The Square Kilometre Array (SKA) radio telescope project is, in terms of scale, undoubtedly the biggest astronomy programme in the world today.

What was once a dream, and then a hopeful project, is now officially structured as the SKA Observatory (SKAO), an international organisation with nine full member countries – Australia, China, Italy, the Netherlands, Portugal, South Africa, Spain, Switzerland and the UK – with several other countries involved in the global endeavour and in the process of joining it.

It has a headquarters overseeing two major arrays, covering different but complementary frequency ranges, which are currently being developed. The head office is in the UK (at the world-famous Jodrell Bank Observatory, in Cheshire), while the two world-class telescope arrays are being rolled out in South Africa and Australia.

The physical scale of, and the wide frequency ranges that will be covered by, the SKA telescopes are because, to put it simply, the greater the area of the radio receivers, working as an interferometer (that is, together, as if they were one instrument), the fainter the signals they can detect, and with greater accuracy. This will allow the SKAO to look far deeper into the universe than any previous radio telescope, which translates into looking far deeper back into time, right back to near the birth of the universe.

The Australian array, located at Inyarrimanha Ilgari Bundara/the Commonwealth (of Australia) Scientific and Industrial Research Organisation Murchison Radio astronomy Observatory, will cover the frequency range from 50 MHz to 350 MHz, and take the form of 131 072 log-periodic antennas, and is designated the SKA-Low array.

The South African array will cover the frequency range from 350 MHz to 15.4 GHz, with a goal of 24 GHz. The South African array is designated SKA-Mid, will be composed of 197 dish antennas, located in the Karoo region of the Northern Cape province, and is the focus of this report. Both arrays will be deployed in phases, known within the SKAO as ‘Array Assembly’ (AA) phases, and identified numerically.

Coming Together

SKA-Mid is currently in the AA 0.5 phase (and the SKA-Low in Australia is progressing at the same pace). This involves the construction of four dishes. “The very first dish is being shipped to South Africa at the moment,” SKA-Mid telescope director Dr Lindsay Magnus tells Engineering News & Mining Weekly exclusively. “It will arrive in the country in December and be shipped to the site. But the builders’ summer holidays mean that its erection should start in January.”

The very first SKA-Mid dish has actually already been set up, but in China. The dish antennas for the SKA-Mid are being produced in that country, as one of the “fair work return” work packages assigned to that country, as a full member of the SKAO. All the member countries have been assigned SKA work packages. The dish in China will serve as a test and evaluation unit.

“Each of the dishes will be fully fitted out with all its detection, operational and control systems as it is built,” he explains.

“The idea is to test the dishes as much as we can, so it’s in our interests to deploy as much functionality in each dish as we can. For example, we need to see if we have to change any aspect of the dish design or its systems. Anything we might detect will be communicated to the engineers in China, where they will try and replicate the issues on their dish, and then adjust the design to deal with the issue. The AA 0.5 dishes are really preproduction units; the final design will only be signed off following the test and evaluation of the four initial dishes.”

Each of the AA 0.5 dishes will be subject to individual test and evaluation immediately after it is completed. The expectation is that there will be no need to adjust the design of the dishes, or that only minor adjustments will be required. But the only way to be confident that a dish design is truly optimised is to build it for real.

What the SKAO desires is dishes that can be pointed as quickly as possible (to be able to respond to important but transient astronomical phenomena) while maintaining their observational stability and performance. The length of this test and evaluation period will be determined next year.

Once two dishes have been completed, it will be possible to start testing their pointing accuracy and the accuracy of their dish surfaces (this latter is known as holography). These tests require the two dishes to be operated together, as a single unit.

Distortions in the dish panels, imperceptible to the human eye, can degrade performance. Such distortions could result from manufacturing or shipping errors but could also be the result of stresses imposed by design flaws in the dish support structures. And once all four dishes are completed, it will be possible to undertake the signal processing tests.

Phase AA 0.5 will be completed, with all four dishes erected and fully equipped, by early 2025. “The schedule has lagged a little, but it’s still within the project limits, so the final delivery date for the total project is unchanged,” he notes.

Fair Share

As previously mentioned, each full member country of the SKAO gets work packages on the project. South Africa has received four main work packages.

The main local work package is the construction of the infrastructure for the SKA-Mid array.

This includes roads, power connections, and fibre and other signal networks. Also important is that the South African Radio Astronomy Observatory (SARAO) has been contracted to do the testing of both the individual dishes and the different phases of the array, as they are deployed. This is a services contract. “SARAO has a lot of experience in this, from MeerKAT, so they are the best people for this,” highlights Magnus. (MeerKAT is South Africa’s radio telescope array and will eventually be integrated into the SKA-Mid telescope.) SARAO is also contributing to different software design packages.

The fourth major South African work package relates to the dish receivers. Each dish has several of these radio receivers, and EMSS Antennas (a South African company) has been contracted to build one receiver for each and every SKA-Mid dish (this company has built two of the receivers fitted to all the MeerKAT dishes). The company will also provide the associated receiver systems services.

Regarding the infrastructure work, the SKA-Mid team established the policy of involving as many local stakeholders as possible, in the region around the telescope site. “It’s a rather poor region,” he points out. “We’ve negotiated that any contractor who does work for us in the Karoo adheres to our local procurement and labour supply requirements. This is part of our determination to be ‘a good neighbour’ and uplift the local economy around the telescope site the best we can.”

The team has set up a priority hierarchy of four ‘tiers’. Tier 1 is the four towns surrounding the array site (Carnarvon, Brandvlei, Vanwyksvlei, and Williston); whenever possible, contractors should source as much as they can from Tier 1. Should they be unable to do so, then they should try Tier 2, which covers the districts in which the four towns are located. Then Tier 3, which is the Northern Cape province. Finally, Tier 4 is the whole of South Africa.

“This has involved the establishment of a comprehensive labour database and a procurement database,” he reports. “We’re currently meeting all the labour deliverables from the four towns.”

Follow-up Phases – and MeerKAT

AA 0.5 will be followed, from the start of 2025, by AA 1.0. This phase will see the erection of a further four dishes, to the production standard (to take the total number of dishes to eight). The completion of AA 1.0 (by early 2026) will take the SKA-Mid instrument to the point where its scientific performance can be validated, and so this will mark the start of scientific testing. (AA 0.5 test and evaluation will very much be focused on engineering tests.) Scientific testing will involve pointing the dishes at ‘test’ parts of the sky, where the scientists expect to get certain results, and see if the array delivers those results.

“After AA 1.0 is formally completed, we’ll just add more and more dishes, at quite an impressive pace,” he explains. “Each new dish will first be plugged into a test array, to validate it, and only after that will it be integrated into the operational array. The main constraint to dish deployment will be the development of the ‘back-end’ signal processing systems.”

At the beginning of 2026, AA 2.0 will start. The main focus of this phase will be on the transfer of the array computing system from the Karoo site to Cape Town. This process will take place throughout the whole of the year, to avoid disrupting the telescope’s operations. The AA 2.0 system will be configured to work with 64 dishes and should be operational by early 2027. At that point, the first data will be released to the science community for science verification.

The design baseline is for the SKA-Mid to have 197 dishes (including the MeerKAT dishes), the achievement of which would mark the completion of AA 4.0. But the funding for AA 4.0 has not yet been secured. So, an interim array design, in-between AA 3.0 (an early defined milestone with 133 dishes) and AA 4.0, has been developed, and should be complete by early 2028. This interim design is designated AA *, and corresponds with what can be built with the funding that should then be available.

In the meantime, the SKAO director- general and the SKAO leadership are actively seeking additional funding, as the ambition remains to build the full design baseline for both telescopes. “Indeed, it has been decided to establish all the infrastructure required for AA 4.0, that is, for 197 dishes, so that, when the funding is secured, we can simply add dishes,” he notes.

As intimated above, it is, and always has been, the plan to integrate MeerKAT into SKA-Mid. Originally a 64-dish array, a joint project between SARAO and Germany’s Max Planck Institute for Radio Astronomy is adding up to 20 dishes, to bring MeerKAT’s total up to 84. A highly successful and much regarded instrument, it is intended to be integrated into the SKA at some point after the completion of SKA-Mid’s AA 2.0.

“It is imperative that the scientific work being done by MeerKAT is not disrupted,” affirms Magnus. “So, the incorporation of MeerKAT into SKA-Mid will be a phased process and something that we actively discuss with our partners at SARAO. We will have to manage the transition period as well as possible.” The transition will include the refitting of the MeerKAT dishes with SKA-Mid systems, to make them compatible.

Down Under, and up North

Work on SKA-Low in Australia is moving forward in parallel with that on SKA-Mid. The Australian team is, like their South African colleagues, busy with the building and deploying of the support infrastructure for their antennas. The third and last prototype for SKA-Low, known as AA Verification System (AAVS) 3, is now complete. The SKA-Low instrument will be composed of a series of “stations”, each with 256 antennas. AAVS 3 is a complete station and is currently being tested. The AA 0.5 deployment of the SKA-Low will start early next year and the subsequent phases will be rolled out in parallel with those of the SKA-Mid telescope.

And, in the UK, the SKAO head office is becoming fully operational. It is responsible for programme delivery, science operations and engineering operations. Programme delivery includes overseeing of the manufacture of the telescopes, ensuring the correctness of the designs, and that the contracting is done correctly. The science operations and engineering operations teams are starting to take shape. The core science team will be run out of head office. “We have learnt from MeerKAT that you need a science operations team when the instrument is being built, and involved in its construction,” he highlights. “That’s the way they learn how to use it!”