Ex-climatologist Jim Salinger has penned an article for Australia’s The Conversation regarding his 30-year-old version of New Zealand average temperature trends and recent efforts to have the High Court order its removal from NIWA’s official website.
Salinger’s version was initially published in 1980, when he was a student at Victoria University. It relied upon seven geographically representative stations with long-term data (‘Seven-station Series’, or ‘7SS’). The article acknowledged the temperature data had been ‘homogenised’ but offered no details. The graph showed a warming trend of 1.1 °C from 1853 to 1975.
In 1999, 20 years later, Salinger was a principal scientist at NIWA. He then put his old 7SS graph up on NIWA’s official website under a “New Zealand Temperature Record” (NZTR) headline. Ever since, it has been used by the NZ Government as proof that NZ temperatures have increased at a rate of about 1 °C/century since the 1850s.
History of constant national temperature
In about 2006, the NZ Climate Science Coalition became aware that the Dominion Meteorologist had published an official NZTR in 1868 and again in 1920. Both these series clearly showed that the country’s average temperature has remained exceptionally stable ever since records began.
It then became apparent that Salinger’s 1980 “homogenization” of the data was a very big deal. His process hadn’t just tidied things up, but was responsible for the entire alleged warming trend. The Coalition tried in vain to obtain enough details (dates, locations, techniques, values) to allow replication of the adjustments he had made.
In 1999, the Coalition published a paper “Are we feeling warmer yet?”. This graphed the historical temperature data (daily maxima and minima) collected and recorded by hundreds of public servants over 165 years. The temperature trend for this official archive was dead level. The paper contrasted this with the Salinger graph showing a steep warming trend – the unsupported opinion of a single public servant.
By this time, Salinger had been summarily sacked by NIWA and was suing them in the Employment Court. Despite this, NIWA climate manager David Wratt brought him back in as a consultant to defend the use of his NZTR. He attempted to do this by hastily devising the infamous 11SS – but that is a story for another day.
NIWA resists data release
During 2010, the Coalition went all out to obtain the information necessary for replication. As a result of Official Information Act requests, Parliamentary Questions and correspondence with the NIWA Board Chairman, more information was slowly and painfully pieced together. This included:
- A Schedule of Adjustments devised by NIWA (it took them two months of detective work) showing the dates, locations and values of the Salinger adjustments. This Schedule was tabled in Parliament together with a paper illustrating the adjustment techniques (the “Hokitika Example”).
- The source of both these key documents was Salinger’s 1981 doctoral thesis. Unfortunately, the Thesis showed outcomes only – the actual calculations of all the adjustments had been lost in a computer mishap at Victoria University in the mid-1980s.
- The Thesis adjustments had never been checked, reviewed or recalculated by NIWA (or by anybody else). NIWA believed Salinger had applied measurement techniques which were later described in a peer-reviewed paper known as Rhoades & Salinger (1993) (“RS93”).
NIWA’s Minister, recognizing that this mess was indefensible, promised three remedial measures for the 2010-11 Science Plan.
- NIWA would recalculate (“the Review”) the 7SS adjustments using internationally recognized techniques. This would involve several months’ work by 5 or 6 scientists.
- The recalculations would be externally peer-reviewed by two scientists from NIWA’s sister organization in Australia, the Bureau of Meteorology (“BoM”).
- NIWA would describe their adjustment calculations in a scientific paper to be independently peer-reviewed and published in an international science journal.
NIWA thwarts promises
The Minister expected that these three documents would collectively ensure public confidence in the adjustments emerging from the Review. NIWA reneged on all three undertakings.
1. After NIWA’s Review (December 2010) quoted RS93 at least 20 times, the Coalition published an audit (June 2011) showing the RS93 techniques were not correctly applied. Just before the Court hearing in 2012, NIWA finally admitted this was the case. They had used the outdated calculation techniques described in Salinger’s 1981 thesis.
2. When the BoM peer-review report failed to support NIWA’s recalculations, NIWA claimed it was confidential. The Coalition immediately appealed to the Ombudsman under the Official Information Act – but NIWA has succeeded in stalling his decision for over three years.
3. When no scientific paper appeared in 2011, NIWA’s Minister told Parliament it would be submitted by June 2012. That didn’t happen.
Totally frustrated, the Coalition turned to the Courts to hold NIWA accountable. But the Judge was not prepared to adjudicate what he saw as a scientific dispute.
NZTR remains unexplained
Although wholly discredited in the Court of Public Opinion, the 7SS-based NZTR remains stubbornly in place. Its many unexplained features include:
Random flaws require random adjustments, which should tend to cancel out over time and space. The Salinger adjustments are unique in that all 30 go in the same direction — to create a steady warming trend of 1 °C/century until 1975.
Almost all of the Salinger-created warming occurred during 1930-55. During this period, which coincided with slight cooling across the globe, no significant increase of global greenhouse gas concentrations can be detected.
A peer-reviewed journal paper, Hessell (1980), found that no significant warming occurred in New Zealand during 1930-80. Authored by the country’s most senior meteorologist, this paper carries much greater scientific weight than Salinger’s contemporary unpublished student thesis.
The Hessell (1980) paper demonstrated that the Auckland and Wellington temperature data were seriously contaminated by sheltering trees, while Auckland and other 7SS stations were also impacted by urban heat island (UHI) effects. Yet Salinger made no attempt to test or correct for any of these known pitfalls.
All of the worldwide surface-based temperature series (HadCRUT, GHCN or GISS) include grid-boxes for New Zealand and surrounding seas. They accept the historical data and are unconvinced by the Salinger adjustments.
The Salinger-adjusted New Zealand warming trend is almost twice as steep as the IPCC’s official global warming trend of 6.1 °C. Yet NIWA has constantly predicted that future New Zealand warming will be one-third less than the global average.
The official mean New Zealand-wide temperature in the 1860s was 13.1 °C. Today it is a little lower than that – which is clearly incompatible with Salinger’s story. Perhaps this is why NIWA insists that the 7SS is an “unofficial” series?
You say it’s recognised: so tell us by whom
During the course of the recent Court proceedings, NIWA was endlessly invited to cite some authority for its hollow claim that Salinger’s 1981 measurement technique (used in the Hokitika Example and the Review) is “internationally recognised”. Four important aspects of that technique were rejected by RS93 for the very good reasons spelled out in that seminal paper.
It would be progress if some of these mysteries could be resolved. But the 7SS will never shake off its dodgy reputation until Salinger or NIWA can point to at least one independent statistician who is prepared to endorse its obsolete measurement technique.
Or until NIWA finally provides a reference to the technique in the scientific literature.