Investigation of NOAA climate scientists finds bupkis

Image of a magnifying glass.

During his run as chair of the House Science Committee, recently retired Texas Congressman Lamar Smith made a habit of accusing US National Oceanic and Atmospheric Administration scientists of manipulating data to exaggerate global warming. Smith did so almost entirely on the basis of his own wishes that data would stop showing warming, relying on fever swamp bloggers and disregarding published research.

His favorite target for both accusations and subpoenas was a 2015 paper published in the journal Science that incorporated new research to update NOAA’s global temperature dataset. The updates to the dataset made warming in the early 21st century more pronounced, and they indicated that a supposed “pause” in warming was a statistically insignificant event.

After the paper was published, the UK’s Mail on Sunday and its digital sibling The Daily Mail published allegations of improper behavior at NOAA, courtesy of a “whistleblower” inside the agency. The source, John Bates, claimed that the study had been rushed through without following proper data archiving protocol and that questionable choices had exaggerated the warming trend.

Most of the claims failed to make any sense right from the start, especially given the fact that the NOAA dataset showed no more warming than those run by NASA, the UK Met Office, or anyone else. To give one example, Bates claimed that the NOAA team led by Tom Karl made dodgy adjustments to ocean temperature measurements, but they were actually using a previously published ocean dataset completed two years earlier. And as Ars learned at the time, Bates had some personal issues with Karl—who had demoted him in 2012.

The Daily Mail/Mail on Sunday were eventually forced to publish corrections and the science moved on, but the Department of Commerce (which runs NOAA) commissioned a firm to do a third-party evaluation of Bates’ claims. Although the resulting report is dated July 2018, the Department of Commerce released it to the public just before the holidays—and just as the government shutdown took NOAA’s websites offline. As such, climate scientists only discovered it a few days ago when a retired NOAA researcher shared it on Twitter.

The report details interviews and a review of emails that settle the questions raised by John Bates. The report’s topline conclusion is hardly a surprise: “After carefully reviewing internal NOAA email correspondence, the MITRE Committee found no evidence that the Karl Study falsified, or intentionally distorted climate data. The Karl Study data were subsequently used in multiple peer-reviewed scientific publications.”

(Incidentally, in his written answers to the firm’s questions about the team’s scientific decisions, Tom Karl quotes Ars’ original coverage of the controversy.)

However, there is one surprise in the report, and it relates to Bates’ accusation that the team failed to follow internal protocols. It turns out that Bates himself would be partly responsible for that: “The MITRE Committee learned that the internal review, later criticized by Bates, was conducted and approved under his own authority. The MITRE Committee found no evidence that Bates ever mentioned this fact in his blog, email, or anywhere else in his discussion of the matter in public.”

The only possible issue the outside review identified is that the researchers sent their paper to Science on December 23, while Bates’ official written approval was emailed the following day (Christmas Eve). It’s not stated whether there was offline discussion about this or if the holiday played a role, but NOAA’s rules state that the approval should have come in first.

The report did find that NOAA’s rules for handling and archiving data are ambiguous. The agency has different rules for “research” data and the “operational” data that are going straight into live products provided to the public, but identifying which is which can sometimes be fuzzy. While the data in the study was available, there were apparently debates about process because the updates described in the study weren’t going live in the “operational” data quite yet. Basically, the report says NOAA needs to sit down and figure out what its standard procedure is going to be in situations like this.

The Department of Commerce press release on the report states that NOAA has until February 1 to provide a plan along those lines, though the ongoing government shutdown may alter that timeline.

Unless you find clarification of office policies to be unusually engrossing, there isn’t much excitement to be found in the report, which will receive much less attention than the splashy tabloid allegations it debunks. The 2015 Science study has held up, and researchers at NOAA and elsewhere have continued their slow work to make the tiny tweaks to global datasets that make them as accurate as possible. Their work rarely makes headlines, and after this episode they may be thankful for that fact.

Leave a Reply

Your email address will not be published. Required fields are marked *