David Thomson, Lauréat du prix pour impact de la SSC 2013
The 2013 recipient of the Statistical Society of Canada Award for Impact of Applied and Collaborative Work is Professor David J. Thomson, Professor in the Department of Mathematics and Statistics at Queen’s University, Kingston. The award recognizes outstanding contributions by members of the SSC in collaborative research and applied work, the importance of which derives primarily from its relatively recent impact on a subject area outside of the statistical sciences, on an area of application, or on an organization.
David was born on 1942 in St. John, New Brunswick. He attended Acadia University in Wolfville, obtaining a Bachelor of Science (Honours) in Mathematics and Physics in 1965. Soon after, he moved south to New Jersey, having been hired by Bell Telephone Laboratories into a development division, and to take part in their Graduate Study Program. Accordingly, through this program David completed an MS (1967) and a PhD (1971) in Electrical Engineering, both degrees granted by Brooklyn Polytechnic Institute (now part of NYU), while completing a set of internal courses on topics so new that textbooks did not yet exist.
When David originally joined Bell Labs, he was placed in the “Outside Plant” division, which was concerned with elements of telecommunications networks that are placed outside. This included long-line transmission systems, and from 1966 to 1977, while simultaneously working on his graduate degrees, David was involved with the WT4 Millimeter Waveguide Project, a work that was seminal in his eventual discovery and publication of the multitaper method of spectrum estimation. Following the conclusion of the WT4 project, David spent 1977 through 1984 on a team involved with the development of the first commercial cellular phone systems. As part of this work, David was awarded five of his eventual 27 patents. During this time, he was also promoted to Distinguished Member of Technical Staff, at the time the highest possible title at Bell Labs for a scientist or engineer.
In parallel with his work on cellular phone systems, David found time to write his best-known statistical work, the 1982 paper entitled Spectrum Estimation and Harmonic Analysis, published in the Proceedings of the IEEE. This paper has received over 1900 external citations, and has been highly influential in the related fields of geophysics, seismology, helioseismology and sig-nal processing. The key accomplishment of this paper was to explain why multiple data tapers are needed for spectrum estimation. This work gave a consistent estimator for power spectra and practical tools for harmonic analysis. Multitaper anal-ysis converted spectrum estimation from something in the realm of “black magic” to a rigorous scientific tool, correcting problems that had been known as far back as Lord Rayleigh and Sir F. Arthur Schuster in the late 1800s.
In 1983 David moved to the Mathematics of Communications Department within the mathematics research area of Bell Labs. He was fortunate enough to have a position that allowed tremendous flexibility in his choice of research, and from 1984 to 2000 David worked on projects in the areas of digital signal processing, robust statistics, electrical engineering circuit design, phase tracking and time delay problems, communications satellite failure analysis, magnetotellurics, and externally (to Bell Labs) inspired fields of seismology, paleoclimatology, gravitational lensing (astrophysics) and medicine.
David was recruited by Queen’s University in 2001, and since then has been a Professor and Canada Research Chair (Tier 1) in Statistics and Signal Processing in the Department of Mathematics and Statistics. In his 11 years with Queen’s, he has supervised 10 MSc and eight PhD students, and has seven additional students at various stages of their studies. Through his career at Bell Labs and Queen’s University, David authored or co-authored 130 peer-reviewed journal papers or conference proceedings, wrote over 300 technical reports, has been granted 27 patents, contributed 15 book chapters, and has received many thousands of citations.
David has also received a number of honours in his long career. He was a Green Scholar at the Scripps Institution of Oceanography (UCSD, 1983), a Steinbeck Visiting Scholar at the Wood’s Hole Oceanographic Institute (1985), a Visiting Professor at Princeton (Statistics, 1993), an Adjunct Professor at Scripps (UCSD, 1993–1998), a Visiting Professor at Stanford (Statistics, 1996) and a Houghton Lecturer (MIT, 1996). He has also served as Associate Editor for IEEE Transactions on Information Theory (1984–1990), Chair for the Signals and Systems panel of the Commission C of URSI (1991–1999), Associate Editor for Radio Science (1996–2000) and was on the Panel on Physical Sciences (National Research Council, US Army Research Labs, 1997–2000). He has been elected as a Fellow of the IEEE (1991), as a Chartered Statistician (Royal Statistical Society, 1993), a Killam Fellow (Canada Council for the Arts, 2009–2011), a Fellow of the Royal Society of Canada (2010), and is a Professional Engineer (PEO). Since 2008 he has also been a member of the Space and Atmospheric Advisory Committee (SAEAC) for the Canadian Space Agency.
The Impact Award was given to David both for his creation of the multitaper spectrum estimator and its subsequent application of the associated theory to a multitude of scientific and technological problems. The technological applications of the multitaper method are myriad, including applications in Nuclear Test Ban Treaty verification, and ensuring accuracy of touch-tone phone signals against noisy backgrounds (for which several patents were granted). In science, David has contributed in significant ways to climatology, geophysics, and space and solar physics through detailed statistical analysis of large-scale time series data sets. His 1990 Nature and 1995 Science papers were the first statistical analyses of climate data that conclusively showed a relationship between global temperature and atmospheric CO2, and have been key contributors to further discussions about global warming.
Since 1990 David has tirelessly pursued a two-decade-long work exploring data series from a number of spacecraft (especially, but not limited to, the Ulysses project, on which he was Co-Investigator) and has had a profound impact on our understanding of interplanetary magnetic fields and the Sun. His recent discovery of the existence in numerous scientific data sets of signatures of buoyancy (g-) modes that are theorized to exist in the Sun is contributing to a strong change in scientific understanding of the structure of the Sun, its magnetic fields, and their relationship to our climate and environment on Earth.
Finally, David also contributed to robustness and improved error bounds in the magnetotelluric method, a geophysical technique that measures magnetic and electric fields on the Earth’s surface (especially the seafloor) and uses these measurements to determine electrical conductivity. From a statistical point of view, this problem can be viewed as regression with complex predictors and response, with the predictors being highly collinear and both the predictors and response being often contaminated by outliers. The magnetotelluric method has become standard in the field of geophysics, and is an important tool for scientists working to understand the structure of the sub-surface of the Earth.
David currently lives in Kingston, Ontario with his wife, Maja-Lisa, and their two cats, Loki and Selkit. They have two grown children and one grandson, and in his spare time David enjoys woodworking, reading about history and attending classical music performances. His love of the last, especially chamber music, produces both relaxation and inspires his creativity. David will deliver the SSC Impact Award Address at the 2014 Annual Meeting of the SSC.