Career destroyed over battle of the planets

16:00, Feb 12 2012
Steve Swaving
DISPUTE: Steve Swaving applied to do PhDs on the composition of stars and introducing water to Mars

In 2006, media reported that New Zealand astronomers had helped discover a "super earth" – an icy planet the size of Neptune, 9000 light years away.

Unlike Neptune with its thick layer of gas, this planet, unimaginatively named OGLE-2005-BLG-169Lb, was more like a large version of Earth with a rocky, icy interior. According to those involved, it was a step closer to finding an Earth-like planet that could support life.

"Ten years ago, such a finding would have been unthinkable," said Auckland University associate professor Phil Yock, part of the international collaboration.

But a few months later, a member of the three-person team that worked on a computer analysis of the discovery quit, alleging the analysis was flawed.

Masters student Stephen Swaving, who had helped develop the computer codes used to analyse data from telescopes around the world, claimed Yock had misrepresented the team's findings, their results had been manipulated, a vital code was malfunctioning, and they should not have been included as co-authors on a paper published in The Astrophysical Journal.

They were explosive allegations, and the fallout for Swaving was devastating. He was blocked from doing his PhD, and his pleas for an independent inquiry were ignored.


Swaving has the support of several academics, some now retired, including senior members of the physics department. Professor Howard Carmichael, holder of the Dan Walls Chair in theoretical physics, believes Swaving has been discriminated against.

"[The allegations] strike me as very plausible and worthy of an investigation. The university has rules about how they expect their researchers to behave," he says.

Carmichael chaired the committee which approved Swaving's PhD application, but it was blocked by Vice-Chancellor Stuart McCutcheon.

"It's unheard of. It's entirely unacceptable for a university to treat a student this way."

Carmichael suspects the university doesn't want to dig too deep because of Yock's involvement with the MOA (Microlensing Observations in Astrophysics) project, a joint initiative with Japan which has received more than $1 million of government funding. He believes the university fears any scandal impacting on its image, or even the country's.

Indeed Yock says Swaving has brought the country's scientific community into disrepute with his allegations. "They have no basis in scientific fact," he says.

"All of the work done by myself, [the other researcher] and Stephen Swaving when he was working with us, was done with the best of intentions following the best scientific principles."

Six years on and Swaving has not had the chance to put his evidence before an independent inquiry. He has finally decided to go public with his claims.

Swaving, 56, studied astrophysics late in life, after careers as a commercial pilot and a property developer.

"I love astrophysics," he says. "I've been looking at stars ever since my grandfather gave me a telescope when I was 13 or 14."

In 2005 he joined Yock's research team, and Yock became his master's supervisor. The team was involved in international collaborations to find new planets outside our solar system, using a technique, first proposed by Albert Einstein, known as gravitational microlensing.

Put simply, when a massive object such as a star crosses in front of another star, the object's strong gravity bends the light rays from the more distant star and magnifies them like a giant lens.

The magnification can be up to 1000 times, increasing the chances of finding smaller planets like Earth.

When a microlensing event happens, telescopes from around the world produce practical data, measuring the light, which is then compared with theoretical data to determine if there is a planet, its relative mass and distance from the particular star.

Swaving was tasked with finishing computer codes to produce theoretical light curves and magnification maps of areas of space, which had been started by other researchers.

In early 2006, the team was sent results from Ohio State University of a microlensing event the previous year, one of five collaborations around the world that had observed the event.

The Auckland team's job was to help verify the data.

Swaving claims Yock wanted the results analysed by a certain deadline so the team could become co-authors on an article about the discovery, along with more than 30 others.

But one of the analysis codes appeared not to be working properly. "This caused huge problems because we were now desperate to produce results," Swaving says.

For analysis the group had 420 files of data, containing 320 million permutations of the parameters used to find the undiscovered planet.

The team had been sent seven plots by the Ohio astronomers, describing where they thought the planet was. At this point, Swaving says, those plots should have been compared to plots produced from all 420 of Auckland's files.

Instead, he says, Yock and the third member of the team, a post-doctorate student, "cherry-picked" files that contained data that would match the Ohio results. An email told the main authors "all possible solutions" had been searched for and the results were "quite similar".

Swaving says: "For us to say we had analysed this event we had to analyse all 420 files. We analysed six and ignored the rest.

"This is astronomically bad. To verify somebody else's work, you can't take their work and pick your results to match. You have to analyse it independently."

Swaving believes the failure to analyse the full set of files resulted in an apparent match of the team's results to those of Ohio, but he claims that was not the case. A code developed by the third researcher had malfunctioned, he says, which caused a significant amount of data to be corrupted. The researcher did not respond to requests for an interview.

Swaving did not raise his concerns immediately, as it was not until later that he realised what was going on and, as the junior member of the team, was reluctant to question Yock's methods.

He says he is not questioning the work of the Ohio team.

"I've seen nothing to convince me...that they had done anything wrong. I believe the planet was where they say it was – roughly. What I'm saying is we misrepresented our support for that claim."

Yock rejects the criticism.

"The Ohio State group chose the set of values for these parameters, which are perfectly reasonable, and to compare their results with our results of course we set those parameters at the same values, otherwise we couldn't make a comparison. Anyone doing this research would do exactly the same thing."

And he says the fact there were problems with the codes is not unusual. Whenever data is computerised, it introduces error, which is why quality control parameters are used, he says. The errors were solved, leading to agreement between the results.

"You try various procedures and see whether the code runs fast or slow, whether it's accurate or inaccurate. It's a trial and error process. All the other people doing these analyses have gone through similar stages. For Stephen to leave the group and start broadcasting his claims that we're not following scientific procedures ... brings science in New Zealand into terrible disrepute."

Former Auckland University tutor Graham Hill, an astrophysicist with 29 years' experience as a research scientist at the National Research Council in Canada, believes Yock's contention that it was a valid search is "laughable".

"A lot more should have been done. It's like you've got a room full of suspects, and somebody says `those are the two over there'. Your intention was to look at the whole room, but someone's told you to look over there and that's what you've looked at."

Chris Tindle, head of the physics department at the time the dispute arose, also believes it wasn't a genuine search. "Cutting corners is probably a good way of putting it."

Yock has had a celebrated career, helping develop methods for discovering planets, black holes and other celestial "dark matter". He was made a Member of the New Zealand Order of Merit for services to astronomy in 2010.

But he also has a history of conflict at the university. In 1990, he made a string of complaints against the head of the physics department. The investigating committee found most arose out of Yock's "chagrin" at being overlooked as leader of a project to investigate a supernova.

It upheld none of the complaints and recommended Yock no longer be a member of the physics department and not supervise graduate students unless invited to.

Yock sent the Star-Times a copy of a 2008 Royal Society report into Swaving's allegations, which cleared him of any wrongdoing.

Swaving had complained to the society because of the large sums its Marsden Fund had given to microlensing projects Yock was involved in.

The report said there was no evidence that Yock had tried to modify his team's data to fit the Ohio results and said he was an "experienced and distinguished scientist ... who cannot have achieved his international reputation without peer scrutiny".

However, it appears the investigating committee did not look at Swaving's data or interview other academics with knowledge of the case, speaking only to Yock, and the report was later withdrawn on the society's own legal advice.

A new investigating panel was convened, but the inquiry has stalled because Swaving is concerned it is not independent.

Swaving claims that after the incident with the Ohio group's data, Yock indicated they would run a more thorough search using better magnification maps.

But then he changed his mind. Swaving left the team, and continued his Masters thesis with new supervisors. But his troubles with Yock were just beginning.

A university internal inquiry later found Yock had obstructed Swaving's access to a bank of computers, and deleted Swaving's research material, including hundreds of magnification maps, from the database.

Ron Keam, an associate professor in the physics department who examined Swaving's Masters thesis, wrote a letter to the vice-chancellor in which he said Yock had been specifically warned not to delete the material as it would constitute an "act of war".

Keam had many years earlier supported Yock's application for an academic position in the physics department, but he wrote that he now believed Yock should not be allowed to supervise research students. He believed the destruction of the material was a dismissable offence.

The university says disciplinary action was taken against Yock, but it will not give details.

Explaining why he deleted the material, Yock says his team was running out of storage space, he had advised the head of department several times he was going to do it but got no reply, and the maps were easily replicated.

Swaving believes the material was deleted because it could call into question the team's earlier results.

In 2008 Swaving wrote to the Astrophysical Journal about his concerns, asking that he be removed as an author. The Journal replied that most of Swaving's claims were a matter for the university, and as it was two years since the paper was published, "your protest comes too late".

The Journal raised the issues with the paper's main authors, and was assured "that the senior members of the collaboration acted in good faith in doing the research". Swaving does not dispute this.

Swaving gained his Masters in Astrophysics with first class honours in 2007, and applied to enrol for his PhD, studying the composition of stars. The new head of the physics department, Stuart Bradley, declined the application because he claimed Yock was the only suitable supervisor and the pair could not work together.

Swaving applied again, this time proposing to study what would happen if water was introduced to Mars, and nominated supervisors. Now McCutcheon stepped in, saying the application was to go no further until Swaving's disputes with the university were sorted out.

Swaving appealed to the ombudsman, who eventually said it was a matter for the university.

Over the ensuing years Swaving's supporters have written to the university council, successive ministers of tertiary education, the Royal Society – even the prime minister – without getting the independent inquiry they want.

McCutcheon told the Star-Times Swaving had launched at least five complaints against members of the university. He says Swaving "generally failed to co-operate" with the reviews of those complaints, or when a review was completed, "disagreed with the outcome and added the reviewer to his list of villains".

Because of this, McCutcheon says, he told Swaving he was not prepared to let him enrol in the PhD programme until he co-operated with the review of his outstanding complaints. "This he has failed to do."

Swaving denies he lodged five complaints, saying that at the time McCutcheon stopped his PhD, there were only two outstanding.

He also says he would co-operate with an impartial investigation.

Carmichael says Swaving is obstinate, and his approach has allowed the university to cast aspersions on him while allowing the dispute to continue festering, but he admires him for refusing to back down.

"I think initially they thought he would simply go away as most students would have done. But given his age and his financial situation, being more or less self sufficient, he's decided he's not going to be screwed."

Swaving says he wants to be able to do his PhD, and wants independent investigations into his claims.

"I want justice for me, and, just as importantly, for the many academics who have had the courage to stick their necks out to help me."

Sunday Star Times