Whats the point of forecasts that dont turn out to be correct ?
 

The probability of correct forecasting is something that has been
rattling around in my head for a while.

If on Day 1 of a 5 day forecast I'm told Rain is on the way only for
this forecast to change between rain/dry (for example) over the next
three days and then its actually dry on day 5, the forecast on day 1
was utterly worthless.

What probability of a likely event happening triggers it becoming a
'forecast'. Surely it needs to be 75% likely as that would mean it is
twice as likely to happen, as not.

Also why is that probabilty of a forecast never mentioned on tv/radio/
media forecasts. Is it because Jo Public has dumbed down so much over
the last few years that he has no concept what Porbability is?

Whats the point of forecasts that dont turn out to be correct ?
 

Because a lot of the time they are correct.

Whats the point of forecasts that dont turn out to be correct ?
 

On Jun 5, 6:25*pm, ned flanders wrote:
The probability of correct forecasting is something that has been
rattling around in my head for a while.

If on Day 1 of a 5 day forecast I'm told Rain is on the way only for
this forecast to change between rain/dry (for example) *over the next
three days and then its actually dry on day 5, the forecast on day 1
was utterly worthless.

What probability of a likely event happening triggers it becoming a
'forecast'. Surely it needs to be 75% likely as that would mean it is
twice as likely to happen, as not.

Also why is that probabilty of a forecast never mentioned on tv/radio/
media forecasts. Is it because Jo Public has dumbed down so much over
the last few years that he has no concept what Porbability is?

To be fair, as Victoria Coren wrote in the Observer last Sunday:
The Met Office recently had its budget cut by £4.3 million. That may
not sound like much, but it was a loss of more than 20 million pine
cones.

Len Wood
Wembury

Whats the point of forecasts that dont turn out to be correct ?
 

On 5 June, 19:25, Len Wood wrote:
On Jun 5, 6:25*pm, ned flanders wrote:

The probability of correct forecasting is something that has been
rattling around in my head for a while.

If on Day 1 of a 5 day forecast I'm told Rain is on the way only for
this forecast to change between rain/dry (for example) *over the next
three days and then its actually dry on day 5, the forecast on day 1
was utterly worthless.

What probability of a likely event happening triggers it becoming a
'forecast'. Surely it needs to be 75% likely as that would mean it is
twice as likely to happen, as not.

Also why is that probabilty of a forecast never mentioned on tv/radio/
media forecasts. Is it because Jo Public has dumbed down so much over
the last few years that he has no concept what Porbability is?

To be fair, as Victoria Coren wrote in the Observer last Sunday:
The Met Office recently had its budget cut by £4.3 million. That may
not sound like much, but it was a loss of more than 20 million pine
cones.

Len Wood
Wembury

Pine cones? Down here we're more up to date!
http://www.redbubble.com/people/bcim...074-7-you-what

Graham
Penzance

http://www.turnstone-cottage.co.uk/

Whats the point of forecasts that dont turn out to be correct ?
 

On Jun 5, 5:25*pm, ned flanders wrote:
The probability of correct forecasting is something that has been
rattling around in my head for a while.

If on Day 1 of a 5 day forecast I'm told Rain is on the way only for
this forecast to change between rain/dry (for example) *over the next
three days and then its actually dry on day 5, the forecast on day 1
was utterly worthless.

What probability of a likely event happening triggers it becoming a
'forecast'. Surely it needs to be 75% likely as that would mean it is
twice as likely to happen, as not.

Also why is that probabilty of a forecast never mentioned on tv/radio/
media forecasts. Is it because Jo Public has dumbed down so much over
the last few years that he has no concept what Porbability is?

Yes, why don't they tell us which of their forecasts are going to be
wrong?

Cheers, Alastair.

Whats the point of forecasts that dont turn out to be correct ?
 

In message
, ned
flanders writes

The probability of correct forecasting is something that has been
rattling around in my head for a while.

If on Day 1 of a 5 day forecast I'm told Rain is on the way only for
this forecast to change between rain/dry (for example) over the next
three days and then its actually dry on day 5, the forecast on day 1
was utterly worthless.

What probability of a likely event happening triggers it becoming a
'forecast'. Surely it needs to be 75% likely as that would mean it is
twice as likely to happen, as not.

Also why is that probabilty of a forecast never mentioned on tv/radio/
media forecasts. Is it because Jo Public has dumbed down so much over
the last few years that he has no concept what Porbability is?


Yeah right! Ever thought that perhaps the Met Office (MO) is dealing
with nature and nature doesn't lend itself readily to prediction?

Overall the MO does a pretty good job, read the excellent contributions
from Will Hand, he works for the MO and really knows his stuff, do you?
I doubt it.

Here on the Isle of Man our Met Office is excellent. Civil services are
an easy target for criticism from simplistic, right wing dickheads, my
best suggestion for you is to get a life and stop believing the Daily
Mail.


--


Jim

Whats the point of forecasts that dont turn out to be correct ?
 

On Sat, 5 Jun 2010 at 10:25:50, ned flanders wrote in
uk.sci.weather :


The probability of correct forecasting is something that has been
rattling around in my head for a while.

If on Day 1 of a 5 day forecast I'm told Rain is on the way only for
this forecast to change between rain/dry (for example) over the next
three days and then its actually dry on day 5, the forecast on day 1
was utterly worthless.

What if showers are forecast - some people are going to see the forecast
as correct, while others won't...
--
Paul Hyett, Cheltenham (change 'invalid83261' to 'blueyonder' to email me)

Whats the point of forecasts that dont turn out to be correct ?
 

Alastair wrote:
On Jun 5, 5:25 pm, ned flanders wrote:
The probability of correct forecasting is something that has been
rattling around in my head for a while.

If on Day 1 of a 5 day forecast I'm told Rain is on the way only for
this forecast to change between rain/dry (for example) over the next
three days and then its actually dry on day 5, the forecast on day 1
was utterly worthless.

What probability of a likely event happening triggers it becoming a
'forecast'. Surely it needs to be 75% likely as that would mean it is
twice as likely to happen, as not.

Also why is that probabilty of a forecast never mentioned on
tv/radio/ media forecasts. Is it because Jo Public has dumbed down
so much over the last few years that he has no concept what
Porbability is?

Yes, why don't they tell us which of their forecasts are going to be
wrong?

They sometimes do.
But only *after* they go wrong.
What's the point of that, bloody Met Office with their 'oh so clever
with hindsight' attitude.....
--
Col

Bolton, Lancashire
160m asl

Whats the point of forecasts that dont turn out to be correct ?
 

On Sun, 6 Jun 2010 07:40:40 +0100, Paul Hyett
wrote:

On Sat, 5 Jun 2010 at 10:25:50, ned flanders wrote in
uk.sci.weather :


The probability of correct forecasting is something that has been
rattling around in my head for a while.

If on Day 1 of a 5 day forecast I'm told Rain is on the way only for
this forecast to change between rain/dry (for example) over the next
three days and then its actually dry on day 5, the forecast on day 1
was utterly worthless.

What if showers are forecast - some people are going to see the forecast
as correct, while others won't...

It's not just rain though, yesterday evening the Met Ofice website and
TV broadcasts forecast a maximum temperature here in S Hants of 19
degrees C., with maybe some showers.

It's turned out to be a dry and lovely warm day here, essentially the
same feel as yesterday (max 25.5 yesterday, 23.8 today). Now that
wasn't a problem as i suspected that this could happen but the bulk of
the population would just think the forecast was completely wrong. It
was pretty obvious that here, if the showers and cloudiness didn't
materialise (always distinctly possible in these conditions) the
warmth would hang on except on West facing coasts (Solent MRSC).

So WHY didn't the forecasts mention this as a caveat? Sometimes, I
despair.

--
Dave
Fareham

Whats the point of forecasts that dont turn out to be correct ?
 

On Sun, 06 Jun 2010 18:38:09 +0100, Dave Ludlow
wrote:


It's turned out to be a dry and lovely warm day here, essentially the
same feel as yesterday (max 25.5 yesterday, 23.8 today). Now that
wasn't a problem as i suspected that this could happen but the bulk of
the population would just think the forecast was completely wrong. It
was pretty obvious that here, if the showers and cloudiness didn't
materialise (always distinctly possible in these conditions) the
warmth would hang on except on West facing coasts (Solent MRSC).

So WHY didn't the forecasts mention this as a caveat? Sometimes, I
despair.

Just to add to this (and i rarely criticize forecasts), yesterday
evening on TV and the Met Office website, the local forecast for
today, here, was for dry conditions, pleasantly warm (21 deg C) and no
rain until tomorrow morning. I planned accordingly for a late
afternoon t-shirted bike ride. Well it's been raining on and off
(albeit lightly) since before 4pm and there's moderate rain close by.
Max temp 19.4 deg C so at least that was close enough to the forecast.

So the 12-24 hour regional forecasts *for here* have been
significantly wrong for the last 2 days. I wonder if they are going
for a hat-trick? ;)

--
Dave
Fareham

Whats the point of forecasts that dont turn out to be correct ?
 

On Jun 5, 7:54*pm, Graham Easterling
wrote:
On 5 June, 19:25, Len Wood wrote:





On Jun 5, 6:25*pm, ned flanders wrote:

The probability of correct forecasting is something that has been
rattling around in my head for a while.

If on Day 1 of a 5 day forecast I'm told Rain is on the way only for
this forecast to change between rain/dry (for example) *over the next
three days and then its actually dry on day 5, the forecast on day 1
was utterly worthless.

What probability of a likely event happening triggers it becoming a
'forecast'. Surely it needs to be 75% likely as that would mean it is
twice as likely to happen, as not.

Also why is that probabilty of a forecast never mentioned on tv/radio/
media forecasts. Is it because Jo Public has dumbed down so much over
the last few years that he has no concept what Porbability is?

To be fair, as Victoria Coren wrote in the Observer last Sunday:
The Met Office recently had its budget cut by £4.3 million. That may
not sound like much, but it was a loss of more than 20 million pine
cones.

Len Wood
Wembury

Pine cones? Down here we're more up to date!http://www.redbubble.com/people/bcim...074-7-you-what

Graham
Penzance

http://www.turnstone-cottage.co.uk/- Hide quoted text -

- Show quoted text -

OK Graham,
I know you've got superior seaweed in Cornwall!
But I am not sure the Mousehole Barometer is foolproof.

Len Wood
Wembury, SW Devon

Whats the point of forecasts that dont turn out to be correct ?
 

On Jun 7, 8:07*pm, Len Wood wrote:
On Jun 5, 7:54*pm, Graham Easterling
wrote:





On 5 June, 19:25, Len Wood wrote:

On Jun 5, 6:25*pm, ned flanders wrote:

The probability of correct forecasting is something that has been
rattling around in my head for a while.

If on Day 1 of a 5 day forecast I'm told Rain is on the way only for
this forecast to change between rain/dry (for example) *over the next
three days and then its actually dry on day 5, the forecast on day 1
was utterly worthless.

What probability of a likely event happening triggers it becoming a
'forecast'. Surely it needs to be 75% likely as that would mean it is
twice as likely to happen, as not.

Also why is that probabilty of a forecast never mentioned on tv/radio/
media forecasts. Is it because Jo Public has dumbed down so much over
the last few years that he has no concept what Porbability is?

To be fair, as Victoria Coren wrote in the Observer last Sunday:
The Met Office recently had its budget cut by £4.3 million. That may
not sound like much, but it was a loss of more than 20 million pine
cones.

Len Wood
Wembury

Pine cones? Down here we're more up to date!http://www.redbubble.com/people/bcim...074-7-you-what

Graham
Penzance

http://www.turnstone-cottage.co.uk/-Hide quoted text -

- Show quoted text -

OK Graham,
I know you've got superior seaweed in Cornwall!
But I am not sure the Mousehole Barometer is foolproof.

Len Wood
Wembury, SW Devon- Hide quoted text -

- Show quoted text -

Think I probably didnt put my questions clearly enough. What I was
trying to ask was what is the accepted probability that a weather
event will occur for it to become a forecast??

Whats the point of forecasts that dont turn out to be correct ?
 

On Jun 7, 10:10*pm, ned flanders wrote:
On Jun 7, 8:07*pm, Len Wood wrote:





On Jun 5, 7:54*pm, Graham Easterling
wrote:

On 5 June, 19:25, Len Wood wrote:

On Jun 5, 6:25*pm, ned flanders wrote:

The probability of correct forecasting is something that has been
rattling around in my head for a while.

If on Day 1 of a 5 day forecast I'm told Rain is on the way only for
this forecast to change between rain/dry (for example) *over the next
three days and then its actually dry on day 5, the forecast on day 1
was utterly worthless.

What probability of a likely event happening triggers it becoming a
'forecast'. Surely it needs to be 75% likely as that would mean it is
twice as likely to happen, as not.

Also why is that probabilty of a forecast never mentioned on tv/radio/
media forecasts. Is it because Jo Public has dumbed down so much over
the last few years that he has no concept what Porbability is?

To be fair, as Victoria Coren wrote in the Observer last Sunday:
The Met Office recently had its budget cut by £4.3 million. That may
not sound like much, but it was a loss of more than 20 million pine
cones.

Len Wood
Wembury

Pine cones? Down here we're more up to date!http://www.redbubble.com/people/bcim...074-7-you-what

Graham
Penzance

http://www.turnstone-cottage.co.uk/-Hidequoted text -

- Show quoted text -

OK Graham,
I know you've got superior seaweed in Cornwall!
But I am not sure the Mousehole Barometer is foolproof.

Len Wood
Wembury, SW Devon- Hide quoted text -

- Show quoted text -

Think I probably didnt put my questions clearly enough. What I was
trying to ask was what is the accepted probability that a weather
event will occur for it to become a forecast??- Hide quoted text -

- Show quoted text -

That's actually an extremely good question. Ally it to why and how the
decisions are made as to how those probablilities are calculated (an
80% chance of a weather event occurring leads to a red warming on the
charts - how did they get to 80%?) and you have the "secret" of
weather forecasting.

I always accept that the bottom line to a forecast is skill and
experience and the MetO forecasters are the best in this part of the
world, but skill and experience actually equals subjectivity; but
here's the PR achilles heel of the MetO. Instead of sharing the
"secret", in terms of forecast outcome probabilities and admitting
that there are times when the forecast is unlikely to achieve outcome
(especially in timescales of over a week) it is presented as if it
will come true and there is no analysis, except in extreme instances,
of why forecasts turned out to have gone mammaries skywards.

Why can't the MetO PR department see this and include their target
audience so much better? If they did, people in general, would feel so
much better about their national weather service provider. So much
scientific talent, so much superb, world-leading research and so often
a PR disaster area. It must make Met Office employees cringe!

Whats the point of forecasts that dont turn out to be correct ?
 

On Jun 7, 10:36*pm, Dawlish wrote:
On Jun 7, 10:10*pm, ned flanders wrote:





On Jun 7, 8:07*pm, Len Wood wrote:

On Jun 5, 7:54*pm, Graham Easterling
wrote:

On 5 June, 19:25, Len Wood wrote:

On Jun 5, 6:25*pm, ned flanders wrote:

The probability of correct forecasting is something that has been
rattling around in my head for a while.

If on Day 1 of a 5 day forecast I'm told Rain is on the way only for
this forecast to change between rain/dry (for example) *over the next
three days and then its actually dry on day 5, the forecast on day 1
was utterly worthless.

What probability of a likely event happening triggers it becoming a
'forecast'. Surely it needs to be 75% likely as that would mean it is
twice as likely to happen, as not.

Also why is that probabilty of a forecast never mentioned on tv/radio/
media forecasts. Is it because Jo Public has dumbed down so much over
the last few years that he has no concept what Porbability is?

To be fair, as Victoria Coren wrote in the Observer last Sunday:
The Met Office recently had its budget cut by £4.3 million. That may
not sound like much, but it was a loss of more than 20 million pine
cones.

Len Wood
Wembury

Pine cones? Down here we're more up to date!http://www.redbubble.com/people/bcim...074-7-you-what

Graham
Penzance

http://www.turnstone-cottage.co.uk/-Hidequotedtext -

- Show quoted text -

OK Graham,
I know you've got superior seaweed in Cornwall!
But I am not sure the Mousehole Barometer is foolproof.

Len Wood
Wembury, SW Devon- Hide quoted text -

- Show quoted text -

Think I probably didnt put my questions clearly enough. What I was
trying to ask was what is the accepted probability that a weather
event will occur for it to become a forecast??- Hide quoted text -

- Show quoted text -

That's actually an extremely good question. Ally it to why and how the
decisions are made as to how those probablilities are calculated (an
80% chance of a weather event occurring leads to a red warming on the
charts - how did they get to 80%?) and you have the "secret" of
weather forecasting.

I always accept that the bottom line to a forecast is skill and
experience and the MetO forecasters are the best in this part of the
world, but skill and experience actually equals subjectivity; but
here's the PR achilles heel of the MetO. Instead of sharing the
"secret", in terms of forecast outcome probabilities and admitting
that there are times when the forecast is unlikely to achieve outcome
(especially in timescales of over a week) it is presented as if it
will come true and there is no analysis, except in extreme instances,
of why forecasts turned out to have gone mammaries skywards.

Why can't the MetO PR department see this and include their target
audience so much better? If they did, people in general, would feel so
much better about their national weather service provider. So much
scientific talent, so much superb, world-leading research and so often
a PR disaster area. It must make Met Office employees cringe!- Hide quoted text -

- Show quoted text -

I am not sure Met Office bods know why on some occasions the numerical
model output goes mammaries skywards as you so delicately put it Paul.
There are a number of possible reasons (handling of the physics,
maths, feedbacks, quality and quantity of data...) and it is all to do
with the construct of the model and its ability to forecast some
weather situations better than others.
Do the public just want more honesty when forecasts go wrong? This is
more often than Met Office care to admit if you want a very local
detail forecast, and one for more than 5 days ahead. I think the
public might just view the Met office as whingeing if they tried each
time to explain the uncertainties when the forecast goes haywire.
Ned, you started the discussion talking about the forecasting of rain.
The timing of cloud and quantity of rain is one of the more difficult
things to forecast. So many feedbacks across all time and space
scales.
Especially showery rain.

Len Wood,
Wembury, SW Devon