二阶Born近似有限频率走时敏感核

有限频率层析成像考虑了非均匀介质中波的散射、衍射、波前愈合等物理性质,使得其对速度异常体的分辨能力远大于射线层析成像.推导和计算有限频率敏感核是进行有限频率层析成像的关键,当前推导有限频率敏感核多借助一阶Born近似,但这只适用于弱散射介质的情况.本文基于二阶Born近似并利用傅里叶变换推导了三维均匀介质情况下有限频率敏感核的解析表达式,并将其推广到非均匀介质中得到了三维非均匀介质中有限频率敏感核.研究表明:当介质中速度扰动小于2%时,基于二阶Born近似的有限频率敏感核与基于一阶Born近似的有限频率敏感核差别很小,可近似认为相同;当介质中速度扰动大于5%时,基于二阶Born近似的有限频率敏感核与基于一阶Born近似的有限频率敏感核有较大不同,表明此时已不能忽略二次散射.     

基于地震成像数据稀疏反演的不连续及非均质地质体检测方法

本文针对油气藏储层预测中的不连续及非均质地质信息识别问题,研究基于地震成像数据的稀疏反演方法.由于该类地质体的地震响应特征为弱信号,因此利用平面波破坏滤波器由地震成像数据中去除强反射同相轴.在此基础上,对剩余的地震数据进行非线性加强滤波,并构建L1稀疏反演模型.为有效求解L1模型,采用非光滑泛函L1范数逼近和拟牛顿求解算法.该方法考虑稀疏先验信息,能够提高反演结果信噪比.缝洞模型测试验证该方法在检测断点、微断裂、散射点等小尺度地质体上的有效性,塔北缝洞型碳酸盐岩储层预测的应用效果进一步证实该方法的实用性.     

基于delta波包叠加的时间域深度偏移

delta波包可由高斯波束经傅里叶逆变换得到,是高斯波束在时空域的表达.它最早出现在合成理论地震图的研究中,本文将其应用于偏移领域.通过delta波包叠加表达时间域格林函数,可将高斯波束偏移由频率域转换到时间域,再结合Rayleigh积分和激励时间成像条件,本文给出了基于delta波包叠加的深度偏移算法.该偏移算法可在时间域直接计算,但因包含褶积运算,成像时将耗费大量的计算时间.针对这一问题,本文提出了将褶积简化为乘积的近似公式.近似后的偏移算法,不仅保留了高斯波束偏移的优点,而且计算效率得到显著提升.文中通过两个数值算例验证了上述结论.     

On planar seismic wavefront modeling for estimating rotational ground motions: Case of 2-D SH line-source

At large hypocentral distances, it is convenient to approximate the curved transient seismic wavefronts as planar to estimate rotational ground motions from the single-station recordings of translational ground motions. In this paper, we investigate whether and when this approximation, referred to as the ‘plane-wave’ approximation, can be considered adequate close to the source. For this, we consider a simplistic source model comprising a two-dimensional, kinematic shear dislocation SH line-source buried in a homogenous, elastic half-space and assume this to be an equivalent representation of a finite-sized fault. The ‘plane-wave’ rotational motion is then synthesized from the exact translational motion solution to the assumed model and compared with the exact rotational motion solution for this model. The comparison between the two sets of rotational amplitudes in frequency domain suggests that the plane-wave approximation may be adequate, when the wavelengths of the seismic waves are much smaller than the source depth. When this is not true, the plane-wave approximation is seen to underestimate the Fourier amplitudes close to the source by several orders, particularly when the fault planes are vertically oriented. A similar comparison in the time domain indicates that a severe underestimation may also occur when the source rise time is longer than the shear-wave arrival time at the epicenter. Significant discrepancies are also observed between the waveforms of the exact and plane-wave rotational motions.     

TTI介质弹性波相速度与偏振特征分析

相速度和偏振方向是研究地震波传播规律和描述介质特性的重要参数,在理论研究和实际应用中有重要作用.本文假定倾斜横向各向同性(TTI)介质对称轴位于观测坐标系XOZ面内,在此观测坐标系下直接推导了TTI介质弹性波相速度和偏振方向的解析表达式,再进一步利用Thomsen弱各向异性理论,推导了弱各向异性近似条件下弹性波相速度以及qP波和qSV波偏振方向表达式.理论分析和数值试例表明,在相速度方面,随着各向异性介质参数改变,qP波和qSH波速度变化较为平缓,qSV波速度变化较为剧烈.弹性波相速度近似式误差均较小,能较好地近似精确相速度.在偏振方向方面,SH波偏振方向只是传播方向和对称轴倾角的函数,而与各向异性参数无关,SH波偏振方向既垂直于传播方向,又垂直于TTI介质对称轴方向.除特定方向外,qP波和qSV波的偏振方向与传播方向均成一定角度,并且随TTI介质对称轴倾角的改变而改变;在精确和近似情况下,qP波和qSV波的偏振方向始终垂直;在精度允许范围内,偏振方向的弱各向异性近似式与理论解析式吻合较好.     

Geopotential Reconstruction,Decomposition, Fast Computation,and Noise Cancellation by Harmonic Wavelets

Harmonic wavelets are introduced within the framework of the Sobolev-like Hilbert space H of potentials with square-integrable restrictions to the Earth's (mean) sphere _R . Basic tool is the construction of H-product kernels in terms of an (outer harmonics) orthonormal basis in H. Scaling function and wavelet are defined by means of so-called H-product kernels. Harmonic wavelets are shown to be building blocks that decorrelate geopotential data. A pyramid scheme enables fast computations. Multiscale signal-to-noise thresholding provides suitable denoising. Multiscale modelling of the Earth's anomalous potential from EGM96-model data is illustrated by use of bandlimited harmonic wavelets, i.e. Shannon and CP-wavelets.     

水平形变场的函数描述及其实现方法

本文利用近代数值逼近理论,给出了描述水平形变场的函数形式及其实现方法。这一方法可削弱通常数据处理中的模型误差。最后给出了算例,验证了本方法的可行性和有效性。     

Comparison of Ray Methods with the Exact Solution in the 1-D Anisotropic “Simplified Twisted Crystal” Model

The exact analytical solution for the plane S-wave, propagating along the axis of spirality in the simple 1-D anisotropic simplified twisted crystal model, is compared with four different approximate ray-theory solutions. The four different ray methods are (a) the coupling ray theory, (b) the coupling ray theory with the quasi-isotropic perturbation of travel times, (c) the anisotropic ray theory, (d) the isotropic ray theory. The comparison is carried out numerically, by evaluating both the exact analytical solution and the analytical solutions of the equations of the four ray methods. The comparison simultaneously demonstrates the limits of applicability of the isotropic and anisotropic ray theories, and the superior accuracy of the coupling ray theory over a broad frequency range. The comparison also shows the possible inaccuracy due to the quasi-isotropic perturbation of travel times in the equations of the coupling ray theory. The coupling ray theory thus should definitely be preferred to the isotropic and anisotropic ray theories, but the quasi-isotropic perturbation of travel times should be avoided. Although the simplified twisted crystal model is designed for testing purposes and has no direct relation to geological structures, the wave-propagation phenomena important in the comparison are similar to those in the models of the geological structures.In additional numerical tests, the exact analytical solution is numerically compared with the finite-difference numerical results, and the analytical solutions of the equations of different ray methods are compared with the corresponding numerical results of 3-D ray-tracing programs developed by the authors of the paper.     

Simulation of summer storms of differing recurrence intervals in a semiarid environment using a kinematic routing scheme

A kinematic flood routing procedure has been devised for a small dendritic headwater gully network on the Western slope of Colorado. the program is spatially-distributed, incorporating lateral inflows from 103 field sites on the network for which channel geometry variables are known. This model, in which a lateral inflow algorithm for the sideslopes between each channel site is convoluted into a Freeze-type (1978) numerical scheme, is fully developed in this paper. Although the field basis of the lateral inflow algorithm has been tested elsewhere (Faulkner, 1990), sensitivity tests were needed for the roughness and hillslope velocity estimates used in the routing procedure. After these successful tests, a suitably precalibrated run of the model was compared with a field-monitored runoff event on the watershed, and results again were encouraging. However, peak attentuation downstream was more pronounced in reality than on the simulation, so the model was also modified by inclusion of allowances for transmission loss. the tendency that the model had displayed for peak size attenuation downstream was considerably enhanced. Using the model, the geomorphic role of the flashfloods which affect the watershed in the summer months is briefly considered by applying the model to existing records of local summer storm rainfall events as a basis for event simulation. These simulations show that downstream attenuation of the flood wave on concave networks in steep semiarid terrain was likely to be a common occurrence, possibly resulting in down-net deposition and differences in geomorphic behaviour between upstream and downstream sites. the discussion is finally broadened to consider the relative importance of ‘common’ as compared to ‘freak’ watershed events in maintaining these differences.