Seismic Velocity Structure of the Crust and Uppermost Mantle beneath the Northeastern Japan Arc Estimated from Receiver Functions

Seismic Velocity Structure of the Crust and Uppermost Mantle beneath the Northeastern Japan Arc Estimated from Receiver Functions
复制标题

从接收函数估算日本东北部弧下地壳和上地幔的地震速度结构

DOI:
10.4294/zisin1948.54.3_347
复制
发表时间:
2001
期刊:
影响因子:
--
通讯作者:
A. Hasegawa
A. Hasegawa
中科院分区:
--
文献类型:
--
作者:
Yasue Soda;T. Matsuzawa;A. Hasegawa

文献摘要

被引文献

相似文献

我们估计了地震速度结构的地壳和上地幔下的日本东北弧接收函数。针对传统的接收机函数分析方法存在的问题,本文首先对这些方法进行了改进。1)虽然水位法常用于稳定接收函数,但这种方法有时会使结果产生相当大的失真。此外,在选择最合适的水位方面没有客观标准。本文采用汉宁窗平滑频谱的方法进行稳频,因为它在频域FFT分析中的意义比水位法更清楚,对接收函数的失真也更小。2)接收器函数的时域叠加是非常普遍的,但是该过程需要许多波形来获得良好的S/N。此外,从这个过程中得到的结果是不是最小二乘估计。本文提出了一种与加权最小二乘法相对应的新的频域叠加方法,它能根据信噪比对数据进行适当的处理。3)通常,即使采用最小粗糙度约束,由接收函数分析估计的结构也具有较差的唯一性。我们已经开发了一种反演方法,其中的解决方案是约束到初始模型结构。如果我们使用一个适当的初始结构,满足直达波旅行时,预计从反演估计的结构也满足旅行时在相当程度上。我们将这些方法应用于日本东北部弧形宽带台站记录的地震图,以估计台站下方的结构。主要结果如下:1)在日本东北弧的东部,识别出了康拉德不连续面,其深度与Zhao等(1990)的结果一致。2)HSK和DIT台下地壳存在低速带,低速带位于HSK台下Conrad面正下方和DIT台下莫霍面正上方。3)KGJ站下地幔上部存在LV带。这一结果与Nakajima等人(2001 b)的结果一致,但我们的结果表明LV带薄至约5 km。4)在HOJ站下面,表层的速度非常低。
We estimated the seismic velocity structure of the crust and uppermost mantle beneath the northeastern Japan arc using receiver functions. Since the conventional methods for analyzing the receiver functions have several problems, we improved the methods at first. 1) Although the water level method is frequently used in stabilizing the receiver functions, this method sometimes distorts the results considerably. Moreover there are no objective criteria in selecting the most appropriate water level. In this paper, we adopt smoothing of spectra using Hanning window for the stabilization because its meaning in the frequency-domain FFT analysis is clearer than the water level method and it causes less distortion of the receiver functions. 2) Time-domain stacking of the receiver functions is very popular but this procedure needs many waveforms to obtain good S/N. Moreover, the result obtained from this procedure is not the least-squares estimation. We formulate a new frequencydomain stacking procedure that corresponds to the weighted least-squares method and can treat the data properly according to their S/N. 3) Usually, the structure estimated from the receiver-function analysis has poor uniqueness even if the minimum-roughness constraint is adopted. We have developed an inversion method in which the solution is constrained to the initial model structure. If we use a proper initial structure which satisfies the direct wave travel times, it is expected that the structure estimated from the inversion also satisfies the travel times to a considerable extent. We applied these methods to the seismograms recorded at broadband stations in the northeastern Japan arc in order to estimate the structure beneath the stations. The main results are as follows: 1) In the east of the northeastern Japan arc, the Conrad discontinuity is identified and its depth is consistent with the result of Zhao et al. (1990). 2) There exist low velocity (LV) zones in the lower crust beneath stations HSK and DIT; the LV zones are located just below the Conrad beneath HSK, and just above the Moho beneath DIT. 3) There exists a LV zone in the uppermost mantle beneath station KGJ. This result is consistent with Nakajima et al. (2001b) but our result indicates that the LV zone is as thin as around 5 km. 4) Beneath station HOJ the velocity in the surface layer is extremely low.