The form of the reflection operator and extended image gathers for a dipping plane reflector in two space dimensions

The reflection operator for a simple flat‐lying interface can be thought of as the set of all its plane‐wave reflection coefficients or as the set of virtual surveys with sources and receivers along the interface. When there is dip, however, it is necessary to include the varying effects of propagation between the virtual‐survey level and the interface. Hence, step one in this paper is to derive the reflection operator for a dipping plane interface as observed at a datum level some distance away. The key assumption is that the aperture at the datum level is sufficient to characterize the reflector properties around a particular point. This translates into an assumption that the dip is moderate, though no explicit small‐angle approximation is required. The second step is to find the apparent reflection operator that would relate data that have been extrapolated from the datum towards and possibly beyond the reflector using an assumed migration velocity. This apparent reflection operator is closely related to extended common‐image gathers. The apparent reflection operator may be analysed asymptotically in terms of rays and other signals, shedding light on the structure of extended image gathers. In keeping with the virtual‐survey idea, the results are considered in a subsurface space‐time or slowness‐time domain at various extrapolation levels around the interface. An important distinction is drawn between using subsurface midpoint‐offset coordinates and the wavefield coordinates of the incident and reflected waves. The latter reveal more clearly the effects of dip, because they lead to a more asymmetric apparent reflection operator. Properties such as an up‐dip shift of a traveltime minimum and its associated curvature theoretically provide information about the reflector location and dip and the migration‐velocity error. The space‐time form of the reflection operator can be highly intricate around the offset‐time origin and it was described for a simple flat interface in a background paper. To avoid a layer of mathematics, the reflection‐operator formulas presented here are in the intermediate space‐frequency domain. They are analysed by considering their stationary‐phase and branch‐point high‐frequency contributions. There is no Born‐like assumption of weak reflector contrast and so wide‐angle, total reflection and head‐wave effects are included. Snell’s law is an explicit part of the theory. It is hoped that the work will therefore be a step towards the goal of unifying amplitude‐versus‐offset, imaging and waveform inversion.     

叠前地震资料提高分辨率处理方法及应用

对于隐蔽性油气藏和复杂构造油气藏,地震反射资料难以满足精度要求,需要开展地震资料提高分辨率处理.为此,使用反Q滤波、广义S变换、谱整形技术及谱白化技术等方法处理了涠西南地区叠前地震资料,并且分析了四种提高分辨率方法的应用效果.结果表明,四种方法都能提高主频、拓宽频带,其中反Q滤波的效果最好,广义S变换效果次之.反Q滤波结果中波组的低频和中高频能量分配合理,并且在中高频段反Q滤波结果的信噪比明显高于其它三种方法.此外,使用不同Q值对反Q滤波结果影响比较大.随着Q值增加,主频和带宽都稍有降低的趋势,但是信噪比提高了.当Q值大于200.0时信噪比改善的空间已经很小,这时再提高Q值,换来的只是主频降低和带宽缩小.     

基于Zoeppritz偏导方程精确解的地层密度多角度反演

基于Zoeppritz方程对介质密度偏导数所建立的偏导方程的精确解,构造了多角度反演地层介质密度的反演方程,在偏导数求解过程中考虑了介质密度对波速度的影响因素,并由此实现了利用反射系数梯度精确解计算地层密度的多角度联合反演.通过数值算例考察了计算方法,结果显示:反演方法对层状地层模型不论反射波是否存在相干现象均获得了较好的反演结果,反演迭代10次后计算结果的最大相对误差能够收敛到1%之内;随着反演角度的增加地层介质密度反演的精度逐步提高,反演具有自动校正能力,有快的计算速度.本方法克服了传统AVO(Amplitude Versus Offset)基于Zoeppritz方程近似所遇到的困难,不受反演角度大小及反射界面对波反射强弱的限制,为地层介质密度的多角度包括大角度反演提供了一种新的快速有效的计算方法.     

无源自适应磁流变阻尼器设计与研究

针对防震、航空航天、野战军械等领域对能源的限制,以及车辆减振中对阻尼器拉压行程不同阻尼力大小的要求,设计制作了一个无源自适应磁流变阻尼器。该阻尼器采用磁致伸缩逆效应及磁流变液的特性,把负载力的变化转化为磁流变液间隙处磁场的变化,解决了常规的磁流变阻尼器需要外配能源装置及控制电路的问题。对该阻尼器进行了建模,并进行了阻尼特性试验,结果表明:试验数据与模型符合得较好,阻尼力大小和负载有关,拉力越大,阻尼力越大,压力越大,阻尼力越小,具有负载自适应的特点,验证了设计的可行性。     

预测多次波的逆散射级数方法与SRME方法及比较

逆散射级数方法和SRME方法是典型的两类基于波动方程的无需地下结构信息的自由表面多次波预测方法,本文详细讨论了这两类方法的基本思想、方法原理和实现思路,并对2维逆散射级数方法进行了降维简化,推导给出了1.5维逆散射级数衰减自由表面多次波的方法,减少了计算量,并降低了对3维规则观测系统的要求.利用经典SMAART模型数据测试比较了2维逆散射级数方法、1.5维逆散射级数方法及2维SRME方法预测自由表面多次波的效果及优缺点,比较了三者在方法实现、并行计算效率及对数据要求上的异同,并结合分析比较结果,给出了实际应用中方法选择的建议.     

利用基于子波估计的频谱校正方法提高Q值估计精度(英文)

用Q值刻画的地震衰减在地震信号处理和解释中具有很广泛的应用。利用反射地震资料进行Q值估计需要解决地震子波和反射系数序列耦合的问题。从反射地震资料中去除反射系数序列的影响,这个过程称为频谱校正。本文提出了一种基于子波估计的求取Q值的方法,进而设计了一个反Q滤波器。该方法利用反射地震资料的高阶统计量进行子波估计,并利用所估计子波实现频谱校正。我们利用合成数据实验给出了质心频移法与频谱比法这两种常用的Q值估计方法在不同参数设置下的性能。人工合成数据和实际数据处理表明,利用本文提出的方法进行频谱校正后,可以得到可靠的Q值估计。经过反Q滤波,地震数据的高频部分得到了有效地恢复。     

重力异常图中的边界增强和提取技术

对比了重磁勘探领域中4种典型的边界增强方法：Tilt导数、Tilt导数的水平梯度、Theta角和归一化的标准偏差,它们都能放大与边界有关的高频异常。增强后异常幅值极大值与地质体边界对应的情况可用近似边界法自动提取,而异常幅值拐点与边界对应的情况可以引入Canny边界检测算子来提取。理论模型试验表明：用Canny算子提取Tilt导数图边界与用近似边界检测法提取其它边界增强图的边界相比,前者提取的边界位置与真实位置更吻合,虚假边界更少。应用该方法对梅山铁矿的局部重力异常进行了处理,勾画出铁矿的大致范围。     

岩石有限应变测量反向轮法的计算机CSD软件设计

用计算机实现反向轮法测岩石有限应变的过程, 可提高有限应变测量的准确性和效率.开发出了用Visual Basic 5.0编写的CSD软件, 利用岩石薄片图像或者矿物颗粒轮廓图测出岩石有限应变的大小及应变椭圆长轴方向, 同时测出矿物颗粒分布的优选方位.其操作相当简便, 首先统计出图像中0°~180°各方向上矿物颗粒的边界数目; 然后在方位-边界坐标系中投点, 并利用最小二乘法进行数据点的多项式曲线拟合; 最后求出曲线的极值点坐标, 并根据坐标绘制相应的应变椭圆.软件运行中, 统计出的边界数据和拟合曲线以及应变椭圆图形都是可视的, 并能进行相应的保存.      

Research on multiple elimination methods in inverse data space based on randomized singular value decomposition

Based on surfaced-related multiple elimination (SRME),this research has derived the methods on multiples elimination in the inverse data space.Inverse data processing means moving seismic data from forward data space (FDS) to inverse data space (IDS).The surface-related multiples and primaries can then be separated in the IDS,since surface-related multiples will form a focus region in the IDS.Muting the multiples energy can achieve the purpose of multiples elimination and avoid the damage to primaries energy during the process of adaptive subtraction.Randomized singular value decomposition (RSVD) is used to enhance calculation speed and improve the accuracy in the conversion of FDS to IDS.The synthetic shot record of the salt dome model shows that the relationship between primaries and multiples is simple and clear,and RSVD can easily eliminate multiples and save primaries energy.Compared with conventional multiples elimination methods and ordinary methods of multiples elimination in the inverse data space,this technique has an advantage of high calculation speed and reliable outcomes.     

地表水污染源识别方法研究进展

地表水污染源的准确识别对于突发水污染事件中制定补救策略、确定责任方具有重要的指导意义。从污染源识别的基本问题出发,综述了地表水污染源识别的两类方法,一类是数学模拟法,包括直接求解法和间接求解法,直接求解法又包括解析法和正则化方法,间接求解法又包括模拟优化法、概率统计法及耦合算法;另一类是足迹分析法。在此基础上,系统地讨论了污染源识别的不确定性,包括污染源信息、观测数据、地表水模型、水动力条件、污染物性质5个方面。并指出地表水污染源识别的不确定性和时效性是制约其应用价值的关键因素,需要在方法耦合、识别效率提高、不确定性分析及污染物性质等方面做进一步的研究。