July 2022 C.E.T. and EWP forecast contests

[Scorcher]

44 minutes ago, reef said:

That hasn't been the case here. Every August from 2013 to 2019 had above average sunshine:

2013: 205hrs (111%)
2014: 201hrs  (109%)
2015: 191hrs (104%)
2016: 207hrs (113%)
2017: 207hrs (113%)
2018: 190hrs (103%)
2019: 217hrs  (118%)

The last two were pretty dull though.

Probably illustrates that they've been westerly-type months for the most part with areas further east doing better sunshine-wise.

None of those are really outstandingly sunny though- we definitely haven't had a May 2020/2018 equivalent in August!

Edited August 1, 2022 by Scorcher

[Roger J Smith]

July appears to be confirmed at 18.2 thanks to a 20.4 average on 31st. 

Richie3846 and rwtwm had that for a 1-2 finish in CET. Rest of top ten from table of forecasts, order of entry to be added ... official confirmation comes when J10 shows the scoring updates. 

 

 

[Roger J Smith]

__ July 2022 CET scoring results __ (to be confirmed by J10) ...

 

Rank _ CET _ EWP _ Forecaster (order of entry)

 

_01 __ 18.2 _ --- ---__ richie3846 ( 5 ) 

_02 __ 18.2 _ 42.0 __ rwtwm ( 13 ) _____________ 2nd best combined (2nd CET, 6th EWP)

_03 __ 18.3 _ 39.0 __ Emmett Garland ( 15 ) ___ best combined (3rd CET, 3rd EWP)

_04 __ 18.1 _ 43.0 __ Reef ( 28 ) ________________ 3rd best combined (4th, 8th)

_05 __ 18.1 _ 49.0 __ February1978 ( 46 ) 

_08*__ 18.1 _ 80.0 __ Norrance ( L 2-1 ) 

_06 __ 18.0 _ 55.0 __ Polar Gael ( 4 ) 

_07 __ 18.0 _ 55.0 __ seaside60 ( 36 )

_09 __ 18.4 _ 47.7 __ Roger J Smith ( 39 ) 

_12*__ 18.0 _ 54.0 __ stewfox ( L 2-3 ) 

_10 __ 17.9 _ 75.0 __ EdStone ( 11 ) 

_11 __ 17.9 _ 72.0 __ summer18 ( 30 ) 

_13 __ 17.8 _ 45.0 __ Midlands Ice Age ( 31 ) 

_14 __ 17.8 _ 51.0 __ Feb1991Blizzard (41,13.5)

_15 __ 17.7 _ 54.5 __ Bobd29 ( 2 )

_16 __ 17.7 _ 47.0 __ Wold Topper ( 16 )

_17 __ 17.6 _ 60.0 __ dancerwithwings ( 19 ) 

_18 __ 17.6 _ 57.0 __ Mr Maunder ( 32 ) 

_22*__ 18.8 _ 29.0 __ daniel* (  L3 - 1 ) 

_19 __ 17.5 _ 60.0 __ DR(S)NO ( 7 ) 

_20 __ 17.5 _ 45.0 __ DiagonalRedLine ( 33 ) 

_21 __ 17.5 _ 62.0 __ Don ( 42 ) 

_23 __ 18.9 _ 35.0 __ Godber 1 ( 35 ) 

_24 __ 17.4 _ 45.0 __ Freeze ( 18 )

_25 __ 17.4 _ --- ---__ Summer Sun ( 10 )

_26 __ 17.4 _ 40.0 __ The PIT ( 14 )

_27 __ 17.4 _ --- ---__ damianslaw ( 30 ) 

_28 __ 17.4 _ --- --- __ Mark Bayley ( 45 )

_29 __ 17.3 _100.0__ Weather26 ( 9 )

_30 __ 19.1 _ 65.0 __ Stationary Front ( 17 ) 

_31 __ 17.3 _ --- ---__ Walsall Wood Snow ( 37 )

_32 __ 17.2 _ 56.0 __ SteveB ( 26 )

_33 __ 17.2 _ 45.0 __ Mulzy ( 40 )

_34 __ 17.1 _ 52.5 __ snowray ( 24 )

_35 __ 17.1 _ 42.0 __ jonboy ( 34 )

_36 __ 17.1 _ 75.0 __ J 10 ( 43 ) 

_38*__ 17.1 _ 69.0 __ davehsug ( L1-1 )

_37 __ 17.0 _ 44.0 __ NeilN ( 6 )

_39 __ 17.0 _ 80.0 __ East Lancs Rain ( 8 )

_40 __ 17.0 _ 60.0 __ Frigid ( 22 )

_43*__ 17.0 _ --- ---__ Duncan McAlister ( L 2-2 )

_41 __ 16.8 _ 79.0 __ I Rem Atl 252 ( 27 )

_42 __ 16.6 _ 73.0 __ noname_weather ( 44 )

_44 __ 19.9 _ 20.0 __ Earthshine ( 3 ) 

_45 __ 16.5 _ 58.0 __ virtualsphere ( 12 )

_46 __ 20.0 _ 44.0 __ Shillitocettwo ( 1 ) 

_47 __ 16.4 _ 55.0 __ weather-history ( 20 )

_48 __ 16.4 _ 60.0 __ summer blizzard ( 21 )

_49 __ 16.2 _ --- ---__ Kentish Man ( 38 ) 

_50 __ 16.1 _ 59.7 __ Jeff C ( 23 )

_51 __ 15.0 _ 60.0 __ syed2878 ( 25 )

_____________________

* note: jogs in rankings are to reflect scoring adjustments for late penalties, but the raw scoring will go in the order shown before those adjustments happen inside the CET scoring tables. These adjustments are one rank per day late. The only place where it matters is to determine "best combined" rankings. 

Similar rank adjustments are made to EWP scores. EWP ranks to be confirmed, will probably run from 1st for lowest EWP forecast, etc. 

[Earthshine]

Maximum CET an impressive 23.4°C (10th highest on record).  Even more impressive is that the maximum CET anomaly for Jan-July has been 2.43°C above average (15.13°C compared to normal 12.7°C).

[damianslaw]

5 hours ago, Scorcher said:

They just don't happen much these days Damian, so very hard to say. I can see August also being comfortably above average.

Sunshine-wise though we are due a very sunny August with above average sunshine- even the warm ones haven't tended to be very sunny over the last 15 years or so.

The odds of above rather than below certainly has increased over recent years. Mind last year brought a few, notably April. Jan, May and possibly October were below.. anyone have stats. One will turn up again I'm sure.. 

[Walsall Wood Snow]

At 18.2c July 2022 shares the same mean CET with July's:

1759

1818

1911

1934

 

[Roger J Smith]

While tied with those years (above post) in one decimal, apparently, 2022 is behind all of them in second decimal (how they separate them in the table now) so it ranks 20th warmest. The order is 1934, 1911, 1818, 1759, 2022 (16th to 20th assuming none are tied in two decimals).

[J10]

9 hours ago, Roger J Smith said:

July appears to be confirmed at 18.2 thanks to a 20.4 average on 31st. 

Richie3846 and rwtwm had that for a 1-2 finish in CET. Rest of top ten from table of forecasts, order of entry to be added ... official confirmation comes when J10 shows the scoring updates. 

 

 

Hoping to get the scoring charts out tomorrow, fingers crossed.

In addition to the monthly winners though. A special shout out to

 

On 20/06/2022 at 21:27, Shillitocettwo said:

  Am I first?

 

  Starting quite warm and quickly developing into a month with protracted warm or very warm even hot spells, with temperatures in the mid to high 30.Cs and 39.C will be reached at some point in the month. 

    So I'm going to slap a straight CET of 20.0.C on this although I feel this will be the prelude to an August completely off the charts which will go even higher!

   Dryish overall but what rain falls will mostly come from thunderstorms at 44mm.

With the quote above.

[Roger J Smith]

I would imagine there were places in the greater London region that had a monthly average near 20.0. I would bet the three locations they use for CET probably had averages like 17.2, 18.2 and 19.2 to get their blended 19.2 (ignoring the 0.2 urban heat island correction, think I have a handle on why they do that, Rothamsted is near Luton and is at least somewhat suburban while Pershore is quite rural and from what I could see on google earth, Stonyhurst is in a very small village. So it's probably 0.6 off Rothamsted that averages out to 0.2 off the CET. 

I know shillitocettwo was mentioning severe heat this summer earlier in the contest (or maybe in another thread), it wasn't just a call made in early July by them. Without that cool start July 2022 would have been right up there with the big players. Even so it is quite similar in outcome to 1868 and 1911 which are regarded as hot summer months and in both cases were followed by more record heat in August. 

[Roger J Smith]

EWP tracker has ended its work at 22.5 mm. There was in fact the same value posted in the tables on the 5th.

Only those with at least 5/8 contests entered are ranked for average error. Forecasters with fewer than 8 entries have their number of entries in brackets after their name. (e.g., seaside60 has entered seven of eight). I have added the stats of two forecasters who did not enter July but have an average error ranking. 

 

Annual EWP scoring after July ___

Rank _ Forecaster _ Points ___ Avg err (mm) _ Rank

 1. The PIT __________ 56.30 ____ 21.08 __ 3

 2. Freeze ___________ 55.85 ____ 20.85 __ 2

 3. Feb1991 Blizzard _53.62 ____ 25.03 __ 4

 4. Polar Gael ________50.96 ____ 27.15 __ 7

 5. Bobd29 __________ 50.87 ____ 25.16 __ 5

 6. Mulzy ____________ 48.11 ____ 26.53 __ 6

 7. Reef ______________46.63 ____ 27.48 __ 8

 8. NeilN ____________ 44.06 ____ 29.08 __11

 9. seaside60 (7) _____ 43.72 ____ 28.96 __ 9

10. SLEETY (7) ________42.76 ____ 29.01 __10

11. DR(S)NO _________ 42.58 ____ 29.95 __16

12. EdStone _________ 42.04 ____ 30.40 __17

13. DiagonalRedLine (7)_40.83 ___ 29.91__t14

14. weather-history __ 40.78 ____ 30.95 __20

15. Mr Maunder ______40.41 ____ 29.28 __13

16. dancerwithwings _ 39.99 ____ 30.90 __19

17. Roger J Smith _____ 39.93 ____ 30.68 __18

18. Godber 1 _________ 39.36 ____ 35.10 __29

19. Don _______________39.25 ____ 33.35 __25

20. Emmett Garland __ 39.24 ____ 34.88 __28

21. JeffC ______________ 38.87 ____ 31.91 __22

22. daniel* (7)^ _______ 38.40 ____ 29.91 __t14

23. J 10 _______________ 38.17 ____ 31.23 __21

24. Frigid _____________ 37.89 ____ 37.08 __33

25. rwtwm ____________37.53 ____ 35.40 __30

26. virtualsphere _____ 37.28 ____ 33.90 __26

27. Midlands Ice Age __ 36.92 ____ 37.18 __34

28. Wold topper (5) ___ 36.88 ____ 17.64 __ 1

29. syed2878 _________ 36.79 ____ 35.99 __31

30. February 1978 ____ 35.82 ____ 32.23 __23

31. summer blizzard __ 35.59 ____ 36.83 __32

32. davehsug _________ 34.51 ____ 37.59 __35

33. prolonged SL (7) ___34.39 ____ 32.77 __24

34. summer8906 (6) __ 30.86 ____ 41.33 __42

35. noname_weather__ 30.53 ____ 40.03 __40

36. jonboy ____________ 30.36 ____ 38.00 __37

37. Weather26 ________ 30.12 ____ 44.40 __44

38. shillitocettwo ______ 29.53 ____ 45.58 __47

39. SteveB (6) _________ 29.25 ____ 45.32 __46

40. Stargazer (6) ______ 28.76 ____ 34.15 __27

41. stewfox (6) ________27.63 ____ 42.80 __43

42. Norrance (7) ______ 27.28 ____ 37.91 __36

43. snowray __________ 26.36 ____ 39.95 __39

44. Earthshine (5) _____ 25.95 ____ 29.24 __12

45. Stationary Front ___25.59 ____ 39.55 __38

46. I Rem Atl 252 ______22.43 ____ 49.15 __49

47. summer18 ________ 20.32 ____ 48.83 __48

50. Blast from the Past (5) 16.28 __ 44.64 __ 45

55. Coldest Winter (5) __ 12.70 ____ 40.96 __ 41

________________________

^ daniel* would be 12th if 3.7 points from too late forecast for May added.

... average error would be 27.28 mm which would rank 8th

 

Edited August 5, 2022 by Roger J Smith

[Roger J Smith]

Updated list of top eleven dry months of July 1766 to 2022

Rank ____ Year ____ Prec (mm)

01 ______ 1825 _____ 8.2

02 ______ 1800 _____ 9.1

03 ______ 1911 ____15.8

04 ______ 1868 ____20.0

05 ______ 1869 ____21.6

06 ______ 2022 ____22.5

07 ______ 1977 ____23.7

08 ______ 1885 ____24.5

09 ______ 1898 ____25.2

10 ______ 1955 ____26.1

11 ______ 1999 ____26.3

__________________________

The driest July since 1999 was 34th driest 2006 (39.1 mm) ... 2018 was 39.6 mm. 

[J10]

PDF -> July 2022 Summary.pdf

Excel -> July 22 CET.xlsx

Monthly

2 players got it spot on this month richie3846 and rwtwm.

Seasonal
All change in the Seasonal competition with a brand new 1-2-3.

February1978, Summer18 and richie3846

Overall

The Top 4 players remain the same as last month, but some of the positions have changed.

richie3846 1st (from 2nd)
Freeze 2nd (from 1st)    
February1978  3rd (from 4th)
Stationary Front  4th (from 3rd)

Edited August 2, 2022 by J10

[Roger J Smith]

The EWP table value has turned out to be 22.5 mm, same as the tracker had on 2nd. The scoring table a few posts back now has the average error and ranking in addition to the points totals (which remain unchanged). 

[Roger J Smith]

Here is an update on how consensus and the two normals are doing. The contest was scored using legacy CET values Dec to Apr and v2.0 (all that now exists) in May to July. These tables show the actual values, then the errors, and the contest ranks of the entrants who have similar values for average error. This gives an approximate idea of how the "robots" are doing although their scores might have added up somewhat differently as the forecasters with those average errors ranked differently in specific months. This will be adjusted when average errors for the contest through June are available.  I have only compared the contest entrants who have not missed any months. A handful of others might rank better than these robots with one or two missing forecasts.

 

(a) actual values

ROBOT ______ DEC _ JAN _ FEB _ MAR _ APR _ MAY _ JUN _JUL

Consensus __ 4.8 __ 5.4 __ 5.6 __ 7.1 __ 8.4 _ 12.3 _ 15.1 _ 17.5

1981-2010 ___4.6 __ 4.4 __ 4.4 __ 6.6 __ 8.5 _ 11.7 _ 14.4 _ 16.6

1991-2020 ___5.0 __ 4.7 __ 4.9 __ 6.7 __ 8.9 _ 11.9 _ 14.6 _ 16.8

contest CET _ 6.4 __ 4.6 __ 6.8 __ 7.9 __ 9.1 _ 13.1 _ 14.9 _ 18.2

 

(b) absolute errors

ROBOT _____ DEC _ JAN _ FEB _ MAR _ APR _ MAY _ JUN _JUL ___ average

Consensus __ 1.6 __ 0.8 __ 1.2 __ 0.8 __ 0.7 __ 0.8 __ 0.2 __ 0.7 ___ 0.85

1981-2010 ___1.8 __ 0.2 __ 2.4 __ 1.3 __ 0.6 __ 1.4 __ 0.5 __ 1.6 ___ 1.23

1991-2020 ___1.4 __ 0.1 __ 1.9 __ 1.2 __ 0.2 __ 1.2 __ 0.3 __ 1.4 ___ 0.96

SUMMARY: Consensus was warmer than both normals for all months except Dec (where it was between them) and Apr (when it was below both). The actual value was warmer than both normals in Dec, Feb, Mar, Apr, May, June, and July, so the contest consensus was correctly placed in Feb, Mar, May, (marginally so) June, and July. In December and April, our consensus was too low, and in January and June it was too high. The overall performance of consensus was slightly better than the 1991-2020 normal and considerably better than the cooler 1981-2010 normal value. The lowest average error would have been achieved by any strategy of making forecasts around 1.2 above 1991-2020 averages (the results drop steadily from the 0.90 of 1991-2020 for each 0.1 increment, and the gain reverses starting at 1.3 above). The average error of those forecasts (1991-2020 + 1.2) would have been around 0.54 deg.  

 

(c) Contest ranks (comparing only entrants with seven of seven contests entered and notes to show rank including those with one missing forecast) ... I did not add two or three forecasters missing two contests in these comparisons but some of them are also better than consensus and the normals.

 

optimal strategy (91-20)+1.2 (avg error 0.54) __ 2nd (richie3846 0.44, freeze 0.55, February1978 0.56)

consensus (avg error 0.85) ____________________ 9th (two others with one missing forecast are better, so 11th)

1991-2020 (avg error 0.96) ____________________t-14th (two others with one missing forecast are better, t-18th)

1981-2010 (avg error 1.23) ____________________t-37th (no additional from one missing forecast, t-41st)

(Average error is well correlated with total score but not perfectly so, however it can be inferred that contest ranks overall for the three "robots" would be similar to the above.)

Obviously playing an optimal strategy is a good one, if you can guess in advance what that is; at the moment, any optimal strategy between adding 0.5 and 1.5 to 1991-2020 would put you pretty close to the lead held by Richie3846, and also to the close second-third scores of freeze and February1978, and using that  sort of strategy, you could be well into the top ten, even just using 1991-2020 is good for 14th place in average error. Using 1981-2010 gets you no further into the table than middle ranks (55 are ranked based on at least six of eight entries). Since average error is not the only component of CET contest scoring, it would take a more rigorous analysis to work out the points totals for these automatic type strategies but they are bound to be close in rank to the average error. 

An equally good (or marginally better) strategy is to predict consensus plus 0.7, this results in an average error of just 0.50, better than 91-20 plus 1.2, still second compared to forecasters. 

Edited August 5, 2022 by Roger J Smith