Are you a geophysicist that processes seismic data, or someone who uses the processed results of that data? If so, you probably think of seismic data as something that arrives on a tape or "from the cloud." If you're lucky it arrived already sliced into chunks called "traces"; sometimes you might have to cut up the data yourself. In either case the job of a processor is to run the traces through algorithms that turn the seismic data into the best image they can make. You might even aspire to getting some usable quantitative information out of the data along the way (depending on what kind of geophysicist you are). Generally you are satisfied to get "a good image" out of the data.
However, your data also has other, hidden stories to tell—stories that likely were shredded and lost when you chopped the data into traces and fed it into your algorithms for processing. Traditional processing also loses quantitative detail. What do the numbers in that trace really mean? How were they recorded? Scientists and engineers are trained to keep careful track of units. Isn't a geophysicist supposed to be a kind of analytical scientist or engineer, someone who deals in the hard sciences of math and physics? OK then, what units are your data in? Can you tie the numbers in your traces back to what happened at the source? If you can't, how do you know when your source is fit for purpose?
We typically call anything our algorithms are not designed to deal with "noise." Some of it comes from the limitations of the approximations we make. However, some of it comes from what is happening during recording. For example, your source may be varying from shot to shot. There are always other sources of sound in the environment that your receivers are recording along with the desired signal. These details are in your data, but typically ignored. Can we make use of such "noise," or at least better understand it? If we understood it, could we do something useful with it (or at least have a better idea of how to suppress it)?
The goal of this course is to get you thinking more critically about your data: how was it recorded, what is in it, and what happened to it on the way from the field to numbers in a file. It should give you the basic concepts you need to wring more from your data by teaching you how to analyze your data quantitatively, like a scientist or engineer performing a forensic investigation. The concepts will be supported both by model and real-world examples. Although there will be a bit of math here and there, the emphasis will be on understanding the general principles -- what the math means, not the calculation. Much of the material will be familiar to those who attended my 2016 Distinguished Lecture of the same name. The material will be presented in a way that is accessible to as broad an audience as possible, while still providing rigorous detail to those who need it.
Joe Dellinger was born in the SEG home town of Tulsa, Oklahoma, and learned to ride a bicycle practicing in the Amoco Tulsa Research Center parking lot. Dellinger’s father, Tom, led a research group at the Mobil Field Research Lab in the 1970s-1980s, which coined the term “Extended Reach Drilling” to describe to management what it was they were doing. With this background, it is not surprising that Dellinger majored in Geophysics at Texas A&M. He received a PhD in 1991 from Jon Claerbout’s Stanford Exploration Project. He then did a three-year post-doc at the University of Hawaii before joining Amoco in Tulsa in 1994. He moved to BP in Houston in 1999 and has worked there since. In his career he has specialized in anisotropy, multi-component algorithms and processing, and most recently investigating the problem of how to record ultra-low frequencies with the goal of enabling inversion algorithms like FWI to resolve complex velocity-model-building challenges in deep-water marine environments.
This last challenge required Dellinger to look closely at “useful information in our seismic data that is normally ignored,” i.e., “forensic data processing.” This has included studying the 2006 “Green Canyon” earthquake, investigating how the Valhall Ocean-bottom-cable array might be used between seismic surveys, and characterizing seismic sources and noise in deep-water ocean-bottom Gulf of Mexico data. In the course of that project BP created a new vibratory low-frequency marine source, Wolfspar®, which proved to be particularly amenable for these studies because it has a precisely known source signature. These learnings became the basis for his Spring 2016 Distinguished Lecture and will be the core of the follow-up 2021 short course.
Dellinger was awarded Lifetime Membership at the SEG in 2001 for his services in helping the SEG to successfully adapt to the internet age, honorary membership in 2016, and the Kauffman award in 2021 for his efforts in developing the industry’s abilities to record ultra-low frequencies. Dellinger ’s hobbies include attending the Houston Symphony, photographing birds, recording frog calls in the swamps around Houston, and astronomy at the George Observatory. Asteroid “78392 Dellinger” was named in his honor.
If you are interested in hosting our 2022 DISC, please contact us.