Quote: "The forecast was wrong"

[Jack Harrison]

I have become a little tired of people forever going on about “the forecast being wrong”. Yes they do go significantly wrong once or twice a month but in general if the user doesn’t expect too much (eg scattered showers: hit-and-miss: rain arriving in the afternoon: these deliberately are meant to imply a degree of imprecision) then forecasts today are stunningly accurate. But it is essential to listen for all the nuances in a forecast. Media presenters don't help here (specialist aviation forecasts use percentages which is far better). They seem to think the public want certainties - impossible in weather forecasting.

And as for being a know-all, well yes, to an extent that is true. For some ten years, I ran a website, weatherjack, that gave specialised forecasts for gliding (and I earned a few bob doing so). I also gave live (in person) forecast briefings at gliding competitions including National Championships. I received awards from the British Gliding Association and the Royal Aero Club for my work.

I am defending the Met Office and other forecasters and criticising those who expect more from a forecast than is possible. I suspect that human nature comes into this – people hear in a forecast what they want to hear. My recent criticism might have appeared to be aimed at one particular person but that was not the intention.

A few years ago, I made a slimmed down forecast website to help people like butterfly enthusiasts. I published the link on ukb. It got very few hits. So I have tried to help people understand the weather but if people are not interested, then there is nothing I can do. You can take a horse to water, but you can’t make it drink.

Jack

[Simon C]

Hi Jack,

I agree entirely about the benefits of percentages. Why don't they get used here? I spent 12 days walking and sightseeing in Switzerland earlier this summer, at a time when the weather as a whole was rather unsettled and changeable. However, by using the meteonews.ch web site I was able to catch a train and travel either north, south or east to ensure good weather, such that in the whole time there were only two short periods when we experienced rain whilst out walking. Maybe it is something to do with the geography, but the forecasts were remarkably accurate and reliable, e.g. with the weather on one side of Lake Biel correctly predicted different to that on the other (about 3 miles away). If can't help thinking that it had been necessary to rely on a Met Office or heaven-forbid BBC style weather forecast the vacation could easily have been a wash-out.

The great thing is, meteonews.ch gives forecasts (with percentages) for places in the UK too

Simon

[Jack Harrison]

Thanks SimonC. I wasn't familiar with meteonews.ch - it's very good.

This too is pretty accurate

http://cab.inta-csic.es/rems/marsweather.html

Windy place - 25 metres/sec is not far short of 60 mph. But it wouldn't feel like in the thin atmospehre.

Jack

[Simon C]

Excellent. I'll bookmark that. Will be useful next time I am heading that way. Maybe if I am after this:

http://cintos.org/SaginawManifold/Obliq ... age65.html

Simon

[selbypaul]

Hi Jack
Totally agree with what you're saying, I have to defend the Met Office all the time too.

An interesting insight into why some people "just don't get it" comes from watching weather forecasts with my mum. We both watch the same tv forecast and get a completely different idea of what was just presented. I suspect its because my mum can't read maps, and so focuses on what is actually being said. I instead focus on precise map locations at price time periods.

As a result, I've forced my mum out for many a walk that she thought she was going to get soaked on, and actually got severely sunburnt instead! I suspect this is just an example of why psycologists call "different learning styles" - basically different people interpret the same information in different ways simply because they use different senses and/or whether they are evidence based or instead are more intuitive.

As an evidence based analyst (my job), clearly I agree with you Jack. But hopefully I've provided an insight into why others might not agree.
Paul

[David M]

Generally, I find that following day forecasts are surprisingly accurate. The problem is that too many people take for granted those mid to long range forecasts and make plans around them as a result.

You CAN'T do that in this country. You've just got to accept whatever's thrown at you at short notice.

[millerd]

I would also agree wholeheartedly with Jack. The Met Office (accept no substitutes) forecasts are pretty accurate up to five days ahead, but you have to look at all the info available to see how much variability there may be. Unsettled weather patterns are just that - there is a higher likelihood of change. Being a bit of a weather geek, I always look at the synoptic charts ("real" weather forecasts!), satellite pictures and the rainfall radar when deciding whether to go out on a given day. I've several times left cloud, rain and unspectacular temperatures at home with a high level of confidence that my destination will be bathed in sunshine - and not been disappointed. At least a geography A Level meant something when I did it.

Dave

[essexbuzzard]

Well,i can't agree,not entirely,anyway. For a butterfly watcher,i am unusurally on sunshine,and i wont a long way to a site unless the weather is looking reliable.So if i am going a long way,and sunny skies are forecast before i go,i'm going to be less than pleased if there is no sun when i get there. This was the case on Sunday,and again today,when a sunny morning was forecast,with cloud gradually building,chance of a late shower. In fact the sun hardly appeared until 3pm,due to high cloud which blocked out the sun,which meant most of my day was wasted,from a butterfly point of veiw at least.

[MikeOxon]

On the 'Today' programme the other morning, I heard a discussion in which the Met Office person said that the various computer models often produce different results and the forecaster has to make a judgement. This indicates to me that there are still a lot of unknowns in the models.

From my own work, I know that predicting precipitation from a cloud, based on its water density, is very uncertain - if the water density is very high, it will definitely rain and, if very low, it won't but there is a wide 'grey area' in between. I suspect that global changes also affect the reliability of models, as these changes alter some of the assumptions on which the models are based.

Mike

[Simon C]

Which is, of course, what the percentage is. I would much rather know that the percentage chance of rain is 25%, then have a forecaster make a judgement that it is not going to rain and say that. I am perfectly relaxed about the idea that different models give different predictions, and would be surprised given the chaotic nature of fluid dynamics were it any different. Maybe it's because I work in quantum physics where everything is a probability, but I find the percentages and the implied degree of confidence in the forecast easy to understand, and a valuable piece of information.

Simon

[NickMorgan]

Here is a chance for me to learn something that I have always wanted to know. I have never understood what was meant by the percentage chance of precipitation. OK if it said 100% (which I have never seen) then I guess they think it is definitely going to rain. 50% means what? There's a 50/50 chance of rain at some point during that day, or it is predicted to rain for half of the day?

[Jack Harrison]

Nick asked:

50% means what? There's a 50/50 chance of rain at some point during that day, or it is predicted to rain for half of the day?

Not for half the day but a 50% chance during the periods specified. But I agree Joe Public forecasts need a lot of improvement in their presentation.

Compare the aviation forecast for my local airfield, Marham:
PROB30 TEMPO 2008/2014 6000 -RA SCT012
The interpretation of the aviation code is that there is forecast to be a 30% probability that temporarily on 20th of month between 0800 hours and 1400 hours there will be -RA (meaning light rain)

That is unambiguous and clear (Whether or not the forecast turns out to be correct is another matter. However, there is in fact intermittent light rain at present - so far, so good)

Jack

[CJB]

Good morning UK b'ers,
I have always had a bug-bear about the weather and forecasts and consequently pay little to no attention.
I know it sounds facetious, but I simply wake up and look out of the window, like the quote from Good Morning Vietnam 'You got a window? Open it!'
Flutter on!
CJB

[Goldie M]

British weather needs a sense of humour, thank goodness British people have one Goldie

[Lancastrian]

Good morning UK b'ers,
I have always had a bug-bear about the weather and forecasts and consequently pay little to no attention.
I know it sounds facetious, but I simply wake up and look out of the window, like the quote from Good Morning Vietnam 'You got a window? Open it!'
Flutter on!
CJB

This is generally very sound advice , especially when one is considering venturing out locally, ie down to the "local patch". However, when one is planning a trip out further afield it is essential that some general idea of the days weather is known beforehand. I, for instance, need to travel c50 miles just to get to the "promised land" of the Morecambe Bay area, and as fuel prices are what they currently are (extortionate ! ) checking out the forecast is an absolute must....

[essexbuzzard]

My point exactly,as i said in my earlier post. The forecast for rain or no,seems fairly reliable,but the forecast cloud cover much less so-for any butterly enthusiast traveling afar,it is absolutely essential they get this right.
And they don't!

[Jack Harrison]

Essexbuzzard:

The forecast for rain or no, seems fairly reliable, but the forecast cloud cover much less so

Even wearing my retired forecaster's hat, I have to agree (I was not a professional meteorologist, but an experienced gliding weather forecaster).

The extent and amount of medium and high cloud always gave me problems. Although I wasn't following the forecasts in detail for last Monday, I am aware that the gliding forecasters were caught out by the extent of the high cloud that drifted in from the west. Medium and high cloud from upper weak frontal systems remains a nightmare for forecasters.

One more straightforward scenario that occurs often is the spread out of cumulus as the day progresses. After a fine clear promising morning, cumulus cloud develops (which glider pilots love and at this stage might even benefit butterfliers as there are brief cloudy moments that calm the butterflies down before the sun comes out again). But often, the cumulus spreads out into a layer of stratocumulus that cuts off the sun for many hours (to the despair of glider pilots and not usually helpful for butterfliers). Here is a small photo from a tutorial I wrote many years ago and show the early stages of spread out - butterfliers be prepared for cloud that will last several hours.

Hint: if the first wisps of cumulus form as early as 0930 hours BST, expect spread out. If the first puffs don't occur until after 1100 hours, then you are probably OK.



Jack

[David M]

Hint: if the first wisps of cumulus form as early as 0930 hours BST, expect spread out. If the first puffs don occur until after 1100 hours, then you are probably OK.

Thanks, Jack. I shall bear that in mind for future trips.

[David M]

One last 2012 hurrah this Saturday, perhaps?

Looks like practically the entire country is going to be bathed in uninterrupted sunshine and 13-17c temperatures!

Not quite as good as the corresponding weekend last year when records were set and everywhere generally was in the 70s, but in 3 or 4 weeks time we would happily pay £50 for these conditions!

Go out and make the most of it. April warmth is a long way away.