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.
A team led by Rodrigo Ferrer-Chavez, a PhD student in the Department of Physics & Astronomy at Northwestern University, has developed a method called Physics-based Learning for Imaging Planets (FLIP). This computational technique essentially creates a digital twin
of the James Webb Space Telescope (JWST), allowing scientists to remove the blinding glare of stars and reveal faint exoplanets that were previously hidden. The researchers applied FLIP to existing JWST observations and improved the telescope’s sensitivity for detecting these distant worlds by two to fifteen times.
A Digital Telescope to Remove Stellar Glare
Directly imaging exoplanets is a monumental challenge. Planets can be more than 10,000 times fainter than their host stars, and even the most advanced telescopes like JWST struggle to overcome this contrast.
For years, astronomers have used a technique called KLIP, which subtracts a model of a star’s glare based on reference images of other stars. However, this method has limits.
The FLIP method tackles this problem differently. Instead of relying on reference stars, it builds a highly detailed, physics-based model of the JWST itself, specifically its Near Infrared Camera (NIRCam). This model includes parameters for star position, optical distortions, and even tiny pointing errors. The team then uses optimization techniques—similar to those used in training AI models—to tune about one million different parameters in this virtual telescope until its simulated output closely matches the real, messy observation from the actual JWST.
Recovering a Hidden Planet
To test FLIP, the team re-analyzed three sets of JWST observations. Two were of star systems known to host planets: HIP 65426 and 51 Eridani. The third was TYC 5899, a star with no known planets, which served as a control case.
The results were striking. On the star HIP 65426, FLIP improved the signal-to-noise ratio by a factor of 1.5 to 2. For 51 Eridani, the improvement was even more significant. The method successfully recovered the signal of the giant exoplanet 51 Eridani b at a specific infrared wavelength (4.3 microns) where conventional processing had failed to distinguish it from residual starlight. This is a critical finding because carbon dioxide in the planet’s atmosphere absorbs light at that band, making its signal even fainter.
It’s important to note that 51 Eridani b was not a new discovery; it was first spotted in 2015 using the Gemini Planet Imager. However, this is the first time it has been clearly detected in JWST data at this wavelength, opening the door for new atmospheric analysis. The most dramatic sensitivity gain was seen with the star TYC 5899, where FLIP boosted the detection sensitivity by a factor of 15.
It’s not that they don’t exist. It’s that we cannot see them. Every time we improve our techniques, we discover new types of planets. By finding more planets, we can better understand how planetary systems form and evolve, and ultimately how typical our own solar system truly is.
A More Flexible Future for Planet Hunting
The implications of this new technique extend far beyond re-analyzing existing JWST data.
This is particularly important for the future of exoplanet research. NASA’s proposed Habitable Worlds Observatory (HWO), planned for the 2040s, is designed to directly image Earth-like planets around Sun-like stars. The task is incredibly demanding; such planets are roughly 100 to 1,000 times fainter than their stars. For HWO to succeed, it would need to be 100 to 1,000 times more stable than JWST—a level of precision that is extremely difficult and expensive to engineer.
FLIP offers a potential solution. By allowing software to correct for the tiny instabilities in a telescope’s optics, the method could relax the unprecedented stability requirements needed to make these a success,
as the Nature paper suggests. This means future telescopes like HWO might not need to be perfectly rigid, potentially making them more feasible to build and launch.
The technique is not limited to space telescopes. The researchers believe FLIP could be adapted for ground-based observatories as well. This adaptability could accelerate the discovery of smaller, colder, and fainter planets that have so far remained invisible.
The study, titled Imaging exoplanets below the telescope stability limit,
was published in the journal Nature. The method proves that computational innovation can expand the limits of existing hardware and enable more exoplanet discoveries.
