大陆内的长余震序列和对地震危险性评估的意义

板块构造学说最强大的特点之一,是通过已知的板块运动可以了解板块边界未来大地震的地点和平均复发间隔。然而板块构造学却不能探知板块之内何时何处发生地震,因为理想的板块内部不发生形变。因此,在板块内部,进行地震危险性评估主要依靠的假设是,短期历史记录给出的小地震的位置反映出连续形变从而引发未来大地震（Shed—locketal,2000）。然而在此我们提出,最近的这些地震有许多可能是数百年前大地震的余震。我们提出一简单模型可预测：余震序列的长度和断层应力加载的速．率成反比。     

中国大陆中强地震余震序列的部分统计特征

依据1970年以来记录相对完整的294次5.0级以上地震序列资料,研究中国大陆中强地震余震序列统计特征,探讨序列类型、最大余震震级、强余震活动持续时间等与主震震级及主震断层性质之间的关系.中国大陆孤立型、主余型及多震型地震余震序列分别约占23%、59%及18%.其中走滑型、具有倾滑分量的走滑型、具有走滑分量的倾滑型及逆冲型分别占48%、24%、17%及11%. 余震序列1年内最大余震震级与主震震级正相关,但主震震级较低时相对离散,孤立型序列离散程度较高,主余型及多震型序列线性相关性较好.绝大多数序列最大余震均发生在震后200天内,少数具有晚期强余震的序列主要属主余型序列,孤立型及多震型序列通常没有晚期强余震发生.68%的序列1年内最大余震发生在震后10天内,77%发生在震后30天之内,95%发生在震后120天之内.序列最大余震发生时间及5、6级强余震活动持续时间与序列类型及主震震级大小有关,多震型序列最大余震发生最快,孤立型次之、主余型最长.若仅就主余型序列而言,当主震震级较高时最大余震与主震间时间间隔相对较短,主震震级较低时最大余震与主震间时间间隔相对较长.     

中国大陆及其分区余震序列b值的统计特征分析

该文研究1966—1999年中国大陆及其分区的浅源主震型序列余震强度分布b值的统计特征。对于中国大陆整体而言,88个震例样本的b值平均为0.73±0.19。进一步将中国大陆划分为四个分区:西北地区(29个震例样本)、西南地区(29个震例样本)、华北地区(25个震例样本)和华南地区(5个震例样本)。选取样本多的前三者作为本文研究区。它们的b值平均()分别为0.62±0.17、0.86±0.17和0.69±0.14。根据平均值之间差异显著性的Z检验,研究分区之间b值平均在90%以上的显著性水平上的差别是明显的。该文还对比研究了不同板块构造背景的中国大陆(88个震例样本,=0.73±0.19)、美国南加洲(39个震例样本,=1.06±0.26)和日本地区(57个震例样本,=0.85±0.26)之间,浅源余震序列b值平均的差异显著性。     

中国大陆及边邻地区6 级以上地震序列的最大余震统计特征

依据1966 年以来中国大陆及边邻地区记录相对完备的6 级以上地震序列资料,统计研究了地震的破裂类型和不同破裂类型下余震序列特征等,得到了不同破裂类型下不同余震序列比例和优势分布、主震震级与最大余震震级差的经验关系。主震与最大余震的震级关系线性较好,得到的经验公式可以为中国大陆及边邻地区同类地震的最大余震震级判定提供参考;最大余震发生在主震后10 天内的比例为43. 8% ～ 81. 8% ,正断型和走滑型中的孤立型的最大余震发生在主震后10 天内的比例高达80% 。     

中国大陆MS≥7.8大震的余震活动差异性特征及其成因研究

文章以20世纪60年代以来中国大陆发生的4次MS≥7.8大震,即1970年云南通海7.8级、1976年河北唐山7.8级、2001年青海昆仑山口西8.1级和2008年四川汶川8.0级地震为研究对象,探讨了8级左右巨大地震的余震活动及分类,并以流变模型讨论了序列类型差异性的物理力学成因。初步给出:①4次大震的余震序列可归为两种类型,即低活动序列类型和高活动序列类型;②余震的差异性特征与主震孕育过程密切相联,理论模型和实际震例资料均给出,余震活动持续时间长短与大震孕震时间长短之间呈正相关变化;③根据对汶川8.0级大震的研究,孕震区介质流变性和构造变形速率的高低可能是造成大地震余震序列差异性特征的重要原因。     

汶川地震余震序列的地震各向异性

利用2008年5月12日汶川地震震源区及周边地震台站记录的余震序列资料,使用剪切波分裂系统分析法,分析了汶川地震发震构造龙门山断裂带及周边地区的地壳各向异性特征,推断了地壳最大主压应力方向及空间分布特征.研究结果表明：大致以安县为界,位于龙门山北东段的台站快剪切波的偏振方向为北东向,与断裂带走向一致;而位于龙门山西南段的台站快剪切波的偏振方向为北西向,与断裂带走向垂直;这个特征同样揭示出龙门山断裂带西南段逆冲、北东段带有明显走滑性质的特征.研究还显示,靠近龙门山与鲜水河、安宁河小江断裂交汇区附近的台站快剪切波的偏振方向表现比较离散,这可能是由震源区局部的复杂地质构造引起,与该地区复杂的主压应力方向特点一致.     

传染型余震序列模型震后早期参数特征及其地震学意义

研究中国大陆地区中强地震序列震后早期阶段（震后15天）ETAS模型参数的平均统计特征,据此讨论不同统计条件下的序列衰减及余震激发问题.宏观而言,模型参数b、p、α数值分布较为离散,不同统计条件下模型参数平均值的差异显著性不十分突出.详细对比不同统计条件下模型参数平均值的微小差异,b值随主震震级增大而增大,但b值随不同区域、不同主震断层类型或不同序列类型的变化不明显.p、α具有一定的区域特征,西南、西北p值略低于新疆及华北,表明西南、西北序列衰减相对较慢而新疆、华北序列衰减相对较快,华北α较低而西北α相对最高,意味着尽管华北序列衰减相对较快,但其激发高阶余震的能力却相对强,西北尽管序列衰减较慢,但序列结构单一,激发高阶余震的能力弱.p与主震断层类型关系不明显,即主震破裂性质不是决定序列衰减快慢的主要因素;α与主震断层类型有一定关系,走滑-近走滑型破裂所导致序列的α值最小、斜滑型次之、倾滑-近倾滑型最大,表明走滑-近走滑型序列激发高阶余震的能力最强、逆冲型最弱、斜滑型居中.p、α随主震震级增大而减小,意味着主震震级越高则序列衰减越慢、激发高阶余震的能力越强.不同类型序列p、α有一定差异,主余型序列p最小、孤立型p最大,表明相对而言主余型序列衰减最慢、孤立型序列衰减最快、多震型序列衰减速率居中.孤立型序列与主余型序列α大体一致、大于多震型序列的α值,即多震型序列激发高阶余震的能力相对最强,孤立型及主余型序列则相对较弱.     

中国大陆中强以上地震余震分布尺度的统计特征

依据1970年以来记录相对完备的MSge;5.0地震序列资料，统计研究了中国大陆余震分布尺度与序列最大地震震级M0及最大地震破裂形式之间的关系. 在95%置信概率下，考虑主震破裂形式， 分序列类型给出了余震分布尺度与M0之间的统计关系. 定性而言，余震分布尺度的对数lgR与M0正相关， 但数据分布较为离散. 分类型来看， 孤立型序列余震分布尺度与M0之间统计相关程度低， 余震分布尺度介于5~60 km之间； 主余型序列lgR与M0正相关； 多震型序列当M0le;6.2时lgR与M0之间相关性不显著，余震分布尺度介于5~70 km之间， 当M0ge;6.3时lgR与M0线性相关. 统计结果还表明， 走滑近走滑及斜滑型主震所导致序列类型比例之间没有显著差异，而倾滑近倾滑型主震（主要为逆断型破裂）所导致的主余型序列所占比例较高， 孤立型及多震型序列所占比例则相对较低. 对比研究显示， 当M0较高时，余震分布尺度主要取决于主震大小而与主震破裂形式关系不明显.      

云南地震序列最大余震的估算

利用1965~2012年云南地区的181次地震序列资料,分别用4种不同的方法得到主震和最大余震的震级关系式分别为:b值截距法Mmax=-0.44+1.03Mb±0.4;最大后续余震法Mmax=-0.23+1.00Mm±0.4;主震震级估算最大余震法Mmax=-0.71+0.89M±0.4;主震与最大余震震级差法:主震震级在5.0~5.9级之间,Mmax=M-1.2,在6.0~7.9级之间,Mmax=M-1.0。用这4种方法对2009年7月9日姚安6.0级地震和2012年9月7日彝良5.6、5.7级地震序列中的最大强余震进行估算,结果符合较好。统计结果显示最大强余震发生在主震后前3天的比例高达60%,这一结果可为最大强余震的时间判定提供依据。     

Long aftershock sequences in North China and Central US: implications for hazard assessment in mid-continents

Because seismic activity within mid-continents is usually much lower than that along plate boundary zones, even small earthquakes can cause widespread concerns, especially when these events occur in the source regions of previous large earthquakes. However, these small earthquakes may be just aftershocks that continue for decades or even longer. The recent seismicity in the Tangshan region in North China is likely aftershocks of the 1976 Great Tangshan earthquake. The current earthquake sequence in the New Madrid seismic zone in central United States, which includes a cluster of M ~ 7.0 events in 1811–1812 and a number of similar events in the past millennium, is believed to result from recent fault reactivation that releases pre-stored strain energy in the crust. If so, this earthquake sequence is similar to aftershocks in that the rates of energy release should decay with time and the sequence of earthquakes will eventually end. We use simple physical analysis and numerical simulations to show that the current sequence of large earthquakes in the New Madrid fault zone is likely ending or has ended. Recognizing that mid-continental earthquakes have long aftershock sequences and complex spatiotemporal occurrences are critical to improve hazard assessments.     

2017年九寨沟MS7.0地震余震危险性概率分析

本文介绍了余震危险性概率分析的概念,完整列出其相关公式的解析表达式并阐述了余震危险性概率分析的具体方法。以2017年九寨沟M_S7.0地震为例,首先,求得该地震余震序列的相关参数,结果显示:九寨沟M_S7.0地震余震序列理论最大余震震级约为M_L5.3;b值约为0.784 1,明显低于中国西南地区同类型的其它地震,表明九寨沟地震余震区应力水平相对较高;p值约为1.109 7,明显高于中国西南地区同类型的其它地震,表明九寨沟地震余震序列随时间衰减较快。其次,计算此次地震余震与主震释放能量的比例关系,结果显示:在九寨沟M_S7.0地震事件中,截止到10月22日约99.69%的能量为主震所释放,0.31%的能量为余震所释放。最后,利用九寨沟M_S7.0余震序列参数结果并结合衰减关系,分别计算了主震后0—1天、1—10天、10—30天和90—100天内不同断层距内不同水平的峰值加速度和峰值速度的超越概率,结果显示:随着断层距增加,在相同时间间隔内超越概率的值呈下降趋势;在震后第一天内余震危险性最高,随着震后时间增加,相应的超越概率值呈现出明显下降趋势,表明九寨沟M_S7.0地震余震危险性主要来自早期余震。本文的相关成果可以为震后地震危险性分析提供参考,并为短时间内应急救援及灾后重建提供辅助决策意见。      

Probabilistic earthquake hazard assessment for Peninsular India

In this paper, a new probabilistic seismic hazard assessment (PSHA) is presented for Peninsular India. The PSHA has been performed using three different recurrence models: a classical seismic zonation model, a fault model, and a grid model. The development of a grid model based on a non-parameterized recurrence model using an adaptation of the Kernel-based method that has not been applied to this region before. The results obtained from the three models have been combined in a logic tree structure in order to investigate the impact of different weights of the models. Three suitable attenuation relations have been considered in terms of spectral acceleration for the stable continental crust as well as for the active crust within the Gujarat region. While Peninsular India has experienced large earthquakes, e.g., Latur and Jabalpur, it represents in general a stable continental region with little earthquake activity, as also confirmed in our hazard results. On the other hand, our study demonstrates that both the Gujarat and the Koyna regions are exposed to a high seismic hazard. The peak ground acceleration for 10 % exceedance in 50 years observed in Koyna is 0.4 g and in the Kutch region of Gujarat up to 0.3 g. With respect to spectral acceleration at 1 Hz, estimated ground motion amplitudes are higher in Gujarat than in the Koyna region due to the higher frequency of occurrence of larger earthquakes. We discuss the higher PGA levels for Koyna compared Gujarat and do not accept them uncritically.     

Rapid earthquake hazard and loss assessment for Euro-Mediterranean region

The almost-real time estimation of ground shaking and losses after a major earthquake in the Euro-Mediterranean region was performed in the framework of the Joint Research Activity 3 (JRA-3) component of the EU FP6 Project entitled “Network of Research Infra-structures for European Seismology, NERIES”. This project consists of finding the most likely location of the earthquake source by estimating the fault rupture parameters on the basis of rapid inversion of data from on-line regional broadband stations. It also includes an estimation of the spatial distribution of selected site-specific ground motion parameters at engineering bedrock through region-specific ground motion prediction equations (GMPEs) or physical simulation of ground motion. By using the Earthquake Loss Estimation Routine (ELER) software, the multi-level methodology developed for real time estimation of losses is capable of incorporating regional variability and sources of uncertainty stemming from GMPEs, fault finiteness, site modifications, inventory of physical and social elements subjected to earthquake hazard and the associated vulnerability relationships.     

Probabilistic aftershock hazard assessment I: numerical testing of methodological features

Probabilistic aftershock hazard assessment (PAHA, Wiemer, Geoph Res Lett 27:3405–3408, 2000), provided for California within the frame of the STEP project, is based on a methodology, two features of which are addressed in detail: (1) the parameters of Omori’s law and (2) application of attenuation relations in evaluating peak ground velocity exceedence probability. Concerning the first point, we perform a simple parametric study. We assume the generalized Omori’s law by Shcherbakov et al. (Geoph Res Lett 31:L11613, 2004), in which characteristic time c scales with aftershock magnitude. The study shows that, among all the parameters, the hazard is most sensitive to the choice of m* (controlling the overall aftershock productivity) and least sensitive to the scaling of c. We also conclude that the hazard is mainly due to very early (less than 1 day) aftershocks. As regards the second point, we employ various attenuation relations from different tectonic areas to study their effect on the hazard analysis. We conclude that the resulting variations are relatively large, comparable to those obtained for varying m*.     

Probabilistic aftershock hazard assessment II: application of strong ground motion simulations

Probabilistic aftershock hazard assessment (PAHA) has been introduced by Wiemer (Geophys Res Lett 27:3405–3408, 2000). The method, in its original form, utilizes attenuation relations in evaluating peak ground velocity (PGV) exceedence probability. We substitute the attenuation relations together with their uncertainties by strong ground motion simulations for a set of scenarios. The main advantage of such an approach is that the simulations account for specific details of the aftershock source effects (faulting style, slip distribution, position of the nucleation point, etc.). Mean PGVs and their standard deviations are retrieved from the simulation results obtained by the new hybrid k-squared source model, and they are used for the PAHA analysis at a station under study. The model chosen for the testing purposes is inspired by the Izmit A25 aftershock (M _w  = 5.8) that occurred 26 days after the mainshock. The PAHA maps are compared with (1) those obtained by the use of attenuation relations and (2) the peak values of ten selected strong-motion recordings written by the aftershock at epicentral distances <50 km. We conclude that, although the overall hazard decay with increasing fault distance is similar, the PAHA maps obtained by the use of simulations exhibit remanent radiation pattern effect and prolongation in the strike direction due to the directivity effect pronounced for some of the scenarios. As regard the comparison with real data, we conclude that the PAHA maps agree with observed peak values due to appropriate attenuation model adopted in the analysis.     

Seismic hazard assessment for derivation of earthquake scenarios in Risk-UE

The main features of the Risk-UE project approach to assessing the ground-shaking (and related hazards) distribution within urban areas are described, as a basis for developing seismic damage scenarios for European cities. Emphasis was placed in the project on adoption of homogeneous criteria in the quantitative treatment of seismicity and in constructing the ground-shaking scenarios, despite wide differences in amount and quality of data available for the cities involved. The initial steps of the approach include treatment of the regional seismotectonic setting and the geotechnical zonation of the urban area, while the hazard assessment itself takes the form of both a deterministic analysis, and of a probabilistic, constant-hazard spectra analysis. Systematic 1D site response analyses were used, mostly in the softer soil zones, to modify (when needed) the obtained ground motion maps. Earthquake induced hazard effects, such as liquefaction and landsliding, are also briefly dealt with at the end.     

Considering potential seismic sources in earthquake hazard assessment for Northern Iran

Located on the Alpine-Himalayan earthquake belt, Iran is one of the seismically active regions of the world. Northern Iran, south of Caspian Basin, a hazardous subduction zone, is a densely populated and developing area of the country. Historical and instrumental documented seismicity indicates the occurrence of severe earthquakes leading to many deaths and large losses in the region. With growth of seismological and tectonic data, updated seismic hazard assessment is a worthwhile issue in emergency management programs and long-term developing plans in urban and rural areas of this region. In the present study, being armed with up-to-date information required for seismic hazard assessment including geological data and active tectonic setting for thorough investigation of the active and potential seismogenic sources, and historical and instrumental events for compiling the earthquake catalogue, probabilistic seismic hazard assessment is carried out for the region using three recent ground motion prediction equations. The logic tree method is utilized to capture epistemic uncertainty of the seismic hazard assessment in delineation of the seismic sources and selection of attenuation relations. The results are compared to a recent practice in code-prescribed seismic hazard of the region and are discussed in detail to explore their variation in each branch of logic tree approach. Also, seismic hazard maps of peak ground acceleration in rock site for 475- and 2,475-year return periods are provided for the region.