Models for chaotic systems are not very accurate I'm afraid, after all the Met office use a 80million pound super-computer to run their models and in March it predicted that April, May and June would be warmer and much drier than normal.

Looking at the Jet-stream predictions for the coming week Jetstream Forecast - Jetstream Map Updated Four Times Daily - Netweather.tv it would appear to hold out little hope for a change from wet and unsettled weather.

What we need is for everyone in the South of France and in Spain to face north and blow like mad to send the jet-stream back up to the North of Scotland where it should be at this time of year :-))

"Granity" wrote in message
...

kay;963812 Wrote:
Not quite true. You collect info on past events, use it to find a model
that fits, then use the model to assess the probability of future
events. So if your model is based on a data set which includes several
"once in 10 year" events, you may well be able to predict the size of a
"once in a hundred years" event.

Models for chaotic systems are not very accurate I'm afraid, after all
the Met office use a 80million pound super-computer to run their models
and in March it predicted that April, May and June would be warmer and
much drier than normal.

Looking at the Jet-stream predictions for the coming week 'Jetstream
Forecast - Jetstream Map Updated Four Times Daily - Netweather.tv'
(http://tinyurl.com/cuz46cs) it would appear to hold out little hope
for a change from wet and unsettled weather.

What we need is for everyone in the South of France and in Spain to face
north and blow like mad to send the jet-stream back up to the North of
Scotland where it should be at this time of year :-))

g That would be a sight to behold!

--
--

http://www.shop.helpforheroes.org.uk/

"Martin" wrote in message
...
On Sun, 8 Jul 2012 16:26:01 +0100, "Ophelia"
wrote:



"Granity" wrote in message
...

kay;963812 Wrote:
Not quite true. You collect info on past events, use it to find a model
that fits, then use the model to assess the probability of future
events. So if your model is based on a data set which includes several
"once in 10 year" events, you may well be able to predict the size of a
"once in a hundred years" event.

Models for chaotic systems are not very accurate I'm afraid, after all
the Met office use a 80million pound super-computer to run their models
and in March it predicted that April, May and June would be warmer and
much drier than normal.

Looking at the Jet-stream predictions for the coming week 'Jetstream
Forecast - Jetstream Map Updated Four Times Daily - Netweather.tv'
(http://tinyurl.com/cuz46cs) it would appear to hold out little hope
for a change from wet and unsettled weather.

What we need is for everyone in the South of France and in Spain to face
north and blow like mad to send the jet-stream back up to the North of
Scotland where it should be at this time of year :-))

g That would be a sight to behold!

It seems to be working. It's gone from being humid and hot to cold and
windy. There's a strong smell of garlic in the air too, but that could
be this years crop drying.

Bring plenty with you, I just read the the EU are going to charge us a
garlic tax!

--
--

http://www.shop.helpforheroes.org.uk/

"not very accurate" is not the same as "not possible", which was the bit I was challenging.

And working out the possible size of an extreme rainfall event isn't quite as complex as working out the amount of rainfall on a particular day. You don't need to model what's actually happening to cause the weather, you simply need a distribution that fits the observed frequencies of amounts of rainfall over as long a period as you have data for.

Unfortunately we only have good data for just over 100 years which is unlikely to have the matchable data sequence required, so far all we have are the records of the last(?) part of a warming period coming out of a little ice-age, I suspect that after another say 1000 years we might find that there is indeed some previous matching data that we could match up.

In the mean time it should be quite possible to calculate what size drains would be needed to cope with a downpour of the highest amount ever likely to fall over say a one hour period, and this be insisted upon as part of the planning consent for every new housing estate. It would certainly help to stop some of these flash floods that happen.

But that's the same problem that I was talking about and what Harry seemed to be saying wasn't possible. What problem were you talking about?