“Aggregating all the [mistakes]…one estimates Mr. Loblaw [SkepticalScience blog author] to have made about 24 direct analytical mistakes and several arguments from misdirection in this one disquisition. That may be some sort of record.” – Dr. Patrick Frank
On 27 June, 2023, Dr. Patrick Frank had his 46-page paper with 284 references published in the journal Sensors. It exhaustively detailed the magnitude of uncertainty and error in calculating global-scale temperature measurements.
Image Source: Frank, 2023
A blogosphere introduction to the paper appeared a few days later. It generated over 1,200 responses on the WUWT site, where debate is welcomed by those who agree or disagree, including direct discussion with the author himself in comments. (We highlighted Frank’s paper here a few weeks later, but without knowledge of the attention it had generated elsewhere.)
Contrast this open debate format with what occurs at climate blogs like RealClimate or SkepticalScience. Comments challenging the “consensus” position on anthropogenic global warming (AGW) are deleted and/or fail to emerge from moderation. Debate is effectively banned, as this leaves the impression that “consensus” AGW positions are irrefutably correct.
“I’ve deleted a number of his …comments, but I feel no obligation to explain to him why they disappear.” -SkepticalScience moderator
In light of the attention the Frank’s paper was propagating in the skeptical blogosphere, SkepticalScience contributing author, Bob Loblaw, decided it was time to take matters into his own hands earlier this month. His article, “A Frank Discussion About the Propagation of Measurement Uncertainty,” appeared on 7 August, 2023.
Image Source: SkepticalScience
In the vein of actual scientific debate (that is not allowed at SkepticalScience), Dr. Frank has provided a detailed 29-point response to Bob Loblaw’s critique attempt.
Frank accuses Loblaw of presenting a careless, mistake-filled manifesto that “violates the ethics of scientific review.”
In the first five pages of Frank’s response he identifies over 20 mistakes Loblaw made in his analysis.
For example, Frank catches Loblaw (1) confusing the uncertainty of the average (Equation 4) with the average of the uncertainty (root-mean-square, Equations 5&6), as well as (2) assuming uncertainty can simply be reduced by averaging (Loblaw: “Frank’s calculations make the astounding claim that averaging does not reduce uncertainty!”) even though uncertainty can only be reduced by averaging if the uncertainty results from random error. (The paper itself clearly describes why shipboard temperature measurements derived from pulling wooden buckets out of the water are fraught with errors that are not random.)
The last two pages of Frank’s response address the rather self-contradictory Loblaw claim that scientists do not prejudicially assume that measurement error is randomly distributed so they can profess to reduce uncertainty by averaging. (Loblaw: “Nobody actually makes this assumption.”) Clearly the scientific literature is replete with “Assumers of Random Measurement Error.” Frank originally provided seven pages of the random-error-assumption examples in scientific papers dating from the 1920s to present. The list has been substantially parsed down here for brevity’s sake.
Finally, Dr. Frank was interviewed on Tom Nelson’s podcast on 21 August, 2023. He addressed uncertainty and error in temperature measurements and climate model projections. A YouTube link to Dr. Frank’s “Nobody Understands Climate” 90-minute presentation appears below.