My wife keeps telling me I should be more dog — I should just live in the moment, stop reflecting on the past and let the future take care of itself. But I can’t help it. My past is too full of mishap to be comfortably ignored, and I have no reason to believe that the future will be any different; therefore, I worry about it. And whilst I have no great expertise to bring to bear that enables me to predict the precise nature of my fate, there is certainly no shortage of experts around who delight in filling in the picture for me. One such expert is Professor Friederike Otto of the World Weather Attribution (WWA) group, whose party piece is to reflect upon past calamity and come out with statements such as, “This is definitely what we will see much more of in the future”. In fact, that’s precisely what she told the World this week after storm Boris had just finished ravaging Poland, the Czech Republic, Romania, Austria and Italy, leading to at least 24 deaths and billions of pounds of damage. But how can she know this? The BBC explains:

Scientists at WWA work out how much of a role climate change played in an extreme weather event by comparing it with a model of how bad that storm, drought or heatwave might have been in a world where humans hadn’t been burning fossil fuels for nearly 200 years.¹

And these models are very good, don’t you know, because they can fill you in with numbers:

But if warming reaches 2C, similar episodes will become an extra 5% more intense and 50% more frequent, the WWA warned.

The trouble is, however, that specific events such as storm Boris cannot be predicted by these models, and so the art is to wait until they happen, and then attach the reality of the experience to a set of scary future statistics, resulting in a mind-set that couldn’t possibly be any less dog. It is an approach that has worked very well for Professor Otto and her colleagues, and it has made a huge contribution to the zeitgeist that seems to have given the UK’s Foreign Secretary, David Lammy a free pass in declaring that “nothing could be more central to the UK’s national interest than delivering global progress on arresting rising temperatures.”²

But just how good have these experts been in predicting the increasing number of unpredictable events? Is anybody keeping track? Well, there has been one recent prediction relating to the numbers of the unpredictable that has thrown quite a light on the skill of these models, and the light proved to be quite unflattering. According to a logic that Otto would readily recognise, warm oceans provide the ideal conditions for tropical storms. For example, in 2023, the hottest Caribbean sea temperatures on record led to 20 named storms in the Atlantic, the fourth most active hurricane season on record. This year the seas have been just as hot and so, according to the models, there would be at least as many named storms – somewhere between 15 and 25 were confidently predicted. The season is now nearing its end, and the count is in:

Seven. That’s right, a measly seven named storms.³ So what went wrong? Where are all of the missing hurricanes?

Well, from the perspective of the experts, nothing went wrong. Their models were correct but the climate made a few mistakes that they couldn’t possibly have been expected to foresee. At least, that’s the explanation that has now been provided:

With the moist monsoon air reaching further north than usual, easterly waves are emerging into the ocean via Mauritania or Western Sahara rather than Senegal or the Gambia. There, the ocean is cooler, and, as the waves rotate, they bring in cooler, drier air from the north, so there is not as much energy for waves to turn into storms. Effectively, if these waves are the seeds of major storms then they have been planted in the wrong soil: unable to receive the warmth and moisture they need to develop into hurricanes.

Okay, I’ll buy that. But I still want to know why the climate models were unable to anticipate this ‘wrong soil’ development. This isn’t a case of a single storm failing to materialise, where the capriciousness of the weather gods can be blamed for a no-show. This is virtually a whole storm season being wiped out. Science earns its spurs in predicting this sort of thing – not coming up with neat, after-the-fact explanations.

The reality is that life is very complicated and experts rarely have a sufficient understanding to enable accurate predictions. But that doesn’t seem to bother them all that much. If one restricts oneself to statistical forecasting, rather than the prediction of specific events, then the chances of a failed prediction coming to the public’s attention are greatly reduced. As far as storm prediction is concerned, it’s just a sequence of specific predictions that were never actually made, and so were never wrong. And by the time someone has twigged that something did go wrong, the experts will have already come up with their expert explanation as to why the statistics didn’t play the game. That’s why we call them experts. Meanwhile, the models will have been tweaked:

State-of-the-art climate models suggest that in a warming climate, the west African monsoon may become wetter and shift further northwards, potentially resulting in similarly quiet hurricane seasons in future.

And then, when that doesn’t happen, we’ll have even newer models. But what we will never learn, it seems, is that the experts are still hedging their bets:

Researchers at Colorado State University, widely considered some of the most accurate hurricane forecasters, have estimated a 50% chance of a return to normal hurricane activity in the next two weeks.

Have they now?

Footnotes:

[1] Actually, Mr BBC man, it’s not one model. It’s a group of models cherry-picked by Otto, supposedly selected according to their suitability for the given attribution exercise. This collection is then treated statistically as though it were a random sample taken from the space of all possible models.

[2] How the UK can deliver global progress given the decidedly non-global scale of its emissions is not entirely clear. It appears to depend upon some form of dark energy that enables Lammy to force China into submitting to his personal charm.

[3] Admittedly, the season started off strongly. Hurricane Beryl was the earliest category 5 storm on record.