复杂山地随机介质GMM-ULTI法射线追踪

对复杂山地介质的非均质性以及介质中地震波运动学特征进行深入研究,对于提高复杂山地区域地震勘探的效果有着重要的理论意义和实际价值.为了研究复杂山地非均质性和该介质中地震波的一些运动特性,提出了一种复杂山地随机介质的建模方法和一种新的射线追踪算法.与常规算法相比,复杂山地随机介质的生成方法采用更贴近实际介质特点的梯度介质作为背景介质,并在模型生成过程中加入地形修正步骤;新提出的GMM-ULTI射线追踪算法,充分融合群推进法、迎风思想、走时插值法的优势,采用先计算走时后追踪射线路径的两步策略完成射线追踪.算法分析与计算实例表明:复杂山地随机介质的生成方法能灵活、精细且更贴近实际地刻画复杂山地介质的非均质特点;新射线追踪算法兼顾精度和效率、能无条件稳定且灵活地适应复杂山地随机介质的特点;同时基于对几个模型试算结果的分析也得出了复杂山地随机介质中的地震波的一些传播规律.

Ray tracing using GMM-ULTI method in the random media with complex topography

Studying on the complex heterogeneous media and the kinematics characteristics of seismic wave in the model with complex topography are very important in theoretical significance and practical application to improve the effect of the seismic exploration in complex mountainous region. To research the heterogeneity of complex mountainous media and some kinematical characteristics of seismic wave in this region, we propose a new method for generating the complex mountainous random media and a new scheme for ray tracing. Firstly, Comparing with the conventional method, the terrain correction is introduced into conventional random media and the background of random media is changed from homogeneous media into continuous media. Then, combining the advantages of some general ray tracing algorithms include fast marching method, upwind algorithm, and linear traveltime interpolation method, we present a new ray tracing method called GMM-ULTI method. The new method is implemented by using the following two steps: forward computing the traveltime and backward tracing the ray path. Numerical evaluations and examples show that the new generating scheme of random model with complex topography is flexible and has the ability to describe the heterogeneity of complex topography more carefully and practically. The new ray tracing method is accurate, efficient, unconditionally stable, and flexible in the complex mountainous random media with topography. Finally, we obtain some propagation laws of seismic wave in complex mountainous region by analyzing the computational results in some complex mountainous random models.