Imposing interpretational constraints on a seismic interpretation CNN

Geophysics ◽  
2021 ◽  
pp. 1-36
Author(s):  
Haibin Di ◽  
Cen Li ◽  
Stewart Smith ◽  
Zhun Li ◽  
Aria Abubakar
Keyword(s):  
Seismic Data ◽  
Recent Progress ◽  
Facies Analysis ◽  
Three Dimensional ◽  
Seismic Interpretation ◽  
3D Seismic ◽  
Geologic Time ◽  
Seismic Amplitude ◽  
Subsurface Geology ◽  
3D Seismic Data

With the expanding size of three-dimensional (3D) seismic data, manual seismic interpretation becomes time consuming and labor intensive. For automating this process, the recent progress in machine learning, particularly the convolutional neural networks (CNNs), has been introduced into the seismic community and successfully implemented for interpreting seismic structural and stratigraphic features. In principle, such automation aims at mimicking the intelligence of experienced seismic interpreters to annotate subsurface geology both accurately and efficiently. However, most of the implementations and applications are relatively simple in their CNN architectures, which primary rely on the seismic amplitude but undesirably fail to fully use the pre-known geologic knowledge and/or solid interpretational rules of an experienced interpreter who works on the same task. A general applicable framework is proposed for integrating a seismic interpretation CNN with such commonly-used knowledge and rules as constraints. Three example use cases, including relative geologic time-guided facies analysis, layer-customized fault detection, and fault-oriented stratigraphy mapping, are provided for both illustrating how one or more constraints can be technically imposed and demonstrating what added values such a constrained CNN can bring. It is concluded that the imposition of interpretational constraints is capable of improving CNN-assisted seismic interpretation and better assisting the tasks of subsurface mapping and modeling.

Using relative geologic time to constrain CNN-based seismic interpretation and property estimation

Geophysics ◽  
2021 ◽  
pp. 1-44
Author(s):  
Aria Abubakar ◽  
Haibin Di ◽  
Zhun Li
Keyword(s):  
Large Scale ◽  
Three Dimensional ◽  
Seismic Interpretation ◽  
Seismic Survey ◽  
Geologic Time ◽  
Seismic Amplitude ◽  
Property Estimation ◽  
Constrained Learning ◽  
Seismic Amplitudes ◽  
Multiplicative Regularization

Three-dimensional seismic interpretation and property estimation is essential to subsurface mapping and characterization, in which machine learning, particularly supervised convolutional neural network (CNN) has been extensively implemented for improved efficiency and accuracy in the past years. In most seismic applications, however, the amount of available expert annotations is often limited, which raises the risk of overfitting a CNN particularly when only seismic amplitudes are used for learning. In such a case, the trained CNN would have poor generalization capability, causing the interpretation and property results of obvious artifacts, limited lateral consistency and thus restricted application to following interpretation/modeling procedures. This study proposes addressing such an issue by using relative geologic time (RGT), which explicitly preserves the large-scale continuity of seismic patterns, to constrain a seismic interpretation and/or property estimation CNN. Such constrained learning is enforced in twofold: (1) from the perspective of input, the RGT is used as an additional feature channel besides seismic amplitude; and more innovatively (2) the CNN has two output branches, with one for matching the target interpretation or properties and the other for reconstructing the RGT. In addition is the use of multiplicative regularization to facilitate the simultaneous minimization of the target-matching loss and the RGT-reconstruction loss. The performance of such an RGT-constrained CNN is validated by two examples, including facies identification in the Parihaka dataset and property estimation in the F3 Netherlands dataset. Compared to those purely from seismic amplitudes, both the facies and property predictions with using the proposed RGT constraint demonstrate significantly reduced artifacts and improved lateral consistency throughout a seismic survey.

Amplitude and frequency anomalies in regional 3D seismic data surrounding the Mallik 5L-38 research site, Mackenzie Delta, Northwest Territories, Canada

Geophysics ◽  
2006 ◽  
Vol 71 (6) ◽  
pp. B183-B191 ◽  
Author(s):  
M. Riedel ◽  
G. Bellefleur ◽  
S. R. Dallimore ◽  
A. Taylor ◽  
J. F. Wright
Keyword(s):  
Northwest Territories ◽  
Gas Hydrate ◽  
Seismic Data ◽  
Ratio Method ◽  
Unfrozen Water ◽  
3D Seismic ◽  
Mackenzie Delta ◽  
Data Set ◽  
Seismic Amplitude ◽  
3D Seismic Data

Amplitude and frequency anomalies associated with lakes and drainage systems were observed in a 3D seismic data set acquired in the Mallik area, Mackenzie Delta, Northwest Territories, Canada. The site is characterized by large gas hydrate deposits inferred from well-log analyses and coring. Regional interpretation of the gas hydrate occurrences is mainly based on seismic amplitude anomalies, such as brightening or blanking of seismic energy. Thus, the scope of this research is to understand the nature of the amplitude behavior in the seismic data. We have therefore analyzed the 3D seismic data to define areas with amplitude reduction due to contamination from lakes and channels and to distinguish them from areas where amplitude blanking may be a geologic signal. We have used the spectral ratio method to define attenuation (Q) over different areas in the 3D volume and subsequently applied Q-compensation to attenuate lateral variations ofdispersive absorption. Underneath larger lakes, seismic amplitude is reduced and the frequency content is reduced to [Formula: see text], which is half the original bandwidth. Traces with source-receiver pairs located inside of lakes show an attenuation factor Q of [Formula: see text], approximately half of that obtained for source-receiver pairs situated on deep, continuous permafrost outside of lakes. Deeper reflections occasionally identified underneath lakes show low-velocity-related pull-down. The vertical extent of the washout zones is enhanced by acquisition with limited offsets and from processing parameters such as harsh mute functions to reduce noise from surface waves. The strong attenuation and seismic pull-down may indicate the presence of unfrozen water in deeper lakes and unfrozen pore water within the sediments underlying the lakes. Thus, the blanking underneath lakes is not necessarily related to gas migration or other in situ changes in physical properties potentially associated with the presence of gas hydrate.

Space-time continuum in seismic stratigraphy: Principles and norms

2018 ◽  
Vol 6 (1) ◽  
pp. T97-T108 ◽  
Author(s):  
Farrukh Qayyum ◽  
Christian Betzler ◽  
Octavian Catuneanu
Keyword(s):  
Seismic Data ◽  
Seismic Stratigraphy ◽  
Seismic Interpretation ◽  
Space Time ◽  
Practical Significance ◽  
3D Seismic ◽  
Close Link ◽  
Time Continuum ◽  
3D Seismic Data ◽  
Stratigraphic Succession

Seismic stratigraphy is not only a geometric understanding of a stratigraphic succession, but it also has a close link to the space-time continuum started by H. E. Wheeler (1907–1987). The science follows the fundamental principles of stratigraphy, and the norms that govern seismic interpretation play a fundamental role due to their practical significance. The birth of computer-aided algorithms paved a new platform for seismic interpretation. The ideas from A. W. Grabau (1870–1946) and Wheeler were brought to a new level when space-time continuum was represented using 3D seismic data. This representation is commonly referred to as the Wheeler transformation, and it is based on flattening theories. Numerous algorithms have been introduced. Each suffers from its own problem and follow some assumption. The hydrocarbon industry, as well as academia, should seek a solution that is globally applicable to a stratigraphic succession irrespective of resolution, geologic challenges, and depositional settings. We have developed a review of the principles and norms behind these algorithms assisting in developing the space-time continuum of a stratigraphic succession using 2D/3D seismic data.

Facies analysis with merged 3D seismic data

2005 ◽  
Author(s):  
Laisheng Cao ◽  
Yingxin Xu ◽  
Jingbo Yu
Keyword(s):  
Seismic Data ◽  
Facies Analysis ◽  
3D Seismic ◽  
3D Seismic Data

Reservoir characterization using microseismic facies analysis integrated with surface seismic attributes

2016 ◽  
Vol 4 (2) ◽  
pp. T167-T181 ◽  
Author(s):  
Aamir Rafiq ◽  
David W. Eaton ◽  
Adrienne McDougall ◽  
Per Kent Pedersen
Keyword(s):  
Seismic Data ◽  
Principal Curvature ◽  
Facies Analysis ◽  
Shape Index ◽  
Seismic Attributes ◽  
Seismic Survey ◽  
3D Seismic ◽  
Frequency Distributions ◽  
Scaling Properties ◽  
3D Seismic Data

We have developed the concept of microseismic facies analysis, a method that facilitates partitioning of an unconventional reservoir into distinct facies units on the basis of their microseismic response along with integrated interpretation of microseismic observations with 3D seismic data. It is based upon proposed links between magnitude-frequency distributions and scaling properties of reservoirs, including the effects of mechanical bed thickness and stress heterogeneity. We evaluated the method using data from hydraulic fracture monitoring of a Late Cretaceous tight sand reservoir in central Alberta, in which microseismic facies can be correlated with surface seismic attributes (primarily principal curvature, coherence, and shape index) from a coincident 3D seismic survey. Facies zones are evident on the basis of attribute crossplots, such as maximum moment release rate versus cluster azimuth. The microseismically defined facies correlate well with principal curvature anomalies from 3D seismic data. By combining microseismic facies analysis with regional trends derived from log and core data, we delineate reservoir partitions that appear to reflect structural and depositional trends.

Sign in / Sign up

Export Citation Format

Share Document