Roman Telescope Hits 0.00001-Degree Stability in Coronagraph Tests, Validating Spectral Guiding

Engineers at NASA's Goddard Space Flight Center ran the fine-guidance system from September 15-21 using sections of the Wide Field Instrument's 18 detectors to track guide-star spectra four times per second. On September 22 and 27 the same loop fed the coronagraph, which added its internal tip-tilt control. The combined system held the line of sight steady enough that a laser spot would remain on a dime from 240 km—performance already near the 370 km goal after further tuning. No separate guider telescope is used; spectral features supply the error signal.

This approach removes a mass and complexity penalty carried by Hubble and JWST while directly supporting the coronagraph's 10^-9 contrast requirement. The stability numbers sit between current ground-based extreme-AO systems and the 10^-10 contrast needed for Earth-like planet spectroscopy, closing a key technology gap identified in the 2020 Astrophysics Decadal Survey. Because the same detectors serve both science and guiding, any future degradation in one channel immediately affects the other—an operational coupling not present in prior observatories.

Next steps include on-orbit validation of spectral guiding by late 2026 and a six-month coronagraph technology demonstration beginning in 2027. If both succeed, the mission can shift from technology test to a 2,000-hour exoplanet imaging survey that will set target lists for the Habitable Worlds Observatory. Ground-based extremely large telescopes will use Roman's yield statistics to optimize their own observing queues.

The primary remaining uncertainty is long-term thermal drift of the detector mounts once the observatory reaches its L2 halo orbit; thermal-vacuum data suggest the drift can be calibrated to 0.2 mas, but flight data are required.