Removing glare from James Webb image reveals a previously hidden planet

Show summary Hide summary

A team of astronomers has found a way to pry faint planets out of the bright glare of their host stars in James Webb Space Telescope data. Their new technique blends physics-based modelling with machine-learning tools to produce a virtual replica of the telescope and reveal companions that earlier analyses missed.

Building a virtual James Webb

Detecting an exoplanet in direct images is notoriously difficult because a planet can be roughly a million times dimmer than its star. As study lead author Rodrigo Ferrer-Chavez of Northwestern University put it, “Planets are a million times fainter than their stars. So, to image exoplanets, we need to remove that glare.”

Computer screen showing simulation of telescope optics and diffraction pattern
A virtual model of telescope optics helps separate starlight from faint planetary signals.

To do that, the researchers created a “digital twin” of the James Webb Space Telescope. This virtual instrument models how light passes through JWST’s detectors and optics while excluding small, movement-driven imperfections that blur real images. The team combined established physical models of the telescope with machine-learning methods used during model training, though they emphasize the outcome is grounded in physics rather than being a black-box AI system.

As co-author Jason Wang, also at Northwestern, said: “Because we already know the physics, we don’t need AI to understand what’s happening.”

Unearthing a hidden giant: 51 Eridani b

Applying their approach—called Physics-based Learning for Imaging Planets (FLIP)—the researchers reanalyzed JWST observations of three nearby stars. One target, 51 Eridani, lies about 97 light-years from Earth and had long been suspected to host a giant planet.

Researcher examining a processed telescope image revealing a faint companion near a bright star
FLIP processing revealed the previously hidden planet 51 Eridani b in JWST data.

Previous JWST images had failed to clearly show the object because residual starlight and tiny telescope motions concealed its signal. Using FLIP, the team removed those systematic effects and the planet known as 51 Eridani b emerged from the data. Ground-based telescopes had detected this planet before, but JWST observations had so far been overwhelmed by stellar glare.

Why the technique matters

The authors published their results on Nov. 4 in Nature (Ferrer‑Chavez et al., 2026; doi: 10.1038/s41586-026-11213-z). They say FLIP could help astronomers find more faint worlds that current methods miss, particularly planets that resemble those in our own solar system.

Historically, exoplanet surveys have found many large, highly reflective planets because those are easier to spot. The team argues that similar-sized but less reflective planets—more like the worlds around the Sun—may simply remain hidden beneath instrument noise.

“It’s not that [solar-system-like planets] don’t exist,” Ferrer-Chavez said. “It’s that we can’t see them.”

Practical limits and next steps

The technique addresses limitations that affect even advanced instruments: coronagraphs can block a lot of starlight, but they cannot remove it entirely, and tiny vibrations or pointing instabilities shift the apparent pattern of residual light. By reconstructing how a stable telescope would form an image, FLIP reduces those mismatches and improves sensitivity to faint companions.

The researchers hope the method will be applied to more JWST datasets and to other instruments, enabling more confident detections of hard-to-see exoplanets. If successful, the approach could change the kinds of planets astronomers are able to study with direct imaging, and help close the gap between the planets we can detect and the planets that actually exist.

Give your feedback

Be the first to rate this post
or leave a detailed review



The Pacific Tribune is an independent media. Support us by adding us to your Google News favorites:

Post a comment

Publish a comment