Weather 'super' computer fail

[Locutus]

Super-computer designed to predict the weather for the next 100 years gets it wrong inside 48 hours

MET OFFICE’S 100-YEAR FORECAST (BUT IT’S WRONG 2 DAYS LATER)



THE boss of the cash-rich Met Office has boasted that its £41million super computer will deliver accurate forecasts “a century ahead”.


But critics say the money would have been better spent getting the next day’s forecast right.
John Hirst said taxpayers would be well served by the massive computer because it would ensure Britain leads the world in climate science.

He said £20.7million was invested in the project last year alone, taxpayers’ money “which will enable the Met Office to deliver more accurate forecasts, from hours to a century ahead”. Mr Hirst said this referred to climate change predictions.

A Met Office forecast issued last Thursday for yesterday predicted hours of sunshine in central London. But there was hardly any, with rain clouds covering the capital for much of the day.

Mr Hirst’s claim was made in the Met Office’s annual report for 2011-12, which reveals he was paid £215,000, including a £45,000 bonus.


more jocularity


Express.co.uk - Home of the Daily and Sunday Express | UK News :: Met Office’s 100-year forecast (but it’s wrong 2 days later)

 

[lone wolf]

...and we wonder why the world's broke....

 

[gerryh]

If they can't predict the weather 100 years down the road, or even tomorrow accurately. How can they predict "climate change"?

 

[TenPenny]

If they can't predict the weather 100 years down the road, or even tomorrow accurately. How can they predict "climate change"?

Probably because they are two separate things.

Plus, the predictions for climate change are very general in nature. Akin to a weather forecast that says, 'in the next week, we might get some cloudy days, and it's possible we could have rain'.

 

[lone wolf]

Tonight's forecast: dark ... turning to partly light....

George Carlin 'The Hippy Dippy Weatherman' - YouTube

 

[gerryh]

Probably because they are two separate things.

Plus, the predictions for climate change are very general in nature. Akin to a weather forecast that says, 'in the next week, we might get some cloudy days, and it's possible we could have rain'.

I see, thank you.

 

[Locutus]

If they can't predict the weather 100 years down the road, or even tomorrow accurately. How can they predict "climate change"?

charts 'n' graphs

lottsa charts 'n' graphs

 

[IdRatherBeSkiing]

Probably because they are two separate things.

Plus, the predictions for climate change are very general in nature. Akin to a weather forecast that says, 'in the next week, we might get some cloudy days, and it's possible we could have rain'.

Very easy to make a prediction 100 years from now. Nobody will be alive then who heard you make the prediction to realize you were wrong or full of s-hit.

 

[Locutus]

Very easy to make a prediction 100 years from now. Nobody will be alive then who heard you make the prediction to realize you were wrong or full of s-hit.

Like 50 year warranty on caulking or some other crap.

 

[lone wolf]

...or the 25-year warranty shingles whose warranty becomes null and void the moment you use the de-icing salt they sell you....

 

[SLM]

Very easy to make a prediction 100 years from now. Nobody will be alive then who heard you make the prediction to realize you were wrong or full of s-hit.

Or you could be the next Nostradamus! Think about, you could have your very own History Channel series of specials in 2112.

 

[Tonington]

Was it wrong? No computer can predict the stochastic (random) processes, and for that reason weather forecasts are always probabilistic. If the forecast assigns a 70% chance of clear skies, and the actual weather for the day is rain, does that mean the forecast was wrong? In other words, was the 70% probability assigned to clear skies wrong? That's not an easy question to answer...