Jim Dale: Why the Met Office got it wrong – again

The weather-watchers are too business-minded

And so we enter another mild, globally warmed week in the quiet backstreet of largely pedestrian weather here in the UK, where nothing ever hardly happens. Wait a sec, did I say mild and quiet? Sorry. I was just taking my cue from the Met Office's robust end-of-autumn prediction, suggesting we were all in for yet another mild temperate winter.

Curiously, that came right on the back of the Met Office's other empty and forlorn forecast of a glorious "barbecue summer" – which, if you recall, dished up a very mixed salad topped off by an appallingly soggy burger. Given that two out of two important outputs have gone spectacularly wrong, in a country where the weather makes a huge impact on commerce, shouldn't we be asking some very serious questions about how and why these flip-of-a-coin, long-term seasonal forecasts tend to end butter side down for the Met Office?

Its meteorologists will say in their defence that their outputs are based on probabilities, and that their smaller probabilities won the day over their much publicised larger probabilities. But the "don't blame us, we covered all the permutations" argument apart, that response does prompt a question about their own PR machine and its failure to explain that the old horse with asthma and a gammy leg does occasionally win the big race. (And they can't fall back on blaming easy-on-the-eye TV weather forecasters: most of them are perfectly capable of explaining percentage probabilities.)

We all know that a forecast is a forecast, but if the will to be heard first and loudest in order to gain commercial advantage is paramount, then that is when the snow can hit the fan. What we actually need from the Met Office's PR machine is a detailed explanation of forecasters' thought patterns before the media get a hold of the big juicy bone ("It's going to be a barbecue summer"/"It's going to be a mild winter") thrown obligingly to them.

You see, there was a time in the not so distant past when the Met Office would decry any private forecasting organisation that attempted to predict beyond a week, let alone predict a season ahead. Nowadays, with some 30-odd independent UK weather firms out there, market share means having to push the boat out into deep and perilous icy waters, where floods, blizzards and other tempests lurk, waiting for the unwary, the greedy and the foolhardy. For the Met Office, that has meant using its multimillion-pound number-crunching supercomputers to pick the winners. Which, given their record to date, poses serious questions about the merits of supercomputers that managed to choose the pig's ear of a forecast... twice running.

It's not rocket science, and it doesn't take supercomputers to predict that a colder than average continental type winter was far more likely than not this year. The main reason for this is that we had endured the mild and wet Atlantic for several months, and higher air pressure over our latitude, and a switch change was badly overdue, by way of balance.

That's not hindsight either: our own outputs for the season ahead did predict some bitter winter spells – more than we have become used to over the past 30 years or so. That's not to say that anyone out there in the private sector predicted the multibillion-pound polar slammer that we have right now; but maybe councils and commercially effected companies might just like to take a second opinion in the future?

So yep, something isn't quite right in the Met Office stable – be it the PR jockey or the super-duper-computers. Thankfully there are other horses in the meteorological stable – like mine, and we more than welcome the competition.

Jim Dale is senior risk meteorologist with British Weather Services