论文部分内容阅读
本文提出一种由初动到时数据确定康拉德界面和莫霍界面二维深度分布以及台站校正和震源参数的新方法。在本研究中,速度结构用双曲函数tanh来模拟,而各个界面的形状则用以未知参数为系数的切比雪夫函数级数来表达。反演时使用了Tarantola和Valette的非线性反演方法。对复杂地区计算地震射线路径和走时则利用了具有足够精度的射线追踪法。本方法使用国家灾害预防研究中心管理的地震台网中所收集到的地方震和爆炸的P波初动到时数据,以揭示日本关东-东海地区下部的地壳结构。本文所得的研究结果颇为重要,同时这些结果也表明了这种方法在研究地壳结构时的有效性。在有浅部结构资料的那些地方台站校正的分布是合理的。在东京湾和Omae角附近出现低速带。而在山区则出现高速带。康拉德界面的深度约为12—16km,横向变化较为平缓。相对来说,莫霍界面的深度横向变化较大,而且存在一个莫霍界面深度超过40km的狭小地带。除中部山区之外,本文所得到的地壳结构与前人用折射方法所得的结果相当一致。将我们的结果与Ishida和Hasemi用Aki和Lee的方法导出的第一层(0—32km)中的速度扰动相比,我们的工作对引起地壳速度异常的特征提供了技好的信息,而在Ishida和Hasemi的工作中已经不能对其作出约束了。
This paper presents a new method to determine the two-dimensional depth distribution of Conrad interface and Moho interface, as well as station calibration and source parameters from the time-of-arrival data. In this study, the velocity structure is modeled by the hyperbolic function tanh, while the shape of each interface is expressed by the Chebyshev series with unknown parameters as coefficients. Tarantola and Valette’s nonlinear inversion method was used for inversion. The calculation of seismic ray paths and travel times in complex areas uses ray tracing with sufficient accuracy. This method uses the first-arrival-time-of-arrival data of local earthquakes and explosions collected from seismograph networks managed by the National Center for Disaster Prevention Research to reveal the crustal structure in the lower part of Japan’s Kanto-Tokai area. The results obtained in this paper are quite important, and these results also show the validity of this method in studying the crustal structure. The distribution of station corrections is reasonable in those places with shallow structural data. Low speed belts appear near Tokyo Bay and Omae. In the mountains there is a high-speed zone. Conrad interface depth of about 12-16km, horizontal changes more moderate. Relatively speaking, the depth of the Moho interface varies greatly laterally, and there is a small area where the Moho interface depth exceeds 40km. Except for the central mountainous area, the crustal structure obtained in this paper is quite consistent with the previous results obtained by the refraction method. Compared with the velocity perturbations in the first layer (0-32 km) derived by Ishida and Hasemi using Aki and Lee’s method, our work provides good information about the characteristics that cause the anomaly of crustal velocity. However, Ishida and Hasemi’s work can no longer be bound by it.