GeoPhyZViZ - cs.ubc.ca

18
GeoPhyZViZ Exploration of seismic attributes

Transcript of GeoPhyZViZ - cs.ubc.ca

Page 1: GeoPhyZViZ - cs.ubc.ca

GeoPhyZViZExploration of seismic attributes

Page 2: GeoPhyZViZ - cs.ubc.ca

Seismic acquisition

Page 3: GeoPhyZViZ - cs.ubc.ca

Seismic traces

• Energy reflection interfaces are recorded by receivers at the surface

Page 4: GeoPhyZViZ - cs.ubc.ca

Seismic data

Page 5: GeoPhyZViZ - cs.ubc.ca

Seismic analysis

Wref = f(✓,�, ⇢, ..., )Win

Reflected wavelet is a function of the rock properties at the reflection interface.

Page 6: GeoPhyZViZ - cs.ubc.ca

Signal analysis

• In reality, the transfer function is unknown

• Loose empirical relationships based on signal analysis

• Decompose the signal into “attributes”

• Close ties to black magic

Page 7: GeoPhyZViZ - cs.ubc.ca

Seismic attributes

Every attribute creates another data volume increases the dimensionality

Page 8: GeoPhyZViZ - cs.ubc.ca

Attribute analysis

• Crossplots: Looking for relationships between attributes

• Most reflections have similar properties, forming background trends

• Reflections from resource reservoirs are outliers

Page 9: GeoPhyZViZ - cs.ubc.ca

Broken workflow• Sometimes done using excel spreadsheets

• Using multiple software packages to hack an analysis result together

• Large information space, need interactivity to do efficient analysis

• Need linked displays in the same tool

Page 10: GeoPhyZViZ - cs.ubc.ca

https://geophyzviz.appspot.com

/

Page 11: GeoPhyZViZ - cs.ubc.ca

What?Seismic data is a field where each cell contains a position (lat, lon, depth) and attributes (amplitude, similarity, coherence, phase, etc..)

Page 12: GeoPhyZViZ - cs.ubc.ca

Why?Actions and targets

• Discover outliers, trends, features

• Need to identify and compare anomolies

• Relate to spatial data

Page 13: GeoPhyZViZ - cs.ubc.ca

How - encode• Field data as a heatmap, using the colour channels to encode two

attributes

• Attributes as scatter plots, which uses point marks to encode attributes via the spatial channel

Page 14: GeoPhyZViZ - cs.ubc.ca

How - manipulate

• Look for how clusters move through attribute space using change animations

• Select outliers using brushes

• Navigate views by clicking on summary plots

Page 15: GeoPhyZViZ - cs.ubc.ca

How - facet• Small multiples to summarize and juxtapose the

data

• Superimpose the selected anomalies on the spatial plot

Page 16: GeoPhyZViZ - cs.ubc.ca

https://geophyzviz.appspot.com

/

Page 17: GeoPhyZViZ - cs.ubc.ca

Epilogue• Optimize colour maps

• Use on larger real datasets

• How does d3 scale for massive scatter plots?

• Will interpreters find it useful?

Page 18: GeoPhyZViZ - cs.ubc.ca

ReferenceVisualization Analysis and Design, Tamara Munzner (A K Peters Visualization Series, CRC Press, 2014)

http://www.capefarewell.com/climate-science/the-science/seismic-profiling.html

Successful AVO and Cross-plotting, Satinder Chopra, CSEG workshop http://cseg.ca/technical/view/successful-avo-and-cross-plotting

http://wiki.seg.org/wiki/AVO_equations

http://www.allaboutcircuits.com/technical-articles/an-introduction-to-digital-signal-processing/