IViS: Interferometric Visibility-domain inversion Software - A GPU-accelerated Python framework for joint deconvolution with ASKAP
Abstract
Wide-field spectral-line imaging with modern radio interferometers remains challenging when the emission is extended, multiscale, and distributed over many overlapping pointings. Accurate reconstruction requires joint treatment of the calibrated visibilities, control of image-domain regularity, and recovery of missing short spacings from single-dish data. We present IViS (Interferometric Visibility-domain Inversion Software), a GPU-accelerated Python framework for visibility-domain joint deconvolution, and assess its performance for wide-field HI imaging with the Australian Square Kilometre Array Pathfinder (ASKAP). The current model, Classic3D, reconstructs a non-parametric sky cube directly from calibrated visibilities by minimizing a regularized least-squares criterion. The forward model relies on a non-uniform fast Fourier transform (NUFFT), supports joint deconvolution across mosaic pointings and positivity constraints, and can incorporate single-dish data through a fusion term. We validate the method using point-source, noise-only, and multiscale diffuse-emission simulations, and apply it to ASKAP observations toward the Large Magellanic Cloud (LMC). The simulations show that IViS recovers point-source flux accurately without regularization and that regularization sets the trade-off between noise suppression and effective resolution. Joint deconvolution can recover more large-scale power than linear mosaicking of independently deconvolved pointings. With single-dish information, the reconstructed power spectra closely reproduce the input sky statistics. Applied to ASKAP data, IViS yields 10h mosaics with an effective resolution of 22'' and reduced residual side-lobe structure. Compared with ASKAPSoft multi-scale CLEAN for a single observing block, Classic3D exhibits fewer residual side-lobes and recovers substantially more power at low spatial frequencies.
Related
- Denoising Interferometric Observations Using Visibility-Informed Neural Networks
- Generalised Visibility ILC: a new foreground mitigation strategy for interferometric HI intensity mapping in the low-redshift Universe
- Aurora Hunter: A Two-Stage Framework for Probabilistic Visibility Forecasting
- Inversion Framework of Internal Mass Distribution Parameters of Asteroid Apophis from Dynamical Observations
- jax-SNANA: A GPU-accelerated Supernova Framework for the SBI Era
BibTeX
@misc{marchal2026ivis,
title = {{IViS: Interferometric Visibility-domain inversion Software - A GPU-accelerated Python framework for joint deconvolution with ASKAP}},
author = {A. Marchal and N. M. McClure-Griffiths and J. Dempsey and H. Nguyen and H. Dénes and J. Dickey and C. Lynn and Y. K. Ma and D. McConnell and M. -A. Miville-Deschênes and N. Pingel and J. Th. van Loon},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.17636},
eprint = {2609.17636},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.17636},
note = {CC BY-NC-SA 4.0},
}
RIS
TY - UNPB
TI - IViS: Interferometric Visibility-domain inversion Software - A GPU-accelerated Python framework for joint deconvolution with ASKAP
AU - A. Marchal
AU - N. M. McClure-Griffiths
AU - J. Dempsey
AU - H. Nguyen
AU - H. Dénes
AU - J. Dickey
AU - C. Lynn
AU - Y. K. Ma
AU - D. McConnell
AU - M. -A. Miville-Deschênes
AU - N. Pingel
AU - J. Th. van Loon
PY - 2026
DA - 2026/09/15
PB - arXiv
DO - 10.48550/arxiv.2609.17636
UR - https://arxiv.org/abs/2609.17636
AB - Wide-field spectral-line imaging with modern radio interferometers remains challenging when the emission is extended, multiscale, and distributed over many overlapping pointings. Accurate reconstruction requires joint treatment of the calibrated visibilities, control of image-domain regularity, and recovery of missing short spacings from single-dish data. We present IViS (Interferometric Visibility-domain Inversion Software), a GPU-accelerated Python framework for visibility-domain joint deconvolution, and assess its performance for wide-field HI imaging with the Australian Square Kilometre Array Pathfinder (ASKAP). The current model, Classic3D, reconstructs a non-parametric sky cube directly from calibrated visibilities by minimizing a regularized least-squares criterion. The forward model relies on a non-uniform fast Fourier transform (NUFFT), supports joint deconvolution across mosaic pointings and positivity constraints, and can incorporate single-dish data through a fusion term. We validate the method using point-source, noise-only, and multiscale diffuse-emission simulations, and apply it to ASKAP observations toward the Large Magellanic Cloud (LMC). The simulations show that IViS recovers point-source flux accurately without regularization and that regularization sets the trade-off between noise suppression and effective resolution. Joint deconvolution can recover more large-scale power than linear mosaicking of independently deconvolved pointings. With single-dish information, the reconstructed power spectra closely reproduce the input sky statistics. Applied to ASKAP data, IViS yields 10h mosaics with an effective resolution of 22'' and reduced residual side-lobe structure. Compared with ASKAPSoft multi-scale CLEAN for a single observing block, Classic3D exhibits fewer residual side-lobes and recovers substantially more power at low spatial frequencies.
ER -