A team led by Michael Collier and Steven Giacalone has statistically validated 64 new exoplanets discovered by NASA's TESS satellite, marking one of the largest exoplanet validation studies to date. The research, formally accepted by The Astronomical Journal, used an automated AI-driven data processing pipeline to analyze hundreds of gigabytes of TESS data, integrating disparate systems that typically require manual work. The first author noted having no prior astronomy experience but was guided by astrophysicists through the peer-review process.

Using TRICERATOPS, the team computed false-positive probabilities (FPP) for 443 TESS planet candidates. For 264 candidates with high-resolution imaging observations, they ran controlled experiments comparing FPP with and without contrast curves. They found that 72% of 68 contrast-curve bearing validated planets would fail validation without their imaging data. The impact was most pronounced for smaller planets: 100% of planets below 1.7 R⊕ required imaging, compared to 33% above 4 R⊕.

Four of the newly validated planets are highly suitable for James Webb Space Telescope (JWST) observations, each achieving validation only through their imaging constraints. This work demonstrates that high-resolution imaging directly limits the yield of small-planet validation and reduces the pool of targets available for atmospheric characterization.