“Understand what you do in depth and from every angle”

An interview with Daniel Trad

Coordinated by: Satinder Chopra | Photos courtesy: Tyler MacFarlane
Daniel Trad

Daniel Trad is an upcoming and promising geophysicist who has contributed generously to the field of seismic data processing through his research efforts in multi-dimensional algorithms for regularization of data. Commonly referred to as 5D interpolation, it is becoming more of a necessity for doing accurate AVO analysis and other attributes studies.

Daniel received a Ph.D. in Geophysics in 2001 from the University of British Columbia, Canada, where he later also pursued post-doctoral work. His mentor and guide there was Dr. Tad Ulrych. Since 2003, he has been working at Veritas, now CGGVeritas, as a research geophysicist and focusing on problems on regularization and noise attenuation of data, amongst others and has published many papers and abstracts in conference proceedings.

As recognition of his many outstanding contributions, Daniel was awarded the CSEG Technical Achievement Award in 2011, which indeed was a well-deserved honour.

The RECORDER approached Daniel for an interview, which he sportingly agreed to. Following are excerpts from the interview.

Daniel, I would like you to speak a little about your educational background and work experience.

I did my undergraduate diploma on geophysics in the west of Argentina with a stronger emphasis on geology than geophysics. My education was missing depth in math and physics so I switched for some time to a very difficult physics program in the south of Argentina, but ended up returning to geophysics after two years. When I finished my undergrad, I joined a government research institution, where I worked on electromagnetic methods for three years. Also, while finishing my undergrad studies, I joined a group from the Observatory National of Rio de Janeiro with whom I worked for five years, spending a couple of months in that city every year. Then I left Argentina to come to Canada to do my Ph.D. at the University of British Columbia on "Applications and Implementations of the Sparse Radon transform". By the end of my Ph.D. and during my postdoc, also at UBC, I taught a course on time series analysis for two semesters. After my postdoc I joined Veritas, which three years later became CGGVeritas. In 2010 I accepted a position at Pau in France, where I worked for the CGGV dedicated center for Total. There I started collaboration with the CGGV research marine acquisition group in Paris which still continues. After almost two years in France I returned to Calgary where I then joined the multicomponent group. That is where I am at now.

After getting your Geophysics degree from Argentina, how did you decide to come to UBC?

I had begun a Ph.D. in Argentina. I felt a bit isolated and limited, and wanted to move to a prominent University in North America. I heard of Tad Ulrych so I asked him by email if there was any possibility for me to study with him. To my surprise he accepted me as student at UBC right away, perhaps because of his positive experience with his two previous Argentinean students, Mauricio Sacchi and Danilo Velis. Thus I changed my topic of research to start fresh in Vancouver. As soon as I landed in Canada, I knew I wanted to stay. Vancouver was a dream come true. I felt at home, a beautiful green country full of trees, ocean and snowy mountains, studying in a university where the only limit for learning was my own mind. I was starving for learning, and at UBC there was no shortage of possibilities.

Since you joined Veritas in 2003, you have essentially stayed there after it became CGGVeritas. What is it that appealed to you so much that you decided to settle down there?

Many things. The merger brought a lot of challenges, as usually happens, but also made the company more diverse and interesting. Besides, I always felt appreciated and well treated. Working with a large international company gave me a lot of opportunities, like being able to work in Europe or communicating freely with top researchers across the world. There are lots of opportunities for learning and growing at a company that large.

Who have been some of your mentors?

Dr. Jandyr Travassos in Brazil and Dr. Tad Ulrych at UBC - both of them really great scientists and friends. Jandyr taught me to think as a scientist and opened my eyes to the possibility of working anywhere in the world. At his side I felt everything was possible. Tad taught me to think deeply on the fundamentals. Often he asked me to explain my work to him from a deeper angle than usual. Most importantly he gave me self-confidence by helping me to feel like an expert on my topic of research. Tad, when I arrived at UBC, suggested I continue the work of his previous student, Mauricio Sacchi, who had just left to become a professor at the University of Alberta. That is how Mauricio and I started a continuous collaboration over the years. We connected really well, partly because we are both from Argentina, but mostly because we looked at problems in a similar manner. Often we independently thought of the same solution to problems, even though we only communicated from time to time on the phone.

You have received many awards for your work. What personal qualities do you think enabled you to achieve the professional status that you enjoy today? Is it hard work, ambition, or anything else? Do you also think you had a firm grounding in mathematical concepts that has kept you in good stead or is it something else?

Mostly commitment and hard work I think. Maybe a combination of having a good optimization and signal processing background, combined with interest in programming and luck! I have a simple strategy that works for me. I spend some time thinking of experiments which can be run quickly so I can test many things in a short time. Also I write my programs with a flexible design so that I can change them easily. When you try many things often you can find good ideas by chance.

Tell me about Daniel, the person- i.e. your habits, your likes/dislikes, etc.

I love the outdoors, spending time with my family, skiing, reading about particle physics and programming, and fiction stories in a language I don't know well, years ago English, today French. I tend to analyze things over and over until all of the pieces fit together, something I inherited from my father. I set goals for the long term, and feel a great satisfaction when I can achieve them. Most of all I like to make progress in whatever I am doing.

Tell us about some of memorable moments in your professional life and also a success story you might want to share with us, if the two are different.

Some key moments were the completion of my Ph.D. at UBC and the CSEG technical achievement award. Most success stories are the result of difficult times. Often breakthroughs come from failed projects. For example, the work on 5D was somewhat suspended because I was busy doing interpolation shot by shot and receiver by receiver which was working fine. One day we had bad results with an important project so I ran a lot of tests which showed me the domain of interpolation was the problem. Therefore I put all the missing pieces to complete the 5D in just a few days. I have many stories like this in my career. Often failure is the motivation we need to innovate successfully.

What has been the most difficult challenge in your professional life?

In terms of topics I have worked on, I think tomography is quite tough. I worked on that topic only for a short time but I came to appreciate how difficult it is because of the diversity of techniques involved and the multitude of parameters to solve. The size of the problem makes testing very time consuming as well. These days I am also finding converted wave multicomponent processing quite challenging. In converted wave everything is connected with everything, time variant conversion points, split velocities, multiple versions of the same data, polarity changes and many more issues.

What would you say about how your career has shaped up so far?

I have been really lucky in many ways. I came to Canada in 1997 with just a couple of suitcases to start my Ph.D. and today I have a good job that I enjoy and a lot of recognition. That has been a long transformation from my first humble job in Argentina. On the negative side, I have been kept very busy over the years with urgent matters that often prevented me from working on things I would have liked to. Many years ago, a friend gave me this advice: "Do not put important things aside to do only what is urgent." Unfortunately I did a lot of that in my career. The good news is that is changing for me now.

Satinder Chopra and Daniel Trad

Tell us about your research contributions to Geophysics, as I believe that is what is close to your heart.

Since the beginning of my career I have been interested in transformations that allow us to represent information in sparse domains, that is, decomposing the data with the minimum number of ingredients. In my efforts on electromagnetic methods I worked on ways to create basis functions that match the noise or the data, mostly using wavelet transforms and wavelet packets. I picked that up from some of David Donoho's papers from Stanford. In seismic, guided by Mauricio's work, I pursued the same goal using Radon transforms but this time in the frame of optimization. Much of the work Tad, Mauricio and I did together is similar to compressive sensing which is now on stage. I did interpolation with the Radon transform for some time at the university but the results were quite limited. Mauricio Sacchi and Bin Liu had the key realization that we needed to look at the data with large multidimensional operators. This required solving many technical issues and my contribution was to make this idea operational. In multidimensional interpolation we capture all the essential information without missing degrees of freedom or independency of the information. In addition, we look at better ways to represent the acquired information. Later I started to look more and more into acquisition issues, realizing that if interpolation can help, better acquisitions help much more.

So what areas of geophysics do you like better than others? Do you plan to focus on research in these areas only or you would like to move into untouched territory as well?

I am interested in many topics like velocity analysis, tomography, migration and acquisition. In the past I spent most of my time on signal processing. For quite a while now I have been working on marine acquisition and multicomponent processing. Now I am trying to focus on Least Squares Migration. Migration has a physical meaning of wave propagation that it is often difficult to have with other topics like signal processing. For example, interpolation is very much an image processing problem; it doesn't matter if the data are waves or pixels. Migration is different. You need to follow the waves to where they originated and compensate for all propagation distortions. In migration of multicomponent data, the physics is even more critical.

We all basically have a driving goal in life. What has it been for you?

I guess my goal has been achieving my own personal aspirations in a balanced and healthy way. I enjoy learning deep scientific concepts and creating tools that allow me to expand my horizons. The more powerful a tool the more of a joy it is to create it or to learn to use it. I have a driving need to complete unfinished projects or to improve on what I have done before. My driving force is to set goals and make progress in achieving them. Also I feel a need to understand more of everything to make more connections across apparently unrelated topics. In applied geophysics, more in the industry than in universities, we have a tendency to add more and more details to make things work, but not too often we have the time and understanding to step back and replace fudge factors by a better physical description of the problems that would make those factors unnecessary.

What are your aspirations for the future?

My aspirations are to never stop learning and, if possible, to make contributions to the field in as many areas as I can. I have had a good start but I would like to cover other topics. I would like by the end of my career to have a wide and deep knowledge of many of the major topics in geophysics. I would like to have less anxiety about the future and take more joy in learning. Further along the future I would like to retire with enough health and mental energy to spend time learning about particle physics. Nowadays physics is changing deeply, and I sometimes regret not having the time to follow this evolution in detail.

Tell us about any particular technical article that caught your eye and had an impact on your thinking?

Many articles had a significant impact on my work, in particular at the beginning of my career. Much of my work was shaped by Mauricio Sacchi, Jon Claerbout, Dave Nichols, Sergei Fomel and others. The best work I ever read is "Signal vs. Processing" by Jon Claerbout. He and his students created a framework to solve many geophysical problems with outstanding simplicity and flexibility. Geophysicists put a large effort towards compensating for the distortions we introduce during processing through often empirical correction factors. If the forward operator contains the right physics, the forward-adjoint formulation automatically calculates most of these corrections into a combined operator which can be solved automatically. Even if in practice this is often not easy to do perfectly, the simplicity and beauty of this idea is stunning.

In 2011 you received the CSEG Technical Achievement Award for 'various contributions to geophysics'. For the benefit of our members, please tell us what those contributions are.

Most of my past work dealt with different aspects of sparse Radon transforms. Two varieties in particular have important applications, apex-shifted and hybrid Radon. The 5D interpolation was the main contribution, and it has many aspects, including geometry design, amplitude preservation, denoising and multicomponent processing. More recently I have made some contribution for deblending of marine data.

What are the directions R & D in our industry is focused? Any particular revolutionary technology around the corner that will help us in a big way?

The industry has realized how important it is to have appropriate sampling, not only on density but also in variety and quality. In land seismic we are seeing a trend towards multicomponent acquisition and processing and also blended acquisitions. In marine the revolution is really fast, moving from narrow streamer to wide azimuth, long offsets, simultaneous sources, nodes, broad band, multicomponent streamers, ocean bottom cables and nodes. This explosion of improved data has brought capabilities to extract information details which were impossible before. The big challenge now is to adapt our processing fast enough to deal with these data. We have more data but more complicated models, multi-component, anisotropic and anelastic. Luckily, the evolution in hardware has made possible brilliant ideas which never saw the light in the past. Full Waveform Inversion is the most obvious example, but many other are around the corner.

What type of research problems are you working on at present?

I am working on several unrelated topics between land and marine: interpolation for converted waves, amplitude preserving denoising, onboard deblending of marine records using migration operators, and I am starting to look at benefits of compressive sensing also for marine acquisition. Also I just starting to look at least squares migration with the feeling that it might serve better than interpolation for complex structures. It seems to me the industry skipped this topic, perhaps because implementations are expensive and not easy to use. Probably also because most people with expertise on this topic went directly to full waveform inversion.

I would like to know your perception about writing; I see you have many papers published in Geophysics and have made many presentations at professional society meetings.

Both presentations and papers are very important for our field. Personally I feel more inclined to give presentations than to write papers because most of the papers I have submitted have taken a long time to be published. Sometimes the review process is slow with two or three iterations, and other times it is just my failure to converge to the final version of a manuscript. For example, I wrote the first version of the geophysics paper on 5D interpolation early in 2006, but it was not published until the end of 2009. In spite of the known success of the technology the paper had two or three rounds of reviews.

What do you look for first when you get your copy of the RECORDER?

Anything that helps me to understand more about the problems we face in geophysics. I look for new breakthroughs and I love to find simple explanations for complicated problems. Perhaps more than in other journals I look to find out what my friends in other companies are doing. The CSEG is a large family and often we know each other even if we work within different organizations.

What are your other interests?

Travelling, exploring new places and the outdoors. I make spending time with my family a priority. I also love to read. I have a large collection of textbooks on theoretical physics to go through for many years to come! Also I spend time reading about programming and other computer-related topics. However, on a sunny day in winter or after a snowfall, I can't think of anything else but skiing.

What would be your message to young entrants who have just taken geophysics as a profession?

Understand what you do in depth and from every angle, and attack the difficulties with persistence while trying everything possible. Carefully plan experiments that help you to understand your problems and check your intuition.

One last question, do you think I missed out on any question, that you expected me to ask and I did not?

What I would do different if I could start again? I would change my research topic more often. I would worry less and trust that everything works out at the end.


Share This Interview