Obituaries: Charles Lawson

By Fred Krogh

Charles Lawson, who worked at the Jet Propulsion Laboratory (JPL) at Caltech from 1960 until his retirement in 1996, died on July 2, 2015, in Laguna Woods, California.

Chuck was born in 1931 in Weiser, Idaho. He went to the University of California at Berkeley for his undergraduate degree, a BS in optometry. He then joined the U.S. Army, after which—thankfully—he enrolled at UCLA, receiving an MSc in mathematics under Peter Henrici and, in 1961, a Ph.D. under Theodore Motzkin. While at UCLA he had experience with the SWAC (the Standards Western Automatic Computer), built (completed in 1950) by the West Coast branch of the National Bureau of Standards.

Chuck began to support the use of computers at JPL shortly before receiving his PhD. In 1968, when I joined his group, it was clear to people all over the lab that Chuck was the person to see about issues connected with the use of computers, especially with respect to numerical mathematics. It was interesting that people would often come in with one problem and leave with a solution to a different problem—the problem they should have been trying to solve in the first place. If you are involved in consulting, this is a good thing to keep in mind. Of course it also helps if you have a good background, a clear understanding of the issues involved, and an ability to explain things in understandable terms. Chuck had all of these qualities to a high degree, together with a very agreeable personality.

When I joined Chuck’s group, our primary charter was to do research in numerical mathematics, but all of us were interested in creating software that could be used by the projects. Although the funding from NASA headquarters was primarily for research in computational mathematics, our work in developing mathematical software was tolerated, and the usefulness of that work later became key to continued support for the group. Chuck did some early work on computer approximation, which together with contributions of others, appeared in the book Computer Approximations. He was responsible for the early use of orthogonal transformations for matrix computations at JPL, work that led to another book, Solving Least Squares Problems, co-authored with Richard Hanson. That book is now part of the SIAM Classics in Applied Mathematics series.

Early in the 1970s, I had occasion to look at the ugly assembly code generated by our UNIVAC Fortran compiler. That motivated me to write assembler routines for what we would now call ddot and daxpy, which showed that significant speedups could be achieved in this way. I had thought that this would be more useful for the software libraries we were developing at JPL. Chuck had a much better idea – he expanded the functionality and got a number of other people involved. The result was the first BLAS paper, which seems to have spawned a minor industry, spearheaded largely by Jack Dongarra. That later work was followed by more significant gains than possible with just the simple vector operations in the original paper.

It was Chuck who decided that we should develop a library of mathematical software for use at JPL. Work he did in 1976 on C1 interpolation to scattered data in the plane was included in the MATH77 library in 1991. Later, he took the lead in making that work available in C as well. These libraries, MATH77 and mathc90, are still in use at JPL.

In later years, with a change in management came the requirement that a charge number be attached to all work. An engineer or scientist, instead of just dropping by to get some help, had to ask a supervisor for a charge number to cover our group’s time. After the new procedure was adopted, our group was much less effective, and funding gradually dried up. In what I regard as an act of extraordinary generosity, Chuck started to take on other jobs at JPL, in effect giving me the position he had created.

Chuck was a competitive swimmer in college, and his later interests were varied and intense. He enjoyed folk dancing to the point of teaching classes in it, and in his later years the ukulele was his passion. He is survived by his wife Dottie, children Michael, Brian, Melanie, and Marcella, and numerous grandchildren.

I feel blessed to have had Chuck as a boss and as a friend.

The author acknowledges helpful input from Brian Lawson, Richard Hanson, Van Snyder, and Barbara Horner-Miller.

Fred Krogh ([email protected]) is president of Math à la Carte, prior to which he worked in the Jet Propulsion Laboratory at the California Institute of Technology.