July 2013 CET Forecasts / Competition

[virtualsphere]

For reference, these are the July 18 to 31, and 1 to 15 August, daily records in the CET daily means:

 

...

03 Aug ... 24.4 (1990) ... 10.5 (1912)

04 Aug ... 23.4 (1975) ... 11.3 (1812)

05 Aug ... 23.1 (2003) ... 11.3 (1812)

...

12 Aug ... 22.2 (1911) ... 10.5 (1912)

...

14 Aug ... 21.3 (1911) ... 10.8 (1912) 

 

 

Weird to see both 1812 and 1912 in here - it would seem that years ending in '12' are destined to be dire summers.  At least I won't be alive in 2112!

[Summer Sun]

18.4c to the 19th

[The PIT]

Sunny Sheffield up to 18.3C no change for today unless tonight is cold. GFS goes for m ore Heat ECM not so keen on the idea so god knows which way it will go.

[Roger J Smith]

Looking at all 00z model output and factoring in 06z GFS, would say that if the mean holds through today at 18.4, then 20.8 average for 21st-23rd brings us to 18.7 and after that an average of 17 brings it back down to 18.2 ... after adjustments will be lucky to hold on to 18 but finishing range likely 17.5 to 18.5.

Edited July 20, 2013 by Roger J Smith

[Summer Sun]

18.5c to the 20th

[The PIT]

Sunny Sheffield still 18.3C yesterdays low max offset by the night time low so no change. Today will be drop though unless it warms up overnight.

[I remember Atlantic 252]

18.1C, but hopefully cooling off for summer hols (july 22nd)

 

May not be a bad guess that! cooling off a little for summer hols, but maybe not CET wise due to mild nights

[Summer Sun]

18.5c to the 21st

[SomeLikeItHot]

18.6 to the 22nd (21.5)

 

Surprising that wasn't the hotest day this year in CET terms or even max CET.

 

Anyway with a min today of 16.1 and max's around 25, - using the met forecasts for 5 days we should see something like.

 

23rd 18.7 (20.6)

24th 18.7 (19.2)

25th 18.7 (19.0)

26th 18.7 (19.3)

27th 18.8 (19.6)

 

After that GFS ensemble mean there after falling slightly perhaps to around 18.6 by the 31st.  Certainly at this point something in the range 18.4 - 18.9 is looking most likely.

Edited July 23, 2013 by SomeLikeItHot

[BornFromTheVoid]

I'd go with maxima a little closer to 26C today, but similarly an increase to 18.7C on tomorrows update.

 

After that, with the 06z GFS operational

 

18.7C to the 24th (19.0)

18.7C to the 25th (18.6)

18.7C to the 26th (18.3)

18.7C to the 27th (18.4)

18.7C to the 28th (19.4)

18.7C to the 29th (19.6)

18.7C to the 30th (18.7)

 

Little change over the next week. Mid 18s after corrections most likely. This should still have us well within the top 10 warmest July's on record, with a chance of a top 5 finish also. In comparison, March this year was the 12th coldest on the CET record.

[SomeLikeItHot]

18.7 to the 23rd (20.5)

 

with a min today of 15.3  and max's around 24, - using the met forecasts for 5 days we should see something like.

 

 

24th 18.7 (19.7)

25th 18.7 (18.4)

26th 18.7 (18.4)

27th 18.7 (18.8 )

28th 18.7 (18.8 )

 

With the GFS 0z ensemble mean then taking it to

 

29th 18.7 (18.0)

30th 18.6 (17.0)

31st 18.6 (17.0)

Edited July 24, 2013 by SomeLikeItHot

[Scorcher]

The first 18C+ month for some time seems to be odds-on now after corrections, this month looks like it's really going to smash the run of below average months.

[Don]

The first 18C+ month for some time seems to be odds-on now after corrections, this month looks like it's really going to smash the run of below average months.

 

Yes and I wouldn't be suprised if the remaining months of the year all finish above average which would be disappointing after the long run of cold months!

 

 

[SomeLikeItHot]

While I am not ruling out anything its worth noting that in general CET for July is relatively weakly correlated with the other months of the year except the following 2 months of aug-sept.

 

Notably both March and April CETs have higher correlations to Oct-Dec CET than any other month so far this year.If you look at Nov-Dec then March has by far the highest correlation of any month so far.Looking at the two notable CET months this year (so far!)of the 10 Coldest 20th C March's were followed by 6 out of 10 being colder than average winters.

of the 10 warmest 20th C Julys's were followed by 6 out of 10 being warmer than average winters.

 

In the end though I think pretty much anything can happen 3+ months out.

Edited July 24, 2013 by SomeLikeItHot

[BornFromTheVoid]

The 12z GFS upgrades the warmth for the next week.With today likely to average a little over 20C, we'll probably stay on 18.7C tomorrow, but just below 18.75C.

 

After that, the 12z GFS operational run has the CET at

 

18.8C to the 25th (19.7)

18.8C to the 26th (19.3)

18.8C to the 27th (19.2)

18.9C to the 28th (20.2)

18.9C to the 29th (18.3)

18.8C to the 30th (18.4)

18.8C to the 31st (19.1)

 

The 31st is right on the cusp of 18.9C. Either way, high 18s seems the most likely finish before corrections.

 

Here are the top 10 July's for comparison

1st......2006...... 19.7C

2nd.....1983...... 19.5C

3rd......1783...... 18.8C

4th......1852...... 18.7C

4th......1976...... 18.7C

6th......1995...... 18.6C

7th......1921...... 18.5C

8th......1757...... 18.4C

8th......1808...... 18.4C

10th....1701...... 18.3C (joined with1733,1859 and 1868)

 

After corrections, a top 10 CET is extremely likely, with maybe a 30-40% chance of a top 5 finish?

 

 

[The PIT]

Sunny Sheffield back up 18.3C so the cool weekend had a big effect considering we were only 0.1C difference at one stage.

[Weather-history]

10th March- 9th April 2013 CET was 2.4C and 3 months later a July with a CET of 18.0+ is likely that's a 16C difference with a space of just 3 months which is pretty remarkable.

Edited July 24, 2013 by Weather-history

[Lettucing Gutted]

10th March- 9th April 2013 CET was 2.4C and 3 months later a July with a CET of 18.0+ is likely that's a 16C difference with a space of just 3 months which is pretty remarkable.

This is the second time in only 3 years that an exceptional cold month has been followed by an exceptionally warm in just to 2-4 months.

[Don]

This is the second time in only 3 years that an exceptional cold month has been followed by an exceptionally warm in just to 2-4 months.

 

Indeed, having December 2010 the 2nd coldest December on record to be followed by the warmest April on record (By some margin) was pretty remarkable too!

Edited July 24, 2013 by Don

[Cheshire Freeze]

I don't see it as extraordinary to have a very cold month followed by a very warm one a few months later. Each month is independent of one another meaning statistical quirks like this really shouldn't come as a shock.

[Weather-history]

I don't see it as extraordinary to have a very cold month followed by a very warm one a few months later. Each month is independent of one another meaning statistical quirks like this really shouldn't come as a shock.

It depends on the extremeness of the values and the time gap, though. If they happen occasionally fair do but how often is the 2nd coldest of that month on record followed 3 months later by the warmest of that month on record based on records over three centuries?What are the climatic probabilities of that happening? Edited July 24, 2013 by Weather-history

[Scorcher]

Especially when you consider that April 2011 smashed the previous record by some margin, a record that was only set 4 years earlier. April 2007 also easily beat the previous warmest April. I personally think April 2011 was far more exceptional than December 2010 when you consider how far behind the 3rd warmest April is.

[hillbilly]

It depends on the extremeness of the values and the time gap, though. If they happen occasionally fair do but how often is the 2nd coldest of that month on record followed 3 months later by the warmest of that month on record based on records over three centuries?What are the climatic probabilities of that happening?

maybe you need to study statistics a little more and it does not allow for chaos theory,if is very highly unlikely then it will happen.In these parts march was not the second coldest,was third and 1947 was much colder and had a warmer summer.1989 had a cold april and a hot summer 1995 I think was similar and even 2006 had a cold spring

[Don]

maybe you need to study statistics a little more and it does not allow for chaos theory,if is very highly unlikely then it will happen.In these parts march was not the second coldest,was third and 1947 was much colder and had a warmer summer.1989 had a cold april and a hot summer 1995 I think was similar and even 2006 had a cold spring

 

April 1989 was a cold month but not exceptionally so.  Also, Spring 2006 was not particularly cold overall.  March was the only cold month that Spring.  April and May were both warmer than average.

[Roger J Smith]

The weather basically went from one extreme to the other between early March 1947 and late May 1947. Finding monthly values that would match the probable swing in 2013 might prove elusive but as mentioned a large swing occurred from Dec 2010 to April 2011 that will probably prove more significant in rank order terms.

 

If monthly values were truly independent (which correlation studies suggest they are not, quite) then I think the math to answer the question, what are the chances of a top 5 and bottom 5 month falling within four months of each other would be solved something like this:

 

1. chance of top 5 cold month is 1 in 75 (approx) over CET period

2. chance of top 5 warm month is 1 in 75 (approx) over CET period.

 

3. chance of both occurring four months apart (and not 1,2,3) is

 

(1/75) x (73/75) x (73/75) x (73/75) x (1/75) which is 73 cubed over 75 to the fifth power.

 

My excel program says that probability is .000164 or 0.0164% or one in five thousand.

 

The math for how often would it be expected within 4 months is the sum of probabilities for one, two, three and four month return. Since 73/75 is close to unity, this is approximately four times the value of (1/75) squared which is 1/5625 times 4, or 4/5625 (reduced slightly by those 73/75 terms) so about one in fifteen hundred.

 

The odds of seeing this twice in three years (actually in four tries) would be something like one in a million. So that's why the internet was so well timed.

 

Now, I realize that my implied assumption of equal probability of top 5 and bottom 5 is skewed somewhat by climate change although with the recent run of winters being colder than other seasons that is muted since the cold anomalies in question were wintry more or less. Also the assumption of independence is subject to question especially for month one and to some extent month two but actually if there was a study made of reversal of trend I would imagine 3-5 months might show a slight peak over random so that assumption is not all against the flow of the odds derived here.

 

This would also neglect the very slim probability that an intervening month was also top five or bottom five, which was not part of the discussion, so if that ever happened it would increase the overall probability of this outcome very slightly (probably not in the first or even second significant digit).

Edited July 24, 2013 by Roger J Smith