Yesterday - in reference to two non-peer-reviewed climate change studies released over the weekend - I said to ignore studies hyped prior to formal vetting.
Specifically, I was talking about studies released over the weekend by Cal-Berkeley physicist Richard Muller and blogger Anthony Watts.
New information has since emerged suggestive of potential problems with their work reinforcing the pitfalls in accepting their conclusions and further implicating Muller and Watts in scientifically questionable publicity stunts.
Let’s begin with Watts.
His central claim is that NOAA’s temperature data for the U.S. between 1979 to 2008 shows too much warming due to “spurious adjustments.” Using raw data from the highest quality weather stations, he and his colleagues calculate a warming rate of 0.155 degrees Celsius (C) per decade, 50 percent less than NOAA’s adjusted estimate of 0.309 degrees C per decade for all weather stations (of varying quality).
The blogosphere has quickly pointed out two problems with Watts’ estimates:
1) Independent satellite data - which Watts posts on his blog each month and has stood behind - indicate a warming over the U.S. closer to NOAA’s estimate. This point was raised by ClimateAudit blogger Steven McIntyre: “Over the continental US, the UAH satellite record shows a trend of 0.29 deg C/decade (TLT) from 1979-2008,” McIntyre said.
Interestingly, McIntyre is listed as a co-author of the Watts paper but begins a blog post expressing “puzzlement at Anthony’s [Watts’press release] announcement”and qualifies his involvement as “very last minute and limited”. And he admits to not having “parsed” parts of the Watts study.
2) Watts’ failure to make certain adjustments to the raw data, as NOAA has done, is a serious flaw knowledgeable bloggers say. Specifically, Watts did not apply a time of observation bias correction according to Howard University chemistry professor Josh Halpern, who blogs under the pseudonym Eli Rabett. McIntyre also addressed this problem: “There is a confounding interaction with TOBS [time of observation] that needs to be allowed for, as has been quickly and correctly pointed out.”
(For its part, NOAA defended the legitimacy of its peer-reviewed temperature record telling the New York Times’ Andrew Revkin “if anything, we are under-estimating the real world warming trends for the contiguous United States.” Regarding Watts’ work, it said: “we don’t want to comment on the draft pre-submission paper for obvious reasons.”)
Onto the “pre-submission” by Muller...
Muller claims that almost all of the warming observed since the late 1700s is due to human greenhouse gas emissions. His latest work consists of a “curve-fitting” analysis to determine and rule out causes for the long-term course of temperatures using a dataset he and his colleagues constructed
But one of the methods Muller used to construct the dataset submitted to the peer reviewed Journal of Geophysical Research has been rejected, according to Ross McKitrick, a professor of economics at the University of Guelph who served as a reviewer.
The study’s rejection does not mean it’s necessarily fatally flawed. Studies get rejected, reworked and resubmitted (to the originating journal or a different journal) all the time.
But the rejection does mean Muller and his team have work to do to get their house in order. And piling on new work onto a foundation with cracks, while effectively inviting the public in, demonstrates deeply compromised judgment. The cure here is the pillar of peer review; otherwise, the house comes crumbling down.
As I cautioned yesterday: stay away - or wear a hard hat. (The same analogy applies to Watts’ work).
UPDATE, Wednesday, 12:30 p.m.: Skeptical Science has published a throughough critique of the Watts study. Excerpt: “In its current form, the Watts paper contains little in the way of useful analysis. . . . The primary conclusion of the paper, aside from not being supported by the analysis, is simply implausible. . . . The flaws we have identified entirely compromise the conclusions of the paper...”