Interstellar Overdrive

New techniques help astronomers discover new planets faster than ever

The astronauts in Christopher Nolan’s “Interstellar” have the fortuitous appearance of a wormhole to provide a shortcut to potentially Earth-like planets that can sustain the human race as conditions render our planet uninhabitable. Our planet-hunters don’t have that luxury – or Matthew McConaughey to pilot them around the universe – but they have new techniques for identifying potentially habitable worlds. Whether it’s building a better telescope, training machine learning algorithms, or pioneering AI breakthroughs, today’s astronomers are finding planets that might harbor life faster than ever.

The Right Angle

Compared to other methods, Heidi Newberg’s idea might seem low-tech, even simple. But it could prove to be a breakthrough that can “find half of all existing Earth-like planets orbiting sun-like stars within 30 light years in less than three years,” Newberg, a professor of astrophysics at Rensselaer Polytechnic Institute in New York, wrote.

Newberg’s proposal is to use a 1 x 20 meter (65.6 x 3.3 feet) rectangular mirror for a planet-hunting telescope rather than the circular mirror with a diameter of 6.5 m (21.3 ft.) like the James Webb Space Telescope has. The Webb mirror is too small to detect an Earth-size, water-rich planet in the habitable zone of a sun-like star, Space.com reported, and building a version of the mirror large enough to do the job would be too costly.

Caption of Image

Newberg’s rectangular solution would have a larger area than a similarly sized circular mirror and be less expensive and more efficient. Such a telescope would be able to detect light with a wavelength of 10 microns, at which water vapor emits. The mirror would also be able to rotate itself in whatever direction a planet was oriented, using all of its collecting area to image the target planet, something the Webb telescope can’t do. This would make it much easier to survey the areas around the 69 sun-like stars and 300 M dwarfs identified within 32.6 light years of the solar system.

“If there is about one Earth-like planet orbiting the average sun-like star, then we would find around 30 promising planets,” Newberg wrote.

Hunting in a Simulation

Finding planets with a similar mass to Earth located in their stars’ habitable zones used to be up to chance. But by training a machine learning algorithm on simulations of the conditions such planets are found in, astronomers have been able to identify another 44 systems that are likely to contain Earth-like planets. Since only about 6,000 total exoplanets have been found so far, and little is known about many of them, there wasn’t much training data for Jeanne Davoult, an astronomer at Germany’s DLR, and her colleagues to feed to their algorithm.

By using the Bern Model of Planet Formation and Evolution at Switzerland’s University of Bern, they were able to simulate more than 55,000 planetary systems around different types of stars and use them to teach their algorithm how to pick ones that might have Earth-like planets in them. The algorithm found the 44 systems, and further studies showed its picks were promising.

“The results are impressive: the algorithm achieves precision values of up to 0.99, which means that 99% of the systems identified by the machine-learning model have at least one Earth-like planet,” Davoult told Space.com.

The model identified some interesting patterns for promising systems. For instance, around G-type stars like the sun, inner rocky planets tend to share a system with outer gas giants, like Earth does with Jupiter in our solar system. Gas giants that are closer to their star tend to knock potentially Earth-like rocky planets out of their way, making it less likely for them to form.

The algorithm could help sort through data gather by the European Space Agency’s PLATO, which is expected to discover thousands of planets.

“This is a significant step in the search for planets with conditions favorable to life and, ultimately, for the search for life in the Universe,” Davoult collaborator Yann Alibert of the University of Bern said.

Into Heavy Metal

“Polluted” white dwarf stars can tell us a lot about what planets are made of, which in turn can tell us a lot about which ones might support life. White dwarfs are dying stars, and they tend to rip apart and suck in the planets orbiting them as they collapse. Since the stars’ atmospheres are made of hydrogen and helium, any other elements found in them come from somewhere else.

“For polluted white dwarfs, the inside of the planet is literally being seared onto the surface of the star for us to look at,” University of Texas graduate student Malia Kao told UT News. “Polluted white dwarfs right now are the best way we can characterize planetary interiors.”

Finding these white dwarfs is how astronomers learn what planets outside our solar system are made of. In the past, identifying these stars involved sifting through mounds of survey data by hand. Using an AI technique called manifold learning, Kao and her colleagues quickly sorted through data from the Gaia telescope. The model sorted through 100,000 potential white dwarfs, identifying 375 good candidates with heavy metals in their atmospheres, and it did so with 99% accuracy.

“Our method can increase the number of known polluted white dwarfs tenfold, allowing us to better study the diversity and geology of planets outside our solar system,” Kao said. “Ultimately, we want to determine whether life can exist outside of our solar system. If ours is unique among planetary systems, it might also be unique in its ability to sustain life.”

And if there are other habitable worlds out there, we might find them without help from a lucky wormhole.