UCAR > Communications > Staff Notes > December 1998 Search


December 1998

Dick Sato, 1943-1998

Dick Sato, a long-time software designer in SCD, died on 26 October after a long illness. Paul Swarztrauber, a colleague and friend of Dick's, provided this tribute, which will also appear in the upcoming issue of SCDzine.

Dick joined SCD in February 1971 as a scientific programmer. Shortly thereafter he became involved in the development of codes for predicting weather and climate, which became the defining aspect of his career. His work began with NCAR's original general circulation model and continued through the third-generation spectral community climate model (CCM3). From 1973 through 1982, Dick led the implementation of NCAR's weather and climate models on virtually all of the great supercomputers of that time. More recently he directed their implementation on massively parallel computers.

In the mid- to late 1980s, supercomputers were making their way into the universities and the commercial sector but were not yet generally available in university curriculae. NSF selected NCAR as the site to conduct several summer institutes for postdocs and junior faculty. Dick supervised the operations of, and later directed, these successful institutes.

In January 1988, the age of the massively parallel processors began at NCAR with a state-of-the-art computer called the Connection Machine. As before, Dick was again the key player for climate-code development, this time on a new machine with a very different architecture than that of previous NCAR computers. This project was considerably expanded with the involvement of Oak Ridge, Argonne, and Los Alamos--national laboratories that were also interested in developing climate models on these new machines. Dick continued his work on the Connection Machine but also advised and assisted in the porting and implementation of the models on advanced computers at the other laboratories. A group met about every three months, and Dick's skills as both an effective lecturer and a problem solver were evident at those meetings.

Dick became a recognized authority in this country as well as abroad. He was invited to present his work on the Connection Machine at meetings in 1990 and again in 1992 at the European Centre for Medium-Range Weather Forecasts in Reading, England. This work appeared in the proceedings of these meetings, and extensions have appeared in a number of other publications. These contributions laid the groundwork for more recent efforts in multiprocessor modeling at NCAR and elsewhere.

Dick had an exceptional career. He was manager of SCD's User Services Section from 1982 to 1985 and the Computational Support Section from 1993 to 1995. Of course, this was only a part of his contribution. He was also an exceptional human being. When describing Dick the most-used word is "gentleman," and probably the second most used word is "modest." Dick was a man of integrity: dignified, thorough, trustworthy, compassionate, and considerate. His office, where Dick was available for technical problem solving or just to provide a sympathetic ear, was always open. One of Dick's greatest strengths was to treat everybody with respect, and in turn he was highly respected by friends and colleagues. He was able to see and articulate alternate points of view. Finally, he showed incredible strength in the face of adversity.

Dick will be missed! •


In this issue...
Other issues of Staff Notes Monthly


UCAR
NCAR
UOP

Edited by Bob Henson, bhenson@ucar.edu

Prepared for the Web by Jacque Marshall