The forecast was accurate for Herts.
I agree that the forecasters ought to say "highly probable", "possible", etc. However, I guess they have to err on the side of pessimism. Last night's chart was a classic poser: the interaction between a strong front from the west and the cold air sitting over the east -- which would prevail and how quickly, and would the certain precipitation reach the ground as rain or snow, and how wide an area would the snow cover, etc.?
|
One might observe that "Up to 8 inches of snow" clearly includes the number 0.
It does not help at all to issue "severe weather warnings" on a frequent basis, only for nothing to happen.
|
I think the whole TV weather thing has become farcical recently. Last night the local (NE) news went on and on about the 'heavy' snow.
Its not as if it nevers snows in winter is it? And all the usual interviews with the local gritters (patting themselves on the back) and plod warning us not to drive unless its essential (yeah, tell my boss that one) start to get a bit silly.
Its Britain, it rains a lot, and its snows a bit in winter - surely we're used to that by now!
|
|
Weather forceast is a waste of time, but the temperatures are quite accurate, but cloud and precipitations are not at all.
|
|
If someone can tell me with absolute accuracy what the weather will be on the 10th December could they also tell me what the winning lottery numbers will be. :)
|
jbif is precisely correct. Weather forecasts can be, and often are, wrong, but they only mislead those who are easily led.
What is misleading is the way forecasts are often presented. I get annoyed when presenters use the word "will" (as in "Tomorrow a belt of rain will pass over the country in late morning" or "The temperature will reach 15 degrees in London today") which implies complete certainty about these forecasts. They may find it tiresome to say "we expect" . . . but then I suppose I'm a pedant.
Edited by tyro on 04/12/2008 at 15:09
|
It was certainly raining very heavily here on the Cheshire Plain at 5AM - exactly as predicted by the weather radar that I watch.
However it was warmer than the Met Office predicted. A couple of degrees colder and we would have had quite a depth of snow.
See: www.meteox.co.uk/h.aspx?r=&soort=loop24uur&URL
Edited by Bill Payer on 04/12/2008 at 15:30
|
If only we realised weather forecasting is based on a model of the weather system using differential equations. With global warming the model might need amending.
I am sure they are not getting it wrong on purpose. It's weather forecasting based on simulation - they do not know the future.
Yes they may have got it wrong for some of us, but not travelling and finding them wrong is better than travelling and finding they were right...
Edited by rtj70 on 04/12/2008 at 17:53
|
But how much has getting it wrong cost British Business's today ?
Police warnings not to travel, people staying away from business and commerce , we're in the middle of a damaging recession for crying out loud. A High Street devoid of shoppers today, devoid because they took on the false warnings and stayed at home. For many shops, a days lost trading could be enough to swing the balance between keeping going and closing down.
Edited by Mr X on 04/12/2008 at 17:58
|
Well they're damned either way as forecasters. If they'd said light rain and 4 degrees and the public believed it and turned out to be black ice and accidents...
Weather forecasting is based on a simulation of what they believe will happen. There is a mass of data for points around the world and it's run through a simulation of the weather (based on differential equations) and they try to guess what it will be like later. It's not guaranteed.
I wonder if any of us could do better? The Met Office I believe are getting a new computer system by 2011 from IBM for £33m. It will be 30 times more powerful. But never powerful enough. Ever heard of the butterfly affect?
|
I wonder if any of us could do better?
You only have to think of the Formula 1 weather forecasters who are forecasting the weather minute-by-minute, using local radar as well as helicopters flying near the approaching rain clouds.
Yet, in 2008, on many occasions, many teams got their forecasts wrong.
Just one example: This was demonstrated spectacularly in Brazil when just as the formation lap was about to set off, there was a short sharp downpour which resulted in panic and the race start being delayed by 10-15 minutes.
|
|
But the F1 forecast is using local info. The Met Office have to crunch data for the world. For the local forecast it will be run to be more detailed with a smaller grid but it's still a lot of number crunching going. At the end of the day its all solving differential equations. We could do it manually but not quick enough.
|
But how much has getting it wrong cost British Business's today ?
There you go again.
The forecast was right.
The people who misunderstood it were the ones who got it wrong.
|
>>But how much has getting it wrong cost British Business's today ?
Considerably less than getting it wrong the other way.
The met office website, yesterday, was suggesting 60% chance of disruptive heavy snow north of about Luton.
|
.. 60% chance of disruptive heavy snow north of about Luton
And yet, as Channel 4news has just headlined, motorists have been "stranded" in heavy snow. Full report later in the news.
www.metoffice.gov.uk/science/creating/daysahead/en.../
Edited by jbif on 04/12/2008 at 19:08
|
But how much has getting it wrong cost British Business's today ?
I think for the good of the nation and our hard-pressed retailers Tellytubbies should replace all weather forecasts.
|
A microlight pilot friend pays no attention to weather forecasts more than 24 hours in advance, and even then reckons 12 hours is about the limit if you want to be fairly certain of avoiding a nasty surprise.
If someone whose life could depend on the accuracy of the forecast is only willing to trust it that far ahead, it speaks volumes as far as I'm concerned.
Computer modelling for weather and climate (including long term climate trends) is still based on guesswork and assumption. The output from the former is not accurate, and we'll all find out in the coming years if the output from the latter is.
Cheers
DP
|
Look you only need to see what happens at the British grand Prix nearly every year.
The teams send up helicopters, post lookout men a few miles away, all to try and predict within the next 15 minutes what's going to come over the circuit.
95% of the time they get it wrong. And this from guys who can SEE what's happening.
|
The teams send up helicopters, ....
exactly, see my post above - Thu 4 Dec 08 20:50
|
A friend of mine is a senior forecaster at the MET. If you ring her up, tell her where you're going to be, how high, and when, then you can get truly excellent forecasts.
The poor forecasts are around for a number of reasons, as I understand it.
First, there are days when they can tell you with virtual certainty what will happen and when, but there are other days when they really don't know. The media never tell you which is which.
Plus, local variations can be really quite large, and can introduce huge variances to the 'general' picture.
Finally, the TV forecasts tend to me massive generalisations. A lot of the non-MET forecasts are based on the met computer data, but before it's been analysed by the properly trained and experienced types there. The computer model is only really a starting point by the sounds of it, and sometimes is vetoed completely by the duty forecaster. In any case, any critical forecast is updated every 30 mins as conditions change.
Im summary, the weather is so complex it's almost impossible to forecast 100% right. Normally they can tell you what will happen, but the when is a lot more fluid.
|
"The poor forecasts are around for a number of reasons,"
One of which is the fact that the BBC don't want another "Don't worry, there isn't a hurricane on the way" and therefore exaggerate every forecast. (And have done since 1987!!)As said in another thread - Metcheck is far more accurate, as are other forecasters.
I'm also interested in how these people can forecast the exact temperature rise over the next 100 years and exactly how much is due to our car exhausts and then base a tax regime upon these forecasts.
Phil
|
|
It's going to be unseasonably sunny and warm next week - 60 degrees. There, feel better already.
|
"60 degrees"
Oh, come on, you're as bad as the BBC - is that Fahrenheit or Centigrade (Celsius???)
:-)
|
PhilW. What makes you think that Met Office staff dealing with forecasts, are expert in climate change? Do chemists only deal in organic chemistry, engineers deal only in civil engineering? If you believe that scientific fields are much of a muchness, possibly. Here's a snippet: tinyurl.com/5hzyfv
Note how much the historical record has to be considered. This is not the field of the forecasters, SFAIK.
Edited by nortones2 on 05/12/2008 at 19:48
|
|
Phil - I take it you believe me if you have to ask!!!
|
"Phil - I take it you believe me if you have to ask!!!"
Course I do - as much as the BBC!! Guess I'd better cast of that extra vest, scarf and gloves next week! Oh, and the thermal lined trousers which have been a Godsend this week (esp at 3 am!!)
Phil
|
Terribly sorry Nortones, but since the Met office bungs out a great deal of information on climate change and weather forecasts (on their web page the index is firstly "Weather" secondly, "Climate change") I assumed that there might be a link.
See
www.metoffice.gov.uk/weather/uk/index.html
A click here might suggest that they think they are experts
www.metoffice.gov.uk/climatechange/
but then again, you obviously know that this should be disregarded since this is "not the field of the forecasters"
Damn, should have known that this section of the Met Office site is to be completely disregarded - oh, but hang on, perhaps I should disregard their forecasts since they might be experts in Climate Change rather than forecasts???
Talk about confused, maybe I should just ignore, everything they say??
Phil
|
|
I did realise that the Met Office encompasses aspects of both. But the role is significantly different, as are the challenges. Weather is not the same as climate. The Met Office remit obviously includes climate change, but the task of considering climate change, and analysing in hindsight, over the long run is very different to forecasting small, day to day, changes in the weather.
|
Years ago the weather presenter came on, and said it would either rain, or it would be sunny. Sometimes they were right, sometimes they were wrong. No one really took any notice of it. You looked out the window and dressed accordingly. Anyone who needs to rely on weather, especially by air or by sea already have their own advanced ways and forecasts.
Now it's nearly 2009, we have more technology than ever, more weather stations, more forecasting supercomputers and now the weather presenter comes on and says "there's an 80% chance of rain!" Well either the rain is coming this way or it isn't. So I look out the window and dress accordingly.
Checking the BBC online weather 5 Day Forecast on Wednesday evening, the forecast said "Snow" for Thursday. It said there was a more detailed 24 hour forecast available. So I clicked on that to see exactly when the snow would arrive. Instead it showed Thursday to be full of "Sunny Intervals" and "Partly Cloudy" - no mention at all of snow.
Hmmm.
If I get ice on the car then I consider it cold.
|
"but the task of considering climate change, and analysing in hindsight, over the long run is very different to forecasting small, day to day, changes in the weather. "
Agreed, so tell me, where do we get this "considering climate change, and analysing in hindsight, over the long run" from to be able to predict the exact change in temps over the next 100 years - where does the hindsight come from?
Forecasting seem to involve looking at radar shots/pressure charts/satellite photos of approaching weather systems etc for the next few hours based on 350 year records (in UK)of what has happened before yet that is often wrong. So how come they can look at similar records and yet "be right" within a degree or two for the next hundred years (or more).
Perhaps it depends on what someone has programmed into a computer - do you trust him?? - Look up Steve Hansen (IPCC guru) and his predictions.
I gather the ski resorts are open a couple of months early this year - must be all the warming.
Regards
Phil
|
I gather the ski resorts are open a couple of months early this year - must be all the warming.
That sentence says it all.
I know statistical analysis is a difficult subject on its own before you mix climate science in to it.
It is despairing when people cannot even understand the simple arithmetical difference in the concept of margin vs mark-up. This was demonstrated clearly last week by the large number of people who thought that a reduction in VAT from 17.5% to 15% meant that the previous VAT inclusive would be reduced by 2.5%.
What chance then that you can make them understand the basis of climate change theories or the probabilities of harm from siting a nuclear power station in their backyard?
|
I am suitably rebuffed - after all it was only a year ago that the press and the global warmists were publishing pictures of St Anton in Austria being bare of snow as proof of "global warming" - where are the pictures of it this year??.
It's a "science" based on emotion and sensationalism not statistics; a picture or two of polar bears sitting "stranded" on a melting ice floe - taken as a tourist picture. It's only just over a year ago since Radio 5 sent Julian Warricker (sp??) to Greenland to show how the Arctic Ice was "disappearing" - well guess why they sent him in October - yes, that is the end of the "summer melt". Lets not send him in Winter - it's all frozen. Why have global temps been declining since 1988? Why do they tell us (and this is the basis of all the global warming debate) that temps have risen by 2 deg since 1880 when the global average temp (a meaningless statistic in itself) was measured as 14deg +/- 2deg - in other words temp has increased by 2 deg but we don't know what the temp was within 2 deg -oh, and by the way, who took the comparative temps in 1880 at the South and North Poles?. The statistics about global warming are very debatable. See all the "statistics" Steve Hansen has spouted over the last few years.
Anyway, back to "Misleading weather forecasts" - please explain why 24 hour weather forecasts can be so inaccurate yet we are apparently going to spend billions and billions of pounds on trying to prevent a rise in temp over the next 100 years or so and are even now taxing cars/ fuel etc on the basis of of this temp prediction.
Will it rain tomorrow? will the temps be well below freezing today? (no - unfortunately our prediction was wrong by 5 deg and it was dry and as for the predicted snow showers, no chance) What temp will it be in 100 years time? Definitely 2 degrees warmer than "today" and on that basis we are going to stick some more tax on your petrol and and heating bills.
By the way, I get the VAT thing (though the Costa coffee place today doesn't - but at least my coffee was a bit cheaper). Did Gordon and Alastair Dearest get it? Don't think so.
"harm from siting a nuclear power station in their backyard" - roughly the same as harm from bird flu, CJD, hole in ozone layer, lead in petrol, weapons of mass destruction, drinking alcohol while pregnant, being a totally irresponsible banker/motgager/mortgagee/.
And what do you suggest instead of nuclear power - coal? gas? oil, wind? (except when not windy), tidal (please switch off all electrical stuff at high tide and low tide for a few hours)
|
Mr X said:
>And other countries around the world are already using superior equipment to predict their
>weather, meaning the UK is now lagging behind countries like France
This is a bit misleading. Supercomputer replacement/upgrade is typically every two or three years for these organisations so it's perfectly reasonable that country A's recently commissioned system is more powerful than country B's two year old system.
They also conveniently ignore that the world leader for global forecasting is actually in the UK.
Gordon said:
>Im summary, the weather is so complex it's almost impossible to forecast 100% right.
Absolutely.
Accurate forecasting is incredibly complicated and is far from guesswork as some people believe. There is a huge number of variables to take into account and the slightest measurement error in a single variable can have a very significant impact on forecast accuracy. Satellite data is improving the quality and quantity of input measurements but the CPU power needed to process the additional measurements is increasing exponentially.
A beginner's guide to the problems of weather forecasting: tinyurl.com/5ws5zv
Kevin...
|
And now back to motoring.
DD
|
|
