节点文献
普洱、西双版纳地区壳幔速度结构与强震衰减预测研究
【作者】 毛燕；
【导师】 叶燎原；
【作者基本信息】 昆明理工大学 ， 工程力学， 2011， 博士
【摘要】 进行近场强地面运动估计,是地震工程所面对的一个重要问题,自1933年取得第一条强震记录以来,研究者们就开始深入研究地面运动衰减。在普洱、西双版纳地区架设的27个强震动观测台运行良好,已经成功地获取了一批宝贵的强震观测资料。通过研究该区域的地面运动,可以预测发生大震时在近场引起的地面运动,进而为该区域的地震安全性评价、抗震设防提供参考依据。本文利用远震接收函数反演方法反演了普洱、西双版纳研究区域的S波速度结构：并以此速度结构模型为初始模型,结合面波资料进行反演,得到研究区域的Qβ结构；再基于给出的S波速度结构和Qβ结构模型,利用随机振动理论方法,以2007年发生的MS6.4地震为例,预测了该地震在研究区域产生的强地面运动,将预测结果与实际观测数据、利用回归衰减关系计算得到的结果进行对比研究。通过这几方面的研究工作,主要取得了以下几个方面的成果：(1)收集了普洱、西双版纳地区6个数字台站接收到的2008-2009年发生的震中距在3000-9000km范围内的远震数据,利用径向接收函数,反演得到了普洱、西双版纳地区台站下方的S波速度结构模型。研究结果显示：在靠近北边的景谷和思茅台下方的地壳厚度为36km,其南边的台站下方地壳厚度由北到南逐渐变薄,即孟连、澜沧、勐腊台下方的地壳厚度为大约32km,景洪台地壳厚度最薄,仅为30km。通过平均六个台站下方的速度结构,可得到：普洱、西双版纳地区的下地壳厚度是15km,研究区域的莫霍面大约在32km左右,在莫霍面处S波速度从3.85km/s突变到4.65km/s,在4krn深度处S波速度有突变,也即存在一个速度的不连续界面。(2)收集了研究区域内6个台站于2008年接收到的震中距在1000-3000km的波形数据资料,利用多重滤波方法获取面波群速度频散,再进行相匹配滤波,分离出基阶面波振型,再用维纳滤波方法计算双台间的格林函数,进而测定同一大圆弧上双台间周期为10-75s的面波衰减系数,通过平均求得研究区域的平均衰减系数,结合研究区域的平均速度结构,进行Qβ结构反演。结果显示：该研究区域的Qβ相当低,最高仅在180左右,在浅部地壳,对应低速区,Qβ值不到20,说明区域内地震波衰减快、属于低速区,构造活动相对活跃。(3)基于前面研究得到S波速度结构和Qβ结构模型,以2007年发生的宁洱地震为例,运用随机振动理论方法估算该区域的加速度和速度峰值,将理论预测结果与实际的强震观测值进行比较,可以看出在震源距为200km的范围内,预测值和观测值拟合得比较好。另外,利用回归衰减公式计算得到的加速度峰值与预测结果在近震源处也具有较好的相似性。对比的结果证实了利用随机振动理论进行地面运动预测的可行性,该预测结果可为抗震设防提供参考依据,以减少地震带来的人员伤亡和经济损失。
【Abstract】 Estimation of strong ground motion in near field is an important issue for earthquake engineering. Since the first strong earthquake is recorded in 1933, researchers began to investigate the attenuation of the ground motion. Twenty-seven strong seismological observatories setup in Pu’er ad Xishuangbanna area have successfully acquired some valuable datas for strong earthquakes. Ground motion in near field when a strong earthquake happens can be predicted, through studying the ground motion in the study area, thus providing a referential basis for earthquake-resistant design and earthquake-safety assessment.In this dissertation, the S wave velocity structure beneath the Pu’er and Xishuangbanna area is acquired by a popular teleseismic receiver function inversion. Based on this initial model, a more accurate S wave velocity structure is got using surface wave data. Using this velocity structure, the peak values of acceleration, velocity and displacement can be obtained through a random vibration theory. The predicted values are compared with the observation from a Ninger earthquake that happened in 2007. Also, the result from a regression attenuation formula is compared with the prediction. Following are several specific results that I obtain.(1) Teleseismic data acquired from 2008 to 2009 at six digital stations deployed in the Pu’er and Xishuangbanna area are collected and used to invert for the S wave velocity structure beneath the stations through the receiver function technology. The result reveals that the crustal thickness is 36 km beneath the northern Jinggu and Simao stations, while that beneath the southern gradually thins from the north to the south. The crustal thickness beneath the Menglian, Lancang and Mengla stations is approximately 32 km, with the thinnest place occurring at Jinghong, with a value at only 30 km. Through the even value of the six velocity structure values, we can see that the thickness of lower-crust is 15km in the aera, and the Moho is about 32km deep, where the velocity changes from 3.85km/s to 4.65km/s, and at a depth of 4 km, the S wave velocity changes abruptly, implying for a velocity discontinuity.(2) Seismic data acquired in 2008 having epicentral distances from 1000 to 3000 km are collected from six digital stations in the study area. Using the Multiple Filter Technique to the surface waves, the group velocity was calculated. The fundamental-mode surface wave was calculated through the Phase-Matched Filter method. Then, the interstation Green’s function is also calculated from the two fundamental surface wave signal recordered by the two stations located in the same great circle, and the interstation attenuation coefficients with the period of 10-75s are acquired. The distribution of Qβof the Pu’er and Xishuangbanna area is inversed with the even S velocity and even attenuation coefficients. The value of Qβis pretty low, with the highest Qβat only 180 or so. In the shallow crust, Qβis less than 20, corresponding to a low velocity. Thus the study area, having a fast attenuation and a low velocity, is tectonically active.(3) Using the acquired velocity structure model, the peak values of acceleration, velocity and displacement can be estimated through a random vibration theory. Taking the Ninger earthquake in 2007 as an example, the predicted values are compared with the observation, and the results indicate that they agree with each other very well in the case of the epicentral distances within 200 km. Additionally, the predicted acceleration in near field is similar to that from a regression attenuation formula. It is demonstrated that the random vibration theory is feasible in predicting the peak value of ground motion, thus providing a reference for earthquake-resistant design.