How Cold Has This Winter Season Been...

[snowsure]

I do not wish to be accused of cherry picking the following.

It does seem bizarre to me that the 10 warmest CET years have been since 1989 (except for 1, 1949) link

But when you look at the warmest seasons,

only 3 of the 10 warmest winters have been since 1989.

only 4 of the 10 warmest springs have been since 1989.

only 3 of the 10 warmest summers have been since 1989.

only 3 of the 10 warmest autumns have been since 1989.

rank DJF MAM JJA SON -rank

341 6.13 1935 9.73 1779 16.97 1911 11.43 1995 11

342 6.20 1796 9.73 1961 17.00 1933 11.50 2005 10

343 6.23 1990 9.73 2003 17.03 1947 11.53 1959 9

344 6.33 1686 9.83 1959 17.07 1983 11.53 1949 8

345 6.43 1975 9.87 1952 17.10 1846 11.53 1978 7

346 6.43 2007 9.90 1999 17.23 2006 11.60 1729 6

347 6.50 1989 9.93 1992 17.33 2003 11.60 1818 5

348 6.53 1834 10.07 1945 17.37 1995 11.80 1730 4

349 6.77 1869 10.10 2007 17.60 1826 11.80 1731 3

350 10.20 1893 17.77 1976 12.63 2006 2

link

Does that not suggest that we are having short term "extreme" warming events within a season rather than prolonged warming over a season?

Bit of a head scratcher for me, this one. Comments please...

[Terminal Moraine]

There is more than than one way of looking at this. The last 20 years have produced 3 of the warmest seasons on the CET record and although this doesn't look like many ( compared to the number of warmest years) it is actually 15% of the total.

The last 20 years, expressed as a percentage of the total CET record, is about 5.7% therefore if the warmest seasons were equally spread across the whole record you would expect a 20 year period to produce 1 season in the top ten;

in the event we have a figure almost 3 times that number.

The fact that 9 of the warmest top ten years have occurred in the last 20 years implies a shortage of cold months at any time of year. The warming has been consistent throughout the 12 months, even when not extreme, but with an increased frequency of extreme warm months/seasons thrown in for good measure.

[snowsure]

There is more than than one way of looking at this. The last 20 years have produced 3 of the warmest seasons on the CET record and although this doesn't look like many ( compared to the number of warmest years) it is actually 15% of the total.

The last 20 years, expressed as a percentage of the total CET record, is about 5.7% therefore if the warmest seasons were equally spread across the whole record you would expect a 20 year period to produce 1 season in the top ten;

in the event we have a figure almost 3 times that number.

The fact that 9 of the warmest top ten years have occurred in the last 20 years implies a shortage of cold months at any time of year. The warming has been consistent throughout the 12 months, even when not extreme, but with an increased frequency of extreme warm months/seasons thrown in for good measure.

I agree with para 2 but disagree with para 1.

Of the 20 hottest years, 12 have occured since 1989. This is 60%

Of the 20 hottest winters, 5 have occured since 1989. This is 25%

Of the 20 hottest springs, 8 have occured since 1989. This is 40%

Of the 20 hottest summers, 3 have occured since 1989. This is 15%

Of the 20 hottest autumns, 6 have occured since 1989. This is 30%

How can this have occured? Where is the warming occuring? Is it in the spring? Only 40% of the time since 1989. But 60% of the hottest years have occured since 1989? It must be freak months that do not appear in the seasonal record.

I suggest that these results are counter-intuitive.

edit - inserted underlined text.

Edited February 26, 2009 by snowsure

[Terminal Moraine]

I agree with para 2 but disagree with para 1.

edit - inserted underlined text.

Snowsure, in a rather roundabout way you're agreeing with me, not disagreeing.

In your original post you were stating that the number of top ten warmest seasons in the last 20 years was surprisingly low compared to the number of top ten warmest years over the same period. The figures I came up with indicated the the frequency was actually 3 times what might be expected over a CET period which was static, neither warming or cooling.

In your second post you've altered the criteria to cover the top 20 warmest seasons, rather than the top 10, this increases both the actual number of warmest seasons and the relative percentages over the 20 year period since 1989 and re-inforces the point I was making in paragraph one of my post.

[Essan]

Does that not suggest that we are having short term "extreme" warming events within a season rather than prolonged warming over a season?

No, it suggests the opposite. Instead of short lived monthly or seasonal extreme, we're expriencing steady (but not exceptional) year round warming.

How can this have occured? Where is the warming occuring?

It's happening all year round

The warmest year on record might easily have no month or season that's actually warmest on record. And years with the warmest month or season might easily still be fairly normal when averaged over the whole year.

So, why are all months and seasons getting warming?

Edited February 26, 2009 by Essan

[SP1986]

The winter has been different in different parts of the country.

Here the relative coldness has been as much as the relative cold in the south or the Midlands, but the actual cold has not been as severe due to this area being a warmer climate - this may help explain the fact that we are in a warming age still.

Relatively it's still warming, if you take the long term average, it's still above this so it's still warming over the last few years it's just slowed down (but it's still warming).

Everything here is relative.

If it wasn't for external causes this winter would probably have been a lot colder than it was.

[Yeti]

I cannot believe that no one has pointed out yet the fact that we're looking at 20 years out of 350.

So let's look at those stats again:

only 3 of the 10 warmest winters have been since 1989.

So that's 3 out of 20 years as opposed to 7 out of 350! To do the maths:

Let's say that X is the probability of one of those exceptionally warm winters being produced. That means X = 7/350 = 0.02 (2% chance of a winter like that being produced before 1989).

Therefore out of 100 winters you would expect 3 or 4 to be exceptionally mild, in this case such that it could become one of the "top ten".

However we have got 3 in 20 years! So to work out the probability of this:

²⁰C₃ x (0.02)³(0.98)¹⁷ = 0.006

So the probability of the recent warm winters occurring completely by chance is just 0.6%, as opposed to 2% BEFORE 1989. That's very significant.

A good analogy would be watching cars for a day. Imagine if all the colours were evenly spread until the last 10 minutes of observation, at which point 20 red cars come past. You could say that this 20 is less than the 100 you've seen all day, but that doesn't mean there hasn't been a huge increase in the number of red cars passing!

[Grimsby Snow Lover]

One thing that has suprised me this year (past 2 weeks excluding) is the amount of days where the frost has stuck in shaded areas. Even here on the coast we have had this happening. We have has more air frosts this season than I can ever remember since I moved here 5 years ago. I also was suprised with the amount of days in December, January and February where we had low single figure temperatures with very few days that got into double figure.

[Tucco]

One thing that has suprised me this year (past 2 weeks excluding) is the amount of days where the frost has stuck in shaded areas. Even here on the coast we have had this happening. We have has more air frosts this season than I can ever remember since I moved here 5 years ago. I also was suprised with the amount of days in December, January and February where we had low single figure temperatures with very few days that got into double figure.

Coldest winter for a decade according to the Meto

http://www.metoffice.gov.uk/corporate/pres...pr20090225.html

Mark

Teesdale,Co Durham

[Snowyowl9]

Coldest winter since 1995/96 even with this mild/very mild 10 days/2weeks end of feb.

Looks like it`ll be the 10th coldest in the last 40 years now.

[Tucco]

Coldest winter since 1995/96 even with this mild/very mild 10 days/2weeks end of feb.

Looks like it`ll be the 10th coldest in the last 40 years now.

7 winters in the 60's were colder than this one :o

I think winter 2000/01 was colder in Scotland and I will be very close across far north of England compared to this one.

Mark

Teesdale,Co Durham

[Thundery wintry showers]

In Cleadon, where I take my weather records, December had a mean temperature of 4.3C and January had 3.7C. February looks like coming in at around 4.3C as well, which would make the mean 4.1C (estimated). This will make it the coldest winter since 1995/96, though only marginally colder than 1996/97 and 2000/01. Also the winter of 1993/94 had a similar mean temperature.

However, so far the season of 2008/09 it has produced only 10 days of lying snow, behind 1993/94 (18 days), 1995/96 (18 days), 2000/01 (13 days), 2003/04 (11 days) and level with 2005/06 (10 days). It would take an exceptionally wintry spring to bring it even close to 1993/94 or 1995/96.

However, if you narrow it down to just the winter quarter it comes out more notable snow wise, with 9 days of lying snow- behind only 1993/94 (11 days), and 1995/96 (15 days). It's worth noting, though, that the mild winters of 1998/99 and 2003/04 produced 8 and 9 days respectively, adding a bit of perspective.

The fact that 7 winters in the 1960s were colder puts it into perspective. We did have milder winters in the 1900s through to 1930s, but on the other hand I reckon many of the 1940s and 1950s winters were also colder than this one- and certainly, averaged natioinally, a lot snowier.

Edited February 26, 2009 by Thundery wintry showers

[johnholmes]

The fact that 7 winters in the 1960s were colder puts it into perspective. We did have milder winters in the 1900s through to 1930s, but on the other hand I reckon many of the 1940s and 1950s winters were also colder than this one- and certainly, averaged natioinally, a lot snowier.

for most on here, yourself included Ian, equating this winter to the 60's is a bit like me equating my memory of, say, 1963 or 1947 to 50 years before that. Neither you nor I have the memory of those periods so why compare.

This winter is the coldest and for some the snowiest in at least 10 years perhaps nearer 20 years.

nor is it really true that the 60's were full of cold winters.

to add to that last comment

between Dec 1960 and Feb 1969, 27 winter months, and using the CET data, and the 61-90 averages as being nearer to that period than 1970-2000;

27 months showed 13 above average and 14 below average.

the longest 'cold' run being Feb 62-Feb 65, inclusive.

and another bit, re the 7 'cold' winters

nearly with 6 showing an average below the 61-90 winter average of 4.1C

61-62

62-63

63-64

65-65

67-68

68-69

Edited February 26, 2009 by johnholmes

[Duncan McAlister]

Snowier, but less frosty, than 2005-06. Better than the average for this decade but nothing special really, no great depths of snow and never lasting longer than a day.

[I remember Atlantic 252]

this feb probably the mildest since 2004 on my temps 13 out of the last 14 days in double figures, the month with the most double figure temps since i started my temps dec 2004

[snowsure]

Thanks for comments so far. Any statistical consideration always seems, to me, a bit wooly. It talks about what you might expect. I prefer more empirical evidence.

viz.

Consider 2006 (the hottest CET year.)

The Autumn was the 2nd warmest in 350 years.

The Summer was the 6th warmest.

The Spring was the 111th warmest

The Winter was the 147th warmest.

Seems a little uneven, doesn't it?

The 10 warmest years, and their respective seasons position (1 warm, 350 cold) line up like this:

Year.............Winter...Spring....Summer...Autumn

2006..............147.......111............6...........2......

1990................9..........14...........56..........43.....

1999................37..........6...........78...........13....

1949................27.........91..........31.............8....

2002...............42..........20...........93...........23...

1997...............193.......13.............26...........27..

1995................14.........89.............4...........11..

2003................87..........9..............5...........63..

1989.................5..........58............35..........31..

2004................53.........32............47..........21..

It does not appear conclusive that all year round warming is taking place. If it were, then surely the top 10 warmest years would have many seasons that were in the top 10 warmest. Of the 40 seasons, 9 were in the top 10.

9/80=11%

I'm sure that someone could show how this looks on a distribution curve but it feels wrong that warming is happening all year round.

The 10 warmest years have had a range of winters that goes from 5th warmest to 193 warmest.

Range of warmest seasons :

..Winter.....Spring.....Summer.....Autumn

..5-193......9-111.........4-93.........2-63...

Can it be concluded that the warming is certainly autumnal (range=61years) then summer (range=89) then Spring (range=102) then Winter (range=189).

Ultimately, if this is the real fingerprint of our warming (ASuSpW), which of the 10 warmest years had this fingerprint?

1 - 2006

That is the only one.

Just trying to look at this in a different way. I have framed a hypothesis, that is all.

[Terminal Moraine]

Thanks for comments so far. Any statistical consideration always seems, to me, a bit wooly. It talks about what you might expect. I prefer more empirical evidence.

It does not appear conclusive that all year round warming is taking place. If it were, then surely the top 10 warmest years would have many seasons that were in the top 10 warmest. Of the 40 seasons, 9 were in the top 10.

This isn't necessarily the case. July 2006, for instance, was the warmest month on the entire CET record but to achieve that status it didn't mean that most of the days in the month were the warmest on record, or even in the top 10 warmest on record.

In the same way the warmest years don't necessarily have to claim the lion's share of the warmest seasons or even the warmest months; it's the consistency of the warmth which counts, combined with a lack of cold to counterbalance the warmth when the averages are calculated.

If we reach the point where the top few warmest years contain, say, 60 or 70% of the warmest seasons on record we'll probably be looking at years coming in around 1- 2c warmer than the current warmest.

[Eugene]

I think we have seen the peak of the warming(late 90`s) and we are now on a steady downturn, i expect the next 20 years to be alot cooler overall than the past 20 years.

[johnholmes]

Thanks for comments so far. Any statistical consideration always seems, to me, a bit wooly. It talks about what you might expect. I prefer more empirical evidence.

viz.

I'm not really sure how you can say that 'any statistical consideratiom...a bit woolly' ?

Statistics is a pretty exact form of Mathematics and is used in both old data, you call it empirical, and in forecast data, to show how realistic or how likely something is to occur.

[Thundery wintry showers]

for most on here, yourself included Ian, equating this winter to the 60's is a bit like me equating my memory of, say, 1963 or 1947 to 50 years before that. Neither you nor I have the memory of those periods so why compare.

I'm thinking statistically- I wasn't around during this earlier period but I do have access to the stats, plus some were around at that time. You could use that as an argument to say that I shouldn't compare to the 1978-87 period either, as the first weather event I can even vaguely remember is the deep snow in South Shields in February 1991. I also did an analysis on winter snow events going back to 1900 using a combination of sources (Trevor Harley, Brocanica's snow events, Wetterzentrale, Philip Eden sources) and as with many other sources found a decline in recent years, though I was surprised to see that the 1900s through 1930s were generally less snowy than the mid-20th century.

However, it is fair to say that this winter, averaged nationally, probably beats anything since 1995/96 over England and Wales (and locally back to the 1980s) and 2000/01 in Scotland.

Re. comments on year-round warming, actually, we would expect annual temperatures to break more records than monthly temperatures. This is because monthly values are subject to a much higher standard deviation than annual values. Therefore, it would require much less warming to take mean annual values above the standard deviation for, say, 1971-2000, than for monthly values.

[PB22]

Its been one of the coldest winters that I can remember, and I certainly cannot recall seeing as many frosts as we have had this season in the last 25 years or so. Hardly any snowfall though despite the colder winter. Rather than any really extreme cold spells its notable that its generally been consistently on the cold side rather than two or three intense cold snaps.

[Guest North Sea Snow Convection]

I think we have seen the peak of the warming(late 90`s) and we are now on a steady downturn, i expect the next 20 years to be alot cooler overall than the past 20 years.

Agreed.

This winter in relation to expectations in some quarters and certainly against recent winters has been a big turnaround. I would have hoped to have been rubber stamping that point much more as seemed very likely a couple of weeks ago, but the lowpoint of the winter has been reserved right for the end in this last fortnight.

Nonetheless, even taking into account a Feb that is disappointingly (against overall 08/09 winter form) going to reach around 'par' it has shown that a cold winter, overall, is achievable with ease when you take into account the moderately cold synoptics on offer, ie with not a single high latitude block on offer. I think the consistently sustained below average temps relative to these synoptics is the best thing of all to take from the winter as a whole, and clearly shows that a severe winter could still be very much on the cards when deep winter months do coincide with the coldest synoptics. And sooner or later they will.

Edited February 27, 2009 by North Sea Snow Convection

[snowsure]

I'm not really sure how you can say that 'any statistical consideratiom...a bit woolly' ?

Statistics is a pretty exact form of Mathematics and is used in both old data, you call it empirical, and in forecast data, to show how realistic or how likely something is to occur.

Regarding the use of statistics in forecasting; It would appear there is a 62% snow risk, on Wednesday, in Doncaster. Does that mean that, given the synoptics it would snow 62 times out of 100? Or does it mean something else?

Perhaps this is for another thread? Anyone suggest some reading for me?