British supercomputer botches weather forecasts

The British Meteorological Office has a new weather-forecasting super-computer built at a cost of 41 million pounds sterling (approximately $65 million) . The computer is touted to be more powerful than 100,000 standard PCs, is capable of 1,000 billion calculations every second, and uses 1.2 megawatts of energy to run – enough to power a small town. The head of the Met office claims that this new computer “will enable the Met Office to deliver more accurate forecasts, from hours to a century ahead.”

Let’s see how it is doing so far. On 23 March 2012, the computer produced a forecast for the next three months: “The forecast for average UK rainfall slightly favors drier-than-average conditions for April-May-June as a whole, and also slightly favors April being the driest of the 3 months. With this forecast, the water resources situation in southern, eastern and central England is likely to deteriorate further during the April-May-June period.”

And here is what happened:

April: 2012 had wettest April for 100 years, Met Office says “It has been the wettest April in the UK for over 100 years, with some areas seeing three times their usual average, figures from the Met Office show. Some 121.8mm of rain has fallen, beating the previous record of 120.3mm which was set in 2000.”

June: June on course to be wettest in a century: Flooding, storms and persistent showers have blighted the country in recent weeks putting this June in line to become one of the soggiest in 100 years.

25 June: Spring is wettest in Britain for 250 years – England and Wales are on course for the wettest late spring and early summer for 250 years, experts said yesterday. June has just seen its fourth washout weekend and yet more downpours are forecast. Now it is feared combined rainfall for April, May and June will break the record of 13.2in (336mm) set in 1782 and be the highest since records began in 1766.

I was in Wales and Cornwall in June and can personally attest that June was definitely not dry.

This shows that super-computers and weather/climate models fail when the wrong assumptions are input: Garbage in-Garbage out. The new super-computer will, however, allow the Met Office to make their mistakes faster.

(H/t to Global Warming Policy Foundation)



  1. I live near Oxford in the UK and have listened closely to the met office’s predictions based upon “observation, numerical computer models and experts”, for a number of years.

    The piece you quoted includes further rationale of the percentages of probability. As I understand it, there were five categories that could be correct for the, then, forthcoming quarter: very dry, quite dry, average, quite wet and very wet – and, based upon no other information therefore, a 20% chance of each of the outcomes happening.

    Having employed all their computer power and expertise they issued a prediction that was slightly off centre on probability towards drier rather than wetter. We had similar predictions in the recent past with respect to the prediction of a BBQ summer and a mild winter – followed by a wash out and the coldest winter the uk has seen for years respectively.

    The main conclusion from this evidence is that, in reality, they have no predictive ability whatsoever. It suggests that even when they are right it is probably luck.

    The more sinister aspect of this is that the same supercomputer and experts are the same ones telling us of the coming of rapid increases in global temperatures.

    Anyone who has the temerity to question their ability to forecast global warming is hit with the “consensus” argument and are increasingly labelled as “deniers”.

    Whether one is a denier or a warmist, at the very least one would have to agree that the met office consistently demonstrates an inability to forecast or predict even the next month or quarter’s weather. Based upon that observation, what weight would one put on their ability to predict the next century.

    The other interesting part is the always get it wrong the same way. They seem to be always predicting warmer, milder, hotter, less rain, less snow, etc. this suggests that there is, somehow, a bias in their models that is constantly skewing the accuracy – be that an error or on purpose.

    1. Hello Philip,

      Thanks for the comment.  My wife and I passed through Oxford during our trip through Wales and Cornwall in mid-June.  We actually enjoyed the cool, rainy weather.  It was a great and welcome contrast to the dry, 35-40 degrees C we experience in Arizona during June.

Comments are closed.