Counting down to the launch of ASKAP 36 full survey science, Stage 2, first single-beam image with the full ASKAP array

The countdown to ASKAP 36 full survey survey science is a campaign designed to share ASKAP's final commissioning sequence with the community, in the lead-up to full operations. As you might imagine, the process of bringing to life such a complex science instrument doesn't provide a single moment of victory but rather a series of exciting steps, building towards full capability. The nine stages on the countdown, show a sequence of increasingly complex operations. The milestones are being produced in turn by the ASKAP Survey Science Teams or CSIRO observatory staff.

ASKAP’s greatest strength is its ability to image a large area of sky. The Phased Array Feed (PAF) technology allows this by providing 36 primary beams that can be steered in different directions. However, when testing the basic performance of the telescope after integrating all 36 antennas, it makes sense to start simple – using a single beam pointing along the optical axis of each antenna to image a well-known source. This single-beam image is the second stage of our countdown to full survey science.

The image quality of an interferometer array is determined by the total number and arrangement of its antennas. As described in our report on the first fringes across the full array, having all 36 ASKAP antennas provides 630 baselines, giving a large amount of information with which to reconstruct an image. Our very first observations with the full array were of the trusty calibrator source PKS B1934-638 – a very bright radio galaxy that is also very far from Earth. It is so far away that it appears “unresolved” to ASKAP – it looks like a single bright point, much like an optical star to the human eye. This point-like nature makes PKS B1934-638 a good test of how the telescope performs from a technical perspective – we understand what the signals from the telescope should look like and can verify that everything is working as expected. However, point sources do not make the most interesting pictures! For the first single-beam image release, we decided it would be better to look at something closer to home.

The radio galaxy Fornax A is one of the brightest in the sky and has the advantage of being much closer to our own galaxy the Milky Way. If we could see Fornax A in the sky with our naked eyes then each of its two lobes would be about the same size as a full moon. It is close enough that ASKAP can resolve extended structure, demonstrating the telescope’s ability to detect diffuse emission as well as compact objects. Furthermore, as each of the 36 available ASKAP beams can see a region of sky that is about 4 full-moon-spans across, the source can be easily observed within a single beam. Observations of the source had been requested by our international science teams and the CASS commissioning team decided it would be a good target for our single-beam demonstration.

During these first stages of the countdown process it makes sense to examine ASKAP’s data from a variety of angles. We have developed a custom high-speed imaging software package known as ASKAPsoft to process survey data, but we write the data in a standard format. This can be read by standard tools, allowing small ASKAP data volumes to be imaged without a supercomputer and providing a valuable verification of the data quality. With this in mind, the ASKAP operations team observed Fornax A for 10 hours on the 3rd of April 2019 and commissioning team member Emil Lenc processed the data through standard tools to make the image shown below.

This image was made from a single beam observation of NGC 1316 (Fornax A). It displays the classic “double lobed” shape that we believe arises from oppositely-directed jets of material being ejected from the accretion disk around a central super-massive black hole. The telescope observed NGC1316 for 10 hours at a centre frequency of 944 MHz with 288 MHz of bandwidth. Situated at the Murchison Radioastronomy Observatory, ASKAP has almost no radio frequency interference in this band and very little flagging was required. The bandpass was calibrated using an observation of PKS B1934-638, but no other calibration method has been used. The data were de-convolved using a combination of Hogbom CLEAN and the maximum entropy method. No Taylor terms were used, but some W-terms were required due to the large primary beam size (nearly 2 degrees in diameter). This image contains all of ASKAP’s longest and shortest baselines, making the synthesised beam size 12”x10”. The noise in the background of the image has an RMS of 25 uJy/beam, which is close to the thermal expectation for robust zero weighting.​