Yesterday’s edition of The Times had an opinion piece by Matt Ridley (p.23) entitled ‘Trust experts on anything but the future’, apparently inspired by Michael Gove’s recent comments in the build-up to the EU referendum about the British public having “had enough of experts”. Naturally such an article also got rather bogged down in criticising long-range weather forecasting.
Now, I don’t appreciate the way the article was written with regards to this, and I did let Mr Ridley know of this in a rather protracted Twitter discussion (in hindsight, an unsavoury encounter whilst cycling earlier in the day had made me rather irritable).
I don’t think comments such as “[it] covers all possibilities, like an astrologer” in reference to a roughly equally-weighted long range Met Office probability forecast show a true understanding of its usefulness (and probably suggest that the British public aren’t ready for probabilities and some consider it a ‘cop-out’ of giving a deterministic forecast). The article also lists every major Met Office long-range failure, and then goes on to say “a blindfolded person throwing darts at a chart would have done better”. Hardly.
I see his point – that long term future predictions are so difficult to make accurately, that non-experts could be just as likely to get it correct. I always remember that I expected the winter of 2009-10 to be very cold, based on only a few patterns I’d noticed, and I was correct, whilst the Met Office’s prediction of a warm winter is famously wrong (and indeed, mentioned in the article).
It is true that the Met Office have had seasonal forecast failures. It was bold of them to issue the forecasts to the general public, because back in 2007 and even now, seasonal forecasting (especially in the UK) is inherently very difficult. The backlash they suffered after the 2009 summer (the forecast ‘odds on for BBQ summer’ ended with the wettest July on record) goes on. And articles like this one are probably fuelled by that. It’s not the only time I’ve seen a list of all the UKMO’s failures in an article, either.
However, I feel that an appreciation of the sheer nature of atmospheric chaos should come first before launching attacks on the Met Office (some of which, in the past, have claimed they are waste of money). Ridley notes this late on in his article, but says “in the cliché, a butterfly’s wing flap can lead to a hurricane” – the issue being that whilst the phrasing is cliché, it is just about true, especially for Britain, where ensemble plumes do often diverge markedly after 5-7 days, and in certain circumstances always will because the atmosphere has varying states of predictability.
My point through all this is the British public, famed for being ‘weather-obsessed’, actually have a surprisingly poor understanding of weather and climate (it being, quite literally, over their heads). Instead, the public obsession (and consequently, the media obsession) is proving the forecaster wrong – even when they themselves are wrong. “Are you going to go and work for the Met Office and then they’ll get it right?” is something I’ve been asked countless times. And yet, the Met Office do get it right, day after day. Incredibly so.
The UKV, for example, is oustanding, when compared with the older models that were used over a decade ago. Ridley’s article didn’t praise the Met Office at all, but I did learn his praise for the short term forecasts was cut for size (predictably). There is an innate desire amongst Brits to prove the Met Office wrong – hence the rise of the Daily Express headline-generators James Madden and Piers Corbyn. No matter how accurate the Met Office forecasts become, their Facebook posts are still covered in comments from people criticising them for being useless.
I feel people expect perfect weather forecasts. They expect us to be able to perfectly forecast the behaviour of a chaotic system weeks in advance. In my experience the layperson has little appreciation of (amongst other things) the sheer number of calculations, sources of error, and just quite the level of atmospheric understanding it takes to produce NWP models.
It’s an astonishing reflection of the development of computing technology over the last 20 years that we can forecast as well as we can nowadays – just compare NHC hurricane track cone sizes today compared with 10 years ago – and yet people continue to neglect it. If one day we made a perfect forecast, would the response be “wow, that’s an incredible human achievement” or “well, about time! I’ve been using seaweed all my life and it’s never let me down!”. I fear the latter. Perhaps it has something to do with an older generation. Perhaps it’s British negativity. Perhaps it’s poor scientific understanding by the public. Perhaps it’s about rebelling against what we’re told. Whatever it is, I don’t like it.
Now, let’s all sip our coffees in quiet appreciation as we watch the 12Z GFS roll out.