三种方法分析2017年宁夏固原M_(S)4.6地震的震源机制解

Focal mechanism solution of Guyuan,Ningxia M_(S)4.6 earthquake in 2017 determined by three methods

利用宁夏区域地震台网的波形数据,基于三种不同速度模型分别使用相对稳定的Hash方法、Snoke方法及gCAP方法计算2017年9月2日宁夏固原M_(S)4.6地震的震源机制解,并根据这些结果给出了该地震的震源机制中心解。结果表明,该地震的震源机制中心解为节面Ⅰ:走向:41°,倾角:79°,滑动角:-175°;节面Ⅱ:走向:310°,倾角:85°,滑动角:-11°。基于速度模型3的Hash方法结果和中心解结果之间的最小空间旋转角数值相对最小,震源机制解参数最接近中心解的结果。速度模型分层越精细,结果的精度相对越高,其中Hash方法及Snoke方法的反演结果受速度模型的影响较为明显,与Snoke方法相比,Hash方法对台站分布的要求较低,但对波形质量要求较高,要确保整个波形段和剪切波段具有一定范围的高信噪比。在实际应用中,对于台站方位角覆盖不是很好的区域地震台网来说,或者地震震中位于台站分布相对稀少的地区且可利用台站数目不是很多的情况下,在精细速度模型和波形信噪比的阈值等参数设置较为准确的情况下,可以考虑使用Hash方法来丰富计算结果。

Based on the waveform data of Ningxia regional seismic network and three different velocity models,the focal mechanism solution of Guyuan,Ningxia M_(S)4.6 earthquake on September 2,2017 was calculated by using three methods,i.e.,the Hash method,the Snoke method,and the gCAP method.Then the central focal mechanism solution of the earthquake was given according to the results.The results show that the central focal mechanism solution is that:nodal plane I:strike 41°,dip 79°,slip angle-175°;the nodal plane II:strike 310°,dip 85°,slip angle-11°.The minimum 3-D rotation angle between the results of Hash method and the central solution based on velocity model 3 is the lowest,and their focal mechanism solution parameters are the closest.The fine layered velocity structure model can help to improve the precision of the results,and the inversion results of Hash method and Snoke method are obviously influenced by the velocity structure model.Compared with the Snoke method,Hash method has lower requirements for the station distribution,but higher requirements for the waveform quality,and it requires that the whole waveform and the shear wave band both have a high signal-to-noise ratio in a certain range.In reality,for regional seismic networks with poor azimuth coverage of stations,or for the epicentral area with relatively few stations,the Hash method can be used to enrich the calculation results when the parameters of the fine velocity model and the threshold of waveform signal-to-noise ratio are set accurately.