考虑区域特征的地震动模拟——以2020年伽师<em> M</em><sub>S</sub>6.4地震为例

引用本文

王宏伟, 温瑞智, 任叶飞. 考虑区域特征的地震动模拟——以2020年伽师 M_S6.4地震为例 [J]. 地震地质, 2021,43(2): 430-446.
WANG Hong-wei, WEN Rui-zhi, REN Ye-fei. SEISMIC GROUND MOTION SIMULATION CONSIDERING REGIONAL CHARACTERISTICS: A CASE STUDY OF THE JIASHI M_S6.4 EARTHQUAKE IN 2020 [J]. SEISMOLOGY AND GEOLOGY, 2021,43(2): 430-446. 复制到剪切板

DOI:10.3969/j.issn.0253-4967.2021.02.011
Permissions

《地震地质》编辑部所有

考虑区域特征的地震动模拟——以2020年伽师 M_S6.4地震为例

王宏伟, 温瑞智^*, 任叶飞

中国地震局工程力学研究所, 中国地震局地震工程与工程振动重点实验室, 哈尔滨 150080

*通讯作者: 温瑞智, 男, 1968年生, 研究员, 主要研究方向为强震动观测技术与应用、重大工程地震紧急处置、地震海啸危险性分析等, E-mail: ruizhi@iem.ac.cn。

〔作者简介〕王宏伟, 男, 1990年生, 2017年于中国地震局工程力学研究所获防灾减灾工程与防护工程博士学位, 助理研究员, 主要研究方向为地震动特征及地震动预测, 电话: 15104547806, E-mail: whw1990413@163.com。

收稿日期: 2020-09-21 修回日期: 2020-11-12

基金项目: 中国地震局工程力学研究所基本科研业务专项(2018B03)、黑龙江省自然科学基金(LH2020E021)及国家自然科学基金(51808514)共同资助

摘要

为研究伽师地区地震动的区域特征, 提高地震动预测的可靠性, 文中利用伽师地区近10a收集的强震动观测记录, 采用非参数谱反演方法分离了地震动的震源、传播路径及场地影响, 建立了反映伽师地区地震动区域特征的路径衰减、路径持时、场地线性反应等经验模型, 并给出了伽师地区46次地震的震源参数(矩震级、拐角频率和应力降)。文中确定的伽师 M_S6.4地震的矩震级为5.893、拐角频率为0.362Hz、应力降为6.684MPa。伽师地区地震动的近场路径衰减较慢, 而远场更快的路径衰减与显著的非弹性衰减有关, 文中将伽师地区的路径衰减近似地描述为双段线性几何扩散经验模型和品质因子 Q=60.066 f^0.988 表示的非弹性衰减项的叠加; 伽师地区地震动的路径持时较长, 明显高于四川地区; 对比SS14场地线性反应模型, 伽师地区局部场地对10~30Hz频段地震动的线性放大效应很弱, 其他频段的场地线性反应与SS14模型接近, 在此基础上提出了适用于伽师地区的SS14调整模型。文中进一步将伽师地区地震动区域特征应用于伽师 M_S6.4地震的地震动模拟, 震源破裂过程采用随机产生的震源运动学破裂模型表示, 通过对比观测和模拟记录的峰值和反应谱, 检验了区域特征的可靠性, 最后给出了伽师 M_S6.4地震的地震动场。

关键词: 伽师 M_S6.4地震; 震源参数; 路径衰减; 区域特征; 地震动模拟

中图分类号:P315.9 文献标志码:A 文章编号:0253-4967(2021)02-0430-17

SEISMIC GROUND MOTION SIMULATION CONSIDERING REGIONAL CHARACTERISTICS: A CASE STUDY OF THE JIASHI M_S6.4 EARTHQUAKE IN 2020

WANG Hong-wei, WEN Rui-zhi, REN Ye-fei

Institute of Engineering Mechanics, China Earthquake Administration

Abstract

The regional characteristics of seismic ground motion are of importance for its reliable prediction, and the resulting seismic hazard mitigation and seismic risk controllability. The Jiashi region in Northwest China, which is located in the boundary between the western segment of the South Tianshan and the northwest Tarim Basin, is subject to high seismic hazard due to the active seismicity. This region has frequently suffered from the moderate to strong earthquakes. For the purpose of investigating the regional characteristics of the ground motion in Jiashi region, we separated simultaneously the earthquake source spectra, propagation path attenuation, and site responses from the horizontal S-wave spectra in the 366 three-component strong-motion recordings with high quality after manual detection and selection based on the nonparametric spectral inversion technique. The datasets were recorded at 25 strong-motion stations in 46 earthquakes of M3.0~6.4 from July 2007 to February 2020 occurring in the Jiashi region, mainly including both Jiashi seismic sequences in 2018 and 2020.
We thus developed the region-specific empirical models for the Jiashi region, including the path attenuation, the path duration, and the linear site response models. At the same time, the source parameters theoretically depicting the omega-square source spectral model, including seismic moment M_W, corner frequency f_c, and stress drop Δ σ were retrieved from the inverted source spectra for the 46 earthquakes considered using the grid-searching method. Their reliability was further verified from the normalized fitting residuals between the inverted and theoretical source spectra varied with f_c, and the f_c- M_W plots. We determined that M_W=5.893, f_c=0.362Hz, and Δ σ=6.684MPa for the M_S6.4 Jiashi earthquake in 2020. According to the inverted path attenuation curves against hypocentral distance less than 120km, the ground motions generally show the slower path attenuation at local distances, while the faster path attenuation at regional distances was clearly found, which could be ascribed to the significant anelastic attenuation in the Jiashi region. The path attenuation in the Jiashi region was approximately modeled by the empirical expression, simultaneously including the hinged bilinear geometrical spreading model and the anelastic attenuation term in the function of quality factor Q. The yield Q values were regressed as 60.066 f^0.988 at frequencies of 0.254~30Hz. The strong anelastic attenuation may be attributed to the prominent interaction of the seismic wave propagating within the high inhomogeneous crust. The transition distance R₁=60km, and the geometrical spreading exponents n₁=0.30 and n₂=0.59 before and after R₁ well defined the preferred hinged bilinear geometrical model. The site responses show the dependence on the V_S30 values, and the higher amplifications are generally predominant at sites with the lower V_S30 values, and vice versa. Compared with the linear site amplification model proposed by Syehan and Stewart(2014) for the NGA-West2 project, the local sites in the Jiashi region show the very weak amplification effects on the high-frequency ground motions at 10~30Hz, while the good consistencies at low-to-intermediate frequency bands occur. On the basis of the SS14 model, the modified model applicable for the Jiashi region was proposed in this study. The path durations, defined by an effective 95%~5%significant duration, were calculated for the strong-motion recordings(a total of 502)in the Jiashi region, and were applied for the development of the empirical model. The prolonged path durations were observed in those ground motions recorded in the Jiashi region, significantly higher than those observed in the Sichuan region.
Furthermore, the yield regional characteristics were preliminary examined by the ground motion simulations. The stochastic finite-fault method was further applied to reproduce ground motions in the M_S6.4 Jiashi earthquake with these regional characteristics under consideration. The kinematic source models were first generated stochastically by the filtering method in the two-dimensional wavenumber domain. These stochastic source models can well represent the source rupture process in physical and its sufficient variability. We then verified the reliability of these regional characteristics from the good agreements of the ground motion intensity measures between the observations and the simulations, including the peak ground accelerations and velocities, and the spectra accelerations. Moreover, the key role of the regional characteristics in ground motion simulation was clearly showed by the significant deviations of the long-period ground motions at large distances between the observations and the simulations based on the path attenuation models in the Sichuan-Yunnan region. Finally, the ground motion maps were developed for the M_S6.4 Jiashi earthquake according to the regional characteristics and a best-performing source kinematic rupture model.
The region-specific ground motion empirical models are very helpful for reliably predicting ground motions, and mitigating seismic hazards in the Jiashi region with the active seismicity.

Keyword: Jiashi M_S6.4 earthquake; source parameters; path attenuation; regional characteristics; ground motion simulation

文章图片

0 引言

伽师地区位于南天山西段与塔里木盆地西北部的交界区域, 受南天山向S的挤压作用, 北部发育有近EW向展布的柯坪塔格逆冲褶皱带, 位于迈丹断裂与柯坪塔格逆冲断裂之间, 南部塔里木盆地内则发育有隐伏断裂(刘启元等, 2000; 徐锡伟等, 2006), 诸多研究认为塔里木刚性块体向北部远程传递的印度板块向N推挤欧亚板块的压应力与该地区的构造变形有关(Laborde et al., 2019)。地震活动性特别强烈的伽师地区曾发生过一系列中强以上的破坏性地震, 例如1996年阿图什M_S6.9地震、 1997— 1998年伽师强震群、 2003年巴楚-伽师M_S6.8地震、 2011年阿图什M_S5.8地震、 2018年伽师M_S5.5地震等。 1902年阿图什8 ¼ 级地震是该地区有记录以来发生的最大震级地震。 2020年1月19日柯坪塔格逆冲断裂西段发生了伽师M_S6.4地震。频发的中强以上地震对伽师地区的防震减灾工作提出了广泛需求, 可靠地预测地震动有助于减轻地震灾害与控制地震风险。

地震动是地震震源、地震波传播路径及局部场地综合作用的结果, 因不同区域的地震构造、地质构造、地表土层形成等的差异, 地震动普遍具有明显的区域特征。 Boore等(2014)利用NGA-West2计划的强震动数据集揭示了不同区域(美国西部、日本、意大利、中国台湾、四川等)地震动非弹性衰减的显著差异及其对远场地震动衰减的影响; Xu等(2020)将中国南北地震带6次M_W6.0~6.2地震的地震动残差分解为事件内残差和事件间残差, 发现高频地震动差异主要与地震应力降相关, 而远场地震动差异主要与区域非弹性衰减相关; Kale等(2015)在研究伊朗和土耳其地区地震动预测模型时发现2个地区地表浅层土层的剪切波速结构显著影响地震动强度; Boore等(2015)的研究表明美国东部地震动路径持时明显高于美国西部, 这可能与地壳介质差异导致的地震波衰减差异有关。为提高地震动预测的可靠性, 有必要考虑地震动的区域特征, 强震动观测记录包含了地震动路径衰减、场地效应等区域特征的重要信息。

布设于南天山西段的48个自由场强震动观测台站自2007年以来已陆续在60余次伽师地区M2.8~6.4地震中收集了超过500组三分量强震动观测记录(截至2020年2月), 这些观测记录为研究伽师地区的地震动区域特征提供了重要的基础数据。目前已有学者利用伽师地区单次或几次地震的强震动观测记录分析地震动的路径衰减、持时、场地效应等(张振斌等, 2019; 李文倩等, 2020), 但尚未系统地给出伽师地区地震动的区域特征以及可用于地震动预测的经验模型。本文利用伽师地区近10a来收集的强震动观测记录, 基于两步非参数谱反演方法分离地震震源谱、地震波路径衰减及场地效应, 建立伽师地区的路径衰减、路径持时、场地线性反应等经验模型, 并利用区域经验模型模拟2020年伽师M_S6.4地震的地震动, 通过对比模拟和观测记录检验本文给出的区域特征的可靠性, 最后模拟给出伽师M_S6.4地震的地震动场。

1 伽师地区的地震动区域特征

本文基于强震动观测记录S波的两步非参数谱反演方法估计伽师地区的地震震源参数、地震波路径衰减及场地效应, 建立该地区地震动区域特征经验模型。

1.1 两步非参数谱反演方法

两步非参数谱反演方法由Castro等(1990)提出。第1步, 从地震动水平向S波傅氏谱O_ij(f, M_i, R_ij)中分离路径衰减项A_ij(f, R_ij), 即

$\ln O_{ij} (f_{m}, M_{i}, R_{ij}) = \ln M_{ai} (f_{m}) + \ln A_{ij} (f_{m}, R_{ij})$ (1)

其中, M_ai(f)为与第i次地震的震级大小(M_i)相关的量, f_m为第m个频率点的频率, R_ij为第i个地震震源至第j个台站的距离。求解式(1)时需给定路径衰减项的约束条件, 即参考距离R₀处无路径衰减且路径衰减为距离的平滑曲线。本文计算中将频率f_m所有可用谱划分为N_D_,_m个5km间距的距离段, 式(1)中以A(f_m, R_k_,_m)代替A(f_m, R_ij), R_k_,_m表示第k个距离段内频率f_m可用谱震源距的平均值。第2步, 从S波傅氏谱中移除路径衰减项后分离得到震源和场地项, 即

$\ln [O_{ij} (f_{m}, M_{i}, R_{ij}) / A_{ij} (f_{m}, R_{k, m})] = \ln S_{i} (f_{m}) + l n G_{j} (f_{m})$ (2)

求解式(2)时需分配震源和场地项之间的权衡, 一般采用参考场地作为约束条件。

1.2 处理强震动记录

本文收集了伽师地区59次M2.8~6.4地震中48个强震动观测台站获取的502组三分量加速度波形记录, 记录的震源距主要为20~200km, 大多数记录的水平向峰值地面加速度(PGA)≤ 50cm/s²(图 1)。依次采用基线校正、记录首尾加余弦窗并补零、巴特沃斯非因果带通滤波处理观测记录。低通滤波的拐角频率统一设定为30Hz; 高通滤波的拐角频率(f_hp)随记录变化: 首先根据2个经验关系给出f_hp的预估值, 分别为记录最小可用频率(f_lb)与矩震级M_W的经验关系(Yenier et al., 2015)和f_lb=1.25f_hp(Abrahamson et al., 1997), 近似以中国地震台网中心发布的地震面波震级(M_S)或地方震级(M_L)作为M_W初步估计f_hp; 利用f_hp预估值对记录进行带通滤波处理后, 通过人工检查记录的低频段傅氏谱形状判断f_hp的预估值是否合适, 如果随频率降低呈下降的抛物线形则f_hp预估值合适, 如果随频率降低呈先下降后翘起的抛物线, 则将f_hp调高至翘起前的频率; 重复上述人工检查过程直至得到合适的f_hp, 最终确定f_hp为0.10~1.17Hz。为了尽可能确保场地不出现非线性反应, 剔除了PGA> 100cm/s²的记录; 为了清晰地识别S波部分并尽可能消除触发阈值引起的地震动偏差(Abrahamson et al., 2016), 剔除了R> 120km的记录; 为了确保充分的数据冗余, 剔除了不能同时满足记录不少于3组的台站和地震的记录。之后截取剩余的386组记录的S波窗, 首尾加余弦窗并补零处理后计算S波傅氏谱, 截取与S波窗长度相同的事前噪声窗, 进行相同处理后计算噪声傅氏谱和信噪比(SNR), 根据SNR> 5.0判断S波傅氏谱的可用频段。最终, 本文选用了25个强震动台站在46次M3.0~6.4地震中获取的366组强震动记录进行谱反演分析, 选用的地震及台站分布如图 2 所示。将谱反演选用记录的最小震源距作为参考距离R₀, R₀=20.33km。

	Figure Option View Download New Window
	图 1 a 伽师地区强震动记录的震级-震源距分布; b 水平向PGAFig. 1 Magnitude vs. hypocentral distance(a), and the peak ground accelerations at both horizontal components(b)for strong-motion recordings in Jiashi region.

	Figure Option View Download New Window
	图 2 谱反演选用的地震(红色圆圈)及台站(蓝色三角形)分布黑紫色实线为断层(邓起东等, 2003), 灰色实线为传播路径, 黄色五角星为部分6.0级以上历史地震Fig. 2 The epicenters of earthquakes(red circles)and the locations of strong-motion stations(blue triangles)considered in the spectral inversion analysis.

1.3 路径衰减

第1步谱反演给出了伽师地区不同频率的S波路径衰减曲线, 如图 3a所示。路径衰减随震源距(R)增大而减小, 当R≤ 60km时路径衰减大体上介于(R₀/R)^0.5~(R₀/R)^1.0 之间, 当R更大时路径衰减普遍< (R₀/R)^1.0, 路径衰减曲线上凸说明随R增大其衰减速度明显增大, 这与显著的远场非弹性衰减有关。采用频率无关的几何扩散项和频率相关的非弹性衰减项表示路径衰减, 不同几何扩散模型(线性、双段或三段线性模型等)在拟合地震动的距离衰减时差异不明显(Bora et al., 2015)。本文的几何扩散采用双段线性模型表示, 路径衰减可表示为

lnA(f_m, R_k_,_m)=n₁· ln(R₀/R_k_,_m)+[f_m/Q(f_m)]· [-π (R₀-R_k_,_m)/β ], R_k_,_m≤ R₁

lnA(f_m, R_k,_m)=n₁_·ln(R₀/R₁)+n₂_·ln(R₁/R_k,_m)+[f_m/Q(f_m)]· [-π (R₀-Rk_,_m)/β ], R_k,_m> R₁ (3)

其中, 剪切波速β =3.60km/s(赵翠萍等, 2008), 通过最小二乘法拟合得到R₁=60km, 几何扩散指数n₁和n₂分别为0.30和0.59, 各频率的S波品质因子结果见图3b。以幂函数Q₀f^η拟合品质因子得到Q=60.066f^0.988, 低Q₀高η 的特征与伽师地区强地震活动性有关。与川滇地区的几何扩散和Q值(Xu et al., 2010)进行对比可知, 伽师地区的几何扩散明显更弱, 但低频段的非弹性衰减更强(Q值更小)。

	Figure Option View Download New Window
	图 3 a 伽师地区不同频率S波谱路径衰减; b S波品质因子Fig. 3 S-wave spectral attenuation versus distance at each frequency(a)and S-wave quality factor(b) in Jiashi region.

1.4 地震震源参数

第2步谱反演需权衡震源和场地项, 而本文采用的台站中没有可直接作为参考场地的坚硬岩石场地台站。根据台站场地钻孔资料及建筑抗震设计规范对场地类别划分的规定(中华人民共和国住房和城乡建设部, 2010), 有7个台站场地可划分为Ⅱ 类场地(065AKS、 065ALM、 065HLJ、 065AKT、 065BRM、 065ATS、 065SUF), 其余均为Ⅲ 类场地。利用水平垂直谱比法(Nakamura, 1989)计算了记录不少于5组的3个Ⅱ 类场地台站(065AKS、 065ALM、 065HLJ)的场地反应, 本文将这3个Ⅱ 类场地台站用作参考场地。

第2步反演得到的46次地震的加速度震源谱如图 4a所示, 当f< 10Hz时震源谱大体符合ω ^-2 理论震源谱模型(Brune, 1970), 但震源谱的快速衰减则普遍出现在f> f_max的高频段, f< 10Hz频段的反演震源谱可采用ω ^-2 理论震源模型S_T表示, 即

$S_{T} (f) = {(2 π f)}^{2} C M_{0} / [1 + (f / f_{c})^{2}], C = (R_{ΘΦ} VF) / (4 π ρ β^{3} R_{0})$ (4)

Figure Option
View Download New Window

图 4 a 单个地震的反演震源谱及伽师M_S6.4地震反演震源谱拟合结果; b 伽师M_S6.4地震反演与理论震源谱拟合残差随拐角频率f_c的变化及震源参数f_c-M_W的估计值, c κ 估计值Fig. 4 Inverted source spectra for individual earthquakes and the best-fitted theoretical spectra for the Jiashi M_S6.4 earthquake(a), the normalized fitting residuals between the inverted and theoretical source spectra varying with corner frequency f_c, and the f_c-M_W plots(b), and the estimated κ values(c).

其中, M₀、 f_c分别为地震矩和拐角频率, R_{Θ Φ}、 V、 F分别表示平均辐射模式因子、水平向S波占S波总能量的比例、自由表面放大因子, 取值分别为0.55、 0.707、 2(Boore, 2003), ρ =2.61gcm^-3为地壳介质密度(赵翠萍等, 2008)。利用Abercrombie等(2017)提出的方法估计地震震源参数(M₀、 f_c)并判断估计结果的可靠性, 图4b给出了伽师M_S6.4地震反演震源谱与理论震源谱的归一化拟合残差(拟合残差与拟合残差最小值之比)随拐角频率f_c取值的变化以及得到的震源参数(M_W、 f_c), M_W=()lgM₀-10.7(Hanks et al., 1979)。本文确定的伽师地震矩震级M_W=5.893, f_c=0.362Hz, 该结果介于中国地震台网中心给出的M_W=5.8与美国地质调查局(USGS)给出的M_W=6.03之间。利用M₀、 f_c计算地震应力降Δ σ (Brune, 1970),

$Δ σ = (7 M_{0} / 16) (2 π f_{c} {/ 2.34 β)}^{3}$ (5)

本文选用的46次地震的应力降为1.126~16.455MPa, 其平均应力降为3.942MPa, 对数标准差(lg)为0.284, 其中伽师M_S6.4地震的应力降为6.684MPa, 明显高于赵翠萍(2006)通过反演地震震源破裂过程给出的1998年2次伽师强震(M_S6.0、 M_S6.4)和2003年伽师-巴楚M_S6.8地震的破裂面平均应力降(< 1MPa), 但更接近破裂面上最大滑动位错处的应力降(1.3MPa、 7.7MPa、 3.8MPa), 更小的平均应力降可能与设定的破裂面过大有关。

采用高频衰减项exp(-π fκ )表示震源谱在f> f_max的高频段的快速衰减, 其中κ 为高频衰减系数, 可将高频段震源谱表示为

$S (f) = S_{T} (f) \exp (- π fκ) / \exp (- π f_{\max} κ), f > f_{\max}$ (6)

拟合f> f_max高频段震源谱估算κ 及f_max, 所选用的地震震源谱的f_max为10~20Hz, κ 为0.053~0.094s, 其均值为0.071s(图 4c), 其中伽师M_S6.4地震的f_max和κ 分别为13.8Hz和0.069s。

1.5 地震动路径持时

地震动持时由震源持时和路径持时2部分组成。震源持时以地震拐角频率的倒数表示, 计算得到本文谱反演选用的46次地震的震源持时, 进一步计算了这46次地震中470组强震动记录的5%~95%显著持时(D_5%~95%)。为了确保地震动D_5%~95%的部分主要为S波, 本文以2倍的20%~80%显著持时(D_20%~80%)表示D_5%~95%(Boore et al., 2015)。伽师地区地震动路径持时如图 5 所示, 根据距离分段的路径持时均值, 本文给出了伽师地区地震动路径持时(d_P)的经验模型:

$d_{P} = \{\begin{array}{l} 0.147 R & R = 0 ~ 17 km \\ 0.769 (R - 17) + 2.5 & R = 17 ~ 30 km \\ 0.308 (R - 30) + 12.5 & R = 30 ~ 90 km \\ - 0.187 (R - 90) + 31.0 & R = 90 ~ 165 km \\ 17.0 & R = 165 ~ 250 km \end{array}$ (7)

	Figure Option View Download New Window
	图 5 伽师地区强震动记录的路径持时及经验模型Fig. 5 Path duration of ground motion and its empirical model in Jiashi region.

同时, 对比了利用2008年汶川地震的132次余震、 2013年M_S7.0芦山地震的30次余震及2017年九寨沟M_S7.0地震的强震动记录得到的四川地区地震动路径持时(图 5)。伽师地区与四川地区地震动路径持时存在显著的差异, 伽师地区路径持时整体上高于四川地区, 且在30~150km震源距范围内尤为明显。此外, 还注意到当震源距< 150km时伽师地区与美国东部的地震动路径持时较为接近(Boore et al., 2015)。

1.6 场地效应模型

本文经谱反演给出了25个台站的场地反应(图 6), 谱反演选用的记录均为PGA≤ 100cm/s²的弱震记录, 可近似认为场地不出现非线性反应, 本文得到的场地反应均表示线性场地反应。台站场地的30m土层等效剪切波速(V_S30)为220~290m/s, 按V_S30 将台站分为2组, 分别为220~250m/s和250~290m/s, 2组台站场地V_S30 的平均值为233.77m/s和269.33m/s, 2组台站的平均场地反应及± 1倍标准差的范围如图 6 所示, V_S30 较小的台站组的平均场地反应整体上强于较大组。此外, 对比Seyhan等(2014)根据美国场地资料建立的场地线性反应经验模型(SS14模型)可知, 本文得到的伽师地区高频段场地反应与SS14线性模型差异明显, SS14线性模型显著地高估了伽师地区10~30Hz频段的场地反应。针对伽师地区场地反应结果, 本文调整了SS14场地线性反应模型以适用于伽师地区, 经调整后的模型更有效地反映了伽师地区的场地反应(图 6)。

	Figure Option View Download New Window
	图 6 反演给出的台站场地反应及SS14场地线性反应调整模型Fig. 6 Inverted site responses and the modified SS14 linear site response model.

2 伽师M_S6.4地震的地震动模拟

2.1 震源运动学破裂随机模型

据中国地震台网中心测定的伽师M_S6.4地震震中位置为(39.83° N, 77.21° E), 震源深度为16km。根据USGS提供的震源机制解可知, 此次地震为逆冲型地震, 震源破裂面的走向角和倾角分别为221° 和20° 。根据Wells等(1994)建立的逆冲地震破裂面几何尺寸(长度L和宽度W)、破裂面平均滑动位错(D_Avg)与M_W的经验关系, 将伽师M_S6.4地震破裂面的L、 W、 D_Avg分别设定为10km、 6km、 32.8cm, 将破裂面划分为1km× 1km的60个子断层。地震起始破裂点在破裂面上的位置服从特定的概率分布(Mai et al., 2005), 走滑地震和地壳内倾滑地震的起始破裂点在破裂面上沿倾向和沿走向的位置分别服从韦布尔(Weibull)分布和正态分布, 沿倾向和沿走向的位置分别于0.6W和0.5L时出现概率最大, 本文设定伽师M_S6.4地震的起始破裂点在破裂面上沿倾向和沿走向的位置分别为3.6km和5km, 此时破裂面上边缘的深度为14.77km。

本文采用Graves等(2010, 2015, 2016)提出的方法建立滑动位错不均匀、破裂速度与滑动位错相关的震源运动学随机破裂模型, 该方法基于冯· 卡门相关函数的波数域滤波方法在波数域生成二维随机场, 然后将其转换至空间域, 并规定子断层滑动位错服从均值为D_Avg、标准差为0.8D_Avg的正态分布, 最后得到滑动分布不均匀的震源运动学模型。本文生成了200个伽师M_S6.4地震的震源随机破裂模型, 其中4次破裂模型如图 7 所示, 震源随机破裂模型的明显差异说明生成的破裂模型具有很好的随机性, 可以较为充分地描述破裂过程的不确定性, 每个随机震源模型都有可能代表了伽师M_S6.4地震的破裂过程, 后文将分别采用每个随机震源模型给出地震动模拟结果, 200个震源随机破裂模型的模拟结果可近似代表设定地震的地震动, 并能反映出与破裂过程相关的地震动的不确定性。

	Figure Option View Download New Window
	图 7 伽师M_S6.4地震的震源随机破裂模型Fig. 7 Source rupture models stochastically generated for the Jiashi M_S6.4 earthquake.

2.2 随机有限断层方法模拟结果

地震震源破裂面可划分为N个可看做点源的子断层, 叠加各子断层产生的地震动a_i得到整个破裂面产生的地震动, 即

$a (t) = \sum_{i = 1} a_{i} (t - Δ t_{i})$ (8)

地震破裂自起始破裂点开始并辐射至整个破裂面, Δ t_i表示第i个子断层的地震动相对起始破裂点的延迟, 包括破裂延迟和传播延迟, 采用随机点源方法模拟子断层产生的地震动(Boore, 2003), 第i个子断层产生的地震动的S波傅氏谱以子断层震源谱(式(4)、 (6))、路径衰减经验模型(式(3))及场地效应模型表示, 转换至时域得到a_i, 采用Saragoni等(1973)给出的窗函数约束地震动加速度时程的波形。本文利用随机有限断层方法模拟了伽师M_S6.4地震中震源距≤ 200km的39个触发强震动台站(图 2)的水平向地震动加速度时程, 模拟输入参数见表1。

表1 随机有限断层模拟输入参数 Table 1 Input parameters in the stochastic finite-fault simulation

图 8 给出了200次震源随机破裂过程的水平向模拟记录的PGA、峰值地面速度(PGV)以及部分周期(0.3s和3.0s)的5%阻尼比拟加速度反应谱(PSA)的平均值及1倍标准差范围。模拟记录地震动强度指标整体上可以较好地代表观测值, 能体现出地震动的近场饱和效应及远场衰减规律; 1倍标准差范围体现了与破裂过程相关的地震动不确定性, 地震动的不确定性表现出周期相关性, 长周期地震动的不确定性更明显。对比俞言祥等(2013)建立的新疆地震区基岩场地PGA预测方程的PGA预测值, 长轴方向与破裂面的走向一致(220° N)。为了体现场地条件的影响, 根据中国地震动参数区划图(中华人民共和国国家质量监督检验检疫总局, 2015)中不同场地类别相对于Ⅱ 类场地的地震动峰值加速度调整系数, 得到考虑场地条件影响的PGA预测值, 模拟记录的PGA均值和PGA预测值也具有很好的一致性。模拟记录的PSA均值与观测记录的PSA残差(log₁₀(Y_o)-log₁₀(Y_s))如图 9 所示, 短周期(< 0.3s)的PSA模拟值普遍高估了观测值, 但整体上可以很好地预测0.3~10s周期段的地震动。考虑伽师地区地震动区域特征模型之后, 地震动的模拟结果与观测记录强度指标较好的一致性说明本文给出的伽师地区地震动区域特征较为可靠。

	Figure Option View Download New Window
	图 8 对比模拟与观测记录的PGA、 PGV、周期为0.3s和3.0s的PSAFig. 8 Comparisons of PGA, PGV, and PSAs at 0.3s and 3.0s between observations and simulations.

	Figure Option View Download New Window
	图 9 模拟记录的PSA残差Fig. 9 Residuals for the simulated PSAs at 42 strong-motion stations.

为了说明区域特征对地震动模拟结果的影响, 本文也采用川滇地区的地震动衰减模型(Xu et al., 2010)代替伽师地区衰减模型给出了伽师M_S6.4地震的地震动(图 8), 衰减模型对高频地震动(PGA、周期为0.3s的PSA)模拟结果的影响不明显, 但对远场(> 60km)低频地震动(PGV、周期为3.0s的PSA)影响显著, 川滇地区衰减模型不能有效地反映伽师地区低频地震动的远场快速衰减特征。

本文从200次震源随机破裂过程的模拟记录中选取了与观测记录PSA最接近的模拟记录所对应的一次随机破裂过程(M-65)(图 7)作为伽师M_S6.4地震的震源破裂模型, 采用本文给出的伽师地区地震动区域特征模型, 模拟了M-65随机破裂模型的地震动场, 包含了以震中为中心200km范围内辐射状分布的3 636个虚拟观测点的水平向加速度时程, 虚拟观测点V_S30 均设定为250m/s(相当于表2中Ⅲ 类场地V_S30 的平均值)。图 10 给出了伽师M_S6.4地震模拟记录PGA、 PGV、周期为0.3s和3.0s的PSA的空间分布情况, 近断层区域的PGA> 300cm/s²。由于不考虑局部场地条件的影响, 远场地震动强度指标等值线近似为圆形, 近场区域受破裂过程的影响显著, 地震动强度等值线不规则。

	Figure Option View Download New Window
	图 10 伽师M_S6.4地震模拟记录PGA、 PGV、周期为0.3s和3.0s的PSA的空间分布蓝色矩形框为破裂面的水平投影Fig. 10 Spatial distributions for PGA, PGV, and PSA at 0.3s and 3.0s produced for the M_S6.4 Jiashi earthquake.

表2 震源距≤ 200km的39个触发台站的场地条件 Table 2 Site information for 39 stations that recorded the Jiashi M_S6.4 earthquake

3 结论

本文利用伽师地区强震动观测记录, 基于非参数谱反演方法分离地震动的震源、传播路径和场地影响, 建立了体现伽师地区地震动区域特征的路径衰减、路径持时、场地线性反应经验模型, 根据谱反演得到的地震震源谱估计了46次伽师地区中小地震的地震震源参数。 2020年伽师M_S6.4地震的矩震级M_W=5.893、拐角频率f_c=0.362Hz、应力降Δ σ =6.684MPa, 利用伽师地区地震动区域模型基于随机有限断层方法模拟了2020年伽师M_S6.4地震的地震动场, 取得的主要认识如下:

(1)强震动记录含有地震动的震源、路径衰减及局部场地影响的重要信息, 非参数谱反演方法能够从地震动中分离出这些重要信息, 进而建立体现地震动区域特征的经验模型; 在地震动模拟中考虑地震动区域特征, 可以有效地提高模拟结果的可靠性。

(2)在强震动记录数不足以建立地震动预测模型的地区, 可以利用有限的强震动记录得到地震动区域性特征, 并给出模拟地震动, 结合少量观测记录与大量模拟记录建立地震动预测模型。

(3)伽师地区地震动的近场路径衰减较慢, 由于显著的非弹性衰减, 远场路径衰减较快, 伽师地区的路径衰减可近似地表示为双段线性几何扩散经验模型和非弹性衰减项的叠加, 几何扩散经验模型的转换距离R₁=60km, 前、后2段的几何扩散指数n₁和n₂分别为0.30和0.59, 品质因子Q=60.066f^0.988; 伽师地区地震动的路径持时较长, 明显高于四川地区; 对比SS14场地的线性反应模型, 伽师地区局部场地对10~30Hz频段地震动的线性放大效应很弱, 其他频段的场地线性反应与SS14模型接近。

致谢伽师M_S6.4地震虚拟科考工作队在本研究基础数据获取方面给予了协调; 中国地震局工程力学研究所 “ 国家强震动台网中心” 和新疆维吾尔自治区地震局为本研究提供了强震动观测数据和强震动台站场地钻孔资料。在此一并表示感谢!

参考文献

文献列表

[1]	邓起东, 张培震, 冉勇康, 等. 2003. 中国活动构造与地震活动[J]. 地学前缘, 10(S1): 66—73. DENG Qi-dong, ZHANG Pei-zhen, RAN Yong-kang, et al. 2003. Active tectonics and earthquake activities in China[J]. Earth Science Frontiers, 10(S1): 66—73(in Chinese). [本文引用:1]
[2]	李文倩, 朱皓清, 何金刚. 2020. 2020年1月19日新疆伽师 M_S6. 4地震强震动记录特征初步分析[J]. 内陆地震, 34(1): 63—69. LI Wen-qian, ZHU Hao-qing, HE Jin-gang. 2020. Characteristics of strong motion records of Xinjiang Jiashi M_S6. 4 earthquake on January 19, 2020[J]. Inland Earthquake, 34(1): 63—69(in Chinese). [本文引用:1]
[3]	刘启元, 陈九辉, 李顺成, 等. 2000. 新疆伽师强震群三维地壳上地幔S波速度结构及其地震成因的探讨[J]. 地球物理学报, 43(3): 356—365. LIU Qi-yuan, CHEN Jiu-hui, LI Shun-cheng, et al. 2000. Passive seismic experiment in Xinjiang Jiashi strong earthquake region and discussion on its seismic genesis[J]. Chinese Journal of Geophysics, 43(3): 356—365(in Chinese). [本文引用:1]
[4]	徐锡伟, 张先康, 冉勇康, 等. 2006. 南天山地区巴楚-伽师地震( M_S6. 8)发震构造初步研究[J]. 地震地质, 28(2): 161—178. XU Xi-wei, ZHANG Xian-kang, RAN Yong-kang, et al. 2006. The preliminary study on seismotectonics of the 2003AD Bachu-jiashi earthquake( M_S6. 8), southern Tian Shan[J]. Seismology and Geology, 28(2): 161—178(in Chinese). [本文引用:1]
[5]	俞言祥, 李山有, 肖亮. 2013. 为新区划图编制所建立的地震动衰减关系[J]. 震灾防御技术, 8(1): 24—33. YU Yan-xiang, LI Shan-you, XIAO Liang. 2013. Development of ground motion attenuation relations for the new seismic hazard map of China[J]. Technology for Earthquake Disaster Prevention, 8(1): 24—33(in Chinese). [本文引用:1]
[6]	张振斌, 王宝柱, 朱皓清. 2019. 利用2018年伽师小震群强震动记录分析区域场地特征[J]. 地震地磁观测与研究, 40(6): 22—27. ZHANG Zhen-bin, WANG Bao-zhu, ZHU Hao-qing. 2019. Analysis of regional site characteristics using strong motion records of Jiashi small earthquake swarm in 2018[J]. Seismological and Geomagnetic observation and Research, 40(6): 22—27(in Chinese). [本文引用:1]
[7]	赵翠萍. 2006. 1997—2003年新疆伽师震源区特征的地震学方法研究[D]. 北京: 中国地震局地球物理研究所: 110—117. ZHAO Cui-ping. 2006. Seismological studies on the characteristics of Jiashi source region from 1997 to 2003 [D]. Institute of Geophysics, China Earthquake Administration, Beijing: 110—117(in Chinese). [本文引用:1]
[8]	赵翠萍, 陈章立, 郑斯华, 等. 2008. 伽师震源区中等强度地震矩张量反演及其应力场特征[J]. 地球物理学报, 51(3): 782—792. ZHAO Cui-ping, CHEN Zhang-li, ZHENG Si-hua, et al. 2008. Moment inversion of moderate earthquakes and the locally perturbed stress field in the Jiashi source region[J]. Chinese Journal of Geophysics, 51(3): 782—792(in Chinese). [本文引用:1]
[9]	中华人民共和国国家质量监督检查检疫总局. 2015. 中国地震动参数区划图(GB 18306-2015)[S]. 北京: 中国标准出版社. General Administration of Quality Supervision, Inspection and Quarantine of the People’s Republic of China. 2015. Seismic Ground Motion Parameters Zonation Map of China(GB 18306-2015)[S]. Stand ards Press of China, Beijing(in Chinese). [本文引用:1]
[10]	中华人民共和国住房和城乡建设部. 2010. 建筑抗震设计规范(GB 50011-2010)[S]. 北京: 中国建筑工业出版社. Ministry of Housing and Urban-Rural Development of the People’s Republic of China(MOHURD). 2010. Code for Seismic Design of Building (G B 50011-2010)[S]. China Architecture & Building Press, Beijing(in Chinese). [本文引用:1]
[11]	Abercrombie R E, Bannister S, Ristau J, et al. 2017. Variability of earthquake stress drop in a subduction setting, the Hikurangi margin, New Zealand [J]. Geophysical Journal International, 208(1): 306—320. [本文引用:1]
[12]	Abrahamson N, Gregor N, Addo K. 2016. BC hydro ground motion prediction equations for subduction earthquake[J]. Earthquake Spectra, 32(1): 23—44. [本文引用:1]
[13]	Abrahamson N A, Silva W J. 1997. Empirical response spectral attenuation relations for shallow crustal earthquakes[J]. Seismological Research Letters, 68(1): 94—127. [本文引用:1]
[14]	Atkinson G M, Boore D M. 2006. Earthquake ground-motion prediction equations for eastern North America[J]. Bulletin of the Seismological Society of America, 96(6): 2181—2205. [本文引用:1]
[15]	Boore D M. 2003. Simulation of ground motion using the stochastic method[J]. Pure and Applied Geophysics, 160(3-4): 635—676. [本文引用:2]
[16]	Boore D M, Stewart J P, Seyhan E, et al. 2014. NGA-West2 equations for predicting PGA, PGV, and 5%damped PSA for shallow crustal earthquakes[J]. Earthquake Spectra, 30(3): 1057—1085. [本文引用:1]
[17]	Boore D M, Thompson E M. 2015. Revisions to some parameters used in stochastic-method simulations of ground motion[J]. Bulletin of the Seismological Society of America, 105(2A): 1029—1041. [本文引用:3]
[18]	Boore D M, Thompson E M, Cadet H. 2011. Regional correlations of V_S30 and velocities averaged over depths less than and greater than 30 meters[J]. Bulletin of the Seismological Society of America, 101(6): 3046—3059. [本文引用:1]
[19]	Bora S S, Scherbaum F, Kuehn N, et al. 2015. Development of a response spectral ground-motion prediction equation(GMPE)for seismic-hazard analysis from empirical Fourier spectra and duration models[J]. Bulletin of the Seismological Society of America, 105(4): 2192—2218. [本文引用:1]
[20]	Brune J N. 1970. Tectonic stress and the spectra of seismic shear waves from earthquakes[J]. Journal of Geophysical Research, 75(26): 4997—5009. [本文引用:2]
[21]	Castro R R, Anderson J G, Singh S K. 1990. Site response, attenuation and source spectra of S waves along the Guerrero, Mexico, subduction zone[J]. Bulletin of the Seismological Society of America, 80(6A): 1481—1503. [本文引用:1]
[22]	Graves R W, Pitarka A. 2010. Broadband ground-motion simulation using a hybrid approach[J]. Bulletin of the Seismological Society of America, 100(5A): 2095—2123. [本文引用:1]
[23]	Graves R W, Pitarka A. 2015. Refinements to the Graves and Pitarka(2010)broadband ground-motion simulation method[J]. Seismological Research Letters, 86(1): 75—80. [本文引用:1]
[24]	Graves R W, Pitarka A. 2016. Kinematic ground-motion simulations on rough faults including effects of 3D stochastic velocity perturbations[J]. Bulletin of the Seismological Society of America, 106(5): 2136—2153. [本文引用:1]
[25]	Hanks T C, Kanamori H. 1979. A moment magnitude scale[J]. Journal of Geophysical Research, 84(B5): 2348—2350. [本文引用:2]
[26]	Kale S, Akkar S, Ansari A, et al. 2015. A ground-motion predictive model for Iran and Turkey for horizontal PGA, PGV, and 5%damped response spectrum: Investigation of possible regional effects[J]. Bulletin of the Seismological Society of America, 105(2A): 963—980. [本文引用:1]
[27]	Laborde A, Barrier L, Simoes M, et al. 2019. Cenozoic deformation of the Tarim Basin and surrounding ranges(Xinjiang, China): A regional overview[J]. Earth-Science Reviews, 197: 102891. [本文引用:1]
[28]	Mai M, Spudich P, Boatwright J. 2005. Hypocenter locations in finite-source rupture models[J]. Bulletin of the Seismological Society of America, 95(3): 965—980. [本文引用:1]
[29]	Nakamura Y. 1989. A method for dynamic characteristics estimation of subsurface using microtremor on the ground surface [R]. Quarterly Report, Railway Technical Research Institute. [本文引用:1]
[30]	Saragoni G R, Hart G C. 1973. Simulation of artificial earthquakes[J]. Earthquake Engineering and Structural Dynamics, 2(3): 249—267. [本文引用:1]
[31]	Seyhan E, Stewart J P. 2014. Semi-empirical nonlinear site amplification from NGA-West2 data and simulations[J]. Earthquake Spectra, 30(3): 1241—1256. [本文引用:1]
[32]	Wells D L, Coppersmith K J. 1994. New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface displacement[J]. Bulletin of the Seismological Society of America, 84(4): 974—1002. [本文引用:1]
[33]	Xu P B, Ren Y F, Wen R Z, et al. 2020. Observations on regional variability in ground-motion amplitude from six M_W~6. 0 earthquakes of the north-south seismic zone in China[J]. Pure and Applied Geophysics, 177(1): 247—264. [本文引用:1]
[34]	Xu Y, Herrmann R B, Wang C Y, et al. 2010. Preliminary high-frequency ground-motion scaling in Yunnan and southern Sichuan, China[J]. Bulletin of the Seismological Society of America, 100(5B): 2508—2517. [本文引用:2]
[35]	Yenier E, Atkinson G M. 2015. An equivalent point-source model for stochastic simulation of earthquake ground motions in California[J]. Bulletin of the Seismological Society of America, 105(3): 1435—1455. [本文引用:1]

2003

0.0

2020

0.0

... 李文倩等, 2020), 但尚未系统地给出伽师地区地震动的区域特征以及可用于地震动预测的经验模型 ...

2000

0.0

刘启元, 陈九辉, 李顺成, 等. 2000. 新疆伽师强震群三维地壳上地幔S波速度结构及其地震成因的探讨[J]. 地球物理学报, 43(3): 356—365.

LIU

Qi-yuan

, CHEN

Jiu-hui

, LI

Shun-cheng

, et al. 2000. Passive seismic experiment in Xinjiang Jiashi strong earthquake region and discussion on its seismic genesis[J]. Chinese Journal of Geophysics, 43(3): 356—365(in Chinese).

From January 21 to April 16 of 1997, seven strong earthquakes took place in Jiashi region of Xinjiang, China. This Phenomenon was never happened in Chinese continent since earthquakes were able to be recorded by instruments. To investigate the deep structure environment of the Jiashi strong earthquake sequence, a portable broadband seismic array of 30 stations was set up in Jiashi region over about 11,000 km 2 . The portable array crossed the Tarim basin and the Tian Shan fold belt. The distance between stations is about 5-10 km. The frequency bandwidth of our observation system is 0.05--20 Hz. The 3--D S-wave velocity structure down to 100 km beneath the army was obtained by using the teleseismic receiver function migration and its nonlinear inversion techniques （Liu etc. 1996）. The following conclusions can be obtained from our observations: l) The Moho depth is about 4052 kin on the side of the Tarim basin and becomes 60-76 km deep on the side close to the Tien Shan. The crustal thickness beneath the fold belt nearby the south Tien Shan is much thicker than that of the Tarim basin and the obvious tectonic deformation exists in the crust. 2) The north boundary of the Tarim basin is clearly laterally inhomogeneous. The relationship of the contact deformation between Tarim and Tien Shan reveals the compressional movement to the NNW direction. The topography of the Moho discontinuity beneath our array has obvious correlation with the arctuate tectonics on the surface. The arctuate fold formation is closely related with the deep structure and the compression by the Tarim basin. 3) The Jiashi strong earthquake sequence mainly occurred over the gradient zone of the crust--mantle boundary. The genesis of the Jiashi earthquake sequence probably related to the activity of the conceal faults nearby the sources. This may result from the inhomogeneous deformation at the north boundary and the current tectonic movement of the Tarim basin. 4) The Jiashi strong earthquake sequence mainly occurred in the upper crust beneath the fold belt with a lower shear module. This could be a reason why the stress drop of these events is much smaller than common. All of these demonstrate that the Jiashi earthquake sequence genesis is related not only to the current tectonic movement of the Tarim basin, but also to the deep structure. This will need to be investigated further by the numerical stress modeling.

Institute of Geology, China Seismological Bureau, Beijing 100029, China

１９９７年１月２１日至４月１６日（北京时间）新疆伽师地区连续发生了７次６级以上强烈地震．为了深入研究该强震群形成的构造背景，我们在伽师强震群区及其邻域大约１１０００ｋｍ２的范围内布设了由３０台宽频带数字地震仪组成的流动地震台阵，台站的间距约５－１０ｋｍ．利用远震体波接收函数的叠加偏移分析及其非线性反演技术，研究了台阵下方０－１００ｋｍ深度范围内地壳、上地幔的三维Ｓ波速度结构．结果表明：１)Ｍｏｈｏ界面的深度在塔里木盆地一侧为４０—５２ｋｍ，而在靠近天山一侧的褶皱变形区为６０—７６ｋｍ，南天山山前折皱变形区的地壳厚度明显增大，地壳内部产生了明显变形；２)塔里木盆地北缘存在明显的不均匀性，塔里木盆地与天山之间的接触变形关系显示了塔里木盆地向北北西方向的挤压作用，台阵下方壳幔界面的地形与地表的地貌特征有较好的相关性，山前弧形褶皱带的形成与地下深部结构及挤压作用有关；３)伽师强震群分布在壳幔界面梯度带的上方，该震群的成因可能与震源区附近的隐伏断裂活动有关；４)由于震源处于褶皱变形区的地壳上部，相应的剪切模量较小，这可能是伽师强震群应力降明显偏低的...

... 0 引言伽师地区位于南天山西段与塔里木盆地西北部的交界区域, 受南天山向S的挤压作用, 北部发育有近EW向展布的柯坪塔格逆冲褶皱带, 位于迈丹断裂与柯坪塔格逆冲断裂之间, 南部塔里木盆地内则发育有隐伏断裂(刘启元等, 2000 ...

2006

0.0

徐锡伟, 张先康, 冉勇康, 等. 2006. 南天山地区巴楚-伽师地震( M_S6. 8)发震构造初步研究[J]. 地震地质, 28(2): 161—178.

Xi-wei

, ZHANG

Xian-kang

, RAN

Yong-kang

, et al. 2006. The preliminary study on seismotectonics of the 2003AD Bachu-jiashi earthquake( M_S6. 8), southern Tian Shan[J]. Seismology and Geology, 28(2): 161—178(in Chinese).

Owing to strong and permanent Cenozoic re-orogenic processing, a lot of EW-striking active thrusts and folds have been developed in Tian Shan, resulting in crustal shortening in NS direction. There also exist NW-striking transform-like strike-slip faults that cut the Tian Shan and accommodate uneven crustal shortening larger in the west and smaller in the east. The seismogenic structures in and around the Tian Shan mainly include EW-striking thrust ramps or blind thrusts and NW-striking transform-like strike-slip faults. The 2003 AD Bachu-Jiashi earthquake is located at south of the Kalpintag nappe. A NE-trending deep seismic reflection profile about 50km long across the epicenter has been conducted after the earthquake. From this reflection profile four blind faults are identified. Together with earthquake relocation, these identified blind faults are used in the paper to interpret the seismogenic structures of the 1997 AD Jiashi strong earthquake swarm and 2003 AD Bachu-Jiashi earthquake. The 1997 AD Jiashi strong earthquakes were generated mainly by a NW-striking buried transform-like strike-slip fault, while the 2003 AD Bachu-jiashi earthquake by blind thrusts in front of the Kalpintag nappe.

1. Institute of Geology, China Earthquake Administration, Beijing 100029, China; 2. Geophysical Exploration Center, China Earthquake Administration, Zhengzhou 450003, China; 3. Institute of Crustal Dynamics, China Earthquake Administration, Beijing 100085, China; 4. Earthquake Administration of Xinjiang Uygur Autonomous Region, Urumqi 830011, China

新生代期间强烈而持久的再生造山作用,在天山地区形成了大量近EW向逆断裂-褶皱带,引起地壳强烈缩短,穿插有NW向“类转换断层”,显示出天山地区近NS向不均匀的构造挤压作用;区域上地震构造主要为近EW向逆断裂-褶皱带或盲逆断层,其次为NW向“类转换断层”。巴楚-伽师地震区位于南天山柯坪塔格推覆构造系以南,NE向跨越极震区、长约50km的深地震反射探测表明,1997年伽师强震群的发震构造推测为NW向隐伏“类转换断层”,2003年巴楚-伽师地震( M S 6.8)的发震构造为柯坪塔格推覆构造系南缘尚未出露地表的近EW向盲逆断层系。

... 徐锡伟等, 2006), 诸多研究认为塔里木刚性块体向北部远程传递的印度板块向N推挤欧亚板块的压应力与该地区的构造变形有关(Laborde et al ...

2013

0.0

... 对比俞言祥等(2013)建立的新疆地震区基岩场地PGA预测方程的PGA预测值, 长轴方向与破裂面的走向一致(220#cod#x000b0 ...

2019

0.0

... 目前已有学者利用伽师地区单次或几次地震的强震动观测记录分析地震动的路径衰减、持时、场地效应等(张振斌等, 2019 ...

2006

0.0

... 684MPa, 明显高于赵翠萍(2006)通过反演地震震源破裂过程给出的1998年2次伽师强震(M_S6 ...

2008

0.0

赵翠萍, 陈章立, 郑斯华, 等. 2008. 伽师震源区中等强度地震矩张量反演及其应力场特征[J]. 地球物理学报, 51(3): 782—792.

ZHAO

Cui-ping

, CHEN

Zhang-li

, ZHENG

Si-hua

, et al. 2008. Moment inversion of moderate earthquakes and the locally perturbed stress field in the Jiashi source region[J]. Chinese Journal of Geophysics, 51(3): 782—792(in Chinese).

Moment tensors of 52 moderate earthquakes occurred in Jiashi source region, Xinjiang, from 1997 to 2004 were determined in time domain by using the long period waveforms recorded by regional broadband stations. The results demonstrate significant variations of the orientation and plunge of P, T and N axis from west to east. Therefore, we further divided the whole region into two parts, which are called the western part and the eastern part, and performed the stress tensor inversion respectively. The results show that in the eastern part, the maximum principal axis σ 1 is almost horizontal and trends 321°, the minimum principal axis σ 3 trends 68° with a plunge of 40°, and σ 2 trends 213° with a plunge about 40°. Till 2004, the stress within the western part remained stable, with the σ 1 and σ 3 nearly horizontal and trending to 12° and 282° respectively, the middle axis σ 2 almost vertical. Thus the maximum principal stress σ 1 significantly rotates counterclockwise from N10°E in the west part to N39°W in the east part across the whole studied region. Compared with the stress field in the eastern part, the western part is under a stronger horizontal tension effect. This bilateral perturbation of stress field is consistent with the pattern of crustal deformation rate from GPS observation, as well as the topography map at the structure boundary between Tarim basin and Tienshan. We infer that both the sharp deformation in the deep crust and the large-scale NW-trending Puchang and Selibuya fault structures may play a crucial role in forming the local stress perturbation.

Institute of Earthquake Science, China Earthquake Administration, Beijing 100036,China

使用区域数字地震台站记录的宽频带长周期波形资料，在时间域反演了1997~2004年伽师震源区52次中等强度地震的矩张量.反演结果揭示，在小尺度的伽师震源区内，震源机制解的 P 轴、 T 轴和 N 轴呈现出明显的分区特征.本文进一步把伽师震源区分为东区和西区，分别反演了东区与西区的应力场.应力场反演结果表明，东区的应力场主压应力轴走向为321°，基本水平.最大主张应力走向68°，倾角40°.截至2004年7月，伽师震源区西部的应力场一直较为稳定，最大主压应力方向为12°，最大主张应力方向282°，二者都基本水平，中等主应力轴基本直立.自西向东，伽师震源区最大主压应力轴逆时针旋转了49°，并且西区张应力的水平作用较为显著，东区压应力的水平作用显著.应力场的这种非均匀变化特征与GPS观测得到的地壳运动速率的空间分布以及塔里木盆地边界附近的地形地貌特征有很好的一致性.震源区深部结构的陡变以及位于震源区东部边界规模较大的NW走向的普昌断裂和色力布亚隐伏断裂可能对产生这种横向非均匀的局部应力场起了重要的作用.

... 61gcm^-3为地壳介质密度(赵翠萍等, 2008) ...

2015

0.0

... 为了体现场地条件的影响, 根据中国地震动参数区划图(中华人民共和国国家质量监督检验检疫总局, 2015)中不同场地类别相对于#cod#x02161 ...

2010

0.0

... 根据台站场地钻孔资料及建筑抗震设计规范对场地类别划分的规定(中华人民共和国住房和城乡建设部, 2010), 有7个台站场地可划分为#cod#x02161 ...

2017

0.0

... 利用Abercrombie等(2017)提出的方法估计地震震源参数(M₀、 f_c)并判断估计结果的可靠性, 图4b给出了伽师M_S6 ...

2016

0.0

... 为了清晰地识别S波部分并尽可能消除触发阈值引起的地震动偏差(Abrahamson et al ...

1997

0.0

... 25f_hp(Abrahamson et al ...

2006

0.0

2003

0.0

... 707、 2(Boore, 2003), #cod#x003c1 ...

... t_i表示第i个子断层的地震动相对起始破裂点的延迟, 包括破裂延迟和传播延迟, 采用随机点源方法模拟子断层产生的地震动(Boore, 2003), 第i个子断层产生的地震动的S波傅氏谱以子断层震源谱(式(4)、 (6))、路径衰减经验模型(式(3))及场地效应模型表示, 转换至时域得到a_i, 采用Saragoni等(1973)给出的窗函数约束地震动加速度时程的波形 ...

2014

0.0

... Boore等(2014)利用NGA-West2计划的强震动数据集揭示了不同区域(美国西部、日本、意大利、中国台湾、四川等)地震动非弹性衰减的显著差异及其对远场地震动衰减的影响 ...

2015

0.0

... Boore等(2015)的研究表明美国东部地震动路径持时明显高于美国西部, 这可能与地壳介质差异导致的地震波衰减差异有关 ...

... 为了确保地震动D_5%~95%的部分主要为S波, 本文以2倍的20%~80%显著持时(D_20%~80%)表示D_5%~95%(Boore et al ...

... 150km时伽师地区与美国东部的地震动路径持时较为接近(Boore et al ...

2011

0.0

2015

0.0

... 采用频率无关的几何扩散项和频率相关的非弹性衰减项表示路径衰减, 不同几何扩散模型(线性、双段或三段线性模型等)在拟合地震动的距离衰减时差异不明显(Bora et al ...

1970

0.0

... ^-2 理论震源谱模型(Brune, 1970), 但震源谱的快速衰减则普遍出现在f#cod#x0003E ...

... (Brune, 1970), ...

1990

0.0

... 1 两步非参数谱反演方法两步非参数谱反演方法由Castro等(1990)提出 ...

2010

0.0

... 本文采用Graves等(2010, 2015, 2016)提出的方法建立滑动位错不均匀、破裂速度与滑动位错相关的震源运动学随机破裂模型, 该方法基于冯#cod#x000b7 ...

2015

0.0

2016

0.0

... 本文采用Graves等(2010, 2015, 2016)提出的方法建立滑动位错不均匀、破裂速度与滑动位错相关的震源运动学随机破裂模型, 该方法基于冯#cod#x000b7 ...

1979

0.0

... 7(Hanks et al ...

... 本文采用Graves等(2010, 2015, 2016)提出的方法建立滑动位错不均匀、破裂速度与滑动位错相关的震源运动学随机破裂模型, 该方法基于冯#cod#x000b7 ...

2015

0.0

... Kale等(2015)在研究伊朗和土耳其地区地震动预测模型时发现2个地区地表浅层土层的剪切波速结构显著影响地震动强度 ...

2019

0.0

... 徐锡伟等, 2006), 诸多研究认为塔里木刚性块体向北部远程传递的印度板块向N推挤欧亚板块的压应力与该地区的构造变形有关(Laborde et al ...

2005

0.0

... 地震起始破裂点在破裂面上的位置服从特定的概率分布(Mai et al ...

1989

0.0

... 利用水平垂直谱比法(Nakamura, 1989)计算了记录不少于5组的3个#cod#x02161 ...

1973

0.0

2014

0.0

... 此外, 对比Seyhan等(2014)根据美国场地资料建立的场地线性反应经验模型(SS14模型)可知, 本文得到的伽师地区高频段场地反应与SS14线性模型差异明显, SS14线性模型显著地高估了伽师地区10~30Hz频段的场地反应 ...

1994

0.0

... 根据Wells等(1994)建立的逆冲地震破裂面几何尺寸(长度L和宽度W)、破裂面平均滑动位错(D_Avg)与M_W的经验关系, 将伽师M_S6 ...

2020

0.0

... Xu等(2020)将中国南北地震带6次M_W6 ...

2010

0.0

... 与川滇地区的几何扩散和Q值(Xu et al ...

... 为了说明区域特征对地震动模拟结果的影响, 本文也采用川滇地区的地震动衰减模型(Xu et al ...

2015

0.0

... 高通滤波的拐角频率(f_hp)随记录变化: 首先根据2个经验关系给出f_hp的预估值, 分别为记录最小可用频率(f_lb)与矩震级M_W的经验关系(Yenier et al ...