Unsupervised Generation of High Dynamic Range Solar Images: A Novel Algorithm for Self-calibration of Interferometry Data

Abstract

Abstract Solar radio emission, especially at meter wavelengths, is well known to vary over small spectral (≲100 kHz) and temporal (<1 s) spans. It is comparatively recently, with the advent of a new generation of instruments, that it has become possible to capture data with sufficient resolution (temporal, spectral, and angular) so that one can begin to characterize the solar morphology simultaneously along the axes of time and frequency. This ability is naturally accompanied by an enormous increase in data volumes and computational burden, a problem that will only become more acute with the next generation of instruments such as the Square Kilometre Array (SKA). The usual approach, which requires manual guidance of the calibration process, is impractical. Here we present the Automated Imaging Routine for Compact Arrays for the Radio Sun (AIRCARS), an end-to-end imaging pipeline optimized for solar imaging with arrays with a compact core. We have used AIRCARS so far on data from the Murchison Widefield Array (MWA) Phase-I. The dynamic range of the images is routinely from a few hundred to a few thousand. In the few cases, where we have pushed AIRCARS to its limits, the dynamic range can go as high as ∼100,000. The images made represent a substantial improvement in the state of the art in terms of imaging fidelity and dynamic range. This has the potential to transform the multi-petabyte MWA solar archive from raw visibilities into science-ready images. AIRCARS can also be tuned to upcoming telescopes like the SKA, making it a very useful tool for the heliophysics community.