RAVEN finds 118 exoplanets in TESS

Exoplanet hunting is a data problem now. Telescopes like NASA’s TESS don’t just find a few promising blips — they dump out millions of light curves and thousands of maybe-planets. The gap has been sorting the real worlds from the impostors fast enough to matter. That’s where RAVEN comes in: a University of Warwick-led team used the pipeline on TESS data from its first 4 years and came back with 118 statistically validated planets, plus a much bigger pile of strong candidates. (arxiv.org) ### What is RAVEN actually doing? RAVEN is not a telescope. It’s a vetting and validation system built for TESS candidates. Basically, it looks at the dip in starlight that might mean “planet crossing in front of star,” then asks a harder question — could this same signal be caused by something else, like an eclipsing binary star or instrumental noise? To do that, it combines machine-learning classifiers with a B(arxiv.org)eight false-positive scenarios. (arxiv.org) ### Why was TESS missing some of these? Because transit surveys are messy. TESS watches for tiny drops in brightness, but stars vary on their own, nearby stars can contaminate the signal, and binary stars love to impersonate planets. Earlier pipelines were good, but not uniform across everything TESS collected, especially in the full-frame images covering millions of stars. RAVEN was built to run that search and cleanup pass at scale. (arxiv.org) ### How big was this sweep? Pretty big. The team searched a magnitude-limited sample of more than 2.2 million main-sequence stars, using TESS full-frame image data from sectors 1 through 55 — basically the mission’s first 4 years. They focused on short-period planets with orbits from 0.5 to 16 days, which means these are mostly close-in worlds, not temperate Earth twins. (arxiv.org)number is 118 newly validated planets. Inside that set are 31 planets that RAVEN itself newly detected in the TESS data, rather than just upgrading old ambiguous candidates into “very likely real.” The pipeline also produced more than 2,000 additional high-probability candidates that are not yet validated, and about 1,000 of those are new as well. (arxiv.org)s showed up? A lot of the interesting ones are weird, hot, and fast. The sample includes ultra-short-period planets that whip around their stars in less than a day, planets in the so-called Neptunian desert where theory says they should be relatively scarce, and tightly packed multi-planet systems. That matters because unusual edge-case planets are often the ones that stress-test formation theories. (sciencedaily.com) ### Why does “validated” matter so much? Because follow-up time is expensive. Big telescopes can’t chase every maybe-signal in the archive. RAVEN’s threshold for statistical validation is strict — greater than 99% posterior probability against each false-positive scenario, with an implied radius under 8 Earth radii for this version of the pipeline. In other words, it’s trying to hand astronomers a cleaner shortlist, not just a longer one. (arxiv.org) ### Is this replacing astronomers? Not really — it’s changing where they spend time. Think of RAVEN as a very fast triage nurse for planet signals. It doesn’t make spectroscopy, mass measurements, or detailed atmospheric work unnecessary. But it can shrink the haystack before humans and follow-up instruments go looking for the needles. (arxiv.org) ### What’s the bottom line? T(arxiv.org)s that exoplanet science now has a way to re-mine old survey data and turn “maybe” into “worth chasing.” TESS has already helped push the confirmed exoplanet count past 6,000. Tools like RAVEN mean the archive is still yielding new worlds — and probably will for a while. (science.nasa.gov)