Steve just to repeat in the period 1826 - 2015 only 40% of NAO - ve Octobers were followed by NAO -ve Decembers

You suggested looking at more extreme values.If we take NAO-ve Octobers with values <- 1.5 there  are  47 in the series.Only 32% were followed by NAO -ve Decembers So if anything NAO - Octbers are a predictor of NAO +Decembers! Any how I  doubt the figures are statistically  significant.

Quite honestly I have looked at lots statistics on relationship between Autumn and the individual autumn months and the following winter.None have yielded any worthwhile link.Of course we all have our favourites. I thought for sometime that vry dru warm Autumns were good forecasters of a mild winter but that  has not worked  wellin the last 30 yeaers.

The best predictor I have found on circulation types is that if December is NAO + then the winter as whole  has a higher than average chance of being NAO +  as do the individual months of Jan and Feb.

Not much help in spotting the cold winters we crave!

when was the last time we were above average in our Arctic ice amounts?

More generally you highlight the problem seasonal pattern-matching always faces: we simply lack enough years of records to be sure that any results are significant. Perhaps by the Winter 2116-17 thread we'll have a better idea.

Correct, an article that appeared in Weather Magazine in 1982 discussed the exceptionally low temperatures of the previous December and said satellite images suggested temperatures below -30c over the Welsh Border during mid December and again during the second week in January.

This should not come as a surprise as it's highly unlikely that a weather station happens to be located in such a cold spot!

In 2007 I set up a weather station at my new house in the village of Plumpton, Cumbria and have many times recorded minimum temperatures well below the nearby MetO site, in Decemeber 2010 I recorded a mean monthly temperatures of -3c a new record for the County.

Andy

Forget pattern matching, sophisticated long range forecasting and all the rest - just stick to the sheep's entrails.

I am going for a colder winter than average, because that is what I feel in my waters.

Plus, we are heading into solar minimum after the weakest solar cycle in over a century. I am firmly in the solar camp, when it comes to the biggest determining factor in our regional climate. Weak solar activity and a negative NAO appear to be linked.

Reasearch at Reading University has  confirmed that there is some linkage to colder winters around the solar minimum>however I don;t think we are close enough yet to the solar minimum for this to kick in.

At your age, that may down to needing to get your prostate checked out

What would you have posted if I had written that it is a feeling I get in my bones?

I know  the feeling! Mind you, I think it's more obvious this year given the odd weather we've had over the last month. Feels like a front-loaded winter to me - which the bookies will love if December is a cold one!

October is often a quiet month. The calm before the.... 

Evening All-

So I thought I would just expand on Rogers review of hoping October would hold some relevance into the Winter Months- the overwhelming data since 1950 is that there is very little to suggest that there's any significant correlation, thats because for as many months you find Negative from an AO perspective correlating into Winter, you can find as many positive ones that tip the balance- This certainly fits to the zone (-1 < to >+1) AO

However in more recent October years circa 2001 onward the height profiles over the pole ( relating to the rapid loss of ice & subsequent change in feed-backs) have become more extreme & amplified - From 2001-2016 there have been 2 months below -1.5 AO (2009 + 12) & one rounded to 1.5 ( 2002 ) -  so 50 years with zero & then 16 with 3... whilst being a small sample size seems 'fairly' significant & outside the boundary of what you would label 'fluctuations' -

Adding more meat to the bones & trying to link this into Winter is the following--

The X2 record breaking Octobers for Negative AO are 2009 ( -1.54) & 2012 ( -1.51 ) - Below are the AO values for DJF

2009 -  Dec -3.4 ** ( record Breaking in 66 years ) Jan -2.58 ( 6th lowest in 66 years ) Feb -4,26 ( Record Breaking in 66 years )

2012 -  Dec -1.74  Jan -0.6 Feb -1.00                 +   March -3.185 ( record Breaking in 66 years )

So - Co-incidental ??? I doubt it, 3 record breaking negative AO months in Winters ( March INC ) following the record breaking Negative AO's in October.....

2002 -  Dec -1.59  Jan -0.47 Feb +0.12 -   

Thats a pretty good hit rate for the 3 record months -  9/10 negative AOs & The December / Feb & March records Smashed.

The NOAA say the AO correlation to the NAO is around 0.7 - & along these 10 months the NAO was negative for 6 months which supports that claim.

- 

Whilst I am not spending a great deal of time on the flip side with a positive AO- its worth mentioning that within the years highlighted again ( 2001 onwards ) the December +NAO record was smashed in 2011- @ + 2.52 & challenged again in 2015 with + 2.24, The January +NAO record was beaten in 2005, only to be smashed in 2015 @ + 1.79 & 2016 Sits 4th all time positive @ 1.58 ( only surpassed by 89 & 97) 

The evidence is overwhelming that the polar height field is going from one extreme to the other - & the old fashioned Neutral type patterns ( NAO & AO -1.5 < > +1.5 ) are being forgotten about.

This essentially renders any pattern matching USELESS Pre 2000. Maybe even pre 2007, we are now clearly in what I can only label the POST modern winter - & Ian Browns modern Winter >1988 is RIP. 

So where does that leave Oct 16 - Well the sample size is Small - only 2 months @ sub -1.5, but both Winters combined delivered record breaking polar Statistics.

Lets look where we are month to date & with high confidence to the midpoint of the month the 16th. - To set a new record @ -1.6 we need the total monthly negativity to be -50. (-1.6 over 31 days )

Heres the data so far, 

http://www.cpc.ncep.noaa.gov/products/precip/CWlink/daily_ao_index/ao_index_ensm.shtml

I put us around -35 to the 16th (-2.18) leaving 15 days to average -1. The ensemble Mean is clearly guiding us to a secondary trough starting around the 20th, so *IF* this verifies even to a lower degree the models suggest another -1 to -2 second half of the month.

October 16 is looking to SMASH the record as it stands-  & in my book based on the evidence provided that suggests a STRONG probability that 1 month of our winter will be close to record -AO & the at least 1 other month will be sub -1. The 0.7 Correlation slightly dilutes the impacts, however based on this small snapshot I would be forecasting a 2/3 -NAO winter with a 3/4 probability if you include March - which whilst not a Winter month can deliver spectacular winter weather.

Whilst we are only at the midpoint of October, Im looking for some ( but NOT ALL ) underpinning of -AO pattern in November, this is also where the statistical evidence gathered over 60 years supports a VERY good correlation that a November AO of SUB -1 supports an  84% chance that each winter month will be Negative -

Here is the Evidence.

November Months with sub -1 recorded. ( we haven't had one for 14 years )- 

1952,55,57,59,62,65,68,80,85,98,00 & 02 =   12 Years. 

Of these 36 Months 30 Winter Months were Negative AO -- 84% & the NAO was 19/36 -- 52%. Adding to this across 36 Months there was only 2 +1 NAO months - Dec 65, Jan 86 ...

I believe that Statistically that represents good value that November -AO sets the tone for Winter .... But only again within the bounds where -1 was the monthly result ( the neutral zone doesn't give an even picture )

Reversing the polarity again, for +1 AO Novembers we get

1978,93,94,11,13,15  = 18 Months.   The AO only manages 6 Positive Months, so only 33% -

However the following winter NAO was positive 15/18 Months ( & of those its every year post 78 was positive ) - 84% correlated !!

To Summarise then:

October 16 is currently on course to be a record breaker & from the 2 like for like years we should have moderate expectations that the following Winter should produce a significantly -AO pattern & a similarly but diluted down -NAO pattern. its only moderate confidence because of the small sample size.

If the AO comes in negative for November- Ideally sub -1 then our confidence in the above forecast should move from moderate to HIGH expectations- & at that point I would give a high probability of a below average winter with higher incidence of snowfall.

All eyes on the AO for the remainder of OCT & NOV to hopefully set the tone for the Winter...

S 

Thanks Steve. Really fascinating information there. Definitely one to watch. Will be interesting to see how all the different variables that may impact our winter come together by the end of November. 

Further on the topic of AO/NAO, Tim posted a comment here about WACC:

http://www.theweatheroutlook.com/twocommunity/default.aspx?g=posts&m=828929#post828929

I suggest people look into WACC (Warm Artic Cold Continents), since it looks to be a signal or precusor to a blocked winter pattern.

More here:

https://www.aer.com/science-research/climate-weather/arctic-oscillation

"

• Eurasian snow cover is above normal but has currently stalled.  However both low sea ice and a persistent negative AO are favorable for the resumption of a more accelerated advance of Siberian snow cover. High Eurasian snow cover and low Arctic sea ice in the Barents-Kara seas favor first, a strengthening Siberian high and then a weakened polar vortex in winter.
• The polar vortex both in the stratosphere and the troposphere look to remain relatively weak for the remainder of October, if this trend continues severe winter weather is likely for widespread portions of northern Eurasia, including Europe and East Asia, and the eastern United States (US).

"

Nice Post by MM - with Cohen picking up on the years I have posted ( & have been posting on NW for a couple of weeks )

Whilst he says there isnt a strong correlation from Oct > Winter a few more sub -1.5 oct months may sway that feeling-- that however may be 10 years away!

 As I said in September,this year there is no one extreme indicator.The  ENSO  forecast has shifted to a weak La Nina, slightly favoring a more NAO - circulation ,The QBO is heading for a record run of +ve consecutive months which if it continues suggests a stronger than normal westerly circulation.SST forecasts suggest a cold patch to the west of the UK which maybe contributed to the mild winter(though this years forecast has a colder anomalyl  further east to the north of the UK. )Sunspot numbers are in decline but are unlikely to be a factor until around the solar minimum.

So o although there are some conflicting signs I can at present see no strong cold winter signs.So default mild remains my forecast.

One area I am exploring is the influence of December circulation type on the rest of the winter.Hope to have some figures in the next week or so.

1.ENSO

The latest (October 12th) NOAA discussion shifts from  neutral NAO conditions  for next winter( Sept forecast) to a weak La Nina watch.

Weak la Nina conditions  conditions have occurred 11 times in the 65 years of the data set..The winters DJF in this grouping have a 65:45 split in favour of NAO - ,with Decembers being 55:45 NAO +.

So overall a slight shift towards a colder circulation

2.QBO

Currently the QBO  remains positive increasing from 10.07 to 10.48 in September. There have now been 16 consecutive QBO + months which equals the record set in 85/86 and 99/00.So it is highly likely that the record will be broken.It  favors westerly circulation and a milder than average winter.

3.SST's

May SST's did not display tripole pattern.The cold mid Atlantic patch is now declining with warmer Atlantic temperatures and warm sea south of Greenland.However the METO probability maps  for DJF suggest a continuing cold patch to the west of the UK like last year(although there is the suggestion of some lower SST's spreading to the north and north east of the UK as well. 

4.Sunspot numbers.

The trend is downward  although there is a fair bit of jumping around in the last 3 months.September sunspot numbers fell to 44.7 but that is along way above solar minimum values likely around 2020 .and its  around this time that a colder influence is likely.

5.METO probability maps

These suggest near average conditions in terms of temperature and rainfall.

 6. CRS Probability maps.

These show above average temperature and rainfall for DJF

Edited by moderator 16 September 2016 10:51:39  | Reason: stickied

http://www.telegraph.co.uk/news/2016/10/14/twitchers-flock-to-village-car-park-to-catch-glimpse-of-rare-sib/

Yep, a cold winter must be on the way!

You may scoff, but it takes me back to the early 60s when early sightings of Siberian birds at Slimbridge presaged the bitter winter of 62-63.

Of course, the birds cannot predict the weather, but they respond to the climate in real time and can provide clues about the direction of travel.  We had a flock of migrating geese land in a nearby field yesterday. Let's hope a 62-63 winter is in the offing.

My big problem is the absence of any strongly negative Arctic Oscillations in November to accompany the October numbers. It's not as though they are difficult to achieve. As you show, there were 12 of them in the half century up to 2002. Yet in the subsequent 14 years they have been completely absent, despite record-breaking numbers during that time for the month before and the months after.

I know it's not implausible that there might be some mechanism whereby patterns in October only resolve themselves into similar ones months later, after undergoing a change in November. And it wouldn't surprise me if there were one just waiting to be explained to me. But until then, in view of the small sample size, I'll just file any extreme-negative October AOs under 'well, that's interesting' rather than 'where's me sledge'.

Oh absolutely! However, such conditions could be more favourable to UK cold winters than otherwise.

Hi Steve,

A Fascinating study there. Lets hope that Novembers AO falls into the very negative category!

Andy W, the early winter of 81/82 and it doesn't surprise me one bit that -30C was breached in that winter as the temperatures were truly remarkable.

In the dark days of the little ice age, I reckon -30C was breached a few times especially during the 1739/1740 and 1683/84 winters such was the intensity of the cold. 

I read somewhere, that -26C was recorded in Beckenham in Kent in the 1700's I do believe. Again this is an unofficial reading but goes to show how cold it really was during the little ice age.

Grant