Mysteries of the sun: Fargo scientist researches relationship of sunspots and solar flares

FARGO -- Cherish Bauer-Reich is grappling with an enormous research subject that is both too hot and too far away to touch. The Fargo woman is trying to better understand the relationship between sunspots and solar flares. Unraveling the mysterie...

This extreme ultraviolet wavelength image provided by NASA shows an undated solar flare. Fargo's Cherish Bauer-Reich is researching the relationship between sunspots and solar flares in hopes of finding ways to predict future electromagnetic storms.

FARGO -- Cherish Bauer-Reich is grappling with an enormous research subject that is both too hot and too far away to touch.

The Fargo woman is trying to better understand the relationship between sunspots and solar flares.

Unraveling the mysteries of solar activity is important for anyone who talks on a cellphone or relies on GPS -- global positioning system -- navigation.

That's because the electromagnetic energy emitted from solar flares disturb the Earth's magnetosphere, sometimes wreaking havoc with communications and, in rare instances, the power grid.

Knowing about electromagnetic storms in advance can be as important as predicting blizzards.


That's especially the case, for instance, if you're flying in an airplane and abruptly lose communications or navigation.

Ideally, scientists directly observe their research subjects. But when that isn't practical, they turn to sophisticated simulations.

"You can't really make a lab of the sun," said Bauer-Reich, who is doing doctoral research in geophysics. "The only way you can do it is with a computer model."

As it turns out, a very big computer model.

Her solar model is a two-dimensional grid representing a strip of the sun comprised by 1.5 million data points in a computer model.

Since Bauer-Reich will run the program through 200 million time sequences, each perhaps 10 seconds long, that will require a lot of computing power.

Enter the supercomputing center at North Dakota State University, now called the Center for Computationally Assisted Science and Technology.

Bauer-Reich, a graduate of Fargo North High School, will use NDSU's high-performance computer cluster to run her model. She's still tweaking the program.


Building computer models is nothing new to Bauer-Reich, who works part time as a research engineer at NDSU's Center for Nanoscale Science and Engineering.

She represents a new breed of research scientists who combine expertise in science with sophisticated computer programming skills. She earned an undergraduate degree in physics and a master's degree in electrical engineering at NDSU, but will earn a doctorate at the University of Minnesota.

Dual expertise is increasingly demanded in research, said Martin Ossowski, a computer scientist who directs NDSU's supercomputing center.

That's because today's high-performance computers use "massively parallel" processors to break down and analyze huge quantities of data.

NDSU's computer clusters, for instance, connect 1,300 processors arrayed on racks -- a distributed mechanical brain that hums audibly and must be air-conditioned so it doesn't overheat.

Writing computer code for that type of computing is more involved than the earlier generation of high-performance computers, Ossowski said.

"In the old days, you didn't have to be a very good programmer," he said of the input required by the scientist.

Because of the need for higher-level computing in many areas, scientific research is experiencing a "paradigm shift," in which scientists also will have to become computer programmers or programmers will have to become scientists.


"We have not enough people that are skilled in hard sciences and at the same time computer programming," he said.

NDSU and other research centers will have to recruit a new generation of researchers when sophisticated computing is required for their work, Ossowski adds.

"Cherish is a great example of a scientist who also has skills in programming," he said.

Another approach is to develop multidisciplinary research teams. "It's starting to happen here," Ossowski adds. Two computational scientists recently were recruited to help develop new materials for application in energy.

Ossowski and the others at NDSU's Center for Computationally Assisted Science and Technology are in talks with a number of companies to help with projects.

Supercomputer models help save money when new products are developed, allowing designers to zero in on the most promising possibilities by using virtual prototypes.

The pharmaceuticals industry, for example, has long recognized the value of using supercomputers to develop new drugs, Ossowski said.

Similarly, medical research involving the human genome requires massive computing power for gene sequencing. Computing power, in fact, is a major cost associated with that promising field of medicine, Ossowski said.


"That's how important computation is," he said.

NDSU's supercomputers are of the same caliber as those in leading research centers. "The technology that we are using is top-notch," Ossowski said.

The computing challenges will become even more daunting under an ambitious initiative pushed by the U.S. Department of Energy to increase supercomputing performance by 1,000 times by 2020.

The software required to run those machines will be commensurately sophisticated, Ossowski said.

"This is a huge challenge," he said. "The rate of change is just mindboggling."

As for Bauer-Reich, she plans to begin running her solar model soon. She hopes to get preliminary data in the next couple of weeks.

Springer is a reporter for The Forum of Fargo-Moorhead, which is owned by Forum Communications Co.

What To Read Next
Get Local