Radio astronomers from NRAO and ASTRON presenting at Pawsey
3:30pm - 5:00pm
You are invited to join Pawsey who is hosting Rick Perley from the National Radio Astronomy Observatory (NRAO) in Socorro, New Mexico and André Offringa from the Netherlands Institute for Radio Astronomy (ASTRON). They will provide short talks as follows:
Date: Wednesday 17 October, 2018
Venue: Pawsey Supercomputing Centre, Burvill Court, Kensington WA 6151
Imaging Challenges for Radio Astronomy Interferometers, by Rick Perley
Radio interferometers generate images of the sky by utilizing the Fourier transform relationship between the sky brightness and the cross-products of the antenna voltages. Although the theory is well established, the actual applications are fraught with difficulties, as the sampling of the Fourier plane is necessarily discrete and incomplete. As a result, the direct transform of the data (the ‘right’ answer) is wholly unsuited for astronomical analysis. In order to ‘clean’ up the initial images, astronomers utilize various ‘deconvolution’ techniques — one of which, ‘CLEAN’ is the oldest, simplest, fastest, and by far the most utilized. It also gives the wrong result, in the sense that the output image does not match the observed data. In this short presentation, I will illustrate these points with data taken by the NRAO’s Very Large Array, and discuss the relationship of this problem with the subject of Compressed Sensing, which aims to optimize the information available, although at a high cost in computing.
Methods for detecting the Epoch of Reionization with LOFAR, by André Offringa
André will discuss the LOFAR Epoch of Reionization (EoR) project. The Epoch of Reionization is a cosmological process that has taken place in the early ages of the Universe. Several experiments are set out to detect this era, including the Australian MWA and the Dutch/International LOFAR telescope. In the LOFAR EoR project, the international Low-Frequency Array is used to make a statistical detection of redshifted Hydrogen (21-cm) signals originating from this crucial era in the early evolution of our Universe. The LOFAR EoR project has currently set the lowest upper-limits on these signals, so far using only a single night of observation of LOFAR data.
Approximately 2000 hours of observations are required for a detection, and the EoR project has recently finished the collection of this enormous data volume. Now, the challenge is in finding the EoR needle in the haystack of data.
He will cover the methodology used by the project, and describe some of the associated challenges. This includes the approach for calibration, interference avoidance, foreground modeling and subtraction, imaging and power spectrum estimation. The unprecedented dynamic range requirements to detect the EoR-signals have required several novel processing methods.
All are welcome, if you would like to attend please register below.