Realistic Simulations Reveal Something New Under the Sun
You might think that in today’s world of scientific discovery and technology breakthroughs, modern explorers of the universe would have learned everything there is to know under the Sun. Not even close.
From Earth, our home star may look permanent and unchanging, but the bright yellow orb hides some of the most turbulent environments known to exist. With storms larger than our entire planet and extremely powerful magnetic fields, the Sun’s behavior can be very difficult to explain. But with advances in computing power, scientists are closer than ever before to understanding some of the most elusive phenomena that occur on the Sun and being able to accurately model and predict solar events.
Looking at these cool visualizations sometimes feels like looking directly inside of a star. . .
Andre Stejko, New Jersey Institute of Technology
Recently, researchers producing realistic simulations on the Pleiades supercomputer, located at the NASA Advanced Supercomputing (NAS) facility, made a discovery that has changed the understanding of how the ever-evolving solar magnetic field is generated and stored deep inside the Sun before it emerges on the surface, forms sunspots, and releases its giant flares. These powerful eruptions on the Sun’s surface can disrupt electronic systems on Earth such as GPS, satellites, radio communications, and electrical power grids, and create space weather than can potentially change weather patterns on Earth. From the Sun’s interior, the magnetic field extends into the corona, forming the plume-like structures observed during total solar eclipses.
"It’s exciting to discover something new," said Andrey Stejko, a Ph.D. candidate at New Jersey Institute of Technology (NJIT). "The Sun is one of the most powerful objects in our solar system, and we need to understand its behavior in order to predict how it acts—especially when dealing with solar flares."
Matching Up with Real-World Observations
Using a 3D global magnetohydrodynamics code called EULAG-MHD, Stejko and his advisor Alexander Kosovichev, director at the Center for Computational Heliophysics at NJIT, in collaboration with Gustavo Guerrero, Universidade Federal de Minas Gerais, Brazil, found that the tachocline—the transition area between the radiative and convective layers of the Sun, located about 120,000 miles beneath the solar surface—plays a much larger role in the evolution of the magnetic field than scientists previously thought.
"We found that the magnetic field generated by the solar dynamo is amplified and stored in the tachocline," said Kosovichev. "This explains something we have been seeking to understand: why the duration of the full magnetic cycles is 22 years, and not shorter. These magnetic cycles result in 11-year cycles of sunspots and flare activity."
The importance of the tachocline for the solar cycles was suggested by astrophysicist Eugene Parker soon after its discovery in 1992. "Our simulations have provided the first convincing evidence for this elusive phenomenon, and opened new opportunities for deeper understanding of the mechanism of the solar cycle that has fascinated scientists for more than four centuries," said Kosovichev.
EULAG-MHD is a numerical model for studying the motion of naturally occurring fluid flows—in this case, massive flows of plasma in and around the Sun’s convection zone, where most of the turbulent motion and plasma flow occurs. The NJIT team employed EULAG-MHD to create a naturally developing magnetic field that interacts with charged plasma. The model produced evolving cyclical patterns of solar activity that matched with real-world observations of the solar cycle duration—a huge achievement for the team.
Catching Details Through Supercomputing & Visualization
To simulate the motion of plasma on the Sun, Stejko and Kosovichev ran a complex set of calculations on up to 512 parallel nodes (8 cores per processor and 16 cores per node) of Pleiades, simulating 100 years of solar evolution in just 3 weeks—a process that would have taken a decade on regular computers. Higher resolution simulations that were used for hyperwall visualizations took 10 times longer and needed hundreds of gigabytes of memory to store in NAS’s mass storage system.
The NAS Data Analysis and Visualization team used this data to produce high-resolution videos and images that give the NJIT researchers an incredibly detailed look at how turbulence inside the Sun can create magnetic structures. The visualizations helped the researchers track the magnetic field over several centuries, and improved their understanding of where and how it evolves in the Sun.
"Looking at these cool visualizations sometimes feels like looking directly inside of a star, which is something that’s never been possible before," Stejko said.
The high-resolution videos and images are being used to further improve the NJIT team’s numerical methods, in order to generate realistic models that can be used to predict the effects of space weather on Earth. However, many puzzles remain to be solved about what happens inside the Sun’s deep interior. Stejko says the team is now applying what they have learned to build a new model that will give even more precise results and link the modeled interior flows to measurements of the actual flows obtained from NASA’s Solar Dynamics Observatory, with the ultimate goal of developing a model that will work exactly like the Sun.
—Jill Dunbar, NASA Ames Research Center
The NJIT researchers collaborate with the NAS Division’s Heliophysics Modeling and Simulation (HMS) team, which develops high-fidelity modeling and simulation tools that enables research on the interiors and atmospheres of the Sun and other stars.