A former National Oceanic and Atmospheric Administration scientist has reopened a contentious debate over the validity of a key agency climate change study, asserting that procedures for archiving its data were not properly followed by its authors.
But many scientists, although hesitant to pronounce on the specific charges about data archiving, have pointed out that the research has been independently confirmed by another recent study — and that in any case, none of this raises any significant doubt about human-caused climate change. Meanwhile, the researchers behind the original NOAA paper have disagreed strongly with Bates’s charges, as has at least one scientist who worked with the team.
At issue is a 2015 bombshell report from scientists with NOAA’s National Centers for Environmental Information, led by Thomas Karl, then its director. The paper, published in the leading journal Science, addressed a favorite argument from the climate-doubting camp — the idea that temperature records indicate a slowdown or “pause” in global warming from 1998 on into the 21st century. The Karl paper seemed to lay these arguments to rest by using updated temperature data sets to demonstrate that no pause has occurred and that recent temperature patterns remain in line with the long-term warming trend.
The paper caused an immediate firestorm among conservatives upon its publication, prompting Rep. Smith to issue a congressional subpoena to the NOAA administrator aimed at investigating the adjustments the researchers had made to the historical data sets.
Bates, who did not immediately respond to requests for comment, is a longtime NOAA researcher who, before retiring in 2016, was a principal scientist at NCEI. His charges have instantly fed into a bitter debate that has been playing out over the timing of the release of the NOAA report and its conclusion, challenging a central assertion of those who are skeptical of global warming.
Charges about data ‘archiving’
Bates’s claims are outlined in a guest blog post, which he published on retired Georgia Tech climate scientist Judith Curry’s blog over the weekend. Many are highly technical in nature.
In essence, Bates asserts that the paper’s authors failed to comply with certain NOAA policies involving the managing and archiving of climate data and that the study’s results cannot be replicated or verified as a result. He also raises concerns about some of the software used in the study’s data processing, including issues involving coding errors and certain procedural issues surrounding the development and use of the programs.
Furthermore, Bates implies that the authors manipulated their data to place a greater emphasis on global warming and rushed the paper’s publication to coincide with the 2015 U.N. climate conference, which ultimately led to the adoption of the Paris climate agreement.
But in the few days since the blog post and news reports were released, multiple climate scientists — both involved and uninvolved with the research in question — have come forward to combat these claims. First, they observe, even if there were procedural issues involved with the management of the data sets used, the data itself — as well as the study’s results — appear to be sound. And in fact, they’ve already been independently verified by at least one other research group.
“Were all of NOAA’s internal procedures followed? Answer: We don’t know,” said David Titley, a meteorologist from Pennsylvania State University who was not involved with the 2015 paper, in an email to The Washington Post. “It’s really not that interesting, and there are established processes and procedures, internal and external to NOAA, to address those allegations.”
But, he said, “Is the science bad? Answer: No.”
Constructing new data sets for land and oceans
In the Science paper, two major data sets — which had both been updated or adjusted compared to previous versions — were used to construct a long-term temperature record that would enable the scientists to observe whether a global warming pause had ever occurred. One of these contained temperature measurements taken over land, and the other contained measurements taken at the surface of the sea.
And the result, as NOAA depicted it at the time in one image shared by the agency along with the study:
Skeptics have implied these data sets may have been manipulated to maximize warming. But several scientists have pointed out that both data sets have already been shown to be sound.
For one thing, the updated land data set used in the paper is similar to previous versions of the same data and “was responsible for relatively little of the increase in warming they showed,” wrote Zeke Hausfather, a climate scientist at the Berkeley Earth temperature analysis project, in a recent blog post for the website Carbon Brief. The sea surface temperature data was substantially updated for the 2015 paper, but those changes have since been independently verified in a recent study published by Hausfather and colleagues.
“In a paper published last month in the journal Science Advances, we compared the old NOAA record and the new NOAA record to independent instrumentally homogenous records created from buoys, satellite radiometers, and Argo floats,” Hausfather wrote. “Our results…show that the new NOAA record agrees quite well with all of these, while the old NOAA record shows much less warming.”
As we recently reported, the new study — which was independent from the 2015 paper published in Science — also concludes that there was no global warming pause. These results provide a “nice validation” of the 2015 study, said Karl, the paper’s lead author, in an interview with The Washington Post. Karl is now retired.
And what about the data archiving charges? There appears to be a dispute on this front between Karl, the former director of NOAA’s National Centers for Environmental Information, and Bates.
While not all the data used was permanently archived before the study’s publication, it was all readily available to any scientists who requested it at the time, Karl said. As for long-term archiving, that happened too, he said, after publication.
“The term ‘archive’ means a lot of different things to different people. … In this case, the data were available if anyone asked for it, and then they were archived further down the line after the paper was published,” said Karl.
A rush to publication?
Karl and other scientists involved with the research have also taken issue with the suggestion that the paper was rushed to publication to coincide with the Paris climate conference. They claim that the opposite was true — that, in fact, the procedural issues raised at the time the paper was being put together actually slowed the publication process.
“The real problem is that this work was delayed due to the processing concerns that [Bates] raised,” said Thomas Peterson, a retired NOAA research meteorologist and another of the study’s authors, in an email to The Washington Post. “I argued with those pushing process over science for literally years that we were putting out erroneous information because we weren’t allowed to update with new data and algorithms. … So it really bugs me to hear him say it was rushed when the exact opposite is true.”
In any case, the decisions made at the Paris climate conference hardly turned on one scientific paper. There’s an enormous volume of research published every year on the subject of climate change and a consensus in the research community that humans are warming the planet, regardless of what may have happened with the rate of warming over relatively short time periods.
Multiple other scientists have also come forward and published their own responses to these issues and various other points raised by both Bates and the Mail on Sunday — here, here and here are a few examples. At least one of these responses, published by scientist Peter Thorne, who has been involved with research related to the 2015 Science paper, also addresses some of the more technical software issues raised by Bates, suggesting that the codes used in the study were publicly available and internally approved at NOAA, with any bugs or errors having been documented.
Most important, though, each response defends the scientific integrity of the study, regardless of any disputes about data archiving or other NOAA procedural issues.
Curry — whose blog published Bates’s version of the claims — published another post Monday containing Bates’s emailed responses to some of the criticism his assertions have received. In his responses, he reiterates his concerns about some of the coding and other software issues he raised in his original blog post and suggests that some of the scientists who commented on these issues were not closely or recently enough involved with these NOAA operations to comment.
It’s unclear whether an investigation will be launched by NOAA. In a statement to The Washington Post, a NOAA spokesman said that “NOAA takes seriously any allegation that its internal processes have not been followed and will review the matter appropriately.”
— Chris Mooney and Jason Samenow contributed to this report.
More from Energy and Environment: