The Q structure of the upper mantle: Constraints from Rayleigh wave amplitudes
The Q structure of the upper mantle: Constraints from Rayleigh wave amplitudes
复制标题
上地幔的Q结构：瑞利波振幅的限制
DOI：
10.1029/2001jb000257
复制
发表时间：
2002
影响因子：
--
通讯作者：
J. Woodhouse
中科院分区：
文献类型：
--
作者：
N. Selby;J. Woodhouse
[1]尽管地球同一部分的速度模型越来越精确和复杂，但对上地幔的三维剪切衰减（Qμ）结构仍然知之甚少。本文提出了一套基于70 ~ 170 s周期范围内的小弧瑞利波和主弧瑞利波振幅测量的上地幔三维Qμ模型。模型显示，低衰减区域位于大陆下方，深度约为300公里，而高衰减区域与海洋和板块边界有关，特别是在顶部200公里。在~ 350 km以下，数据对衰减结构的灵敏度迅速下降。众所周知，表面波的振幅会受到除衰减以外的其他因素的强烈影响，特别是由于弹性结构和震源机制误差引起的聚焦。我们通过讨论Selby和Woodhouse[2000]研究结果的含义来考虑聚焦的影响。与使用相速度图的聚焦预测的比较使我们能够识别衰减分布图中由于聚焦而产生的特征。通过在模型反演中加入源项，研究了源对观测到的衰减分布的影响。我们发现，成功地将每个频率的数据组合成一个三维模型需要该源项必须随频率变化，这表明对每个事件的标量矩的简单修正不能解释观测结果。虽然聚焦和源的影响可能是显著的，但我们发现观测到的衰减模式在8度以下是稳健的；然而，波长较短的结构会受到其他控制因素的强烈影响。最后，我们利用其中一个模型预测了250-s瑞利波的2度衰减规律。观测到的模式表明，现有的正模态观测可以用350公里深度以上的结构来解释。
[1] The three-dimensional shear attenuation, or Qμ, structure of the upper mantle is still poorly understood despite the increasing accuracy and sophistication of velocity models of the same part of the Earth. Here we present a set of three-dimensional Qμ models of the upper mantle based on amplitude measurement of minor and major arc Rayleigh waves in the period range 70–170 s. The models show that areas of low attenuation underlie continents to a depth of around 300 km and areas of high attenuation are associated with oceans and plate boundaries, particularly in the top 200 km. Below ∼350 km, the sensitivity of the data to attenuation structure decreases rapidly. It is well known that the amplitudes of surface waves can be strongly influenced by factors other than attenuation, in particular, focusing due to elastic structure and errors in source mechanisms. We consider the effects of focusing by discussing the implications of the results of Selby and Woodhouse [2000]. Comparisons with focusing predictions using phase velocity maps allow us to identify features due to focusing in maps of attenuation distribution. The influence of the source on the observed attenuation distribution is investigated by including a source term in the model inversion. We find that successfully combining the data at each frequency into one three-dimensional model requires that this source term must vary with frequency, suggesting that a simple correction to the scalar moment of each event cannot explain the observations. Although the effects of focusing and the source can be significant, we find that the observed attenuation pattern is robust up to degree 8; however, shorter-wavelength structure can be strongly influenced by other controls. Finally, we use one of the models to predict the degree 2 attenuation pattern for 250-s Rayleigh waves. The observed pattern suggests that existing normal mode observations may be explainable in terms of structure above 350 km depth.