INSIGHT
LESSONS LEARNED FROM BUILDING A TELESCOPE IN THE OUTBACK BY ICRAR AND INAF A few months ago, Italian and Australian researchers completed the deployment of a full 256-antenna SKA-Low station at the Murchison Radio-astronomy Observatory, the future home of SKA-Low in Western Australia. The year-long work on site taught them valuable lessons about the behaviour of antennas on the ground, with a few extra insights thrown in... In late 2018, the International Centre for Radio Astronomy Research (ICRAR) and Italy’s National Institute for Astrophysics (INAF) set themselves an ambitious goal to simulate, deploy and commission a full station of 256 SKA-Log periodic Antennas (SKALA) within a year, in order to mitigate the risks associated with station-level calibration that had been identified in the lead-up to the SKA’s System Critical Design Review. “If you can’t calibrate the individual stations properly, you don’t know how accurately your station beam has been formed. This will impact on the overall SKA-Low array beam, and hence how good your observation is, so it’s a vital part of preparation for the SKA. This is especially true for the sensitive science SKA-Low will conduct, such as work on the Epoch of Reionisation, which seeks incredibly faint signals hidden behind much stronger ones,” said Professor David Davidson, Director of Engineering at ICRAR. By May 2019, a smaller station consisting of 48 SKALA antennas version 4.1 was on site, and
commissioning work commenced almost immediately. By the end of the year, the full station was in place, and data was being captured and processed. The demonstrator, known as the Aperture Array Verification System 2.0 (AAVS2.0), was deployed by researchers from ICRAR and INAF, as part of SKAO-led bridging activities. Its 256 antennas are distributed on a wire-mesh ground, with a maximum distance between them of 38 metres.
“It would be impossible to experimentally characterise more than 130,000 antennas in such wide frequency range, so scientists have to rely on antenna models. At the same time, we cannot run electromagnetic models without having proved at least for some frequencies and for some antennas that their results match with high fidelity to the measured responses. This is where the drone measurement entered in the game.”
The array layout is semi-random, with some antenna positions fine-tuned for mechanical and practical reasons.
The researchers were able to achieve excellent agreement between different simulations for a very large array of antennas using different commercial electromagnetic simulation tools. The simulations took almost a year and close to 20,000 hours of computing to complete.
The team then simulated the electromagnetic behaviour of the array components, to predict the station beams (the method by which SKA-Low will view the sky) and other antenna parameters key to understanding the performance of the telescope. They assessed the interference level between the individual antenna beams due to the large number of antennas within such a small diameter, creating socalled ‘Embedded Element Patterns’. The full station beam generated by these patterns differs from the ideal stand-alone one, which is computed with simplified models not including the interactions between antennas. “Having an accurate field validation of the antenna numerical patterns computed in very realistic scenarios, considering for instance the mutual coupling between antennas, is very important before trusting the simulated model,” explains Dr Pietro Bolli, Antenna Researcher at INAF.
The simulated beam patterns were then verified with the help of a drone following the initial roll-out of the array, in collaboration with the National Research Council of Italy and the University of Malta (see our previous article in Contact 1). The team is now working to determine whether the variation in element patterns will allow sufficiently accurate and rapid station-level calibration for SKA-Low. “The next steps are to take all the data we’ve collected and progress the commissioning of the instrument, including making some observations,“ said Professor Davidson. “We will also direct efforts towards a more suitable design maturation for the industrial production phases, as well as the (...) integration of the other parts of the telescope that are in the competence of various working packages such as Telescope Management, Central Signal Processor, Signal and Data Transport, etc.,” said Jader Monari, SKA-Low Italian Programme Manager. And so the challenge continues! Watch the video summarising the work that’s been happening and visit our Youtube channel for interviews with the team.
Completed AAVS2.0 station from overhead. Credit: ICRAR-Curtin University 18
C O N TA C T | JUN E 2 0 2 0
