See Comments down arrow

The hard part of making predictions

04 May 2022 | OP ED Watch

There’s currently a heat wave in the Indian subcontinent. Which if you’ve ever been there might strike you as old news, or no news at all. But it’s climate change, of course. Al Jazeera kicked off with ‘Hottest summer ever’: Many Indian states under ‘severe’ heatwave/ Heatwaves have killed more than 6,500 people in India since 2010 and scientists say climate change is making them harsher and more frequent.” And you had to read quite a long way down to find the scientific foundation for that headline: “what local media reports said was its ‘hottest summer ever’.” And of course, says Climate Home News, “The heatwaves sweeping south Asia were foretold in climate models. Until we all stop burning coal, oil and gas they will get worse. Nor are temperate zones immune. Just ask the former residents of Lytton, Canada.” OK. Let’s ask them whether there’s been a trend toward more heat waves there and whether this summer it happens again. Because otherwise you’re just Zaqqum-picking.

In case you’re not up on your Qur’an, this devil’s-head-shaped berry awaits you in hell. “Surely the tree of Zaqqum,/ Is the food of the sinful/ Like dregs of oil; it shall boil in (their) bellies,/ Like the boiling of hot water.” Which seems much more appropriate than cherries when people are going about looking for the most dreadful things possible and menacing you with an eternity of them.

Fools do rush into such thickets. The BBC explains that India has always been so hot people resorted to things like rubbing raw mango on themselves to avoid heatstroke, and that odd local weather patterns are involved. (And in fact northern India it averages five to six heat waves a year whereas farther south, in Chennai, the year is divided into warm, hot and sweltering and “and it is hot and oppressive year round.”) But of course it’s all climate change… in which case it should continue. Next year should be worse than this year, 2024 worse still and so on. Trends are like that. Anyone care to make a prediction?

On the other side of the world, more or less, highlighting “São Paulo. Cape Town. Now Santiago and Los Angeles”, the New York Times “Climate Forward” just foretold that “Cities around the world are being forced to live with less water, as global warming melts glaciers, diminishes snow packs and exacerbates decades of water mismanagement. Not just less water for now. Maybe less water forever.” Which gets much of the past wrong already, since it is well-documented that most of the melting of the glaciers that advanced during the Little Ice Age took place before World War I. And that the American Southwest experienced megadroughts frequently over the last millennium. And that the last California drought celebrities told us would never end promptly did.

Still, as with economists who’ve predicted seven of the last four recessions, we’re willing to lend half an ear this time as well. But in return we insist that, if it turns out that cities that engage in sensible water management, admittedly a small sample, don’t have less water forever, we want someone to say sorry for being so obtusely dogmatic. Like the author in National Geographic’s “Planet Possible” who just said “The drought in the western U.S. could last until 2030/ After a brutally hot and dry 2021, the region is now in the worst ‘megadrought’ in 1,200 years. Climate change is to blame.”

Actually his editor. Many readers do not realize that writers not only do not pick their own headlines, they generally find that if they make a suggestion their editor would rather eat Zaqqum berries than take it. But the text says

“Without human-driven climate change forcing Earth’s temperatures up, the ongoing drought would still be painful and parched. But it would be unexceptional in the grand scheme of the past 1,200 years. A new study in Nature Climate Change shows that Earth’s warming climate has made the western drought about 40 percent more severe, making it the region’s driest stretch since A.D. 800. And there’s a very strong chance the drought will continue through 2030.”

There’s that spurious precision again. What does “about 40 percent more severe” actually mean? And if it weren’t, assuming you can know such things, when would it be the driest stretch since?

It turns out math am hard. The actual study, which the piece does at least link to, says “Anthropogenic trends in temperature, relative humidity, and precipitation estimated from 31 climate models account for 47% (model interquartiles of 35 to 105%) of the 2000-2018 drought severity, pushing an otherwise moderate drought onto a trajectory comparable to the worst SWNA megadroughts since 800 CE.” So 53 percent of the event in question was not explained by the models that explain it. Moreover all those intimidating interquartiles basically say the models disagreed massively, by a factor of three, but all said human influence was big and bad, perhaps because the programmers told them to. And the study said “the 2000–2018 SWNA drought was the second driest 19-year period since 800 CE, exceeded only by a late-1500s megadrought” so it turns out it wasn’t the worst since 800 at all. There was a worse one during (drum roll please) the Little Ice Age. Other than that, a fine piece of opinion reporting.

For instance, the National Geographic camp wildfire story says,

“Human-caused climate change, the team found, turned the drought from bad to terrible. Without the extra shove from climate change, the last 20 years would have been run-of-the-mill dry, somewhere in the middle of the pack of the 10 driest periods in the past 1,200 years.”

Which actually means they told a computer to find it and a computer did. It doesn’t mean it really happened that way.

If you’re wondering what would satisfy us, well, for starters, if the prediction that it lasts until 2030 pans out. But climate detonation predictions have a pretty awful track record, so we’re waiting for real evidence. Except for one thing. “Even absent climate change, there’s a very high chance the drought would last through 2023; in 94 percent of their simulations, it goes on through next year, and in 33 percent of their simulations it lasts all the way to 2030.” So in short, they made no prediction you can test.

Thanks a lot. As when The Economist asks “Did you know that the world is on track for a rise of 3°C in global temperatures above pre-industrial levels by 2100?” as a warmup to “Discover our climate-change hub offering rigorous, independent journalism” that is about as independent as it is rigorous, based on a prediction only Aubrey de Grey might think he can test. At least CBS had the decency to depict our demise in detail so lurid the Qur’an would be envious: “Excruciating heat will make summers increasingly dangerous. Agriculture and food supplies will suffer. People will be forced to migrate. Costs of living will skyrocket. All of these factors — and more — will contribute to political and social instability worldwide.” By, they seem to imply, the 2030s.

Leave a Reply

Your email address will not be published. Required fields are marked *