On Thu, 27 Jul 2017 10:00:49 +0100, whisky-dave wrote:

On Wednesday, 26 July 2017 20:17:05 UTC+1, James Wilkinson Sword wrote:
On Wed, 26 Jul 2017 13:22:02 +0100, whisky-dave wrote:

On Tuesday, 25 July 2017 19:18:26 UTC+1, James Wilkinson Sword wrote:

How about showers or inteminant showers or heavy downfalls, storms.

Yes, those are more helpful. But "20% chance of rain" means nothing,


Well my BBC weather app tells me that in east London in the next hour theres will be heavy rain 18C 12 mph wind.
Symbols for UV pollen and polution .
sunrise 5:15 sunset 20:58
Far more useful than your pathetic % idea.

I've looked at hourly forecasts before, they're terribly inaccurate.

Because of the small time scale yuo've not meant to look at them a week in advance, they are usualy OK for that day buit for teh next day 2 or 3 hour is the best yuo can hopem for and even that depends on previous conditions.

I was looking at the current day. If they can't work that out they need to stop bothering.

This is were proability comes in.
The storm off the south coast in 1987 had a probability of about 20%.
Not sure what the probabiolity of it being teh worst storm in 300 years was, but that doesn;t mean it could never had happened.

If a forecaster uses probabilities, he's guessing. Wait until we can actually predict it properly, then make forecasts.

--
Why don't Siamese cats come in pairs?