Abstract

Contributed Talk - Splinter DataManage

Thursday, 12 September 2024, 15:05   (S22)

Bayesian self-calibration and imaging in very long baseline interferometry

Jong-Seo Kim, Aleksei Nikonov, Vanessa Pinto, Jakob Roth, Torsten Ensslin, Michael Janssen, Philipp Arras, Hendrik Mueller, and Andrei Lobanov
Max Planck Institute for Radioastronomy

Calibration and imaging are closely interconnected in radio interferometry. The conventional CLEAN algorithm has been widely employed for imaging and calibration. However, forward modeling and Bayesian imaging algorithms have recently outperformed CLEAN, and these new imaging methods can also be utilized for various aspects of data calibration. This talk describes a new approach we have developed for extending the Bayesian imaging framework RESOLVE to include self-calibration of VLBI data. Applications of this approach to imaging of 43 GHz VLBA and 86 GHz GMVA+ALMA observations of M87 show that high-resolution images and antenna-based gain solutions with uncertainty estimation can be reconstructed jointly. Furthermore, we performed self-calibration and imaging for multiple MOJAVE VLBI survey data using resolve software. Joint inference of calibration solutions and images by using multiple data sets could be beneficial for future arrays, such as SKA and ngVLA. Automated and data-driven Bayesian calibration and imaging pipeline will be able to obtain robust and reproducible results from large amounts of data from next generation radio telescopes.