People who work for climate science denial thinktanks tend not to spend all that much time worrying about getting stuff into scientific journals.
Perhaps because it’s easier, people who are paid to tell the public and policy makers that human-caused climate change is overblown bunk would rather pump out newspaper columns, do softball interviews or push out their own self-published reports. There’s a lot less scrutiny in that kind of public relations.
So when two staffers at Australia’s Institute of Public Affairs managed to get some “science” into a journal earlier this month, there was much delight in conservative media outlets, together with a distinct lack of any genuine scepticism.
“Global Warming Is Almost Entirely Natural, Study Confirms,” wrote Breitbart. “Advanced Computer Models Suggest Most Global Warming Is From Natural Forces,” said the Daily Caller.
One of the authors, Jennifer Marohasy, took to the Spectator to claim her research had shown that recent global warming was almost entirely natural. The web traffic behemoth the Drudge Report also linked to Marohasy’s article.
None of the writers bothered to ask a single other genuine climate scientist for their view on the paper.
I asked five. They variously summarised the research as “junk science” and seriously flawed. Oh dear.
Scientists including Dr Gavin Schmidt, director of the Nasa Goddard Institute for Space Studies; Prof Steven Sherwood, deputy director at the University of New South Wales climate change research centre; and Prof Piers Forster, director of the Priestley international centre for climate at the University of Leeds, have pointed out to me serious flaws and errors in the paper’s methodology.
It also appears that the lead author, the IPA’s John Abbot, claimed an academic affiliation to James Cook University that, according to that university, had expired more than six months before the research was submitted to the journal.
I emailed Abbot and Marohasy to ask them to respond to the key points but didn’t hear back.
The gory details
So what did Abbot and Marohasy do? Here’s where we get into the gory detail.
First, the pair used what’s known as proxy temperature records. These are estimates of past temperatures going back as far as 50AD that have been estimated from analysis of things such as tree rings and lake sediments.
There are hundreds of these temperature records but the authors chose only six, and they don’t say why they chose the ones they did.
They then fed these temperature records into several pieces of software and finally into another piece of software that performs a “machine learning” process. Out the other end comes a series of temperature reconstructions, from which Abbot and Marohasy make two claims.
First, they claim that their resulting data shows the world would have warmed by almost as much as it already has, even if the industrial revolution had not happened and we had not added any extra greenhouse gases to the atmosphere.
Second, they claim that if you were to double the amount of CO2 in the atmosphere then you would eventually see planet-wide warming of just 0.6C. Their estimate of this “equilibrium climate sensitivity” (ECS) is much lower than other studies (the UN’s Intergovernmental Panel on Climate Change says, for example, that studies show the likely range will be between 1.5C and 4C).
So what did the scientists think?
Schmidt told me by email the paper was worthless “on a number of measures” and in his opinion demonstrated “what happens when people have their conclusions fixed before they start the work”.
Schmidt wrote that “conceptually this methodology can’t possibly work” because the way the authors had calculated the climate’s sensitivity had assumed that all the natural variability was part of the planet’s internal systems, rather than being “forced” externally by volcanic eruptions or changes in the output of the sun.
This, Schmidt said, was “in contradiction to their claims elsewhere in the paper.”
Schmidt also says “something went wrong” when Abbot and Marohasy digitised their results, meaning that for the northern hemisphere the data had shifted by about 35 years “so what they think is 2000, is actually 1965”. This meant that a huge part of modern warming had been missed.
Dr Benjamin Henley, of the University of Melbourne, has published several studies using proxy data to understand ancient climates. He says the paper should never have been published and should be withdrawn.
“The paper is seriously flawed and should be retracted by the journal,” Henley told me by email, pointing out several serious issues with the way the data had been used.
Henley questioned why only six “paleoclimate” records had been used, when a recent paper identified some 692 proxy records that could be used to determine temperatures.
He said the authors had not tried to verify their approach by comparing their data to actual temperature measurements – an “extremely unscientific approach”.
“The results are incorrectly interpreted and are not verified or even compared to observed instrumental data. The conclusions are not supported by the results.”
Methodology flaws
Forster reviewed the paper and told me he thought “the methodology is unphysical” because it simply took data and then extrapolated it rather than accounting for what was actually known to be happening in the real world.
Forster said the paper contained “fundamental errors” and gave me a detailed rundown.
For example, Forster says Marohasy and Abbot’s methodology assumes that all previous natural swings, or oscillations, in temperature that happened before the industrial revolution would continue up to present day.
In reality there a very few periodic or quasi-periodic oscillations in the Earth system. In reality most are random and not periodic. Volcanoes are the biggest issue. Their method assumes that any periodicity caused by volcanoes prior to 1880 has simply continued though to 2000. We know this is not the case and have a good handle on what volcanic activity has been – it was nothing like it was prior to 1880. Their approach is therefore completely unphysical. They also assume that oscillations seen at a given proxy location will also cause the same oscillation on global average or hemispheric average temperatures. We know this is not the case.
He said the authors had assumed most of the 1C warming that the planet had seen was caused by random variation, “but we know this is not the case.” He said the paper’s conclusion the ECS was only 0.6C could be ruled out entirely.
He said: “Attribution studies with careful statistics put random variability (of the climate) as contributing a maximum of around 0.2C. With 1C of warming already, and only halfway to doubling CO2, we can rule out ECS below about 1.5 C already.”
Sherwood wrote: “The analysis by the authors seems to work like magic.
“What is interesting about this fancy curve-fitting exercise is that the authors are doing exactly what mainstream climate scientists have falsely been accused of doing: extrapolating into the future from short past records.
“There is much evidence that recent warming is unprecedented, for example ancient ice in various mountain regions such as Peru that is now melting for the first time in millennia. Thus the authors’ conclusion is contradicted by direct physical evidence. Also, the authors are alleging that the climate can exert large natural swings in temperature but is insensitive to heating. This is a contradiction.”
Prof David Karoly, a climate scientist at the University of Melbourne, told me the study “appears to be junk science” and listed several major issues with the paper’s methodology.
He said the authors had not bothered to describe the detailed method they had used to calculate ECS, adding: “ECS is based on global mean temperature changes and cannot be estimated without the globally averaged temperature estimates first.”
“In my view the claimed conclusions are not supported by the methods used and the results are likely to be unreliable,” he said.
So none of this is a surprise.
The IPA, which does not have to reveal its funders, has long pushed climate science denial, while promoting fossil fuels and denigrating renewable energy. It’s what they do.
According to the journal manuscript, Marohasy and Abbot list their affiliations as the IPA and the Climate Modelling Laboratory.
The Climate Modelling Laboratory is a trading name linked to Marohasy’s personal business number.
James Cook University
Abbot also adds James Cook University (JCU) to his list of affiliations.
But a JCU spokesperson told me: “John Abbot is no longer affiliated with James Cook University. He was an adjunct senior research fellow between October 2015 and September 2016.”
According to the journal’s website, and confirmed by a journal editor, Abbot and Marohasy submitted their manuscript to the journal in 22 April 2017 – more than six months after Abbot’s affiliation with JCU had ended.
I also asked Abbot and Marohasy about the nature of the Climate Modelling Laboratory and the affiliation to JCU but have not had a response.
The journal editor who handled the Abbot paper, Dr Vasile Ersek, of Northumbria University in the UK, said he was “sorry to see it involved in a controversy” but said the article “was reviewed by two independent referees and neither found major flaws with the manuscript”.
As I’ve written on Planet Oz before, Marohasy has repeatedly claimed that Australia’s Bureau of Meteorology is deliberately conspiring to tamper with its temperature records with the express intention to make global warming look worse than it is.
Marohasy was also the editor of the IPA’s most recent climate book – a collection of essays from a line-up of climate science deniers and contrarians.
The IPA said the book had contributions from “some of the world’s leading experts”. Among these “leading experts” is a New Zealander who has written several New Age-style books on cats (including Pawmistry: How to Read Your Cats Paws) while being the “king of rubber-band magic”.