Tau-p变换若干问题的讨论及其在反演声纳浮标资料中的应用

The discussion on some problems of τ-p transform and its application to the inversion of sonar buoy data,

在时空域进行τ-p变换,由于接收空间范围有限,以及反射波时距曲线在远、近道被截断,必将引起假频和端点效应的干扰。采用相似系数滤波方法可以压制这种假频和端点效应的干扰。为了增加相似系数的平稳性在作相似滤波之前,需对相似系数进行低通滤波,以消除随机干扰。在τ-p变换中,参数选择尤为重要。参数τ(和△τ)可根据有效波出现的范围、不出现假频和计算速度等因素综合考虑。参数p(和△p)可根据最大和最小水平视速度、假频和计算速度等因素考虑。文中还讨论了有偏移距和倾斜界面模型在τ-p变换时出现的问题及τ-p变换在道内插和压制面波干扰中的应用。在τ-p域内应用τ-SUM法反演声纳浮标资料,可获得很好的效果。

Tau-p transform in time-space domain must bring noises due to aliasing and end-point effect not only because of the limited receiving aperture but also because of the truncations at far-trace and near-trace segments of reflection time-distance curve. Similarity coefficient filtering may remove the noises due to aliasing and end point effect. In order to improve the stability of similarity coefficients, low pass filtering of similarity coefficients is needed to remove random noise before similarity filtering is made. The choice of parameters is particularly important in τ-ρ transform. τ and △τ may be chosen by taking a comprehensive consideration of no aliasing, computation speed, the extent in which effective waves appear, and so on. ρ and △ρ can be determined by overall considering the factors such as maximum and minimum horizontal apparent velocities, aliasing, computation speed, etc. The problem is discussed which arises in τ-ρ transform of synthetic seismic data obtained with the model that has dipping reflectors and in-line offset. The paper discribes the applications of τ-ρ transform to trace interpolation and surface wave suppression. τ-SUM method in τ-ρ domain can bring good result in the inversion of sonar buoy data.