|Robert Serafin. (Photo by Carlye Calvin.)|
Search committee forms for new NCAR directorJoseph Klemp, a senior scientist in NCAR's Mesoscale and Microscale Meteorology Division, is chair of the search committee for the new NCAR director. The other members are David Burridge, director of the European Centre for Medium-Range Weather Forecasts; Edna Comedy, UCAR associate vice president; Michael Knölker, director of NCAR's High Altitude Observatory; UCAR Board of Trustees president Lennard Fisk, University of Michigan; and trustee Paola Malanotte-Rizzoli, Massachusetts Institute of Technology.
The job description and other information about the search may be found on the Web. To apply or to suggest a candidate, contact Steve Dickson at UCAR, 303-497-1674 or firstname.lastname@example.org. All applications must be received by 1 October.
Serafin recently discussed the state of NCAR with the UCAR Quarterly. That discussion is reported below. Serafin also reflected on his decade of leadership in the May issue of Staff Notes Monthly .
UQ: Have NCAR's university interactions changed over the decade that you've been director?
RS: NCAR's interactions with university colleagues have increased, and NCAR is more open to interactions with the universities. We have some prominent examples of that: the Climate System Modeling Program and the MM5 [Pennsylvania State University and NCAR's mesoscale model]. We made a major decision quite a few years ago to make the MM5 open, not to protect it through intellectual property rights. That has resulted in a model that has more than 500 users nationally and internationally.
I think the process for allocating facilities at NCAR is also more open, so, in general, the interactions at all levels with the universities are stronger than they were a decade or two ago.
UQ: How did that come about?
RS: I don't think there was a conscious decision at any one time that we needed to have more or better interactions with the universities, but over the period of the last [NSF/UCAR] cooperative agreement [1993-97] it was evident to just about everybody--management, staff, and also NSF--that the new science initiatives were going to require better coordination. Morever, the large programmatic emphasis at NSF indicated that the community needed to work together better on these initiatives, and NCAR was the logical place to do that. I personally have always felt that we needed to be strongly connected to the universities, and I have emphasized that.
UQ: How have the facilities changed?
RS: The observational facilities over the last half decade have been virtually replaced with next-generation systems. We have revamped the aviation fleet; we have worked with colleagues in France to create ELDORA [the Electra Doppler radar]; we have worked with NOAA and the German space agency to develop the GPS dropsonde; and of course UCAR, with COSMIC, is developing a really exciting global observing system that's going to be important for climate research, weather research, and upper atmosphere research. So I think the observational facilities have made a lot of progress. HIAPER [the High-performance Instrumented Airborne Platform for Environmental Research, a midsized research aircraft] was our highest observing-system priority, and we're hoping that HIAPER will soon be supported.
On the computing side, we kept pace with computing very effectively for quite a few years and were successful in establishing the Climate Simulation Laboratory with support from NSF and USGCRP [the U.S. Global Change Research Program]. We've used creative methods to finance computers to stay at the state of the art.
I think we've made substantial progress in the facilities arena. Right now we need some infusion of funds to increase our supercomputing capability, and we obviously need the one-time funds for HIAPER to acquire that facility.
UQ: What do you see as the biggest upcoming challenges at NCAR?
RS: We're facing the transition to new computing technologies--distributed architectures--which is going to imply changes in the computing paradigms and also the programming or software development that goes along with those paradigms. Making the transition will not be easy for our users, but we're going to have to learn to use these new machines. There is no choice. The alternatives are very clear, and I think we're on the right path. I'm comfortable with the decisions we've made.
It seems that NSF's programs are becoming centralized, NSF-wide initiatives. We have examples of that in fiscal year 2000: the NSF biocomplexity initiative, the IT2 [Information Technology for the 21st Century] initiative, and there seems to be emphasis to build an environmental initiative that will clearly be inter- and multidisciplinary. Another challenge will be to develop a clear policy, working with NSF, on how NCAR is going to participate in those initiatives. I think NSF should be defining generic roles for its national centers--not just NCAR but the supercomputing centers--because the centers can, and should, play a very significant role.
This trend at NSF will be a challenge for the entire atmospheric science community because if the funding is centralized, which it may become, then the place to which proposals are written and the review process may be different from what they have been, and the interdisciplinary nature of proposals is likely to receive higher weight than if proposals are submitted to a disciplinary group within a division.
One of the things we strive for and have been unable to achieve is a greater level of stability. I've always said I'd be happy to get a guarantee from NSF that we'd have a certain size of program for five years--a certain size, not a certain budget--but that seems to be impossible with annual budget cycles and so on. So the budgetary instability always creates some instability in planning. I think we've dealt fairly well with it, though, and the staff within the last decade has learned to cope with this and exploit opportunities and still be faithful to our mission.
UQ: What challenges do you see for our field as a whole?
RS: Our five-year plan is quite visionary. I suggest that people take a look at it. [The plan for NCAR's science and facilities, 1998-2003, can be found in NCAR and UCAR at the Millennium: A Vision for Science, Facilities, Service, and Leadership, available on the Web.]
One of the biggest challenges scientifically and technically is going to be to make effective use of very large data bases and computers through modeling and assimilation. I've heard estimates that over the next two or three years the amount of satellite data, because of next-generation satellites, will increase the total volume of satellite information by a factor of almost 100. Obviously that's a challenge, but it also is an opportunity, because it may offer great opportunities for combining these data with numerical models for much better and higher-resolution descriptions of the earth's atmosphere, oceans, and weather system.
Our high-performance computers also increase the data bases. We just passed 200 terabytes in our mass storage system in NCAR's Scientific Computing Division. A few years ago 100 terabytes was our limit, but new technologies have allowed us to store more. But the read-and-write rate on tapes is not increasing at the same rate as the volume of the mass stores. So although we can store the data, access becomes more time-consuming.
I think even more data are going to be available to scientists in the future. Things like real-time weather radar data from the NEXRAD network will be available in the next half decade, probably through this Internet or the next-generation Internet, and that's going to allow for rather substantial opportunities in research. It'll offer interesting opportunities for graduate students in the universities and of course for their faculty mentors.
UQ: Ten years ago, NCAR's educational programs were all at the college and graduate level. Now we have something to offer down to the grade-school level. How did that change come about?
RS: I helped push that a little bit, but I can't say that it was a pressing initiative on my part. There was a lot of interest several years ago in the Atmospheric Sciences Directorate at NSF to get some projects going on K-12 education, and we were hearing from NSF directors, particularly Neal Lane, that research and education were inseparable. We worked very hard to get Project LEARN started, which was one of NCAR's first formal activities aimed at K-12.
There were many people at the time who weren't sure this was the right thing to do. They felt that NCAR should focus only on graduate education and research, and K-12 could be the responsibility of others. But I think the record's been good and we've accomplished quite a lot, so it was certainly worthwhile doing. The argument I find very persuasive is that even though a center like NCAR cannot solve all of the nation's problems in science education, as a significant funded activity through the National Science Foundation, a national center, we have an obligation to be part of the solution.
UQ: What are your plans after you leave?
RS: I'm not leaving. I'll step down sometime around the first of February, but I'm going to stay on full time for another year. I may take a couple of short sabbaticals; I've been here [at NCAR] over 25 years and I've never taken a sabbatical. I plan to stay active--I serve on a lot of committees, and I chair a few. Currently my plan after retirement is to keep doing the things I have been doing but at a reduced rate.
UQ: So there's no point in asking you what you'll miss, since you'll still be here.
RS: I'll miss the job for sure. It's a great job. It's really hard work; you never stop working--weekends, vacations, whatever; you're always working one way or another. But it's been a great experience. I wouldn't have chosen anything else. One bit of advice I'd give young people is not to back away from responsibility and opportunity.