【作者】 白志明；

【导师】 王椿镛；

【作者基本信息】 中国地震局地球物理研究所 ， 固体地球物理学， 2002， 博士

【摘要】 本文首先回顾了深地震测深(DSS)方法的发展。在分析深地震测深剖面层析成像基本理论的基础上提出了利用走时、振幅和重力数据，通过有限差分反演和射线反演方法对地壳上地幔结构进行层析成像的研究思路。 本文第三章研究了相对振幅比的获取方法，探讨了利用振幅和重力资料模拟来约束地壳速度结构的有效性，从而形成本文层析成像工作的总体结构。 本文的第四章通过一系列模型试验详细探讨了利用射线反演方法反演速度和深度参数的实际步骤，认为走时反演对速度层内顶部和底部速度变化的分辨能力很弱，仅仅依靠走时数据难以反演出层内顶部和底部的速度变化，故一般情况下应采用速度层底部和顶部速度相同的纵向均匀模型。在参数反演过程中应采用先反演速度参数、再反演深度参数的方案，这样的步骤使得反演结果较稳定，得到的解更为可信。 利用检测板试验的方法以及与其它二维和三维成像结果的对比验证了本文成像结果的可靠性。检测板试验的结果表明，HQ-13线速度地壳速度结构横向变化的最佳绝对分辨应在30km左右，该测线上界面形态横向变化的绝对分辨可以达到10km以内；取速度节点间距和界面深度节点间距均为20km的检测板试验表明，云南地区四条测线地壳内界面形态的横向分辨良好。 本文第五章利用有限差分反演和射线反演的方法获得了符离集—奉贤地震测深剖面(HQ-13线)精细的地壳上地幔结构剖面，揭示出下扬子地区一些重要的结构和动力学特征： 地壳速度结构在纵向上大致可分上地壳、中地壳和下地壳三部分，横向上可划分为6个块体，各块体的P波速度沿测线方向呈现高速—低速—高速的组合。上部地壳高速异常和低速异常分别与地表的隆起和坳陷高度一致，高速异常和低速异常的接触带往往与地表的断裂位置有良好的对应关系，沉积盖层底部的基底界面横向上起伏变化较大。中地壳的中间层速度分的基本特征类似上地壳底部，横向均匀性显著；中地壳下部速横向不均匀性明显，在灵壁隆起和湖苏隆起的下部有显著的高速异常。下地壳的两构造层速度差异不明显。莫霍界面深度为30～36km。上地幔顶部速度值为8.06～8.29km／s。 成像结果显示符离集大断裂、嵩沟坳陷南界断裂、洪泽-流均沟断裂等均延伸至上地幔，反之郯庐断裂带缺乏深大断裂构造的迹象。就地壳深部结构而言，苏北坳陷区板块的划分办法、各板块的接触边界以及相应的动力学背景，有进一步求证的必要。 根据所得结构剖面，分析了该测线经过区域的地震构造环境。镇江块体容易发生中强地震的解释是：由于地壳结构偏刚性及壳内界面整体上隆，块体中活动性显著的大断裂很可能将地壳深部或上地幔的能量源源不断地传输至中上地壳，在有利部位积聚并诱发地震。苏北坳陷、无锡块体和上海块体的下部存在低速、塑性的中间层。这些块体内部结构之间互动的位能容易被塑性中间层吸收，且由于区域构造活动不显著，从而不容易发生大的地震。 本文第六章对云南地区四条DSS剖面进行层析成像研究，得到以下一些重要认识： 云南地区地壳大致可分为包括沉积层和基底层的上地壳、包含两个弱反射层的中地壳以及下地壳构成。各个反射层的界面形态有相似性，各层的速度分布密切联系，存在显著的深浅构造的一致性，意味着浅部物质活动有深部背景，同时说明云南地区是典型的构造活动区。云南地区莫霍界面总体形态南端浅、北端深，地壳厚度约33刀～54刀km。上地慢速度明显偏低，可能与贯穿整个新生代的热过程相联系。 根据界面形态和速度异常分布特征，判别出云南地区断裂构造的深部特征。作为一级构造单元分界线的红河断裂、怒江断裂是超壳断裂，而昌宁－双江断裂显示为低角度铲式断层，中下地壳无与之对应的速度异常，可能意味着该断裂切割不深。澜沧江断裂南段为基底层断裂。中甸断裂、剑q断裂和丽江－宁落断裂组成的断裂系延伸至下地壳。 通过对龙陵地震区、思茅一普洱地震区、建水附近地震带以及剑川］附近等地震构造环境的分析，提出一个云南地区强烈地震孕育发生的模式：在历史上火山活动剧烈的地区，现今深部活动仍未停止，存在较强的应力作用。由于地表火山活动的停止，来自中下地壳和上地慢的大部分能量得不到释放，从而在有利的构造部位积聚，进而在流体的作用下，或在其它外动力诱发下，导致孕震系统失稳而发生地震。 本文第七章总结了本文的主要工作和意义，并指出一些需要深入研究的问题更多还原

【Abstract】 This paper reviews the development of the Deep Seismic Sounding (DSS) method firstly. Based on the analyzing of the DSS’s basic theory of tomography,an inversion scheme that can be described as the combination of the finite-difference inversion and ray inversion using traveltime,amplitude and gravity data is proposed.The third chapter of this paper explored the method of acquiring amplitude-ratio,and also the significance of the amplitude and gravity data for the probing of the crust and upper mantle’s structure is investigated. Thus the framework of inversion adopted by this paper is formed.Several numerical experiments are carried out to detect the actual steps and limitation for the velocity and depth parameter’s ray inversion in the fourth chapter,and the we find that:a. the resolution capability of the layer’s velocity change along both the top and bottom boundary is very poor. Thus generally the vertical homogeneous layer model whose top velocity equal to that of its bottom boundary should be considered. During the inversion process of velocity and depth parameter,the velocity parameter should be inversed firstly and then do the depth so as to assure the stability of the final results and the reliability of the solution.Take advantage of the checker-board-test method to test the horizontal resolution of velocity and depth parameter on HQ-13 profile and the four profiles of Yunnan region. It’s found that the best absolute lateral resolution of the velocity on the HQ-13 profile is about 30 kilometer,while that of the depth parameter on this profile can be smaller than 10 kilometer. Both of the velocity and depth parameter points are of 20 kilometers interval on the four profiles in Yunnan region,their lateral resolution are also detected by the checker-board-test method. It shows that the resolution of the crust interface’s shape on this four profiles is very good,nevertheless that of the velocity structure is relatively poor.The crust and upper mantle’s fine configuration on the HQ-13 profile is obtained by our tomography,which reveals some important structure and dynamic features of this region:The velocity structure of the crust consists of three layers,the upper crust,the middle crust and the lower crust. However it also can be divided into six lateral blocks,each of them ’s velocity is always higher or lower than that of its neighboring blocks. The high or low velocity anomalies’s locations are consistent with the uplift and depression respectively,and their contacting boundary always correspond the faults exposed on the ground’s surface. The basement interface that is the bottom interface of the sedimentary cover undulates strongly. The characters of velocity’s distribution within the mid-layer of the middle crust resemble that of the upper crust’s bottom layer,and the velocity strongly disturbed. The prominent higher velocity anomalies exist below the Linbi uplift and theHu-Su uplift within this layer. The velocity difference between the two layers of the lower crust is not obvious. The depth of Moho interface is about 30-36 km,and the velocity of the upper mantle is nearly 8.06-8.29 km/s.The final structure also shows that the Fuliji fault,the Songgou basin south fault,the Hongze-Liujungou fault among with other deep faults stretch into the upper mantle. While there is no distict evidence to prove the Tanlu fault’s noted deep and dynamic behavior within the lower crust or even on the upper mantle. Thus there still exists a necessary to research and check the understanding about the boundary’s location and the dynamic background.Base on the velocity profile,the seismotectonic environment along this profile is studied. The reason for the moderate and large earthquake of Zhengjiang block could be attribute as such features:Because of the structure’s more rigidity and the rising of the several interfaces in the crust,the energy from the deep crust of upper mantle is transported to the middle or upper crust persistently. Thus it always accumulates at some locations and this may更多还原