BRISK: Towards Scalable Wide-Field Imaging for the Next Generation of Radio Telescopes

Over the last 18 months, I have been developing a software package called BRISK (Baseline-aware Rank-compressed Interferometric Sky-imaging Toolkit) as part of my PhD research at the University of Oxford. BRISK is an end-to-end imaging pipeline that processes data collected by radio telescope arrays and produces highly accurate images of astronomical objects such as stars, galaxies, and other sources across the Universe.

The research behind BRISK focuses on reducing the computational cost of radio interferometric imaging. This is becoming increasingly important as next-generation facilities, such as the Square Kilometre Array (SKA) and DSA-2000, will generate enormous volumes of data, measured in petabytes per day. Processing this information requires vast computing resources and significant energy consumption, making efficient imaging algorithms a key challenge for modern astronomy.

Earlier this year, I submitted my first research paper to the Monthly Notices of the Royal Astronomical Society (MNRAS), presenting the algorithms behind BRISK and the results of extensive performance testing. We compared BRISK against two of the most widely used imaging packages in radio astronomy, WSClean and CASA, evaluating both image quality and computational performance. Our results showed that BRISK can be up to 1,000 times faster while maintaining, and in some cases improving upon, the image accuracy achieved by existing methods. We also demonstrated imaging at scales that have not previously been achieved on a single CPU node.

Thanks to the generous support of the Tanglin Trust School Foundation, I have been able to pursue several exciting research directions over the past few months, leading to novel results that are now being prepared for publication. I am extremely grateful for Tanglin's support, which has helped me advance my research and contribute to the development of the next generation of astronomical imaging technology.