Estimating Slab Earthquake Response Spectra from a 3 D Q Model by-猫眼课题宝

从 3 D Q 模型估计板地震响应谱

DOI：

发表时间：

2003

期刊：

影响因子：

通讯作者：

中科院分区：

文献类型：

作者：

D. Eberhart‐Phillips;G. McVerry

估计非均匀岩石圈的响应谱问题可以通过直接计算物理模型的衰减来解决。在俯冲带中，板块地震通过地幔楔的衰减与通过板块地震的衰减不同。本文主要关注通过新西兰北岛火山带下的衰减地幔的非常高的损失路径，其中低Q需要修改“标准”新西兰工程响应谱模型。强震数据的缺乏阻碍了对深板地震经过高度衰减的地幔带的路径进行标准回归分析。取而代之的是，利用北岛俯冲带2-20 Hz当地地震t*数据开发的3D频率无关Q模型进行了修改。通过将t*地壳结果校准到标准新西兰模型，可以比较标准模型和3D Q模型之间的振幅。确定每个源和站对的附加路径平均衰减率系数（CQ）。这导致CQ作为质心深度函数的简单表达式，用于修改标准模型。这种修正将火山区以下地幔楔路径的模式光谱降低了约2-4倍。这种减少与160公里深的6.0级地震在0.4秒以下的周期内观测到的反应谱变化相似。对于通过浅层火山区传播的浅层地震，新模型给出的结果与回归分析得出的火山路径衰减项相似。新西兰深板地震的非均质传播模式对地震地震动的工程预测具有重要意义。北岛位于Hikurangi俯冲带（图1）之上，这导致了陶波火山带（TVZ）的火山活动和伸展。由于板块运动和板块界面性质的沿走向变化，TVZ仅限于北岛的中部和北部，尽管俯冲的板块继续延伸到南岛北部（Reyners, 1998； Eberhart-Phillips等，2002；Upton等，2002）。来自地幔路径的地震图表明，与板块路径相比，强震数据衰减较大，但没有足够的强震数据来使用标准回归方法来模拟地幔中的衰减。深地震的几何图形如图2a所示。在40公里以下的深度，地震往往只发生在俯冲板块，因此位于TVZ正下方。俯冲的太平洋板块板块衰减极低，澳大利亚板块地壳衰减中等，板块与上覆地壳之间的地幔楔衰减极高。此外，TVZ中的地壳具有非常高的衰减。因此，在较远的东部有低损失的板块路径的地区，比如纳皮尔，深层地震的感觉会比在较近的有穿过地幔楔的路径的地区更强烈。深震已经在东部地区产生了触发强震的记录，但对于工程应用来说，包括高损失路径的深震响应，如TVZ的陶波和罗托鲁瓦以及西北部的其他地区，也必须进行估计。高衰减对地幔和TVZ的影响在实测资料和仪器资料中都可以清楚地看到。新西兰等震地图集（Downes, 1995）包含几张深震地图集，其中符合等震模式的椭圆长轴向震中以东偏移得很好（图2b）。这些有震感的地图为缺乏强震记录的地震提供了深入的了解。虽然没有正式的映射，但同样的效应可以在强震加速度计的记录中看到。在TVZ附近的几次深层地震中，在Napier和Gisborne附近记录到了最强的加速度，而在震中上方或西部的仪器要么没有被触发，要么返回了较弱的运动。[2650] D. Eberhart-Phillips和G. McVerry等人，他们发现，在新西兰，在新西兰，在澳大利亚，在新西兰，在澳大利亚，在澳大利亚，在澳大利亚，在澳大利亚，在澳大利亚，在澳大利亚，在澳大利亚，在澳大利亚，在澳大利亚

The problem of estimating response spectra for a heterogeneous lithosphere can be addressed by directly computing attenuation from physical models. In a subduction zone, slab earthquakes will have different attenuation through the mantle wedge than the slab. This article is primarily concerned with very high loss paths through the attenuating mantle underlying the volcanic zone of the North Island, New Zealand, where low Q requires modification of the “standard” New Zealand engineering response spectrum model. A lack of strong-motion data prevents a standard regression analysis for paths from deep slab earthquakes through the highly attenuating mantle zone. Instead, modifications have been derived using a 3D frequency-independent Q model that has been developed for the North Island subduction zone from 2–20 Hz local earthquake t* data. By calibrating the t* crustal results to the standard New Zealand model, amplitudes can be compared between the standard model and the 3D Q model. Additional path-averaged attenuation rate coefficients, CQ, for each source and station pair are determined. This results in simple expressions for CQ as a function of centroid depth, for modifying the standard model. This modification reduces the model spectra by a factor of approximately 2–4 for mantle wedge paths below the volcanic region. This reduction is similar to the observed variation in response spectra, at periods below 0.4 sec, for a 160-km-deep Mw 6.0 earthquake. For shallow earthquakes propagating through the shallow volcanic region, the new model gives results that are similar to a volcanic-path attenuation term derived by regression analysis. Introduction It is important for engineering prediction of earthquake ground motion to model the heterogeneous transmission of shaking from deep slab earthquakes in New Zealand. The North Island lies above the Hikurangi subduction zone (Fig. 1), which has resulted in volcanism and extension in the Taupo volcanic zone (TVZ). Because of along-strike variation in plate motion and plate-interface properties, the TVZ is limited to the central and northern portions of the North Island, although the subducted slab continues to the northern South Island (Reyners, 1998; Eberhart-Phillips et al., 2002; Upton et al., 2002). Seismograms from mantle paths indicate high attenuation compared to slab paths, yet there are not enough strong-motion data to use a standard regression approach to model the attenuation in the mantle. The geometry for deep earthquakes is illustrated schematically in Figure 2a. Below 40-km depth, earthquakes tend to occur only in the subducting slab and hence are located directly below the TVZ. The subducting Pacific plate slab has very low attenuation, the crust of the Australian plate has moderate attenuation, and the mantle wedge between the slab and the overlying crust has very high attenuation. Additionally the crust in the TVZ has very high attenuation. Thus deep earthquakes will be more strongly felt at more distant eastern sites with low-loss slab paths, such as Napier, than at closer sites with paths through the mantle wedge. Deep earthquakes have produced triggered strong-motion records at eastern sites, but for engineering applications the response for deep earthquakes at sites including high-loss paths, such as Taupo and Rotorua in the TVZ and other locations to the northwest, must also be estimated. The effect of high attenuation in the mantle and TVZ can be clearly seen in both felt and instrumental data. An atlas of isoseismal maps for New Zealand (Downes, 1995) contains several deep-earthquake maps for which the major axis of the ellipse that fits the isoseismal pattern is displaced well to the east of the epicenter (Fig. 2b). These maps of felt effects provide insight for earthquakes that lack strongmotion records. Although not formally mapped, the same effect can be seen in records from strong-motion accelerographs. For several deep earthquakes in the general vicinity of the TVZ, the strongest accelerations have been recorded in the vicinity of Napier and Gisborne, whereas instruments at sites above or west of the epicenter have either not been triggered or have returned weaker motions. 2650 D. Eberhart-Phillips and G. McVerry W ha ng an ui B as in N I D ex tra l F au lt Be lt Hi ku ra ng i T ro ug h Alp ine Fa ult PACIFIC PLATE AUSTRALIAN PLATE

Estimating Slab Earthquake Response Spectra from a 3 D Q Model by

从 3 D Q 模型估计板地震响应谱

文献摘要

被引文献

相似文献