Wavelets and sparsity for Faraday tomography. (4th August 2022)

• Record Type: Journal Article
• Authors: Cooray, Suchetha; Takeuchi, Tsutomu T; Ideguchi, Shinsuke; Akahori, Takuya; Miyashita, Yoshimitsu; Takahashi, Keitaro
• Is Part Of: Publications of the Astronomical Society of Japan. Volume 75(2023)Supplement 1
• Page Start: S85
• Page End: S96
• Publication Date: 2022-08-04
• Subjects: magnetic fields -- methods: data analysis -- polarization -- techniques: interferometric -- techniques: polarimetric; Astronomy -- Periodicals; 520.5
• DOI: 10.1093/pasj/psac052
• Languages: English
• ISSNs: 0004-6264

Abstract

Abstract: Faraday tomography through broad-band polarimetry can provide crucial information on magnetized astronomical objects, such as quasars, galaxies, or galaxy clusters. However, the limited wavelength coverage of the instruments requires that we solve an ill-posed inverse problem when we want to obtain the Faraday dispersion function (FDF), a tomographic distribution of the magneto-ionic media along the line of sight. This paper explores the use of wavelet transforms and the sparsity of the transformed FDFs in the form of wavelet shrinkage (WS) for finding better solutions to the inverse problem. We recently proposed the Constraining and Restoring iterative Algorithm for Faraday Tomography (CRAFT; Cooray et al. 2021, MNRAS, 500, 5129), a new flexible algorithm that showed significant improvements over the popular methods such as Rotation Measure Synthesis. In this work, we introduce CRAFT + WS, a new version of CRAFT incorporating the ideas of wavelets and sparsity. CRAFT + WS exhibits significant improvements over the original CRAFT when tested for a complex FDF of a realistic Galactic model. Reconstructions of FDFs demonstrate super-resolution in Faraday depth, uncovering previously unseen Faraday complexities in observations. The proposed approach will be necessary for effective cosmic magnetism studies using the Square Kilometre Array and its precursors. The code is made publicly available at 〈https://github.com/suchethac/craft 〉.