Peter Cary is an experienced geophysicist who specializes in the processing of surface seismic data and in writing algorithms for both conventional P-wave and multi-component data.
Peter has a Ph.D. in geophysics in 1987 from Cambridge University, England, where his research focused on nonlinear inverse theory. Soon after completing his Ph.D., Peter started working at Pulsonic Geophysical Ltd in Calgary and remained there for 8 years.
While at Pulsonic some of his most fruitful efforts went towards deconvolution and multicomponent processing. He has now been at Sensor Geophysical Ltd. for the last 11 years, and is engaged in work on deconvolution, migration, Radon transforms and multicomponent seismic data, among many other efforts. Along with his professional work, Peter has contributed greatly to the geophysical community in other ways. In 1994, he lectured at the University of Calgary on 3D seismic analysis. He served on the Technical Committee for the CSEG Convention for three years (1991, 92, 93) and was chairman of that committee in 1993. He organized workshops on the topics of AVO, complex imaging and phase for those conventions. He was 2nd Vice-President of the CSEG in 1996-97, and CSEG president in 2004-2005. He has been the Associate Editor (seismic processing) of Geophysics from 1998-2001 and is the recipient of the 2008 CSEG Honorary Membership Award.
We approached Peter cautiously for an interview, as he believes that awards and other forms of recognition are not necessarily appropriate, and prefer that his work speak for itself. We were delighted that he consented to the interview, which follows.
(Photos courtesy: Penny Colton)
Please tell us about your educational background and your work experience.
I’m originally from Alberta, growing up in Edmonton, but I finished high school in Timmins, Northern Ontario, where my father worked for several years as a chemical engineer in the mining industry. So I ended up going to university in Ontario, at the University of Toronto getting a B.Sc. in physics with a specialization in geophysics. Thereafter I had a brief stint at UBC where I began studying glaciology before returning to the University of Toronto to do a B.A. in philosophy. After that two year departure from science I came back to my senses and continued in geophysics first by getting an M.Sc. at U of T and then a Ph.D. at the University of Cambridge. Between my master’s degree and Ph.D. I worked for Chevron, first in Calgary and then in their research lab in La Habra, California. After my Ph.D., I joined Pulsonic Geophysical for 8 years. In 1996 I moved to Sensor Geophysical where I am today.
So after getting your B.Sc. degree in physics, were you not interested in continuing with science, as you chose to do a B.A. in philosophy instead?
I suppose I originally started studying physics because I wanted to understand the fundamentals of how the universe, or nature, works. For quite a while before I reached university age I had an interest in popular science topics like astronomy, and that led to a general interest in physics because it led to subjects like relativity, Newton and Einstein and all that stuff. A desire to understand the fundamentals of how the universe works must sound grandiose, but that probably accurately describes the naïve passion that I and other first-year students in physics have.
So I started studying physics and math and enjoyed it tremendously for a few years. I loved, and still do love, the university environment that fosters intellectual freedom and pursuit of knowledge for the sake of knowledge, so my interests kept broadening while I got to know more people in other disciplines and did more reading and attending public lectures, etc. I gradually became interested in even more fundamental questions, questions that were more philosophical than scientific. Perhaps it was just taking me a long time to grow up, but I gradually became aware that science, even physics which I considered to be the most fundamental science, had a human aspect to it. It was around the time that the philosophy of science was being rocked by Thomas Kuhn’s “The Structure of Scientific Revolutions”, a book that fundamentally challenged any notions that science should be described as a slow, steady growth in human knowledge. Kuhn pointed out how data, evidence and experiments that do not happen to fit well with the prominent theory of the day can be ignored and even scorned by the scientific community, or just as easily they can be used to justify a shockingly new theory, if that happens to be your point of view. This kind of idea sort of shook me up as I came to understand that he was at least partly right.
I don’t want to imply that Thomas Kuhn’s book is why I decided to study philosophy. But it was those types of issues that made me want to set science aside and study philosophy for a while. To make a long story short, I studied philosophy, as much of it as I could consume, for two years, and loved every minute of it. I think my parents were worried that I had gone over the deep end, or that I was squandering my youth, but I think I may have enjoyed that part of my education more than any other. My science background was actually excellent training for the analytical thinking that is required for philosophy, at least for western philosophy. But after a couple of years I began to tire of the endless argumentation without much hope of every settling the argument, which is what most philosophers do, unless you decide to accept one side of the argument or the other. I could never help but see all sides of every argument, which was leading to endless frustration. I yearned to have that feeling that scientists enjoy that comes from making progress, even if it is minor progress and even if it is just progress within a system of knowledge that would someday be overthrown by another. So I went back to studying our wonderfully down to earth science, geophysics. Sorry for going on so long with my answer to your question. I guess just mentioning philosophy gets me going again.
How did you choose University of Cambridge after doing your Masters from U of T? What differences did you perceive as you changed universities?
I did my master’s degree in geophysics at U of T under the supervision of Chris Chapman. I worked on Radon transforms, which was a really new topic at the time. It was a thoroughly enjoyable experience working under a geophysicist of the caliber of Chris Chapman. I was awarded an NSERC postgraduate scholarship just as I was finishing my M.Sc. program but I had already accepted a job with Chevron in Calgary. I was allowed to defer the scholarship for up to two years so I decided to go to work and decide later as to whether I wanted to do a Ph.D. During those two years Chris Chapman moved from U of T to Cambridge to take up the position as the head of Bullard Labs. The chance to attend Cambridge and once again study under Chris Chapman was too much to resist. It was a tough decision to leave Chevron because they had transferred me to their research lab in California where I was able to do some really interesting work just when 3D seismic was starting out. I was extremely fortunate to meet and work with some fantastic people, such as Dave Hale. It was a great place to work.
What kind of research work did you do for your Ph.D.?
I worked on a general approach to inverting marine refraction data using waveform fitting of WKBJ seismograms and inversion ideas from Bayesian and Monte Carlo inversion. It was a very ambitious approach to inversion that did not have much of an immediate application to exploration geophysics. The kind of approach I was working on has much in common with the waveform inversion work that Gerhart Pratt and his students have been doing for many years now and been getting such good results with. It’s still an ambitious approach to inversion of reflection data because of the computational requirements and the sensitivity to the starting model, but perhaps now it is becoming more practical. It was an excellent topic to do a Ph.D. on because it trained me in many aspects of inverse theory, and after all, inversion is basically all that we geophysicists do.
What do you have to say about how your career has shaped up so far?
I feel extremely fortunate to have always had a job that I enjoyed doing.
What more can you ask for than to have a job that you look forward to doing each day?
If I were to ask you to list three of your qualities that would reflect Peter’s personality, what would they be? I know one for sure: you are shy about certain things, right?
I believe that I said at some point that I was shy about doing this interview, so that must be where your comment about me being shy comes from. But I’m not sure that I chose the right word. Reluctant is probably a better word to describe how I felt about agreeing to do this interview rather than shy. I was reluctant because I feel that readers of The RECORDER should not be more interested in reading an interview with me than with anyone else in the geophysical industry. Why me rather than anyone else? I realize that I am in some sense “recognizable” in the Calgary geophysical community but that is because I am one of those people who frequently gets up and gives talks at conferences or writes articles about the work I’ve been doing. But I am just doing my job. It should not mean that I am in any way more interesting as the subject of an interview than any one else who has a job that makes them less recognizable. As for your original question about qualities that reflect my personality, I’ve found that there can be two quite different opinions about a person: first is the person’s own opinion of himself and then there is the right opinion. If you don’t mind, I am going to shy away from answering your question.
After getting your Ph.D. from University of Cambridge, you started working at Pulsonic Geophysical as Manager of R and D. How would you describe your 8 years there?
I finally decided to leave the academic world and go to work in Calgary in 1988, the winter that the Olympics came to Calgary. I wanted a job that would challenge me in terms of what I was going to be working on. Working at Pulsonic was great because I was responsible for keeping the company up to date on all aspects of data processing. The responsibilities were wide-ranging but I was always trusted and given freedom to work on whatever I considered to be important at the time. Of course I soon learned that what I worked on was usually forced upon me by forces outside my control but I’ve never really minded that. I always enjoy working on things that will find an immediate application. It makes the problem more real, in a sense. I had the opportunity to work with many fine individuals when I was at Pulsonic, too many to name here. It was an excellent place to meet many people in the Calgary community. Pulsonic was also the place where my work on multicomponent seismic began. It was funny how it started…another topic of research that was thrust upon me. Rob Stewart put a call out for membership in his new consortium at U of C, the CREWES Project. Hugh Stanfield, the owner of Pulsonic, without hesitation decided that Pulsonic should join. Before I knew it, I had a 3-C line from Unocal, and I was trying to figure out how to process it. That was about 20 years ago. I’ve been trying to figure out how to process 3-C data ever since, thanks originally to Hugh Stanfield and Rob Stewart.
Thereafter, you started working at Sensor Geophysical in 1996. How did the focus of your research change as you moved from Pulsonic to Sensor?
I’m not sure that my research has ever had much focus, to tell you the truth! I kept working on things that were thrust upon me, as well as finding time to work on things that I just wanted to work on. Multicomponent processing continued, and I did some more work on Radon transforms that grew out of the ideas of Mauricio Sacchi and Tad Ulrych at UBC. I became very interested in the problems to do with spatial sampling, especially as they had to do with the interaction between design, acquisition and the imaging steps in processing such as DMO and prestack time migration. I was always concerned during the growth in popularity of prestack time migration that prestack migration was going to do more harm than good because we were not even capable of migrating flat events as well with prestack migration as with poststack migration. Since so much Canadian data is flat-lying plains data, that seemed like a legitimate concern. That was what led me towards thinking about breaking up 3D datasets into common-offset vector gathers. I was looking for a way of reducing migration artifacts from the impact of irregular spatial sampling.
Looking back on your geophysical career, would you share with us one or two of your most exciting successes?
When I started my geophysical career I was naïve enough to believe that what I was accomplishing was hugely important and therefore exciting. But let’s face it, seismic processing is an excruciatingly boring subject to everyone else but seismic processors. I completely understand those people who look at you when you tell them that you are a geophysicist, and they have a completely blank look on their faces. What could be exciting about a subject as dry as deconvolution for goodness sake! Whatever excitement we experience while studying a subject like deconvolution could only be due to the emotion that we can share with others. So by far the most exciting times that I have experienced as a geophysicist have come from personally working with other geophysicists. There are so many of them that I have had the privilege to work with through the years that I hesitate to start naming them because I fear that I would leave out too many names.
What personal and professional vision are you now working towards?
Well, as usual what I am now working on right now is not the result of any vision but what is being forced upon me by immediate circumstances. I suppose that I am working towards things, or at least I have wishes about what I would like to accomplish, if only I knew how. I wish I could improve the resolution and quality of P-S data so that it was something closer to what we get from P-P data. I wish I understood what makes one dataset so much better or worse than another, or even why the quality varies within one survey so much. It is always such a mystery and we always point at the near-surface as the culprit. But we sometimes don’t even know that for sure. It would be nice to tie together some of the analysis of P-P AVAZ and VVAZ with the shear-wave splitting observations on P-S data. There is so much obvious work to be done there. And of course, if I had the time, I would like to work on fundamental problems like multiple attenuation, deconvolution and phase stability. There is basically no aspect of seismic processing that does not deserve more attention. When it comes right down to it, almost nothing that we do to the data works completely satisfactorily in all situations.
Looking at your list of publications, I notice you have worked in quite a few areas of geophysics. Beginning with the discrete Radon transform, nonlinear Bayesian inversion, 3D migration, neural networks for editing, four-component deconvolution and thereafter you switched over to multicomponent processing, inverse Q filtering, etc. You seem to be all over the place. Your comments?
I guess my comment is that seeming to be “all over the place” is mostly in the eye of the beholder. Yes, I have worked in many areas, but they have all had to do with seismic data processing. Working for a small company, I tend to encounter all aspects of data processing all the time. They all need attention because they are all important in their own way, and that is what I have done through the years…tried to pay attention to all these different aspects of processing.
What according to you is your most important contribution to geophysics?
To be perfectly honest, any accomplishments that I have been fortunate enough to make have been incremental at best, and would certainly have been made by someone else soon afterwards if not done by me.
Are there other areas of geophysics that fascinate you in particular?
Other than seismic processing? At one point a long time ago I was determined that I was going to be a glaciologist. I worked with Garry Clarke at UBC on some glaciers up in the Yukon one summer. The physics of glaciers and ice sheets still fascinates me. And I worked for many summers in the Canadian bush doing mining geophysics, EM, magnetics, IP, gravity, etc. I enjoyed that work very much, and it gave me an understanding and appreciation of that branch of exploration geophysics. It is amazing how there are some areas of geophysics, like global geophysics, that operate in virtual isolation from exploration geophysics, and vice versa. I remember being interested in some aspects of global seismology when I was in grad school, but I never interact with those types of geophysicists any more.
Tell us about inverse Q-filtering. How and how much is it being used in the industry? What are the challenges that we still face?
That is an interesting question to be asked because I did not expect it. I’m not sure that I have much to say about it either. My perception is that inverse Q filtering is not being done very much at all, to my knowledge. I know that we don’t use it very often at Sensor. It is sometimes done, especially on marine data or on land data with longer record lengths to try to reverse the obvious drop in frequency content with time. Time-varying deconvolution or time-varying spectral whitening are theoretically poor, but practically successful, substitutes. Gary Margrave has of course been pushing the incorporation of inverse Q filtering as part of his Gabor deconvolution. Q is a funny thing. Everybody agrees that our data suffers from its influence, and there are various very good algorithms for doing inverse Q filtering, but still it is not a popular thing to apply to the data. I went through a period years ago of trying to encourage people to use it more. That was back when Roger Hawthorne first joined Pulsonic. The idea was stillborn. People want high frequency data, of course, but once Q has knocked the frequencies below the noise level, no filter, whether it is an inverse Q filter or any other filter, will ever retrieve that signal. Of course there are inversion methods that retrieve higher frequencies, but those are outside the realm of what I would call normal signal processing. And inverse Q filtering can change the phase of data as well. But phase is such a strange indeterminate animal. Interpreters don’t seem to mind if their data has the wrong phase. They almost expect it. Even if the phase changes a bit with time, nobody gets upset. Lateral changes of phase at the zone of interest are the only thing they don’t want happening. And inverse Q filters won’t affect that, at least if they are run with the same values of Q across the data. I think there is probably more of a future in detecting changes in Q in time-lapse data, or in measuring differences in Q in fast and slow split shear waves than in trying to remove the effects of Q from the data.
Your name is often associated with four-component surface-consistent deconvolution. Were you the originator of this idea? How did you think about it?
I certainly was not the originator of the idea of four-component surface-consistent deconvolution. Many people had worked on it before me, and people were applying it to data, although it involved a lot of data sorting. If you read my paper carefully, you will realize that all I did was point out that summation and multiplication commute in the application of the surface-consistent equations. By rearranging the equations I realized that surface-consistent deconvolution with multiple components could be applied in just two passes over the data, the first to design the operators from the data, and the second to apply the operators to the data. That made it easy to apply to large datasets. That’s basically all I did.
Your company is known for multicomponent seismic data processing. Tell us about the promise of multicomponent data. I ask this because it has been around for quite some time and the industry has not embraced it like it adopted 3D seismic in the late 1980s and early 1990s. Do you think the benefits of multi-component seismic data are commensurate with the investment that the oil companies make?
I would say that the benefits of multicomponent data must be close to being commensurate with the investment because the oil companies are not investing that much into multicomponent seismic! The extra cost of acquiring multicomponent land data these days is minimal, if it is any added cost at all. And the cost of processing, as we know, is still minimal compared to acquisition. So it appears to me that most oil companies are not investing much at all into multicomponent data. Several acquisition companies have invested a lot, however, into modern multicomponent acquisition systems. So there is a fair amount of 3-C data being acquired out there, and a good deal of it is being processed. The benefits of 3-C data don’t just fall in your lap though. The benefits come with knowledge, experience and effort. The potential benefits of multicomponent data have been known for so long now, much longer than 3D. Those potential benefits are the same today as they were 50 years ago. I used to think that there was a lack of multicomponent case studies, but there are a surprising number of published case studies around now. Bob Tatham at UT Austin keeps a database on the web of all the published case studies. Everyone likes to use the 3D seismic revolution for comparison for new technologies, of course, because it was such a revolution. The benefits of 3D were so easily visible, and they were applicable everywhere. They don’t use the AVO revolution as an example because it was not much of a revolution. It was slow to be accepted, and may still not be accepted by everyone. Undoubtedly AVO analysis is more successful in some cases than others, and that is partly dependent on data quality. The growth of multicomponent seismic is more like the AVO revolution than 3D. Multicomponent seismic is undoubtedly more applicable in some areas than in others. It’s most popular use right now is for heavy oil plays. It may not ever be applicable for every type of play, and like most technologies, it cannot be expected to work 100% of the time. But it does seem to have gained a firm foothold in the Canadian market and it is not going to go away. I am quite excited about some of the results I have seen with time-lapse multicomponent data recently, especially the effects on shear-wave splitting. We will gradually learn what multicomponent data can tell us if we just keep listening to it.
What are the directions in which future R & D worldwide will be focused in our industry? Any important developments that we will see in the next five years?
Gee, that’s a big question. What I expect is that the type of projects that people are doing now that are in the form of science projects will become more and more mainstream. More and more time-lapse, multicomponent, AVO and AVAZ analysis, anisotropy analysis, perhaps Q measurements. Interest in heavy oil areas will grow despite all the environmental concerns because the activists have to drive cars and heat their homes just like the rest of us. But environmental concerns will not go away either and oil companies will respond in a responsible way because people who work in oil companies care about the environment too. I expect that all the things that we are starting to do now, like CO2 sequestration and microseismic monitoring will just grow.
You have been associated with teaching at the U of C as well as teaching courses. Tell us about that. Do you teach any courses at present?
I have occasionally taught courses at the university or for Doodletrain but it has not been a regular thing that I have done. The latest course I taught was a one-day course on multicomponent seismic in Brazil last summer. Putting together a course and teaching it well is a lot of work, and my work schedule has not allowed for much of that.
You have been actively volunteering your time with the professional societies like the CSEG and the SEG. You were the CSEG President during 2004- 2005. Could you tell us why you do this and what the benefits are?
All of us benefit from these professional societies, especially people like me who give talks at the meetings, publish in the magazines, teach courses, etc. The least that I can do is to give some time and effort back to the society that has done so much for me to try to keep the societies running for the benefit of all of us. Fortunately the CSEG and SEG have many people who feel the same way as I do, so the societies thrive on volunteer contributions. As you well know, serving on the various societies in these organizations exposes you to the many splendid individuals who are in these societies. I don’t think that we always recognize how fortunate we are to work in this type of highly-educated, respectful and pleasant environment.
What other interests do you have? I notice you are active in yoga, skiing, canoeing, etc.
Yes, outdoor activities like canoeing and skiing, camping and hiking have always been a part of my life since doing them with my family while growing up. They have probably shaped me into the type of person I am more than I am aware. For the last 10 years or so I have been doing yoga, and I’m even doing a bit of teaching of yoga lately.
What would be your message for young geophysicists entering our profession?
The most important aspects of your geophysical career will likely have little or nothing to do with geophysics. Try always to be respectful towards the people you work with, especially when you feel that you are not being respected. Work hard, but not so hard that you lose sight of what is most valuable. Progress comes when you are humble enough to admit your mistakes and apologize for them. Those are some of the things that I’d probably say.
Is there anything else you’d like to say?
I’d like to thank you, Satinder, for conducting this interview, and for all the many hours of work that you put into making The RECORDER such a topnotch publication.