Thursday, July 30, 2009

IS THERE A DIFFERENCE BETWEEN CLIMATE AND WEATHER?

This is the tactic that the 'shit happens'-mongers are using to stop criticism of the Meteorological Office's very public failures of their long-term weather forecasts.

They say, with great pride, that weather is different from climate, so because the weather was not predicted correctly does not mean that the climate predictions are wrong too.

OK, let's look at this again.

I've saved the link to The Meteorological Office website in my favourites/bookmarks, and when I open the favourites/bookmarks it says Met Office : Weather and Climate Change.

So the Meteorological Office is making a clear link between the two.

But which is easier to predict? The weather over a small portion of the planet over a period of several months, or the climate over the whole planet over a period of several decades?

In the former the timescale is smaller, the physics is less complex and the computational domain considered is smaller.

In the latter, the timescale is much larger, there is much more complex physics to consider, e.g. solar effects, and the computational domain is much larger.

I am researching much smaller scale fluid flow and am aware of the physics, the computational methods and computer systems used in these calculations, and I'm telling you that weather and climate prediction are not an exact science. Even in my research, in which I deal with fluid flow in a volume of say 20 metres cubed, I am having to use approximate methods on a very fast computer with limited precision, and previous research has never been able to agree with experimental results. Even in such a small computational domain we don't have the methods to calculate exactly the behaviour of the fluids involved. As a simulation runs the initial errors build up and up and up.... There are only a very, very small number of situations in which exact solutions to the governing equations can be found, and weather and climate do not belong to that tiny subset.

It does not fill me with confidence in the long range predictions of climate when our weather over the UK alone cannot be predicted accurately over a period of a week, never mind months. I've been doing this for a while now and the initial predictions are never correct. On The Meteorological Office website I look at the weather prediction for my area six days in advance. It's sometimes fun to watch the prediction change over those six days from glorious sunshine to shower to dark rain to sunshine again. It's often correct only the day before when a large region of high pressure is sitting directly over the UK and the nearest clouds are over the North Atlantic and it's obvious that there ain't going to be any rain.

There is always the probability that the military has the methods and computational resources to predict the weather and climate to a high degree of accuracy, but why would they tell us? The military is probably 40 to 50 years ahead of what's in the public domain.

No comments: