面向需求多元化的强震动记录Flatfile重要参数识别与反演研究

In order to establish the strong-motion Flatfile meeting for the pluralistic demands in the multiple disciplines such as the seismology, earthquake engineering, and civil engineering, we aim to study and develop some methods used to extract, identify, and inverse some key parameters associated with the seismic source, site, and recording. The following researches are carried out:.(1) Developing the identification methods to exclude the low-quality strong-motion recordings with different types, and proposing the calibration criteria in evaluating the quality of recordings in China..(2) Exploring the optimum filtering parameters and high-pass corner frequencies applicable for the China strong-motion recordings, then proposing a self-adapting method for the baseline correction of strong-motion recordings and designing a stable, high-efficiency and automatic flowchart for the data processing..(3) Developing the standard spectral ratio curves for various site classes defined in the code of seismic design in China, then proposing a new horizontal-to-vertical spectral ratio method for site classification which comprehensively consider the spectral shapes of sites characterized by the bedrock and soft intercalation and the priority weights of spectral shapes in different periods..(4) Based on the study of stable solution of linear equations, developing a non-parameter generalized inversion technique for separating simultaneously seismic source, propagation path, and site contributions to ground motions for the purpose of solving the vital problem of unstable inversion in the parameter generalized inversion technique..(5) Establishing the parameter-complete strong-motion Flatfile of China.

为建立面向地震学、地震工程以及土木工程领域多元化需求的强震动记录Flatfile，保证参数完备性，研究确定地震、场地与记录三个维度重要参数的提取、识别与反演方法。主要工作包括：(1)研究给出多种类型低质量强震动记录的识别排除方法，设计我国强震动记录质量评价准则；(2)研究强震动记录数据处理的最佳滤波参数以及高通截止频率确定标准，提出一种自适应的基线校正方法，设计一套稳定高效的强震动记录自动化数据处理流程；(3)构建我国抗震规范中不同场地类别的标准谱比曲线，给出同时考虑硬基岩与软弱夹层场地谱比特征的分类标准、以及兼顾不同周期谱形优先权重的曲线匹配方法，提出一套新的H/V谱比场地分类方法；(4)为解决传统广义反演方法反演不稳定的问题，研究线性方程组的稳定求解方法，给出一种用以分离地震动震源、路径和场地项的非参数化广义反演方法；(5)利用上述方法建立参数完备的我国强震动记录Flatfile。

为建立我国强震动Flatfile，尽可能保证其参数完备性，本项目研究低质量强震动记录识别排除技术，研究Flatfile记录基础信息、台站场地信息、地震震源信息重要参数的提取、识别和反演方法。取得的重要成果有：给出了我国强震动记录质量评价方法与准则；给出了我国强震动记录的滤波适用参数及高通滤波截止频率选取原则，提出了一套完整的面向工程应用的强震动记录数据处理流程；针对目前已有强震动记录基线校正方法大都依赖于主观经验易造成结果不确定性的问题，提出了一种自适应的基线校正新方法；建立了我国规范定义的三类工程场地的H/V标准曲线以及标准差，提出了基于H/V谱比的场地分类新方法及流程；提出了非参数化的地震动参数广义反演方法，解决了传统方法反演不稳定问题；建立了我国强震动Flatfile，涵盖事件、震源、台站、距离、波形和强度指标6类元数据。研究成果可充实现有区划图理论与方法体系，同时将服务于我国下一代地震区划图的编制工作。