Geophysical Insights at SEG 2017 in Houston, Booth #301
Using Attributes to Interpret the Environment of Deposition - A 1 day course. Taught by Kurt Marfurt, Rocky Roden, and ChingWen Chen
Dr. Kurt Marfurt and Dr. Tom Smith featured in the July edition of AOGR on Machine Learning and Multi-Attribute Analysis
Rocky Roden and Ching Wen Chen in May edition of First Break - Interpretation of DHI Characteristics using Machine Learning
Seismic interpretation and machine learning by Rocky Roden and Deborah Sacrey, GeoExPro, December 2016

Use SOM to define geology via natural patterns/clusters

A Presentation by Rocky Roden

An excerpt:

Initially, the complexity of the Seismic data model was a concern. There are many ways of representing the same information. This discrepancy can result in disconnects between different implementations of the same PPDM module.
However, we have been able to choose a way of representing Survey, Geometry, Trace, and storage information for Post Stack trace data (3D Seismic Attribute volumes and 2D data types) in PPDM that satis es our needs.
The goal was to be able to collect Seismic Attribute volumes (3D) and data types (2D) from Seismic Surveys from several different Oil & Gas Seismic projects via OpenSpirit.

...read more >

Presentation Slide Text...

Machine Learning and Interpretation

Intro Slide:

Slide 2

Seismic Attributes

What is a Seismic Attribute?

•A measurable property of seismic data, such as amplitude, dip, frequency, phase, and polarity. Attributes can be measured at one instance of time/depthor over a time/depth window, and may be measured on a single trace, on a set of tracesor on a surfaceinterpreted from seismic data. Schlumberger Oilfield Dictionary

•Seismic attributes reveal features, relationships, and patterns in the seismic data that otherwise might not be noticed. Chopra and Marfurt, 2007

Slide 3

Objective of using Seismic Attributes

•To take advantage of seismic attribute analysis and today’s visualizationtechnology and to minepertinent geologic informationfrom an enormous amount of seismic data.

•The ultimate goal is to enable the geoscientist to produce a more accurate interpretation and reduceexploration and development risk.

Slide 4

Most Common Seismic Attributes for Interpretation

  • Instantaneous Attributes
  • Amplitude Defining Attributes
  • Coherency/Similarity
  • Curvature
  • AVO Attributes
  • Inversion
  • Spectral Decomposition

Slide 5

Instantaneous Attributes

Instantaneous Attributes

Reflection Strength (trace envelope, instantaneous amplitude)

  • Lithological contrasts
  • Bedding continuity
  • Bed spacing
  • Gross porosity
  • DHIs

Instantaneous Phase

  • Bedding continuity
  • Visualization of unconformities and faults

Instantaneous Frequency

  • Bed thickness
  • Lithological contrasts
  • Fluid content (frequency attenuation)

Slide 6

Amplitude Accentuating Attributes

These attributes help define how the amplitude stands out against surrounding reflectors and background events.

  • DHI characteristics
  • Stratigraphic variations
  • Porosity
  • Lithology variations

Slide 7

Coherency/Similarity

Coherency, similarity, continuity, semblance and covariance are similar and relate to a measure of similarity between a number of adjacent seismic traces (multi-trace analysis). They convert data into a volume of discontinuity that reveals faults, fractures, and stratigraphic variations.

  • Crosscorelation-Based Coherence
  • Semblance-Based Coherence
  • Variance-Based Coherence
  • Eigenstructue-Based Coherence
  • Gradient Structure Tensor-Based Coherence
  • Least-Squares-Based Coherence

Slide 8

AVO Attributes

AVO attribute volumes are computed from prestackdata (gathers) . They include combinations of near, mid, and far offset or angle stacks and depending on approximations of the Knott-Zoeppritzequations, various AVO components. Most of the AVO attributes are derived from intercept and gradient values or equivalents. Employed to interpret pore fluid and/or lithology.

  • Intercept (A)
  • Gradient (B)
  • Curvature (C)
  • A*B
  • A-C
  • ½ (A+B)
  • ½ (A-B)
  • Far-Near
  • (Far-Near)Far
  • Poisson Reflectivity
  • Fluid Factor
  • Lambda-Mu-Rho

Slide 9

Seismic Inversion

  • Inversion transforms seismic reflection data into rock and fluid properties.
  • The objective of seismic inversion is to convert reflectivity data (interface properties) to layer properties.
  • To determine elastic parameters, the reflectivity from AVO effects must be inverted.
  • The most basic inversion calculates acoustic impedance (density X velocity) of layers from which predictions about lithologyand porosity can be made.
  • The more advanced inversion methods attempt to discriminate specifically between lithology, porosity, and fluid effects.

 

  • Recursive Trace Integration
  • Colored Inversion
  • Sparse Spike
  • Model-Based Inversion
  • PrestackInversion (AVO Inversion)
    • Elastic Impedance
    • Extended Elastic Impedance
    • Simultaneous Inversion
  • Stochastic Inversion
    • Geostatistical
    • Bayesian

Slide 10

Spectral Decomposition

Use of small or short windows for transforming and displaying frequency spectra (Sheriff, 2005 Encyclopedic Dictionary of Applied Geophysics). In other words, the conversion of seismic data into discrete frequencies or frequency bands.

  • Layer thickness determinations
  • Stratigraphicvariations
  • DHI characteristics (e.g. shadow zones)

 

  • Discrete Fourier Transform
  • Fast Fourier Transform
  • Short Time Fourier Transform
  • Maximum Entropy Method
  • Continuous Wavelet Transform
    • Gabor
    • Gabor-Morley
    • Gaussian
  • Spice
  • Continuous Wavelet Packet-Like Transform
  • Wigner-Ville Distribution
  • Smoothed Wigner-Ville Distribution
  • Matching Pursuit
  • Exponential Pursuit

 

Slide 11

Curvature

Curvature is a measure of bends and breaks of seismic reflectors. Another way to describe curvature for any point on a seismic reflecting interface is the rate of change of direction of a curve.

  • Fractures 
  • Folds 
  • Faults

 

  • Mean Curvature
  • Maximum Curvature
  • Minimum Curvature
  • Gaussian Curvature
  • Most Positive Curvature
  • Most Negative Curvature
  • Shape Index Curvature
  • Dip Curvature
  • Strike Curvature
  • Curvedness

Slide 12

Even though seismic attributes can provide important information in the interpretation process, how efficiently can dozens of seismic attributes be interpreted and how do they relate or not relate to each other?

New technology is required to address these issues. 

Slide 13

New Terminology

  • Big Data
  • Machine Learning
  • Unsupervised Learning
  • Principal Component Analysis
  • Artificial Intelligence
  • Neurons
  • Self-Organizing Maps
  • Pattern Recognition
  • Cluster Analysis
  • 2D Colormap 

Slide 14

Big Data - Entering the era of data deluge, where the amount of data outgrows the capabilities of query processing technology.

How do geoscientists efficiently interpret all this data and understand how these different data types relate to each other?

Machine Learning has evolved to address Big Data. 

Slide 15

Machine Learning - Uses computer algorithms that iteratively learn from the data and independently adapt to produce reliable, repeatable results. 

Machine learning employing current computing technology and visualization techniques addresses two significant issues in seismic interpretation:

  1. The Big Data problem of trying to interpret dozens if not hundreds of volumes of data and
  2. The fact that humans cannot understand the relationship of several types of data all at once. 

Slide 16

Types of Machine Learning

Supervised Learning

  • Multi-layer Perceptron NN
  • Probabilistic NN
  • Generalized Regression
  • Radial Basis Function NN

Supervised Learning - the correct/desired answers are known from training on a dataset with known results, then this training is applied to a dataset with unknown results. 

Unsupervised Learning

  • K-means
  • Projection Pursuit
  • Principal Component Analysis
  • Self-Organizing Maps
  • Vector Quantization
  • Generative Topographic Mapping

Unsupervised Learning - The correct/desired answers are not know from a previous dataset, the training adapts to the data, identifying natural patterns or clusters. 

Slide 17

Principal Component Analysis (PCA)

A linearmathematical technique to reduce a large set of variables (seismic attributes) to a small set that still contains most of the variation in the large set.

In other words, find the most meaningful seismic attributes!

Slide 18

What is the largest variation in the data?

The eigenvectoris the direction of the line showing the variance or spread in the data.

The eigenvalueis the value showing how much variance there is in the eigenvector.

Slide 19

How PCA relates to finding the most significant seismic attributes (12 seismic attributes were employed)

inlineThe first principal component accounts for as much of the variability in the data as possible, and each succeeding component (orthogonal to each preceding) accounts for as much of the remaining variability as possible.

Slide 20

Principal Component Analysis (PCA)

The seismic attributes contributing the most to the first few principal components often indicate the most important seismic attributes to define geological features in the data.

The interpretation of the most important seismic attributes from PCA can be employed in a Self-Organizing Map (SOM) Analysis

Slide 21 Self Organizing Maps

Slide 22

Biological Neural Network

The brain is a collection of 10 billion interconnected neurons, each is a cell that uses biochemical reactions to receive, process, and transmit information.

Slide 23

Artificial Neural Networks (ANN)

(An information processing technology pertaining to the area of machine learning in artificial intelligence)

An Artificial Neural Network is a computational simulation of a Biological Neural Network, composed of a large number of highly interconnected processing elements (neurons)

Slide 24

Self-organizing Maps (SOM)

Self-organizing Maps (SOM)

  • Self-Organizing Maps are artificial neural networksemploying unsupervised learning methods.
  • A SOM is a cluster analysisand pattern recognitionapproach.
  • A SOM analysis produces a collection of neuronswhich classify data samples into categories, patterns or clusters based on their properties.
  • A neuronis a point that identifies a natural cluster of attributes.

Slide 25

SOM Classification for 10 Attributes

SOM analysis maps the natural clusters from the numerous attributes’ data points to a 2D Colormap

AttributesEach sample in the 3D survey will have 10 valuesassociated with each attribute

Multi-Dimensional Space = Data from Numerous Attributes we cannot visualize multiple dimensions at the same time (10 attributes = 10 dimensional space)

Slide 26

Self-organizing maps (SOM)

The neurons from the Classification on the 2D Colormapare interpreted for geologic significance

Slide 27

Additional Information

  • Well Logs
  • Production
  • Interpretation Knowledge
  • etc.

Seismic Attributes

  • Instantaneous
  • Geometric
  • AVO
  • Spectral Decomposition
  • etc.

Predictions

  • Facies
  • Porosity
  • Thickness
  • Faults/Fractures
  • DHI’s
  • Stratigraphy
  • Channels
  • Data Quality Artifacts
  • Pressure
  • etc.

Slide 28 - What are the Interpretation Steps to use PCA and SOM?

Slide 29 - Step 1

  • Differential Compaction
  • Facies
  • Porosity
  • Thickness
  • Fractures
  • Pressure
  • Faults
  • Channels
  • DHI's
  • Acquisition Footprint
  • Stratigraphy
  • Depositional Environments
  • Processing Artifacts
  • Karsts

Slide 30 - Step 2 Which attributes to choose

Slide 31 - Step 2

Select seismic attributes for SOM based on:

  1. Principal Component Analysis
  2. Previous knowledge and experience of appropriate attributes for the geologic feature of interest. 

Slide 32

Instantaneous Attributes

  • Reflection Strength, Instantaneous Phase, Instantaneous Frequency, Quadrature, Instantaneous Q
  • Lithology Contrasts, Bedding Continuity, Porosity, DHI's, Stratigraphy, Thickness

Geometric Attributes

  • Semblance and Eigen-Based Coherency/Similarity, Curvature (Maximum, Minimum, Most Positive, Most Negative, Stike, Dip)
  • Faults, Fractures, Folds, Anisotropy, Regional Stress Fields

Amplitude Accentuating Attributes

  • RMS Amplitude, Relative Acoustic Impedance, Sweetness, Average Energy
  • Porosity, Stratigraphic and Lithologic Variations, DHI's

AVO Attributes

  • Intercept, Gradient, Intercept/Gradient Derivatives, Fluid Factor, Lambda-Mu-Rho, Far-Near, (Far-Near)Far
  • Pore Fluid, Lithology, DHI's

Seismic Inversion Attributes

  • Colored inversion, Sparse Spike, Elastic Impedance, Extended Elastic Impedance, Prestack Simultaneous Inversion, Stochastic Inversion
  • Lithology, Porosity, Fluid Effects

Spectral Decomposition

  • Continuous Wavelet Transform, Matching Pursuit, Exponential Pursuit
  • Layer Thicknesses, Stratigraphic Variations

Slide 33 - Step 3

Run SOM  with combinations of attributes to resolve the defined geologic problem. 

A generic 8X8 (64 neurons) neuron count may be a good first run, however; depending on the scale of the geologic feature to identify, a lower count such as 3X3 or 4X4 may be warranted.

Slide 34 - Step 4

Interpret SOM results with 2D colormapIsolate neurons or combinations of neurons to identify geologic features

Slide 35 - Step 5

Refine Interpretation:

  • Modify attribute list employed in SOM analysis
  • Use different neuron counts (3X3, 8X8, 20X20, etc.)
  • Re-evaluate most meaningful neurons associated with geologic significant features (develop new 2D colormaps)

Slide 36 - Case Study

Slide 37

Offshore Gulf of Mexico Case Study - Class 3 AVO

3900' Reservoir

  • UpthrownFault Closure
  • Approximately 100’ Reservoir Sand
  • Two Producing Wells
    • #A-1 (gas on oil)
    • #A-2 (oil)

 

Slide 38

Gulf of Mexico Case Study Seismic Attributes Employed in PCA

  • Acceleration of Phase
  • Instantaneous Frequency Envelope Weighted
  • Average EnergyInstantaneous Phase
  • Bandwidth
  • Instantaneous Q
  • Dominant Frequency
  • Normalized Amplitude
  • Envelope Modulated Phase
  • Real Part
  • Envelope Second Derivative
  • Relative Acoustic Impedance
  • Envelope Time Derivative
  • Sweetness
  • Imaginary Part
  • Thin Bed Indicator
  • Instantaneous Frequency
  • Trace Envelope

Slide 39

SOM A

  • Trace Envelope
  • Envelope Modulated Phase
  • Envelope 2ndDerivative
  • Sweetness
  • Average Energy

SOM B

  • Instantaneous Frequency
  • Thin Bed Indicator
  • Acceleration of Phase
  • Dominant Frequency

Slide 40

Attributes for Attenuation
1.Envelope Second Derivative
2.Envelope Modulated Phase
3.Trace Envelope
4.Average Energy
5.Sweetness

Slide 41

Attributes for Attenuation
1.Envelope Second Derivative
2.Envelope Modulated Phase
3.Trace Envelope
4.Average Energy
5.Sweetness

Slide 42

Attributes for Flat Spots
1.Instantaneous Frequency
2.Thin Bed Indicator
3.Acceleration of Phase
4.Dominant Frequency

Slide 43

Attributes for Flat Spots
1.Instantaneous Frequency
2.Thin Bed Indicator
3.Acceleration of Phase
4.Dominant Frequency

Slide 44

Summary and Conclusions

  • Today’s enormous amount of interpretation data and the desire to derive more meaningful information, requires implementation of New Technology.
  • Machine Learningin the form of Principal Component Analysis (PCA)and Self-Organizing Maps (SOM)are resolving Big Data issuesand extracting more meaningful information from our data.
  • PCA identifies the most prominent attributesin our seismic data and helps determine appropriate attributes for SOM.
  • SOM (ANN) identifies the natural patterns/clustersfrom various combinations of seismic attributes.
  • These natural patterns/clusters help define geologic featuresthat are difficult to interpret or can not be identified otherwise.
  • Using modern computing and visualization technology, these Machine Learning approaches appear to be the Next Advancement in Seismic Interpretation.

Member Login
Welcome, (First Name)!

Forgot? Show
Log In
Enter Member Area
My Profile Not a member? Sign up. Log Out