Visibility analysis for target-oriented reverse time migration and optimizing acquisition parameters

Abstract

To obtain the best possible image quality of complex geologic structures, prestack depth migration has evolved over the past several years from using one-way wave-equation migration (WEM) to using reverse time migration (RTM) in production seismic processing (Etgen et al., 2009). RTM extrapolates both source and receiver wavefields in time (Baysal et al., 1983; McMechan, 1983), rather than along a spatial (usually vertical) axis and is thus able to handle waves that propagate long distances horizontally, turning waves that propagate beyond 90°, and some special classes of multiple reflections, such as duplex waves and prism waves. The huge computational burden of RTM for high-frequency wavefields, especially in tilted anisotropic (TTI) media often prevents its use as an iterative seismic imaging tool. While certainly some imaging projects run multiple iterations of RTM, these projects often require long time frames and substantial expense, something justified only by the highest-profile projects. A wide class of seismic imaging problems being faced today are mostly “target-oriented” in their nature; they usually involve imaging particular details of salt features or other velocity complexities, or they involve creating the best possible image of the reservoir itself. In these cases, it is not necessary to migrate full-volume data to better reconstruct the image around the area of interest for each iteration because most of these data have no physical contributions to the target events and may even introduce noise and, thus, degrade the image quality of the target events. Rather, we should migrate those data having physical contributions to the target events. The visibility analysis we describe provides a solution to quantitatively select the data that need to be migrated for reconstructing the target events (Jin and Xu, 2009; Jin et al., 2010).