论文部分内容阅读
在大多数地震面波的研究中,观测到的群走时被解释为全部由波沿传播路径产生的时间延迟,并且认为震源效应非常小而可忽略不计。这与观测的震相时间不同,因为对震源相进行校正一般认为是必须做的。一个重要的但尚未解决的问题是宽频带面波研究中对震源群走时(SGT)的忽略如何影响测量的群速度曲线和利用观测数据进行层析成像的精度。本文考虑SGT对基阶瑞利波群速度测量的影响,并论述其对周期(10～200s)、震源机制和震源深度的依赖关系。这些参数的变化强烈地影响SGT偏移的幅值和方位角图象,本文将展示这种相关特性的统计性。SGT对于周期小于75s和震源深度小于25km的浅震是可以忽略的。对于较长周期和较深的地震事件,平均的SGT校正在幅值上将大于10s,在大陆尺度的研究中,相当于1％～2％的群速度扰动。估计了在跨越欧亚大陆的瑞利波群速度图象反演中由于未作SGT校正引起的偏差。图象的最大扰动(周期为50s时达到1％～2％,周期为100s时达到5％)出现在射线较稀疏的大陆边缘。根据这些结果,即对邻近的波路径(簇)的统计估计结果,以及SGT校正对地震深度非常敏感的结果,我们的结论是:SGT对群速度层析成像的影响在周期小于75s和浅源地震时是可以忽略不计的。虽然对震源深度大于25km的地震,在较长周期处的校正是明显的,原则上必须进行,但目前矩心矩张量解和群速度测量的固有精度使得这些校正对现今群速度层析成像的研究实际上不是必须的。
In most studies of seismic surface waves, the observed group travel time is interpreted as the total time delay caused by the wave propagation path, and the source effect is considered to be negligible. This is not the same as the observed seismic facies time, since correction of the source phase is generally considered to be a must. An important but unresolved issue is how the omission of SGT in wideband surface wave studies affects the measured group velocity curve and the accuracy of tomographic imaging using the observed data. In this paper, we consider the influence of SGT on the velocity measurement of the fundamental Rayleigh wave group, and discuss the dependence on the periodicity (10 ~ 200s), focal mechanism and focal depth. Changes in these parameters strongly affect the magnitude and azimuthal image of the SGT shift, and this article shows the statistical nature of this correlation. The SGT is negligible for shallow earthquakes with periods less than 75 s and focal depths less than 25 km. For longer periods and deeper seismic events, the average SGT correction will be greater than 10s in magnitude, which is equivalent to 1% -2% of the group velocity perturbation in continental scale studies. The deviations due to no SGT correction were estimated during the image retrieval of Rayleigh wave velocities across the Eurasian continent. The largest perturbation of the image (1% -2% at 50s and 5% at 100s) occurs at the edge of the sparsely-contoured land. Based on these results, that is, the statistical estimation results of adjacent wave paths (clusters) and the results of SGT correction very sensitive to the depth of the earthquake, we conclude that the effect of SGT on group velocity tomography is less than 75 s in period and shallow source Earthquake is negligible. Although the correction over longer periods is obvious for earthquakes with focal depth greater than 25 km, it must in principle be done, but the inherent accuracy of moment-centric moment-tensor solutions and group-velocity measurements makes the correction of current group velocity tomography In fact not necessary.