Credit: National Weather Service/NOAA
Fittingly, perhaps, a former researcher with expertise in snowstorms is taking the reins of the U.S. National Weather Service (NWS) just as a major
blizzard bears down on New England. Yesterday, the National Oceanic and Atmospheric Administration (NOAA)
announced that 24-year NWS veteran Louis W. Uccellini will replace former director John "Jack" Hayes, who retired in May 2012 amid a controversy over
spending practices at the agency. Uccellini has led NWS's National Centers for Environmental Prediction since 1999, and "is known for coauthoring the widely acclaimed two-volume book, Northeast Snowstorms," a NOAA press release notes. Yesterday, Uccellini spoke to ScienceInsider about upcoming forecasting challenges and
his hopes for the research community. Here is an edited version of the conversation.
Q: What one bit of science would you like to see happen that could help improve weather forecasting?
Just one? I'm only half-joking! I think in terms of numerical modeling, as an example, we're entering an era of interdisciplinary research and modeling. So
it's not just the atmosphere anymore; we have to couple the atmosphere with an ocean prediction system, hydrology, with land models and with the cryosphere
[ice surfaces]. So this interdisciplinary approach is broadening our associations and interactions with the science community in a very important way. And
we have to depend on that increasing interaction to move our prediction capabilities forward.
Q: Is there any development in computer forecasting that would be particularly helpful to NWS?
I think it's known that the European center has the best scores in their model forecasts beyond week one. We know that we need to run our models and our
data assimilation systems at a higher resolution. We're in a process of transitioning our computing code to the new IBM iDataPlex [a supercomputer], which
gives us about a threefold increase in computing capacity. So we're anticipating that we'll be advancing our models.
Q: Can NWS catch up to the European forecasting ability? What would it take to pull even?
Well, I want to emphasize that the European center has a mission of only running their models and supporting other weather services around Europe. The NWS
not only runs its models but it makes weather forecasts out to day 7, and short-range climate forecasts out to seasonal range. We're actually doing quite
well. From a weather service perspective, the local forecast and the warnings and watches that are put out in extreme events are the best in the world
right now. What we need to do is build up our modeling capabilities and make those forecasts even better.
Q: Is there anything that might stall the progress you're making with forecasting?
There are always challenges to keep on moving forward. We all recognize that budgets will be tight. We're concerned about sustaining the satellites, which
are a fundamental backbone for the entire global observing system. There are issues involved in ensuring there are no gaps in the satellite data. … We are
concerned about that, but we're also working hard to take mitigating steps.
Q: What sort of mitigating steps?
We're trying to partner with other satellite agencies around the world to help fill in if those gaps develop in the [operations of U.S. polar orbiting
satellites]. Right now we work with the Europeans, so we already have partnerships in place, but we would be looking towards other countries and other
satellite systems to help. But it's clear that our forecasts will be degraded with the loss of the new high-resolution infrared data and the advanced
microwave data that we know these satellites can bring to bear. The other thing is to continually improve the data assimilation system to take advantage of
data sets like aircraft data. But we still need the polar orbiting satellites to give us the full measure of global observations.
Q: What direction would you like to see the research community and the weather service go to solve problems that can't be fixed by improved computing
ability, like hurricane intensity and tornado forecasting?
What we're attempting to do and what we need to do better is to get university researchers, academic researchers—people who are interested in improving
the forecast—[into] relationships and better partnerships with forecasters in local offices. The weather service now has forecasters and science
operation officers who have bachelor's, master's, and even Ph.D. degrees that are interested in working with the research community. If we can get the
research community working with the forecasters at the local levels, and even support that at a higher level, we'll get a better flow of research results
Q: And is there anything you wish the research community would pay more attention to?
I was in the research community for an extended period of time before I joined the weather service in 1989, and I know that times have changed even in the
24 years since I've been here. The research community has gotten more involved with us from an applied research perspective and really is working with us
to help improve our capabilities. I'm optimistic, actually, that we'll be able to build on existing partnerships and work with the research community to
move the entire weather service forward.