【作者】 王海云；

【导师】 陶夏新；

【作者基本信息】 中国地震局工程力学研究所 ， 防震减灾工程及防护工程， 2004， 博士

【摘要】 强地震动研究是工程科学与地球科学交叉的基础研究领域中一个关键的科学问题，也是保障人类社会稳定和发展、减少地震灾害的迫切需求。大量震害调查、统计资料表明，人身伤亡和经济损失主要是由建、构筑物等工程结构的破坏、倒塌造成的，而工程结构的破坏、倒塌主要是由强烈的地震动造成的。此外，强烈地震动也是地基失效、滑坡等其它地震破坏作用的外部条件。 在最近的十几年中，强地震动数据积累得很快。在一次地震中获取的空间覆盖面相当大、又有一定密度的观测数据的研究结果显示强地震动不仅受场地条件的影响，而且受震源破裂面上破裂子源的空间分布特点、子源破裂的先后顺序的控制。我国强地震动工程预测常用的经验性衰减关系过于简单，不能很好地描述强震、近场的地震动工程特征，只有用有限断层才能更好地描述震源的特点。当前主要的难点是如何在未来地震发生之前恰当的估计震源滑动的不均匀分布，进展相当艰难。 本文在大量文献调研的基础上采用基于运动学的矩形有限断层震源网格模型，通过以下三方面艰苦地推进，创造性的成果集成了一个有特色的近场地震动预测的有限断层模型，与地震动的随机合成相结合，形成了一套可实际应用的工程预测方法，以满足我国活断层地震动估计的迫切需求。 1．对于有限断层的全局参数，在归纳遥感图像识别，地震地质调查，人工地震勘探，余震分布范围研究的基础上，重点研究定量化的定标律。为了克服许多研究者基础数据各不相同的分散研究得到的经验关系难以得出表达参数之间的制约关系和震级的不一致带来使用中的困难，本文在迄今最系统的破坏性地震震源数据库的基础上向前推进，进一步验核、补充震源参数数据，根据数据分布趋势表现的具体变化改进分析方法，同时，建立断层尺度和断层面上平均滑动与矩震级之间的关系。 (1) 本文从Wells和Coppersmith(1994)的数据库中严格选择了149个地震的震源数据，用Hanks和Kanamori(1979)的公式校正了地震矩，另补充收集了9个1993年以后的地震震源数据。系统研究了地震断层破裂尺度和断层面上平均滑动与矩震级之间的关系，建立了一套可实用的地震定标律。根据地震学的相似性原理把地震破裂尺度和断层面上平均滑动与矩震级之间的关系归纳成统一的简单形式logY=αM_W-C_y，当α=1.0时，Y代表断层破裂面积；α=0.5时，Y代表断层破裂长度和宽度，以及断层面上的平均滑动，C_y为相应的参数值。 (2) 根据数据分布趋势显示的规律指出上式中的常数项C_y在不同的震级范围内可能是不同的，在约束斜率不变的同时，放松对截距的限制。这一震级范围上的分段不仅适应性更好，而且可以表达一些物理上的限制，例如破裂宽度不能无限制地随矩震级增大而增大。 (3)本文对三种断层类型中不同的震级段，系统地得出了一套预测地震断层破裂长度、宽度、破裂面积、平均滑动等四个参数的公式。实际上涵盖了己有的各种关系式，可以用来确定有限断层模型全局参数的上、下限和平均值。 (4)本文研究发现，走滑断层的破裂宽度达到饱和的临界矩震级为7 .0。 2.近断裂地震动，尤其是高频部分，十分强烈地受到地震断层破裂面上滑动分布不均匀性和破裂过程的影响，目前数据十分有限，认识到了复杂性、尚未有可靠的规律性。许多有限断层模型被称为随机模型，就是由于重点用随机方法描述了不确知的这一方面，最经典的是k平方模型。至少有一点在最近得到共识，断层破裂面上有一些部分的滑动量明显高，是高频地震动的主要激发源，有的研究者用凹凸体描述、有的用障碍体描述。本文抓住凹凸体的概念，从有限的数据中探索规律性，在Somerville等人(1999)开创性工作的基础上补充数据，分类分析，提取凹凸体的特征参数、建立其定标关系。 (1)本文在Somerville等人(1999)的长周期地震动波形反演的滑动分布数据中选择了13个地震的数据，补充收集了16个，总数增加了近一倍，扩大了涵盖的范围。提取了浅源地震凹凸体的特征参数、建立了其定标关系。地震断层破裂面上是仅有一个凹凸体，还是有多个凹凸体，有原则区别，参数的个数和数值都会有所不同。本文根据地震断层面上凹凸体的数量，将凹凸体模型分成所有凹凸体、单凹凸体和多凹凸体三类模型。所有凹凸体模型包括所有地震数据，是为了与Somerville等人(1999)整体分析的结果比较而设置的;单凹凸体模型地震占地震总数的40%左右，多凹凸体模型地震占地震总数的60%左右。为便于应用，本文建立了倾滑断层凹凸体数量与矩震级、断裂长度的关系。 (2)众所周知，在其它条件完全相同时一般逆断层引起的地震动最强，走滑断层次之，正断层更弱，为了反映断层滑动类型对强地震动的影响，分析中进一步区分所有断层、倾滑断层和走滑断层三种类型。前者包括所有数据，是为了与Somervi 11e等人(1999)整体分析结果的比较而设置的;后两者是面向实际应用的。 (3)提取的凹凸体参数，包括最大、所有和其它凹凸体的面积及其平均滑动;最大凹凸体的长度、宽度及其中心位置，地震破裂初始点位置等。最大凹凸体中心位置和地震破裂初始点更多还原

【Abstract】 Study on strong ground motion is a key research field in the intercross area of engineering and earth science, and is also an urgent requirement of ensuring stabilization and progress of mankind society & reducing earthquake disaster. A lot of investigation and statistic result of earthquake disaster show that casualty and economic losses are both mainly from the damage and collapse of buildings and other engineering structures that are further mainly caused by strong ground motion. Strong ground motion is also an inducing factor of other earthquake destructiveness, such as ground failure, landslip, and so on.In the past ten odd years, strong ground motion records were accumulated very much faster than before. The research results from motion records that were obtained in one earthquake, covered quite large area with a certain density, show the fact that ground motion is not only affected by local site condition, but also governed by spatial distribution of slip and the rupture process on the source plane. In general, it is believed that the empirical attenuation relationships widely adopted in China are too simple and cannot describe the engineering features of the motion very well at near field during strong earthquake, but the finite fault model FFM can do it better. The most difficult problem at present on FFM is how to estimate the in inhomogeneous distribution of slip before an earthquake, and the progress is quite slow.In this paper, a FFM, rectangle finite fault grid model based on kinematics, is adopted after a comprehensive review of plentiful published literatures, the creative achievements are integrated into a specific FFM for predicting near field strong ground motion, and formed a set of doable and applied engineering approach by combined the model with stochastic synthesis of ground motion, in order to satisfy urgent requirement from the project of exploration the assessment of earthquake active fault in urban areas in China, by means of the following three hard works.1. The quantitative scaling laws between global parameters of FFM and moment magnitude are studied in detail while other methods such as remote sensing image recognition, seismo-tectonic investigation, study of aftershock distribution, et al., are summarized. The most systemic database of destructive earthquake source parameters so far is checked and supplemented, analysis methods on relationships between faulting size parameters and moment magnitude are improved from the inherent trends in the data, and a relationship between average slip on the fault planeIVABSTRACTand moment magnitude is established, in this paper, in order to overcome the disadvantages in application of the empirical relationships derived from various source data by some researchers separately, such as the difficulty to express the constraint between each other of some source parameters and the difficulty from different earthquake magnitude scales.(1) The seismic source parameter data of 149 history earthquake world wide from the database by Wells and Coppersmith (1994) are selected, the moment magnitude in the database are corrected by Hanks and Kanamori (1979)’s formula, in addition, the source parameter data of 9 earthquakes since 1993 are supplemented. The relationships between fault size parameters on the fault plane and moment magnitude, and between the average slip on the fault plane and moment magnitude, are systemically studied, and a set of doable and applied scaling laws are worked. The theoretic relationships are simplified to a uniform form log7=aMw-Cy, in which Yis for fault rupture area if o=1.0, but Y for fault rupture length, width or average slip, if a=0.5, Cy is corresponding constant.(2) The different Cy values in the above mentioned equation for different magnitude intervals mean that the limitation on intercept is unbent while the slope is constrained as 1.0 or o.5, from the inherent trend in the data. This subsection of moment magnitude is better to match the data, and to express some physical constraints, for example, rupture更多还原