华北地区瑞利面波噪声层析成像研究
|
摘要
华北具有我国大陆最古老的克拉通构造,自中生代以来,华北克拉通活化,发生大规模的伸展和裂解,岩石圈减薄,并成为中新生代构造强烈活动的地区。华北地区不但是我国的政治、经济和文化中心,也是地震灾害多发区,历史上曾发生多次破坏性地震。近代发生的唐山、滦县、宁河强震造成了巨大的经济损失和人员伤亡,因此该区一直是我国的地震重点监视防御区。区内断层分布纵横交错,地质构造非常复杂。有鉴于此,许多学者利用多种地球物理方法对华北地区的地壳上地幔速度结构进行了研究,取得了许多重要成果,使人们对华北地区的地质构造背景,地壳上地幔深部构造环境,地震活动性及破坏性地震的成因和机理有了一定程度的了解和认识。
尽管前人就华北地区的地质构造和地壳上地幔速度结构进行了很多研究,但由于地震定位精度、台站分布的不均匀性、台站密度和数据质量等多方面因素的限制,以前的研究也存在一些不足。如:P波速度结构研究较多,而S波速度结构的研究相对较少;早期面波层析成像研究多采用1°×1°以上的网格划分,只能揭示较大尺度的构造框架和主要构造单元的基本性质,分辨率较低;射线平均路径长,短周期面波信号不足,对地壳浅部速度结构的约束较差。
为研究华北克拉通的破坏和岩石圈减薄,深入探讨华北地区强震的孕震背景,获取高分辨率的地壳上地幔三维速度结构,为该区域中长期破坏性地震预测和华北地区的防震减灾提供科学依据,中国地震局地球物理研究所于2006年11月份在华北地区开展了大规模流动地震台阵观测,在该区域布设了250套流动地震观测设备。作为华北地区宽频带流动地震台阵综合研究的一部分,本文对噪声层析成像的理论和方法进行了研究,并将该方法成功应用于华北地区。利用布设在华北地区的190个宽频带、10个甚宽带台的2007年1月份至2008年2月份垂直分量的连续波形数据,采用互相关方法提取出了瑞利面波的格林函数,利用多重滤波和相位匹配滤波方法提取了基阶瑞利面波的群速度频散曲线,得到5630条高质量的频散曲线,继而对研究区进行0.25°×0.25°网格剖分求得了每个网格的群速度纯路径频散,分辨率比以往的研究大有提高。利用遗传算法反演得到了每个网格下方0-50km深度范围的横波速度分布,构建了华北地区的三维S波速度结构。迄今为止,这是中国大陆分辨率最高的群速度层析成像结果。
主要研究内容及结果概述如下:
一、对从背景噪声提取格林函数的理论,频散曲线的测量和反演,以及面波层析成像的理论和方法进行了研究。开发了噪声层析成像的数据处理程序和群速度频散曲线的自动测量程序。测量由地震背景噪声提取出的瑞利面波的频散曲线时,使用多重滤波方法大多能获得较为满意的结果,对于信噪比较差的格林函数,可配合使用相位匹配滤波方法。
二、对提取出的格林函数分析后发现,从地震背景噪声中提取的频散曲线的最大周期与台阵孔径有关,台阵孔径越大,提取的频散曲线频带越宽;通过引入最大可信周期可使测得的频散数据更可靠;发现格林函数的信噪比与观测时间的平方根成正比,使用对称分量可以提高信噪比;噪声源的不均匀分布,导致格林函数的不对称性;使用一年以上的观测数据,可以降低噪声源分布不均匀性对格林函数的影响,使提取出的格林函数更加对称,信噪比更高。
三、对噪声特征进行了研究,发现不同周期段的噪声特征是不一样的。4-10s周期段,互相关函数在零时刻附近出现一组振幅强、震相连贯的波组。由于短周期面波强烈的衰减和吸收效应,提取出的这一周期段的瑞利面波格林函数的信噪比较低。在10-20s周期段,噪声源存在季节性变化,噪声强度与全球海浪高度较一致。在20-50s周期段,瑞利面波格林函数基本对称且不随季节性变化,表明这一周期段的噪声源分布较均匀。在4-20s周期段,尽管互相关函数的正负分支在振幅上存在很明显的不对称性,但是到时却基本一致,说明噪声源分布对互相关函数振幅的影响较大,对到时的影响较小。
四、将研究区域划分为0.25°×0.25°的网格,用面波层析成像方法,反演得到了4s-40s周期的群速度分布图像。层析成像结果较好地揭示了地壳内部、尤其是浅部地壳的横向速度变化。短周期的群速度分布同地表地质结构、地形密切相关;华北盆地及山间的沉积盆地,如延庆-怀来盆地,大同盆地,显示低速异常,而基岩广泛出露的太行山和燕山隆起区,呈现高速异常;我们计算了9s-12s的群速度分布图的速度梯度场,将速度梯度划分为三个区间:＜0.05km/s,0.05-0.15km/s,＞0.15km/s,并同时给出了公元前780年以来发生的M≥6.0级的强震的震中分布,发现约2/3的强震(M≥6.0级)都发生在中等速度梯度带(0.05-0.15km/s)附近。
五、使用光滑约束的遗传算法反演了432个节点的瑞利面波群速度纯路径频散曲线,得到了华北地区0-50km深度的三维S波速度结构。反演结果较好地揭示了华北地区地壳内部的S波速度结构、结晶基底埋深和莫霍面埋深的横向变化。反演结果表明太行山、燕山隆起区结晶基底埋深一般约＜2km,延庆-怀来盆地约5km,大同盆地约3km。华北盆地内部结晶基底厚度一般＞6km,且不同构造单元存在差异,沧县隆起约8km,冀中坳陷和黄骅坳陷约11km。研究区域的地壳厚度横向变化较大,基本形态是西部厚、东部薄,由东南向西北逐渐加深,莫霍面的区域构造方向呈北东或北北东向展布。地壳厚度从东部渤海湾附近的28km逐渐向西北方向变厚,进入山西高原、张家口地区达到44km,其横向变化达16km。在太行山山前断裂附近,莫霍面厚度由东部的36km左右,至西部陡变为40km左右。燕山隆起带地壳厚度由渤海边的32km,向西逐渐加深至约44km。
六、S波速度横向分布图揭示:在0-12km深度范围,太行山山前断裂以东表现为低速异常,而以西表现为高速异常,13km以下,东西两部分的高低速速度异常开始变得模糊,仅在部分区域较明显。据此推测,太行山山前断裂可能仅在部分区域延伸至莫霍面。在0-8km深处,燕山隆起带为大范围的高速异常;在燕山隆起带南侧、与张渤地震带相邻的地方,高低速异常交替分布,可能是受北北东-北东向断裂的切割所致;沿张渤地震带附近,在10km深处有一条较为连续的北西向低速异常,这一低速异常带和燕山高速区的南缘勾画出了张渤断裂带的位置和北部边界。北京-天津-唐山区域在中下地壳(15km-32km)存在显著的低速异常,可能是由地幔热物质侵入造成的。
七、对唐山、滦县、宁河强震区的S波速度结构分析发现,唐山地震和滦县地震发生在高速体内部,且震源下方存在低速体,宁河地震发生在S波低速体与高速体的过渡带。震源区上地幔顶部的S波速度较低,莫霍面在宁河西南和滦县存在局部隆起。根据本文反演得到的高分辨率的S波速度结构,我们认为这三个强震的成因与上地幔的垂直形变、壳幔物质的交换和地壳中的流体有关。幔源岩浆沿壳幔边界的断错侵入,导致上地幔顶部的速度较低。同时岩浆的侵入使下地壳增热,粘度下降,某些矿物发生脱水作用,脱出的水上移并存储于中地壳内。这些流体长时间的存在影响了断层带的结构和组成,进而改变了断层带的应力状态,造成了孕震区的弱化,从而触发地震。
North China (NC) has the oldest craton of China mainland. NC had experienced significant tectonic rejuvenation in the late Mesozoic and Cenozoic, which is evidenced by the widespread lithospheric extension and thinning. NC is not only the political, economical and cultural center of China, but also a region with strong seismic activity. Many strong earthquakes have occurred in this region and cause great damages to our country and people. There are large numbers of faults in this region. The tectonic and geologic setting is extremely complex. So, many researchers studied the velocity structure of crust and upper mantle and the cause of destructive earthquakes using multiple geophysical methods, and obtained many significant results.
Extensive geophysical investigations have been conducted in NC. However, these previous results are mainly concentrated on P wave velocity structure. Many surface wave tomography studies have also been conducted, but most of them are at China mainland scale and the lateral resolution is in the order of 100-200km. Yet few studies of surface waves in NC have been undertaken to date, which is due to the relatively low level of seismicity and the consequent logistic limitation suffered by earthquake-based surface wave tomography. Previous studies almost exclusively used teleseismic earthquakes. It is difficult to obtain reliable short-period (<10s) dispersion measurements from distant earthquakes due to intrinsic attenuation and scattering along the ray paths, and it is these short-period waves that are most useful for constraining the structure of the crust and uppermost mantle. Moreover, the long paths also result in broad lateral sensitivity kernels which limit resolution to hundreds of kilometers. For these reasons, high-resolution surface wave tomography results are scarce.
In order to study the formation and evolution of NC, to obtain crust and upper mantle structure and to verify various proposed mechanisms for the interpretation of the lithospheric process occurred in North China caton, a seismic experiment (North China Seismic Array, NCSA) has been carried out since the winter of 2006. 250 portable stations were deployed in North China, of which 190 are broadband stations, 10 are very broadband stations and 50 are short period stations. In the thesis, we studied the theory and methodology of ambient noise tomography and applied this method to NC sucessfully. Continuous vertical-component seismograms, spanning the period from January 1, 2007 to February 28, 2008 recorded by 190 broadband stations and 10 very broadband stations, are used in this study. We apply the cross correlation technique to the ambient noise data for each station pairs recorded by NCSA. Rayleigh wave group velocity dispersion curves are measured at periods between 4s and 40s by multiple filter technique. We obtain 5630 high quality dispersion curves. Surface wave tomography is conducted to generate group velocity maps with a grid spacing of 0.25°x0.25°. These maps display higher resolution and span to shorter periods than previous surface wave tomography maps. Then genetic algorithum was used to invert pure path dispersion curves. The 3-D shear wave velocity structure from 0 to 50km depth was readily constructed. To the authors' knowledge, the resolution presented here is, so far, the highest one in China mainland.
Follows are the original points of this thesis:
1, Studied the theory about extracting Green Function's from ambient seismic noise, dispersion curve measurement and inversion, and surface wave tomography. Developed ambient noise tomography processing codes and dispersion curve measurement program. We can obtain very good dispersion measurement with multiple filter technique. For Green Functions with lower signal-to-noise ratio (SNR), we can use phase matched filter to get continuous and smooth dipsersion curves.
2, We can obtain more reliable dispersion measurment by introducing maximum cutoff period. The maximum period extracted from ambient seismic noise is related to the aperture of the seismic array. The larger the aperture, the wider the spectrum band is. The SNR of Green Function is propational to the square root of observation time and can be enhanced by using symmtric component. The inhomogeneous distribution of seismic noise gives rise to the asymmetry of Green Function. Using more than one year's data, we can get more symmetric and higher SNR Green Function.
3, We analized the characteristic of ambient seismic noise and found the characteristics are different for different period band. Between 4-10s period band, there is a coherent phase with large amplitude near zero lag time. Because of absorbing and attenuation of seismic waves, the Green Function's SNR is relative lower. In 10-20s period band, the sources of ambient seismic noise has a very clear seasonal variability. The azimuthal distributions of noise share a great similarity with the map of average ocean wave height map obtained by TOPEX-Poseidon. In 20-50s period band, Rayleigh wave Greeen Functions are near symmetrical and have less seasonal variation in both signal strength and directivity, which indicate the distribution of noise is almost homogeneous. In 4-20s period band, the amplitudes of positive and negtive components of Green Functions are obviously asymmetrical, but the arrival times are near identical, which indicate the distribution of noise has much influnce on the amplitude of Green Function, but less influnce on arrival time.
4, Tomographic maps, with a grid spacing of 0.25°×0.25°, are computed between 4s and 40s period band. The maps at short periods reveal an evident lateral heterogeneity in the curst of North-China, quite well in agreement with known geological and tectonic features. The North China Basin is imaged as a broad low velocity area, while the Taihangshan and Yanshan uplifts and Ordos block are imaged as high velocity zones, and the Quaternary intermountain basins show up as small low-velocity anomalies. We computed the velocity gradients of the group velocity maps at 9s, 10s, 11s and 12s, using 0.25°×0.25°cells, and compare them with the space distribution of all the large historic earthquakes (M≥6.0) occurred in this region since 780 BC. We grouped the velocity gradients into three intervals: <0.05 km/s, 0.05-0.15 km/s, >0.15 km/s, we find that most of the large earthquakes (about two-thirds) occurred in regions with moderate velocity gradients.
5, We constructed the 3-D shear wave crustal velocity model in North China by inverting the pure path Rayleigh wave dispersion curves at 432 nodes using genetic algorithum. The inversion results revals the shear wave velocity structure in North China crust, the thickness of sedimentary cover and the lateral variation of Moho interface very well. Our results show that the thickness of sedimentary cover is less than 2km in Taihangshan and Yanshan uplifts, about 5km in Yanqing-Huailai basin and about 3km in Datong basin. The thickness of sedimentary cover in North China basin is more than 6km and has lateral variation in different tectonic units. The thickness of sedimentary cover is about 8km in Cangxian uplift and about 11km in Jizhong and Huanghua depressions. The depth of Moho interface is thicker in westert part than those in the eastern part. The crustal thickness increases from 28km in Bohai Bay to 44km in Zhangjiakou. Near Taihangshan fault belt, the depth of Moho interface increases from 36km in the east to 40km in the west. In Yanshan uplift, the crustal thickness increases from 32km in Bohai Bay to 44km in the west.
6, S-wave velocity maps at different depths show that Taihangshan fault is a boundary of high and low velocity anomaly in 0-12km dpeth. Below 13km, the boundary of high and low veloity anomaly is visible only at some parts of Taihangshan fault.We infer that Taihangshan fault only extends to Moho interface in some regions. From 0 to 8km, Yanshan uplift is mapped as broad high velocity anomaly, while its southern margin is mapped as high and low velocity anomalies alternatively, which may be caused by the NNE-NE trend faults. There is a distinct low velocity belt with NW trend at 10km depth near Zhangjiakou-Bohai seismic zone. This low velocity belt and the southern margin of Yanshan high velocity anomaly draw the outline of Zhangjiakou-Bohai seismic zone and its northern boarderline. There is a well-defined low velocity zone in middle to lower crust (15-25km) in the Beijing-Tianjin-Tangshan region, which may be caused by intrusion of mantle materials.
7, We analized the S-wave velocity structure near Tangshan, Luanxian and Ninghe earthquake region. We find the focal depth of Ninghe earthquake is located in the trasition zones of high and low velocity bodies. The focal depths of Tangshan and Luanxian earthquake are located in high velocity bodies. There is a low velocity zone beneath the focuses of Tangshan and Luanxian earthquake. Moho interface is locally uplifted in Ninghe and Luanxian. The S-wave velocity is lower in the uppermost mantle of these three earthquake regions. We infer that these three earthquakes are mainly caused by vertical deformation of upper mantle and material exchange between crust and upper mantle. The magma intrudes crust along faults near the boundary of crust and upper mantle, which leads to the low velocity anomaly in the uppermost mantle. The magma intrusion heats up the lower crustal material and drops the viscosity. Some minerals are dehydrated. The water move up and store in the middle crust. The exsitence of liquid affects the structure and composition of fault zone, further changes the stress state, weaken the seismotectonic region and triggers the earthquakes.
尽管前人就华北地区的地质构造和地壳上地幔速度结构进行了很多研究,但由于地震定位精度、台站分布的不均匀性、台站密度和数据质量等多方面因素的限制,以前的研究也存在一些不足。如:P波速度结构研究较多,而S波速度结构的研究相对较少;早期面波层析成像研究多采用1°×1°以上的网格划分,只能揭示较大尺度的构造框架和主要构造单元的基本性质,分辨率较低;射线平均路径长,短周期面波信号不足,对地壳浅部速度结构的约束较差。
为研究华北克拉通的破坏和岩石圈减薄,深入探讨华北地区强震的孕震背景,获取高分辨率的地壳上地幔三维速度结构,为该区域中长期破坏性地震预测和华北地区的防震减灾提供科学依据,中国地震局地球物理研究所于2006年11月份在华北地区开展了大规模流动地震台阵观测,在该区域布设了250套流动地震观测设备。作为华北地区宽频带流动地震台阵综合研究的一部分,本文对噪声层析成像的理论和方法进行了研究,并将该方法成功应用于华北地区。利用布设在华北地区的190个宽频带、10个甚宽带台的2007年1月份至2008年2月份垂直分量的连续波形数据,采用互相关方法提取出了瑞利面波的格林函数,利用多重滤波和相位匹配滤波方法提取了基阶瑞利面波的群速度频散曲线,得到5630条高质量的频散曲线,继而对研究区进行0.25°×0.25°网格剖分求得了每个网格的群速度纯路径频散,分辨率比以往的研究大有提高。利用遗传算法反演得到了每个网格下方0-50km深度范围的横波速度分布,构建了华北地区的三维S波速度结构。迄今为止,这是中国大陆分辨率最高的群速度层析成像结果。
主要研究内容及结果概述如下:
一、对从背景噪声提取格林函数的理论,频散曲线的测量和反演,以及面波层析成像的理论和方法进行了研究。开发了噪声层析成像的数据处理程序和群速度频散曲线的自动测量程序。测量由地震背景噪声提取出的瑞利面波的频散曲线时,使用多重滤波方法大多能获得较为满意的结果,对于信噪比较差的格林函数,可配合使用相位匹配滤波方法。
二、对提取出的格林函数分析后发现,从地震背景噪声中提取的频散曲线的最大周期与台阵孔径有关,台阵孔径越大,提取的频散曲线频带越宽;通过引入最大可信周期可使测得的频散数据更可靠;发现格林函数的信噪比与观测时间的平方根成正比,使用对称分量可以提高信噪比;噪声源的不均匀分布,导致格林函数的不对称性;使用一年以上的观测数据,可以降低噪声源分布不均匀性对格林函数的影响,使提取出的格林函数更加对称,信噪比更高。
三、对噪声特征进行了研究,发现不同周期段的噪声特征是不一样的。4-10s周期段,互相关函数在零时刻附近出现一组振幅强、震相连贯的波组。由于短周期面波强烈的衰减和吸收效应,提取出的这一周期段的瑞利面波格林函数的信噪比较低。在10-20s周期段,噪声源存在季节性变化,噪声强度与全球海浪高度较一致。在20-50s周期段,瑞利面波格林函数基本对称且不随季节性变化,表明这一周期段的噪声源分布较均匀。在4-20s周期段,尽管互相关函数的正负分支在振幅上存在很明显的不对称性,但是到时却基本一致,说明噪声源分布对互相关函数振幅的影响较大,对到时的影响较小。
四、将研究区域划分为0.25°×0.25°的网格,用面波层析成像方法,反演得到了4s-40s周期的群速度分布图像。层析成像结果较好地揭示了地壳内部、尤其是浅部地壳的横向速度变化。短周期的群速度分布同地表地质结构、地形密切相关;华北盆地及山间的沉积盆地,如延庆-怀来盆地,大同盆地,显示低速异常,而基岩广泛出露的太行山和燕山隆起区,呈现高速异常;我们计算了9s-12s的群速度分布图的速度梯度场,将速度梯度划分为三个区间:＜0.05km/s,0.05-0.15km/s,＞0.15km/s,并同时给出了公元前780年以来发生的M≥6.0级的强震的震中分布,发现约2/3的强震(M≥6.0级)都发生在中等速度梯度带(0.05-0.15km/s)附近。
五、使用光滑约束的遗传算法反演了432个节点的瑞利面波群速度纯路径频散曲线,得到了华北地区0-50km深度的三维S波速度结构。反演结果较好地揭示了华北地区地壳内部的S波速度结构、结晶基底埋深和莫霍面埋深的横向变化。反演结果表明太行山、燕山隆起区结晶基底埋深一般约＜2km,延庆-怀来盆地约5km,大同盆地约3km。华北盆地内部结晶基底厚度一般＞6km,且不同构造单元存在差异,沧县隆起约8km,冀中坳陷和黄骅坳陷约11km。研究区域的地壳厚度横向变化较大,基本形态是西部厚、东部薄,由东南向西北逐渐加深,莫霍面的区域构造方向呈北东或北北东向展布。地壳厚度从东部渤海湾附近的28km逐渐向西北方向变厚,进入山西高原、张家口地区达到44km,其横向变化达16km。在太行山山前断裂附近,莫霍面厚度由东部的36km左右,至西部陡变为40km左右。燕山隆起带地壳厚度由渤海边的32km,向西逐渐加深至约44km。
六、S波速度横向分布图揭示:在0-12km深度范围,太行山山前断裂以东表现为低速异常,而以西表现为高速异常,13km以下,东西两部分的高低速速度异常开始变得模糊,仅在部分区域较明显。据此推测,太行山山前断裂可能仅在部分区域延伸至莫霍面。在0-8km深处,燕山隆起带为大范围的高速异常;在燕山隆起带南侧、与张渤地震带相邻的地方,高低速异常交替分布,可能是受北北东-北东向断裂的切割所致;沿张渤地震带附近,在10km深处有一条较为连续的北西向低速异常,这一低速异常带和燕山高速区的南缘勾画出了张渤断裂带的位置和北部边界。北京-天津-唐山区域在中下地壳(15km-32km)存在显著的低速异常,可能是由地幔热物质侵入造成的。
七、对唐山、滦县、宁河强震区的S波速度结构分析发现,唐山地震和滦县地震发生在高速体内部,且震源下方存在低速体,宁河地震发生在S波低速体与高速体的过渡带。震源区上地幔顶部的S波速度较低,莫霍面在宁河西南和滦县存在局部隆起。根据本文反演得到的高分辨率的S波速度结构,我们认为这三个强震的成因与上地幔的垂直形变、壳幔物质的交换和地壳中的流体有关。幔源岩浆沿壳幔边界的断错侵入,导致上地幔顶部的速度较低。同时岩浆的侵入使下地壳增热,粘度下降,某些矿物发生脱水作用,脱出的水上移并存储于中地壳内。这些流体长时间的存在影响了断层带的结构和组成,进而改变了断层带的应力状态,造成了孕震区的弱化,从而触发地震。
North China (NC) has the oldest craton of China mainland. NC had experienced significant tectonic rejuvenation in the late Mesozoic and Cenozoic, which is evidenced by the widespread lithospheric extension and thinning. NC is not only the political, economical and cultural center of China, but also a region with strong seismic activity. Many strong earthquakes have occurred in this region and cause great damages to our country and people. There are large numbers of faults in this region. The tectonic and geologic setting is extremely complex. So, many researchers studied the velocity structure of crust and upper mantle and the cause of destructive earthquakes using multiple geophysical methods, and obtained many significant results.
Extensive geophysical investigations have been conducted in NC. However, these previous results are mainly concentrated on P wave velocity structure. Many surface wave tomography studies have also been conducted, but most of them are at China mainland scale and the lateral resolution is in the order of 100-200km. Yet few studies of surface waves in NC have been undertaken to date, which is due to the relatively low level of seismicity and the consequent logistic limitation suffered by earthquake-based surface wave tomography. Previous studies almost exclusively used teleseismic earthquakes. It is difficult to obtain reliable short-period (<10s) dispersion measurements from distant earthquakes due to intrinsic attenuation and scattering along the ray paths, and it is these short-period waves that are most useful for constraining the structure of the crust and uppermost mantle. Moreover, the long paths also result in broad lateral sensitivity kernels which limit resolution to hundreds of kilometers. For these reasons, high-resolution surface wave tomography results are scarce.
In order to study the formation and evolution of NC, to obtain crust and upper mantle structure and to verify various proposed mechanisms for the interpretation of the lithospheric process occurred in North China caton, a seismic experiment (North China Seismic Array, NCSA) has been carried out since the winter of 2006. 250 portable stations were deployed in North China, of which 190 are broadband stations, 10 are very broadband stations and 50 are short period stations. In the thesis, we studied the theory and methodology of ambient noise tomography and applied this method to NC sucessfully. Continuous vertical-component seismograms, spanning the period from January 1, 2007 to February 28, 2008 recorded by 190 broadband stations and 10 very broadband stations, are used in this study. We apply the cross correlation technique to the ambient noise data for each station pairs recorded by NCSA. Rayleigh wave group velocity dispersion curves are measured at periods between 4s and 40s by multiple filter technique. We obtain 5630 high quality dispersion curves. Surface wave tomography is conducted to generate group velocity maps with a grid spacing of 0.25°x0.25°. These maps display higher resolution and span to shorter periods than previous surface wave tomography maps. Then genetic algorithum was used to invert pure path dispersion curves. The 3-D shear wave velocity structure from 0 to 50km depth was readily constructed. To the authors' knowledge, the resolution presented here is, so far, the highest one in China mainland.
Follows are the original points of this thesis:
1, Studied the theory about extracting Green Function's from ambient seismic noise, dispersion curve measurement and inversion, and surface wave tomography. Developed ambient noise tomography processing codes and dispersion curve measurement program. We can obtain very good dispersion measurement with multiple filter technique. For Green Functions with lower signal-to-noise ratio (SNR), we can use phase matched filter to get continuous and smooth dipsersion curves.
2, We can obtain more reliable dispersion measurment by introducing maximum cutoff period. The maximum period extracted from ambient seismic noise is related to the aperture of the seismic array. The larger the aperture, the wider the spectrum band is. The SNR of Green Function is propational to the square root of observation time and can be enhanced by using symmtric component. The inhomogeneous distribution of seismic noise gives rise to the asymmetry of Green Function. Using more than one year's data, we can get more symmetric and higher SNR Green Function.
3, We analized the characteristic of ambient seismic noise and found the characteristics are different for different period band. Between 4-10s period band, there is a coherent phase with large amplitude near zero lag time. Because of absorbing and attenuation of seismic waves, the Green Function's SNR is relative lower. In 10-20s period band, the sources of ambient seismic noise has a very clear seasonal variability. The azimuthal distributions of noise share a great similarity with the map of average ocean wave height map obtained by TOPEX-Poseidon. In 20-50s period band, Rayleigh wave Greeen Functions are near symmetrical and have less seasonal variation in both signal strength and directivity, which indicate the distribution of noise is almost homogeneous. In 4-20s period band, the amplitudes of positive and negtive components of Green Functions are obviously asymmetrical, but the arrival times are near identical, which indicate the distribution of noise has much influnce on the amplitude of Green Function, but less influnce on arrival time.
4, Tomographic maps, with a grid spacing of 0.25°×0.25°, are computed between 4s and 40s period band. The maps at short periods reveal an evident lateral heterogeneity in the curst of North-China, quite well in agreement with known geological and tectonic features. The North China Basin is imaged as a broad low velocity area, while the Taihangshan and Yanshan uplifts and Ordos block are imaged as high velocity zones, and the Quaternary intermountain basins show up as small low-velocity anomalies. We computed the velocity gradients of the group velocity maps at 9s, 10s, 11s and 12s, using 0.25°×0.25°cells, and compare them with the space distribution of all the large historic earthquakes (M≥6.0) occurred in this region since 780 BC. We grouped the velocity gradients into three intervals: <0.05 km/s, 0.05-0.15 km/s, >0.15 km/s, we find that most of the large earthquakes (about two-thirds) occurred in regions with moderate velocity gradients.
5, We constructed the 3-D shear wave crustal velocity model in North China by inverting the pure path Rayleigh wave dispersion curves at 432 nodes using genetic algorithum. The inversion results revals the shear wave velocity structure in North China crust, the thickness of sedimentary cover and the lateral variation of Moho interface very well. Our results show that the thickness of sedimentary cover is less than 2km in Taihangshan and Yanshan uplifts, about 5km in Yanqing-Huailai basin and about 3km in Datong basin. The thickness of sedimentary cover in North China basin is more than 6km and has lateral variation in different tectonic units. The thickness of sedimentary cover is about 8km in Cangxian uplift and about 11km in Jizhong and Huanghua depressions. The depth of Moho interface is thicker in westert part than those in the eastern part. The crustal thickness increases from 28km in Bohai Bay to 44km in Zhangjiakou. Near Taihangshan fault belt, the depth of Moho interface increases from 36km in the east to 40km in the west. In Yanshan uplift, the crustal thickness increases from 32km in Bohai Bay to 44km in the west.
6, S-wave velocity maps at different depths show that Taihangshan fault is a boundary of high and low velocity anomaly in 0-12km dpeth. Below 13km, the boundary of high and low veloity anomaly is visible only at some parts of Taihangshan fault.We infer that Taihangshan fault only extends to Moho interface in some regions. From 0 to 8km, Yanshan uplift is mapped as broad high velocity anomaly, while its southern margin is mapped as high and low velocity anomalies alternatively, which may be caused by the NNE-NE trend faults. There is a distinct low velocity belt with NW trend at 10km depth near Zhangjiakou-Bohai seismic zone. This low velocity belt and the southern margin of Yanshan high velocity anomaly draw the outline of Zhangjiakou-Bohai seismic zone and its northern boarderline. There is a well-defined low velocity zone in middle to lower crust (15-25km) in the Beijing-Tianjin-Tangshan region, which may be caused by intrusion of mantle materials.
7, We analized the S-wave velocity structure near Tangshan, Luanxian and Ninghe earthquake region. We find the focal depth of Ninghe earthquake is located in the trasition zones of high and low velocity bodies. The focal depths of Tangshan and Luanxian earthquake are located in high velocity bodies. There is a low velocity zone beneath the focuses of Tangshan and Luanxian earthquake. Moho interface is locally uplifted in Ninghe and Luanxian. The S-wave velocity is lower in the uppermost mantle of these three earthquake regions. We infer that these three earthquakes are mainly caused by vertical deformation of upper mantle and material exchange between crust and upper mantle. The magma intrudes crust along faults near the boundary of crust and upper mantle, which leads to the low velocity anomaly in the uppermost mantle. The magma intrusion heats up the lower crustal material and drops the viscosity. Some minerals are dehydrated. The water move up and store in the middle crust. The exsitence of liquid affects the structure and composition of fault zone, further changes the stress state, weaken the seismotectonic region and triggers the earthquakes.
引文
陈顒,李丽.2003.地震科学的几个发展趋势.国际地震动态,289(1):1-6.
陈顒.2007.城市地震灾害及其应对:过去、现在和将来.中国民主同盟网站,http://www.mmzy.org.cn.
邓启东,等.2004.中国活动断层分布图(1:400万).北京:科学出版社.
邓晋福,魏文博,等.2007.中国华北地区岩石圈三维结构及演化.北京:地质出版社,31-32.
丁志峰,曾融生.1994.用近震资料反演京津唐地区的地壳三维速度结构.华北地震科学,12(2):14-20.
丁志峰.1999.近震层析成像的理论及应用.[博士论文].北京:中国地震局地球物理研究所.
房立华,吴建平.2009.背景噪声频散曲线测定及其在华北地区的应用.地震学报,已接受.
国家地震局《一九七六年唐山地震》编辑组.1982.一九七六年唐山地震.北京:地震出版社,71-79.
国家地震局《地学断面》编辑委员会.1992.上海奉贤至内蒙古阿拉善左旗地学断面.北京:地震出版社.
国家地震局分析预报中心编.中国东部地震目录.北京:地震出版社,1980.
高文学,马瑾.1993.首都圈地震地质环境与地震灾害.北京:地震出版社.
何正勤,丁志峰,叶太兰,等.2001.中国大陆及其邻域地壳上地幔速度结构的面波层析成像研究.地震学报,23(6):596-603.
何正勤,丁志峰,贾辉,等.2007.用微动中的面波信息探测地壳浅部的速度结构.地球物理学报,50(2):492-498.
黄金莉,赵大鹏.2005.首都圈地区地壳三维P波速度细结构与强震孕育的深部构造环境.科学通报,50(4):348-355.
黄忠贤,胥颐,郝天珧,等.中国东部海域岩石圈结构面波层析成像.地球物理学报,2009,52(3):653-662.
嘉世旭,齐诚,王夫运,等.2005a.首都圈地壳网格化三维结构.地球物理学报,48(6):1316-1324.
嘉世旭,张先康.2005b.华北不同构造块体地壳结构及其对比研究.地球物理学报,48(3):611-620.
金安蜀,刘福田,孙永智.1980.北京地区地壳和上地幔的三维P波速度结构.地球物理学报,23(2):172-182.
陈顒.2007.城市地震灾害及其应对:过去、现在和将来.中国民主同盟网站,http://www.mmzy.org.cn.
邓启东,等.2004.中国活动断层分布图(1:400万).北京:科学出版社.
邓晋福,魏文博,等.2007.中国华北地区岩石圈三维结构及演化.北京:地质出版社,31-32.
丁志峰,曾融生.1994.用近震资料反演京津唐地区的地壳三维速度结构.华北地震科学,12(2):14-20.
丁志峰.1999.近震层析成像的理论及应用.[博士论文].北京:中国地震局地球物理研究所.
房立华,吴建平.2009.背景噪声频散曲线测定及其在华北地区的应用.地震学报,已接受.
国家地震局《一九七六年唐山地震》编辑组.1982.一九七六年唐山地震.北京:地震出版社,71-79.
国家地震局《地学断面》编辑委员会.1992.上海奉贤至内蒙古阿拉善左旗地学断面.北京:地震出版社.
国家地震局分析预报中心编.中国东部地震目录.北京:地震出版社,1980.
高文学,马瑾.1993.首都圈地震地质环境与地震灾害.北京:地震出版社.
何正勤,丁志峰,叶太兰,等.2001.中国大陆及其邻域地壳上地幔速度结构的面波层析成像研究.地震学报,23(6):596-603.
何正勤,丁志峰,贾辉,等.2007.用微动中的面波信息探测地壳浅部的速度结构.地球物理学报,50(2):492-498.
黄金莉,赵大鹏.2005.首都圈地区地壳三维P波速度细结构与强震孕育的深部构造环境.科学通报,50(4):348-355.
黄忠贤,胥颐,郝天珧,等.中国东部海域岩石圈结构面波层析成像.地球物理学报,2009,52(3):653-662.
嘉世旭,齐诚,王夫运,等.2005a.首都圈地壳网格化三维结构.地球物理学报,48(6):1316-1324.
嘉世旭,张先康.2005b.华北不同构造块体地壳结构及其对比研究.地球物理学报,48(3):611-620.
金安蜀,刘福田,孙永智.1980.北京地区地壳和上地幔的三维P波速度结构.地球物理学报,23(2):172-182.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |