UT researchers using AI to find 'polluted' white dwarf stars

Astronomers at the University of Texas at Austin are using artificial intelligence to find and identify white dwarf stars.

Researchers say these stars are actually gobbling up planets outside our solar system. The team at UT Austin published their findings in July in the Astrophysical Journal.

Malia Kao, fourth-year PhD student and team leader, spoke with FOX 7 Austin's John Krinjak about this cutting-edge research

JOHN KRINJAK: Talk to us a little bit about this discovery. What are polluted white dwarfs and how many of them did you find?

MALIA KAO: So a white dwarf is the remnant core of dead stars, like the sun. And about 95% of all stars in the universe will become white dwarfs at the end of their lives. And so if a planet or an asteroid gets too close to the white dwarf, it'll be torn apart by tidal forces. And so, since the white dwarf's atmosphere becomes polluted by these metals from the infalling debris, that's why we call them polluted white dwarfs.

JOHN KRINJAK: You see that the planets are essentially gobbled up by these stars. How does that happen?

MALIA KAO: It's a lot like how Saturn's, Saturn's rings are. So if you didn't know, Saturn's rings are eventually going to disappear because the rings are constantly feeding the planet. It's the same kind of idea with the pleated white dwarf. So you end up with, like, this ring of planetary debris around the white dwarf. And so that debris disk, like kind of like that ring of, you know, the dead planet is constantly feeding the white dwarf in that sense.

JOHN KRINJAK: And so, yeah, talk about this discovery. How many of these stars did you find?

MALIA KAO: Yeah. So we found 375 of these new polluted white dwarfs out of 100,000 indigo space telescopes today. And so what's special about the polluted white dwarfs that we've found is that they contain different types of metals in their atmospheres. And these specific ones, we call them heavily polluted white dwarfs. They're very rare because only a few dozen have been discovered up until now.

MORE FOX 7 FOCUS

JOHN KRINJAK: And so you're talking about hundreds of stars here. How did these researchers from UT actually find and identify all these? I understand AI was a big part of the process. Can you talk about that a little bit?

MALIA KAO: Yeah, yeah. So given the large volume of stars, it would be nearly impossible for us to identify the polluted white dwarfs by eye. That would take probably a really long time. And so that's why we employ AI to do this for us. And so we used an unsupervised machine learning technique called U map. It creates a very readable 2D map for us, where the white dwarfs with similar atmospheric features are clumped together, and those with dissimilar features are kind of separated out.

JOHN KRINJAK: From this information that you're learning about the white dwarfs, you can actually learn about planets as well, right? Planets outside our solar system. Can you explain how does that work? How do you learn about those?

MALIA KAO: So one of the main questions in astronomy is whether our solar system is unique in sustaining life, and what the polluted white dwarfs can tell us is how similar other planetary systems are to our own solar system. And so if other planetary systems evolve similarly, similarly to our own, the possibility of life existing beyond our souls may become more concrete.

MORE AIR AND SPACE NEWS

JOHN KRINJAK: How exciting is it for you to be part of this research and knowing that all this is happening, you know, with people right here in Austin?

MALIA KAO: Yeah. It's honestly a dream come true. I mean, I never thought that I would be in this position. So. Yeah, I'm completely stoked. Okay. Yeah. This is amazing.

JOHN KRINJAK: All right. Malia Kao, a fourth year PhD student in the astronomy department at UT. Malia, thanks so much for being here and explaining all this to us. We appreciate it.

MALIA KAO: Yeah. Of course. 

FOX 7 FocusUT AustinAir and SpaceArtificial Intelligence