Published on:

Dodgy long term forecasts spoil the reputations of good short-term forecasters

Though I am writing this from Texas, from tomorrow I will be
back in the UK and I have been checking the weather forecast for my
home at the Met Office’s excellent website.

By excellent, I mean both clear and accurate. I find the Met
Office’s forecasts for a day ahead very good and the ones for up to
five days ahead pretty reliable and the ones for 30 days ahead
mostly not worth reading. That is not a criticism; it’s a fair
reflection of the unpredictability of chaotic systems: short-term
certainty rapidly decaying into long-term luck.

Which is why I wonder what the excellent Met Office weather
forecasters feel about the way their organisation’s leaders have
got them into dire reputational trouble, by obsessively trying to
pretend they can forecast climate as well as they can forecast
weather. Weather people, are you not getting just a bit disgusted
with your climate collagues? By predicting a “barbecue” summer that
was a washout in between two “mild” winters that were colder than
average, did they not committ the cardinal sin of weather
forecasting — long term overconfidence? And drag you down with
them?

This reputational trouble is deepening. December was the coldest in the UK since records began in
1910 with an average of -1.7C [-0.7C in England, and the CET series is longer and has a colder December
in 1890
.]. The Met Office is now claiming it privately told the
government in October to expect severe cold. Yet at around the same
time it put this image on its website, clearly predicting a mild
winter:

 

 

Was it really saying one thing to government and another to
readers of its website? How very convenient.

Autonomous Mind has this to say:

Ask yourself is this: Does it seem reasonable
or probable that the publicly funded meterology department of the
UK provided the government with a secret forecast about exceptional
cold, at the same time it was publishing the opposite forecast to
the public, but did so because it was previously ridiculed for
getting seasonal forecasts wrong? And that the government conspired
to keep it secret, took no action to prepare to keep the highways
clear and maintain a safe driving environment and let its Winter
Fuel Allowance budget be used up with only a fraction of the winter
gone?

Where is the logic in the Met Office thinking
it would avoid ridicule by telling the public on its own website
that there was a circa 80% probability of a warmer than average
winter if it was actually predicting the exceptional cold as it
claims to have told the government?

If I ran the Met Office, I would tell it to get back to its
knitting: concentrate on short-term weather forecasting and deny
any expertise on climate. The reason it won’t is money: the Met Office
gets huge dollops of grants from taxpayers to study climate. And it
is chaired by an ideologue on the subject, Robert Napier.

By Matt Ridley | Tagged:  Uncategorized