Speaking of long-term trends and data, the cool kids or at this point the hot kids are all independently stampeding to attribution science, a form of hocus-pocus that bypasses evidence to get the verdict in question. For instance “Recent Canadian heat waves made much more likely by human-caused climate change” from Environment and Climate Change Canada. Which you might think means they have data proving they are happening more often. Pfft. This is attribution science, where you only need models, not data.

It’s one thing to cherry-pick and we don’t encourage it. Of course it is necessary to separate signal from noise, which means designating certain data and data series as more significant than others which takes a certain amount of judgement as well as calculation. But it must be done according to some sound statistical principle and “If I start there it seems to prove me right” isn’t really on that list. However it’s quite another to show up with an empty basket and say “Eat, eat”.

For instance Bloomberg Green serving up this Barmecidal tidbit:

“Climate change fueled scorching summer temperatures that killed an estimated 16,500 people in European cities this summer, pushing overall heat deaths to more than three times what might otherwise have occurred, according to a new study.”

What “more than three times” offers, “might otherwise” snatches away. And of course what you get instead is Friederike Otto, the flavour of the month, not data, the climate-science equivalent of boiled Brussel sprouts.

Meanwhile that Environment and Climate Change Canada study seems to suffer from some kind of catastrophe envy. We recently rubbished a study claiming to pin various global heatwaves on various specific energy firms including Canadian ones to as much as four decimal places. But in keeping with the classic Canadian self-obsession with finding a Canadian angle to any story including, famously, prompting one columnist to ask in 1999 “Exactly what is the Canadian perspective on a mudslide, anyway?”, well, this study chirps:

“Canadian climate scientists recently analyzed ten of the hottest heat waves that impacted the country in July and August of 2025, using the Rapid Extreme Weather Event Attribution system. The system continuously monitors for extreme heat and extreme cold events across Canada. The system shows that of these heat waves, nine were made much more likely because of climate change. This means that human influence on the climate made the following heat waves at least 2 to 10 times more likely to occur…”

And then it lists them. But before we get to that, surely the “at least 2 to 10 times more likely” should have you rending your lab coat. If it’s “at least” then the upper limit is more than 10; otherwise what they meant was “at most”.

And when they say “more likely”, you might wonder, more likely than what. They do eventually stagger around to saying, starting with:

“The system uses climate models to compare today’s climate to a pre-industrial one and demonstrate how CO2 and other emissions from human activities increase the chances of some extreme weather events while decreasing the chances of others.”

Ah, so a model-simulated fictional world. A preindustrial one, but which one? Surely there are several. Or indeed millions, since we had a “preindustrial” world from the time the Earth coalesced around 4.5 billion years ago until James Watt dramatically improved the steam engine in 1776. Are they comparing it to the Cretaceous? The Oligocene? The Eemian?

Heck no. These guys and gals don’t do long data series. Actually they don’t do data series at all. They do computer models of imaginary worlds. Still, they narrow down preindustrial considerably, basically to the tail end of the Little Ice Age:

“The Rapid Extreme Weather Event Attribution system uses climate models to compare two different climates: the climate of the 1800s, based on levels of atmospheric gases that existed before the Industrial Revolution; and the climate of today, based on observed levels of greenhouse gases and other results of human activity.”

Climate models versus data. Not really a fair fight. But again, instead of getting better it’s here that things get dramatically worse. Their first such heatwave was:

“Alberta (August 25 to August 31) Peak daily high temperature during the heat wave: 28.8 °C* Degrees above normal daily high temperature: 9.3 °C”.

OK. We’ll bite. How many of them were there in 2025 and how many in 1825? After all, they say directly that:

“Some events, like heat waves, are happening more frequently because of climate change.”

Show us. How often were there heatwaves in Alberta in the 19^th century that peaked at 28.8°C? Egad. You have no idea whatsoever?

OK then. How many such heatwaves were there in the 19^th century, assuming it’s what they meant by “the 1800s”? And how many during which, instead of the 28.8 mark, it was simply 9.3C above whatever normal daily high temperature you’ve concocted for, say, six days or whatever definition they’re using? Egad. You have no idea whatsoever?

No. None. No data. Just a computer simulation that made up that there were fewer. But how many fewer? After all, if this thing was two to ten times as likely and there was one in 2025, then presumably there had to be .5 of one in 1825, or on average in “the 1800s” and if it was ten times as likely, then .1 in 1825 or on average in the 19^th century. Was there?

They have no idea. Literally none. They don’t just have dodgy data. They have no data. Instead they told a computer to say something got more likely and when it duly said so, they hollered that we’re setting the planet on fire.

Oh, by the way, they also say:

“This past winter, scientists began analyzing the connection between human-caused climate change and extreme cold temperature events. They will begin analyzing extreme precipitation events later in 2025. As with extreme hot events, human influence on the climate increases the likelihood of extreme precipitation events (for example, heavy rainfall).”

So as usual with attribution science, they didn’t wait for the result to announce the result. Which makes sense, at least in their terms, because the methodology is guaranteed to finger humans for anything they feed in. Including, if they ever did it, heat waves in the Cretaceous.