How Australia is helping sharpen the James Webb Space Telescope
When NASA launched the James Webb Space Telescope (JWST) in 2021, its ability to resolve faint celestial objects depended on preserving an exceptionally sharp view from over a million kilometres away. A team of Australian scientists from the University of Sydney has now pushed that capability further by developing a software-based correction that recovers lost detail in JWST’s imagery. The advancement hinges on the telescope’s Aperture Masking Interferometer (AMI), a small metal plate inside one of JWST’s cameras that detects optical distortions in the optical path.
The problem at the heart of the images
Led by PhD student Louis Desdoigts, the researchers discovered that bright pixels in JWST images were leaking light into nearby darker pixels. This leakage blurred fine details and degraded the telescope’s ability to distinguish dim planets from the glare of nearby stars. The effect limited JWST’s potential to unveil the faint companions and features that scientists chase in distant systems.
How the correction works
To tackle the issue, the team built a detailed model of how AMI behaves under real observing conditions. They then trained a machine-learning system to reconstruct the true signal by reversing the leakage patterns observed in the corrupted images. The result is an open-source correction pipeline that restores AMI’s full resolution power, effectively peeling back the blur without requiring any physical adjustments to the telescope.
From theory to sharper science: notable results
Applying the correction to JWST data yielded immediate, tangible improvements. One of the first success cases was the star HD 206893, where the enhanced processing brought a faint planetary companion and a reddish-brown dwarf into clearer view—a discovery not as readily discernible with earlier processing.
Encouraged by this success, co-researcher Max Charles extended the technique to more complex targets. The upgraded pipeline sharpened features on Jupiter’s volcanic moon Io, enabling more precise monitoring of eruptions. It also refined images of a supermassive black hole jet in galaxy NGC 1068 and revealed dust structures surrounding the unusual star WR 137.
Why this software fix matters for space astronomy
JWST operates at an extraordinary distance from Earth—about 1.5 million kilometres—making direct physical repairs impractical. Software-based calibration and image reconstruction thus become essential tools for ensuring the telescope delivers on its scientific promise. The Australian work demonstrates how data-driven, machine-learning techniques can complement hardware design, extending JWST’s lifetime and scientific reach without extra on-orbit servicing.
Looking ahead
The team stresses openness: the correction software is open source, inviting collaboration from researchers around the world. As JWST continues to collect data across the infrared spectrum, refined AMI corrections will help astronomers hunt for faint companions, subtle disk structures around young stars, and the signatures of distant exoplanets. The Australian contribution underscores how international partnerships, informed by rigorous modelling and modern ML, can sharpen the eye of one of humanity’s most powerful space telescopes.
Conclusion
In a field governed by precision, a small software insight can yield big returns. By decoding and reversing AMI leakage with machine learning, Australian scientists have sharpened JWST’s gaze, enabling clearer views of the cosmos and opening new pathways to discover the faint phenomena that illuminate our understanding of the universe.
