论文中英文摘要

论文中英文摘要 论文中英文摘要 ,InSAR , .... Niudown.COM Niudown.COM 2 Niudown.COM Niudown.COM 3 作者姓名：陈国浒 论文题目：长白山天池火山形变监测与模拟研究 作者简介：陈国浒，男，1981年10月出生，2004年7月师从于中国地震局地质研究所单新建研究员，于2007年7月获硕士学位。 中 文 摘 要 本文结合差分合成孔径雷达干涉测量技术（D-InSAR）和计算机数值模拟技术对长白山地区的形变监测进行了研究。采用成熟的D-InSAR技术，获取了位于长白山东南侧的间白山火山1995~1998年间的干涉形变场；在长白山地区架设角反射器，开展 CR-InSAR技术在长白山地区的试验研究；利用2002~2003年的GPS和水准获取的形变数据采用Mogi点源模型和Okada拉张位错源模型进行反演，获取了长白山地区岩浆囊 的最佳参数；最后利用反演结果正演了D-InSAR获取的形变场；分析了研究区火山活 动的规律及其与断裂、地震活动等的关系，对研究区火山活动的态势和危险性作了简单 评估。在吸收和借鉴前人工作成果的基础上，本论文取得了如下的结果和认识： , 地形变是监测岩浆活动和预测火山喷发的主要手段之一，虽然一些地体活动和形变 在喷发前并没有处于峰值点，但多数火山在喷发前确有很明显的地形变。精确的地表 形变监测可用于火山活动和火山爆发预测等方面的研究。地形变数据的数值模拟可以提 供关于火山岩浆囊（规模、压力、形态和深度等）及其相关的水热系统的地下构造方面 的重要信息。 本论文详细介绍和讨论了目前火山形变监测手段的应用现状和不足，提出InSAR技术在火山区形变监测的优势，并由此提出了本文的选题意义和科学目标。 , “九五”、“十五”期间国内学者对火山活动开展了比较全面的研究，包括火山地 质、地球物理、形变监测、矿物学、岩石学、地球化学、年代测定和火山活动的环境效 应、灾害预测与监控等。研究表明：长白山火山正处于由初始扰动状态向动荡状态演化阶段，是我国目前最具喷发危险性的火山。 通过对长白山地区的地质构造背景和研究进展的分析，为我们后续开展工作提供借 鉴、参考，并能对本文研究得出的结论进行检验和论证。 , D-InSAR1995~1998 长白山火山地区地形复杂、气候条件恶劣，目前常规的火山形变监测手段（水准测 量和GPS形变监测）很难从较大区域来研究火山的形变特征，InSAR技术凭借其独特的优势可以直接获取大范围的、连续空间覆盖的火山形变的定量化基础数据。 本论文分别对2景ENVISAT数据（20041015、20051104）和5景JERS-1数据（19930615、19940829、19950816、19980707、19980820）进行了分析处理。其中ENVISAT数据因为相干性差而未能获得形变场信息。对JERS-1数据进行处理，获取了一系列干 涉图、差分干涉图、相干性图像和强度图等产品。其中19950816/19980820这对JERS数据的计算结果比较理想，其差分干涉图显示在1995~1998年期间长白山天池火山东南角位于朝鲜境内的间白山火山大约有6~12cm左右的隆起。尽管没有其他可以进行相互对比验证的资料和数据，但间白山地区较为频繁的地震活动，可从一定程度上证实这一 区域的活动性。 , CRInSAR 虽然InSAR技术弥补了以往各种测量手段（如常规水准测量，GPS 测量）的不足，能够对地表进行大面积形变监测，具有很大的优势，但其局限性却也不容忽略，如时间 去相干、空间去相干和大气扰动造成的延迟等问题，都会给干涉形变测量带来误差。 CRInSAR 和PSInSAR 技术的出现有效地提高了重轨图像系列的时间相干性和空间相 干性，实现了植被茂密地区的微量形变观测，测量精度也较常规InSAR测量精度有所提高。 在进行了一系列前期准备工作（包括相关地质背景资料的分析、光学影像的解译、 地形地貌的分析等）后，于2006年8月在长白山天池火山周围布设了10个角反射器点， 5 并于2007年4月13日获取了架设角反射器后的第一景ENVISAT数据，通过影像特征分析发现了7个角反射点位，没发现的三个点位海拔都在1600m以上，可能是因为影像 获取时长白山积雪尚未融化所致。 , MogiOkada 为了研究不同时段地下岩浆活动特点，笔者便利用1995-1998年InSAR资料，以及2002~2003年的水准和GPS观测数据进行模拟研究。 尝试了不同模型和不同模型组合，分别利用Mogi单源、Mogi双源、Okada三轴拉张模型、Mogi双源和Okada拉张模型的联合等进行反演计算。最终通过误差分析发现， Mogi双源模型拟合效果最佳，两个点源一个位于长白山天池老火山口下方7.9km处，另一个位于间白山火山下方5.5km处。Mogi双源模型还跟InSAR结果具有一定的一致性，利用双源模型正演InSAR形变场也取得了比较好的效果。创造性地利用Okada三轴拉张模型进行了反演计算，计算结果表明：三轴拉张源位于天池火山口偏北处，NNE和NWW方向拉张明显，跟深部探测资料结果对应较好，被认为是天池下方压力源更详 细的状态描述。天池火山和间白山火山岩浆囊位置的连线方向与该地区一条地震活动性 很强隐伏断裂的走向基本一致，推测两岩浆囊之间可能存在岩浆通道，岩浆的活动和运 移可能是造成该断裂和地震活动的原因。长白山地区火山活动在时间上具有间歇性，在 空间上存在迁移性。该火山区活动可能正在从初始扰动期开始向动荡期过渡，应加强火 山地震、形变、水化活动性监测尤其是固定站连续观测，以预测其喷发的危险性。 关键词： D-InSAR CR-InSAR 长白山天池火山 Mogi点源模型 Okada拉张模型 火山岩浆囊参数 6 Deformation monitoring and simulation research of Mt.Changbai Tianchi volcano Chen GuoHu ABSTRACT This thesis is a study on the deformation monitoring of the Tianchi volcano in the Changbai Mountains by using the techniques of differential interferometry synthetic aperture radar (D-InSAR) and computer numerical simulation. Using D-InSAR technology the deformation field of 1995～1998 of Jianbaishan volcano which lies in the southeast of Mt. Changbai is obtained. Ten corner reflectors have beed installed around Mt. Changbai in order to carry out test research of CR-InSAR technique. Using the data of GPS and leveling surveying of 2002-2003, the Mogi's Point Source model and Okada's tensile point source model inversion is made to obtain the optimum parameters of magma chamber in the Mt. Changbai volcano; andthe inversion results are applied to forward simulation of the deformation field from D-InSAR. This work is on the basis of previous the studies, and the main results and innovation are as follows: Investigating the means of deformation monitoring around volcano areas and the current situation of research. Deformation observation is the main means to monitor the activity of magmatic and forecast volcano eruptions. Although some tectonic activity and deformation cannot be at the peak before eruption, the majority of the volcanic eruptions did have obvious surface deformation. Accurate measurements of surface deformation can be used for the research of volcanic activity and volcanic eruption prediction. The surface deformation data can be used for numerical simulation in order to get parameters of volcanic magma chambers (size, pressure depth, etc) and important information of the underground structure of related hydrothermal systems. In this thesis, the status quo and deficiency of the means of deformation monitoring at 7 the volcanic areas are presented and discussed, the advantage of the technology of InSAR is introduced. At the end of this part, the significance and background of this research are given. Investigating geological conditions and present situation of research in the Mt. Changbai volcano area. During the Ninth Five-year and Tenth Five-year Project period, comprehensive researches have been made on the activity of the Mt. Changbai volcano, including volcanic geology, geophysics, deformation, mineralogy, lithology, geochemistry, dating, environment effect and volcanic hazard prediction and monitoring-control, and so on. The results show that Mt. Changbai volcano is an active volcano with most potential eruption hazard in China, and it is being in the stage from an initial disturbance state to a roilyt stage. Analyzing these research results provides some information for following research work. Using D-InSAR technology, the deformation in the near field of the Mt. Changbai is obtained. Because of the complex terrain and adverse climatic conditions, conventional volcanic deformation monitoring methods (such as leveling, GPS etc) are very difficult to gain volcano deformation of a larger regional. Using the technology of InSAR with its unique advantages can directly obtain the quantitative data of volcanic deformation in the large scope, with continuous coverage of the space. This work collected two scences of ENVISAT SAR data (200401104, 20051015) and five scenes of JERS-1 SAR data (19930615, 19940829, 19950816, 19980707, 19980820). These SAR data are processed using two-pass and three-pass differential interferomertry mode. But using the ENVISAT SAR data can not obtain the deformation result because of the poor coherence of the data. I obtained the perfect deformation result using the pair JERS-1 SAR data (19950816/19980820). The results show that during the three years from 1995 to 1998, there were 6-12cm LOS deformation in the area of Mt. Jianbai volcano which is located in the southeast of Mt. Changbai volcano. Although there is no other information can test this result, the frequent earthquakes can demonstrate the activity of this area. Experimental research of CR-InSAR technology in the area of Mt. Changbai volcano. Although InSAR makes up for the lack of many previous measuring methods (such as 8 leveling, GPS etc), and can monitor large-scale surface deformation, having immense advantages., its limition of application can not be neglected, which results in some uncertainties of results, such as temporal unrelated, spatial unrelated and delay caused by atmospheric effect. All these influential factors can bring some errors to interferometry deformation measure. The appearance of CR-InSAR (Corner Reflector InSAR) and PS-InSAR (Permanent Scatteres, InSAR) enhances effectively temporal coherence and spatial coherence of these repeat-pass image series, realizing observations of small deformation in vegetative cover areas. Moreover, accuracy of CR-InSAR monitoring is better than accuracy of conventional InSAR. After the earlier preparation, including analysis of the geological background, interpreting remote sensing images, analysis of topography and morphology, ten corner reflectors were installed around the Mt. Changbai volcano in August 2006, in order to make experimental research in the area of Mt. Changbai volcano. On April 13, 2007 the first ENVISAT SAR data after installing the corner reflectors are acquired. Based on the SAR data analysis seven feature points were found except the three points at the altitude above 1600m, the reason may be the snow has not melted. Simulation calculation was carried out by means of Mogi point source model and Okada tensile point source model. In order to study the activity of the magma chamber at different periods, this work uses the JERS-1 data of 1995~1998 and the data of GPS and leveling of 2002-2003 to simulate the parameters of the magma at the Mt. Changbai volcano. In this thesis, I try to use different models and different combination models to simulating calculation, such as Mogi single point source, Mogi two-points source, Okada triaxial tensile model, the combined model of Mogi two-points source and Okada tensile model. Through comparative analysis of the errors it is found that the Mogi two-points source model is the optimum fitting methods. A point source is located at 7.9km below of the Mt. Changbai Tianchi volcano, another is located at 5.5km below of the Mt. Jianbai volcano. It is found that the inversion result of the Mogi double point source and the result of the D-InSAR are in good agreement, and the forward simulation result of the deformation field obtained by D-InSAR technology is also satisfactory. The inversion calculation result of Okada triaxial 9 tensile model shows that the source is located at the north of the Tianchi crater and there is an obviously tensile component in the direction of NNW and NWW. This result accords well with the result of deep geophysical probing. The triaxial tensile source is considered to give the detailed parameters of the magma chamber which is located below in Mt.Changbai volcano. From these results we can infer that there may be a channel between the two magma chambers because of the direction of this two magma chamber is consistent with the strike of the hidden fault which has seismic activity in this area. Magmatic activity and migration may be the reason of activity of fault activity and seismic activity. The volcanic activity in the area of the Mt. Changbai has intermittence in time and migration in space. The activity of the volcano is likely in the stage from an initial disturbances state to a roilyt stage. In order to forecasts the eruptive risk of the volcano, the monitoring of earthquakes, deformation, hydration, especially continuous measurement at fixed station should be enhanced. Key s: D-InSAR, CR-InSAR, Mt. Changbai volcano, Mogi point source model, Okada tensile point source model, volcano magma chamber parameter. 10 李红光 KH积分方法合成地震图的研究 李红光，男， 1982年12月出生，2004年09月师从于中国地震局地球物理研究所吴庆举研究员，于2007年11月获硕士学位。 中文摘要 本论文系统简洁地介绍Kirchhoff-Helmholtz积分方法（简称KH积分方法）的基本原理、推导了基本计算公式并将其应用到计算反射波、转换波和回折波的地震学问题中。 KH积分方法是一种边界积分方法，它从严格的波动方程出发，将体积分转化为边界积 分，当用于计算反射或透射波时，KH积分方法把界面上的每个点都看作一个点源，认 为每个点源对反射或透射振幅都有一定的贡献，把界面上每个点的贡献相加就得到了反 射或透射响应。 本文首先用KH积分方法计算单层水平界面的反射波响应，并与反射率法和有限差 分方法的计算结果进行了对比，验证了方法及程序的正确性。在此基础上，进一步模拟 了弯曲界面的反射波响应，与有限差分模拟的结果对比表明：KH积分方法在计算不规则界面时有很好的精确度，而且其计算速度明显优于有限差分方法。考虑自由界面的反 射，我们应用KH积分方法计算了包含深度震相在内的所有反射波：PP、PS、pPP、pPS、 pSP、pSS，通过与反射率法的结果对比发现，六种震相都吻合的很好。 进一步地把KH积分方法推广到计算近震/远震转换波响应，推导出转换波的KH积 分公式，分别计算了水平界面与弯曲界面上的近震/远震转换波响应，通过与反射率法、 动力学射线追踪及有限差分方法的对比，表明KH方法能够很好地模拟不规则界面上的 近震/远震转换波。 最后我们用KH方法分别计算了含有多层界面介质中的反射波响应和速度梯度变化 介质中的回折波响应，与反射率方法结果非常相近。 我们的研究表明，KH积分方法是一种较好地模拟横向非均匀介质的合成地震图方 11 法，其计算精度不亚于其他方法，且计算效率高。 : KH积分方法， 合成地震图， 转换波， 反射波 Study on the “Kirchhoff-Helmholtz Synthetic Seismograms” Li Hongguang ABSTRACT In this paper, author systematically and succinctly introduced the basic theory and the basis equations of Kirchhoff-Helmholtz (KH) integral, and applied the method to the seismological problems of reflected wave, converted wave and refracted wave. The KH integral is one of the boundary integral methods, it is derived in terms of wave equation and transforms body integral to the boundary integral. When computing the reflected wave and transmitted wave, every point over the surface of the boundary can be regarded as a point source which contributes to the reflected wave or transmitted wave and the wave responses are the summation of the contribution of every point over the boundary. First, we computed the response of reflected wave on a single horizontal interface using the method of KH integral. The numberical results are good agreement with those of reflectivity method and finite-difference method. Furthermore, the response of reflected wave on a curved interface are calculated, comparison with the results from the finite-difference method shows that KH method can work quickly on curved or irregular interface, its computing speeds is superior to finite-difference method. In addition, the reflected waves from the free surface are obtained by the method of KH integral. From computing the six kinds of phases: PP, PS, pPP, pPS, pSP, pSS , my result indicates that both of the amplitude and the wave form for the model of layered medium are basically similar to the reflectivity method. From above basis, the KH integral theory is also applied to compute the conversion wave responses. With the derivation of the conversion wave of KH integral equation, we simulatted the near seism conversion wave response and the teleseismic conversion wave response 12 respectively. The results compared with that of the reflectivity method and the dynamic ray tracing gives satisfactory result. Finally, we try to calculate the theoretical seismogram in the multi-layered medium and the situation that the material parameters are linear below the surface. The result is consistent with that from the reflectivity method. Our analysis indicates that KH integral method is a preferable synthetic seismogram method in a laterally inhomogeneous elastic medium. The accuracy of this approximation is excelled other methods and it has the high computing efficiency also. Key Words: KH integral method, synthetic seismogram, conversion phase, reflection phase 13 作者姓名：杜海林 论文题目：2004年苏门答腊-安达曼大地震能量辐射源的时间域台阵 技术分析 作者简介：杜海林，男， 1978年11月出生，2004年9月师从于地球物理研究所许力生研究员，于2007年11月获硕士学位。 中 文 摘 要 首先，介绍了两种基于地震射线理论的台阵资料分析方法——聚束能量分析和迁移叠加分析方法的理论背景。前者适用于平面波分析，后者既适用于平面波分析，也适用 于球面波分析。 利用北京首都圈台网的短周期台站和宽频带台站，重构了两个台阵，讨论了两个台 阵的波数响应、增益和视慢度矢量求解时各参数的设定方法。 基于Matlab语言自主编写了追踪远震能量辐射源的软件，简要介绍了软件的功能 和计算流程。并通过数值试验，测试了程序的可靠性，分析了台阵几何形状对定位结果 的影响。 2004年12月26日苏门答腊-安达曼大地震以后发生了许多中等震级的余震，我们 分别应用聚束能量分析方法和迁移叠加方法对余震进行了重新定位，以美国国家地震信 息中心的结果为基准，得到了台阵在苏门答腊-安达曼震区的定位误差，并且针对不同 的情况，提出了两种校正主震能量辐射源的方法——插值校正和拟合曲线校正。 最后，将聚束能量分析方法和迁移叠加方法分别应用于北京首都圈台网构造的短周 期台阵和宽频带台阵，分析了2004年12月26日苏门答腊-安达曼地震的能量辐射源， 14 得到了震源破裂长度、方向、持续时间和破裂速度等有关破裂过程的信息。 关键词： 台阵技术；2004年苏门答腊-安达曼地震；能量辐射源 Analysis of the Energy Radiation sources of the 2004 Sumatra-Andaman Earthquake Using Time-Domain Array Techniques Du Hai-lin ABSTRACT Firstly, theoretical background of two seismic array techniques on a basis of the ray theory, the Beam Power Analysis and the Migration Stacking, are introduced. The former is suitable for plane waves, and the later is suitable for both plane waves and spherical waves. Two arrays are formed with stations of the Capital Region Digital Seismograph Network (CRDSN) of Beijing for this study. One is equipped with short-period seismometers, and another is equipped with broad-band seismometers. Different array has different array transfer function (ATF), and different array has different array gain. The ATF and array gains of the above two arrays are discussed. Also the skills of setting parameters in determination of the apparent slowness vector are simply illustrated. The software is made using Matlab computer language as a tool of processing array data in time domain for determination of the radiation sources of large earthquakes based on the principle presented in the previous chapter, so a brief introduction of the main functions and structures of the software is made. The reliability of the software is tested as well as the effects of the geometry of arrays on the locating results by means of some numerical tests. The 2004 Sumatra-Andaman huge earthquake was followed by a number of 15 moderate-magnitude aftershocks. It allows us to correct the radiation sources of the main-shock using the aftershocks' locations under condition of assuming the locations of these aftershocks published by NEIC were correct. It is discussed that how the locations of the radiation sources of the main-shock are corrected and two approaches of correcting locations, named Interpolation Correction and Fitting Correction, respectively, are proposed for different situations. Finally, the two techniques are applied to the short-period recordings and broadband recordings of CRDSN generated by the 2004 Sumatra-Andaman huge earthquake for the locations of the energy radiation sources, and the source information regarding of rupture duration, tempo-spatial process, fault geometry, average rupture velocity and others is obtained. Key words: the array techniques, the 2004 Sumatra-Andaman earthquake, the energy radiation sources 16 作者姓名：曹振中 论文题目：基于可靠性理论的砂土液化判别方法研究 作者简介：曹振中，男， 1982年3月出生，2004年9月师从于中国地震局工程力学研究所袁晓铭教授，2007年6月获硕士学位。 中 文 摘 要 17 上个世纪，新泻、邢台、海城、唐山、阿拉斯加、阪神、集集等国内外的地震中都 有典型的液化破坏现象。液化所带来的一系列问题越来越被人们关注，抗震设计中首先 要解决就是砂土在给定地震作用下能否液化的问题，因此砂土液化预测就成为工程抗震 研究中全球性的重要课题。 经过四十多年现场和室内试验的研究，国内外在液化判别预测研究上已取得较大成 果和进展，出现了基于不同指标的确定性判别方法。但到目前为此，对液化机制和影响 因素仍未完全了解和掌握，工程上使用的判别方法都是通过经验手段和直观方法得到。 并且，主要解决工程场地是否会液化的问题。 随着工程建设迅猛发展，对液化判别的要求也不断提高，要求不仅回答场地是否会液化的问题，还要给出液化判别的可靠性分析结果。探索将可靠性理论应用于液化判别之中，还只是刚刚起步，还有许多关键性问题需要解决。 本文主要工作和成果包括： 1．讨论了砂土液化的机制、影响因素，总结了近几年国际上几个大地震液化研究 新成果，分析了存在的问题，取得了一些新认识。 2．对国内外典型的确定性液化判别方法进行了总结，并对人工智能和专家系统在液化判别中的应用进行了归纳，指出了存在的问题和需要改进的地方，得到了一些新认 识。 3．阐述了可靠性分析的一般性理论，对应用于砂土液化判别时遇到的关键问题进 行了提炼。 4．研究直接基于液化现场调查资料建立液化发生概率的新方法，就其离散程度、回判成功率和预测成功率进行了分析，并与现有规范进行了对比。 5．深入研究我国抗震规范液化判别式的可靠性问题，提出其液化判别的概率分析结果。可靠性分析中包括利用国内资料检查其回判成功率、对台湾集集地震液化预测成 功率、其标贯比与液化概率关系的离散程度检查以及规范自身保守程度判定等。 关键词：砂土液化，判别方法，可靠性，抗震规范，概率 Study on Reliability-Based Procedure for Evaluating Liquefaction Potential Cao Zhenzhong ABSTRACT Liquefaction-induced damage phenomena occurred during many earthquakes last 18 century, such as Niigata, Xingtai, Haicheng, Tangshan, Alaska, Kobe, Chi-Chi earthquake etc. A series issues induced by liquefaction were paid more and more attention to, the first issue should be solved is that sand liquefies or not after experiencing the given earthquake. So the prediction of sand liquefaction has been a global important study subject in engineering seismology. After field and laboratory test studied for more than four decades, assessment liquefaction has been made great progress, and many different parameter-based deterministic methods were proposed in China and abroad. However, many procedures used in engineering were developed from empirical or heuristic method because of incomplete understanding of the sand liquefaction mechanism, and could only provide the assessing result of „yes? or „no?. As the rapid development of engineering construction, the demands of evaluating liquefaction increase gradually, which should not only provide the assessing result of “yes” or “no”, but also give the reliability analysis result. Adopting reliability theory for evaluating liquefaction is just at the beginning by now, there are still many key issues should be solved. The main contents and achievements of this paper include: 1. The paper has discussed the liquefaction mechanism and effect factors, summarized the new achievements of recent research, analyzed the existent issues, and got some new realizations. 2. The paper has summarized some domestic and abroad typical deterministic methods of predicting liquefaction, and introduced the artificial intelligence and specialist system adopted in evaluating liquefaction, pointed out the existent issues and some aspects need to be improved, and got some new realizations. 3. The paper has introduced the general theory of reliability analysis, concentrated the key issues during using in assessing liquefaction. 4. The paper has studied the new method of getting the probability of liquefaction based on the field investigated data, analyzed the proposed formula?s discrete extent, succeed ratio of judging and predicting, and compared the proposed formula with the seismic code procedure. 5. The paper has thorough studied the reliability issue of the seismic code procedure, and given its probability results. The succeed ratio of judging, the discrete extent of correlation between the ratio of SPT-N value and the probability of liquefaction, the conservative extent of the code procedure, were checked by using the domestic data, and also the succeed ratio of predicting was checked by using Chi-Chi earthquake data during the reliability analyzing. Key words: liquefaction, ground motion, acceleration, displacement, model of water pressure, modified response spectrum, incremental response spectrum 19 作者姓名：荣棉水 论文题目：粘弹性场地地形对地震动谱特性的影响分析 作者简介：：荣棉水，男， 1982年3月出生，2004年9月师从于中国地震局地球物理研究所李小军研究员，于2007年07月获硕士学位。 20 中 文 摘 要 大量的震害表明，地表地形对地面运动有很强的放大和缩小作用，因而直接 影响到震害的分布。对地形条件进行研究的目的有两个，一是解释地震时地面运动的放 大作用和空间变化，再就是预测在将来受地震时地面运动的特性。因此地形对地面运动 的影响分析是地震工程学中非常重要且具吸引力的领域。通过大量研究，局部不规则地 形对地震动峰值加速度的影响我们已有相当的了解，但是地形对地震动谱特性影响的研 究相对较少。本文就以探讨地形对地震动谱特性的影响规律为研究目标，通过数值计算结果分析，着重开展了以下几个方面的研究工作，并初步得到了一些有意义的结果。 1．数值计算方法的选择与分析 本文选择的数值计算方法为李小军等提出的粘弹性场地地震反应分析的一种时域 高精度显式有限元-有限差分方法。第二章系统回顾了显式有限元有限差分方法的理论 框架。另外，这种时域高精度的显式有限元有限差分方法的优点在于采用了李小军等的 显式逐步积分方法，而逐步积分方法的基本特性除了计算格式的收敛性、稳定性、精度 以外还必须考虑数值计算的高频能耗特性(dissipation)和计算起步的幅值超越性 （overshoot），因此，作为前人工作的补充，本文还在第二章中分析了中心差分格式、中心差分结合单边差分的格式以及李小军等的积分格式这三种不同显式格式的能耗特 性（对高频成分的抑制和消除作用）与幅值超越性（初始计算数据的放大特性），并对 三种显式格式显示出来的这两种特性做一综合评判。研究结果表明：李小军等的显式差 分格式在物理阻尼比为0到1的范围内均具有较强的能耗特性，这种特性优于另外两种 显式积分格式，能有效地克服透射边界的高频失稳，另外，这种格式也具有较小的速度 幅值超越性，因此，在处理复杂场地地震反应方面是一种比较理想的显式积分格式。 2．地形对地震动谱特性的影响分析 由于实际工程遇到的场地地形条件是复杂多变的，地形对输入地震波的谱 特性的影响也与很多因素有关，这些因素大体上包括场地特性和输入地震波特性两个 部分。场地特性包括地形的特征尺寸如坡度和高度、介质的性质等，输入地震动特性 包括输入地震波的频谱特性、入射角度等。为了便于总结规律，本文控制了若干影响 因素以简化问题的分析。首先研究了坡地地形在高宽比不同、体波入射角度不同时出 21 平面运动谱特性的差异，接着进一步研究了简单平台地形及凹陷地形对入射体波谱特 性的影响，给出了若干定性及定量的结果。这些研究结果表明：地形的特征尺寸和入 射角度不仅对地震动的峰值有影响，而且对地震动的谱特性有较大的影响，其中入射 角度的影响尤为显著。另外，研究还表明基于脉冲输入研究粘弹性地形谱特性的影响 仍不失为一种有效的方法。 3．地形对输入地震动反应谱加速度峰值及特征周期影响规律的分析 若干震害调查结果表明地形影响地震动的峰值及频谱特性，即影响反应谱的形状。 现有的场地反应谱对地形的考虑是以地震力的形式给出的，实际工程中碰到局部不规则 地形的问题时，工程师们无法从反应谱中直接获得特征周期或峰值加速度的调整， 本文以一个具体的平台地形为算例，通过拟合的人造地震动输入，计算获得有地形影响的场地观测点的加速度时程，再利用美国抗震设计样本规范ATC3-06建议的方法以及双参数标定反应谱的方法确定观测点加速度时程所对应的场地反应谱特征周期值，进而简 要分析了地形对输入地震动反应谱峰值加速度及特征周期的影响规律，并尝试着给出了特定条件下地形对观测点反应谱的调整方案。 关键词： 地表地形， 地震动谱特性， 显式有限元有限差分方法 22 The effect analysis of topography on the spectrum properties of viscoelastic ground Rong Mianshui ABSTRACT Many investigations of earthquake damages manifest that surface topography has significant amplification or reduction effects on ground motion, and, therefore, it can influence the distribution of earthquake damages directly. The purpose of studying on the influence of surface topography is to explain the reasons for the large amplifications and spatial variations of the ground motion and to predict some of the features of the ground motion during future earthquakes. So analysis of the effect of topography is one of the most important and attractive field in earthquake engineering. After a lot of investigation, we have know quite a few about the effect of surface topography on peak ground acceleration, but comparatively, the effect of surface topography on spectrum property of earthquake is still obscure for us. The objective of this thesis is to discuss effects of topography on spectrum property of earthquakes, through numerical experiments and result analysis, the following problems are analyzed, and some applicable results are obtained. 1. the choice and analysis of numerical integration methods A numerical integration method executed in this thesis is the explicit finite difference method suggested by Li Xiaojun et al. for analyzing the earthquake response of two-dimensional visco-elastic site. The systematic review of this method is made in the second chapter. Besides, the virtue of this method is using the explicit integration formula also suggested by Li Xiaojun. We know for a explicit integration formula, basic characteristics including astringency、stability、precision、dissipation and overshoot characteristic. So, as complementarity of former people?s work, dissipation and overshoot characteristics of three different integration formulas are discussed in detail. The conclusions show that the explicit integration formula suggested by Li Xiaojun and others has stronger numerical dissipation 23 characteristic than the other 2 formulas while the real damping varies from 0 to 1, and the larger the real damping is, the phenomenon is more obvious, this kind of formula can depress or eliminate the high-frequency instability induced by Local Transmitting Boundary, and, under the condition of the same precision, this kind of formula has lesser velocity overshoot characteristic, so it is an ideal integration formula for solving the earthquake response of complex site. 2. effect analysis of topography on spectrum property of earthquake Surface topography is complex and various, actually, effect of topography on spectrum property refers to many factors, such as characteristic dimensions of topography (including height and slope angle et al.)、characters of soil or rock、spectrum properties of input wavelets、angle of incidence et al. These factors can be classified into two sorts, characters of site and characters of earthquake. In order to summarize conveniently, firstly, we discussed the difference of spectrum property of anti-plane movement under the conditions of different topographic factors and direction of the incident waves, then research expands to mesa and valley. The conclusions show that the presence of topographic factors and the direction of the incident waves not only significantly affect the peak ground acceleration of site, but also affect the spectrum property of the incoming waves, and the effect of direction of the incident waves is especially remarkable. The study also find that it is reasonable to analyse spectrum properties with the input of impulse wavelets in the finite element simulation of wave motion。 3. the effects analysis of topography on peak-acceleration and characteristic periods Investigations after earthquakes show that surface topography affects the peak-accleration and spectrum property of earthquake, so, it affects the shape of design spectra. Considerations of topographic effects in the seismic design code for buildings are suggested in form of earthquake force, while it is hard for engineers to get adjusting project through seismic design code if they have to face topographic problems. In this thesis, a model of mesa is established, a lot of artificial ground motions are generated and have been used as input movements, time histories of observation points are gained, finally, two methods (one is suggested by ACT3-06 24 and another is suggested by Liao Zhenpeng and Li Dahua)are used to get the characteristic periods of sites. With these results, the effects analysis of topography on peak-acceleration and characteristic periods is executed and corresponding adjusting project for design spectra is suggested. Key words: Surface topography，Spectrum property of earthquake，Explicit finite- -difference method 25 作者姓名：姜文亮 论文题目：PS InSAR技术监测断层活动性应用研究 作者简介：姜文亮，男， 1982年5月出生，2004年9月师从于中国地震局地壳应力研究所张景发教授，于2007年6月获硕士学位。 中 文 摘 要 PS InSAR（Permanent Scatterer InSAR）技术即永久散射体干涉合成孔 径雷达技术，它是基于传统InSAR技术，对在时间序列上表现出稳定后向 散射强度或相位特征的目标点进行识别，进行差分计算，以研究较长时间 序列上目标点位移规律的一门技术。相比传统InSAR技术，PS技术受空间及时间失相关因素的影响极为有限，并且可以较好地消除大气影响，从而 提高数据的使用效率，达到对较长时间运动事件检测的效果。然而，PS技术在理论及应用方面还存在许多问题，如PS点的有效识别、PS点三维相位分析等。受可识别天然PS点分布密度的影响，PS成功应用实例也多局限于城市、矿山地面沉降检测。本文就PS技术应用中的关键问题进行了研 究，并进一步将该技术应用于无人区活动断裂带，对地震间歇期间活动断 层运动引起的地壳形变进行检测研究。 对传统InSAR技术中所存在的问题进行了分析，主要包括时间去相干、 空间去相干以及大气效应对InSAR技术的影响，继而对PS InSAR技术原理、优势及其发展现状作了介绍。 可靠的PS候选点是进行PS干涉处理的关键前提，本文重点对PS候选点识别方法进行研究，对已有的相干性、相位分析法、波谱差异法进行对 比总结，对后向散射强度法判别条件进行发展补充，基于PS点提取方法对人工角反射器识别作了研究，并对公共主图像选取进行了研究。 分析了传统InSAR相位的组分结构，对各组分的时空行为进行了研究， 26 进一步分析了PS点相位与时间及空间二维向量的关系，研究表明PS点差分干涉相位与干涉像对的时间基线、空间垂直基线（相对于公共主图像） 呈线性关系，与时间的比例斜率为PS点相对线性形变速率的常数倍，与空 间垂直基线的比例斜率为各PS点对应的高程修正值，在此分析基础上建立 了PS点干涉相位二维线性回归分析模型；与此同时对残余相位进行了分析。 在PS候选点提取方面，不再基于单一的统计分析方法，将波谱差异法 与强度稳定性法相结合，在对试验区构造环境及影响因素进行分析的基础 上，对PS候选点提取中关键参数进行最优选择，利用二维线性回归分析模 型对PS干涉相位进行分析，通过两次回归分析逐步计算出PS点高程修正值及相对线性形变速率，对差分干涉图进行修正，最终得到试验区时间序 列形变图。基于试验区形变图像对PS点位移特性进行分析，对区域形变特 征及形变规律进行讨论，对PS InSAR技术理论及应用再次进行总结。 研究表明盆地西北部念青唐古拉山处于上升阶段，沿卫星视线方向相对 位移速率约3mm/a，盆地内部处于沉降阶段，由北向南运动幅度具有一定 差异，靠近盆地南部山前地带沉降速率比较大，沿卫星视线方向相对位移 速率约3～4mm/a。随着PS点在盆地内部分布向西北推移，其位移形式逐 渐复杂，受青藏高原特有季节性冻融因素影响，许多PS点位移随季节表现出一定的往复式位移方式，但总体位移趋势沿卫星方向呈负向运动。 关键词： PS InSAR 波谱差异 后向散射强度 二维线性回归分析 时间序列形变图 27 Research on Application of PS InSAR Technology on Movement Monitoring of Active Fault Jiang Wenliang ABSTRACT The permanent scatterers approach (PS InSAR), which is based on conventional InSAR technique, put emphasis on processing time series of SAR interferograms by recognizing and analyzing single scatterers with a stable backscatter intensity or reliable phase behavior in time, to study the deformation histories of the earth's surface in a long time series. Compared with the conventional InSAR technology, the PS approach can better conquer problems of temporal and spatial decorrelation, also the atmospheric delay effect, which will improve the efficiency of datum utilization, to measure large time scale deformation affair. However, there are still some problems hamper the successful application of PS approach, which are identification of reliable point targets, three-dimensional analysis of PS interferometric phase. At present, the successful application of PS approach is mainly limited to ground surface sedimentation of urban area or diggings, where abundance of natural permanent scatterers can be identified. By contrary, fewer natural point targets will be recognized in Damxiong area, which is selected as the test area of the paper, for lack of human building and structure, only naked rocks are identified as permanent scatterer candidate. The paper will put emphasis on the study of some key aspects of PS technology, applying the PS approach on active fault, to detect the micro-movement of earth surface among seismic diapauses induced by active fault. The limited factors of conventional InSAR are analyzed in paper, which are 28 temporal decorretaion, spatial decorrelation and atmospheric delay. The basic principle, advantage and status are introduced. Methods of point targets? identification are investigated, including coherence, spectral diversity, phase and backscatter intensity. Moreover, new criterions are compensatory for method of backscatter intensity. Identification of manual corner reflector is researched based on these methods. Selection of common master image is studies subsequently. The phase model of convention InSAR are computed and analyzed, especially the temporal and spatial behavior of it. Furthermore, relations of unwrapped interferometric phase with two dimensions which corresponding to the perpendicular baseline and time difference are analyzed, which indicates a linear dependence of phase on the two terms. The related slopes of them correspond to relative terrain height correction and relative linear deformation rates. On base of this theory, a model of two-dimensional linear regression analysis is established for PS interferometric phase. Residual phases are studies at last, which is composed of non-linear deformation phase, atmospheric delay phase and phase noise. Methods of spectral diversity and stable intensity are combined to identify point targets candidate. This is quite different form former studies, for which single method is used only. After analysis of geological structure and influencing terms of test area, the optimal parameters are selected and combined for PSC selection. Interferometric phase of PSC are analyzed using two-dimensional linear regression analysis model. Maps of deformation histories are obtained after differential processing. Characteristics of typical point targets are analyzed, including deformation histories, influencing terms. Characteristics and disciplines of test area deformation are discussed. Key points of PS InSAR technique are summarized again. Research indicates, Nyainqêntanglha Range in northwest of Damxiong basin 29 raise at a relative rate of 3mm/a along the line of sight of satellite, at the same time, the Damxiong basin descends in a certain degree, but the descending rates is a few different from the south to the north. In southeast of basin, the permanent scatterers descend at a relative rate of about 3-4mm/a along the line of sight of satellite. Affected by term of frozen earth, the displacement of permanent scatterers in the north of Damxiong basin are not obvious, but behave a descendent trend as a whole. Key words: PS InSAR, Spectral diversity, Backscatter intensity, Two-dimensional linear regression analysis, Damxiong basin, Deformation of time series. 30 作者姓名：张 彬 论文题目：近年来几次强震在中国南北地震带动态应力触发问题研究 作者简介：：张彬，男， 1981年12月出生，2004年9月师从于中国地震局地壳应力研究所杨选辉教授，于2007年7月获硕士学位。 中 文 摘 要 地震应力触发指一个大震产生的应力场变化对邻近和远程地震活动的影响。地震应 力触发包括静态应力触发和动态应力触发。对静态应力触发研究较早，并且取得很多研 究成果。而对动态应力触发研究开展较晚，真正的动态应力触发研究是1992年美国兰德斯地震后，由于其在解释远程触发中具有很大的优势，所以近年来动态应力触发研究 发展迅速。我国位于环太平洋地震带和亚欧地震带之间，中国大陆及周边强震频发，研 究地震动态应力在中国的触发作用是非常必需的。本文旨在以研究中国南北地震带动态 应力触发作用为例，探讨强震动态应力触发问题。 文中较系统总结了国内外关于地震动态应力触发的研究动态，包括目前动态应力触 发存在的证据、研究现状、触发机理以及存在的问题等。从1992年兰德斯地震后，地震学家在全球范围内寻找动态应力触发证据，目前存在的主要证据包括：从地震频度图 中探寻被触发震群和独立地震、数字滤波后新发现的被触发地震和形变仪器记录的同震 变化和应力阶跃等。目前，提出的触发机理包括库仑破裂模型、速率-状态模型和孔隙压力扩展等。国内对动态应力触发问题研究的较少，很大程度上可能是因为地震动态应 力计算复杂。2004年苏门答腊8.7级地震后，动态应力触发研究升温，我国部分地震学 者也开始进行动态应力触发研究方面的探讨。 计算地震动态应力主要用到的方法是反射率法和离散波数分解法。计算方法涉及到 的数学函数较多：贝塞尔函数、格林函数、反射率函数和Bouchon坡度函数等；涉及到的物理过程也非常复杂：破裂时间的离散、格林函数褶积和频率域内的微分等。我们应 用前人一些公式导出了地震动态应力计算公式，改写了Cotton提供的部分FORTRAN程序。为了计算被触发断层上的库仑应力，建立了和MATLAB的连接，同时实现了MATLAB 31 程序可视化界面计算和绘图。计算地震动态应力分为两步：第一步利用FORTRAN程序计算应力张量的六个分量；第二步计算被触发断层上的动态应力，并绘制动态应力变化曲 线。 我们应用上述地震动态应力计算方法和程序研究了部分强震在中国大陆的触发情 况。南北地震带是中国大陆一条代表性地震活动带，近几年来地震活跃，资料丰富，所 以文中选用南北地震带为研究区域，并主要从两种思路开展了研究。第一，被触发地区 MM应力张量和库仑应力计算。包括2000年苏门答腊8.0地震对甘肃景泰5.8地震SS MM和缅甸6.5地震的触发作用； 2003年中、俄、蒙交界7.9地震对甘肃民乐-山丹SSMMM6.1地震和岷县5.2地震的触发作用以及2004年苏门答腊8.7地震对云南双SSSMM柏5.1地震和思茅5.1地震的触发作用。结果表明，这几次强震在被触发地震上SS 产生的动态应力峰值都超过地震触发阈值。第二，被触发地区小震突然增强。我们利用 小地震频度变化探讨了1990年以来部分地震在南北地震带上的触发作用，并以 2001 M年昆仑山口西8.1地震对四川、甘肃等地地震活动的触发作用为例进行了分析。总S 的看来，南北地震带易于被触发地震活动，但是不同地区的地震对南北地震带的触发作 用不同，这可能与地质构造有关系。 地震动态应力研究之所以复杂，因为它还受到很多因素影响：主震参数和被触发地 区构造，同时还有时间延迟问题。在计算主震在被触发点的应力张量时，主震的参数影 响非常大。改变断层三要素（走向、倾角、滑动角）、破裂时间以及破裂时的上升时间 对最终计算的动态应力大小和变化趋势都有很大影响；改变主震地震矩和发震断层的大 小只是影响了计算结果的大小，其趋势没有改变。大部分主震参数的变化都会影响到最 后的计算结果，所以在计算地震产生的动态应力变化时，应该获得更准确的主震参数。 构造区对地震动态触发影响也非常大，研究结果表明，大部分被触发的地震活动发生在 地热或火山区及扩张性块体上，同时也在挤压性块体上发生动态触发的少量证据。动态 应力对走滑和正断层地震触发更明显。强震面波到达后被触发地区不久可能触发当地地 震活动，但是还是很多被触发的地震具有时间延迟，延迟时间从几天到几个星期不等。 总之，我通过硕士论文研究，改进了动态应力触发计算方法和程序，计算了南北地 震带几次被触发地震的动态库仑应力变化，同时深入探讨了主震参数、触发地区构造特 32 征对地震动态应力触发的影响，还初步研究了地震动态触发时间延迟问题，在研究过程 中也发现了一些值得今后进一步研究的问题。 关键词：地震应力触发；库仑摩尔圆；构造块体；触发机理 Research on Dynamic Stress Triggering at Chinese North-South Seismic Belt in Recent Years Zhang Bin ABSTRACT Seismic stress triggering points to the influence induced by one earthquake to the nearby and remote seismic activity, including static stress triggering and dynamic stress triggering. Scientists have studied static stress triggering for a long time and have got lots of achievement. On the other hand, researching on dynamic stress triggering was scarce. The time we indeed researched on seismic dynamic stress triggering was after the Landers earthquake of 1992, USA. As dynamic stress triggering has superiority in explaining remote triggering, so it is developed rapidly in recent years. China locates between the Pacific Ocean seismic zone and the Asia-Europe seismic zone, so Chinese mainland and its periphery has more strong shocks. Although Chinese seismologists study seismic dynamic stress triggering later, it is necessary to study seismic dynamic stress triggering in China. In order to explore Chinese seismic dynamic stress triggering, we take Chinese North-South seismic belt as an example in this article. First, we summarized the development of seismic dynamic stress triggering at home and abroad systematically, including the existed evidence of dynamic stress triggering, current situation, triggering mechanism and so on. From Landers earthquake in 1992, seismologists began to seek evidence of dynamic triggering on a global scale. The main existent evidence at present includes: seek triggered earthquake swarm and single earthquake in seismic frequency map, find new triggered earthquake by digital filter, co-seism change and stress step recorded by strain instrument. The triggering mechanism includes coulomb failure model, rate-state model and pore pressure expansion etc. The reasons why the study on dynamic stress are so 33 less in China mainly maybe rely on the complexity of the computation. After Sumatra 8.7 magnitude earthquake in 2004, some seismologists start to study dynamic stress triggering in China. The main methods of computing dynamic are reflectivity method and the discrete wavenumber decomposition method. It involves lots of mathematics functions, including Bessel function, Green function, Reflectivity function and Bouchon slope function etc. The physical process related is also complicated: break time scattering, Green function convolution and different in frequency area etc. According to some predecessors? formulae, we deduced seismic dynamic stress calculating formula and rewritted part FORTRAN procedure that Cotton provided. In order to calculate coulomb stress on triggered fault, we connected FORTRAN procedure with MATLAB. At the same time, we realized the MATLAB procedure visualization interface for calculating and drawing. The calculation of Seismic dynamic stress comprised two steps: first, calculate six component of stress with FROTRAN procedure; second, calculate dynamic stress on triggered fault and its curve. With the mentioned method and procedure of calculating seismic dynamic stress above, we researched the triggered situation by some strong earthquakes in Chinese mainland. As the North-South seismic belt is a representative one in Chinese mainland, so we take it as the researching zone, and study mainly from two aspects. First, calculate stress tensor and Mcoulomb stress in triggered area, including 8.0 earthquake of 2000 in Sumatra triggered S MMM5.8 earthquake of Jingtai in Gansu and 6.5 earthquake of Burma, 7.9 earthquake SSS Mof 2003 in border of China、Russia and Mongolia triggered 6.1 earthquake of S MMMinle-Shandan and 5.2 earthquake of Minxian in Gansu, 8.7 earthquake of 2004 in SS MMSumatra triggered 5.1 earthquake of Shuangbai and 5.1 earthquake of Simao in SSYunnan. The results show that the dynamic stress peak value on triggered fault produced by the several strong shocks all exceeds to triggering threshold value. Second, the small earthquakes were strengthened abruptly in triggered area. We discussed the triggering effect on the North-South seismic belt by using the change of some earthquakes frequency since M1990. At the same time, we take west Kunlun mountains pass8.1earthquake as an S 34 example to analyse triggering effect in Sichuan and Gansu provinces. All in all, the earthquake activity is triggered easily in the North-South seismic belt, but the earthquakes in different area have different triggering effect in the North-South seismic belt, it is influenced geologic structure. The research on seismic dynamic stress is therefore complicated, because it is affected by so many factors: the main earthquake parameters and structure in triggered area, there still have the time delay problem at the same time. The main earthquake parameters have big effect when we calculate stress tensor in the triggered point. Change fault three key elements(rake, dip angle and slide angle)、breaking time and rising time have great effect on the final result and trend of the dynamic stress. The change of the main seismic moment and the size of the seismic fault only affect calculated result, but whose trend does not change. The change of most main earthquakes? parameters can affect the final result, so we must get much more accurate parameters when calculating the dynamic stress changing produced by earthquakes. Besides, tectonic setting also has great effect on the earthquake dynamic triggering. Studying indicates that most triggered earthquakes happened in geothermal or volcanic area and extension blocks. We also find a little of triggered earthquakes in nobbing blocks. Strike-silp and normal fault earthquakes are more easily triggerred dynamic stress. Seismic activity will be triggered not long before strong shocks surface wave arrived triggered area, but lots of triggered earthquakes have time delay, the time may change from several days to several weeks. In short, in my thesis, I have improved the calculating method and procedure of dynamic stress, calculated dynamic coulomb stress change, which triggered by some earthquakes in the North-South seismic belt. At the same time, we studied effect of earthquake dynamic stress triggering induced by main earthquakes? parameters and tectonic characteristic of triggered area, and also,we discussed the problem of dynamic triggering time delay . At last, we found several problems that are worth to study in future. Key words: Seismic stress triggering; Coulomb Mohr?circle; tectonic blocks; triggering mechanics 35 作者姓名：廉超 论文题目：地震定位新方法研究 作者简介：廉超，女， 1980年8月出生，2004年9月师从于中国地震局地震研究所李胜乐研究员，于2007年6月获硕士学位。 中 文 摘 要 地震定位是地震研究中最经典、最基本的问题之一。随着数学计算方法、计算机技术的发展，地震定位研究得到很大的发展，陆续出现了多种用各种最优化方法来进行定 位计算的定位方法，如双差地震定位法等。此外，随着多学科的相互交流与渗透，出现了集两种或多种计算方法的优点于一身的综合定位方法。 本论文提出了由三种新的地震定位方法，并结合精确定位的穷举法，形成的综合定 位法。 球面交切法既适合近震也适合远震。该方法在球面直接上进行交切计算，只要知道三个以上的到时差或能根据震相求出震中距，就能够快速定出震中位置，台网布局、震 源深度、发震时刻对定位结果的影响较小。 球面轨迹法是在球面上用初动到时差求得距离差进行定位计算，只要知道4个以上的初动到时就能快速定出震中位置。球面轨迹法采用各个台站的双差平方，即观测初动 到时差与计算初动走时差之间的差值的平方和为目标函数，因此发震时刻不影响震中定 位。 穷举“重复地震”定位法利用已发生的历史地震资料定位新发生的地震。与其它定 位方法相比，台网布局和地壳结构不均匀性对其定位结果的影响比较小。且将新定位的 地震输入数据库，增加历史地震的资料，对未来的地震定位的准确度也将不断提高。 综合三种定位方法的优点，对不同情况分别定位，将影响定位精度因素的影响减至 最低，以取得最优的定位结果。考虑到对定位精度的更高要求，可用“穷举法”对定位 结果作重新定位，更加满足不同实际需要。 36 关键词：地震定位 球面交切法 球面轨迹法 穷举“重复地震”定位法 The Research of New Earthquake Location Method Lian Chao ABSTRACT The Earthquake Location is one of the most classic and basic problems in the Seismologic research. As the development of mathematic method and calculator technique, Earthquake Location research gets big developing. There are appearing various superior methods using to locate earthquake, such as a double-difference earthquake location algorithm. In addition, by communicating and permeating of many courses, it appears to gather kinds of various advantage of calculation method in the whole comprehensive Earthquake Location method. This paper put forward to three new Earthquake Location methods. It is forming a comprehensive earthquake location method by combine them with exhaustion earthquake location method. The method of spherical surface shearing suitable for the location of near and far earthquake was proposed. The shearing calculation can be done directly on the sphere with the method, the epicenter can be rapidly located with only S-P arrival time differences of 3 stations at least. The station distribution, focal depth and origin time affect on the epicenter 37 location less. The earthquake location calculation can be done directly in spherical surface by use of the arrival time difference of initial movement to obtain the difference of distance by the method of spherical surface track. The epicenter can be located rapidly so long as the arrival times of initial movement of more than 4 stations are known. The square sum of double difference (the difference between the arrival time difference of initial movement wave by observed and the travel time difference of initial movement by calculated) of each station is selected as the objective function with the method of spherical surface track, so the origin time doesn?t affect the epicenter location. The method of exhaustion repeated-earthquakes location using known earthquakes to located the new earthquake which happened soon. Comparing with the others, station distribution and the heterogeneity of crustal structure affect on the epicenter location less. As the new earthquake which is located accurately be imported into earthquake database, locating precision of the method would increase continually. Synthesizing the advantages of three new location methods, it could reduce the influence positions the accuracy factor to go to lowest by using different method accounting different circumstance. So it can obtain the superior locating result. In consider of request about higher locating accuracy, it could use exhaustion earthquake location method to do the earthquake relocation. Key words: earthquake location, the method of spherical surface shearing, method of spherical surface track, the method of exhaustion repeated-earthquakes location, exhaustion earthquake location method 38 作者姓名：王涛 论文题目：钢筋混凝土框架－剪力墙结构弹塑性地震反应分 析 作者简介：王涛，男， 1978年9月出生，2004年9月 师从于中国地震局工程力学研究所孙景江研究员，于 2007年7月获硕士学位。 中 文 摘 要 我国现行建筑抗震设计规范(GB50011-2001)要求对于一些高层复杂结构应进行罕 遇地震下的弹塑性变形验算，但是对于目前复杂高层及超高层建筑，如何进行罕遇地震 下的弹塑性变形验算以实现“大震不倒”的设防目标仍是未解决的问题。 钢筋混凝土框架－剪力墙结构是高层建筑的主要结构形式之一。数值模拟技术是研 究建筑结构地震反应的重要手段，然而目前国内外可以进行钢筋混凝土框架－剪力墙结 构弹塑性动力反应分析的计算软件并不多，而且其中绝大部分结构模型属于二维或拟三 维分析模型，而对于大多数框架－剪力墙结构来说，很难将其简化为平面结构，所以建 立三维空间模型来进行三维空间弹塑性地震反应分析是解决问题的根本。 在对框架－剪力墙结构进行弹塑性地震反应分析时，采用适当的剪力墙分析模型， 建立合理的滞回模型是非常关键的。由于剪力墙的滞变性态复杂，工作机理及延性性能 方面的研究也远没有柱和梁研究得深入和成熟，目前还没有提出一个比较完善的数学模 型，而且传统的剪力墙宏观单元模型(如DRAIN－2D和IDARC-2D中所采用的各种模型)难以用到三维分析之中，微观模型同样也是应用困难。目前较有希望解决结构三维弹塑 性分析的单元模型就是纤维模型以及类似的多弹簧模型。纤维模型可以考虑变动轴力和 39 双向弯曲之间的耦合作用，能够表达压弯杆件在多轴荷载作用下的力学特性，但是它们在对框架－剪力墙结构进行弹塑性地震反应分析时，采用适当的剪力墙分析模型，建立 合理的滞回模型是非常关键的。由于剪力墙的滞变性态复杂，工作机理及延性性能方面 的研究也远没有柱和梁研究得深入和成熟，目前还没有提出一个比较完善的数学模型， 而且传统的剪力墙宏观单元模型(如DRAIN－2D和IDARC-2D中所采用的各种模型)难以用到三维分析之中，微观模型同样也是应用困难。目前较有希望解决结构三维弹塑性分 析的单元模型就是纤维模型以及类似的多弹簧模型。纤维模型可以考虑变动轴力和双向 弯曲之间的耦合作用，能够表达压弯杆件在多轴荷载作用下的力学特性，但是它们的模拟性能到底如何，与传统模型相比有何差别，是否可以应用实际结构分析，本文就上述 问题进行了研究。 本文采用纤维模型(CANNY99)和传统单元模型(IDARC-2D4.0)，分别对九层1：6钢筋混凝土框架－剪力墙结构振动台试验模型和美日联合七层足尺框架－剪力墙结构伪动 力试验模型的纵向结构进行地震反应分析，并在结构顶层位移、加速度，基底剪力及各层的最大层间位移角等多个方面与试验结果进行了对比，综合评价了它们的模拟效果。 从中得出这样的结论：基于纤维模型的三维弹塑性分析方法与基于二维平面分析模型IDARC-2D4.0模拟结果十分相似，具有较好的可靠性和稳定性，而且数值模拟与试验结 果比较一致，可以比较真实地模拟结构的弹塑性地震反应。这对研究结构三维弹塑性地 震反应有较大的参考意义。 随着震害经验的不断积累，人们逐渐认识到，地震时的地面运动是多维的，不论是 对称结构，还是偏心结构常常会产生扭转破坏。因此将结构简化成平面模型并只考虑单 向地震动作用的弹塑性分析不能全面反映和揭示结构地震反应的本质。那么，单、双向 输入地震动结构弹塑性反应有多大程度上相似和差异，这是人们十分关心的问题，因此 应进行多维地震动作用下的结构反应分析。本文将利用三维结构非线性分析程序 CANNY99，对一个规则、对称的钢筋混凝土框架－剪力墙结构建立三维空间分析模型， 分别在单、双向输入地震动情况下进行了弹塑性地震反应分析，对两种情况下结构顶层 位移、加速度、扭转角，基底剪力及各层的最大层间位移角、层剪力、层扭转角等反应 进行了比较。通过对比分析表明：对于规则、对称结构在单、双向输入时，最大楼层位移、最大层剪力、最大层间位移角等反应变化不大，顶层位移、加速度时程曲线吻合较好，曲线幅值大小及波形变化也十分相似。但值得注意的是，双向输入时的结构顶层扭转角时程反应和楼层最大扭转角反应都远大于单方向输入时的扭转反应，而且扭转反应 与地震波的选择紧密相关。 关键词：框架－剪力墙结构 弹塑性地震反应 纤维模型 三维分析模型 40 Elastic-Plastic Earthquake Response Analysis of Reinforced Concrete Frame-Shear Wall Structures ABSTRACT ’ 41 ’ 42 It is suggested that the analysis by multi-directional earthquake waves inputted is very important. The author used CANNY99, 3-D non-linear analyzing software, establish 3-D non-linear analyzing model to simulate a regular and symmetrical R.C. frame-shear wall building, process some elastic-plastic response analysis under the conditions of one-direction inputted and two-directions inputted. Comparing the results of top floor displacement, acceleration, base shear response, maximum story drift ratio, maximum story shear response and story torsion angle, the author draw a conclusion as follow: when the regular structures are under the conditions of one-direction wave inputted or two-direction inputted, maximum story displacement, maximum story shear response and maximum story drift ratio transform little. The time history responses of top floor displacement and acceleration are consistent each other under same conditions. There are some phenomenons which are worth to considering. The time history responses of top floor-torsion angle under two-direction inputted condition are greater than the ones under one-direction inputted condition. The style of earthquake wave has great influence upon the-torsion response. 43 作者姓名：潘波 论文题目：北京地区近断层强地震动研究 作者简介：男， 1981年10月出生，2004年9月师从于中国地震局地质研究所许建东研究员，于2007年7月获硕士学位。 中 文 摘 要 近断层强地震动研究是近十多年来地震学和地震工程学中非常活跃的领域。上世纪 末发生的许多破坏性大地震在近断层区域造成了严重破坏，同时也表现了一定的近断层 强地震动特征，这促使研究人员研究近断层强地震动的基本特征和分布规律、近断层强 地震动的模拟和预测方法，从而预测未来发震断层附近的强地震动分布区域和地面运动 时程，为城市和结构抗震以及震害预防服务。本论文作为“城市活断层试验探测” 中地震危险性评价里的子专题，目的为北京地区的城市规划和结构抗震提供参考，同时 也为全国各大城市开展的城市活断层地震危险性评价提供技术储备。 本论文以1679年三河-平谷8级地震的发震断层—夏垫新断裂为震源研究对象，以北京的中东部地区地下地质结构为传播介质研究对象，且分别建立相应的模型；利用日 本活断层中心的有限差分程序和中国地震局工程力学研究所的有限元程序对各自模型 进行了计算，并对计算结果进行了分析检验，验证了模拟结果的合理性，同时也反映了 近断层强地震动的基本特征，以下就所开展的工作做简要介绍。 本次研究范围包括了北京中东部和河北省的小部分区域，面积约7800平方公里。 研究区北部和东北部为山区，地层出露主要为前寒武系变质岩和燕山期的花岗岩，中部和东南部为平原区，大部分被全新世松散沉积物所覆盖。北京市市区位于研究区的西南 44 部，是我们重点关注地区。研究区内活动构造大量发育，为一地震频发区，区内已探测 到的大断裂主要分为NE向和NW向两组。其中NE向的新夏垫断裂被认为是1679年的三河-平谷8级地震的发震断层，是我们此次研究的目标。 发生于康熙十八年（公元1679年）的三河-平谷8级大地震是京津唐地区有历史记 录以来最大的一次地震。对于此次地震，不仅当时的史料给予了详细记载，现代地震工 作者也震开展了大量的科学研究工作，为我们近断层强地震动震源模型的建立提供了丰 富的资料。通过多方面的收集整理，我们得到了： （1）三河-平谷8级地震的历史记录和依据历史记录与中国地震烈度表（GB/T 17742-1999）绘制的历史等震线； （2）地表10余公里地震破裂带的野外实测和利用矢量分解与合成转变对数据的分 析计算结果，以及在此区域内开展的深浅地震勘探资料； （3）同处于华北地块，所受区域构造应力与破裂模式相似的1976年7.8级唐山地震的资料； （4）华北地区乃至世界性的震源参数之间的经验关系及公式； 四个方面的资料。利用这些资料先后建立了有限差分震源模型与有限元震源模型。有限 差分中我们试验性的建立了六个震源模型，并分别进行了计算，模拟过程为有限元模型 建立积累了经验。有限元方法充分利用以上资料建立一个相对合理的震源模型，各项参 数为：断层走向N40?E，倾向倾角SE78?，断层长度56km，宽度22km，平均位错3m， 20距地表高度3km，地震距1.48*10N.m，滑动角219?，破裂速度2.6km/s，上升时间2s， 2凹凸体面积270 km，初始破裂点位于潘各庄附近地下15km处。 传播介质模型建立中，我们收集整理了研究区内的区域地质资料，穿过研究区的13条速度或密度测线和第四纪，上第三纪，下第三纪，结晶基地G界面和中地壳稳定界面 C的等深度线图，综合分析这些资料建立了研究区内的三维地下速度结构模型。并根据 不同介质的不同属性，分别设定了波速、密度、伯松比与阻尼比等参数。 利用日本活断层中心的三维有限差分程序和哈尔滨工程力学研究所的三维有限元程 序对建立的有限差分与有限元震源模型和传播介质模型分别进行了计算。有限差分模拟 中共将传播介质模型划分为220万个网格，对6个不同的震源模型分别在工作站上进行 了计算，计算单个模型耗时约4小时，计算得到了研究区的地震动峰值速度分布。有限元法的模拟是在中国地震局工程力学研究所的40节点并行计算机上进行，共划分2187万个单元格，计算用时81.6小时，计算得到地表各点各时刻的速度时程，并通过后处 理程序获得了研究区的峰值速度分布图和地表各时刻的地震动等值线图。 通过对有限差分六个模型模拟结果的分析处理，发现结果很好的反应了近断层强地 震动的基本特征： （1）近断层效应：地震动高峰值主要集中在沿发震断层两侧的一个狭窄范围内，随着断层距的增加，峰值迅速衰减； 45 （2）上盘效应：对于倾滑断层，处于上盘的点的地震动值比下盘的点要大，同时 上盘点的地震动衰减相对于下盘要缓，强地震动的分布区域也比下盘大； （3）方向性效应：当断层面破裂时，从某一点开始朝着另一个方向开始破裂，位 于破裂方向前方的点的地震动值大于后方的点； （4）盆地效应：盆地中松散沉积物对地震波的放大作用和盆地边缘效应，使盆地 内的地震动值明显高于了盆地周边地区的值。 计算结果与近断层强地震动基本特征的吻合，验证了模拟结果的合理性。另外， 模拟中也暴露了一些存在的问题，为后续有限元模拟提供了经验。 通过对有限元法模拟得到的地表地震动速度值的分析处理，绘制了两条测线24个点的速度时程图和峰值速度等值线图。首先我们利用首波到时和弹性回跳理论，检验了两 排测线点的正确性；同时，从测线点上观察到（1）盆地效应：北京凹陷内点的地震动 峰值大于距离断层更近盆地外的点的值，且地震动持时明显大于盆地外各点的值；（2）方向性效应：位于断层一端平谷地区测点的地震动速度值达到了1.0m/s，而在距初始破裂点更近点的地震动速度值却约0.5m/s。另外，通过历史等震线与据模拟结果和“中国 地震烈度表”绘制的等震线的吻合检验了模拟结果的正确性。最后，我们绘制了每一时 刻的速度等值线图，并连接成小影片对震后60s内的地表地震动进行了演示。在影片中 再次体现了近断层强地震动的方向性效应、边界效应、盆地效应等基本特征，同时，在 55s后出现了边界失稳问题，且这一问题目前仍未得到解决。 通过对有限差分与有限元模拟结果的分析检验，验证了我们模拟结果的合理性和方法的 可信性，同时也证明了震源模型与传播介质模型建立和参数设置的合理性。另外，模拟 中还存在需要进一步改进和加强研究的地方。例如，发震断层的破裂过程和其参数的设 定没有清楚地解决；模拟中高频部分的计算和边界失稳等问题有待理论和方法上的突 破。但作者相信，这些都会随着资料的积累和研究的深入逐步得到解决，使近断层强地 震动模拟更加完善。 关键词： 近断层强地震动 右旋走滑正断层 震源参数 凹凸体 传播介质参数 并行运算 峰值速度等值线图 弹性回跳理论 方向性效应 盆地效应 46 A Study on Near-fault Strong Ground Motions in the Beijing area Pan Bo ABSTRACT Near-fault strong ground motions have become a very attractive subject of research in seismology and earthquake engineering in the last decade. In the end of the last century, many destructive earthquakes have caused great damages and shown some common characteristics in near-fault zones, which impel researchers to study the basic characteristics and the prediction method of near-fault strong ground motions. Based on these studies, people can predict the spatial distribution and the time history of near fault strong ground motions for city planning, seismic structural design and earthquake damage prediction. This thesis is the sub-topic of the seismic hazard evaluation which belongs to the project of the active fault detection in cities. This study will be helpful for the seismic structural design and city planning in the Beijing area, and to provide the technological storage for seismic hazard evaluation in other big cities of China. The research area of this work is about 7800 square kilometers, which includes the mid-eastern Beijing and small part of Hebei Province. Mountains are dominant in the north and northeast part of the study area, and plains which are covered with unconsolidated Holocene sediments are located in other part of the study area. In the research area, there are a lot of earthquakes and active faults which seriously affect people's life safety and economical construction. The major regional faults are divided into two groups of NE and NW direction in the study area, and the NE Xiadian fault that is the seismogenic fault of the 1679 Sanhe-pinggu M=8 earthquake is the objective of this research. The study process of near-fault strong ground motions is composed of establishment of 47 the earthquake source models and transmitting medium models, calculation of models and analysis of the calculated result. First is establishment of the seismic source models. The 1679 Sanhe-pinggu M=8 earthquake is the biggest event in the Beijing-Tianjin-Tangshan region with detailed historical records. Seismologists have carried out a great deal of work on this earthquake. In this thesis, reasonable earthquake source models are established and the corresponding parameters are given according to the following four data. 1) Historical records of the 1679 Sanhe-Pinggu M=8 earthquake and the isoseismal map of historical earthquakes which is compiled on the basis of historical records and “The Seismic Intensity Table of China”. 2) The field survey parameters of the more than 10 kilometer long earthquake surface rupture zone and the data of the deep and superficial seismic prospecting in research areas. 3) The parameters of the 1976 Tangshan M=7.8 earthquake for reference, because the two earthquakes happened in North China block and are similar in the regional tectonic stress field and of the earthquake rupture mode. 4) The empirical relations and formulas between the earthquake source parameters in the North China block and world. Utilizing the above mentioned data, I established 6 different earthquake source models for the finite-difference method(FDM) and 1 earthquake source model for the finite-element method(FEM). Source parameters of the FEM are given in Table 1. Table 1. Parameters of earthquake source of the Sanhe-Pinggu earthquake M=8.0 in 1679 Variable Strike Dip Length Width Average dislocation Depth of top Value N40?E SE78? 56km 22km 3m 3km Variable Seismic moment Rake Rupture velocity Rise time Area of asperity Initial rupture point 202 Value 1.48*10N.m 219? 2.6km/s 2s 270kmPangezhuang/15km Meanwhile, by using the regional geological data, thirteen profiles and the contour map of Quaternary and Tertiary deposits, interface G in the upper crust and interface C in the middle crust in study area, I established every transmission medium model and endowed with corresponding parameters by utilizing the FDM and FEM. Secondly, using the FDM established on previous work and a three-dimensional finite-difference computing program of near-fault strong ground motions, the peak ground velocity(PGV) caused by the destructive earthquake is simulated. In this simulation, a total of 62.28×10 grids are generated with the minimum grid spacing of 0.15km, and one computation took about 4 hours by workstation. In addition, applying a three-dimensional finite-element program and parallel computation technique developed by the Institute of Engineering Mechanics, China Earthquake Administration, the PGV and the velocity history for 60 48 seconds duration time of 32,400 observation points on the ground are calculated. In the simulation of the FEM, the earthquake source is divided into 4928 0.5km×0.5km grids , and the transmission medium model which comprises 36 different character parts is divided into 72.187×10 elements with the grid spacing of 0.08km. Finally, some contour maps of the PGV and velocity histories of observation points are drawn according to the simulation results. The analysis of some PGV contour maps of the FDM reveals, some basic characteristics of near-fault strong ground motion, for example: 1) Concentration effect. Near-fault strong ground motions with high amplitude are located in a narrow belt area along the projection of the fault on the ground； 2) Hanging wall effect. For a dip-slip fault, the ground motion in the hanging wall is stronger than that in the foot wall, and the attenuation of the ground motion in the hanging wall is slower than that in the foot wall； 3) Rupture directivity effect. If rupture starts from one end of the fault and sweeps to another, the PGV in front of the rupture will be larger and the frequency will be higher and the duration will be shorter； 4) Basin effect. Because of the amplification effect of the loose sediments in the basin and the interaction of the seismic wave with basin margins, the ground motion and duration time in the basin are more obvious than that in the rim of the basin. Then, I selected and analyzed 24 velocity histories of observation points which were along two beelines and a contour map of horizontal PGV in order to verify the simulation result of the FEM. Utilizing the first wave arrival and the elastic rebound theory, the correctness of 24 velocity histories is validated. Meanwhile, he basin effect and rupture directivity on 24 velocity histories of observation points are observed. In addition, it is found that the contour map of horizontal PGV is well in accordance with the isoseismal map of the 1679 Sanhe-Pinggu M=8 earthquake. The results from the simulation and analysis suggest that the source parameters and transmit medium parameters in the models are suitable, and the staggered-grid FDM and the explicit and decoupled FEM are applicable to estimate the distribution of strong ground motions in the study region. Key words: near-fault strong ground motion, right-lateral strike-slip normal fault, earthquake source parameter, asperity, transmitting medium parameter, parallel computation, contour map of horizontal peak ground motion, elastic 49 rebound theory, rupture directivity, basin effect 50