June 2017 C.E.T. Forecasts

[The PIT]

Sunny Sheffield now 16C will fall below that for tomorrow.

[Roger J Smith]

10 hours ago, Relativistic said:

The 21st is currently showing 21.7C (the 20th is showing 20.1C), but the 21st was definitely showing 22.5C a few days ago. It seems that the value for the 22nd has also been upped very slightly, from 17.9C to 18.0C.

I suspect part of the scenario is that I lost track of what day it was, when I posted and when the records fell etc, sometimes the eight hour time zone difference messes with my brain. But whatever the situation, I am happy for good old 1846 and its hard-fought records (not an SUV in sight).

Between May 1833 and June 1846 there's a bit of a problem for the IPCC in there somewhere. 

 

[NorthYorksWeather]

15.9c here to the 28th, 2.4c above the 1981-2010 average.

Couple more drops likely for the month to finish around 2c above average.

[Summer Sun]

16.3 to the 28th

2.2 above the 61 to 90 average

1.9 above the 81 to 10 average

__________________________________________

Current high this month 17.1 to the 1st

[Relativistic]

1 hour ago, Larger than average Hobo said:

Are any of the days this week getting close to the record daily lows? It's been cold and wet here up north so I suspect it may have come close. 10.2C on 27th might have been under threat. Has it been as cold in the CET zone?

Unfortunately, we're not even close. Yesterday was the coldest day of the last few, at a provisional 13.9C. The record for the 28th is 9.4C in 1812.

[BornFromTheVoid]

The nights are much too mild for any daily mean records to fall. I'll check the record lows for the daily maxima later this evening and see if any of those are under threat.

[The PIT]

Sunny Sheffield at 15.8C looking like a 15.7C finish for the month. Anyway at the moment it's +1.7C above normal. Rainfall at 99.6mm 136.4% above normal and will go above 100mm by tomorrow 64mm of which has fallen in three days.  

[snowray]

So we have a well above average first half of the year Jan-June, I wonder if we could end up with a colder than average 2nd half? Charts for July looking consistently bleak.

[BornFromTheVoid]

So the maximum yesterday of 14.5C is, provisionally, the lowest on record for June 28th. The previous record was 14.7C in both 1945 and 1981

Edited June 29, 2017 by BornFromTheVoid

[Weather26]

11 hours ago, BornFromTheVoid said:

So the maximum yesterday of 14.5C is, provisionally, the lowest on record for June 28th. The previous record was 14.7C in both 1945 and 1981

From one end of the spectrum (the warmest June day since 1976 on the 21st) to the other in the space of a week!

[Summer Sun]

16.2 to the 29th

2.1 above the 61 to 90 average

1.8 above the 81 to 10 average

__________________________________________

Current high this month 17.1 to the 1st

[Roger J Smith]

21 hours ago, snowray said:

So we have a well above average first half of the year Jan-June, I wonder if we could end up with a colder than average 2nd half? Charts for July looking consistently bleak.

No year so far in CET records has seen a colder second half of the year than first half (I understand what you were saying about relative to normal), although some first half values have been higher than second half values of other years fairly close to them in time.

Maybe this year we can break that 358-year drought. Here's why second half is always warmer in each calendar year, the second listed months have to be colder than the first listed months on average, the actual average difference (over the entire period of record) is shown in brackets.

DEC - JAN (0.8)

NOV - FEB (2.2)

OCT - MAR (4.4)

SEP - APR (5.4)

AUG - MAY (4.4)

JUL - JUN (1.7)

As you can see that's not a fair fight but 1846 managed to come within 1.5 deg of tying both halves of the year (9.5, 10.8). The first half of 2007 is the warmest on record at 9.70. This month needed to finish 17.1 to tie that mark. Using 16.0 as a provisional it comes out to 9.52, same as 1846 and tied for second warmest. The last time a second half was cooler than 9.6 was 1892 (9.5) so the chances are slim to none probably. These are the five closest attempts to match first half with second half ...

YEAR ___ 1st half __ 2nd half ___ difference

1676 ____ 8.58 _____ 9.08 ____ 0.50

1896 ____ 8.89 _____ 9.79 ____ 0.90

1683 ____ 8.00 _____ 8.92 ____ 0.92

1822 ____ 9.48 ____ 10.62 ____ 1.14

1796 ____ 8.43 _____ 9.60 ____ 1.17

 

In more recent years, 1993 managed 1.28 (8.85, 10.13) and that ranks sixth slightly ahead of 1878 and 1840. Those were the only years to come within 1.4 deg of matching .

If we then look at the first half of the following year and how that compares with the second half of any given year, a slightly smaller difference is found (0.48 in 1845 (second half) to 1846 (first half) ) from 10.0 to 9.52. Waiting two years, that same first half was only 0.1 colder than 1844's second half (9.62). 

As to the differential from year A first half to year B second half, first half of 1828 (8.83) and second half of 1829 (9.15) provide the smallest difference, 0.32 deg.

The closest separation of a colder second half two years after any first half in time was 0.09 (9.33 to 9.42) from first half 1686 to second half 1688.

The closest separation in time that provides a warmer first half appears to be second half 1683 to first half 1686, the increase there was 0.41 (8.92 second half 1683 to 9.33 first half 1686). This was the only increase in that set. 

 

 

[Summer Sun]

Confirmed as 16.0c

[The PIT]

Sunny Sheffield finished on 15.7C +1.4C above normal. Rainfall 102.6mm 140% of normal.

[Roger J Smith]

Odd stat here, it would appear that eleven days were higher than the monthly mean (16.0) and nineteen were lower.

The confirmed numbers change one detail in a table I posted earlier, as the 20.0+ interval is extended to five days; only twice (in the daily temp records 1772-2017) has that been outperformed. in 1846 and 1976. 

I will update on the three robot forecasters when I have a bit more time but would say that consensus did quite well compared to the two Norms. 

[Roger J Smith]

J10 will confirm the results but according to my table of forecasts, Relativistic wins top points for June at 16.0.

Consensus at 15.5 finished around 15th place and the two norms (14.5) for recent 30-year periods were well back.

 

At 14.6, the May-June average CET was the warmest since 14.65 in 1992. It was tied 10th warmest overall, behind this pack:

1846 (15.25), 1726,1762,1798 and 1822 (14.9), 1833 (14.85), 1804 and 1970 (14.7), 1992 (14.65)

and tied with 1776, 1788

Notice that most of these happened before 1850. 

At 11.7, it was the warmest March to June average CET. Second warmest there was 11.55 in 1893.

And as noted earlier, it was tied second warmest January to June, with 1846, and behind 2007.

[Man With Beard]

Oh finally! In 2016, I got within 0.5C for 7 months in a row. This time, it's taken my 7 months to get anywhere near within 0.5C!! Think I used up all my luck in one go last year!

[J10]

On 01/07/2017 at 19:13, Roger J Smith said:

J10 will confirm the results but according to my table of forecasts, Relativistic wins top points for June at 16.0.

Consensus at 15.5 finished around 15th place and the two norms (14.5) for recent 30-year periods were well back.

 

At 14.6, the May-June average CET was the warmest since 14.65 in 1992. It was tied 10th warmest overall, behind this pack:

1846 (15.25), 1726,1762,1798 and 1822 (14.9), 1833 (14.85), 1804 and 1970 (14.7), 1992 (14.65)

and tied with 1776, 1788

Notice that most of these happened before 1850. 

At 11.7, it was the warmest March to June average CET. Second warmest there was 11.55 in 1893.

And as noted earlier, it was tied second warmest January to June, with 1846, and behind 2007.

Indeed Relativistic won June getting it spot on on and leads the Seasonal comp, LetItSnow! and Roger J Smith were 0.1c out for June, and are 2nd and 3rd respectively in the seasonal competition.

Overall the 1-2-3 is

1. davehsug (1st last month)
2. The PIT (5th last month, only 0.2c out this month)
3. I remember Atlantic 252 (2nd)

As ever any problems let me know.

June 2017 CET.xlsx

June 2017 CET.pdf

[Roger J Smith]

Updated report for the two recent 30-year normals and Consensus in the scoring.

 

_________________ APRIL 2017 ________ MAY 2017 __________ JUNE 2017

FORECASTER ____ Fx _ Err __ Pts (rk) ___ Fx _ Err __ Pts (rk) ___ Fx _ Err __ Pts (rank)

Normal 1981-2010 __8.5_-0.4 _ 86.8 (9)___ 11.7_-1.5 _ 49.7 (31) __ 14.5_ -1.5 ___ 7.5 (58)

Normal 1987-2016 __8.8_-0.1 _ 98.4 (2) __11.8_-1.4 _ 53.0 (29) ___ 14.5_ -1.5 ___ 7.5 (58)

Consensus ________9.8_+0.9 _57.1 (27)__11.9_-1.3 _ 56.4 (27) ___15.5_ -0.5 __ 72.4 (18)

(previous stats below)

_________________ DEC 2016 ________  JAN 2017 ________  FEB 2017 _______ MAR 2017

FORECASTER ____ Fx _ Err _ Pts (rk) __ Fx _ Err _ Pts (rk) __ Fx _ Err _ Pts (rk) __ Fx _ Err _ Pts (rk)

Normal 1981-2010 __4.6_-1.4_79.6 (21) __4.4_+0.4_69.9 (29) __4.4_-1.7_56.1 (36) __ 6.6_-2.1_59.0 (30)

Normal 1987-2016 __5.0_-1.0_89.8 (11) __ 4.6_+0.6_58.0 (40) __ 4.8_-1.3_67.4 (27) __ 6.7_-2.0_61.8 (28)

Consensus (median)_3.9_-2.1_50.0 (50)__ 3.9 _--0.1_95.7 (5) __ 4.3 _-1.8_52.4 (39) __ 6.5_-2.2_56.2 (32)

As explained earlier, ranks are equal to the first actual forecast submitted of equal error, in June this was always the same value as the mirror images (16.5 for Consensus, 17.5 for the two normal 30-year values) were not predicted by anyone. The last submitted entry of equal error for Consensus (scoring 18th above) was 24th so this gives you some idea of the advantage these robots gain from being around from the very beginning. (in this case, ten points)

Annual rankings against only those who have entered all six months ... (in brackets, including those who have missed one month).

 

FORECASTER ________ TOTAL POINTS __ Rank ______ Average abs error __ Rank

1981-2010 normal _________ 408.6 _______ 14th (15th) _____ 1.29 ________ t 12th (t 15th)

last 30 years average ______ 435.9 ________ 8th ( 9th) ______ 1.13 _________t 7th (t 9th)

Consensus (median) _______ 440.2 ________8th ( 9th)_ _____ 1.27_________ t 12th (t 15th)

This is the first month that our consensus has overtaken the two normals in overall scoring. The last thirty years dropped from 2nd overall in total points after May, to 8th after June. Consensus gained from 12th overall to 8th (same rank as 1987-2016 because no actual forecasters are between 435.9 and 440.2, to reach 7th they both need to overtake 457.2, while to get ahead of the one six-forecast total ranking ahead of them, 450.7.