Autumn And Winter Lrfs

[johnholmes]

Paul you seem to be falling into the same error of perception of what they actually said in September. yes there forecast was not spot on but tell me a better main player forecast?

Comments about

' error spread (two standard deviations, typically) did not actually include the outcome, therefore, technically, it wasn't a great forecast.'

sound good but do you not think they are not aware of this.

I do despair of some people ever accepting a pretty accurate forecast for what it is. They gave you the odds as a betting man, 2 in 3 chance of it being correct, explained how they arrived at it.

Care to do better this year? Whether they will be correct in September we will all have to wait until the end of February to judge.

John

Edited July 17, 2006 by johnholmes

[Damien]

The winter was actually milder than the Met Office NAO forecast predicted! Some of the coldest and snowiest weather occurred last Autumn, in November, which is not recorded as winter and this does seem to have clouded memories on the severity of the winter.

I rest my case.

With one light snow shower at the end of December and another in mid-February I want my winter back!

That said March had some *quite* snowy spells though, but they were still very unremarkable by even recent standards - winter 2000/01 and late February 2005.

[TEITS]

It wasn't bad; certainly the headline forecast looked good, but the error spread (two standard deviations, typically) did not actually include the outcome, therefore, technically, it wasn't a great forecast. The winter was actually milder than the Met Office NAO forecast predicted! Some of the coldest and snowiest weather occurred last Autumn, in November, which is not recorded as winter and this does seem to have clouded memories on the severity of the winter.

Paul

Depends where you live because for my region the coldest & snowiest weather was just after Xmas when I max temp of only -4C was reached.

[West is Best]

I rest my case.

What, because someone on net-weather agrees with you? Is that forensically dependable?

With one light snow shower at the end of December and another in mid-February I want my winter back!

That said March had some *quite* snowy spells though, but they were still very unremarkable by even recent standards - winter 2000/01 and late February 2005.

So the Met O forecast was about snowfall was it? No, it wasn't. They actually forecast this:

"The Met Office continues to predict a two in three chance of a colder-than-average winter for much of Europe, and that if Europe were to experience below-average temperatures, parts of the UK - especially southern regions - would also be affected. There is also an indication for a drier-than-average winter over much of the UK."

Drier than average. Not snowy. They were spot on with their forecast and only the most churlish person would deny them their rightful moment of glory.

[johnholmes]

not often I am in deep agreement WIB but printing what they actually said 3 months in advance of the event starting does show that it was extremely near to be spot on, as they admit themselves, I think, 3/3/06, it was less cold than predicted away from the south, Europe was spot on, it was still correct in my view to predict months in advance that we were going to experience the coldest winter for some years, and also a dry one. Nowhere did they predict snow. Quite why Damien carries on as he does I have no idea. I am a little surprised at Paul(Dawlish) encouraging him as an academic but perhaps he is not yet aware of how Damien posts.

John

Edited July 17, 2006 by johnholmes

[SMU]

What, because someone on net-weather agrees with you? Is that forensically dependable?

So the Met O forecast was about snowfall was it? No, it wasn't. They actually forecast this:

"The Met Office continues to predict a two in three chance of a colder-than-average winter for much of Europe, and that if Europe were to experience below-average temperatures, parts of the UK - especially southern regions - would also be affected. There is also an indication for a drier-than-average winter over much of the UK."

Drier than average. Not snowy. They were spot on with their forecast and only the most churlish person would deny them their rightful moment of glory.

I do agree that the METO forecast was good & at a level managable with the current knowledge in the field of LRF's- in terms of not over diluting the forecast- but keeping it simple with 2 parameters-

What does stick in my mind to enhance their accuracy was the European forecast in terms of it being well below ave- this turned out equally accurate-

The NAO output was however called incorrectly- the Call for negative was wrong, it was a Positive Winter-

Interestingly as I mentioned earlier If they are basing this Winter on the NEwfoundland SSTA's then the forecast will most definatly be for a POS NAO winter-

Whether they will elude to any other estimates based on this one measure is very unlikely-

but this should give the Meto Winter seasonal forecast a steer to mild -

Whether we agree especially with the alleged changing of patterns ever lurking in the background is another matter-

Good discussion points though-

S

[johnholmes]

hi

This will go round and round, some agree with the Met O forecast some don't. I'm getting of the round a bout as there is a limit to repeating ourselves.

I will just finish with this below which is the most data I can find from the Met Office site and gives a pretty fair coverage of what they said initially and how near or far they were by their own comments in March this year.

No apologies for it being long.

This forecast is produced using a combination of statistical models, such as the North Atlantic Oscillation (NAO), and complex climate models with interpretation by operational forecasters.

Our latest predictions indicate a colder than average winter for much of Europe. If this holds true, parts of the UK — especially southern regions — are expected to have temperatures below normal.

The last eight winters have been relatively mild and perhaps have given the impression that these are 'normal'. The balance of probability is for a winter colder than those experienced since 1995/6.

Confidence in long-range predictions is relatively low compared to our normal short-period forecasts. Typically the broad features of the forecast for the European region are correct two out of every three occasions. Using a traffic light analogy we consider that people should be on amber alert for the prospect of a colder-than-average winter.

There is also an indication for a drier-than-average winter over much of the UK.

It is not possible to be specific about forecast temperatures or precipitation for winter 2005/6,

but the tables below show a comparison of recent winters and 1995/96 to the long-term average

(1971-2000).

as an ex Met man myself I am quite certain that publishing this in full will not bring any retribution on the head of Paul in Net Weather. I'm sure they will think it a good idea to have the full text printed.

I do not normally agree with highlighting but to try and show the most relevant parts, I have put them in red.

John

26 September 2005

A prolonged, severe winter is one of the biggest threats to the efficient day-to-day running of the country. With this in mind, the Met Office has given advanced warning to many of its customers and partners to plan for a 'colder-than-average winter'. Using a traffic light analogy, the organisation has written to contingency planners in the Government — including the NHS and Highways Agency — and in the energy industry, as well as many others to put them on Amber Alert.

Our daily lives are inextricably linked to the weather — from the ease of travelling to and from work, to the health of individuals with certain conditions or the efficiency and demands on the UK energy industry. It is critically important that planners are aware of long-term forecasts issued by the Met Office and are able to act on the information.

Since 1995-96, winters in the UK have been mild, giving many the impression that mild winters are now the norm. Even an average winter could come as a surprise to many, although at this stage, it is not possible to issue specific details about a particular day, or location, over the winter season.

The long-range forecast also signals a dry winter across much of the UK. A cautious approach needs to be taken at the moment, but, bearing in mind the lack of rain for the UK over the last 12 months, this factor may be more significant in the long run.

Winter forecast Dec-Feb 2005/6

This forecast is produced using a combination of statistical models, such as those used to predict the North Atlantic Oscillation (NAO), and complex climate models with additional interpretation by operational forecasters.

Forecast updated 24 Oct 2005

Our predictions continue to indicate a colder than average winter for much of Europe. If this holds true, parts of the UK — especially southern regions — are expected to have temperatures below normal.

There is also an indication for a drier-than-average winter over much of the UK.

The last eight winters have been relatively mild and perhaps have given the impression that these are 'normal'. The balance of probability is for a winter colder than those experienced since 1995/6.

Confidence in long-range predictions is relatively low compared to our normal short-period forecasts. Typically the broad features of the forecast for the European region are correct two out of every three occasions. Using a traffic light analogy we consider that people should remain on amber alert for the prospect of a colder-than-average winter.

There will be regional variations across the UK this coming winter. For example, historically, when southern and central England experience a harsh winter, it is quite common for parts of Scotland and Northern Ireland to have a mild one.

It is not possible to be specific about forecast temperatures or precipitation for winter 2005/6,

but the tables below show a comparison of recent winters and 1995/96 to the long-term average

(1971-2000).

Mean Temperature (degrees C)

Long-term average

2004/5

2003/4

1995/6

Scotland

2.7

4.0

3.2

1.8

Northern England

3.5

4.6

4.2

2.4

Southern England

4.5

5.2

5.3

3.5

Wales

4.2

5.2

4.8

2.8

Northern Ireland

4.3

5.4

4.8

3.7

Precipitation (millimetres)

Long-term average

2004/5

2003/4

1995/6

Scotland

464

543

484

276

Northern England

264

227

298

226

Southern England

220

130

229

290

Wales

447

335

474

326

Northern Ireland

326

303

280

287

The winter forecast will be updated at the end of November.

This is the full text of their comments in March about how they viewed their success or otherwise.

Winter weather summary

3 March 2006

In September 2005, the Met Office made predictions about the coming winter. The final figures are now in for the months of December 2005 through to February 2006 and it is now possible to assess the winter of 2005/6.

In summary; Europe has experienced below-average temperatures over a wide area through the winter 2005/6. The winter has been very dry across the whole of the UK, warmer-than-average in the north, colder-than-average in the south.

In the original forecast there were five main issues that it is now possible to verify. It can be said that in four out of these five cases, the predicted most likely event happened:

WINTER REVIEW

Prediction

Outcome

Mean temperature across Europe

Most likely colder than average

Colder than average for many areas

UK mean temperature

Most likely colder than average

Warmer than average

Southern UK mean temperature

Most likely colder than average

Colder than average

UK precipitation

Most likely drier than average

Drier than average

North Atlantic Oscillation (NAO) for the winter season

Negative

Negative

• Europe has experienced below-average temperatures over a wide area through the winter 2005/6, including southern parts of the UK. This is the outcome we predicted was most likely
  
• Northern parts of the UK did not follow the outcome we predicted as being most likely, and had a mild winter
  
• The UK as a whole experienced a slightly warmer-than-average winter, but which was still the coldest since 2000/1
  
• England has had a colder-than-average winter - the coldest since 1996/7
  
• Wales has had a colder-than-average winter - the coldest since 1996/7
  
• The coldest region relative to the long-term (30-year) average was southern England, at 0.4 °C below normal; the coldest since 1996/7
  
• Scotland and Northern Ireland had a mean temperature above normal (0.77 and 0.48 °C respectively), which continued a run of five mild winters for these parts of the UK
  
• The average of this winter's Central England Temperature (the longest existing temperature record in the world), was below the average for 1971-2000
  

Below is the link to moe info on max/min/mean etc.

http://www.met-office.gov.uk/climate/uk/20...inter/maps.html

sorry some of this has not come out how it copied but refer to the Met O site for the actual charts and text please.

John

[The PIT]

I'm stunned.

One very impressive forecast and some poeple are complaining. I hope you're not confusing it with the media hype.

Steve I thought NAO was close to neutral most of the time????

Thanks for doing the Homework John

Edited July 17, 2006 by The PIT

[johnholmes]

according to the slightly different method of assessing the NAO by the Met Office it was negative. Again it depends on what method anyone is actually referring to. So long as its made clear what method then its that one must compare accuracy with.

John

[Dawlish]

Paul you seem to be falling into the same error of perception of what they actually said in September. yes there forecast was not spot on but tell me a better main player forecast?

Comments about

' error spread (two standard deviations, typically) did not actually include the outcome, therefore, technically, it wasn't a great forecast.'

sound good but do you not think they are not aware of this.

I do despair of some people ever accepting a pretty accurate forecast for what it is. They gave you the odds as a betting man, 2 in 3 chance of it being correct, explained how they arrived at it.

Care to do better this year? Whether they will be correct in September we will all have to wait until the end of February to judge.

John

Easy John! I was just explaining the technicalities of their NAO forecast and the results! They gave me odds of 2 in 3 and technically got it wrong. In their hindcast, it will not add to their correct dataset and will actually slightly rduce the hindcast accuracy of their forecast.

The forecast was probably the best of the "major players", who use even iffier prediction techniuqes, in my opinion, as you rightly say, but that doesn't change my own view that this experimental forecast shouldn't be relied on. I can assure you I am not falling into any "error of perception" having written extensively to defend the Met Office last winter and recently on here.

I will happily give you a forecast for the coming winter. It is the same one that I've used for winters, summers and years since 1990:

A 70% chance of the winter being warmer than the long term Manley CET, a 20% chance of it being colder and a 10% chance of it being within 0.25C of the mean.

Last year, it fell into the last category, being very slightly milder than the long-term Manley sequence. This way of forecasting has a much higher accuracy that the Met Office's experimental one, over the last 16 years, but the Met Office don't use percentages and probabilities like that.

Paul

Depends where you live because for my region the coldest & snowiest weather was just after Xmas when I max temp of only -4C was reached.

Yes, of course it does, eye, I agree.

Paul

[Damien]

What, because someone on net-weather agrees with you? Is that forensically dependable?

Yes - because this is one of the "heavyweights" on the forum IMHO.

So the Met O forecast was about snowfall was it? No, it wasn't. They actually forecast this:

"The Met Office continues to predict a two in three chance of a colder-than-average winter for much of Europe, and that if Europe were to experience below-average temperatures, parts of the UK - especially southern regions - would also be affected. There is also an indication for a drier-than-average winter over much of the UK."

Drier than average. Not snowy. They were spot on with their forecast and only the most churlish person would deny them their rightful moment of glory.

Then so be it. But I am going by the "cold and dry" winters of the past - like 1995/96 and many of the 1980s winters - that brought my region of the world a fair bit of snowfall. Wasn't even February 2005 "colder and drier" than normal?, and that brought my area a fair bit of snow. (Compare with January 28(?), 2004, which didn't.)

Edited July 17, 2006 by Damien

[Bottesford]

This debate still going on?!

The Met Office forecast was spot on. Considering it was long range it spelt out the winter exactly how it turned out to be. Colder than the previous ten winters and dry. Which is exactly what it was.

Anyway back to summer talk...

[Dawlish]

I am a little surprised at Paul(Dawlish) encouraging him as an academic but perhaps he is not yet aware of how Damien posts.

John

errrrr?? I really didn't encourage Damien as an academic. Honestly! He may be an academic for all I know, but what he actually did was to use something of mine to rest a case on. It may have been a chair. You can't have read my replies to Damien very carefully!

Paul

Edited July 17, 2006 by Dawlish

[Damien]

What does stick in my mind to enhance their accuracy was the European forecast in terms of it being well below ave- this turned out equally accurate-

The NAO output was however called incorrectly- the Call for negative was wrong, it was a Positive Winter-

Interestingly as I mentioned earlier If they are basing this Winter on the NEwfoundland SSTA's then the forecast will most definatly be for a POS NAO winter-

Whether they will elude to any other estimates based on this one measure is very unlikely-

but this should give the Meto Winter seasonal forecast a steer to mild -

Whether we agree especially with the alleged changing of patterns ever lurking in the background is another matter-

Good discussion points though-

You're right on all three points, Steve. Yes, the MetO European forecast was very good. In fact, it was excellent, for yet another year. And yes, the winter NAO forecast for the UK failed because it was, as two experts have now said, a predominantly positive NAO winter, and there was very little "winter" in my area. Kent and the South-East got hit pretty hard I understand again though, as did some parts of the East Coast.

I think the last post by Dawlish summed it up just nicely.

Edited July 17, 2006 by Damien

[Dawlish]

You're right on all three points, Steve. Yes, the MetO European forecast was very good. In fact, it was excellent, for yet another year. And yes, the winter NAO forecast for the UK failed because it was, as two experts have now said, a predominantly positive NAO winter, and there was very little "winter" in my area. Kent and the South-East got hit pretty hard I understand again though, as did some parts of the East Coast.

I think the last post by Dawlish just summed it up quite nicely.

My last post talked about you resting your case on a chair. I don't get this any more. Help me out. What did I say that you agree with so much and somehow proves your point?

Paul, completely lost....

Edited July 17, 2006 by Dawlish

[Damien]

What did I say that you agree with so much and somehow proves your point?

You said the same as Steve. You were the two experts I indirectly identified earlier; and I agree with you both on the assessment of the NAO forecast for last winter.

That said John H. also makes a good point about the MetO concluding that it was a negative NAO winter, to whatever extent, using their methods of post-February 28 assessment.

The MetO's performance last year was fair but was a bitter failure for many areas of England, including mine. But maybe I am just judging it by my expectations of other "cold and dry" winters like 1995/96, as I have already said.

Anyhow, moving on and looking to the future, a weakly positive NAO winter 2006/07 (which 1990/91 was) with above average precipitation, from a major UK weather body like the Met Office, sets me looking with great interest to next season, as Bottesford said.

[SMU]

Just reading between the few posts here & picking out the messages about the NAO-

Im very interested to know where the Met office import the data from to get an Negative NAO'd Winter-

Clearly the NOAA (CPC) historic shows Positive for 05/06 Winter-

http://www.cpc.noaa.gov/products/precip/CW...ent.ascii.table

I guess the differences were the point of measurements.....

From my perspective whoever & whatever the measurement was there wasnt any pronounced negativity until March- this is also backed up by the NE US +ve temp anomaly & Philip Edens MSLP Anomalies for the same period-

Dec 05- http://www.climate-uk.com/monpre/0512.htm neutral ( anomalies wise )

Jan 06- http://www.climate-uk.com/monpre/0601.htm Must be 100% pos Nao..

So then just to Clarify-

Met office ACTUAL forecast- Good & well Called-

NAO forecast/Model - Wrong to call negative despite the Metoffice claiming it was negative-Never read that information before & I despute it-

Perhaps they use Tom osbourn whoever he is-

http://www.cru.uea.ac.uk/~timo/projpages/nao_update.htm

Still cant believe JAn was Negative- What a crock-

S

Edited July 17, 2006 by Steve Murr

[summer blizzard]

The NOAA give the foolowing NAO values for the winter...

December - -0.4

January - 1.2

February - -0.5

Unless they are saying that it was a negative NAO winter because the majority was, they should of said it was a weakly positive NAO winter.

[johnholmes]

hi Steve

I said I would make no further comment about winter 2005-6, did it did it not etc etc. However re your comments about the Met O and their NAO -ve comment

There was a full explanation of how they use these NAO -/+ indices but am unable to find it. All I have is this which suggests that they use 'their' version amended as seen fit by a human being!

This forecast is produced using a combination of statistical models, such as those used to predict the North Atlantic Oscillation (NAO), and complex climate models with additional interpretation by operational forecasters.

I can only suggest you send an e mail to them asking just what it was they used, I'm sure they will respond.

regards

John

[Dawlish]

You said the same as Steve. You were the two experts I indirectly identified earlier; and I agree with you both on the assessment of the NAO forecast for last winter.

That said John H. also makes a good point about the MetO concluding that it was a negative NAO winter, to whatever extent, using their methods of post-February 28 assessment.

The MetO's performance last year was fair but was a bitter failure for many areas of England, including mine. But maybe I am just judging it by my expectations of other "cold and dry" winters like 1995/96, as I have already said.

Anyhow, moving on and looking to the future, a weakly positive NAO winter 2006/07 (which 1990/91 was) with above average precipitation, from a major UK weather body like the Met Office, sets me looking with great interest to next season, as Bottesford said.

Expert?? Me?? You want to pick yourself a decent expert, if you want an expert and not one who sits in my chair. I would never, EVER claim to be an expert, in either field of meteorology, or climatology. I have no research qualifications and no professional experience. I am an amateur with a bit of knowledge. No more.

Paul

Edited July 17, 2006 by Dawlish

[SMU]

hi Steve

I said I would make no further comment about winter 2005-6, did it did it not etc etc. However re your comments about the Met O and their NAO -ve comment

There was a full explanation of how they use these NAO -/+ indices but am unable to find it. All I have is this which suggests that they use 'their' version amended as seen fit by a human being!

This forecast is produced using a combination of statistical models, such as those used to predict the North Atlantic Oscillation (NAO), and complex climate models with additional interpretation by operational forecasters.

I can only suggest you send an e mail to them asking just what it was they used, I'm sure they will respond.

regards

John

Hi John - The problem here is ALL entities are using different recording/validation methods-

This then becomes a pigs ear when trying to varify any one particular forecast-

Considering its the hardest Teleconnection to forecast then I think at least for now the MSLP anomalies rather than the NAO are the way forwards in giving us a tool to forecast at a seasonal level-

The weather which arrives via those pressure anomalies is 'generally' forecastable from that perspective- especially when referring to seasonal parameters such as temperature & rainfall-

S

[The Enforcer]

Im very interested to know where the Met office import the data from to get an Negative NAO'd Winter-

Clearly the NOAA (CPC) historic shows Positive for 05/06 Winter-

Well whether the NAO was positive or negative, the result was the same - cool, dry and nondescript. I doubt there is much difference between weak negative or weak positive anyway.

Surely the actual weather that is predicted is more important than the statistics? What would have been the point in everyone agreeing that NAO was negative if the tangible winter weather had turned out the opposite of what had been predicted (e.g. very mild and very wet)?

The MetO's performance last year was fair but was a bitter failure for many areas of England, including mine. But maybe I am just judging it by my expectations of other "cold and dry" winters like 1995/96, as I have already said.

Anyhow, moving on and looking to the future, a weakly positive NAO winter 2006/07 (which 1990/91 was) with above average precipitation, from a major UK weather body like the Met Office, sets me looking with great interest to next season, as Bottesford said.

I agree that Winter 2005/2006 was another failure to add to the preceding 9 winters, in terms of snowfall. But, as the Met Office were not forecasting snowfall (unless you interpret dry = snowless) you were clearly expecting too much from that LRF - it cannot be right to blame them for this. Letdowns were only created by several poor short-range forecasts, which is another debate entirely.

You cite 1990/1991, but how many recent weakly positive NAO winters have not produced widespread snowfalls? Have we even had any recent weakly positive NAO winters (using the Met Office measurement)? How does the overall long-term percentage mild:cold winters stack up when NAO is weakly positive?

Edited July 17, 2006 by The Enforcer

[The Enforcer]

Considering its the hardest Teleconnection to forecast then I think at least for now the MSLP anomalies rather than the NAO are the way forwards in giving us a tool to forecast at a seasonal level-

MSLP - that's a new one on me. What's that when it's at home?

[Dawlish]

To defend the Met Office, which is what I usually do - they are the equal of any agency in the world at what they do and we started meteorology anyway! The Americans would pale at the difficulty or trying to forecast our weather on a daily basis and get it right (I'm probably hopelessly wrong there, they'd probably do as good a job, but a bit of jingoism sometimes helps!). The Met Office are doing as well as they can, but face a dreadfully difficult task in terms of public opinion.

Take 2 winters ago. metcheck issued a long-range forecast saying that it was going to be the coldest winter since goodness knows when. Of course it wasn't, but who got the flak? The Met Office got the flak for someone else saying it would be a dreadfully cold winter. The general public just hadn't caught up with Internet weather sites and other providers of equally iffy long-range forecasts.

It was the same with last winter's forecast. The fact that there was an eventual perception that they were correct saved them (I would dispute the correctness of that forecast to my dying day - well, at least until I go to bed tonight). Their use of the NAO as a winter forecast could have been a public relations nightmare and it almost was, but they only get this bad press because of their prominence and the nations' metorological ignorance, which creates an expectation that the Met Office should always be correct! I really hope, for their sake, that a max weather record for July is set in the next two days.....or they will get crucified in the Sunday press, yet again!

Paul

[hammerb32]

Dawlish et al

Another thing that struck me was how quickly the media latched on to any impending cold spell, some of the tabloid headlines, especially in the Daily Mail/Express/Star were laughable, it was almost as though they were getting their info from the maddest snow/cold rampers on this board and reporting them as fact incorectly attributing them to 'met office sources'.

In general I find the MO and BBC to be pretty near the mark both long term and short term, some of the mid term forecasts ie the month ahead or the next 16 days are questionable, but we all know how difficult that is to forecast.