Scientists estimate billions of planets orbit other stars in the Milky Way. Finding them is one thing; photographing them is another — a planet is a faint speck next to a star that outshines it by orders of magnitude. Now researchers with ties to Caltech and JPL have found a way to erase the star from the picture, and planets that were hiding in plain sight are starting to show up.
The method is called FLIP — Physics-based Learning for Imaging Planets — described in a paper published online November 4 in Nature. In tests on Webb telescope images, it improved sensitivity to faint planets by two to 15 times over standard techniques.
A digital twin of the telescope
The problem is not the star alone. It is the telescope itself. Even the tiniest imperfections in a telescope’s optics smear a star’s light across the image, and that smeared glare drowns out anything faint nearby. “Directly imaging exoplanets is incredibly challenging because even the tiniest imperfections in a telescope can make it difficult to separate the faint light of a planet from the overwhelming glare of its star,” said Katie Bouman, a Caltech professor of computing and mathematical sciences, electrical engineering and astronomy, and a co-author of the paper.
Instead of building a more perfect telescope, the team built a perfect model of an imperfect one: a “digital twin” of Webb’s Near Infrared Camera, NIRCam. Machine-learning optimization then solved for millions of possible physical imperfections in the real instrument’s data, so the team could model exactly what a given star’s glare should look like — and subtract it, cleanly, from the image.
A planet that was already in the picture
With the standard method, an observation of the star 51 Eridani A showed no sign of the planet 51 Eridani b. Run the same observation through FLIP, and the planet appears — a significant detection, right where it was expected to be. The planet was in the data the whole time. It took a better way of reading the data to see it.
“We now have a tool that lets us discover and study planets that we were not able to ‘see’ before, even if that planet was already in an image in the JWST archive,” said lead author Rodrigo Ferrer-Chavez, a graduate student at Northwestern University. “We can try to find planets that may have been missed there, and also planets that perhaps we tried to image, but we could not actually detect at the time.”
Why it matters here in Pasadena
The Pasadena connection runs through the whole project: Bouman is at Caltech, and co-author Dimitri Mawet is Caltech’s David Morrisroe Professor of Astronomy and a senior research scientist at the Jet Propulsion Laboratory, which Caltech manages for NASA. Ferrer-Chavez first took on the problem as an undergraduate at the Autonomous University of Yucatán in Mexico, then worked on it during a Summer Undergraduate Research Fellowship in Mawet’s lab. His first attempt, as an undergraduate thesis, failed — and the team credits that failure with making the eventual approach possible.
The work also points at what comes next. NASA’s recently launched Nancy Grace Roman Space Telescope carries a Coronagraph Instrument designed by scientists at Caltech, JPL and IPAC to block out starlight and capture exoplanets. The researchers say FLIP is not limited to Webb: it could be used on Roman, on ground-based telescopes, and on future missions such as NASA’s planned Habitable Worlds Observatory, which aims to take direct images of Earth-like planets around Sun-like stars.
Imaging an Earth-like planet means photographing something roughly 10 billion times fainter than the star sitting right next to it. “I feel we just unlocked a new frontier,” said Jason Wang, the Northwestern astronomer in whose lab Ferrer-Chavez works. “We’ve now gone beyond the telescope stability limit for the first time.”
For now, the work is running FLIP over the Webb archive to see what else was missed the first time around. The ultimate dream, Bouman said, is pictures of Earth-like planets around other stars — and, perhaps, one day signs of life. “Sometimes, the key to seeing something new isn’t just building a better instrument but finding a better way to make sense of the light it collects.”
Photo: The exoplanet HIP 65426 b, the first exoplanet Webb ever imaged directly, shown in four bands of infrared light. NASA/ESA/CSA, A. Carter (UCSC), the ERS 1386 team, and A. Pagan (STScI) — public domain, via Wikimedia Commons.
Sources: Caltech, “Erasing the Stars to Find Exoplanets,” November 4, 2026; Nature, “Imaging exoplanets below the telescope stability limit,” published online November 4, 2026; NASA/JPL, “NASA’s Webb Takes Its First-Ever Direct Image of Distant World.”
