Early winter hopes and chat

[BornFromTheVoid]

I think what some people question is if you are going to say you are uncertain in every forecast then why bother doing them at all.

 

Would the resources not be better directed to improving forecasts in the mid range 5 - 8 day period? We have more chance of putting a man on Mars than we do forecasting the weather to any great degree of accuracy beyond 15 - 20 days.

 

They're not always so uncertain, but I'd agree to giving the 5-8 day range greater focus, but then, who says they're not already focusing on that area? My main gripe was the comparison of expert scientific analysis with made-up tabloid sensationalism, and the idea that admitting things are uncertain somehow shows a lack of ability or expertise.

[BornFromTheVoid]

That's a fair point BFTV, but we all know LRF can never be accurate per say so they could add that caveat at the end of any LRF. I just want them to produce LRF, I for one wouldn't knock them if they got wrong.

 

I think the main issue with the MO producing seasonal forecasts for the general public, is that the probability method they use just isn't much use (though most LRFs aren't much use in general!). Lets say, for example, that for 4 different seasons in a row the MO predict a 50% chance of above average temps, a 25% chance of average temps and a 25% chance of below average temps. Most people will see this as being a forecast for above average temps in each season, because the greatest likelihood in each forecast was for above average temps. Therefor, if only 2 of the 4 seasons end up above average, to the general public, that's a 50% failure rate and really not much better than chance.

The reality is, that with a 50% chance of above average temps over 4 seasons, one would expect only 2 of those seasons to be above average, with 1 (25%) season average and 1 (25%) season below average. That would be an entirely accurate prediction... but not to the public. To most people, if the most likely temperature range with the highest probability doesn't occur each and every time, then they consider the forecast wrong. Hence, it's just not of much use to them to predict seasons for the public.

Making more general forecasts, such as predicting a mild winter based on a 50% probability of above average temps, isn't going to be much use. When the probabilities increase, to around 80 or 90%, then I think it would be time to start issuing seasonal forecasts for the general public. Until then, probabilities will have to do, even if it's not what most people want!

[Hocus Pocus]

I think the main issue with the MO producing seasonal forecasts for the general public, is that the probability method they use just isn't much use (though most LRFs aren't much use in general!). Lets say, for example, that for 4 different seasons in a row the MO predict a 50% chance of above average temps, a 25% chance of average temps and a 25% chance of below average temps. Most people will see this as being a forecast for above average temps in each season, because the greatest likelihood in each forecast was for above average temps. Therefor, if only 2 of the 4 seasons end up above average, to the general public, that's a 50% failure rate and really not much better than chance.

The reality is, that with a 50% chance of above average temps over 4 seasons, one would expect only 2 of those seasons to be above average, with 1 (25%) season average and 1 (25%) season below average. That would be an entirely accurate prediction... but not to the public. To most people, if the most likely temperature range with the highest probability doesn't occur each and every time, then they consider the forecast wrong. Hence, it's just not of much use to them to predict seasons for the public.

Making more general forecasts, such as predicting a mild winter based on a 50% probability of above average temps, isn't going to be much use. When the probabilities increase, to around 80 or 90%, then I think it would be time to start issuing seasonal forecasts for the general public. Until then, probabilities will have to do, even if it's not what most people want!

Thanks for the reply, it makes perfect sense in what you saying and I agree that until such a time accuracy in LFR improves ( if ever ) then there's no point issuing them, more so with the public flogging they receive when they get it wrong occasionally.

[johnholmes]

I can see what both the above posters are suggesting and tend to agree with both. You might find that odd being an ex senior forecaster with them. The atmosphere behaves according to the laws of thermodynamics. However the further one gets from T+00 then the more inaccuracies creep in initially becoming over riding the further out one goes. Models get totally confused by background signals that have nothing to do with the laws of thermodynamics and attempts to wipe this background mush out have so far not been successful. To predict for the UK 24 hours ahead is relatively (note the word!) straightforward with professional training and a decent understanding of maths and physics. No human brain can cope with anything further out than perhaps 3 days, the complications are just way to complex

. The arrival of computers was a godsend and we quickly made a lot of progress in improving 24 hour accuracy and more slowly as computers increased in speed and ability to deal with the immense complexity of the needs to predict further ahead. Beyond 3-5 days then even with the latest computer for the main Met centres and closer grid lengths the accuracy tails off. Indeed even now the accuracy of getting the moisture profile correct at relatively close time scales seems to defeat most models beyond 24 sometimes 48 hours. That is other than a general picture of where the heaviest may fall and when, and even that is at times not as good as most professional meteorologists would hope for. It has always, be it a human using their knowledge and hand drawn charts, or a model, easier to get the upper air pattern more correct than the surface be it 12 hours ahead or 168 hours (for example). I suspect it will remain so for some time to come no matter how small the grid lengths can become, and that on either global or much more local area models.

 When we turn to trying to predict out to a month let alone 3 months ahead then obviously the problem magnifies by a large factor even to get the approximate type of weather pattern. I wrote the other day how in the mid 70's in total honesty and scientific belief the Met Research folk felt that the answer to the 'new' month long prediction was all down to the jet stream. Look how far and much more complex our knowledge has become in the past 40 years. It is a dilemma for Met what to do.

 The method outlined by BFTV is probably fine for the recipients wanting to know several months in advance what type of winter in terms of overall temperature or precipitation require. The Office is then stuck with folk publishing what they are privy too within their work place and parliament insisting that this data is made available to the public. Those responsible for making public outputs are very probably aware of the pitfalls and as scientists are also not very good at pr as we have seen on more than one occasion from Met. So between the devil and the deep blue sea they sit, wrong whatever they do I think. As to when will the accuracy improve-who knows?

The depth of understanding is growing at quite a sharp rate but the ability of how to programme any computer is not rising at the same level in my view. I do understand the absolutely immense complexity of predicting the weather 24 hours ahead and the idea of being able to do this, however much broad brush for 3 months ahead leaves me, one could say breathless. I am too far removed from the operational side of forecasting to be able to understand how 'they' may do this. Again I have to say it will be man decades in my view before, the broad brush idea of colder/.wetter/drier/warmer 3 months in advance is anywhere near as accurate as the much maligned 16-30 Met outlook.

Sorry it’s so long but as some on here know I am passionate about the weather, sometimes quite critical of my ex colleagues but totally fascinated by the subject.

[Winter Cold]

not a good update from fergieweather on the model thread. MOGREPS is showing signs of high pressure to the south.

 

I dont want to use the B word, as im not sure thats what hes suggesting, but it sounds like a winter weather killer if it came? Mild S/SE winds

 

Im bored to tears now with SW mildness, its been there all year mostly.

 

Please please lets get a pattern change soon which is favourable for cold!!

[Festivefreeze]

According to today's Daily Star, we can expect a repeat of last year's winter, with wind and rain likely to be the main feature of the weather for weeks to come. I personally hope that this is wrong, because I love cold and snowy weather, and would love a repeat of December 2010. However, I have a horrible feeling that if the USA gets record breaking cold again this year, it is going to put the UK in the firing line for relentless Atlantic depressions. Even if we are lucky enough to end up with an high pressure-dominated winter, it could turn out to be a very mild one, but at least it would give Britain a change to dry out.

Edited November 14, 2014 by November2005Fan

[starstream]

Well of course the daily star is a scientific font of all knowledge!

So it must be true !

Good grief!

May I suggest, you spend some time reading the very well informed posters on here,and evaluating the prospects of whatever this winter may throw at us from the guidance they offer.

It's been said many many times....headlines of extreme weather,sell papers....even the Daily Star.

Edited November 14, 2014 by starstream

[ColdFront80]

It is quite funny seeing these cold dampers crying like babies because its not showing cold in mid November. Perhaps they should stop thinking you can predict snow 10 days ahead and just enjoy the ride.

Very interesting few months many signs of cold are there but as ever the weather will do what it wants cold or warm I'm looking forward to this winter.

[Hocus Pocus]

I can see what both the above posters are suggesting and tend to agree with both. You might find that odd being an ex senior forecaster with them. The atmosphere behaves according to the laws of thermodynamics. However the further one gets from T+00 then the more inaccuracies creep in initially becoming over riding the further out one goes. Models get totally confused by background signals that have nothing to do with the laws of thermodynamics and attempts to wipe this background mush out have so far not been successful. To predict for the UK 24 hours ahead is relatively (note the word!) straightforward with professional training and a decent understanding of maths and physics. No human brain can cope with anything further out than perhaps 3 days, the complications are just way to complex

. The arrival of computers was a godsend and we quickly made a lot of progress in improving 24 hour accuracy and more slowly as computers increased in speed and ability to deal with the immense complexity of the needs to predict further ahead. Beyond 3-5 days then even with the latest computer for the main Met centres and closer grid lengths the accuracy tails off. Indeed even now the accuracy of getting the moisture profile correct at relatively close time scales seems to defeat most models beyond 24 sometimes 48 hours. That is other than a general picture of where the heaviest may fall and when, and even that is at times not as good as most professional meteorologists would hope for. It has always, be it a human using their knowledge and hand drawn charts, or a model, easier to get the upper air pattern more correct than the surface be it 12 hours ahead or 168 hours (for example). I suspect it will remain so for some time to come no matter how small the grid lengths can become, and that on either global or much more local area models.

 When we turn to trying to predict out to a month let alone 3 months ahead then obviously the problem magnifies by a large factor even to get the approximate type of weather pattern. I wrote the other day how in the mid 70's in total honesty and scientific belief the Met Research folk felt that the answer to the 'new' month long prediction was all down to the jet stream. Look how far and much more complex our knowledge has become in the past 40 years. It is a dilemma for Met what to do.

 The method outlined by BFTV is probably fine for the recipients wanting to know several months in advance what type of winter in terms of overall temperature or precipitation require. The Office is then stuck with folk publishing what they are privy too within their work place and parliament insisting that this data is made available to the public. Those responsible for making public outputs are very probably aware of the pitfalls and as scientists are also not very good at pr as we have seen on more than one occasion from Met. So between the devil and the deep blue sea they sit, wrong whatever they do I think. As to when will the accuracy improve-who knows?

The depth of understanding is growing at quite a sharp rate but the ability of how to programme any computer is not rising at the same level in my view. I do understand the absolutely immense complexity of predicting the weather 24 hours ahead and the idea of being able to do this, however much broad brush for 3 months ahead leaves me, one could say breathless. I am too far removed from the operational side of forecasting to be able to understand how 'they' may do this. Again I have to say it will be man decades in my view before, the broad brush idea of colder/.wetter/drier/warmer 3 months in advance is anywhere near as accurate as the much maligned 16-30 Met outlook.

Sorry it’s so long but as some on here know I am passionate about the weather, sometimes quite critical of my ex colleagues but totally fascinated by the subject.

Excellent post and thank you for taking the time out to respond in such a professional and dignified manner John.

[Festivefreeze]

I'm on the autistic spectrum. Please don't be too hard on me. I never said that I believe the Daily Star's headlines. Nevertheless, I find it an interesting read. Some newspaper has to be right surely...?

[Bobby]

I'm on the autistic spectrum. Please don't be too hard on me. I never said that I believe the Daily Star's headlines. Nevertheless, I find it an interesting read. Some newspaper has to be right surely...?

 

It's best to avoid any newspaper weather predictions, they're in the business of selling headlines after all. I'd also be wary of these "independent forecasters" such as James Madden etc. Hell, I'd just not pay much attention to any long range forecast to be honest, even the best are just stabs in the dark and are only of curiosity value in my opinion.

 

If looking for an idea about the weather longer term you can't go wrong with the Met Office further outlook (6-30 day forecast) for reliability and accuracy. And of course following the models and seeing what they're up to for yourself can be fun too..

[Festivefreeze]

I've decided to leave the site, because I know nothing about the weather...bye

[Barry12]

I've decided to leave the site, because I know nothing about the weather...bye

 

If you know nothing about the weather, surely this is the best place to be as you'll learn?

[Festivefreeze]

All I know is that Britain has a temperate climate, which means that it is harsh from extremes. Britain's summers aren't as cold as other places at its latitude due to being the influence of the Atlantic Ocean. This also means that summers are not as warm as those on the continent? Am I right?

[Bobby]

All I know is that Britain has a temperate climate, which means that it is harsh from extremes. Britain's summers aren't as cold as other places at its latitude due to being the influence of the Atlantic Ocean. This also means that summers are not as warm as those on the continent? Am I right?

 

Yes, several thousand miles of ocean to our west with the prevailing 'default' wind direction also being W/SW (driven by low pressure to our N and high pressure to our S). This water is also unusually warm due to the Gulf Stream.. so another thing that counts against us in winter. The ocean has a much higher heat capacity than land meaning it warms very slowly and cools very slowly - which moderates any temperature extremes in areas affected by this oceanic air mass. Less extreme cold in winter, less extreme heat in summer.

 

Also unlike the USA we don't have a landmass to our N in which to import frigid polar air... any cold air that comes down from the pole is again moderated by the Atlantic by the time it gets to us (although we can get very cold air from the N of course but nothing like the USA).

 

So to get extremes of cold, heat we usually have to get our weather from the continental landmass to our E where such extremes occur. For this we usually need a blocking anticyclone to our north somewhere to reverse the normal westerly flow (because air travels clockwise around an anti-cyclone, anti-clockwise around a cyclone, so high pressure to our north will bring air from the east.)

 

More info on our climate type http://en.wikipedia.org/wiki/Oceanic_climate

Edited November 15, 2014 by Bobby

[Mark wheeler]

Not sure where to ask this question so here goes . I often read the model thread and understand about 40% what's going on and understand the posts. But when I'm not sure and need to understand what's happening I'm so reluctant to ask. This is for several reasons. A. Cluttering the thread. B.other threads that should help I can't find a example at that time.(difficult to explain).C. The very knowledgable personal on here are in these threads and seems not in learning areas as they are occupied by the here and now (models).

Is there a forum where there is coaching/commentary on current output. With future possibilities explained and previous output also reflected.

I have been an avid follower of this site for 4 years and have learnt plenty just trying to grasp and expand my knowledge which has been a obsession since i was a child.

Here is to a lovely and snowy winter. PLEASE

[johnholmes]

Not sure where to ask this question so here goes . I often read the model thread and understand about 40% what's going on and understand the posts. But when I'm not sure and need to understand what's happening I'm so reluctant to ask. This is for several reasons. A. Cluttering the thread. B.other threads that should help I can't find a example at that time.(difficult to explain).C. The very knowledgable personal on here are in these threads and seems not in learning areas as they are occupied by the here and now (models).

Is there a forum where there is coaching/commentary on current output. With future possibilities explained and previous output also reflected.

I have been an avid follower of this site for 4 years and have learnt plenty just trying to grasp and expand my knowledge which has been a obsession since i was a child.

Here is to a lovely and snowy winter. PLEASE

hi Mark

Please do not be afraid to ask questions. That is how we all advance our understanding of the fascinating subject the weather. have you had a look at the Net Wx Guides. Lots of information for a beginner in there. Not sure then please pm me and perhaps we can chat via pm rather than the thread if you like.

[Festivefreeze]

I'm still an amateur here, so I was wondering if anyone can help explain what caused the cold second half to November in 2005? I was 17 at the time and I remember some very thick fogs forming and often lingering through much of the day. Was there a rise in pressure over Greenland back then or something? I'd be happy if anyone could message me and try to explain this. Thanks

[johnholmes]

I'm still an amateur here, so I was wondering if anyone can help explain what caused the cold second half to November in 2005? I was 17 at the time and I remember some very thick fogs forming and often lingering through much of the day. Was there a rise in pressure over Greenland back then or something? I'd be happy if anyone could message me and try to explain this. Thanks

 

have a look at the charts for that month, they are available on wetter or within Net Wx. That will show you what the surface pattern was over most of the north Atlantic/Europe region.

[BLAST FROM THE PAST]

I'm still an amateur here, so I was wondering if anyone can help explain what caused the cold second half to November in 2005? I was 17 at the time and I remember some very thick fogs forming and often lingering through much of the day. Was there a rise in pressure over Greenland back then or something? I'd be happy if anyone could message me and try to explain this. Thanks

We had high pressure over the UK keeping it calm and with clear nights it would be chilly with fog forming.  This High Pressure backed west to the Atlantic as a low pressure system move east to our north.  A Greenland High pressure didn't form, what happened was the high pressure in the Atlantic ridged north towards Greenland and the low pressure moved SSE into the north sea and brought us an Arctic northerly.

If you go into charts homepage on this site and bottom left of that page you will see archived charts.  You can then call up that period and view how it happened.

here we go, this is 25th Nov

 

 

HP ridged to Greenland BUT NOT A true Greenland HP, and low pressure to our east in the North Sea.  Follow isobars in north sea northwards....air coming down from the arctic

 

 

 

BFTP

Edited November 15, 2014 by BLAST FROM THE PAST

[Cloud 10]

The end of November 2005 is a good example of how quickly things can change,as shown in the 3 Northern hemisphere charts below:

 

20th.. 23rd.. 26th..

[Bobby]

Here's a good resource to flick through with information on the weather, models and how to understand them.. and lots more

 

http://weatherfaqs.org.uk/

http://weatherfaqs.org.uk/node/15

http://weatherfaqs.org.uk/node/214

http://weatherfaqs.org.uk/node/76

http://weatherfaqs.org.uk/node/140

 

This another great resource, you could spend weeks reading through all the goodies here.

 

http://www.theweatherprediction.com/

Edited November 15, 2014 by Bobby

[summer blizzard]

https://forum.netweather.tv/topic/80936-autumn-2014-forecasts-thoughts-predictions/page-15

 

Made a post here about that month as its the November i prefer the most. 

 

Stupidly sunny in the end here as well.

[johnholmes]

Here's a good resource to flick through with information on the weather, models and how to understand them.. and lots more

 

http://weatherfaqs.org.uk/

http://weatherfaqs.org.uk/node/15

http://weatherfaqs.org.uk/node/214

http://weatherfaqs.org.uk/node/76

http://weatherfaqs.org.uk/node/140

 

This another great resource, you could spend weeks reading through all the goodies here.

 

http://www.theweatherprediction.com/

 

 

yes I would recommend both sites, our own Net Wx Guides are well worth a look at as well, the main link is

https://forum.netweather.tv/forum/5-the-netweather-guides/

[Winter Cold]

Thanks for those links, great sites, been reading them a little bit this morning already

 

Thats a good post from Cloud10 above about Nov 05, just shows you that it can all fall into place quickly.

 

Id like to think that with the state of play this year, we would be very unlucky not to have seen a snowflake by the time the year is out. Fingers crossed!