 |
| GCMT | 20 | 104 | 89 |
The following plot show the ground motion records that was used by adjusted and new developed GMPEs in GMC model, the colored point are the selected records, the gray points are whole GMC GM database.
The following table shows the event number using in each GMPE, the number of BATS(to GCMT)Affted is the event number that have affated by the technical problem of Antelope in BATS. That about 6%, 14% and 6% of events respectively used in ajdustment and development GMPEs in GMC model.
| Events | Adjusted GMPE | Chao18 | Phung18 |
|---|---|---|---|
| Total | 49 | 118 | 48 |
| BATS(to GCMT)Affted | 3 | 17 | 3 |
| BATS(to GCMT)OK | 10 | 25 | 9 |
| Convet | 4 | 13 | 6 |
| GCMT | 32 | 63 | 30 |
The following plot show the ground motion records that was used by adjusted and new developed GMPEs in GMC model, the colored point are the selected records, the gray points are whole GMC GM database.
The impact of the BATS MW issue caused by the technical problem of Antelope seems not impact the GMC model, becase the events with problematic MW only take a relative small amont of the events that are use in the adjustment and development of GMPEs for the GMC model.