Amazing New Planet Haul
In a great leap forward in detecting (possible) exoplanets around other stars, a team of astronomers announced yesterday that they have detected 11,554 new planet candidates from data collected by NASA’s Transiting Exoplanet Survey Satellite (TESS). Previously, TESS data have been used to confirm 882 exoplanets, but also there are 7,927 candidates that have been at least provisionally integrated into NASA’s Exoplanet Archive and a further 4,787 candidates that have not been either confirmed or identified as false positives. Since TESS’ launch in 2018, it has been finding exoplanets by observing periodic dips in the brightness of distant stars that are induced as exoplanets moving along their respective orbits pass in front of the stars (from our perspective) and dim their light slightly and briefly.
To accomplish this feat, the team created the “T16 project” to look at 83,717,149 light curves, the variations of brightness over time, of objects down to a visual magnitude of 16, i.e., much dimmer than the light curves TESS’ pipeline normally processes. (Magnitudes are the historical units by which astronomers measure brightness. For context, our eyes can see stars down to a magnitude of 6 in the darkest of skies but this project includes stars that are 10 magnitudes, or 10,000 times, fainter. Note that we wacky astronomers use a scale for magnitudes that is both backwards and logarithmic.) To analyze nearly 84 million light curves within a human lifetime, the team employed machine learning, enabling them to examine the light curves for periodic dips in a matter of seconds rather than years of astronomers’ lives. To be fair, 411 of the candidates showed single-transit events, leaving “only” 10,091 candidates for exoplanets with orbital periods between 0.5 and 27 days. That’s still quite a lot!
Of course, these are exoplanet candidates still and have not been vetted by independent observation. Nevertheless, the team did do some follow-up observations with the 6.5-m Magellan Telescope in Chile of one candidate host star, TIC 183374187 with a magnitude of 14.5, by looking to see if it exhibited also the back-and-forth velocity shuffle expected when a star is orbited by an exoplanet. In this particular case, the exoplanet was confirmed, having a mass of 0.56 Jupiter masses with a period of 5.06 days. With a size of 1.25 Jupiter radii, the resulting density is about 0.35 g cm-3, making the exoplanet consistent with being a “hot Jupiter,” a gas giant whose size has been puffed up by its close proximity to its host star. So it seems the machine learning technique that was used to identify this one object as hosting an exoplanet worked; welp, 10,090 objects to go I guess! For more information, here’s a link to the story on universetoday.com and a link to the preprint of team’s paper, which has been accepted for publication soon in The Astrophysical Journal Supplement Series journal.
