Combining IT and science to monitor the weather
“With the ice storms on the East Coast, we’ve seen a real increase in people using our eWall,” he says.
The eWall, or electronic wall, was developed by Fred Gadomski, well known "Weather World" host and senior lecturer in meteorology, as a way for students, forecasters and weather enthusiasts to have a one-stop online location for current weather information.
To the novice, the eWall’s home page is a dizzying digital display of numbers and computerized weather data from various countries, including the United States. But when used as a utility page, the eWall is a major resource for meteorologists around the world for viewing computer-simulated models of the weather.
Pavloski, director of meteorological computing and research associate in Penn State’s Department of Meteorology, spends most of his day monitoring the eWall as well as a plethora of servers, switches, applications and services for which his IT team is responsible. A quick glance at his status screen lets him know immediately of any service outages so he can rectify the issue before faculty and others experience any problems.
“This is how fast the eWall is continuously getting hit from different forecasting sites. If it shuts down for more than two minutes, I can guarantee you I’ll get phone calls,” he says as he points to a steady stream of Internet traffic flowing across his monitor.
The department's eWall is a major resource for meteorologists around the world for viewing computer-simulated models of the weather.
As director of meteorological computing, Pavloski’s job of keeping things running smoothly is his No. 1 priority. In addition to his knowledge of IT, Pavloski received his Ph.D. in meteorology from Penn State and he also is well-versed in the science of weather. His office shelves are adorned with books about meteorology, physics, calculus, quantum theory and computer science. The books huddle with several weather measurement gadgets, including a deflated weather balloon Pavloski used during his graduate study of atmospheric remote sensing.
Throughout his 21-year career at Penn State, Pavloski has taken part in numerous scientific field experiments, taught meteorology courses and is currently involved in the SCRiM initiative, a National Science Foundation-sponsored team of research scholars focused on sustainable climate risk management strategies. As a result, he understands the needs of the department, its students and faculty members on both an IT and scientific level.
Pavloski was always interested in weather and refers to himself as a full-blown “weather weenie.”
“There’s a lot of science involved in meteorology, so I always joke around and say I’m a scientist gone bad,” he says. His interest in meteorological scientific research led him to his work today, away from the bright lights of television where only about 10 percent of meteorology graduates end up. “I also joke around and say I have a face made for radio,” he adds.
While constantly dabbling in new technologies for the classroom, Pavloski is more of a “weather wizard” and often has a hand in the faculty or graduate student research projects taking place in the department to make sure they have the necessary IT resources for their research.
“If our faculty have an initiative where they’re going to be collecting or generating 50 terabytes of climate data, then I’ll go over their projects with them so they can get their models running scientifically and to make sure they’re computationally sound,” he says.
It’s a big job. Penn State’s Department of Meteorology is one of the oldest and largest meteorology and atmospheric science programs in the country — and is a good fit for a weather weenie like Pavloski.
“Meteorology is an IT-dependent field,” says Pavloski. Forecasters rely on real-time data on temperatures, winds and other observations from radar, satellite and surface weather stations from around the globe, as well as the latest computer models to mold their own predictions. Since Penn State is a tier one distribution site for the U.S., the University receives the collection of data for the eWall instantaneously 24 hours a day — far beyond the weather depictions available via the latest smartphone weather app.
“Students who go through the meteorology program at Penn State are basically getting the equivalent of an atmospheric engineering degree,” says Pavloski. “It’s not just about getting on TV and pointing to maps on a green screen.” Today’s students are learning math, physics, chemistry, computer programming and fluid dynamics.
Before the use of computer-simulated weather modeling, most college meteorology departments had a map room with walls lined in hard-copy weather maps — typically purple squiggles generated on facsimile paper — a technology that lingered into the 1990s. Today’s technology allows the data to be easily displayed as computer-simulated animations, which show up on the department’s eye-catching digital display that spans more than 20 feet of the Joel N. Myers Weather Center. The data also are analyzed on personal computers that enable students and faculty to develop forecasts and carry out innovative research.
These types of computerized modeling and high-performance science applications generate massive amounts of data, and the department has tapped into Penn State’s new “Big Data” research network, a sub-network on the Internet devoted to large data application areas, to manage it all.
Modelling enables meteorologists to simulate a storm and then look at how variations to the model change the amount of precipitation formed, wind speeds, and other conditions to simulate weather patterns.
For graduate research assistants like Burkely Twiest, this high-performance network will help her conduct research on nighttime temperature and energy balance at the Shale Hills Critical Zone Observatory in Penn State’s Stone Valley Forest, about 10 miles south of the University Park campus. Twiest, who received her bachelor of science degree in meteorology last December, takes 10 wind measurements per second from instruments she leaves in the field for a period of three months and stores the data on the departmental network. “As you can imagine, three months of this sort of information from several different instruments creates quite a lot of data,” she says. “Without the technology, I would be unable to perform my research.”
Twiest, who is also well-versed in computer information systems, knows programming allows meteorologists to model the atmosphere to forecast the weather and also enables researchers to conduct experiments that they would otherwise be unable to perform in the real world. Modelling enables meteorologists to simulate a storm and then look at how variations to the model change the amount of precipitation formed, wind speeds, and other conditions to simulate weather patterns.
“It’s very hard, for example, to have the same storm occur over and over again with differences in wind shear as the only variable,” says Twiest.
Even with all of this technology and high-end computing at hand, Pavloski doesn’t think we’ve reached the outer limits when it comes to forecasting the weather. He sees the development of further improvements in the way meteorological forecasts are delivered with the use of smartphones and other devices, as well as improvements on remote sensing capabilities and satellite technologies that will provide exact measurements.
For now, Pavloski continues to monitor and support the infrastructure that’s already in place. On a break from his computer monitor, he meets Gadomski at the digital map wall where they discuss possible enhancements to the wall including interactivity and the ability to display images that come directly from geographical satellites with resolutions in the tens of millions of pixels.
“The current generation is much more into their personal devices,” says Gadomski as he turns to admire an ultra-high resolution global satellite picture on the digital map wall. “I’m an old-fashioned weather person, and I like to stand here and look directly at the wall. I use it to get a general sense of the large-scale weather pattern. Then if I really want to think about weather and weather forecasting, I will go to my personal workstation and do more detail. But I can stand here, relax, have my eyeballs massaged and learn a little bit about the large-scale weather. And I like that.”
Source: Penn State Univ.