Autumn And Winter Lrfs

[ribster]

PS Ribster, I have no idea what you are whistling about. You have obviously been on too many courses!

Courses? Interesting that you thought the comment was directed at you!

[West is Best]

Hi Paul- There are only 3 people on this site who spend a considerable time preparing Winter Forecast-

I think I can see what you were getting at Steve, though it's rather infelicitously phrased (well hopefully it was just an error of language).

I think we need to be careful here. Some people and organisations may spend a lot of preparing winter forecasts (and indeed others) but put their thoughts succinctly, rather than in minutiae. That's sensible because the moment you go into intricate detail in a LRF you are asking for trouble.

What I particularly liked about the Met Office forecast was that despite aeons of research they made it 1. nice and short and 2. in plain English. They also stuck to generalities and posted % chances. Oh, I nearly forgot, they also got it right.

Others on here also spend a lot of time preparing LRFs but are able to put their thoughts in a few sentences. When like the Meto they also get it right they shouldn't be ignored for failing to dress everything up in complexities.

On the teleconnections issue - I keep telling myself that they should work. But I return to a point I made a season or two back: unless you take seriously the local factors it's possible to make the upstream markers the be all and end all, and end up with egg on one's face. Looking to the Atlantic is obviously very important, and the jet too etc. etc.. But I think frankly that once you start heading out west past the Atlantic the links begin to become tenuous at best. Too many downstream factors seem to make the teleconnection forecasters less than successful. Again, I keep telling myself that's merely because the 'science' is still undeveloped, not that teleconnections are like looking at the universe through the wrong end of the telescope.

Edited July 20, 2006 by West is Best

[pureasthedriven]

Others on here also spend a lot of time preparing LRFs but are able to put their thoughts in a few sentences. When like the Meto they also get it right they shouldn't be ignored for failing to dress everything up in complexities.

I find myself agreeing with this post, West. I think what is important is that we all recognise that there are a blend of methods and approaches to the weather out there and until one single method is proved flawless then none should be disregarded. The range of methods makes for interesting reading, though I would say that in tracking forecasts and models for the last few years, it's often the case that the 'hunchers' (such as myself) can be as accurate as the 'scientists' on occasions, in the very short term (last nights projected storms in the SE being a prime example - one could just tell by looking at the sky and sniffing the air that they were never going to happen in my area).

I think the danger can be that yes, the UK weather can be unpredictable and changeable (that's a laugh at the moment!) but we most certainly can also make reasonable assumptions based on a small amount of sciene coupled with a healthy slab of experience and common sense.

Edited July 20, 2006 by pureasthedriven

[johnholmes]

re this comment Steve

In terms of the MET office you CANNOT just build a Winter forecast around the NAO- I think this is why this remains as their secondary Forecast rather than Primary-

are you really trying to say that you as an enthusiastic amateur, no offence intended, know more than they do? Anyway, as they have explained, if you read it, that is NOT what they do.

their quote on their method is below,

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

Re the Paul post(Dawlish), not often I totally agree with you Paul but the post about stats and the 95% prob being required is totally correct. Many on here do not seem to appreciate that this kind of stats discipline is needed to show any improvements or accuracy is real not chance. All the major weather centres do this as a matter of course. One only has to look at the Met O site with their stacks of tables showing how much they have improved, or rather how little, to the average man in the street, their forecast accuracy has improved.

John

Edited July 20, 2006 by johnholmes

[West is Best]

it is possible to read such posts and use them as an invaluable technical baseline for the way ahead (what to look for in the models etc) whilst also applying your own anecdotal, commonsense, looking out of the window, previous experiences of a given weather pattern etc

Yes I agree with you here. However, I do want to caution on two things (not to you but to the whole debate). First, I wouldn't want it to be thought that 'professional' metereologists are the ones who use complex upstream language to forecast, and that those who don't aren't. That's not a division I'm comfortable with. Don't forget that the two or three most avid teleconnectionists on here are, like you and me, amateurs.

Second, I don't want to think the obverse of teleconnections is less scientific and more 'commonsense'. I really don't think that's true. Scientific analysis of the downstream data may simply appear less impressive but may contain just as much vital scientific rigour and, as importantly, produce just as accurate forecasting imho.

At the end of the day mother nature knows best and we will never reach her standards - as much as we my aspire to IMO. I think that there will always be a definitive limit of time wherepon a reasonably accurate forecast can be made, irrespective of advances in science and technology,

I'm not so sure I can agree to this. If metereology is a study of something with empirically verifiable facts then eventually it must surely be possible to nail it? The only reason for thinking otherwise is to take a mystical/religious approach to it (Mother Nature, Gaia, God, gods etc.). I don't want to give weather the power of an outside supernatural force, however majestic and awesome it can be and, therefore, it must theoretically be possible eventually to forecast with 100% accuracy. (Note the word theoretical!)

[pureasthedriven]

There is much that can be learned from Glacier Point in his posts, and for the likes of me as well as you pureasthedriven (as one of those 'healthy slab' people! ) it is possible to read such posts and use them as an invaluable technical baseline for the way ahead (what to look for in the models etc) whilst also applying your own anecdotal, commonsense, looking out of the window, previous experiences of a given weather pattern etc to make your own decisions as to what may/may not happen

SP: quite.

I also think that irrespective of major long-term factors (such as GW as the biggest example) it is possible to see that our weather has actually produced 33 to 35 C heat on many occasions before in history (in years regarding as very pre GW)and perhaps we tend to over emphasise the significance of these events when we apply them in forecasting

And again, SP; quite. You should have heard Ken Livingston yesterday, telling us all that we can expect year on year of this now, becoming hotter and hotter. Since when did Ken also become Stratos Ferric!

Edited July 20, 2006 by pureasthedriven

[The Enforcer]

I think we need to be careful here. Some people and organisations may spend a lot of preparing winter forecasts (and indeed others) but put their thoughts succinctly, rather than in minutiae. That's sensible because the moment you go into intricate detail in a LRF you are asking for trouble.

Agree with most of what's been said this morning. The only points I would add are:

1. Even when an LRF gets it right or when LRFs agree, they can only realistically predict overall trends. I am thinking here about whether a period is westerly or easterly etc and a broad temperature range (e.g. v-mild/mild/average/cool/cold). What they definitely can't predict is snowfall. Last winter was synoptically very different to all the others this century, but just as snowless.

2. As far as I can recall, winter has never been an autonomous season. By this, I mean that even the coldest winters will have mild spells and vice versa. Therefore, it would be disingenious to try to summarise the forthcoming winter season under one heading (e.g. cold/mild) as mild could be mostly very mild but with two very cold periods or slightly mild with only one cool spell. To draw comparison with Dawlish and JH's %'s debate, I believe the ideal LRF should be to suggest how much of the winter season would be taken up by various synoptics (e.g. 60% westerly, 10% southerly, 10% static, 10% easterly, 10% northerly).

[The Enforcer]

SP: quite.

And again, SP; quite. You should have heard Ken Livingston yesterday, telling us all that we can expect year on year of this now, becoming hotter and hotter. Since when did Ken also become Stratos Ferric!

Having to run the Capital explains why SF spends long periods off the forums.

On the other hand, SF doesn't give any weight to LRFs so he can't be KL.

[summer blizzard]

Don't worry about it Steve. I think it's time you and I moved on with this or we'll incur wrath ... in due course when I see the 2006 archives I'll do a demonstration. I don't think anywhere publishes the charts that have just occured do they? NW archive charts only go up to 2005, so I have no option but to use model output charts.

West Is Best, the archives on TWO and Wetter Zentrale go upto and include June 2006.

Folks

Can someone explain the METO stance this winter. I take it the METO are basing this forecast on the SST in May? Were they not very similar to last year if not more inviting? I see that big Stevie MURR has mentioned the warm SSTs off Newfoundland as a possible reason as it suggests a positive NAO.

I find it ironic that the Met Office have forecast a positive NAO next winter and now nearly two months after the data has been taken in, the sea surface temperature setup in July 2006 is now more conducive to a negative NAO than July 2005, unless the Met Office have programmed this into their calculations, in which case the reason for the positive NAO forecast is that there are expecting the warm anomolies in the Mediterrannain to persevere creating a Barlett Senario.

Another aspect is the upper level weakness in the eastern Atlantic and the mean NE'ly flow there likley to result in cold anomalies during the winter from residual SSTAs - potentially inducing a trough over western Europe and displaced jet flow to the south.

Glacier Point, are you describing the type of Jet Fow that i have illustrated on the chart below...

Looking at SSTs:

Last year:

http://weather.unisys.com/archive/sst/sst_anom-050718.gif

This year:

http://weather.unisys.com/archive/sst/sst_anom-060716.gif

Notable differences - the emergence of warm anomalies off Newfoundland which were not present at this time last year and GP has already stated what the impact of those would be, were they to persist.

I also see that whereas a year ago there were significant warm anomalies in the North Pacific, these do not exist this year - what affect will this have? Slower Jet in this area?

Am I to understand that under the above set of circumstances, we would have a strong jet, but this would be deflected further south than of even larger teapot ilk?

The Enforcer, the joined warm anomolies in the eastern Pacific were indicative of a strong Jet Stream, so yes, current indications are for a weaker Jet Stream in the eastern Pacific however, because there are both warm and cold anomolies along the eastern USA, current indications are for a fairly strong Jet Stream leaving the USA, ideally we would like to see either a dominant warm or cold anomoly.

As for your last question, i believe that would be the case and if so, this winter could be quite snowy at times.

ALL opinion are welcome-

As for the pacific SST's they will have a big effect on this Winter- THese SST's play a part in determining the Strength of the Pacific jet-

For information look for the teleconnection called the PDO if you are really interested-

However of the 2 factors contributing to the pacific jet the bigger influence is the QBO- we need it to be Westerly rather than easterly this year to minimise its effects-

S

Is this because high pressure would be diplaced to mid or high lattitudes with a trough at low lattitudes?

[BLAST FROM THE PAST]

I find it ironic that the Met Office have forecast a positive NAO next winter and now nearly two months after the data has been taken in, the sea surface temperature setup in July 2006 is now more conducive to a negative NAO than July 2005,

SB,

This is precisely what I was aiming at...but then again I probably am looking at the set up incorrectly

[Glacier Point, are you describing the type of Jet Fow that i have illustrated on the chart below...]

SB that is how I read into it too....would be nice

BFTP

.

Edited July 20, 2006 by BLAST FROM THE PAST

[Damien]

July's IRI has been released. I think the charts speak for themselves:

We should take comfort from the fact that this forecast has been really poor in the past. For comparison, this was its previous run:

http://iri.columbia.edu/climate/forecast/n...6_Eur_temp.html

http://iri.columbia.edu/climate/forecast/n...6_Eur_temp.html

http://iri.columbia.edu/climate/forecast/n...6_Eur_temp.html

http://iri.columbia.edu/climate/forecast/n...6_Eur_temp.html

If anything the former showed the "warming trend", as it became average after August: and the latter the cooling trend as December is still set to be only slightly above average ultimately.

I quite like events in Europe and Western Russia on the last chart though were it not absolutely unreliable.

[Damien]

We have an NCEP update today as well. Summer is still expected to be very hot but with an interesting August chart now for England. September is now cool for most of us from the north and north-west. October now looks mild from the west generally, and quite dry too. November is very cool principally from the North and now with a Scandinavia High; December is expected to be very mild from the Atlantic with the mild air tearing right up into Scandinavia.

In the less reliable longer term this scenario is not really backed up with a mild September from the south and a generally cool October from the north. November is now expected to be mild on here, as is December. January is coming clearly into the same vein with a weakly positive NAO now starting to form once agin over Southern Europe. February keeps us cold but only just - with the mild air now just starting to tear slowly further into the British Isles from Ireland, Western Scotland, and Northern Cornwall. Southern Europe remains cold, especially over Sardinia as opposed to the Balkan region though.

Winter is over.

[SMU]

re this comment Steve

In terms of the MET office you CANNOT just build a Winter forecast around the NAO- I think this is why this remains as their secondary Forecast rather than Primary-

are you really trying to say that you as an enthusiastic amateur, no offence intended, know more than they do? Anyway, as they have explained, if you read it, that is NOT what they do.

their quote on their method is below,

This forecast is produced using a combination of statistical models, such as the North Atlantic Oscillation (NAO), and complex climrational forecasters

Re the Paul post(Dawlish), not often I totally agree with you Paul but the post about stats and the 95% prob being required is totally correct. Many on here do not seem to appreciate that this kind of stats discipline is needed to show any improvements or accuracy is real not chance. All the major weather centres do this as a matter of course. One only has to look at the Met O site with their stacks of tables showing how much they have improved, or rather how little, to the average man in the street, their forecast accuracy has improved.

John

John in terms of overall Meteorology Certainly not- Their expertise in local & short term forecasting is Second to none-

However I do feel that the field of Seasonal & LRF's in general is something that actually the Met office are NOT the Pioneers-

I seem to remember the Met office abandoning their research into LRF's a number of Years back- & I 'Doubt' the investment is at a level that it was before-

You can take this next paragraph how you wish- but for Some of Winter 03/04 but mainly All of 04/05 & 05/06 I spent hours- maybe overall days studying the LRF preperation/data that the Professionals in America use- to say they are in a different league is an understatement OR the met office actually fail to detail the level at which they study the global paterns-

As for the NAO statistical model quoted Im sorry but that ISNT robust- & I havent seen any extracts from its Skill level/Accuracy-

If the data was there somebody Over on the 'other' forum I work/post with would have published it & used it-

Believe you me Im really not trying to come across condesending in any way ( and do make an appology to those I didnt mention last night) But I will give you an example on how this topic makes me feel- which you can probably relate to-

When you see a post in Winter & Sometimes in the Summer slagging off the Met office for Inaccuracies you feel like & often do jump to their defence- this I guess Is because Partially yuu worked there & feel that a negative comment to them still encapsulates you, but as well you feel equipped to set straight any innaccuracy in the aforementioned statement-

Alos John remember we are privvy to ALL of the met office monthly forecasts & to say they are generally perceived as poor would be fairly accurate-

Well its a bit like that for me with details & forecasts on LRF's-

Firstly I think it could & should be a seperate/Specialist part of the Meteorolgical Science, but I actually feel entirely equipped to stand up alongside the big boys and argue a case- and that does include the Metoffice- Just because I havent got a job at the Met office doesnt underqualify me in a field that actually the most qualified people arent the Met-office-but the ones investing a huge proportion of their time studying that small segment of forecasting-

thats not to say its great by any stretch of the margin & something that will no doubt be discussed to the N'th degree in Oct/Nov-

The crux of the Matter is if someone wants to be credible they put there arguement across and issue a forecast- THe challenge that will ALWAYS come back is whether the 'issuer' has put the background study in on All the Paramenters or whether they may have just dismissed them as 'noise' & just focussed on what they think is most important......

Anyway- best regards

Steve

Edited July 20, 2006 by Steve Murr

[kold weather]

Very intresting post Steve. I can't say I've really put in a massivly detailed LRF before but I'm tempted to give one a go.

However I do know where you come from, I spent quite a few hours over the lasat few months looking at data to try and guess exactly what the hurricane season will be like. In some regards even short term hurricane forecasting is like long range forecasting because its so complicated and not all is known about how they work and form.

Its also the case that there are some people here who are very bit as knowleadgeable on thier chosen feild as some of the experts who have studied in Uni's and so on. The main difference between the Met-office and amuters like most of us on here is simply the amount of data. The pros usually have far more data to use therefore a far higher chance of being right, whilst people on here have to make the most of what they have got and depend on the models that are publically present.

(ps, Steve, the forum your talking about is probably Eastern us I believe, been a floater around there sometimes to find out data like the SOI and others to help with long range hurricane forecasts.)

[SMU]

Very intresting post Steve. I can't say I've really put in a massivly detailed LRF before but I'm tempted to give one a go.

However I do know where you come from, I spent quite a few hours over the lasat few months looking at data to try and guess exactly what the hurricane season will be like. In some regards even short term hurricane forecasting is like long range forecasting because its so complicated and not all is known about how they work and form.

Its also the case that there are some people here who are very bit as knowleadgeable on thier chosen feild as some of the experts who have studied in Uni's and so on. The main difference between the Met-office and amuters like most of us on here is simply the amount of data. The pros usually have far more data to use therefore a far higher chance of being right, whilst people on here have to make the most of what they have got and depend on the models that are publically present.

(ps, Steve, the forum your talking about is probably Eastern us I believe, been a floater around there sometimes to find out data like the SOI and others to help with long range hurricane forecasts.)

yes Kold- I saw you posting the other day

Luckily I have a Proforecaster tag ( red) so I can see the pro exchanges as well away from th 'Weenie' as they put it eyes....- Good mates with Nineinchnails & a few others-

Valuable assets-

S

I'm glad you're so optimistic; today's LRF models in general have left me feeling physically sick. No really, my stomach *really* does feel quite queasy right now, and it seems to have gotten worse after doing today's updates.

Damien- they are poor in terms of accuracy- very poor infact-

Pinch of saxa at this stage- Id be concerned if today was Nov 20- chill- weed smoking smiley...

S

Edited July 20, 2006 by Steve Murr

[Damien]

I'm glad everyone else is so optimistic; today's LRF models in general have left me feeling physically sick. No really, my stomach *really* does feel quite queasy right now, and it seems to have gotten worse after looking at today's runs.

[SMU]

I'm glad everyone else is so optimistic; today's LRF models in general have left me feeling physically sick. No really, my stomach *really* does feel quite queasy right now, and it seems to have gotten worse after looking at today's runs.

how did my reply get before your post???

S

[The Enforcer]

how did my reply get before your post???

The boy's a bit special.

I'm glad everyone else is so optimistic; today's LRF models in general have left me feeling physically sick. No really, my stomach *really* does feel quite queasy right now, and it seems to have gotten worse after looking at today's runs.

Don't worry, they'll show a different picture in a few days.

[The Enforcer]

THe challenge that will ALWAYS come back is whether the 'issuer' has put the background study in on All the Paramenters or whether they may have just dismissed them as 'noise' & just focussed on what they think is most important......

It is the easy option to pick on one factor and say that it's the most important variable and it's saying X therefore winter will be Y. Unfortunately, you'll find that most people on here, whilst having the interest haven't got either the ability, time or inclination to spend days wading through lots of technical data.

This begs the question, is it actually wrong to produce and LRF without going through the rigorous, robust process you have outlined? My take on it is that it is unfair to compare the type of LRF that you or GP might produce with the type of LRF that KW or SP might produce - that's unfair on both sides. This is one of the main reasons things get overly personal. Difficult to say where one could draw the line to say which LRFs are 'technical-based' and which are 'pattern-based', though. Nice little headache for OON.

[johnholmes]

hi Steve

I'm not getting into an argument on the basis of your last comments. You have your view on what the Met O do or do not do, I have my view. Just leave it at that. Perhaps you could send an e mail with your points to them and publish their reply. I'm sure they would reply.

Just one point about your answer.

I do hate lots of bold print! I did NOT use ANY, so why have you or was it just a slip as you copied it?

John

[SMU]

hi Steve

I'm not getting into an argument on the basis of your last comments. You have your view on what the Met O do or do not do, I have my view. Just leave it at that. Perhaps you could send an e mail with your points to them and publish their reply. I'm sure they would reply.

Just one point about your answer.

I do hate lots of bold print! I did NOT use ANY, so why have you or was it just a slip as you copied it?

John

John I was just highlighting the part of your post that I was referring to-

It seems you are making more of this than need be- Why write about an Arguement- I politly tried to put a point across- at no point did I I try to put anyone down-

If we have a difference of opinion thats fair enough- I dont want to Email the met office Id rather stay observing the people that I think are the leaders-

Whilst I have acceptped your arguement you do seem to be ever so slightly tunnel visioned in as much as the Met-office can be the only organisation that can produce something with credibility-

Like I said no offence to anyone- Im trying to be ( with everyone) quite conscious regarding the way msges come across & I didnt see it as being offensive- and for the record- I didnt say the Met office was bad- just thatwith some things I think there are better expertise elsewhere-

Steve

[Weather-history]

Here's are the hottest summers on record in order and the amount of rain that fell during the subsequent September. Some summers are missing off the list because the rainfall series started in 1766 whilst CET records go back to 1659.

Does a hot summer lead to a wet September?

1976: 150.8mm

1826: 135.6mm

1995: 123.3mm

2003: 34.7mm

1846: 53.8mm

1983: 100.8mm

1947: 46.7mm

1933: 60.6mm

1781: 86.7mm

1911: 64.2mm

1899: 92.9mm

1975: 106.5mm

1868: 90.2mm

1779: 100.2mm

1808: 90.4mm

1818: 115.6mm

1959: 8.0mm

1997: 30.8mm

1798: 95.7mm

1857: 94.0mm

1955: 75.7mm

1778: 61.1mm

1949: 47.1mm

Mixed I would say

[Damien]

Summer is largely the same and still very hot overall. September is now much cooler from the north. So far so good. October is largely much warmer and drier now from the Atlantic. November remains quite cool from Scandinavia in particular but mild air close by in Northern Europe. December is still very mild - and even more mild than yesterday as the air continues to dig in from the Atlantic. But remember that many of the classic 1980s-style negative NAO winters started with a mild(er) December with the real cold only hitting in the January-March period so there is no need to worry just yet.

The longer term generally reflects this but with a very unusual-looking September chart and the October and November overall periods still largely cool and mild respectively. December and January continue to look atrociously mild on these charts - January now looking like January 2004 which may please some people. (And who said I was not altruistic?) The mild also wins in February but with the very mild air slightly further out in the Atlantic now slowly retreating with some below average air still to our east in the North Sea.

http://www.cpc.ncep.noaa.gov/products/peop...bT2mMonNorm.gif

http://www.cpc.ncep.noaa.gov/products/peop...bT2mSeaNorm.gif

[reef]

One thing I keep noticing when looking at these long range charts is that there is always a constant very warm anomoly directly to the west of Ireland in the middle of the Atlantic. Is there any particular reason why this is always there? It seems to me like it is an error in the model, as the months closer to present never seem to have it yet as you go further out it always appears, no matter what the month.

Something to think about with regards to how useful these LRFs are maybe?

[Damien]

One thing I keep noticing when looking at these long range charts is that there is always a constant very warm anomoly directly to the west of Ireland in the middle of the Atlantic. Is there any particular reason why this is always there? It seems to me like it is an error in the model, as the months closer to present never seem to have it yet as you go further out it always appears, no matter what the month.

Something to think about with regards to how useful these LRFs are maybe?

Yes, I have noticed that too.

I can also tell you that the mid-Atlantic warm *anomaly* has consistently expected to be at its weakest in November this year/winter, be it materialise again.