Dice Question Streamline Icon: https://streamlinehq.com

Adequate convergence of DISCO-based reconstruction for Io is unresolved

Ascertain whether the DISCO approach (Method 2) can achieve an adequately converged image reconstruction for JWST/NIRISS AMI observations of Io, and identify the conditions under which visibility-based forward-modelling fails for heavily resolved sources.

Information Square Streamline Icon: https://streamlinehq.com

Background

The paper compares two reconstruction methods: Method 1 fits directly in the image plane, while Method 2 uses forward-modelled complex visibilities transformed into DISCO observables that are designed to be robust to wavefront errors. Io is a heavily resolved target in JWST/NIRISS AMI data.

While Method 1 with total variation regularization produces consistent, time-resolved maps of volcanic hotspots on Io, the authors explicitly state that Method 2 failed to converge adequately despite varied initializations and hyperparameter tuning, indicating an unresolved question about its applicability to strongly resolved scenes like Io.

References

Io is heavily resolved as seen by jwst, and visibility extraction in the forward-modelling approach has performed poorly; we have been unable to get any adequately converged image using Method 2 with a variety of initialisations and hyperparameter tunings.

Image reconstruction with the JWST Interferometer (2510.10924 - Charles et al., 13 Oct 2025) in Section “Results”, Subsection “Io Results”