Improving Analysis of Astronomy Data
Editor's Note: This story accompanies the Science Journal feature article The Disks that Make Solar Systems Flat
To understand the dynamics of the disks of gas and dust swirling around stars, Assistant Professor of Astronomy and Astrophysics Ian Czekala has turned to the Atacama Large Millimeter/submillimeter Array (ALMA) in Chile. Not unlike the radio telescope featured in the movie Contact, ALMA uses a network of interconnected antennas to improve the resolution of its images.
“A larger diameter telescope typically yields better spatial resolution,” he said. “ALMA has 66 antennas spaced up to 15 kilometers apart, effectively mimicking a telescope with a 15-kilometer diameter. We have to use some complicated math to help fill in the gaps, but these interferometric techniques have allowed us to get a very high-resolution look at the young planetary systems that I study that we had not been able to achieve before. It has literally changed the picture.”
Czekala also studies protoplanetary disks in systems with multiple stars. Though it sounds counterintuitive, he suggests that by complicating the picture by introducing another gravitating body, you can sometimes get a better handle on the more difficult-to-study processes of star and planet formation.
“Binary stars can really throw a wrench into the system in some interesting ways,” he said. “If we’re seeing things that theories would predict should never happen, we have to go back and reexamine our understanding of the processes.”
Czekala has also developed tools to disentangle the light emitted from each of the stars in a multiple-star system so that astronomers can study the stellar orbits. Perhaps unsurprisingly, he is interested in improving methods of data analysis for these and other astronomy-related data.
“Some of my work is dedicated to developing new algorithms to push the level of inference we can get from large and multifaced datasets,” he said.
Czekala, who joined Penn State in August 2020, is eager to take advantage of the collaborations and supercomputing resources at Penn State and its Institute for Computational and Data Sciences (ICDS), where he is a Faculty Fellow.
“It may seem like a specialized problem, but a lot of the image processing questions we are facing crop up in many different fields,” he said. “Ultimately we’re being pushed to use more and more computational power to try to make better images with the data we have. There’s a lot of natural collaborations that I’m looking forward to starting within ICDS.”