A new software trick is helping the Webb telescope exoplanets hunt see faint worlds that were buried in glare, and it works on pictures the telescope has already taken. A team led by Rodrigo Ferrer-Chavez of Northwestern University published the method in Nature on November fourth, according to Caltech.
The problem: stars are way too bright
Spotting a planet around another star is a bit like trying to see constellations at noon. According to Scientific American, even the biggest planets can be more than ten thousand times fainter than the stars they circle, and their light gets lost in the glare.
Webb already carries a mask called a coronagraph that blocks most of a star's light, which the lead author compares to an artificial eclipse. A little glare still leaks through, so astronomers photograph a similar reference star and subtract its glare from the target image.
That template is only a snapshot, though. Stars sit slightly off-center, temperatures shift, and the six and a half meter mirror flexes a tiny bit. Those mismatches leave smudges of starlight that can hide a small planet.
The fix: a digital twin of the telescope
The team built a virtual copy of Webb's Near Infrared Camera, then tuned it until it matched real observations. The approach is called Physics-Based Learning for Imaging Planets, or FLIP. Instead of borrowing glare from another star, FLIP simulates how the target star's light travels through the instrument, wobbles included.
Caltech says the researchers used methods from machine learning, with millions of adjustable knobs, to solve the physics. Ferrer-Chavez describes watching the glare shrink in bizarre ways as the model improves, until a clean image appears and, with luck, a glowing blob that is a planet.
What it found
FLIP was tested on Webb observations of three star systems. The method was two to fifteen times more sensitive than standard techniques, and the gains were biggest close to the star, where Earth-like planets tend to orbit.
In one system, the software pulled out the giant planet 51 Eridani b, which had not been visible at the wavelengths Webb can detect. That makes the Webb telescope exoplanets story more than a lab demo, since a real world popped out of old data.
Dimitri Mawet of NASA's Jet Propulsion Laboratory, who was not involved, called it a significant result. He told Scientific American that model-based glare removal has been discussed for more than a decade, but until now it could not beat methods built on physical reference stars.
Why Gen Z should care
Today Webb can detect worlds about the size of Saturn that orbit farther out than Pluto does around our sun. Imaging an Earth-sized planet would take a telescope far steadier than anything built so far. FLIP does not make that easy, but it moves the line.
The bigger prize is NASA's planned Habitable Worlds Observatory, a Webb successor aimed at finding Earth-like planets and maybe signs of life. It is planned to launch in the twenty-forties. Software that cleans up glare means the hardware can wobble a little more, which could ease the build.
Even so, the optics will still need to hold steady to within a trillionth of a meter to photograph planets about ten billion times fainter than their stars. As Ferrer-Chavez put it, building the observatory may be really hard, but not really, really, really hard.
What comes next
Because the technique is not tied to one instrument, the team plans to apply it to NASA's Roman Space Telescope and to ground-based telescopes. Co-author Jason Wang says the method could be used on any telescope.
Old archives could also get a second look. Ferrer-Chavez says astronomers can now search images already sitting in Webb's archive for planets that were missed or never detected. For more stories like this, check our science section and our tech coverage.
Katie Bouman of Caltech sums up the idea: sometimes the key to seeing something new is not building a better instrument but finding a better way to make sense of the light it collects.
Comments 0
No comments yet. Be the first to share your thoughts!
Leave a comment
Share your thoughts. Your email will not be published.