Ross Hill is receiving the Kauffman Gold Medal for his work in developing, implementing, and promoting the process of Gaussian beam migration imaging in numerous difficult geologic areas around the world. In both public and private, by conversations, presentations, and papers, he has presented both theory and results of this continuously improving work, providing excellent insight with clear communication and significantly advancing the state of seismic imaging for the profession. Perhaps the most important aspect of his efforts has been the effective manner in which he has compared and contrasted his work with alternative imaging approaches like those using conventional Kirchhoff and wave-equation methods, so that geophysical professionals gain clear insights and can continually make improvements in the art of imaging.
Biography Citation for the SEG Virgil Kauffman Gold Medal Award
Contributed by Robert Langan
It is a privilege to honor my friend and colleague Ross Hill for his contributions to seismic imaging and multiple removal. Ross has exhibited incredible focus, passion, and determination in taking the theory of Gaussian beams and turning it into a practical solution for seismic imaging in complex areas. His original interest in physics was kindled by his high school physics teacher in Metairie, Louisiana, U.S., who, in Ross' words, "saved" him from some unspecified, horrible fate. Ross received a BS at Louisiana State University in electrical engineering and an MS in physics two years later at the University of New Orleans, where he worked with Professor George Ioup. He then worked with Professor Vittorio Celli at the University of Virginia, where his PhD dissertation was Scattering of Waves from Rough Surfaces. Ross was an assistant professor at Virginia when he became intrigued by a Gulf Research ad in Physics Today seeking research physicists interested in scattering wave phenomena.
After joining Gulf in 1980, Ross developed a 2D wave equation-datuming program for areas of rough topography and complex geology. Around this time two papers triggered Ross' long-term fascination in Gaussian beams, which built on his interests in asymptotics and the wave equation. The first was Chapman and Drummond's 1982 paper in the Bulletin of the Seismological Society of America on Maslov asymptotic theory, and the second was Cerveny's 1983 publication on Gaussian beams in the Geophysical Journal of the Royal Astronomical Society.
In 1984 Ross and colleague Alan Levander left the turmoil at Gulf to join the faculty at Rice University. In addition to teaching, setting up a geophysics program, and working on a recalcitrant Elexi computer, Ross found time to work on wave-equation refraction imaging. This period may have marked the beginning of Ross' unusual ability to work at the cutting edge of computer hardware and software, and to apply both to geophysical problems. Ross continued to learn about Gaussian beams and became intrigued by their potential to solve the multivalued traveltime problem in traditional Kirchhoff migration.
In 1986 Ross joined Chevron Oil Field Research in La Habra, California and began incorporating Gaussian beams into seismic migration. His first success occurred in early 1987, and he soon had an effective 2D poststack depth migration (PostSDM) algorithm, which led to his 1990 paper in GEOPHYSICS. By 1991 he had an effective 2D prestack (PreSDM) algorithm that handled turning rays, but the general approach was too inefficient for 3D PreSDM. At Rice, Ross had taught from a draft of Oz Yilmaz's book, and he was drawn to the sections on predictive deconvolution. In 1991 it occurred to Ross that predictive deconvolution applied to Gaussian beams might remove long-period multiples. Because subsalt demultiple problems are difficult to solve in 2D, Ross had an additional incentive to solve the computational and technical hurdles for 3D beams, and he soon found solutions.
By 1994 Ross had 3D PreSDM (GBMig) working and by 1997 had 3D Gaussian beam demultiple (GBDecon), with both sharing a common ray tracer. He also had, in collaboration with Mark Meadows, a production strength version of GBMig working for anisotropic media. After porting the migration code to Beowulf clusters in 2000 (with Chuck Mosher, Yue Wang, and Ming Zhao), he found a way to avoid any form of decimation of traces, so that each trace could be migrated from the exact coordinates at which it was collected.
In the last four years there has been continual improvement in the computational efficiency, accuracy (e.g., preservation of amplitude relationships), and interactivity of both GBMig and GBDecon, and these algorithms have led to large-scale, routine use by ChevronTexaco around the globe.
Beginning in 2000, Ross committed his work on 3D migration and demultiple to the public domain in a series of talks at EAGE and SEG workshops, regional society meetings, and a 2001 paper in GEOPHYSICS. In talking about his work, Ross has always been very modest and careful not to overstate or oversell, yet he has always been eager to share his knowledge, and indeed is most in his element when discussing wave-theoretical concepts with colleagues.
- Hill, N. R. (2001) Prestack Gaussian-beam depth migration, GEOPHYSICS 66(4):1240.