The Seasonal Forecast Thread

[VillagePlank]

Updated my forecast given the unfolding November temperature.

Essentially, this is the analogue method on statistical steroids. I look at anomolies of 2015 from the preceding 30 year average of Jan-Nov of the year, and then compare it to the anomalies of the average for every year in the CET record. The closest match, by the Pearson Correlation Coefficent wins, and then I extrapolate the following Dec, Jan, and Feb from the top five years, and average them.

Assuming Nov will post a CET of 9.5 the years, the closest years are 1994, 1661, 1894, 1913, and 1938, in descending order, registering a coefficient between 0.72 to 0.54 respectively.

Here's the data



Here's the chart comparing the average of the five top years to what we currently know of the data for 2015,



Here's the prediction,



So .. warm December, cold January, average February; essentially a mixed bag.

For interest, the closest modern year, 2009, which came in 6th at a coefficient of 0.53 given the +1.53 November anomaly, and cool July that year. The following Dec, Jan and Feb were all cold and/or much colder than normal each posting anomalies of -1.77 or below (-3, for Jan 2010). But it didn't make the top 5, so wasn't included.

I may well redo this so that the average contains every year that posts a correlation of >= 0.5, which, in this case would be the top ten years so at least years such as 2009/2010 get a 'vote'

Edited November 25, 2015 by VillagePlank

[johnholmes]

The data you have there and the result is quite interesting.

It is not that far from either the lrf done by Ed and Tony or the base line model outputs from UK Met?

We will see on 1 March 2016.

Personally I have never really rated analogue methods but maybe I should take them more seriously.

[VillagePlank]

The data you have there and the result is quite interesting.

It is not that far from either the lrf done by Ed and Tony or the base line model outputs from UK Met?

We will see on 1 March 2016.

Personally I have never really rated analogue methods but maybe I should take them more seriously.

 

It really is just another idea put into the mix. And it really doesn't have any skill if you restrict it to post 1948 data.

[BornFromTheVoid]

Updated my forecast given the unfolding November temperature.

Essentially, this is the analogue method on statistical steroids. I look at anomolies of 2015 from the preceding 30 year average of Jan-Nov of the year, and then compare it to the anomalies of the average for every year in the CET record. The closest match, by the Pearson Correlation Coefficent wins, and then I extrapolate the following Dec, Jan, and Feb from the top five years, and average them.

Assuming Nov will post a CET of 9.5 the years, the closest years are 1994, 1661, 1894, 1913, and 1938, in descending order, registering a coefficient between 0.72 to 0.54 respectively.

Here's the data

data.png

Here's the chart comparing the average of the five top years to what we currently know of the data for 2015,

chart.png

Here's the prediction,

predicted.png

So .. warm December, cold January, average February; essentially a mixed bag.

For interest, the closest modern year, 2009, which came in 6th at a coefficient of 0.53 given the +1.53 November anomaly, and cool July that year. The following Dec, Jan and Feb were all cold and/or much colder than normal each posting anomalies of -1.77 or below (-3, for Jan 2010). But it didn't make the top 5, so wasn't included.

I may well redo this so that the average contains every year that posts a correlation of >= 0.5, which, in this case would be the top ten years so at least years such as 2009/2010 get a 'vote'

 

Cheers for posting this. Using your data, I'd have approached things a little differently. I'd have forecast a milder than average February, seeing as 4/5 of the years were at least 1C above average, and a close to average January as 4/5 were within 1C of the average.

[VillagePlank]

Cheers for posting this. Using your data, I'd have approached things a little differently. I'd have forecast a milder than average February, seeing as 4/5 of the years were at least 1C above average, and a close to average January as 4/5 were within 1C of the average.

 

Yes, completely unweighted. A way of achieving what you suggest is to eliminate events that are outside, say, 1.98stdevs from the average which would elimate the very cold 1894, and thus make February warmer. What I've done is basically very basic and very raw with really no finesse at all - at best I'd suggest the pattern would be close - ie the anomaly extent predicted. I'd be very surprised if the temps turned out what I've predicted on the basis that surely someones tried something like this before etc, and found no skill in it otherwise we'd already all know how to do seasonal forecasting!!

Edited November 28, 2015 by VillagePlank

[VillagePlank]

Adjusted the Nov CET to 9.4 and included all years that scored greater than 0.5, which gives us 1994, 1661, 1894, 1938, 1913, 2009, 1931, and 1821 in descending order with a correlation coefficient ranging from 0.71 to 0.51 respectively (so 1994 is the closest year with a correlation of 0.71)

 

Here's the data,

 

 

Here's the chart,

 

 

Here's the prediction,

 

 

So, average to mild December, average to cool January, and completely average February.

[BLAST FROM THE PAST]

Indeed as RJS says the anticipated main winter thrust is late Jan into Feb.  I have done a more indepth Dec outlook and will do same for Jan [later in Dec] and Feb [later in Jan].  Dec has for me thrown up a reasonable cold blast mid section sandwiched between two mild periods.  This may well bring a CET of nothing special of note either way, but CETs don't tell a full story. 

 

BFTP  

[Nigerian Prince]

December 2015

 

  A  mostly mild and dry to start the month, west to southwest winds dominating, but,  around Christmas we could have a cold spell, dry sunny spells  with light North winds and frosts, turning milder around  the New year Guess Cet 2.5

 

Janauary2016

 

 A mild start, wet, winds mainly from the SW, then around the 10th my guess is that North to northeast winds becoming established given heavy snow showers and severe frost to the North and West of uk, later on to the East and south will be affected by this wintery weather,  has the month goes on the cold will moderate but it will still be dry and sunny and cold  Guess cet -2.5

 

February2016

 

A dry cold start, but it wont last before mid  month there will be a sudden change to mild and increasingly wet and windy weather as several deep depression make their way across western Europe Guess cet 3.5

 

 

Reason for my guess forecast

 

For 2 years now the jet stream has been very strong over western Europe. This summer we did not have a break from the jet stream, my guess is that this will come around January, when we are likely to have a colder spell of weather, cold spells in December are becoming less common now, obviously 2010 was an exception, I think given recent history they are less likely to occur in February, but we can take comfort from 2012 another 200 miles east and we would have had a very colder spell of weather, cold spells of weather in February maybe less common but they are not impossible.

 

What are the professional organisations forecasting.

 

Meteo France are forecasting for a ( l hiver standard,)  average winter with mild southwest winds dominating throughout, they do not rule out a 3 or 4 day cold spell with snow during this period. This is similar to what most long range forecast are forecasting, net weather`s forecast is  interesting, just like mine, they are expecting a cold spell of weather in January and only time will tell if we are right, but I think forecasting record breaking  warmth for February is risky, despite climate change record breaking weather in an given month is not as common as we think.

 

Gender Male, location Battersea London.

[VillagePlank]

Indeed as RJS says the anticipated main winter thrust is late Jan into Feb.  I have done a more indepth Dec outlook and will do same for Jan [later in Dec] and Feb [later in Jan].  Dec has for me thrown up a reasonable cold blast mid section sandwiched between two mild periods.  This may well bring a CET of nothing special of note either way, but CETs don't tell a full story. 

 

BFTP  

 

Yes, forecasting a future CET values ignores the granularity of days/weeks, and also any sort of granularity spatially - so, for instance, the north of the CET area can quite often be quite different from the south of the CET area.

[BLAST FROM THE PAST]

Well the cold middle didn't quite get there, indeed very mild for many as the trough didn't move east and we have had some very mild contrasting with more seasonal way up north.  So we look on to Xmas and the Atlantic coming in with stormy as we approach / get to New Year looking quite feasible.  Now for January.....

Well the way Dec panned out its looking very difficult to get real cold to our shores.  This month could be dire for cold or could introduce some contrasting and interesting battles.  I see a continued set up of a stubborn block being 'within vicinity' but could be too far east or south east and I fear that a 'general' theme of flow from S or SW.  If the block plays ball and interacts a bit more then a cold continental feed could develop.  Apart from an anticipated Atlantic assault mid month bringing potential stormy conditions to N/NW I think a 'rut' month is possible.

Towards very end of month entering into Feb is when I'm looking for change to an assault from the North.  Short and sweet but almost as we were in a way.  I hope this reverse psychology of mild to very mild brings a bitter cold blast to really mess it up

 

Merry Christmas to one and ALL, have a great one, enjoy family and friends as they are not always there

 

regards

 

BFTP 

 

 

[iapennell]

Dear Weather Enthusuiasts;

First of all "Happy Christmas" and all the best for 2016.

I am sorry to say that I have not produced a winter forecast for 2015/16; some of you appreciated the forecasts I produced for 2013/14 and 2014/15 but I have unfortunately not had much spare time over recent months so I have not had so much chance to visit this Forum, let alone produce a detailed forecast for the winter.  I have been working as a full-time commis chef and also studying NVQ Level 2 Accounting; I also produce and send monthly weather reports to the local BBC news (Look North for me near Alston, Cumbria) so I have my hands rather full.  This is why I have not actually produced a detailed weather forecast for this winter.

My predictions for the winter would have led me to conclude a mild and stormy season overall, though not with the record amounts of rain we have seen (leading to record flooding across large parts of the North and ruining hundreds of Christmases!); January 2016 is likely to continue mild and stormy (possibly some snow at times in Scotland).  There will be more short snowy snaps with sharp frost due to Arctic airstreams at times during February; interspersed with further milder wetter spells with strong south-westerly winds.  The El-Nino in the Equatorial Pacific is a record strong one and the QBO (Quasi-Biennial Oscillation) of winds in the equatorial stratosphere is in a westerly phase at the end of 2015; both these factors are supportive of mild stormy winter conditions across northern Britain.  The Sunspot Cycle remains in active phase (as it has been since 2012 and likely to remain so through much of 2016), the interactions from solar flares and interactions with the Earth's magnetic field are supportive too, of strong winter westerlies at higher latitudes.  The Arctic ice-cover is near average extent in the Canadian Arctic but retreated north of its seasonal position in the Eurasian Arctic- this encourages the westerlies to blow further north and (to satisfy Conservation of Angular Momentum laws) to thus blow stronger.

However, as with last year, there remains a sizeable patch of the mid-North Atlantic close to 45N with colder-than-normal waters and with sea-ice west of Greenland further south than usual this is likely to (by February) with sea-ice even further south in that sector to lead to high-pressure in that region at times with surges of cold Arctic air sweeping south along the eastern flank.  Such cold air will reach Britain at times following the passage of some of the (likely) deep depressions we have been getting lots of lately.

I am not at all convinced by the forecast for the latter part of the winter produced by Ed O'Toole and Tony Wells; for Sudden Stratospheric Warming (SST) to occur over the high-arctic there needs to be some predisposition towards weaker mid-latitude jet-streams that will allow upper ridges of warm air to buckle far to the north and most such SSTs it seems there is a pre-requisite for strong upper westerlies to impact hard on the Himalayas or high Pamirs of central southern Asia so that the fast-flowing air is forced upwards over great heights leading to ridges extending north over Siberia and into the Arctic.  Currently the Cuircumpolar Vortex is blowing too strongly and too far north for these developments to unfold.  It is thus my assertion that there will be plenty strong winds and rainfall in January and February 2016 as well, particularly for the North West and Scotland (sorry for folk living in Cumbria and Lancashire) and that cold snaps will come from the north-west (off the Greenland Icecap) in February consequent on colder-than-usual conditions in the mid North Atlantic and Canadian Arctic that, after another month of the prevailing NW winds that have been blowing there this winter, will be especially cold: This, combined with the prevailing SW winds over much of northern Europe until then, will shift the trough in the upper Westerlies east towards the UK and that development is likely to bring about cold snaps.  However, the North Atlantic will have cooled down significantly by early February so with less moisture evaporated from the colder ocean surface there should be rather less rain than of late.

 

Ian 

[iapennell]

On 18/12/2015 at 10:52 AM, chionomaniac said:

I think that's its probably safe to say that one is unlikely before mid January now. Still the reasoning behind it occurring still holds true for later in January, and we are seeing the atmospheric Stratospheric responses that we expected to see. Still feel quietly confident that we will get that SSW eventually though.

It is my take that there wont be any SSW event over the high-Arctic this winter:  The Circumpolar Vortex is extremely strong and it is overall north of its usual position.  The Law of Conservation of Angular Momentum (westerlies need to blow stronger to counteract tropical and polar easterlies if they are in higher latitudes) dictates the Westerlies are likely to be too strong and too far north for pronounced troughs and ridges (with ridges combined with stronger winds hitting the Himalayas that are a pre-cursor to SSW's developing since the upper-ridge is "pushed" up into the Arctic):  Cold weather is likely to come from the North West later in the winter due to cold water in the mid-north Atlantic and Arctic ice being further south than usual off Canada- modest upper troughs (nothing excessive) in the Upper Westerlies will be displaced eastwards towards northern Britain as a result. 

Because of this likely development, in my view, I do not buy into the idea of a quiet end to the winter though February will probably be the coldest month of the winter (though less wet than December or January it will still be wetter than usual).

[RajP]

Paul, a really well written and informative piece. I completely agree with knocker above. If you don't mind me being a little cheeky, may I ask, when you next long term report will be available? as I noticed this piece may have been written a little while ago. Also, it may be worth adding the date/time to the page (and a revision date/time) as weather reports may not include recent data, especially when considering long term articles like this one. I hope this suggestion doesn't translate to criticism and I appreciate you may have reasons for not doing this.

Again, one of the best pieces I've read (and I receive many weather reports).

Many Thanks

RajP

[karyo]

When I read Ian Pennell's comments I thought they would have generated some discussion here but instead there is silence...

Karyo

[Polar Maritime]

Not much to discuss until after Winter, Its just a shame some lrf's were posted over a Month late for whatever reason.. As most were produced in October and released before Winter even started. 

Edited December 30, 2015 by Polar Maritime

[karyo]

17 minutes ago, Polar Maritime said:

Not much to discuss until after Winter, Its just a shame some lrf's were posted over a Month late for whatever reason.. As most were produced in October and released before Winter even started. 

Monthly updates would be good. Why wait till after the winter is over?

[John Badrick]

2 hours ago, karyo said:

When I read Ian Pennell's comments I thought they would have generated some discussion here but instead there is silence...

Karyo

I was thinking the same thing Karyo. He speaks a lot of sense and certainly knows his stuff. He has been quite accurate with his predictions the last couple of years and I for one will be reading and taking interest in what he has to say. The last two winters I haven't liked  his predictions because they have been broadly of mild and stormy weather which we have experienced and sadly it is generally how he thinks the next couple of winters will turn out. But from 2020 (only 4 years time) he thinks the switch will come. I think we Brits need to sit tight and be patient. I do believe, soon we will be experiencing much different winters in this country. I hope he doesn't mind but I wanted to quote something from Ian's email I had recently.

"As a result could well set the scene for deviations in the opposite direction towards extremely severe winters in Britain (i.e. a CET of -5C). "

Be carful what you wish for comes to mind

JB

[karyo]

2 minutes ago, John Badrick said:

I was thinking the same thing Karyo. He speaks a lot of sense and certainly knows his stuff. He has been quite accurate with his predictions the last couple of years and I for one will be reading and taking interest in what he has to say. The last two winters I haven't liked  his predictions because they have been broadly of mild and stormy weather which we have experienced and sadly it is generally how he thinks the next couple of winters will turn out. But from 2020 (only 4 years time) he thinks the switch will come. I think we Brits need to sit tight and be patient. I do believe, soon we will be experiencing much different winters in this country. I hope he doesn't mind but I wanted to quote something from Ian's email I had recently.

"As a result could well set the scene for deviations in the opposite direction towards extremely severe winters in Britain (i.e. a CET of -5C). "

Be carful what you wish for comes to mind

JB

Thanks John, that's useful to know. I do value his opinion. I can't say I am pleased with having to wait till 2020 for a cold winter. Might have to make some winter trips to Scandinavia till then.

Karyo

[lassie23]

I follow iapennell's winter forecasts too and very accurate they have been so far. He should charge for them on the net, instead of the likes of Madden and Corbyn.

[I remember Atlantic 252]

20 hours ago, lassie23 said:

I follow iapennell's winter forecasts too and very accurate they have been so far. He should charge for them on the net, instead of the likes of Madden and Corbyn.

can only agree, I found Ian Brown quite accurate too

[feb1991blizzard]

14 minutes ago, I remember Atlantic 252 said:

can only agree, I found Ian Brown quite accurate too

Yes but that's only because we are going through a period of mild winters, he is James Maddens nemesis, you could say James Madden was pretty accurate during the turn of the last decade, Ian kept denying many a cold spell until it was literally guaranteed, I don't think we will get any deep widespread long lasting cold air conducive to convection before February but that its based on all the evidence available, why would I say that if I didn't believe it (I am a huge cold fan) but there will be another time in the future when I do predict a whole month of cold.

 

JANUARY - TRANSITIONAL MONTH - FEBRUARY AND MARCH COLD.

 

A VERY INTERESTING MONTH OF MODEL WATCHING THOUGH, ESPECIALLY THE DEEPER INTO JAN - DONT BE PUT OFF, THIS WINTER COULD HAVE A BIG STING IN THE TAIL.

Edited December 31, 2015 by feb1991blizzard

[Methuselah]

13 minutes ago, I remember Atlantic 252 said:

can only agree, I found Ian Brown quite accurate too

With the glaring exceptions of 2009-10, 2010-11 and 2012-13, he's been spot on!

[I remember Atlantic 252]

James Madden is rubbish! yes he knows more than me, but many may say I'm correct more than him by saying it's going to be mild mild mild, wet wet wet, low pressure all winter

[Weather-history]

18 minutes ago, Ed Stone said:

With the glaring exceptions of 2009-10, 2010-11 and 2012-13, he's been spot on!

2008-09........

He was spot on with northerly blocking has become extinct.......not.

 

Edited December 31, 2015 by Weather-history

[Methuselah]

2 minutes ago, I remember Atlantic 252 said:

James Madden is rubbish! yes he knows more than me, but many may say I'm correct more than him by saying it's going to be mild mild mild, wet wet wet, low pressure all winter

I agree, I tell my family the same thing...Because, in October (when no-one's got a clue) a mild mushy winter is always statistically far more likely than a repeat of 1963, which - strangely enough - has never once been repeated!