In this paper, we approach automated seismic interference (SI) identification for towed-streamer acquisitions in a new way. We show how to teach a neural network to identify SI based on features the human brain would use for such a task. This includes describing geophysical attributes such as amplitudes and wavefield propagation directions with instantaneous and multishot statistical measures. The statistical measures are then passed to a multilayer perceptrons neural network, first for training and then validation using noise-free synthetic data. Following encouraging results on the synthetic data, we applied the neural network to identify SI on field data from the North Sea and Barents Sea without any further training. The SI in a North Sea case study was very similar to the SI used for training; whereas, the SI in a Barents Sea case study was very different in terms of moveout and amplitude. Even though the neural network was only trained on noise-free synthetics, it successfully detected the SI in both cases.
SUMMARYWe propose a novel acoustic decomposition operator for time slices, loosely based on conventional surface decomposition operators. The proposed operators hold for constant velocity models and require two 2D Fourier Transforms (one forward, one backward) per decomposed time slice per decomposition direction. We then demonstrate the capabilities of our operators on a constant velocity model and the Marmousi model. The decomposition results prove that we can decompose into up-, down-, left-and right-going waves for complex velocity media.
SUMMARYThe feasibility of using an air gun near the sea floor as shear-wave source has been investigated. With an air gun near the sea floor, an evanescent P-wave in the water becomes a propagating S-wave in the sea floor, such that it seems that a pure shear-wave source has been used at the sea floor. This type of wave has been called a P*S wave. An experiment with such a set-up has been carried out at the Valhall field. For that case, modelling shows that shear-wave related event of the type of P*S waves can be expected with such a set-up. Especially at larger offsets, P*S waves can be expected. When analysing the field records and Constant-Velocity Stacks, it is hard to find P*S-wave reflected events. On the other hand, P*S-wave refracted events can be discerned in records. These events can be brought up to stack level and imaged, as shown in this paper.
Up–down wavefield decomposition is effectuated by a scaled addition or subtraction of the pressure and vertical particle velocity, generally on horizontal or vertical surfaces, and works well for data given on such surfaces. The method, however, is not applicable to decomposing a wavefield when it is given at one instance in time, i.e. on snapshots. Such situations occur when a wavefield is modelled with methods like finite‐difference techniques, for the purpose of, for example, reverse time migration, where the entire wavefield is determined per time instance. We present an alternative decomposition method that is exact when working on snapshots of an acoustic wavefield in a homogeneous medium, but can easily be approximated to heterogeneous media, and allows the wavefield to be decomposed in arbitrary directions. Such a directional snapshot wavefield decomposition is achieved by recasting the acoustic system in terms of the time derivative of the pressure and the vertical particle velocity, as opposed to the vertical derivative in up–down decomposition for data given on a horizontal surface. As in up–down decomposition of data given at a horizontal surface, the system can be eigenvalue decomposed and the inverse of the eigenvector matrix decomposes the wavefield snapshot into fields of opposite directions, including up–down decomposition. As the vertical particle velocity can be rotated at will, this allows for decomposition of the wavefield into any spatial direction; even spatially varying directions are possible. We show the power and effectiveness of the method by synthetic examples and models of increasing complexity.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.