IMAGE RECONSTRUCTION

STATE OF THE ART INTERFEROMETRIC IMAGING

I have developed and applied sparse image reconstruction algorithms to observations from radio telescopes, to show how these algorithms can change the way we see the Universe. Above is an image of the radio galaxy PKS J0116-473, where the sparse image reconstruction compared with a traditional modelling method. This can also be seen in Pratley et al., 2018. See the software PURIFY (interferometric imaging) and SOPT (convex optimisation) for the open source code, of which I am a development team member.

DISTRIBUTED IMAGE RECONSTRUCTION

The new generation of telescopes are collecting more information than can be processed by traditional methods. By distributing new algorithms over computing clusters, we can create images from these telescopes. The software package SOPT is collection of distributed convex optimisation algorithms. This powers interferometric imaging software PURIFY to operate on computing clusters. See the software PURIFY (interferometric imaging) and SOPT (convex optimisation) for the open source code.

WIDE-FIELD IMAGING METHODOLOGY

For low frequency radio telescopes, the fields of view can be as wide as 30 by 30 degrees or more. The curvature of the sky needs to be modelled to get accurate image reconstruction. Pratley, Johnston-Hollitt & McEwen 2019 shows how to computationally distribute imaging algorithms for wide-field imaging.

POLARIMETRIC METHODOLOGY

Light is the single most important way to learn about the distance galaxies and stars, and the polarisation of light can tell us about their distant magnetic fields. The above picture is the polarised image of a galaxy at radio wavelengths, reconstructed with methodology from Pratley & Jonston-Hollitt, 2016. See this task in ATNF MIRIAD.

I have also developed new formalism for wide/broadband rotation measure synthesis (Pratley & Jonston-Hollitt, 2020).