Overview
- The Nature paper published on Nov. 4, 2026, introduces Physics‑based Learning for Imaging Planets (FLIP), a method that builds a physics‑based digital twin of JWST and tunes it to match real observations.
- Applied to archival JWST coronagraphic images of three targets, FLIP produced sensitivity gains of about 2 to 15 times and revealed the known giant planet 51 Eridani b in data where standard processing failed.
- FLIP works by simulating how starlight travels through JWST’s mirrors and detectors, adjusting model parameters to reproduce tiny thermal and mechanical shifts, and subtracting the modeled glare rather than relying only on reference‑star images.
- Authors and outside experts say the software could inform design trade‑offs for NASA’s proposed Habitable Worlds Observatory by reducing part of the stability error budget, but they caution that imaging Earth analogs still needs instruments far more stable than JWST.
- In practice the advance means astronomers can reprocess existing JWST data to hunt for faint planets closer to their stars and will need further validation across more datasets before missions change hardware requirements.