Horizon extraction using ordered clustering on a directed and colored graph

2020 ◽  
Vol 8 (1) ◽  
pp. T1-T11
Author(s):  
Zhining Liu ◽  
Chengyun Song ◽  
Kunhong Li ◽  
Bin She ◽  
Xingmiao Yao ◽  
...  
Keyword(s):  
Seismic Data ◽  
Spatial Information ◽  
Spatial Distributions ◽  
Global Level ◽  
Clustering Method ◽  
Colored Graph ◽  
Seismic Image ◽  
Level Information ◽  
Novel Method ◽  
Seismic Images

Extracting horizons from a seismic image has been playing an important role in seismic interpretation. However, how to fully use global-level information contained in the seismic images such as the order of horizon sequences is not well-studied in existing works. To address this issue, we have developed a novel method based on a directed and colored graph, which encodes effective context information for horizon extraction. Following the commonly used framework, which generates horizon patches and then groups them into horizons, we first build a directed and colored graph by representing horizon patches as vertices. In addition, edges in the graph encode the relative spatial positions of horizon patches. This graph explicitly captures the geologic context, which guides the grouping of the horizon patches. Then, we conduct premerging to group horizon patches through matching some predefined subgraph patterns that are designed to capture some special spatial distributions of horizon patches. Finally, we have developed an ordered clustering method to group the rest of the horizon patches into horizons based on the pairwise similarities of horizon patches while preserving geologic reasonability. Experiments on real seismic data indicate that our method can outperform the autotracking approach solely based on the similarity of local waveforms and can correctly pick the horizons even across the fault without any crossing, which demonstrates the effectiveness of exploring the spatial information, i.e., the order of horizon sequences and special spatial distribution of horizon patches.

scholarly journals Appraising structural interpretations using seismic data — Theoretical elements

Geophysics ◽  
2019 ◽  
Vol 84 (2) ◽  
pp. N29-N40
Author(s):  
Modeste Irakarama ◽  
Paul Cupillard ◽  
Guillaume Caumon ◽  
Paul Sava ◽  
Jonathan Edwards
Keyword(s):  
Seismic Data ◽  
Structural Model ◽  
Synthetic Data ◽  
Target Region ◽  
Structural Interpretation ◽  
Vertical Seismic Profiling ◽  
Seismic Image ◽  
Vsp Data ◽  
Phase Shift Data ◽  
Seismic Images

Structural interpretation of seismic images can be highly subjective, especially in complex geologic settings. A single seismic image will often support multiple geologically valid interpretations. However, it is usually difficult to determine which of those interpretations are more likely than others. We have referred to this problem as structural model appraisal. We have developed the use of misfit functions to rank and appraise multiple interpretations of a given seismic image. Given a set of possible interpretations, we compute synthetic data for each structural interpretation, and then we compare these synthetic data against observed seismic data; this allows us to assign a data-misfit value to each structural interpretation. Our aim is to find data-misfit functions that enable a ranking of interpretations. To do so, we formalize the problem of appraising structural interpretations using seismic data and we derive a set of conditions to be satisfied by the data-misfit function for a successful appraisal. We investigate vertical seismic profiling (VSP) and surface seismic configurations. An application of the proposed method to a realistic synthetic model shows promising results for appraising structural interpretations using VSP data, provided that the target region is well-illuminated. However, we find appraising structural interpretations using surface seismic data to be more challenging, mainly due to the difficulty of computing phase-shift data misfits.

Edge-aware filtering with Siamese neural networks

2020 ◽  
Vol 39 (10) ◽  
pp. 711-717
Author(s):  
Mehdi Aharchaou ◽  
Michael Matheney ◽  
Joe Molyneux ◽  
Erik Neumann
Keyword(s):  
Neural Networks ◽  
Seismic Data ◽  
Learning Ability ◽  
Seismic Exploration ◽  
Data Sets ◽  
3D Seismic ◽  
New Class ◽  
Seismic Image ◽  
Siamese Networks ◽  
Seismic Images

Recent demands to reduce turnaround times and expedite investment decisions in seismic exploration have invited new ways to process and interpret seismic data. Among these ways is a more integrated collaboration between seismic processors and geologist interpreters aiming to build preliminary geologic models for early business impact. A key aspect has been quick and streamlined delivery of clean high-fidelity 3D seismic images via postmigration filtering capabilities. We present a machine learning-based example of such a capability built on recent advances in deep learning systems. In particular, we leverage the power of Siamese neural networks, a new class of neural networks that is powerful at learning discriminative features. Our novel adaptation, edge-aware filtering, employs a deep Siamese network that ranks similarity between seismic image patches. Once the network is trained, we capitalize on the learned features and self-similarity property of seismic images to achieve within-image stacking power endowed with edge awareness. The method generalizes well to new data sets due to the few-shot learning ability of Siamese networks. Furthermore, the learning-based framework can be extended to a variety of noise types in 3D seismic data. Using a convolutional architecture, we demonstrate on three field data sets that the learned representations lead to superior filtering performance compared to structure-oriented filtering. We examine both filtering quality and ease of application in our analysis. Then, we discuss the potential of edge-aware filtering as a data conditioning tool for rapid structural interpretation.

Pitfall experiences when interpreting complex structure with low-quality seismic images

2015 ◽  
Vol 3 (1) ◽  
pp. SB29-SB37 ◽  
Author(s):  
Bob A. Hardage
Keyword(s):  
Seismic Data ◽  
Structural Model ◽  
Signal To Noise Ratio ◽  
Complex Structure ◽  
Lessons Learned ◽  
Structural Deformation ◽  
Structural Interpretation ◽  
Seismic Image ◽  
Data Volume ◽  
Seismic Images

Structural interpretation of seismic data presents numerous opportunities for encountering interpretational pitfalls, particularly when a seismic image does not have an appropriate signal-to-noise ratio (S/N), or when a subsurface structure is unexpectedly complex. When both conditions exist — low S/N data and severe structural deformation — interpretation pitfalls are almost guaranteed. We analyzed an interpretation done 20 years ago that had to deal with poor seismic data quality and extreme distortion of strata. The lessons learned still apply today. Two things helped the interpretation team develop a viable structural model of the prospect. First, existing industry-accepted formation tops assigned to regional wells were rejected and new log interpretations were done to detect evidence of repeated sections and overturned strata. Second, the frequency content of the 3D seismic data volume was restricted to only the first octave of its seismic spectrum to create better evidence of fault geometries. A logical and workable structural interpretation resulted when these two action steps were taken. To the knowledge of our interpretation team, neither of these approaches had been attempted in the area at the time of this work (early 1990s). We found two pitfalls that may be encountered by other interpreters. The first pitfall was the hazard of accepting long-standing, industry-accepted definitions of the positions of formation tops on well logs. This nonquestioning acceptance of certain log signatures as indications of targeted formation tops led to a serious misinterpretation in our study. The second pitfall was the prevailing passion by geophysicists to create seismic data volumes that have the widest possible frequency spectrum. This interpretation effort showed that the opposite strategy was better at this site and for our data conditions; i.e., it was better to filter seismic images so that they contained only the lowest octave of frequencies in the seismic spectrum.

Potentially Large Subsurface Gas Hydrate Bodies in the Mexican Ridges, Southwestern Gulf of Mexico

2021 ◽  
pp. 1-29
Author(s):  
Papia Nandi ◽  
Patrick Fulton ◽  
James Dale
Keyword(s):  
Gulf Of Mexico ◽  
Gas Hydrate ◽  
Seismic Data ◽  
Poor Quality ◽  
Reflection Point ◽  
Seismic Reflection Data ◽  
Reflection Data ◽  
High Impedance ◽  
Acoustic Velocities ◽  
Seismic Images

As rising ocean temperatures can destabilize gas hydrate, identifying and characterizing large shallow hydrate bodies is increasingly important in order to understand their hazard potential. In the southwestern Gulf of Mexico, reanalysis of 3D seismic reflection data reveals evidence for the presence of six potentially large gas hydrate bodies located at shallow depths below the seafloor. We originally interpreted these bodies as salt, as they share common visual characteristics on seismic data with shallow allochthonous salt bodies, including high-impedance boundaries and homogenous interiors with very little acoustic reflectivity. However, when seismic images are constructed using acoustic velocities associated with salt, the resulting images were of poor quality containing excessive moveout in common reflection point (CRP) offset image gathers. Further investigation reveals that using lower-valued acoustic velocities results in higher quality images with little or no moveout. We believe that these lower acoustic values are representative of gas hydrate and not of salt. Directly underneath these bodies lies a zone of poor reflectivity, which is both typical and expected under hydrate. Observations of gas in a nearby well, other indicators of hydrate in the vicinity, and regional geologic context, all support the interpretation that these large bodies are composed of hydrate. The total equivalent volume of gas within these bodies is estimated to potentially be as large as 1.5 gigatons or 10.5 TCF, considering uncertainty for estimates of porosity and saturation, comparable to the entire proven natural gas reserves of Trinidad and Tobago in 2019.

Fault Detection by Using Instance Segmentation

2021 ◽  
Author(s):  
Donglin Zhu ◽  
Lei Li ◽  
Rui Guo ◽  
Shifan Zhan
Keyword(s):  
Fault Detection ◽  
Seismic Data ◽  
Semantic Segmentation ◽  
Data Interpretation ◽  
Machine Learning Methods ◽  
Segmentation Algorithms ◽  
Direct Recognition ◽  
First Time ◽  
Seismic Images ◽  
Instance Segmentation

Abstract Fault detection is an important, but time-consuming task in seismic data interpretation. Traditionally, seismic attributes, such as coherency (Marfurt et al., 1998) and curvature (Al-Dossary et al., 2006) are used to detect faults. Recently, machine learning methods, such as convolution neural networks (CNNs) are used to detect faults, by applying various semantic segmentation algorithms to the seismic data (Wu et al., 2019). The most used algorithm is U-Net (Ronneberger et al., 2015), which can accurately and efficiently provide probability maps of faults. However, probabilities of faults generated by semantic segmentation algorithms are not sufficient for direct recognition of fault types and reconstruction of fault surfaces. To address this problem, we propose, for the first time, a workflow to use instance segmentation algorithm to detect different fault lines. Specifically, a modified CNN (LaneNet; Neven et al., 2018) is trained using automatically generated synthetic seismic images and corresponding labels. We then test the trained CNN using both synthetic and field collected seismic data. Results indicate that the proposed workflow is accurate and effective at detecting faults.

scholarly journals Language Modeling for Morphologically Rich Languages: Character-Aware Modeling for Word-Level Prediction

2018 ◽  
Vol 6 ◽  
pp. 451-465 ◽  
Author(s):  
Daniela Gerz ◽  
Ivan Vulić ◽  
Edoardo Ponti ◽  
Jason Naradowsky ◽  
Roi Reichart ◽  
...  
Keyword(s):  
Large Scale ◽  
Language Modeling ◽  
Language Models ◽  
Data Sets ◽  
High Type ◽  
Word Level ◽  
Level Information ◽  
Character Sequences ◽  
Novel Method ◽  
Morphologically Rich Languages

Neural architectures are prominent in the construction of language models (LMs). However, word-level prediction is typically agnostic of subword-level information (characters and character sequences) and operates over a closed vocabulary, consisting of a limited word set. Indeed, while subword-aware models boost performance across a variety of NLP tasks, previous work did not evaluate the ability of these models to assist next-word prediction in language modeling tasks. Such subword-level informed models should be particularly effective for morphologically-rich languages (MRLs) that exhibit high type-to-token ratios. In this work, we present a large-scale LM study on 50 typologically diverse languages covering a wide variety of morphological systems, and offer new LM benchmarks to the community, while considering subword-level information. The main technical contribution of our work is a novel method for injecting subword-level information into semantic word vectors, integrated into the neural language modeling training, to facilitate word-level prediction. We conduct experiments in the LM setting where the number of infrequent words is large, and demonstrate strong perplexity gains across our 50 languages, especially for morphologically-rich languages. Our code and data sets are publicly available.

Some Aspects of Seismic Data Reverse Time Migration for Salt Tectonics Geology of the Dnieper-Donets Basin

2021 ◽  
Author(s):  
Pavlo Kuzmenko ◽  
Viktor Buhrii ◽  
Carlo D'Aguanno ◽  
Viktor Maliar ◽  
Hrigorii Kashuba ◽  
...  
Keyword(s):  
Seismic Data ◽  
Full Range ◽  
Donets Basin ◽  
Salt Tectonics ◽  
Imaging Method ◽  
Reverse Time ◽  
Reverse Time Migration ◽  
Time Migration ◽  
Seismic Image ◽  
Complex Geology

Abstract Processing of the seismic data acquired in areas of complex geology of the Dnieper-Donets basin, characterized by the salt tectonics, requires special attention to the salt dome interpretation. For this purpose, Kirchhoff Depth Imaging and Reverse Time Migration (RTM) were applied and compared. This is the first such experience in the Dnieper-Donets basin. According to international experience, RTM is the most accurate seismic imaging method for steep and vertical geological (acoustic contrast) boundaries. Application of the RTM on 3D WAZ land data is a great challenge in Dnieper-Donets Basin because of the poor quality of the data with a low signal-to-noise ratio and irregular spatial sampling due to seismic acquisition gaps and missing traces. The RTM algorithm requires data, organized to native positions of seismic shots. For KPSDM we used regularized data after 5D interpolation. This affects the result for near salt reflection. The analysis of KPSDM and RTM results for the two areas revealed the same features. RTM seismic data looked more smoothed, but for steeply dipping reflections, lateral continuity of reflections was much improved. The upper part (1000 m) of the RTM has shadow zones caused by low fold. Other differences between Kirchhoff data and RTM are in the spectral content, as the former is characterized by the full range of seismic frequency spectrum. Conversely, beneath the salt, the RTM has reflections with steep dips which are not observed on the KPSDM. It is possible to identify new prospects using the RTM seismic image. Reverse Time Migration of 3D seismic data has shown geologically consistent results and has the potential to identify undiscovered hydrocarbon traps and to improve salt flank delineation in the complex geology of the Dnieper-Donets Basin's salt domes.

Full Wavefield Redatuming: Accurate Velocity Modelling for Imaging Beneath Complex Overburden

2021 ◽  
Author(s):  
Farah Syazana Dzulkefli ◽  
Kefeng Xin ◽  
Ahmad Riza Ghazali ◽  
Guo Qiang ◽  
Tariq Alkhalifah
Keyword(s):  
Wave Propagation ◽  
Seismic Data ◽  
Model Building ◽  
Velocity Model ◽  
Surrounding Rock ◽  
Target Area ◽  
Low Density ◽  
Velocity Model Building ◽  
Seismic Image ◽  
Seismic Wavefield

Abstract Salt is known for having a generally low density and higher velocity compared with the surrounding rock layers which causes the energy to scatter once the seismic wavefield hits the salt body and relatively less energy is transmitted through the salt to the deeper subsurface. As a result, most of imaging approaches are unable to image the base of the salt and the reservoir below the salt. Even the velocity model building such as FWI often fails to illuminate the deeper parts of salt area. In this paper, we show that Full Wavefield Redatuming (FWR) is used to retrieved and enhance the seismic data below the salt area, leading to a better seismic image quality and allowing us to focus on updating the velocity in target area below the salt. However, this redatuming approach requires a good overburden velocity model to retrieved good redatumed data. Thus, by using synthetic SEAM model, our objective is to study on the accuracy of the overburden velocity model required for imaging beneath complex overburden. The results show that the kinematic components of wave propagation are preserved through redatuming even with heavily smoothed overburden velocity model.

Relative time seislet transform

Geophysics ◽  
2020 ◽  
Vol 85 (2) ◽  
pp. V223-V232 ◽  
Author(s):  
Zhicheng Geng ◽  
Xinming Wu ◽  
Sergey Fomel ◽  
Yangkang Chen
Keyword(s):  
Seismic Data ◽  
Computational Cost ◽  
Lifting Scheme ◽  
Real Data ◽  
Data Sets ◽  
Relative Time ◽  
New Approach ◽  
New Formulation ◽  
Wavelet Lifting ◽  
Seismic Images

The seislet transform uses the wavelet-lifting scheme and local slopes to analyze the seismic data. In its definition, the designing of prediction operators specifically for seismic images and data is an important issue. We have developed a new formulation of the seislet transform based on the relative time (RT) attribute. This method uses the RT volume to construct multiscale prediction operators. With the new prediction operators, the seislet transform gets accelerated because distant traces get predicted directly. We apply our method to synthetic and real data to demonstrate that the new approach reduces computational cost and obtains excellent sparse representation on test data sets.

Sign in / Sign up

Export Citation Format

Share Document