Gilles Hennenfent is the lead author of three papers in Geophysics and five expanded abstracts. He has been a co-author on one additional paper in Geophysics and three other expanded abstracts. Gilles worked on seismic data regularization using the curvelet transform. During his PhD studies, Gilles managed to turn an initial rough idea into a mature technology, addressing theoretical as well as practical challenges that come with introducing a new technology in the field. During his stay at the University of British Columbia, Gilles (co-)authored 6 journal papers, 16 expanded EAGE/SEG/CSEG abstracts, and 21 presentations at meetings ranging from consortium meetings to invited presentations at international conferences, including professional exploration geophysics meetings, and at the American Mathematical Society von Neumann Symposium on Sparse Representations and High-Dimensional Geometry. Seismic data regularization has been an important topic in exploration seismology. His contribution to this problem has been his unique ability to leverage recent and mathematically very technical and challenging developments (known as compressive sensing) to the solution of the seismic interpolation problem. With this theory (which comprises information theory, theoretical signal processing and nonlinear optimization), Gilles has been able to gain fundamental insights and make contributions to four key areas: the introduction of the curvelet transform; the design of a new Nyquist subsampling scheme; the development and employment of largescale sparsity-promoting solvers; and the development of a test suite for sparsity-promoting programs, including the development of an object-oriented programming environment.
Biography Citation for the J. Clarence Karcher Award 2013
Contributed by Felix Herrmann
Who could have imagined that after our first encounter— Gilles, an undergraduate interning at MIT, and a young assistant professor at UBC (myself) visiting a colleague in Brazil—I would have the honor to write the citation for this well-deserved award. Those who know Gilles will not be surprised, but allow me to share the major impact he has made on our field for the rest of you. Gilles got his undergraduate degree from Telecom Physique Strasbourg, formerly known as the Ecole Nationale Superieure de Physique de Strasbourg. After finalizing his degree specializing in optics, Gilles decided he wanted something else and I was fortunate enough to recruit Gilles to come to UBC. In retrospect, the timing could not have been better since I was about to start a research consortium on the use of curvelets and compressive sensing in exploration seismology. Gilles played a key role in turning this ambition into an unequivocal success where the impact of his work on missing-trace interpolation, large-scale optimization, and his general approach to research is still being felt.
Despite the fact that missing-trace interpolation was a well studied problem at the time, I felt that exploring recent insights on sampling and sparsity in applied harmonic analysis (read wavelets) was worthwhile. Gilles was able to take these abstract concepts and leverage them successfully. This resulted in a PhD with fundamental contributions in four key areas described above. These contributions, in conjunction with his pioneering work on reproducible research, are having a lasting impact on the research of my group and on exploration geophysics in general.
Convincing the wider geophysical community to break with the Nyquist-based paradigm in favor of a sampling paradigm that exploits structure in seismic signals has been Gilles’ major accomplishment. He clearly showed that improvements can be achieved if one deliberately breaks this structure during acquisition, by randomizing the sampling, followed by promotion of structure, through sparsity promotion, during the reconstruction. One only has to think of WesternGeco’s coil sampling and ConocoPhilip’s randomized land and marine acquisitions to recognize the impact of Gilles’ work.
After his graduation, Gilles joined the Geophysics R&D group of Chevron in San Ramon, California. He quickly became the curvelet applications expert. He developed several products and workflows, significantly enhancing Chevron’s capability for noise attenuation, signal separation, and multimodality data integration. He directly impacted high-profile projects not only by implementing production-grade tools but also by providing initial data analysis and significant amount of user training, demonstrating thereby his passion for making his work useful to others.
A couple of months ago Gilles started a new challenge. He became a development geologist working unconventional resources in the Wolfcamp group in the Delaware Basin for the MidContinent Business Unit of Chevron in Houston. All in all Gilles is an extraordinarily talented individual who has made major contributions in technical areas and by communicating these findings to a wide audience.
Aside from this unique ability, Gilles is a natural leader with very strong organizational skills. These well-rounded faculties put Gilles on track for a stellar career in our business. It has been a true pleasure to have worked with Gilles over the years and to have become a close friend.