Target-oriented least-squares reverse time migration with Marchenko redatuming and double focusing: Field data application

Abstract

Recently, the focus of reflection seismologists has shifted to applications for which a high-resolution image of the subsurface is required. Least-squares reverse time migration (LSRTM) is a common tool used to compute such images. Even so, its high computational costs have led seismologists to use target-oriented LSRTM for imaging only a small target of interest within a larger subsurface block. Redatuming the data to the upper boundary of the target of interest is one approach to target-oriented LSRTM, but many redatuming methods cannot account for multiple scattering within the overburden. We apply a target-oriented LSRTM algorithm that integrates Marchenko redatuming and double focusing to a field data set. This redatuming method accounts for all orders of multiple scattering in the overburden, thus improving the accuracy of target-oriented LSRTM. Moreover, we determine the effectiveness of a double-focusing algorithm in reducing the data size by decreasing spatial and temporal dimensions of the model and the data. The algorithm’s performance is evaluated using field data acquired in the Norwegian Sea. The numerical results indicate that the target-oriented LSRTM algorithm with Marchenko double-focusing can reduce the internal multiple effects and increase the resolution of the resulting image.