The issues of prestack reverse time migration and solutions with Graphic Processing Unit implementation

Abstract

A B S T R A C TPrestack reverse time migration (RTM) is a very useful tool for seismic imaging but has mainly three bottlenecks: highly intensive computation cost, low-frequency band imaging noise and massive memory demand. Traditionally, PC-clusters with thousands of computation nodes are used to perform RTM but it is too expensive for small companies and oilfields. In this article, we use Graphic Processing Unit (GPU) architecture, which is cheaper and faster to implement RTM and we obtain an order of magnitude higher speedup ratio to solve the problem of intensive computation cost. Aiming at the massive memory demand, we adopt the pseudo random boundary condition that sacrifices the computation cost but reduces the memory demand. For rugged topography RTM, it is difficult to deal with the rugged free boundary condition with the finite difference method. We employ a simplified boundary condition that avoids the abundant logical judgment to make the GPU implementation possible and does not induce any sacrifice on efficiency. Besides, we have also done some tests on multi-GPU implementation for wide azimuth geometries using the latest GPU cards and drivers. Finally, we discuss the challenges of anisotropy RTM and GPU solutions. All the jobs stated above are based on GPU and the synthetic data examples will show the efficiency of the algorithm and solutions.

Determining finite difference weights for the acoustic wave equation by a new dispersion‐relationship‐preserving method

“…; Liu et al . ). A review of applications of the FD methods to geophysical imaging is given in Virieux, Calandra and Plessix ().…”

“…The FD methods are efficient in calculation since the methods need less memory and are easy to be implemented and hence are widely used in seismic forward problems and reverse-time migration (McMechan 1989;Etgen and O'Brien 2007;Bansal and Sen 2008;Buur and T. Kuhnel 2008;Zhang, Zhang and Zhang 2011;Guan et al 2011;Yang, Tong and Deng 2012). Recent works show that this method can be greatly accelerated using graphics processing unit clusters (Abdelkhalek et al 2012;Liu et al 2012). A review of applications of the * E-mail address: yfwang@mail.iggcas.ac.cn FD methods to geophysical imaging is given in Virieux, Calandra and Plessix (2011).…”

Closing the performance gap between an iterative frequency-domain solver and an explicit time-domain scheme for 3D migration on parallel architectures

“…The classic RTM algorithms in the time domain are known to be computationally and I/O intensive (Ji et al, 2012;Liu et al, 2012) because the forward-and time-reversed wavefields have to be computed and stored. If the correlation between these fields is carried out during the time-reversed computation of the receiver data, only snapshots of the forward wavefield have to be stored.…”

“…In this case, the propagation effort is doubled because the forward wavefield is first computed, after which the backward propagation and the time-reversed forward wavefield are calculated simultaneously. This method has been implemented on a GPU by Liu et al (2012).…”

Salt-related structure geologic model of Pre-Caspian Basin and analysis on imaging capability

“…Naturally, RTM is the first choice to improve the imaging quality of the salt dome structures in Pre-Caspian Basin. However, the verification of RTM is using the salt dome model of the Gulf of Mexico deep-water oil field (Liu et al 2009;Zhang et al 2007a, b;Liu et al 2012). It is necessary to analyze and compare the imaging capabilities of RTM and WPDM based on the actual regional geologic characteristics, and then give some suggestions to apply the migration method to the real seismic data processing.…”