The topic
Radio interferometry builds a telescope from many smaller dishes separated by large distances. In the right arrangement, that array can behave like a single antenna with an aperture as wide as the greatest spacing between elements. The familiar rule applies: larger aperture, finer detail on the sky. That is why ALMA, the VLA, and future arrays such as the SKA are laid out over kilometres or continents instead of being one solid dish.
Each pair of antennas measures only a fragment of the spatial frequencies needed for an image. We never observe the full set, and noise is always present. Mathematically, many different sky brightness patterns can match the same data. In other words, the reconstruction problem does not have a unique solution unless we add sensible assumptions and quantify what remains uncertain. Much of the field is about making that choice explicit so we do not mistake software artefacts for astrophysics.
What I do
I work on imaging algorithms and software that turn raw interferometric visibilities into maps suitable for publication, with emphasis on speed, reproducibility, and checks along the way. I lead development of Pyralysis, a framework aimed at large surveys such as the SKA, and supervise student projects on reconstruction methods and GPU computing.
Selected publications
