Autumn And Winter

[summer blizzard]

In recent years, many people have touted the warm September/cold October combination as the harbringer of doom for winter, so i decided to see if a warm September with a CET of 1C above average or more (14.7C+) does encourage a cold October. Here are the results since 1950, the CET values are the following October value and the following winter CET...

1958: 10.8C - 0.4C above average - 3.6C - 0.9C below average

1959: 12.6C - 2.4C above average - 4.6C - 0.1C above average

1961: 10.9C - 0.5C above average - 3.6C - 0.9C below average

1980: 9C - 1.4C below average - 4.5C - Average

1989: 11.7C - 1.3C above average - 6.2C - 1.7C above average

1991: 10.2C - 0.2C below average - 4.6C - 0.1C above average

1998: 10.6C - 0.2C above average - 6.1C - 1.6C above average

1999: 10.7C - 0.3C above average - 5.4C - 0.9C above average

2000: 10.3C - 0.1C above average - 4.5C - Average

2004: 10.5C - 0.1C above average - 5.2C - 0.7C above average

2005: 13.1C - 2.7C above average - 4.1C - 0.4C below average

As you can see, a September CET that is 1C above average actually favours an October CET of around 10.9C which is slightly above average and also favours a winter CET of around 4.2C, which is slightly below average.

So we know that a mild September actually favours a slightly above average October and slightly below average winter, however i wanted to see if a cold October (9.4<) indicated a mild November and winter period as a whole ahead, here are the results since 1950, figures on the left are the following November CET, figures on the right are the following winter CET...

1951: 8.5C - 1.6C above average - 3.9C - 0.6C below average

1952: 4.2C - 2.7C below average - 3.5C - 1C below average

1955: 7C - 0.1C above average - 2.9C - 1.6C below average

1956: 6C - 0.9C below average - 5.5C - 1C above average

1964: 7.4C - 0.5C above average - 3.3C - 1.2C below average

1973: 6C - 0.9C below average - 5.4C - 0.9C above average

1974: 6.8C - 0.1C below average - 6.4C - 1.9C above average

1980: 6.6C - 0.3C below average - 4.5C - Average

1981: 7.8C - 0.9C below average - 2.6C - 1.9C below average

1992: 7.4C - 0.5C above average - 4.7C - 0.2C above average

1993: 4.6C - 2.3C below average - 4.7C - 0.2C above average

2003: 8.1C - 1.2C above average - 5.1C - 0.6C above average

As you can see from the statistics above, an October with a CET of 9.4C or less favours a November CET of around 6.5C, which is slightly below average and a winter CET of around 4.4C, which is slighly below average.

Based on the data above, i would say that a Mild September/Cold October actually favours a slightly below average winter CET.

Certainly, the winter before last, the Met Office suffered dreadfully from Metcheck's erroneous forecast of a cold winter. The Met Office did NOT forecast a colder than average winter, that year but were roundly criticised, come the spring, for having done so!! The press got metcheck and the Met Office confused. They were really very upset about it and with justification!

Last year the Met Office forecast the coldest winter for a decade and were correct. I think the metcheck forecast was for colder than that.

Paul

The Met Office were actaually wrong, aside from a small segement along the south coast of England, the NAO value was 0.1, neutral not negative and nationally, winter 2006 was the coldest winter since 1997, not 1996, so close but no cigar.

[Stratos Ferric]

Based on the data above, i would say that a Mild September/Cold October actually favours a slightly below average winter CET.

The Met Office were actaually wrong, aside from a small segement along the south coast of England, the NAO value was 0.1, neutral not negative and nationally, winter 2006 was the coldest winter since 1997, not 1996, so close but no cigar.

Not sure I agree with either of the above SB.

Haven't got time to carve the data up, but I'm not sure that there's a real pattern there at all. Averaging is well and good if there's a genuine trend, but I suspect what you have is a lot of noise behind some generalisations, these latter quite possibly being largely factored to general background CET at the time (i.e. was the climate in a warmer or cooler phase).

Re the MO forecast. If ever there's demand for a hair splitting tool, I suggest you give up your long teleconnections forecasts and go invent it. In terms of outcome the MetO were pretty well on the money last year, and were certainly the best forecast I saw. The other skill they have is to keep forecasts at a level of detail that aligns with ability to forecast (at all) at that distance. Some of the ones we get on here that talk about events and dates are essentially random guesses, for all that the individuals concerned may try to wrap them up in method. The imprecision of LRFs is best reflected in forecasts that are general rather than specific; of course, there can be different reasons for a particular avergae outcome, so passing mention to type of synoptics is a helpful arbiter between "right guess for the wrong reasons" and "good guess based on good guess re synoptics".

[Michael Fishermans Friend]

'The Met Office were actaually wrong, aside from a small segement along the south coast of England, the NAO value was 0.1, neutral not negative and nationally, winter 2006 was the coldest winter since 1997, not 1996, so close but no cigar.'

Agreed, the average or mean temp for winter 2005/2006 here was about 0.1c below average - generally making it an 'average winter' and the coldest since the 1996/97 winter, which says it all about the winters since then. In these parts we have not experienced an 'iceday' since January 8th 1997 all though there have been a few close shaves.

Edited September 7, 2006 by Timmy H

[summer blizzard]

Could somebody post a link to the Scottish Met Office, if i remeber correctly, March 2006 was the coldest since 1947.

[Optimus Prime]

March 2006 in Scotland wasn't the coldest since 1947. It had a CET of 2.5 which is 1.4c below average. Others;

2001; 2.5c (-1.4)

1996; 2.9c (-1.0)

1995; 2.4c (-1.5)

1987; 2.3c (-1.6)

1970; 2.1c (-1.

1969; 1.5 (-2.4)

1962; 1.1 (-2.

1947; 0.2 (-3.7)

March 2006 was the 17th coldest in Scotland with records dating back to 1914.

[Guest Viking141]

March 2006 in Scotland wasn't the coldest since 1947. It had a CET of 2.5 which is 1.4c below average. Others;

2001; 2.5c (-1.4)

1996; 2.9c (-1.0)

1995; 2.4c (-1.5)

1987; 2.3c (-1.6)

1970; 2.1c (-1.

1969; 1.5 (-2.4)

1962; 1.1 (-2.

1947; 0.2 (-3.7)

March 2006 was the 17th coldest in Scotland with records dating back to 1914.

Interesting stuff Optimus tks for that. As far as Shetland goes March 2006 was the coldest since 1969 and had a mean max temp of 3.9C, some 2.4C below the average with snow/sleet on 15 days (in fact snow/sleet was the only type of precip during March - there was no rain at all!!)

[Snowyowl9]

In recent years, many people have touted the warm September/cold October combination as the harbringer of doom for winter

As you can see, a September CET that is 1C above average actually favours an October CET of around 10.9C which is slightly above average and also favours a winter CET of around 4.2C, which is slightly below average.

I`ve noticed the the 10 years or so that if we have a cold october that december is relatively mild so I`ll put the cet down for october and december.

1995 oct12.9 - Dec 2.3

1996 11.7 - 2.9

1997 10.2 - 5.8

1998 10.6 - 5.8

1999 10.7 - 5.0

2000 10.3 - 5.8

2001 13.3 - 3.6

2002 10.1 - 5.7

2003 9.2 - 4.8

2004 10.5 - 5.3

2005 13.1 - 4.4

It`s just something I`ve noticed over the recent years.

And of course last october was the warmest of them all and the winter was the coldest for 9 years,it doesn`t mean it`ll happen this time but it`s 11-1 to those figures.

S9

Edited September 7, 2006 by Snowyowl9

[North-Easterly Blast]

Thank you very much summer blizzard for your data on warm Septembers and the Octobers and winter CET. I have certainly noticed even looking further back that a warm September was very rarely followed by a particularly cold winter (below 3.5 CET). To further your contributions - how about a list of warm Octobers (11*C+) and the following winters CET? And also what the CET was for the preceding September? My view is that in any autumn if October is a warm month, but September is not, then you do stand the greatest chance of a cold winter.

[The Enforcer]

Well, I've been suggesting for the best part of two weeks now that a settled spell is likely "next week" (lol). The longer one type of weather persists (after two days), the greater the likelihood of a change.

Trend correctly identified at a range of T+504. Ever thought of becoming a long-range weather forecaster?

IBETHEISALREADYONEONTHESIDESMILEY

[Snowyowl9]

This was exceptional in October 1974 it was very cold at 7.8c and December was warmer at 8.1 that`s the only time that there was a warmer December than october

1988 was the next mildest december at 7.5c

October was 10.4c

This was also incredible dec 1981 just 0.3c and october was very cold at 8.6c so it means nothing that a cold october brings a mild december.

S9.

Edited September 8, 2006 by Snowyowl9

[summer blizzard]

I`ve noticed the the 10 years or so that if we have a cold october that december is relatively mild so I`ll put the cet down for october and december.

1995 oct12.9 - Dec 2.3

1996 11.7 - 2.9

1997 10.2 - 5.8

1998 10.6 - 5.8

1999 10.7 - 5.0

2000 10.3 - 5.8

2001 13.3 - 3.6

2002 10.1 - 5.7

2003 9.2 - 4.8

2004 10.5 - 5.3

2005 13.1 - 4.4

It`s just something I`ve noticed over the recent years.

And of course last october was the warmest of them all and the winter was the coldest for 9 years,it doesn`t mean it`ll happen this time but it`s 11-1 to those figures.

S9

While i would argue that the dataset is too small, of the five below average Octobers, only three were above average as opposed to two, the only truly cold October was 2003, which was followed by a below average December.

I will put up the warm October statistics later tonight ot tommorow.

[Dawlish]

The Met Office were actaually wrong, aside from a small segement along the south coast of England, the NAO value was 0.1, neutral not negative and nationally, winter 2006 was the coldest winter since 1997, not 1996, so close but no cigar.

You are absolutely spot on Summer! In both instances. They did issue the best forecast of all the major players, but you are right - they were not technically correct!

That should, now, leave their NAO hindcast forecast at less than 66% correct, as 66% correct was the figure they claimed before last winter and the figure they still claim on their site after last winter! It should be about 64% correct, as they appear to have a 36 year dataset. Last year's NAO forecast was just outside the confidence limits that they set themselves.

My own view is that less than 2 out of 3 correct with their NAO forecast isn't good enough to leave it so public, though they do preface it with experimental, in big red letters - a fact that is still lost on many people that comment after visiting the site and on all hacks. Last winter, this experimental forecast shook the whole country after Amicus went public with their recommendations to their members to begin preparations for a colder winter, after Met Office advice to local councils and the hacks who interpreted this, then exaggerated it out of all proportion.

The sound of very wet, brown, stuff being blasted off the fan could be smelt by the whole country!

Paul

Edited September 8, 2006 by Dawlish

[Weather-history]

I posted this last year

Does a mild, dry October increases the chances of the following winter being colder than average? I let you decide

1995 Oct 12.9 Rain 52.0mm --------> winter 1995-96 below average

1985 Oct 11.0 Rain 46.7mm --------> winter 1985-86 below average

1978 Oct 11.9 Rain 17.2mm --------> winter 1978-79 severe

1969 Oct 13.0 Rain 17.5mm --------> winter 1969-70 below average

1965 Oct 11.0 Rain 33.9mm --------> winter 1965-66 close to average notable cold spells

1908 Oct 11.8 Rain 56.2mm --------> winter 1908-09 slightly below average

1861 Oct 11.8 Rain 48.2mm --------> winter 1861-62 close to average

1809 Oct 10.2 Rain 10.9mm --------> winter 1809-10 slightly below average#

1806 Oct 10.6 Rain 41.9mm --------> winter 1806-07 slightly above average#

1793 Oct 11.3 Rain 52.9mm --------> winter 1793-94 fairly mild#

1790 Oct 10.3 Rain 56.9mm --------> winter 1790-91 fairly close to average#

1781 Oct 10.6 Rain 8.8mm ---------> winter 1781-82 fairly close to average#

1776 Oct 10.2 Rain 49.6mm --------> winter 1776-77 fairly cold#

1774 Oct 10.3 Rain 40.9mm --------> winter 1774-75 fairly mild#

# For that time period those Octobers could be considered mild.

The results are fairly mixed to be honest, however there is no especially mild winters (>5 CET) in those stats

[Guest Mike W]

I noticed though Mr Data that regarding the 20th Century October's you have listed, only one September that preceeded the October was 14+, which is Sept. 1985. So generally by the looks of it you also should really have a sub 14 CET for September aswell as having a mild and dry October, although these days especially, this would not be enough as 2001 has shown.

Edited September 8, 2006 by Mike W

[Weather-history]

This was exceptional in October 1974 it was very cold at 7.8c and December was warmer at 8.1 that`s the only time that there was a warmer December than october

Even more remarkable is that the first half of October 1974 was 0.5C colder than the second half of December 1974

As for the warm September - mild winter theory, this is what I posted last year

September of 15.0 or greater and mild winter theory

September

2005 15.2 Winter CET: 4.1

1999 15.6 Winter CET: 5.4

1961 15.2 Winter CET: 3.6

1958 15.1 Winter CET: 3.6

1949 16.3 Winter CET: 5.1

1929 15.3 Winter CET: 4.6

1898 15.2 Winter CET: 5.8

1865 16.3 Winter CET: 5.3

1825 15.1 Winter CET: 3.8

1795 16.0 Winter CET: 6.2

1780 15.6 Winter CET: 3.4

1779 15.2 Winter CET: 1.4

1760 15.7 Winter CET: 5.8

1750 15.2 Winter CET: 3.2

1731 15.3 Winter CET: 4.7

1730 15.3 Winter CET: 2.5

1729 16.6 Winter CET: 4.6

1708 15.0 Winter CET: 1.2

1678 15.0 Winter CET: 1.0

Winters with CET <2: 3

Winters with CET between 2 and 3: 1

Winters with CET between 3 and 4: 5

Winters with CET between 4 and 5: 3

Winters with CET between 5 and 6: 5

Winters with CET >6: 1

So there's 9 winters with a CET of less than 4C and 9 winters with a CET of greater than 4C

So does a September with a CET>15C lead to a mild winter?

Well it depends from which year you are observing from, in recent times you would say it lead to a mild winter but pre 1760 you could say it favoured a cold winter

Inconclusive, I would say

[Dawlish]

Even more remarkable is that the first half of October 1974 was 0.5C colder than the second half of December 1974

As for the warm September - mild winter theory, this is what I posted last year

September of 15.0 or greater and mild winter theory

September

2005 15.2 Winter CET: 4.1

1999 15.6 Winter CET: 5.4

1961 15.2 Winter CET: 3.6

1958 15.1 Winter CET: 3.6

1949 16.3 Winter CET: 5.1

1929 15.3 Winter CET: 4.6

1898 15.2 Winter CET: 5.8

1865 16.3 Winter CET: 5.3

1825 15.1 Winter CET: 3.8

1795 16.0 Winter CET: 6.2

1780 15.6 Winter CET: 3.4

1779 15.2 Winter CET: 1.4

1760 15.7 Winter CET: 5.8

1750 15.2 Winter CET: 3.2

1731 15.3 Winter CET: 4.7

1730 15.3 Winter CET: 2.5

1729 16.6 Winter CET: 4.6

1708 15.0 Winter CET: 1.2

1678 15.0 Winter CET: 1.0

Winters with CET <2: 3

Winters with CET between 2 and 3: 1

Winters with CET between 3 and 4: 5

Winters with CET between 4 and 5: 3

Winters with CET between 5 and 6: 5

Winters with CET >6: 1

So there's 9 winters with a CET of less than 4C and 9 winters with a CET of greater than 4C

So does a September with a CET>15C lead to a mild winter?

Well it depends from which year you are observing from, in recent times you would say it lead to a mild winter but pre 1760 you could say it favoured a cold winter

Inconclusive, I would say

Thanks Mr. D.

From your stats banks, have you any instances of any good indicators of a season being presaged by a particular CET in a previous month, or season? I'm sceptical of there being any, as you've probably read, but I'd certainly be interested if there was/were, especially if it was a fairly large dataset and the outcome was 75% agreement, or greater.

Paul

Edited September 8, 2006 by Dawlish

[Big Bear]

I really don't think an average to high September CET means, "right, thats it - Winters written off"

It's a case of roll the dice come December and see what pans out.

[Weather-history]

From your stats banks, have you any instances of any good indicators of a season being presaged by a particular CET in a previous month, or season?

Paul

Not a season Paul but a month. Its a stats quirk that makes you wonder.

A June with a CET of 16 or greater does not favour a mild December

There have been 26 Junes with a CET of 16 or greater, 12 have been followed by a December of less than 3C

21 have been followed by a December that was below the 1961-90 average

24 have been followed by a December that was below the 1971-00 average

Only 2 were followed by a December that had a CET of above 5C: 5.4 (1781) and 5.8 (1826)

----------------------------------------------------------------------------------------------------------------------

One thing I would add though is that a severe winter (CET <2) does not favour a particularly warm summer. Only 3 summers that followed a winter of less than 2C (there are 36 of them) had a CET of 16.0+C

Edited September 8, 2006 by Mr_Data

[Coopsy]

Interesting stat that one - that 24 out of 26 Decembers have been below average if June is above 16oC.

[Dawlish]

Thanks Mr. D. You really are a mine!

Is it really is a quirk.......or have you found the holy grail! This winter will be a cold one because of June's CET being >16C. The stats add up, it is just whether the connection is spurious in the greater scheme of things!

Like you say - it does make you wonder!

Paul

[shuggee]

Hold on - the stats say December will (is more likely to) be colder than the 1971-2000 average, not the winter Paul

[summer blizzard]

Interesting CET developments, it seems that in recent times, a mild October indicates a below average winter and the June CET favours a below average December.

[Snowyowl9]

Even more remarkable is that the first half of October 1974 was 0.5C colder than the second half of December 1974

Wow that must of been a cold 1st half of october 1974,for the 2nd half of december to be 0.5c warmer is unprecidented,I was around then but a little to young to appreciate it then

Interesting CET developments, it seems that in recent times, a mild October indicates a below average winter and the June CET favours a below average December.

What will decide this once and foreall is which is alot to ask for anyone to work out, what the ratio factor is for all the years from say 1900 to the presant day, if a cold october brings a mild winter as far as the June thats interesting but I`m only interested in october.

We`ll know if its a pattern or not or rather even.

S9.

Edited September 8, 2006 by Snowyowl9

[Snowyowl9]

That would explain the contrasts between the 2 months,my mum remembers that december 74 she said it was that mild you could go out in a tee-shirt.

S9

[Weather-history]

Junes with a CET of 16C and the following December CET

1846 Jun 18.2 Dec 0.5

1676 Jun 18.0 Dec -0.5

1826 Jun 17.3 Dec 5.8

1822 Jun 17.1 Dec 1.6

1976 Jun 17.0 Dec 2.0

1762 Jun 16.9 Dec 3.6

1798 Jun 16.9 Dec 1.5

1858 Jun 16.8 Dec 4.8

1775 Jun 16.6 Dec 4.5

1726 Jun 16.4 Dec 1.8

1728 Jun 16.4 Dec 1.6

1818 Jun 16.4 Dec 3.6

1940 Jun 16.4 Dec 3.8

1970 Jun 16.4 Dec 4.3

1781 Jun 16.2 Dec 5.4

1896 Jun 16.2 Dec 3.9

1950 Jun 16.2 Dec 1.2

1772 Jun 16.1 Dec 4.8

1785 Jun 16.1 Dec 2.8

1786 Jun 16.1 Dec 2.8

1804 Jun 16.1 Dec 2.1

1960 Jun 16.1 Dec 3.9

2003 Jun 16.1 Dec 4.8

1672 Jun 16.0 Dec 4.5

1683 Jun 16.0 Dec 0.5

1707 Jun 16.0 Dec 3.5

So for a June with a CET of 16.0 and greater the CET average for the following December is about 3.0C

Interestingly for Junes with a CET of less than 13C, the CET average for the following December is 4.1C