Vemund Stenbekk Thorkildsen

• Huff, Owen Rohwer; Evensen, Andreas Kjelsrud; Lie, Jan Erik; Bugge, Aina Juell; Sablon, Ronan & Thorkildsen, Vemund Stenbekk [Show all 7 contributors for this article] (2025). Automatic conditioning of marine seismic angle stack data with a convolutional neural network. Interpretation. ISSN 2324-8858. 13(1), p. T177–T202. doi: 10.1190/INT-2024-0034.1. Full text in Research Archive Show summary

  Misalignment of reflections is a common problem in migrated prestack seismic data, which occurs due to moveout correction with incomplete knowledge of the velocity model. This issue persists in partial angle stack data often used for amplitude-variation-with offset (AVO) analyses and is traditionally treated with residual moveout (RMO) and trim static corrections. However, these methods require time-consuming velocity selection and parameter tuning and commonly lead to spurious alignment of reflections with noise. Therefore, we train a convolutional neural network (CNN) on synthetic examples to automatically perform conditioning and alignment of angle stack data. First, 2D synthetic common-depth point gathers are created using a convolutional modeling method and then made into input-target pairs, where the inputs are poorly moveout corrected (using inaccurate normal-moveout velocities) and the targets are accurately moveout corrected. These input-target examples are split into angle stacks for the near, mid, and far offsets and then used to train a CNN to transform the misaligned angle stacks into their more-aligned counterparts. In testing, the trained network increases the alignment of unseen synthetic data, improving the correlation with the target far-offset trace from 0.76 to 0.85 on average. The network is applied on two field examples from offshore Norway with class 3 and class 2P AVO responses, respectively. In both cases, the angle stack data predicted by the network indicate improved alignment, greater resolution in the far-offset stacked section, and increased correspondence to a well-tie synthetic seismogram, all while retaining the AVO responses. The CNN predictions are compared with angle stack data after traditional conditioning (RMO and trim static corrections), with the results being of similar or greater quality. This method enables high-quality conditioning of angle stack data at a fraction of the time required for conventional methods and is easily adaptable to different data sets.

• Huff, Owen Rohwer; Thorkildsen, Vemund Stenbekk; Greiner, Thomas Andre Larsen; Lie, Jan Erik; Evensen, Andreas & Bugge, Aina Juell [Show all 7 contributors for this article] (2024). Near offset reconstruction for marine seismic data using a convolutional neural network. Geophysical Prospecting. ISSN 0016-8025. 72(6), p. 2164–2185. doi: 10.1111/1365-2478.13505. Full text in Research Archive Show summary

  Marine seismic data is often missing near offset information due to separation between the source and receiver cables. To solve this problem, a convolutional neural network is trained on synthetic seismic data to reconstruct the near offset gap. The synthetic data is created using a two-dimensional finite difference method within a heterogeneous velocity model. These synthetics are generated with a source-over-receiver acquisition geometry so that they contain complete near offset data. The convolutional neural network is then trained on input-target synthetic pairs where the inputs are common midpoint gathers with the near offset section removed, and the targets are the same gathers with the near offset section retained. Following training, the robustness of the method is investigated with regards to common midpoint data sorting, normal moveout correction and changes in the velocity model. It is found that training on common midpoint-sorted data results in 2.8 times lower error than training on shot gathers, that normal moveout correction of the training data makes no significant difference in error levels, and that the model can reconstruct realistic near offsets on synthetic data generated 10 km away within the heterogeneous velocity model. In field data testing, first a dataset with source-over-cable acquisition geometry from the Barents Sea is used to compare the reconstructed wavefields to ground truth values. Although the reconstructed amplitudes require minor scaling to match the true values, predictions on this dataset yield 2.5 times lower near offset reconstruction error compared to a simple Radon transform interpolation method. Furthermore, amplitude versus offset gradient and intercept sections from the Barents Sea dataset are estimated with half the error when including the convolutional neural network-predicted near offset data, compared to only using the conventionally-acquirable portion of the data (beyond 112.5 m of offset). In a secondary field data test, a conventional northern North Sea dataset is used to demonstrate how the method may be applied in practice. Here, the convolutional neural network generates more realistic predictions than the Radon method, and the gradient and intercept sections calculated using the convolutional neural network-predicted traces have higher signal-to-noise ratios than the sections calculated using only the original data. The combination of high-quality synthetic training data and interpolation in the common midpoint domain enables near offset reconstruction at significant depth (1 s of two-way traveltime or more), which is demonstrated in both synthetic and field examples.

• Thorkildsen, Vemund Stenbekk & Gelius, Leiv-J. (2023). Electromagnetic resolution - a CSEM study based on the Wisting oil field. Geophysical Journal International. ISSN 0956-540X. 233(3), p. 2124–2141. doi: 10.1093/gji/ggad046. Full text in Research Archive Show summary

  SUMMARY We consider marine controlled source electromagnetic (CSEM) data and demonstrate that a typical CSEM survey is associated with significant data redundancy. Thus, it should be possible to obtain a high-quality inversion result by using only a subset of the original data. Moreover, in survey design, effort should be made to optimize the placement of the receivers. This study therefore investigates the challenges of data decimation and survey design in the case of repeated surveys by use of the data resolution matrix and model resolution matrix. A framework of analysis has been developed to efficiently use these quantities. The feasibility of the proposed approach is demonstrated using both synthetic data computed from an interpreted model of the Wisting oil field in the Barents Sea, as well as actual field data from the same oil field.

• Thorkildsen, Vemund Stenbekk & Gelius, Leiv-J. (2023). Resolution enhancement of 2D controlled-source electromagnetic images by use of point-spread function inversion. Frontiers in Earth Science. 11. doi: 10.3389/feart.2023.1101228. Full text in Research Archive Show summary

  The marine controlled-source electromagnetic technique is employed both in large-scale geophysical applications as well as within the exploration of hydrocarbons and gas hydrates. Because of the diffusive character of the EM field, only very low frequencies are used, leading to inversion results with low resolution. In this paper, we calculated the resolution matrix associated with the inversion and derived the corresponding point-spread functions. The PSFs provided information about how much the actual inversion was blurred. Using a space-varying deconvolution can thus further improve the inversion result. The actual deblurring was carried out using the nonnegative flexible conjugate gradient least-squares (NN-FCGLS) algorithm, which is a fast iterative restoration technique. To attain completeness, we also introduced the results obtained using a blind deconvolution algorithm based on the maximum likelihood estimation with unknown PSFs. The potential of the proposed approach has been demonstrated using both complex synthetic data and field data acquired at the Wisting oil field in the Barents Sea. In both cases, the resolution of the final inversion result was improved and showed greater agreement with the known target area.

• Thorkildsen, Vemund Stenbekk; Gelius, Leiv-J. & Robinson, Enders A. (2021). Revisiting holistic migration. The Leading Edge. ISSN 1070-485X. 40(10), p. 768–777. doi: 10.1190/tle40100768.1. Full text in Research Archive

View all works in NVA

• Huff, Owen Rohwer; Evensen, Andreas Kjelsrud; Lie, Jan Erik; Bugge, Aina Juell; Sablon, Ronan & Thorkildsen, Vemund Stenbekk (2024). Efficient Angle Stack Conditioning Using a Convolutional Neural Network. 85th EAGE Annual Conference & Exhibition. 2024, p. 1–5. doi: 10.3997/2214-4609.202410580. Full text in Research Archive
• Huff, Owen Rohwer; Thorkildsen, Vemund Stenbekk; Greiner, Thomas Andre Larsen; Lie, Jan Erik; Bugge, Aina Juell & Evensen, Andreas Kjelsrud (2023). Convolutional Neural Network-based Restoration of Near Offsets in Marine Seismic Data. 84th EAGE Annual Conference & Exhibition. 1, p. 1–5. doi: 10.3997/2214-4609.202310843. Full text in Research Archive
• Thorkildsen, Vemund Stenbekk (2019). Diffraction separation by diffraction stacking. Full text in Research Archive

View all works in NVA