The warmest 12 months ever (May 2006-April 2007)

[North-Easterly Blast]

For around the past three years Kevin Bradshaw has been writing reviews on the UK's major past weather events, including a database of reviews on TWO, and many reviews and reports on the Historic Weather section of this forum, and also on the UK WeatherWorld Past Weather section.

Recent months have included Kevin writing reviews / special reports on the severe winters of 1946-47, 1962-63 and 1978-79, and also the notable cold spells in 1985, 1986, 1987, 1991, 1995-96 and many more, and a most recent review on the cool summer of 1954. In all of these reviews he has written a detailed summary of the weather events together with selected historic charts from the days of the coldest spells of the above winters, and the CETs for the coldest periods of each of those winters.

As we all know that the 12 months from May 2006 to April 2007 was the warmest ever 12 month period in the CET series with a CET mean of 11.63*C, hugely beating the old record of 11.07*C set in Nov 1994 to Oct 1995; and given that exactly this time last year we were nearing the end of the first highlight of the above record 12 month period, the hottest July and hottest ever calendar month on record, I would be most grateful if Kevin Bradshaw would be willing to write a similar review / special report on the May 06 - April 07 record warm 12 month period, to place in the Historic Weather section of this forum, set out in a similar way to the reviews of the severe winters such as 1978-79, the cool summer of 1954 etc, with selected charts from the warmest spells of the last 12 months like last July's heatwave, the very warm spells in Sep and Oct last year, the very mild spells last winter, and the very warm spells in April. The report should also detail that we saw July, September and April set new records on the CET, and October and January were also near record months, and also the whole autumn set a new CET record and that the other three seasons also were in the top five warmest ever. An even more notable point could be that the report details that August 2006 still ended up with a CET close to average despite having northerly winds over the UK for most of the month, when the synoptics should have produced a much cooler month!

A review / special report written by Kevin Bradshaw on the record warm 12 months from May 2006 - April 2007 similar to his other reports on major past UK weather events would be a great addition to this forum's Historic Weather section, as this (the UK's warmest year ever in 348 years of records, by a huge margin of 0.5*C!) has now become one of the most major events in the UK's weather history equally as much as the exceptional winter of 1962-63.

[BLAST FROM THE PAST]

Looks like you did it already! Oh and last August synoptics has been talked about many times, the source of air was not of arctic origin so it shouldn't have been much cooler.

BFTP

[Weather-history]

Here's a small stats summary on July 2006-July 2007 because for me, the period July 2006 to July 2007 has been truely staggering with its exceptional warmth and recently its wetness for England and Wales . Lets look at the stats.

July 2006: Warmest July and month ever recorded with a CET of 19.7, sunniest month ever recorded for Areal series

August cooler than September

September 2006: Warmest September ever recorded with 16.8

October 2006: Joint 3rd warmest on record with 13.0

Warmest September-October combination on record

Warmest Autumn on record by a fair margin with a CET of 12.6

Warmest second half to a year on record with 13.4

January 2007: 5th mildest on record (7.0)

Joint 4th mildest winter on record with a CET of 6.4

April 2007: Warmest on record with 11.2, 4th driest on record, sunniest on record for Areal series.

May 2007: 10th wettest May on record

Spring 2007: 2nd warmest on record with a CET of 10.1

June 2007: 2nd wettest on record

Wettest May-June period on record

Wettest June-July period on record

Wettest May-July period on record by some margin

Warmest first half to a year on record with a CET of 9.7

It really has been a remarkable period we have been through.

Edited July 27, 2007 by Mr_Data

[Bobby]

Here's a small stats summary on July 2006-July 2007 because for me, the period July 2006 to July 2007 has been truely staggering with its exceptional warmth and recently its wetness for England and Wales . Lets look at the stats.

July 2006: Warmest July and month ever recorded with a CET of 19.7, sunniest month ever recorded for Areal series

August cooler than September

September 2006: Warmest September ever recorded with 16.8

October 2006: Joint 3rd warmest on record with 13.0

Warmest September-October combination on record

Warmest Autumn on record by a fair margin with a CET of 12.6

Warmest second half to a year on record with 13.4

January 2007: 5th mildest on record (7.0)

Joint 4th mildest winter on record with a CET of 6.4

April 2007: Warmest on record with 11.2, 4th driest on record, sunniest on record for Areal series.

May 2007: 10th wettest May on record

Spring 2007: 2nd warmest on record with a CET of 10.1

June 2007: 2nd wettest on record

Wettest May-June period on record

Wettest June-July period on record

Wettest May-July period on record by some margin

Warmest first half to a year on record with a CET of 9.7

It really has been a remarkable period we have been through.

Indeed, at least the weather never seems to be boring these days.

[charlton north-downs]

What are the odds of a 51 year old living through an event such as the the record breaking May June July 2007 , which broke say a 250 year old record. Can odds apply to weather anyway as it is not constant . If the same principle applied as the lottery jackpot which runs like this 49x48x47x46x45x44 = 10,068,347,520 divided by 720 because the numbers dont have to be in order. =13,983,816 nearly 14 million to 1 . Am I having a really grey moment here-its along time since I left school but could it be like this 250x249x248 divided by 51 =302,705 to 1

[osmposm]

Stratos is the statistician, I believe (HEEEEEEEELP, Stratos!!), but I think it may work like this (though it's nearly forty years since I did my Probability & Statistics):

Ignoring any questions of the appropriateness of simple statistical analysis to weather (particularly in a climate apparently in a period of significant change), and assuming the probability of it occurring are equal in any year, the chances of the record being broken in any individual year are one in however long the record is at that time. Thus taking this year alone, it would be one in 242 (record began in 1766).

But it gets more complex if you want to know the odds of it happening in 2007 or 2006 or 2005 or or any year back to 1956 (assuming that was the 51 year old's first early summer) - or indeed of it happening in more than one year during the period. To do that you calculate the probability of it NOT happening in any year during the period, and you subtract it from 1. To make it even more confusing, the length of the record has extended by one year for each year of his life....

So, in 1956 the chances of the May-July high rainfall record (or any other individual weather record) NOT falling were 190/191. In 1957 they were 191/192, in '58 192/193 and so on. The chances of the record NOT falling in 56 or 57 or 58 are calculated by multiplying the individual probs together, and the chances that the record would fall one or more times were that fraction subtracted from 1 - i.e 1 minus (192X191X190 divided by 193X192X191) = 1 - (190/193) = 0.0155441, i.e approx 1.55%, or about 63 to 1 against.

For the 52 years of early summers we are saying the man has lived the calculation would be:

1 minus (241X240x239.......X190 / 242X241X241......X191) = 1 - (190/242) = 1 - 0.7851239 = 0.248761, i.e. approx 21.5%, or 3.65 to 1 against.

So, I believe the chances of your 51 year old today having lived through a record-breaking May-June wetness spell (or any other individual yearly-assessed record in 242 years) is rather higher than you think, at slightly over 7-2 against.

In the longer CET record the probability for a record temp is lower.....and of course the probability of seeing as many different records broken as we have seen in the last year is much, much lower still.

Ossie

[Stratos Ferric]

Stratos is the statistician, I believe (HEEEEEEEELP, Stratos!!), but I think it may work like this (though it's nearly forty years since I did my Probability & Statistics):

...

Ossie

I'm not a professional statistician, but I do get involved in a lot of analysis.

Far be it from me to quibble with Os's numbers, apart from anything I'm supposed to be packing. I'd make two observations re CND's questions:

1 - As Os states (and, alas, this proabably invalidates some of his maths), the weather is not like the lottery. The odds in the lottery are fixed and permanent so long as all else remains unchanged. Weather and climate are dynamic, so the total population of possibility changes over different time periods. Over a longtime we can treat the data (e.g. CET) as if normally distributed, though this is really a cheat. As we're seeing at present, the outcomes is not normally distributed at present (re temperature) because the population mean is moving upwards; this skews the odds slightly. Therefore, any calculation of odds at present, based on history, is either going to be an overstatement or understatement (depending on what feature you're calculating against).

2 - The other factor is that assuming the population mean were unchaning, each new record would make the setting of a further record less likely. It is not only because, as Os states, the population grows larger by 1 with each addition (imagine how much harder the lottery would be if, each week, a new numbered ball was added), but because the margin is now pushed further out. Tossing a coin would be a good analogue. It is harder to toss three consecutive heads than two, four then three, and so on. Each new record makes the setting of a new recroder harder by some factor of 1/n. N grows depending on how much the new record adds. At present, again, the moving mean actually compensates for this on one tail, and exacerbates it at the other. Thus, the current run of exceptional warmth looks astonishing in the mirror of history, but in ten years' time (after having waited 350 years) we may have breached it again.

The short answer is that if the climate continues to change as it appears to be doing at present, you won't have to wait as long for a new record to be set, so long as you're not waiting for a cold winter.

[Weather-history]

The contrast between July 2006 and July 2007 is pretty remarkable

July 2006

Rainfall: 41.3mm

Sun: 287.6hrs

CET: 19.7

July 2007

Rainfall: 126.9mm+

Sun: ~140hrs

CET: 15.4-15.6

Julys 2006 and 2007 will go down as a set of contrasting months in the same way Januarys 1796 and 1796 and Augusts 1911 and 1912 have gone down.

[Weather-history]

If you take the period August 06 to July 07, then September 06 is going to be the warmest month in that 12 month period.

The last time September was the warmest month for that period was from August 1865 to July 1866

The only other occasion was August 1729 to July 1730

[North-Easterly Blast]

Not sure that I would rate Julys 2006 and 2007 as contrasting months to the same degree as August 1911 and 1912 were; July 2007 is certainly not going to be anywhere near record breaking cool for July, and not record breakingly wet either, and July 2006 although the hottest on record either, was dry yes but not exceptionally so at 41mm. In fact July 1999 was a drier month than July 2006, and believe it or not the Julys of 1977 and 1979 were drier months than July 2006 even though they were average to slighly below CETwise.