Home ]


SWLGS Luncheon Topics

Updated August 05, 2003

March 2000

Seismic Attributes Past, Present, and Future
A. E. Barnes, Landmark Graphics

Abstract: Seismic attribute analysis, for years a tawdry business, has undergone a makeover and is once again fashionable. Is this something to cheer? To answer this, I reverently cast an eye over its past and present, and fearlessly predict its future. I restrict my survey to attributes derived directly from poststack seismic data.

Seismic attribute analysis was born with the invention of digital recording and the subsequent discovery of bright spots in the 1960's. For the first time, geophysicists recognized that there was more to be learned from seismic data than geologic structure. This led directly to Nigel Anstey's pioneering ideas of attribute analysis in the early 1970's. His remarkable work inspired the no less remarkable work of Tury Taner and Bob Sheriff that resulted in complex seismic trace analysis. Timing is everything: this method, introduced with seismic stratigraphy and at the same time that color plots first became practical, was hugely influential and popularized seismic attributes.

The 1980's saw a proliferation of wondrous new attributes developed by energetic people who hoped the world might make sense of them. Every company had its own private attribute, so that attribute analysis became like that oriental religion in which every village worships its own god. Philosophical geophysicists mused, "What does it all mean?" Many of them concluded: not much.

But the late 1980's also saw the first development of multidimensional attributes, which led to continuity, dip, and azimuth in the 1990's. These new attributes have proven so wildly successful that they have breathed new life and respectability into attribute analysis overall. Their popularity owes much to simple meanings and geologic relationships. They represent a fundamental advance, but are only a first step towards a future in which attributes will become more geological and less geophysical, and will be meaningfully combined to predict stratigraphic properties. This is indeed something to cheer about..

Biographical Sketch: Art Barnes earned a BS in physics in 1974 from Denison University in Ohio, an MS in geophysics in 1980 from the University of Arizona in Tucson, Arizona, and a PhD in geophysics in 1990 from Cornell University in Ithaca, New York. With all that fancy education, you would think he would have learned not to volunteer to give a talk.

Art first entered the oil industry in the halcyon days of 1980, when one could get a job by spelling geophysiks. He worked on a marine seismic crew and in seismic data processing. In 1986, encouraged by market fundamentals to seek new opportunities, he re-entered academics. He worked in both the COCORP and Lithoprobe deep seismic exploration projects, becoming an expert in random seismic noise. During this time, he pursued research in attribute analysis and signal processing, publishing several papers on these topics in leading journals and misleading journals.

In 1995, Art returned to the oil industry as a software engineer, knowing that good times were here to stay. He joined Landmark Graphics in Denver in 1997. He currently maintains PostStackô, Landmark's product for interpretive seismic data processing, to which he is adding new attribute functionality. He is also fond of repairing GUIs and updating IO because it is more glamorous.