health & science
What MBIE told Woods on super-scope
Last-ditch talks between a mega-telescope's supporters and New Zealand government officials seem to have pushed officials more firmly towards withdrawing all funding, despite the estimated membership cost dropping to less than a million dollars a year. MBIE staff told Megan Woods another $3 million annually would have been needed to boost astronomy, and that there were better ways to spend the money, newly-released documents show.
MBIE has revealed what it told science minister Megan Woods before New Zealand dropped all involvement in an international mega telescope, the Square Kilometre Array.
Government officials are in Manchester this week explaining to other countries the decision to pull out of the telescope-building effort after 2020.
The SKA will be a vast array of radio antennae and dishes in the deserts of Australia and South Africa, working together as a single, super-sensitive instrument and generating an almost unfathomable amount of data.
Being involved in the data-sorting effort would have benefitted New Zealand and helped it find a place in the digital world, argued AUT and others who've been involved in the telescope project, following several years of government support for local entities working on the ICT back-end design.
But MBIE staff told Woods before a Cabinet meeting on the subject that designing a taxpayer boost for the IT industry around access to an astronomy telescope was “not viable”, that the benefits would be captured by a "narrow range" of commercial interests, and that efforts by the telescope's IT proponents to persuade officials to stay involved had left officials less convinced of the project's merits than before.
That rather cutting statement was in a briefing to Woods dated May 30 2019, before New Zealand attracted international attention by deciding to pull out completely from the SKA before construction started.
The telescope is currently slated for launch in 2027, about a decade later than originally planned.
When Cabinet's decision to withdraw was announced, last Monday, MBIE said it would release its briefing to the Minister on Friday at the latest. The briefing was originally expected to be shared earlier on Friday but landed at 4.55pm, although the timing, on the cusp of the Friday night news dead-zone, did not prevent the NZ Herald from running a story.
When Woods decided just over a year ago not to take full membership of the international consortium that will build and run the SKA, she told officials to look at the cheaper option of associate membership.
It’s still not quite clear what that would have entailed, since the terms of associate membership haven’t been set by the international telescope organisation, SKAO, which is head-quartered in Manchester.
The latest briefing from MBIE officials to Woods shows that taking full membership - the option Woods already rejected - would have been cheaper than the anticipated $20-30m estimated by MBIE when full membership was discarded.
Apparently, efforts to keep the vast international project under its self-imposed price cap (and get more countries more firmly on board) have resulted in shrinking both the telescope's computing power and number of dishes, to the point where the energy costs to run the finished product will be lower than MBIE originally thought.
The latest estimate is that full membership would cost New Zealand about $18m over a decade, and associate membership maybe half that, $9m a decade or $900,000 a year.
But MBIE advised Woods that even that lesser figure was not worth it.
MBIE staff felt another $2-3m annually would be needed to build New Zealand’s astronomy field to the point where Kiwi astronomers could take full advantage of using the telescope.
That's because SKAO will dish out most of the time using the telescope, and its enormous screeds of data about the universe, to astronomers in countries that are paying to build and run the dishes. Only a small fraction will be left available for projects led by astronomers from other nations.
MBIE has been concerned that New Zealand doesn’t have a sufficiently large pool of world-class experts doing the type of work that would be advanced by using the SKA to justify paying for $2m, $3m or even $1m worth of time on the telescope per year.
Radio astronomy in New Zealand suffered a set-back when a leading Victoria University astronomer, Melanie Johnston-Hollitt, returned to her native Australia and took some of her team with her. While AUT has a radio astronomy team led by a top scientist, Willem Van Straten, and Victoria University of Wellington has also gained new talent, MBIE recently described the field of radio astronomy here as small with “few of any standing internationally”.
One of the main, contentious issues between MBIE and the SKA’s New Zealand supporters (and beneficiaries) was whether the project nevertheless stacked up as an IT investment, independent of astronomy.
The latest briefing says that if New Zealand was going to invest in growing its super-computing prowess, it would be better to do so in another field such as climate science or bioinformatics.
New Zealand is currently spending about $1.2m a year in operating and design costs for the telescope, money MBIE says can now be re-allocated.
For several years, partially government-funded ICT design contracts have been granted to members of a group called the NZ SKA Alliance, comprising AUT, private IT companies including Catalyst IT and Open Parallel, and others, including Massey University. The group's members have been working on projects including the signal-processing powers needed for the finished super-scope to sort through huge amounts of data and make sense of it for astronomers. Had New Zealand’s involvement gone ahead, the group's members would have been in line to bid for construction contracts during the building phase, and continue what the group says have been fruitful international partnerships.
But MBIE said the really valuable work for building up New Zealand’s IT expertise has already happened during the design phase. Joining the construction phase only made sense if New Zealand wanted to buy a long-term share in the radio telescope, MBIE told Woods, adding that the total spending would need to include a $2-3m spending boost for astronomy to make joining worthwhile.
“It is difficult to make a case for a $3 million dollar astronomy investment outside of a contestable process," officials said in the briefing.
"There are a number of spending pressures on the Government, and a number of potential investments we would recommend you make in the science system.”
The briefing was signed by the ministry's international science partnerships manager, Simon Rae, who has been leading the telescope project for the New Zealand government.
MBIE - correctly - anticipated Woods' decision to pull out of the SKA entirely would be criticised by the telescope's "vocal" New Zealand "stakeholders", the members of the NZ SKA Alliance.
The ministry's briefing told Woods that MBIE staff had spent time listening to the local SKA group's members since her first announcement in May, though the discussions had the opposite result to what the telescope's supporters would have wanted. “While we will be criticised for not consulting further, we have had substantial exchanges with stakeholders engaged in SKA design since your earlier decision, and these have not been convincing," MBIE said. “We have had a series of robust exchanges with stakeholders on our advice, particularly from an ICT perspective. If anything, those exchanges have left us less convinced of the merits of the investment from an ICT perspective."
Elsewhere in the briefing MBIE told Woods: “One of the key arguments we used when we provided advice to you on full membership of the SKA Observatory was that investment in the SKA performed poorly relative to alternatives .... we can now express this choice more sharply: if we were to invest a certain sum in high performance computing research capability, would we choose to invest it in this way?”.
MBIE concluded that the answer was no. The share of international construction contracts New Zealand would win if it joined the telescope, and the likely nature of the contracts, were clearer now than when the economic costs were first estimated by the consultancy Sapere, the briefing said. Whether the Sapere report was adequately considered by MBIE has been a major bug-bear for the SKA's supporters, who feel the benefits were under-rated.
“Benefits are likely to be captured by a narrow range of commercial partners ... a significant portion of construction activity will fall outside standard definitions of R&D, and/or will be aimed at developing an essentially saleable product or service. We would normally expect this type of research to be funded by industry,” MBIE said.
“If (high performance computing and data science) was a priority we would arguably be better served by an investment supporting existing science strengths … such as bioinformatics, complex organic system modelling, or climate science,” it said.
“Participating in the procurement process for the construction of a radio telescope is not a viable way to design an ICT research investment.”
AUT says its astronomers are still involved in the design phase.
But New Zealand's official involvement in the telescope is all but over.
New Zealand was a founding member of the non-profit organisation set up to design the radio telescope. Before that, the former Labour, then National, governments hoped this country would host some of the dishes and other infrastructure, but a joint bid with Australia missed out when the prize was split between Australia and South Africa.
New Zealand continued its involvement by contributing to the cost of ICT design, partly to further friendly relations with Australia (a SKA member), and partly on the basis that there were spillover benefits to the ICT sector from the cutting-edge computing collaborations.
MBIE calculates the cost of that investment was justified by the creation of one spin-off company alone, Waikato's Nyriad.
Full membership of the SKA would have been a much longer-term commitment, and would have involved buying a share of the radio telescope for nearly 50 years, according to Woods' notes for the recent Cabinet meeting. "I do not believe that this purchase would be the best investment we could make in ICT research at this time," read her preparatory notes.
Recent years have been fraught with local acrimony over the telescope.
Last week, MBIE told Newsroom that Cabinet’s decision to completely withdraw from the project on June 25 “was not related to the conflict between stakeholders in the project".
That conflict has been intense and personal: Senior AUT staff sent an unusual mental health smear and other emails attacking the conduct of an outspoken critic of the telescope’s funding, the University of Auckland’s Richard Easther, in their efforts to persuade journalists and others that Easther’s criticisms weren’t credible.
The New Zealand SKA Alliance’s director, AUT’s Andrew Ensor, later apologised to Easther for telling a journalist he should seek medical help.
MBIE has always maintained that Easther and other astronomers who questioned the size of the spending weren't the deciding factor.
There was a "full range of views" within New Zealand astronomy, says their latest briefing.
Earlier, MBIE itself copped a number of scathing emails from members of the SKA group, especially over funding and its advice to the minister.
With the decision now made to pull out completely, New Zealand has one final bill to pay: $460,000 for its 2020 contribution to the design phase.
In the meantime, MBIE officials will be fronting the Government's decision at the international SKA board meeting this week, explaining our withdrawal to other countries in person.