www.bbc.co.uk/news/uk-18744189
This woman predicts the weather for the next century.
Yet the same team, using the same computer model, cheerfully predicted a barbecue summer last year. Why are their forecasts 100 years hence more reliable than their forecasts for the coming three months (used to be before they stopped them as they were too inaccurate...)?
|
|
From the same university department whose legendary lies and manipulation of weather stats a few years ago made the news ?
|
|
Bet she's a laugh a minute in a barbie !
|
Well of course, estrapolation from current trends is one way to forecast..
By definition it's always wrong as something changes.
Given that they did not forecast the flooding this year, I can safely say the Government can save a great deal of money by not funding them.
|
|
Isn't there a big difference between predicting general trends over decades and predicting what's going to happen in the next few months?
|
>> Isn't there a big difference between predicting general trends over decades and predicting what's going
>> to happen in the next few months?
Yes. Predicting the weather in 5 minutes time is easy: "the same as now" will be right almost all of the time.
Predicting the weather this afternoon is fairly easy: "the same as now" will be right most of the time.
Tomorrow's weather is a bit trickier. "The same as today's" will be right more often than not.
Next month's weather is impossible.
Next season's weather is entirely impossible.
Yet somehow you've been sucked in by the global warming brigade into thinking that predicting the weather in 100 years' time is easier????!
|
>> Yet somehow you've been sucked in by the global warming brigade into thinking that predicting
>> the weather in 100 years' time is easier????!
No. ( !!!!? )
|
|
Sorry, I've obviously misunderstood your suggestion that forecasting 100 years ahead is not the same as forecasting the coming season.
|
>> Sorry, I've obviously misunderstood your suggestion
More of a question really (hence the '?') - don't know enough about it.
>> that forecasting 100 years ahead is not the same
>> as forecasting the coming season.
I was questioning that, yes, but sorry, I missed the connection between my question and your statement.
|
I'm inclined to think that we've (or more specifically, my parents' generation) has well and truly stuffed up the climate to the extent that the Azores High is unlikely to fall across the UK again.
We need to understand that Britain is not global. Sure, it's chucking it down here but it's still very warm in the South East; Central Europe is suffering temperature in excess of 38C and you only have to look at the news to see the recent severe weather in Russia and the US.
I certainly don't think you can predict accurately more than 48 hours in advance, but you only have to take one look at the charts ( weather.noaa.gov/pub/fax/PGDE14.PNG ) to see how far south the polar jet stream is at the moment. If anything it's moving further south comparing tomorrow's chart with today's.
It's not going to shift north in a hurry, and once it does high pressure and settled weather will not develop instantly. Hence the weather forecasters could well be right when they suggest that the crap weather will persist into August.
Last edited by: Fursty Ferret on Mon 9 Jul 12 at 17:20
|
Thanks FF. I can't read weather charts like you, but I am acutely aware that Britain is a slightly special case. Even when it's not unpredictable here, it's not predictable in the same way as other parts of the world can be.
We also have much more temperate weather than other bits of the the northern hemisphere in the same latitudes.
Like probably everybody here, I learned at school that Britain has a temperate climate, but I didn't appreciate what that meant until I'd been somewhere else!
Even north German winters would be a shock in the south east of England. Brrrr.
|
>I certainly don't think you can predict accurately more than 48 hours in advance,
Forecasts provided by the Met Office and ECMWF currently have a 3 Monthly mean accuracy of around 90% at 5 days - the most accurate available anywhere. Accuracy at 2 days is pretty much spot on.
tinyurl.com/d2mhdx7
Mapmaker's original post confuses weather forecasting with climate prediction and the Met Office with UEA's Climatic Research Unit.
|
Errr Kevin... weather forecasts for what planet are accurate to 90% within five days? And 90% of what?
The Met Office's accuracy results are published here.
www.metoffice.gov.uk/about-us/who/accuracy/forecasts
They offer a 90% accuracy for getting the temperature right in an early morning forecast to within +/- 2 degrees THAT DAY. For the night following that drops to 80%, 84% for the following day, and 75% for the night following the second day. Statistics are not published for their accuracy at forecasting temperature more than 48 hours ahead.
When it comes to predicting rainfall or sunshine, they manage a 70%ish accuracy for the day in question.
>>Mapmaker's original post confuses weather forecasting with climate prediction and the Met
>>Office with UEA's Climatic Research Unit.
No it doesn't. Perhaps you can come up with some backing for your 90% at 5 days and "spot on at 2 days" figures first.
Just to start, there is no way of assessing whether the climate forecasts are any good. The only thing that can be tested is short-term predictions. If the latter are no better, statistically, than picking at random, what faith can we have in the long-term predictions?
Last edited by: Mapmaker on Tue 10 Jul 12 at 09:44
|
Long range weather forecasts never provide a daily forecast. For example, a long range forecast won't say
On 10th Aug 2012, the morning will be sunny, the afternoon cloudy MAx temp 23c, min 12c, 80% chance of rain in the period 21:00 -23:00
It simply isn't provided, so you can't argue they are wrong,
|
>>It simply isn't provided, so you can't argue they are wrong,
Ah, that must be why Kevin says they're "spot on".
BTW, did somebody mention a barbecue...?
For the avoidance of doubt, when I wrote "Just to start, there is no way of assessing whether the climate forecasts are any good" the reason there is no way of testing is because we have to wait 100 years to see whether they were any good.
Last edited by: Mapmaker on Tue 10 Jul 12 at 10:47
|
>Ah, that must be why Kevin says they're "spot on".
>BTW, did somebody mention a barbecue...?
To address your points Mapmaker:
>>Errr Kevin... weather forecasts for what planet are accurate to 90% within five days? And 90% of what?
The 90% is the Anomaly Correlation Coefficient (expressed as a percentage) which is the accepted measure of accuracy or forecast skill.
tinyurl.com/73lo7u5
>>No it doesn't. Perhaps you can come up with some backing for your 90% at 5 days and "spot
>>on at 2 days" figures first.
Yes it does - lookup the difference between climate and weather. You also claimed that "the same team, using the same computer model, cheerfully predicted a barbecue summer last year".
Claire Goodess is from the CRU at UEA not UK MET, and UK MET's seasonal model would be useless for a 100yr climatic forecast.
>>before they stopped them as they were too inaccurate...)?
UK MET have learned their lesson. Despite presenting their seasonal forecasts as probabilities they were seized upon as predictions and suffered the consequences. It's our loss.
Anyway, here you are, no 'perhaps' about it:
The WMO and NOAA maintain stats on the accuracy of the major forecast organisations on a rolling basis. The NOAA stats are here:
tinyurl.com/7w22op5
The ACC fall-off for June 2012 in both hemispheres is here:
tinyurl.com/d4y4ptl
You will see that ACC is very close to 1 at 2 days out, and only falls below .9 at around 5 days (EC on the graph is ECMWF, UKMO is UK Met).
>>Just to start, there is no way of assessing whether the climate forecasts are any good.
I don't have a clue what the CRU at UEA base their forecasts on so I can't comment.
>>If the latter are no better, statistically, than picking at random, what faith can we have
>>in the long-term predictions?
The statistics prove you wrong. See the links above.
If you want a simple introduction to weather forecasting, try this:
tinyurl.com/d5ke3qg
Let me know when you've taken in all in and we can both move on to beginner level.
|
Very funny. Here it is, in plain English from the Met Office's website:
Maximum temperature - first day of forecast
89% of maximum temperature forecasts are accurate to within +/- 2°C on the current day (36-month average).
Target for 2012/13 is 85.0%.
If the Met Office thought they were 90% correct at 5 days out, then they'd trumpet it on their website. As it is, they think they are nearly 90% correct for the same day in temperature terms.
|
>>UK MET have learned their lesson. Despite presenting their seasonal forecasts as
>>probabilities they were seized upon as predictions and suffered the consequences. It's our loss.
In principle, I agree with you, of course. However giving probabilities of 0.3 average, 0.4 hotter than normal, 0.3 cooler than normal as they famously have, tells us absolutely nothing at all and they'd be better off not bothering.
|
>Very funny. Here it is, in plain English from the Met Office's website:
That is for a very small subset (45 sites in the UK) of their complete forecast.
|
>> >Very funny. Here it is, in plain English from the Met Office's website:
>>
>> That is for a very small subset (45 sites in the UK) of their complete
>> forecast.
>>
I'm not sure whether you're trying to impress me. Any fool can predict next week's weather in the Sahara, which is quite a big place really. Getting the average accuracy up in order to convince people like you that they're good at forecasting by largely ignoring the one thing that the Country actually needs just goes to show what a scandal the whole weather forecasting industry is.
It's the British weather that's so hard to predict.
|
>I'm not sure whether you're trying to impress me.
Mapmaker,
you asked, with a hint of sarcasm "Perhaps you can come up with some backing for your 90% at 5 days and "spot on at 2 days" figures first."
When I do exactly that, with stats maintained by a respected third-party source, you think I'm trying to impress you?
How odd. Would you be less impressed if I used Wiki in future?
>Getting the average accuracy up in order to convince people like you..
People like me? Who would they be pray tell?
Last edited by: Kevin on Thu 12 Jul 12 at 19:44
|
>>People like me? Who would they be pray tell?
Blind believers in anthropomorphic global warming pseudoscience.
Hockey stick, anyone?
If the Met Office thought it could convince the average intelligent member of the public that it was producing forecasts that were 100% accurate for the next 48 hours, it would trumpet that fact loudly. It doesn't.
Whatever you suggest it is predicting with 100% accuracy over a 48 hour period, it is not the weather that is being experienced by those of us who pay for it. (Noting that also it receives funding from its commercial activities.)
Last edited by: Mapmaker on Fri 13 Jul 12 at 10:31
|
>Blind believers in anthropomorphic global warming pseudoscience.
And you, of course, would know exactly what my views are about man-made global warming.
Maybe Mystic Mapmaker would be a more apt moniker than just plain old Mapmaker?
>If the Met Office thought it could convince the average intelligent member of the public
>that it was producing forecasts that were 100% accurate for the next 48 hours, it would
>trumpet that fact loudly. It doesn't.
I provided you with links to independent stats showing the accuracy of their complete forecast. Sticking your fingers in your ears and going "La, la, la-la, la" doesn't change that.
>Whatever you suggest it is predicting with 100% accuracy over a 48 hour period..
Did you bother to read the info on the NOAA site?
>..it is not the weather that is being experienced by those of us who pay for it.
A study conducted in 2007 estimated that the UK economy benefits by more than £7 for every £1 of UK MET funding. Admittedly it was commissioned by PWSCG, but that doesn't sound like a bad deal to me.
|
>>I provided you with links to independent stats showing the accuracy of their complete forecast
Like most Britons I don't really worry about the weather in the rest of the world. The met office's website tells me that they can predict whether or not it will rain TODAY in Britain with a 70% accuracy.
A simple question, please answer with a yes or no. "Does that feel anything like nearly 100% accuracy over 48 hours?"
|
>Like most Britons I don't really worry about the weather in the rest of the world.
I'm not sure where you are going with this.
Are you saying that we should judge UK MET's forecast accuracy on a sample size of 45 UK sites and ignore the thousands of other observations from the rest of the world?
>The met office's website tells me that they can predict whether or not it will rain TODAY
>in Britain with a 70% accuracy.
You are cherry picking Mapmaker and very conveniently ignoring something that they make an explicit point of explaining -
"It is difficult to forecast rainfall at a pinpoint location, therefore the target for this element is not as high as for others. For example, one location may have rain and a location a mile away may remain dry."
>A simple question, please answer with a yes or no. "Does that feel anything like nearly
>100% accuracy over 48 hours?"
After driving back from Torquay today, 4hrs in constant heavy rain when this mornings forecast was for cloud and light rain, that's an obvious no.
But that does not alter the fact that when taken as a whole, UK MET's 2 day forecast is "pretty much spot on" and ~90% accurate at 5 days. Links above.
|
So you're telling me that a fancy graph, with standards created by the industry who have vested interests in flattering their own results, proves that they're 90% accurate over a 5 day period, whereas you're also telling me they've no idea whether or not it will be constant heavy rain or a bit of light rain today.
You can point to links above as much as you like. But they're not measuring anything to do with the weather we receive, are they.
You cannot actually tell me what is likely to be 100% right for the next 48 hours, can you? All you do is to point me at graphs.
|
You guys need to scope the question. Asking whether a weather forecast is accurate is too ill-defined.
I'm looking at the local forecast here from the Met Office because I am about to walk 2 miles to the shop and it looks like rain.
The forecast says there is a 40% chance of rain today, and a 20% chance of rain between 10.00 and 13.00. So whether I get wet or not, the forecast could be said to be accurate - you'd need to look at lots of forecasts to see whether their percentage probablilities are reliable but that doesn't tell me with any certainty at all whether it will rain today.
If it turns out they are spot on with their probabilities, then you can still argue that 40% forecasts are wrong 60% of the time if you take 40% as a "rainy day" forecast, or wrong 40% of the time if you take it as a "dry day" forecast.
I guess that's one reason why they chart temperature forecasts, at which they are fairly good short range.
From a practical viewpoint, I'm only interested in weather where I am. Not even 5 miles away. As one of you has already pointed out, it could be raining here when I set off and still dry at the shop when I get there - one mile as the crow flies. That kind of forecasting is always going to be difficult.
|
Mapmaker,
You asked for some "backing" for my statement about UK MET's forecast accuracy. I provided you with links to articles and data from NOAA, the US National Weather Service, the Canadian Met Center and a Prof. from University of Washington.
Because they do not support your position you dismiss them offhand.
You then attempt to imply that they would falsify the data because of "vested interests in flattering their own results".
Your argument is absurd.
>But they're not measuring anything to do with the weather we receive, are they.
You really have no clue how forecast accuracy is measured do you?
>You cannot actually tell me what is likely to be 100% right for the next 48 hours, can you?
>All you do is to point me at graphs.
And even less of a clue of how forecasts are produced.
|
Rhetorical question I take it.
Short range (<= 2 days) forecasts are mostly OK.
The longer the range and the more changeable the weather, the worse it gets. Butterfly's wing and all that. Britain has very changeable weather so we can't hope for much beyond a day or two.
By the time it gets to predicting the next season's weather here, recent experience suggests a very poor correlation between forecast and actual. ISTR we have been forecast severe winters that turned into the mildest on record, and barbecue summers that were among the wettest in recent years.
It's also now official I think that the last quarter has not only been the wettest expected drought ever, but the wettest April-May-June on record, period.
|
I want to the man from Thames Water to choose my Lottery Tickets.
when the hosepipe ban was brought in he said. "We would need it to rain heavily for three months before we can lift it" I didn't realise he was a prophet. He has a direct line to Allah/god/buddha, whoever.
|
>> I want the man from Thames Water to choose my Lottery Tickets.
I'd rather trust the man from Del Monte. He's better dressed too.
|
|
...................and his product tastes better, too!
|
Of course forecasting 100 years is for muppets.
All you need is one BIG volcanic outburst and it will be useless scrap paper.
We are generally overdue a large volcanic explosion throwing ash into the air and leading to a Mini Ice Age lasting a couple of decades.
see:
news.softpedia.com/news/What-Caused-the-Medieval-Little-Ice-Age-video-41383.shtml
and
en.wikipedia.org/wiki/Little_Ice_Age
Funny the Met Office never mention it...
Last edited by: madf on Mon 9 Jul 12 at 17:59
|
>> en.wikipedia.org/wiki/Little_Ice_Age
>>
>> Funny the Met Office never mention it...
we are due a meteorite as well, the met office never mentions that either
|
"we are due a meteorite as well, the met office never mentions that either"
don't tell anybody - it will cause a panic ( I heard it hits Surrey next Wednesday)
Last edited by: CGNorwich on Mon 9 Jul 12 at 19:17
|
>> "we are due a meteorite as well, the met office never mentions that either"
>>
>> don't tell anybody - it will cause a panic ( I heard it hits Surrey
>> next Wednesday)
Really? I shall go on a trip to Kent that day then, should be ok there.
|
|
No - you need to go somewhere really remote - like Swaffham.
|
|
ah well, if it has to be remote, how about Darsham?
|
Should be safe in Basingstoke. No self-respecting meteorite would want to land here if it could help it.
Especially on Wednesday when it would be greeted by the assembled masses in shell-suits welcoming the Olympic torch :-(
|
If you want remote with a train station you can't do better than Berney Arms.
en.wikipedia.org/wiki/File:Berney_Arms_railway_station_1.jpg
|
|
blimey, makes Shippea Hill look like suburbia.
|
There's only one railway station I'd be prepared to be stuck on.
I was stranded there in, IIRC, 2005 in a thunderstorm. Halfway down the first pint of something from the Titanic brewery, the power went off and the towels went on. I was obliged to point out to the clearly non-technical barmaid that hand-pumps needed no power, and normal service was resumed, to general approbation and relief.
The rain was arriving in stair rods, and cascading out of the box gutter between the station building and the platform roof, forming a waterfall curtain at the buffet exit which sadly meant having to drink two more pints while cut off from the outside world.
Happily I had taken a break from narrowboating on the Huddersfield Narrow Canal, whence I and my crew had hoofed it to this legendary watering hole, so didn't have to drive that day. Not at more than 3mph anyway.
www.buffetbar.org/
|
Stalybridge is earmarked to have a chunky bit of money spent on it soon...I don't think the buffet will be affected.
Did you know it's the only station on the Lancashire & Yorkshire Railway which wasn't in either county......being in Cheshire ?
Not a lot of people know that.....or want to !
One of the smallest stations is Damems on the Keighley line......rumour has it that the station building was loaded on a goods train that stopped there, the porter thinking it was a crate !
Ted
|
|
Not that I'm beer-obsessed or anything, but you can often get a pint of 'real' ale on the Keighley & Worth Valley. Nice little line, which includes the well known Oakworth station.
|
>> If you want remote with a train station you can't do better than Berney Arms.
>>
>> en.wikipedia.org/wiki/File:Berney_Arms_railway_station_1.jpg
>>
Gone over the top with the furniture in the waiting room, haven't they?
I bet I paid for that too!
;-)
|
The story is that the when the landowner sold his land to the GNER one of the provisions was that the company provide a stop there 'in perpetuity".
It really is remote - in the Halvergate marshes near Yarmouth and the nearest road is about 3 miles away - You can only access the palce on rail by foot or on a boat. I think the train only stops there twice a day. Mainly used by walkers and visitors to the pub whihc is only open in summer.
|
The last time I was at that pub, two shirt lifters were running the place.
Nice dinner though. bangers and mash at a reasonable price as well.
|
>> The last time I was at that pub, two shirt lifters were running the place.
>>
>> Nice dinner though
Why shouldn't it be?
|
...Why shouldn't it be?...
Focus,
Are you trying to impersonate Bromptonaut?
|
>> Are you trying to impersonate Bromptonaut?
Bromp's talents would be wasted on that one!
|
|
Just remembering things that 'stood out' as it were :-)
|
>> >> Nice dinner though
>>
>> Why shouldn't it be?
Just goes to show that the taste buds of Homophobic people are no different to anyone else's.
|
>> Just goes to show that the taste buds of Homophobic people are no different to
>> anyone else's.
I hope that was irony, or you can consider yourself red-gonged.
I don't think it follows that Diddy etc considers confirmed bachelors in general to be bad cooks, and whilst it might be considered bad form to mention sexual orientation gratuitously, the fact is that many people would remark on it or find it interesting as a point of distinction in, for example, a pub.
Neither am I prejudiced. Some of my best friends are politically correct but I still put up with them.
|
>> Neither am I prejudiced. Some of my best friends are politically correct but I still
>> put up with them.
Nor me, some of my best friends are terrible cooks.
|
|
None of my cooks are friends.
|
>> I don't think it follows that Diddy etc considers confirmed bachelors in general to be
>> bad cooks, and whilst it might be considered bad form to mention sexual orientation gratuitously,
It was the use of 'though' which appeared to imply that the pub was somehow sub-standard due to their sexual orientation. That's how I read it anyway, but happy to be corrected.
|
>> It was the use of 'though' which appeared to imply that the pub was somehow
>> sub-standard due to their sexual orientation. That's how I read it anyway, but happy to
>> be corrected.
I'm in no position to correct you, as I don't know what other people are thinking. It was obviously a construction that would have been a severe gaffe at the EHRC annual conference, but that's not real life is it?
Surprised at the red-gonging going on. I wonder if I have really offended anybody. If so I am irredeemable.
|
>> Surprised at the red-gonging going on.
None from me BTW, like you say could just be down to interpretation. Didn't have a problem with your post, just explaining my reply to diddy's.
|
The simple test is would you use the phrase in question face to face without expecting the person so describe to to be offended
If the answer is no then clearly you should not be using it here. Forget political correctness, just try good manners.
|
>>The simple test is would you use the phrase in question
Ah I see, the language - I'd forgotten about that, oddly enough!
|
>> Nice dinner though. bangers and mash at a reasonable price as well.
well as long as you didn't take the meat and two veg....
|
>> I you can't do better than Berney Arms.
>>
I spent a happy few hours there in 1984 waiting for the tide.
Several pints later, we were nearing Yarmouth when the new bridge under construction across Breydon Water loomed up in the darkness. There seemed to be some sort of obstruction in mid-channel, but nothing to indicate which side to go. Unable to stop in a small boat on a strong tide we guessed, by chance correctly.
Later in daylight we saw the machinery in the other channel, and the hawsers stretched across the river.
|
Had a downpour this morning, of monsoon proportions, tho shorter. So bad its scoured all the moss off the roof tiles and flushed it down my storm water drain, blocking it and flooding the rear and side of the house to a depth of two inches, (just below air brick level fortunately)
SO I had to spend 30 minutes with my hand down the drain digging it out again.
|
>> SO I had to spend 30 minutes with my hand down the drain digging it
>> out again.
>>
Be thankful it was a storm drain then..and not a domestic one...
|
|
Have rods for that one as it clogs up on an annual basis.
|
Would seem to suggest someone who lives there is full of....
:-))
Sorry ! Couldn't resist !!!
|
>> Would seem to suggest someone who lives there is full of....
>>
>> :-))
>>
>> Sorry ! Couldn't resist !!!
>>
Pork pies?
|
|
Jet-stream forecast to be moving North again to it`s proper position after this week-end, Summers here!!! - Yeah right!
|
|
havnt seen a moth or a cranefly yet this summer. keep the rain on...can at least have the windows open in the evenings
|
How about the blimmin midges, 10 or more attack me every morning while I squeegee the car windows!
Never had that before in the 15 years we've lived down ere.
|
Barbecue summer model confidently predicts 100yrs - R.P.
