论文部分内容阅读
为了提高相对定位精度,我们用空间上变化的台站项对南加州地震台网(SCSN)1975～1998年记录的297 400个地震进行了重新定位。我们的方法应用了已有的 SCSN 的 P 波和 S 波到时、平滑一维速度模型和基于 L1范数的迭代网格搜索方法。通过计算作为震源位置函数而连续变化的台站到时校正,我们对三维结构应用了经验校正。通过将地震活动性在天然邻区进行棋盘形布量来平滑相邻地震的残差,获得每个地震的台站项,然后进行迭代直到得到一组稳定的地震位置和台站项。该方法得到的相对定位精度在局部上与主震方法相当,但同样能用于较大范围。我们最终定位的中值估计标准误差,水平位置上为328m,深度上为741m。我们的定位显示出,与 SCSN 目录相比,离散要小得多,特别是在深度上,而且显示出能说明断层结构的排成线性和平面特征的更大趋势。我们的结果和先前 SCSN 用联合震源一速度反演技术重新定位的结果相当,且在某些情况下更好。白天与夜间的事件对比图能分辨出天然地震和人工地震群。我们没有观察到地震活动的最大深度和地表地质间有简单关系。
To improve the relative positioning accuracy, we repositioned 297 400 earthquakes recorded by the SCSN from 1975 to 1998 with spatially varying stations. Our method uses the existing SCSN P-wave and S-wave arrival, smooth one-dimensional velocity model and iterative grid search method based on L1 norm. By calculating station-to-station corrections that vary continuously as a function of source location, we apply empirical corrections to the three-dimensional structure. The stationarity of each earthquake is obtained by smoothing the residuals of the adjacent earthquakes by applying seismicity to the tessellated volume in the natural neighborhood, then iterating until a stable set of seismic locations and stations are obtained. The relative positioning accuracy obtained by this method is comparable with that of the main shock locally, but it can also be used in a large area. The median error of our final location estimate is 328 m at the horizontal position and 741 m at the depth. Our positioning shows that the dispersion is much smaller than the SCSN catalog, especially at depth, and shows a larger tendency to account for the linear and planar features of the fault structure. Our results are comparable to those of previous SCSN relocations using the Joint Source-Velocity Inversion technique, and in some cases are better. Day and night comparison of events can identify natural earthquakes and artificial earthquake swarm. We have not observed a simple relationship between the maximum depth of seismic activity and surface geology.