Abstract: The science of waveform inversion was in its infancy in the early 1980’s. While there was a general understanding of the power of inverse methods for characterizing and solving geophysical inverse problems, the task of “Full Waveform Inversion” seemed to many to be theoretically desirable, but practically impossible. At the time, even the forward waveform modelling problem was a complex computational problem that was more commonly solved with approximate ray methods, approximate 1D methods, or approximate 1-way propagators. It was widely believed (incorrectly) that inversion required the explicit computation of a Frechet derivative, with a computational complexity that was out of the question.
Between 1980 and the present day the FWI problem was progressively solved, piece-by-piece, by a group of researchers who collectively contributed innovations. The most significant of these was certainly the development of the Adjoint state method for computing the gradient of the misfit functional by Lailly and Tarantola in the mid 1980s, but it would be another decade before real data inversions were finally accomplished, and another decade before the FWI became a part of the exploration industry. The talk will cover some of this history, but will of necessity be limited to a myopic and personal view from my limited perspective. Of course, the solution is still not "complete" and the richness of the field is testified to by the quality of the contributions to this conference.
Bio: R. Gerhard Pratt is a Professor of Geophysics in the Department of Earth Sciences, at the University of Western Ontario. Gerhard received a B.Sc (1980) from Queen's University in Canada. He received an M.Sc. (1986) and a Ph.D. (1989) from Imperial College, London. Gerhard was a lecturer at Imperial College from 1992 to 1998, and from 1998 to 2008 he was a Professor of Geophysics at Queen's University; he moved to Western University in 2008. He is a past associate editor of GEOPHYSICS. His interests are the nonlinear seismic inverse problem, numerical modeling, imaging, seismic data processing, and fundamental questions of resolution and accuracy. He is a member of SEG, EAGE, and AGU. He is a recipient of the Conrad Schlumberger award (EAGE) and the Virgil Kauffman Gold Medal (SEG).