Finite-difference boundary conditions for seismic data processing, imaging and immersive wave experimentation
IPGP - Îlot Cuvier
Séminaire de sismologie, de géosciences marines et de géophysique d'exploration
Johan O. A. Robertsson
It is well known that recomputing modelled wavefields after local model alterations can be done efficiently by employing the method of FD-injection proposed by Robertsson and Chapman (2000) or the method of exact boundary conditions (van Manen et al., 2007) along the edge of the truncated model. In this presentation we discuss how these boundary conditions also have important applications outside modelling of wave propagation. First, we show that by carefully considering the models (medium parameters and boundary conditions) for injection, wavefield injection of multicomponent data can be used to solve a number of long-standing challenges in marine seismic data processing by means of conventional time-space-domain finite-difference propagators. We outline and demonstrate several of these applications including up-down separation of wavefields (deghosting), direct wave removal, source signature estimation, multiple removal and imaging using primaries and multiples. Exact boundary conditions provides a robust and stable time-space solution for multi-dimensional deconvolution without the need to invert large matrices. We show how a combination of FD-injection and exact boundary conditions can be used to directly obtain an estimate of all surface-related multiples that then can be adaptively subtracted from recorded seismic data. Second, at ETH we propose to build a new wave propagation laboratory in which a physical experiment is fully immersed inside a virtual numerical environment by means of using a real-time implementation of exact boundary conditions through transmitting and recording transducer surfaces surrounding a target. Specific applications include time reversal and focussing experimentation in 3D and the study of wave propagation in media where the physics of wave propagation is poorly understood such as the effect of fine scale heterogeneity on broadband propagating waves.