from Tim Cornwell, August 3 1998
Members of the AIPS++ group were shocked and saddened by the tragic news that Dan Briggs of the NCSA/BIMA AIPS++ was killed in a sky-diving accident on July 4. Dan was well-known to many of us as a colleague and friend. Here I'd like to say something about my memories of his professional career.
I'd personally known Dan for a long time in a number of different roles: as a stilt-walker in parades in Socorro, as a student at New Mexico Tech, as my Ph.D. student in the early 90's, and then as a collaborator and colleague within AIPS++. He came to Socorro as a graduate student in physics after completing degrees in physics and mathematics at Caltech. My first contact with him was to see him walking on stilts in a parade in Socorro. In the late eighties, he started spending time at NRAO, working with Rick Perley and then Craig Walker on different projects. He eventually came to me to ask about doing a PhD on radio- astronomical imaging, something that I readily agreed to once I realized his great range of talents. For his Ph.D. and then later on in his career, he worked as what I would call a technical radio-astronomer, interested more in the techniques of astronomy rather than the results. Dan had an excellent combination of attributes for a researcher: great basic intelligence, varied technical skills, persistence, a judicious amount of perfectionism, and an enormous appetite for hard work. His Ph.D., conducted at New Mexico Tech, and overseen by myself and Jean Eilek, showed the quality of work that he was capable of producing (his thesis is at http://www.aoc.nrao.edu/ftp/dissertations/dbriggs/diss.html).
It covers the image processing of moderately resolved sources observed with radio-interferometric arrays. It's a stunningly complete piece of work, in which each substantive assertion is carefully described, analyzed and demonstrated. It contains two things that deserve to be called by his name: first, the effect whereby deconvolution errors on moderately resolved sources can masquerade as calibration errors, and second, a new form of weighting, which he called robust weighting, whereby resolution and sidelobe level could be traded one against the other, thus obtaining a compromise of improved resolution and signal-to-noise, thus optimizing the use of telescope time. Robust weighting was immediately recognized by others as a very useful technique and has been adopted in many reduction packages. In recognition of Dan's contribution, Bob Hjellming has called for the term Briggs' weighting to be used in place of robust weighting.
After his Ph.D. here in Socorro, Dan moved on to Washington, D.C., to work at the Naval Research Lab as a postdoc, first on the Big Optical Array, and then on low-frequency imaging with the VLA (in collaboration with Namir Kassim, Kurt Weiler and others). In the latter work, Dan developed an existing, limited package for wide- field imaging from VLA B-configuration 327 MHz data into a well- engineered, very capable package for A-configuration data, that could run in a reasonable amount of time on a parallel machine. This exploited many of his skills: understanding of the physics and mathematics, expertise at meeting the complicated computing requirements, and appreciation of the resulting science.
Following his time at NRL, he moved on to work with the NCSA/BIMA group on parallelization of code, and he was just starting to implement wide-field imaging in AIPS++.
In addition to the personal tragedy, his death is a great loss for radio-astronomy. Dan was only getting started on what I'm sure would have been a very interesting and productive career. I know that I'll miss his quick mind, willingness to get involved, and high standards, qualities that also showed up in his personal pursuits. One very typical memory that I have of him professionally is the Friday afternoon meetings on radio- astronomical imaging that we used to have at the Array Operations Center in the early nineties. In a very short time, he could pick up a new idea, reflect on it, fit it into his understanding, and then offer insightful comments, all while leaning laconically against the whiteboard. I miss the to and fro of the discussions in those meetings.
Dan, we'll all miss you.
Tim Cornwell
Back to the Dan Briggs homepage.