Annual Cet - 2010

[Lettucing Gutted]

One of the reasons why its easier to get a large negative anomaly in winter compared to summer is the wider range of sources of cold air in winter compared to summer. In winter you can get cold Continental Polar air from the Arctic, cold Maritime Polar air from the Atlantic during Cold Zonality and cold Continental Polar air from the Continent. However in summer you can only get cold Maritime Polar air from Atlantic Cold Zonality and Continental Polar air from Arctic originated Northeasterlies, Northerlies and Northwesterlies. The Continental Easterlies always bring warm, humid and sunny weather in the summer rivaling southerly and southwesterly warm, humid and sunny plumes.

[AderynCoch]

It is easier to record a negative anomaly for winter months than it is with summer months. If you look at all the months with the largest negative anomalies, they will all be winter months and therefore winter will be the season with the largest negative anomalies.

For instance, January 2010 had an anomaly of -2.4C to the 1961-90 average. When was the last time that happened with a summer month of that magnitude? Even June 1991 was just 2.0C below the 1961-90 average and it is one of the coldest Junes on record.

Which is why I prefer to compare different months by their ranking in the CET series.

Taking January 2010 as an example (joint 64th coldest with a long-term anomaly of -1.8C): it can be compared with a June averaging 13.4C (anomaly of -0.9C), a July averaging 15.0C (anomaly of -1.0C) and an August averaging 14.7C (anomaly of -0.9C).

The coldest July on record (1816 with 13.4C) has a -2.6C deviation from the long-term average - a January which came in -2.6C below average (0.6C) would only be the 37th coldest in the CET series. The coldest January (1795, with a CET of -3.1C) is a whopping -6.3C below the long-term average.

Edited February 28, 2010 by AderynCoch

[Weather-history]

Highest CET daily mean of the year up to the 1st of March is just 6.7C!

Years with a "lower" highest CET daily mean of the year at this stage

1979: 5.4 (7.0 by 2nd of March)

1963: 2.4 (8.9 by 6th of March)

1895: 4.6 (7.2 by 16th of March)

1814: 6.4 (7.1 by 22nd of March)

Shows how little mild there has been.

[AderynCoch]

Are they the only years dating back to 1772? Impressive!

1979 is about to be beaten, and likely 1963 too.

Surely not 1814 and 1895?

[kold weather]

Yep quite amazing really, certainly shows just how dominating the -AO has been with everything being shunted southwards so even when we don't have cold spells temps are still lower then you'd expect in the synoptics...

Anyway with Feb coming in at 2.8C, thats another 1.4C down on the 71-00 average...I suspect March won't decrease the -ve anomaly we've gained from Jan and Feb much if at all...

So we need around +0.7C months from now onwards to reach 10C by the end of the year and thats without March possibly coming in below average...so odds are decent for now of a sub 10C year.

Also FWIW IF March came out average and the rest of the year matched 2006 we'd end up at 10.6C, so thats the utter top limit IMO we can now reach unless we smash some records...

[Stu_London]

CET 01/03/09-28/02/10 - 9.87C

Has potential to fall further too if the cold/cool theme continues as up against a quite mild spring in 2009. I would say an annual in running CET of 9.4C or 9.5C is possible at some stage.

[Lettucing Gutted]

So we need around +0.7C months from now onwards to reach 10C by the end of the year and thats without March possibly coming in below average...

The last year that didn't produce any months 1.0C or more above their respective CET average was 1996.

[kold weather]

Yeah thats true Craig and I'm pretty confident Summer/Autumn will give a couple but even so any average/below average months from now on is only going to make that figure grow quite alot!

For us to reach the record now we are going to need most of your CET predicitions to come off!

[Weather-history]

Highest CET daily mean of the year up to the 1st of March is just 6.7C!

Years with a "lower" highest CET daily mean of the year at this stage

1979: 5.4 (7.0 by 2nd of March)

1963: 2.4 (8.9 by 6th of March)

1895: 4.6 (7.2 by 16th of March)

1814: 6.4 (7.1 by 22nd of March)

1979 is off the list now, 1963 looks likely to fall

[SteveB]

Mr D,

Has there been any years that have started of this low but ended up above 10c?

If March comes in below average, then the same question applies again.

[stewfox]

Highest CET daily mean of the year up to the 1st of March is just 6.7C!

Years with a "lower" highest CET daily mean of the year at this stage

1979: 5.4 (7.0 by 2nd of March)

1963: 2.4 (8.9 by 6th of March)

1895: 4.6 (7.2 by 16th of March)

1814: 6.4 (7.1 by 22nd of March)

1979 is off the list now, 1963 looks likely to fall

Is this due to the Northern part of the CET at present ?

How many stations count towards the CET is it just 3 ?

[kold weather]

Mr D,

Has there been any years that have started of this low but ended up above 10c?

If March comes in below average, then the same question applies again.

Well there isn't any in the 20th century that were as cold as this Jan-Feb combo, 1959 is probably the closest at the moment which did come in at a warm 10.48, though we are already decently down on that year at this stage and 59 did have a well above average Spring overall.

Of course one record breaking season, which is always possible nowdays if everything sets up perfectly would bring a +10C year right back into play but right now I'd favour a sub 10C thats for sure, possibly the coldest year since 1996?

Who knows!

[BornFromTheVoid]

It would seem that there has never been a yearly CET above the 61-90 average that included both a January below 1.5C and a February below 3C. The highest any year that fit both those parameter got to was 1729 which had a January of 1.2C and a February of 2.3C and the year finished 9.26C.

There's also only been 2 years in the entire CET series that had winters over 1.5C below average and eventually ended up above average, 1947 and 1731, which ended up at 9.57 and 9.85 respectively.

So this year would be quite exceptional if it ended up above 10C

[SteveB]

Even more so then if March comes in -1.0c below average I would have thought.

I suppose if March does indeed come in -1c below average and we still finish the year with an annual CET above 10c, at least we will have one hell of a good Summer.

[Weather-history]

Its 9 weeks since the start of the year and we still haven't recorded a CET daily mean of 7.0 or greater. Its a pretty modest value, nothing difficult about it to achieve.

The last time we achieved was the 10th December, 12 weeks ago.

Its 100 days since we achieved a double digit CET mean, that was on the 24th November

25th November 2009 - 4th March 2010 CET: 2.7C

Edited March 5, 2010 by Mr_Data

[kold weather]

Yeah Steve though one much above average season brings the above 10C right back into play and at this stage its too early to know what March will do, could yet still get a very mild second half which takes this one above average, I'm far from sure though!

Th utter top limit to the years CET outcome now IMO is probably about 10.7C, that would need a well above average March and the rest of the year like 2006...and whilst I'm expecting a warmer second half to the year compared to the cold departures of the last 3 months, I'd be shocked if it was close to that!

That yearly rolling CET must be tumbling downwards yet further...

[virtualsphere]

Never mind reaching 10C, what about recording a sub 9C year? It's easy to forget that until relatively recently (circa 1994) that the majority of years came in between 9 and 10, with 10+ being less likely and two sub-9C years reported together in 1985 and 1986. Both got off to a colder start, but also recorded Decembers in the 6C+ range. If March continues below average, and December comes in between 3 and 4C, then we wouldn't need to go too much below average so I think this year may be as good a shot as we get for a while.

Given that we've waited so long for a sub 3C month and then we not only get two together, but also a sub 2C month to boot, what is to say that it can't happen for the annual series?

[Weather-history]

Highest CET daily mean of the year up to the 1st of March is just 6.7C!

Years with a "lower" highest CET daily mean of the year at this stage

1979: 5.4 (7.0 by 2nd of March)

1963: 2.4 (8.9 by 6th of March)

1895: 4.6 (7.2 by 16th of March)

1814: 6.4 (7.1 by 22nd of March)

Really are in unusual territory

For 1895, it was a gap of 79 days between a daily CET mean of 7.0+ being recorded.

For 1814, it was 86 days.

We are currently at 86 days and tomorrow will be the 87th.

Edited March 7, 2010 by Mr_Data

[Gavin P]

Never mind reaching 10C, what about recording a sub 9C year? It's easy to forget that until relatively recently (circa 1994) that the majority of years came in between 9 and 10, with 10+ being less likely and two sub-9C years reported together in 1985 and 1986. Both got off to a colder start, but also recorded Decembers in the 6C+ range. If March continues below average, and December comes in between 3 and 4C, then we wouldn't need to go too much below average so I think this year may be as good a shot as we get for a while.

Given that we've waited so long for a sub 3C month and then we not only get two together, but also a sub 2C month to boot, what is to say that it can't happen for the annual series?

The way we're going (the unusualness of the situation highlighted nicely by Kevin) I would say sub 9c is more likely than plus ten.

[kold weather]

I've got a feeling Gavin the summer and Autumn will probably ruin any chance of going sub 9C however sub 10C right now is IMO fairly likely, esp if the northern jet starts to take over more in the next few weeks and we keep a -ve AO, thats going to cause more northerly topplers which by late March will be one of the coldest sources we have. Not only that but the area from which we source our SW air is currently running decently below average.

Still more then possible to get above 10C but I'd argue its certainly more likely we will go sub 10C....the rolling CET should be interesting by mid March.

[guitarnutter]

Who has the 1 year rolling avaerage?

[Stu_London]

CET 01/03/09-28/02/10 - 9.87C

Has potential to fall further too if the cold/cool theme continues as up against a quite mild spring in 2009. I would say an annual in running CET of 9.4C or 9.5C is possible at some stage.

CET 07/03/09-06/03/10 - 9.84C

If March comes in at 6C, April at 8C and May at 11C (all around average) the CET in running will be 9.52C by end of May.

[kold weather]

That'd be pretty decent, anyway its got some more dropping to do yet it seems!

[Weather-history]

These rolling annual temperature graphs by NOAA illustrate the mild and cold periods pretty well.

Stornoway

Aberdeen

Aldegrove

Dublin

Valentia

Shawbury

Bournemouth

Edited March 8, 2010 by Mr_Data

[Roger J Smith]

You may find this interesting ... I worked out the correlation stats for the CET monthly anomaly values over the entire period of record (including this past winter) and the month-to-month correlation drops significantly in April to May.

These are the stats:

DEC to JAN (year n to year n+1), rest are all same year

DEC to JAN .363

JAN to FEB .356

FEB to MAR .413

MAR to APR .307

APR to MAY .106

MAY to JUN .273

JUN to JUL .275

JUL to AUG .427

AUG to SEP .370

SEP to OCT .251

OCT to NOV .225

NOV to DEC .245

And since this may be of particular interest,

JAN to JUL .150

JUL to JAN of next year .015

I may look at some others, but clearly, the seasonal temperature persistence factor tends to fade out over 2-4 months on average, rather than month to month. However, April to May has much less persistence than any other adjacent monthly sets, which suggests it is the prime time of the year for regime change to occur. April 2007 comes to mind in that regard, as it seems to mark the end of the warm period in western Europe that may have started around late 1987.

The January to July and July to next January correlations are so weak that it certainly shows that winter-to-summer persistence in either case is much weaker than month-to-month persistence. In fact, the correlation from summer to winter is totally insignificant (explains .02% of variance).

March and August are the most likely months to follow the anomaly pattern of the previous month.

Finally, how about year to year correlation?

This is what I came up with in that regard:

JAN to JAN 0.073

FEB to FEB -.017

MAR to MAR 0.145

APR to APR 0.154

MAY to MAY 0.174

JUN to JUN -.045

JUL to JUL 0.051

AUG to AUG 0.178

SEP to SEP 0.192

OCT to OCT 0.168

NOV to NOV 0.142

DEC to DEC 0.068

None of these are very high, and two are slightly negative (showing a tendency to anomaly sign reversal). The late summer and early autumn show the greatest year-to-year persistence.