绕射波携带了地下小尺度非均质体的高分辨率信息,对岩溶体储层等的识别具有重要意义.能量较弱的绕射波易被反射波场掩盖,导致偏移成像不清、小尺度异常体精细识别与刻画难度大等.传统的绕射波分离方法存在深层绕射波能量损失严重、绕射波与反射波场相交或相切时难以分离的问题.为此,提出一种叠前绕射波分离方法:首先通过共虚震源变换将共炮点道集转换成共虚震源道集,然后在共虚震源道集中进行偏移—反射波去除—反偏移,最后通过共虚震源反变换得到共炮点道集绕射波场.数值模型和实测资料的应用表明,该方法能有效地压制反射波,同时较完整地保留绕射波能量,提升成像剖面上小尺度地质体的识别精度.

Good explosion effect is one of the important conditions for obtaining high-quality seismic data in desert-grassland area. Because of the thickness and velocity change dramatically in the low-velocity zone, the burial depth of the favorable excitation layer is unstable. The velocity and thickness of near-surface stratification can be obtained from near-surface data, but due to the limited distribution of control points and the limitation of surface survey methods, it is difficult to ensure that the location outside the control points is designed in the favorable excitation layer. In this paper, a design method of shot depth under multi-information constraints is proposed. The near-surface velocity model is determined by multi-information constraints, then the excitation layer velocity of surface area is determined according to the parameters of experimental, and then the optimum shot depth of each point is obtained according to the coordinates of the design shot points. This method can better solve the influence of the inadequate of interpolation accuracy and topographic to the design of shot depth, and improve the condition of seismic and the quality of data.

复杂地表区地表起伏大,表层低速带横向变化大,静校正是此类地区地震资料处理的关键和难点.层析静校正方法利用地震初至波射线的走时和路径反演介质速度结构,因不受地表及近地表结构纵横向变化的约束等优势在地震资料处理领域得到广泛应用.笔者针对人机交互和全自动层析静校正技术做了概要描述,并给出了两种技术的实现过程,结合鄂尔多斯北部工区的实际资料,开展初至拾取、反演模型、静校正量、反射波剩余量、叠加剖面、闭合误差和长波长静校正等应用效果的对比研究,论述两者在静校正应用效果上的优缺点和差异,为地震资料处理选择实用高效的静校正技术提供依据.

文章总结了鄂南黄土塬探区的表层地震地质条件和地震资料特点，在对原始地震资料进行全面分析的基础上，准确识别记录中线性干扰、多组面波以及有效反射等波场。以往去除线性干扰采用的方法是f-k倾角滤波，由于该方法存在折叠噪声、混波和挂面条带异常等负效应，因此，资料处理中该方法的应用受到了限制。文章详细分析了线性干扰的线性特性、倾角特性、相干性等波场特点，利用线性噪音的相干性，寻找到去除线性干扰的新方法。通过对比两种方法应用后的叠前道集和叠加剖面的效果分析，以及道集的信噪比、F-K频谱的对比，揭示了新方法去除线性噪音的效果明显好于F-K倾角滤波，同时克服了以往方法的折叠噪声、混波和挂面条等缺点，较好地解决了该区黄土塬的线性噪音问题，为后续的精细处理成像奠定了基础，丰富了研究院的核心处理技术，它以独特的优势广泛应用在鄂南多个工区黄土塬地震资料处理中。

由于鄂尔多斯黄土塬探区的地貌特点和表层复杂的地震地质条件,地震资料存在很严重的静校正问题;层析反演静校正和剩余静校正基本解决了短波长静校正,但仍有部分长波长静校正未解决,因此,识别和解决残留的长波长静校正问题,恢复地层真实形态是地震资料处理的重点内容之一.本文系统地总结了长波长静校正常用方法的三种思路及其适用限制,在分析长波长静校正成因的基础上,详细给出了五种长波长静校正问题的识别方法:水平叠加剖面法、共炮检距剖面法、分偏移距叠加法、地形相关法、交会图法,并对这些方法的应用效果及其适用条件进行分析研究,最后提出采用测井实钻深度约束的井控静校正方法消除剩余长波长静校正量.应用表明,识别方法可以准确地辨别长波长静校正问题,而且井控静校正能有效解决鄂尔多斯黄土塬长波长静校正问题,效果显著,为长波长静校正研究提供了借鉴和启示.

Because of the topography characteristics and the complicated surface seismic geological conditions of the exploration area in Ordos Loess Plateau,the seismic data acquired have strong multiple sets of linear noise.The elimination of the linear noise and the im?provement of SNR of seismic signal is one of the key problems in seismic data processing.In this paper,the authors conducted a compre?hensive analysis and study of all types of linear noise as well as the multiple sets of surface waves. Based on the linear noise's linear characteristics,dip features and coherence properties, the authors used the addition?subtraction method to remove the linear noise. A comparative analysis shows that this means is a high?fidelity method that has a strong capability for suppressing linear noise and impro?ving the quality of data processing,thus showing a significant effect.