When you think of a published scientific paper that makes empirical predictions, you probably think that the authors actually tested the predictions empirically to verify their accuracy, right?

Well, apparently this is not seen as being necessary for scientific research in climate science, including by that prestigious journal Science.

No, in climate science it is apparently common practice to publish papers in which computer models to make predictions about events years and even decades in the future. Naturally, these predictions were not tested prior to publication. Most likely they will not be tested at all, for by the time the years about which predictions were made come around the study will have been forgotten, but not before public policies are crafted based on the unverified predictions of the study.

Roger Pielke, Sr., here discusses a classic paper of this sort. He also points out that the paper attempts to explain away a lack of significant global warming by arguing, and in this I don’t know if the paper is following or leading the pack, that the really bad anthropogenic global warming won’t start until 2009. It appears that disastrous global warming has been postponed…again. It is awfully convenient, yet typical, that this new starting point is several years after the publication of the paper, and gives plenty of time for ruinous public policies to be enacted on the basis of its unverified claims. This tactic also breathes several years extra life into the CAGW movement. How many times will the “due date” for CAGW have to be pushed back before the alarmists lose all credibility?

What happens if the study’s predictions don’t play out? Will the “due date” be postponed yet again? And even if a trend of significantly increasing global temperatures is detected starting in 2009, will it really prove the CAGW thesis correct? Conveniently, for the CAGW crowd, who may be hedging their bets, the upcoming warming period happens, as one of Pielke’s commenters mentions, to coincide with the next solar maximum lasting from 2009 to 2014. So…if significant warming does indeed suddenly occur, it could just be spurious correlation with the study’s predictions.

Pielke seems to have overlooked an important error in the study, however. The study’s authors claim that 1998 was the warmest year on record, but this is false. It was the second warmest year on record, and even this status is attributable to an unusually strong El Niño event. The warmest year on record was 1934, which is notably prior to the time the CAGW crowd claim anthropogenic global warming started. In fact, more than half of the ten hottest years on record occurred prior to this claimed anthropogenic period. (See here.)

{ Comments on this entry are closed }