Supposedly we’re facing the end of winter as we know it, the hottest year ever and so on. So can someone tell us why last week Minnesota experienced the largest early season snowfall ever recorded? To its credit, the news story specifies that “ever recorded” only takes us back 140 years. But still. Why did Elseworth Wisconsin have 9 inches of snow well before Hallowe’en? Why are low temperatures expected to persist and more snow to arrive in the area? Why did it snow in Ottawa on October 26? Of course one explanation is to retreat into vagueness, blaming “extreme weather” on “climate change”. But as we have noted and will continue doing until it becomes as tedious as shoveling snow in February, in real science a theory lives, or dies, by its predictive power. If you say winter is going to disappear it has to, or you are wrong.

We know, of course, that one cold swallow does not make an early winter. We object vehemently to cherry-picking. But Minnesota is not alone, or joined only by Montana. Italy just had its coldest September in 50 years, complete with, yes, snow in the hills around Rome. (For that matter, Tony Heller shows that early October in the United States has been getting cooler for decades.) And lest you fear we are being parochial, China recently had snow too.

There are a great many other predictions that aren’t doing well including for instance crop failure (in fact both staple grains and specialty items are doing very well) and mass extinction. Or the disappearance of snow in the “Australian alps” by 2020 or, if you prefer your predictions unverifiable, 2050 (whereas if you like evidence, they had massive snowfalls this August). In a way it’s amazing how many alarmist predictions manage to fail given how few are testable at all. But to wave one’s arms and say bad stuff coming, can’t say when or where or what, is not to do science. It’s even fairly lousy fortune-telling.

If you’re told to expect a tall dark handsome stranger, and true love walks past unnoticed in the form of a short blond friend, the crystal ball is worse than a mere decoration. And if you’re told it won’t snow and it does, or California will have more heatwaves and it doesn’t, so’s the model.

As we have also said before, if 10 years from now winter has ended, formerly fertile fields are cracked, dusty deserts, and all the good water levels went down and all the bad ones went up, we won’t still be arguing about climate change. But what if none of these things happen?

Can we then ditch the theory? For that matter, if this coming winter is colder and longer than usual not warmer and shorter in much of the world, can we at least question the theory?