Updated CET data base v2.0 posted (9 May 2022) -- an ongoing analysis of changes made to what they now call the "legacy" data base

[Roger J Smith]

Now with the fourth update on high max and low min record adjustments, very few have changed in terms of years, a bit of movement in the actual values here and there. ...

 

High Maximum 1878-2022 _ Legacy vs v2.0 ___ Low Minimum 1878-2022 _ Legacy vs v2.0

DATE ___ Legacy High max __ v2.0 High max___ [] __Legacy Low min __ v2.0 Low min, year

 

01 Jul __ 32.7 _ 2015 ______ 32.8 _ 2015 _______ ___ 5.7 _ 1888 ______ 5.4 _ 1888

02 Jul __ 31.8 _ 1976 ______ 31.7 _ 1976 _______ ___ 6.4 _ 1921 ______ 6.5 _ 1938 (7.0 _ 1921)

03 Jul __ 33.2 _ 1976 ______ 33.1 _ 1976 _______ ___ 6.4 _ 1988 ______ 6.3 _ 1988, 1990

04 Jul __ 31.4 _ 1976 ______ 31.3 _ 1976 _______ ___ 4.7 _ 1965 ______ 4.7 _ 1965

05 Jul __ 31.4 _ 1976 ______ 31.3 _ 1976 _______ ___ 7.2 _ 1898 ______ 7.0 _ 1921 (7.4 _ 1898)

06 Jul __ 31.6 _ 1976 ______ 31.6 _ 1976 _______ ___ 6.3 _ 1954 ______ 6.4 _ 1954

07 Jul __ 30.4 _ 1976 ______ 30.3 _ 1976 _______ ___ 5.5 _ 1954 ______ 5.4 _ 1954

08 Jul __ 30.0 _ 1984 ______ 29.9 _ 1984 _______ ___ 5.2 _ 1929 ______ 5.1 _ 1929 

09 Jul __ 28.8 _ 1921 ______ 28.9 _ 1921 _______ ___ 7.0 _ 1919 ______ 7.1 _ 1919

10 Jul __ 30.1 _ 1921 ______ 30.2 _ 1921 _______ ___ 6.7 _ 1890 ______ 6.7 _ 1890

11 Jul __ 29.2 _ 1983 ______ 29.1 _ 1983 _______ ___ 5.1 _ 1888 ______ 4.7 _ 1888

12 Jul __ 31.1 _ 1923 ______ 31.0 _ 1923 _______ ___ 5.1 _ 1902 ______ 5.1 _ 1902

13 Jul __ 30.6 _ 1923 ______ 30.6 _ 1923 _______ ___ 5.6 _ 1993 ______ 5.4 _ 1993

14 Jul __ 29.6 _ 2003 ______ 29.4 _ 2003 _______ ___ 6.5 _ 1898 ______ 6.4 _ 1898

15 Jul __ 31.1 _ 2003 ______ 30.9 _ 2003 _______ ___ 6.5 _ 1977 ______ 6.4 _ 1977

16 Jul __ 29.3 _ 1983 ______ 29.2 _ 1983 _______ ___ 6.9 _ 2001 ______ 6.9 _ 2001

17 Jul __ 29.6 _ 2006 ______ 29.7 _ 2006*_______ ___ 6.2 _ 1965 ______ 6.2 _ 1965

18 Jul __ 30.9 _ 2006 ______ 30.9 _ 2006*_______ ___ 5.8 _ 1971 ______ 5.8 _ 1971

19 Jul __ 32.9 _ 2006 ______ 33.0 _ 2006*_______ ___ 7.2 _ 1892 ______ 7.2 _ 1892

20 Jul __ 29.6 _ 1901 ______ 29.4 _ 1901 _______ ___ 7.3 _ 1890 ______ 7.4 _ 1890 (7.6 _ 2020)

21 Jul __ 29.7 _ 1989 ______ 29.6 _ 1989 _______ ___ 7.0 _ 2020 ______ 7.2 _ 2020 (7.4 _ 1906)

22 Jul __ 29.4 _ 1989 ______ 29.3 _ 1989 _______ ___ 7.2 _ 1970 ______ 7.2 _ 1970

23 Jul __ 30.2 _ 2019 ______ 30.3 _ 2019 _______ ___ 8.1 _ 1883 ______ 8.2 _ 1932 (8.3_1883, 8.4_2011)

24 Jul __ 28.5 _ 2019 ______ 28.6 _ 2019 _______ ___ 6.9 _ 1973 ______ 7.0 _ 1973

25 Jul __ 34.1 _ 2019 ______ 34.2 _ 2019 _______ ___ 6.8 _ 1920 ______ 7.0 _ 1938 (7.1_1920)

26 Jul __ 30.7 _ 2018 ______ 30.8 _ 2018 _______ ___ 6.9 _ 1884 ______ 6.8 _ 1884

27 Jul __ 29.7 _ 2018 ______ 29.7 _ 2018 _______ ___ 6.9 _ 1882 ______ 7.2 _ 1882, 1963

28 Jul __ 30.8 _ 1948 ______ 30.8 _ 1948 _______ ___ 5.7 _ 1881 ______ 5.6 _ 1881

29 Jul __ 31.6 _ 1948 ______ 31.5 _ 1948 _______ ___ 7.8 _ 1896 ______ 7.4 _ 1896

30 Jul __ 30.5 _ 1948 ______ 30.5 _ 1948 _______ ___ 6.6 _ 1898 ______ 6.7 _ 1898

31 Jul __ 33.2 _ 2020 ______ 33.2 _ 2020 _______ ___ 5.4 _ 2015 ______ 5.6 _ 2015

Note also that some higher values may have occurred  in 1868 and in 1808, 1825 and 1826.

* Higher values in July 2022 are not shown here as this is a comparison thread. But note that v2.0 records were broken for 17th, 18th and 19th with values of 30.0, 34.8 and 37.3 on those three dates

________________________________________________

 

High Maximum 1878-2022 _ Legacy vs v2.0 ____ Low Minimum 1878-2022 _ Legacy vs v2.0

DATE ___ Legacy High max __ v2.0 High max ____[]__Legacy Low min __ v2.0 Low min, year

 

01 Aug __ 31.9 _ 1995 ______ 31.9 _ 1995 _______ ___ 6.2 _ 1976 ______ 6.2 _ 1976

02 Aug __ 32.8 _ 1990 ______ 32.8 _ 1990 _______ ___ 6.5 _ 1966 ______ 6.5 _ 1966

03 Aug __ 33.2 _ 1990 ______ 33.4 _ 1990 _______ ___ 5.2 _ 1912 ______ 5.6 _ 1912

04 Aug __ 31.9 _ 1975 ______ 31.9 _ 1975 _______ ___ 7.3 _ 1934 ______ 7.2 _ 1891 (7.3 _ 1934)

05 Aug __ 29.1 _ 2003 ______ 29.1 _ 2003 _______ ___ 6.8 _ 1956 ______ 6.8 _ 1956

06 Aug __ 28.5 _ 2003 ______ 28.5 _ 2003 _______ ___ 6.1 _ 1987 ______ 6.0 _ 1987

07 Aug __ 31.0 _ 1975 ______ 31.1 _ 1975 _______ ___ 6.4 _ 1961 ______ 6.4 _ 1961

08 Aug __ 30.1 _ 1975 ______ 31.1 _ 1975 _______ ___ 6.8 _ 1955 ______ 6.7 _ 1895 (6.8 _ 1955)

09 Aug __ 32.8 _ 2003 ______ 32.8 _ 2003 _______ ___ 6.7 _ 1931 ______ 6.7 _ 1931

10 Aug __ 30.3 _ 1997 ______ 30.5 _ 1997 _______ ___ 6.0 _ 1892 ______ 6.1 _ 1892

11 Aug __ 30.7 _ 2020 ______ 30.7 _ 2020*_______ ___ 5.1 _ 1892 ______ 4.9 _ 1892

12 Aug __ 32.0 _ 2020 ______ 32.0 _ 2020 _______ ___ 5.0 _ 1912 ______ 4.8 _ 1912

13 Aug __ 30.1 _ 1911 ______ 30.5 _ 1911*_______ ___ 6.3 _ 1993 ______ 6.3 _ 1993

14 Aug __ 28.3 _ 1911 ______ 28.7 _ 1911*_______ ___ 5.2 _ 1885 ______ 4.9 _ 1885

15 Aug __ 28.4 _ 1893 ______ 28.0 _ 1893 _______ ___ 4.4 _ 1885 ______ 4.2 _ 1885

16 Aug __ 29.2 _ 1947 ______ 29.2 _ 1947 _______ ___ 5.9 _ 1966 ______ 5.9 _ 1966

17 Aug __ 29.0 _ 1893 ______ 28.7 _ 1893 _______ ___ 7.0_ 1908, 63, 71 __7.0 _ 1908,63

18 Aug __ 29.5 _ 1893 ______ 29.0 _ 1893 (28.9 _ 1996) _5.7 _ 1946 ______ 5.8 _ 1946

19 Aug __ 30.3 _ 1932 ______ 30.3 _ 1932 (29.7 _ 1997) _5.1 _ 1885 ______ 5.1 _ 1888

20 Aug __ 28.8 _ 1995 ______ 28.8 _ 1995 _______ ___ 4.7 _ 1964 ______ 4.7 _ 1964

21 Aug __ 29.2 _ 1995 ______ 29.2 _ 1995 _______ ___ 5.5 _ 1964 ______ 5.4 _ 1904, 1964

22 Aug __ 29.2 _ 1995 ______ 29.3 _ 1995 _______ ___ 5.3 _ 1964 ______ 5.4 _ 1964

23 Aug __ 28.1 _ 1955 ______ 28.1 _ 1955 _______ ___ 5.9 _ 1903 ______ 5.9 _ 1903

24 Aug __ 28.9 _ 1976 ______ 28.9 _ 1976 _______ ___ 5.5 _ 1966 ______ 5.6 _ 1966

25 Aug __ 29.9 _ 2019 ______ 29.9 _ 2019 _______ ___ 5.9 _ 1904 ______ 5.8 _ 1904 

26 Aug __ 27.2 _ 1930 ______ 27.1 _ 1930 _______ ___ 4.8 _ 1931 ______ 4.8 _ 1931

27 Aug __ 30.2 _ 1930 ______ 30.5 _ 1930 _______ ___ 5.9 _ 1896 ______ 5.9 _ 1896

28 Aug __ 29.8 _ 1942 ______ 29.8 _ 1942 _______ ___ 4.8 _ 1935 ______ 4.7 _ 1935

29 Aug __ 29.6 _ 1961 ______ 29.6 _ 1961 _______ ___ 6.3 _ 1959 ______ 5.9 _ 1979 (6.4 _ 1959)

30 Aug __ 27.7 _ 1886 ______ 26.9 _ 1886 _______ ___ 4.1 _ 1890 ______ 4.1 _ 1890

31 Aug __ 30.4 _ 1906 ______ 29.9 _ 1906 _______ ___ 4.0 _ 1921 ______ 3.7 _ 1921

___________________________________

* A higher value of 30.9 was recorded in 2022 on the 11th, then 31.4 on the 13th and 31.7 on the 14th.

 

Edited August 16, 2022 by Roger J Smith

[Roger J Smith]

Fifth and penultimate listings of the changes (very few) to record high max and low min values. The late September 1895 warm spell gained two additional days at either end with a general upward movement of its already impressive numbers. Some recent cold days that had just missed out in the legacy CET managed to get ties with previous record holders. But otherwise most of the changes have been slight.

 

High Maximum 1878-2022 _ Legacy vs v2.0 ____ Low Minimum 1878-2022 _ Legacy vs v2.0

DATE ___ Legacy High max __ v2.0 High max ____[]__Legacy Low min __ v2.0 Low min, year

 

01 Sep __ 31.0 _ 1906 ______ 31.2 _ 1906 _______ ___ 2.5 _ 1890 ______ 2.6 _ 1890

02 Sep __ 31.3 _ 1906 ______ 31.1 _ 1906 _______ ___ 3.0 _ 1909 ______ 2.9 _ 1909

03 Sep __ 26.3 _ 1906 ______ 26.0 _ 1906 _______ ___ 3.4 _ 1923 ______ 3.7 _ 1923

04 Sep __ 27.9 _ 1880 ______ 28.1 _ 2023a _______ ___ 3.6 _ 1907 ______ 4.1 _ 1907 (4.2_1915, 4.4_2000)

05 Sep __ 26.6 _ 1999 ______ 27.7 _ 2023b __________ ___ 3.6 _ 1915 ______ 3.8 _ 1908 (4.2_1915)

06 Sep __ 25.5 _ 1898 ______ 28.1 _ 2023c _______ ___ 4.1 _ 2015 ______ 4.2 _ 1952, 2015

07 Sep __ 28.3 _ 2021 ______ 28.8 _ 2023d_______ ___ 4.7 _ 1952 ______ 4.2 _ 1952

08 Sep __ 29.6 _ 1911 ______ 29.4 _ 1911e_______ ___ 2.5 _ 1931 ______ 2.4 _ 1931

09 Sep __ 25.9 _ 1895 ______ 28.9 _ 2023f _______ ___ 3.1 _ 1986 ______ 3.0 _ 1986

10 Sep __ 26.5 _ 1959 ______ 27.1 _ 2023g_______ ___ 3.2 _1897, 1944 __ 3.1 _ 1897

11 Sep __ 27.3 _ 1919 ______ 27.2 _ 1919 _______ ___ 2.9 _ 1944 ______ 2.9 _ 1944

12 Sep __ 25.5 _ 1891 ______ 25.4 _ 1891 _______ ___ 2.6 _ 1908 ______ 3.1 _ 1908

13 Sep __ 26.4 _ 2016 ______ 26.3 _ 2016 _______ ___ 2.8 _ 1925 ______ 3.0 _ 1925 (3.1_1908,31)

14 Sep __ 26.4 _ 2016 ______ 26.4 _ 2016 _______ ___ 3.5 _ 1996 ______ 3.4 _ 1996 (3.5_1882)

15 Sep __ 27.4 _ 2020 ______ 27.4 _ 2020 _______ ___ 3.4 _ 1903 ______ 3.4 _ 1903, 1979 (3.5_1882 4.0 _ 1986)

16 Sep __ 26.3 _ 1898 ______ 26.2 _ 1898 _______ ___ 3.2 _ 1903 ______ 2.9 _ 1903 (3.4 _ 1986)

17 Sep __ 27.4 _ 1898 ______ 27.1 _ 1898 _______ ___ 3.1 _ 2005 ______ 3.1 _ 2005

18 Sep __ 25.4 _ 1926 ______ 25.6 _ 1926 _______ ___ 2.9 _ 1952 ______ 2.3 _ 1889

19 Sep __ 27.1 _ 1926 ______ 27.1 _ 1926 _______ ___ 2.0 _ 1986 ______ 2.0 _ 1986

20 Sep __ 22.9 _ 1970 ______ 22.8 _ 1970 _______ ___ 1.4 _ 1952 ______ 1.3 _ 1952

21 Sep __ 27.1 _ 2006 ______ 27.1 _ 2006 _______ ___ 2.2 _ 1964 ______ 2.2 _ 1964

22 Sep __ 22.9 _ 1956 ______ 22.9 _ 1956 _______ ___ 1.8 _ 1911, 79 __ 1.7 _ 1979

23 Sep __ 23.3 _ 2000 ______ 23.4 _ 1895 (23.2_2000) _ 2.2 _ 1914 ______ 2.0 _ 1914

24 Sep __ 25.1 _ 1895 ______ 25.7 _ 1895 _______ ___ 1.5 _ 1878 ______ 2.1 _ 1878

25 Sep __ 24.2 _ 1895 ______ 24.5 _ 1895 _______ ___ 2.1 _ 2018 ______ 2.2 _ 2018

26 Sep __ 25.4 _ 1895 ______ 25.4 _ 1895 _______ ___ 1.1 _ 1885 ______ 1.1 _ 1885

27 Sep __ 24.9 _ 1895 ______ 25.7 _ 1895 _______ ___ 1.1 _ 1885 ______ 1.0 _ 1885

28 Sep __ 24.8 _ 2011 ______ 25.0 _ 1895 (24.7_2011) _0.5 _ 1885, 1919 __ 0.5 _ 1885 (0.7_1919)

29 Sep __ 26.4 _ 2011 ______ 26.4 _ 2011 _______ ___ 0.3 _ 1919 ______ 0.3 _ 1919

30 Sep __ 27.3 _ 2011 ______ 27.2 _ 2011 _______ ___ 0.8 _ 1914 ______ 0.9 _ 1914

_________________

a 4 Sep ... 27.4 _ 1880 previous record

b 5 Sep ... 26.5 _ 1999 previous record also 26.4 in 1949. 

c 6 Sep ... 25.6 _ 1898 previous record

d 7 Sep ... 28.3 _ 2021 previous record 

e 8 Sep ... 28.6 in 2023 failed to break 1911 record

f 9 Sep ... 25.7 _ 1895 previous record

g 10 Sep ... 26.5 _ 1959 previous record

==========================

 

High Maximum 1878-2022 _ Legacy vs v2.0 ____ Low Minimum 1878-2022 _ Legacy vs v2.0

DATE ___ Legacy High max __ v2.0 High max ____[]__Legacy Low min __ v2.0 Low min, year

 

01 Oct __ 27.1 _ 2011 ______ 27.0 _ 2011 _______ ___ 0.1 _ 1928 ______ 0.0 _ 1928

02 Oct __ 24.2 _ 2011 ______ 24.0 _ 2011 _______ ___ 0.1 _ 1888 ______-0.3 _ 1888

03 Oct __ 25.5 _ 1959 ______ 25.4 _ 1959 _______ ___ 0.1 _ 1888 ______-0.3 _ 1888

04 Oct __ 23.7 _ 1959 ______ 23.7 _ 1959 _______ ___-0.5 _ 1912 ______-0.6 _ 1912

05 Oct __ 23.4 _ 1921 ______ 23.9 _ 1921 _______ ___-0.4 _ 1888 ______-0.4 _ 1888

06 Oct __ 24.6 _ 1921 ______ 24.4 _ 1921 _______ ___-0.3 _ 1888 ______-0.4 _ 1888

07 Oct __ 21.5 _ 1965 ______ 21.5 _ 1959  (21.4_1965,2023) _-0.2 _ 1888 _ -0.5 _ 1888

08 Oct __ 23.3 _ 1995 ______ 23.1 _ 1995 _______ ___ 0.7 _ 1912 ______ 0.6 _ 1990 (0.9_1912)

09 Oct __ 24.0 _ 1921 ______ 23.9 _ 1921 _______ ___ 0.8 _ 1925 ______ 1.1 _ 1925

10 Oct __ 22.8 _ 2018 ______ 22.6 _ 2018 _______ ___-0.1 _ 1925 ______ 0.0 _ 1925

11 Oct __ 23.0 _ 1978 ______ 22.8 _ 1978 _______ ___-1.0 _ 1952 ______-1.0 _ 1952

12 Oct __ 22.3 _ 1978 ______ 22.0 _ 1978 _______ ___-0.9 _ 1887 ______-1.2 _ 1887

13 Oct __ 22.0 _ 2018 ______ 21.8 _ 2018 (21.7_1990)__-0.7 _ 1964 ______-0.7 _ 1964

14 Oct __ 19.2 _ 1990, 2017__ 19.0 _ 1990, 2017 ___ ___ 0.3 _ 1917 ______ -0.3 _ 1888 (0.5_1917)

15 Oct __ 19.1 _ 1908 ______ 18.8 _ 1908 _______ ___-1.5 _ 1917 ______-1.3 _ 1917

16 Oct __ 20.5 _ 2017 ______ 20.4 _ 2017 _______ ___-2.7 _ 1993 ______-2.8 _ 1993

17 Oct __ 19.1 _ 1969 ______ 19.2 _ 1969 _______ ___-2.2 _ 1905 ______-2.4 _ 1881 (-2.2_1905)

18 Oct __ 21.8 _ 1921 ______ 21.7 _ 1921 _______ ___-2.2 _ 1955 ______-2.3 _ 1955

19 Oct __ 20.2 _ 1997 ______ 20.1 _ 1997 _______ ___-3.0 _ 1926 ______-2.9 _ 1926

20 Oct __ 18.5 _ 1969 ______ 18.5 _ 1969 _______ ___-1.6 _ 1905 ______-1.6 _ 1905

21 Oct __ 18.9 _ 1968 ______ 18.8 _ 1968 _______ ___-2.2 _ 1935 ______-2.3 _ 1935

22 Oct __ 19.4 _ 1971 ______ 19.3 _ 1971 _______ ___-2.8 _ 1931 ______-2.9 _ 1931 (-2.7_1905)

23 Oct __ 18.7 _ 2011 ______ 18.6 _ 2011 _______ ___-2.6 _ 1951 ______-2.7 _ 1951

24 Oct __ 18.2 _ 1995 ______ 18.1 _ 1995 _______ ___-3.7 _ 1895 ______-3.6 _ 1895

25 Oct __ 17.8 _ 2013 ______ 17.7 _ 2013 _______ ___-1.5 _ 1895, 1931__-1.5 _ 1931 (-1.4 _ 1895 -1.3 1887)

26 Oct __ 17.0 _ 2005 ______ 17.7 _ 2022*(16.8 _ 1888) _-3.6 _ 1887 ______-4.0 _ 1887

27 Oct __ 19.6 _ 2005 ______ 19.6 _ 2005 (18.7 _ 1888) _-3.9 _ 1931 ______-3.9 _ 1931

28 Oct __ 18.1 _ 2014 ______ 17.9 _ 2014 (17.4 _ 1888) _-3.8 _ 1895 ______-3.5 _ 1895

29 Oct __ 17.2 _ 1908 ______ 19.5 _ 2022*_______ ___-3.9 _ 1895*______-4.0 _ 1895, 1997

30 Oct __ 17.8 _ 2005 ______ 17.8 _ 2005 _______ ___-1.9 _ 1895 ______-1.9 _ 1895 (-1.7_1988)

31 Oct __ 20.6 _ 2014 ______ 20.5 _ 2014 _______ ___-3.2 _ 1881 ______-3.3 _ 1881

26th_ * 17.1 (2005) had been the record previously.

29th _ * 17.5 (2005) had been the record previously.

============================================

 

Edited October 8, 2023 by Roger J Smith
2023 entered tie (1965) 7 Oct _ 0.1 below record

[Roger J Smith]

Final segment shows the changes to November and December daily record values ... Nov-Dec 2010 and 2015 have held on to all of their large share of the daily records. Changes are infrequent in general. While 1904 was dethroned as coldest mean daily, it retains coldest minimum on the 24th. That value was raised by 1.0 C however. Late Dec 2000 was rendered several degrees colder and this has moved 31 Dec 2000 into the record table (tied with 1892 despite its own slight colder turn). I have included a few 2010 readings that are only "recent" lows as well, the smaller type face is for secondary values that are not quite record values. 

 

High Maximum 1878-2022 _ Legacy vs v2.0 ____ Low Minimum 1878-2022 _ Legacy vs v2.0

DATE ___ Legacy High max __ v2.0 High max ____[]__Legacy Low min __ v2.0 Low min, year

 

01 Nov __ 16.7 _ 2020 ______ 16.6 _ 2020 _______ ___-3.8 _ 1926 ______-4.3 _ 1926

02 Nov __ 17.7 _ 1927 ______ 17.6 _ 2005 _______ ___-4.0 _ 1880 ______-4.2 _ 1880

03 Nov __ 16.8 _ 2011 ______ 16.7 _ 2011 _______ ___-2.2 _ 1985______ -3.9 _ 1985

04 Nov __ 18.7 _ 1946 ______ 18.7 _ 1946 _______ ___-2.4 _ 1901 ______-2.1 _ 1880,1896

05 Nov __ 17.6 _ 1938 ______ 17.7 _ 1938 _______ ___-3.6 _ 1930 ______-3.4 _ 1930

06 Nov __ 16.2 _ 2015 ______ 16.2 _ 2015 (16.1_1972)__-3.6 _ 1901 ______-3.8 _ 1901

07 Nov __ 16.1 _ 2015 ______ 16.1 _ 2015 _______ ___-2.6 _ 1896 ______-2.5 _ 1896

08 Nov __ 15.5 _ 1998 ______ 15.2 _ 1983, 1985, 1998__-4.1 _ 1923_____-4.0 _ 1923

09 Nov __ 15.9 _ 1977 ______ 15.6 _ 1977 _______ ___-5.4 _ 1921 ______-4.8 _ 1921

10 Nov __ 16.2 _ 1977 ______ 16.0 _ 1977, 2015 ___ ___-4.7 _ 1908 ______-4.5 _ 1908

11 Nov __ 15.3 _ 1938 ______ 15.9 _ 2022* _______ ___-3.0 _ 1921 ______-2.6 _ 1921

12 Nov __ 16.3 _ 1938 ______ 16.3 _ 1938 _______ ___-5.1 _ 1919 ______-5.3 _ 1919

13 Nov __ 16.1 _ 1938 ______ 16.2 _ 1938 _______ ___-4.4 _ 1979 ______-4.8 _ 1921 (-4.6_1979)

14 Nov __ 14.8 _ 1994, 2015__ 14.7 _ 1938, 1994, 2015 ___-4.1 _ 1925 ______-4.0 _ 1925, 1965

15 Nov __ 15.4 _ 2015 ______ 15.4 _ 2015 _______ ___-4.7 _ 1965 ______-4.7 _ 1965

16 Nov __ 15.7 _ 1895 ______ 15.8 _ 1895 _______ ___-6.4 _ 1901 ______-5.8 _ 1901

17 Nov __ 14.8 _ 1997 ______ 14.7 _ 1997 (14.6 _ 2015)_-6.1 _ 1901, 30 __ -6.1 _ 1901 (-5.8_1930)

18 Nov __ 15.2 _ 1978 ______ 14.9 _ 1978 _______ ___-4.6 _ 1929 ______-4.4 _ 1882

19 Nov __ 15.6 _ 1994 ______ 15.4 _ 1994 _______ ___-2.6 _ 2005 ______-2.7 _ 2005 

20 Nov __ 15.3 _ 1947 ______ 15.3 _ 1947 _______ ___-3.7 _ 1971 ______-3.8 _ 1971

21 Nov __ 15.7 _ 1947 ______ 15.7 _ 1947 _______ ___-3.9 _ 1880 ______-4.1 _ 1914 (-4.0_1880)

22 Nov __ 15.4 _ 1906 ______ 15.4 _ 1906 (15.3_1947) _-5.0 _ 1988 ______-5.1 _ 1988 (-4.9_1880)

23 Nov __ 14.3 _ 1947 ______ 14.4 _ 1947 _______ ___-6.6 _ 1983 ______-6.7 _ 1983*

24 Nov __ 14.0 _ 1980 ______ 13.8 _ 1916, 1980___ ___-8.8 _ 1904 ______-7.8 _ 1904

25 Nov __ 13.8 _ 1983 ______ 13.7 _ 1983 (13.6_1970) _-5.6 _ 1952 _____ -5.9 _ 1923 (-5.7_1952, -5.5_1904)

26 Nov __ 14.5 _ 1979 ______ 14.3 _ 1979 _______ ___-5.1 _ 1989 ______-5.3 _ 1989

27 Nov __ 13.6 _ 2000, 2006 __13.5 _ 2006 (13.4_2000)_ -6.3 _ 1915 ______-6.3 _ 1923 (-6.1_1904, 5.8_1915)

28 Nov __ 14.7 _ 2000 ______ 14.4 _ 2000 _______ ___-7.0 _ 2010 ______-7.0 _ 2010

29 Nov __ 14.0 _ 1939 ______ 14.0 _ 1939 _______ ___-6.3 _ 2010 ______-6.3 _ 2010

30 Nov __ 14.3 _ 2001 ______ 14.1 _ 2001 _______ ___-5.8 _ 1919 ______-5.5 _ 1919 (-4.5 _ 2016)

______________________________ Notes:

* 11th _ 2022 (15.9) replaced 15.5 in v2.0 (1938).

* 23rd _ Minimum possibly lower in 1858 (daily mean -3.9) before daily minima began.

==============================================

High Maximum 1878-2022 _ Legacy vs v2.0 ____ Low Minimum 1878-2022 _ Legacy vs v2.0

DATE ___ Legacy High max __ v2.0 High max ____[]__Legacy Low min __ v2.0 Low min, year

 

01 Dec __ 13.8 _ 1985 ______ 13.6 _1985 (13.4_1939) _ -6.4 _ 1947 ______-6.4 _ 1947 (-5.2_1973)

02 Dec __ 14.4 _ 1985 ______ 14.3 _ 1985 (14.2_1954) _-7.1 _ 1879 ______-7.1 _ 1879

03 Dec __ 13.7 _ 1953 ______ 13.7 _ 1953 _______ ___-7.5 _ 1879, 2010 _-7.8 _ 1879 (-7.5_2010)

04 Dec __ 13.8 _ 1888, 1979 __ 13.7 _ 1888 (13.6_1979) _-6.7 _ 2010 ______-6.8 _ 2010

05 Dec __ 13.9 _ 1898 ______ 14.0 _ 1898 _______ ___-7.8 _ 1879 ______-8.3 _ 1879

06 Dec __ 14.0 _ 2007 ______ 14.0 _ 2007 _______ ___-8.1 _ 1879 ______-8.3 _ 1879 (-5.8_2010)

07 Dec __ 14.0 _ 2015 ______ 14.0 _ 2015 _______ __ -11.4 _ 1879_____-11.0_ 1879 (-6.6_2010)

08 Dec __ 13.8 _ 1964 ______ 13.8 _ 1964 _______ ___-7.5 _ 1879 ______-6.9 _ 2010

09 Dec __ 13.5 _ 2016 ______ 13.5 _ 2016 _______ ___-5.6 _ 1967 ______-5.6 _ 1967 (-4.8 _ 2010)

10 Dec __ 14.4 _ 1994 ______ 14.3 _ 1994 _______ ___-6.8 _ 1981______ -7.0 _ 1981

11 Dec __ 14.5 _ 1994 ______ 14.3 _ 1994 _______ ___-6.7 _ 1879*_____ -7.4 _ 1882 (-6.1_1879)

12 Dec __ 13.9 _ 1994 ______ 13.8 _ 1994 _______ __-13.2 _ 1981 ____ -13.5 _ 1981

13 Dec __ 13.5 _ 1972 ______ 13.5 _ 1972 _______ __-15.9 _ 1981_____ -16.1 _ 1981

14 Dec __ 13.9 _ 1912 ______ 13.6 _ 1972, 1998_ ___ -8.4 _ 1890 ______-8.6 _ 1878 (-8.5_1899, -7.7_1890)

15 Dec __ 13.2 _ 2015 ______ 13.2 _ 2015 _______ ___ -8.4 _ 1878 ______-9.3 _ 2022*

16 Dec __ 13.9 _ 2015 ______ 13.9 _ 2015 _______ ___ -5.5 _ 1899 ______ -8.0 _ 2022* 

17 Dec __ 14.1 _ 2015 ______ 14.1 _ 2015 _______ ___ -7.3 _ 1878 ______ -7.8 _ 1878*

18 Dec __ 15.1 _ 2015 ______ 15.1 _ 2015 _______ ___ -7.2 _ 1981 ______ -7.5 _ 1981*

19 Dec __ 15.3 _ 2015 ______ 15.3 _ 2015 _______ __ -10.3 _ 2010 _____ -10.4 _ 2010 (-9.0 _ 1981)

20 Dec __ 13.9 _ 1971 ______ 13.9 _ 1971 _______ __ -12.4 _ 2010 _____ -12.5 _ 2010

21 Dec __ 13.7 _ 2015 ______ 13.7 _ 2015 _______ __ -10.6 _ 2010 _____ -10.6 _ 2010 

22 Dec __ 13.8 _ 2015 ______ 13.7 _ 2015 _______ __ -12.6 _ 1890 _____ -11.7 _ 1890

23 Dec __ 14.7 _ 1977 ______ 14.5 _ 1977 _______ ___-8.4 _ 1935 ______ -8.4 _ 1935

24 Dec __ 13.0 _ 1997 ______ 13.4 _ 2023*_______ ___-9.8 _ 1878 _____ -11.1 _ 1878

25 Dec __ 13.1 _ 2015, 16 ___13.2 _ 2016 _______ __-11.2 _ 1878 _____-11.8 _ 1878* (-9.3 _ 2010)

26 Dec __ 13.8 _ 2015 ______ 13.8 _ 2015 _______ ___ -9.7 _ 2010 ______-9.7 _ 2010 (-8.0 _ 1962)

27 Dec __ 13.9 _ 1974 ______ 13.7 _ 1974 _______ ___ -9.1 _ 1892 ______-9.2 _ 1892 (-6.7 _ 2010)

28 Dec __ 14.3 _ 1974 ______ 14.2 _ 1974 _______ ___ -8.1 _ 1892 ______-7.9 _ 1892

29 Dec __ 13.8 _ 1925,2021 ___13.9 _ 2021 _______ ___ -9.2 _ 1964 ______-9.3 _ 1964 (-8.2_1995)

30 Dec __ 13.9 _ 2021 ______ 14.0 _ 2021 _______ __ -10.7 _ 1908 ______-9.5 _ 1908 (-8.3_2000)

 

31 Dec __ 14.0 _ 2021 ______ 14.1 _ 2021 _______ ___ -6.4 _ 1892 ______-6.8 _ 1892, 2000

Note:

* 15 Dec _ -9.3 2022 replaced -8.3 1878.

* 16-18 Dec _ -8.0 2022 replaced -5.6 1899 on 16th; 1859 likely had lower minima around -10 to -12 16th-18th

* 24 Dec _ 2023 (13.4) replaced 1997 (12.7) in v2.0. It was 13.0 in CET legacy. 

* 25 Dec _ 1796 likely around -15 as daily mean was -10.8. Also some lower minima likely in Dec 1798, 1844, 1859 (see daily means)

Edited December 25, 2023 by Roger J Smith
new high max Dec 24, 2023 (13.4 C)

[Roger J Smith]

On 21/05/2022 at 05:52, SummerShower said:

Does anyone know the rationale behind the database being changed?  And is this new version now 'the official' one, or merely just an alternative data set?

I guess the new data set doesn't adjust figures downwards for urban warming (hence many recent figures being corrected upwards) but I believe there is more to it.

Just interested to know really  

The rationale is at least partly explained in their news release which was linked in the thread in the first few posts. They wanted to have fewer changes in station selections from year to year and did some work on comparing stations used in the legacy CET with stations they inserted as substitutes in this version (stations that were later used in the legacy CET). The changes are not all upward, I would say they are mixed but more down than up from about 1970 to 2004 then more up than down since, although 2005 and 2011 seem to be the main beneficiaries of any warming in terms of additional records acquired. 

As to the urban corrections, those were standard before and remain the same overall, the changes are more due to changes in the stations blended together in different eras. As I've noted several times, there is no change to the data base at all up to 1852 other than the addition of daily data for previously missing Dec 1786. Different rounding protocols seem to have been used, however, since a couple of months as far back as 1788 changed by 0.1 despite showing no changes in daily data. 

Now on the subject of "official" data, the MetOffice will clearly want these v2.0 numbers to be used as official data replacing the legacy set. And the legacy set is unlikely to be updated going forward, I would think. It will terminate with April 2022. What happens around net-weather and other weather enthusiast sites is pretty much up to individuals posting here (or there). Some people are going to continue to reference the legacy CET records and others will make an effort to look at the v2.0 records. Some may even go forward unaware that this change has been made and I'm sure we will see things for years to come like "What happened to the CET?" and "what is v2.0 and what is legacy CET?" as one by one people with a small interest find out something happened. 

In the grand scheme of things, these are fairly minor changes. I found very few of the high max and low min dates for records changing, despite some fairly robust alterations. For example both 1890 and 1904 saw daily records become less extreme by about 1 C deg but as there was no close competition in the legacy set on the dates involved, the years kept their daily records despite those changes. 

If anyone from the MetOffice happens to read this thread, we would welcome participation, I don't feel like I can or should be justifying any of the changes, I am just reporting on them for general interest. It would be good to have an insider answer some of these questions of interest. But to those asking, please find that document that I mentioned and read through it, that may answer some of your questions too. 

[Roger J Smith]

The attached graph shows the number of daily maximum (CET) records for each year. The ties are counted as 0.5 or where three years tied (this happened twice in November) 0.33. The year with the highest count was 2015 (15.83) and the pace picks up considerably around 1989. Before that, peaks can be seen around the 1890s, 1916 to 1922, 1930s and 1940s, and a prominent spike for 1976 which had a count of 12.0. The data interval for this is 1878 to 2022. The last data point (3.0) is for 2022 and covers only 40% of a year so pro-rated it should be taken as equivalent to 7.5. Top years include 2015 (15.83), 2019 (13.0), 2020 (12.5), 1976 (12.0), 1990, 2011 and 2016 (each 10.0) ... half of the daily maximum records were set after 1985 in the last 38 of the 145 years.

Edited May 27, 2022 by Roger J Smith

[Midlands Ice Age]

2 hours ago, Roger J Smith said:

The attached graph shows the number of daily maximum (CET) records for each year. The ties are counted as 0.5 or where three years tied (this happened twice in November) 0.33. The year with the highest count was 2015 (15.83) and the pace picks up considerably around 1989. Before that, peaks can be seen around the 1890s, 1916 to 1922, 1930s and 1940s, and a prominent spike for 1976 which had a count of 12.0. The data interval for this is 1878 to 2022. The last data point (3.0) is for 2022 and covers only 40% of a year so pro-rated it should be taken as equivalent to 7.5. Top years include 2015 (15.83), 2019 (13.0), 2020 (12.5), 1976 (12.0), 1990, 2011 and 2016 (each 10.0) ... half of the daily maximum records were set after 1985 in the last 38 of the 145 years.

Roger 

An interesting graph..

To set my mind straight could you provide the above alongside the same legacy graph.

If poss it would also be possible to show the differences  graph?.

That would highlight where the changes have occured - to the maxs anyway.

Any chance of the same graphs for the mins also??

MIA

[Roger J Smith]

Yes, I can work those up, the legacy version of daily max is not as different as the daily means because for the maxima (and minima) we are only looking at 1878 to present, and a lot of the mean daily transfers were away from the years before 1878 which would not apply here. Still, there was a slight difference. It won't look like much on a graph. As to the daily minima, what I posted earlier is my first draft and I have been working through the tables using a more thorough search function, finding one or two changes to make. So it will be a few days before I can post a final graph of the daily minima. 

The comparison graph for the maxima will be along shortly, I think, as I just need to create the graph, the numbers are already in my file. 

[Roger J Smith]

So here's a graph of the differentials. The positive numbers in the graph are gains from legacy to v2.0, the negative values are losses (to legacy totals). The basic analysis is that the middle portion of the record, from about 1900 to 1940, saw the greatest attrition, while the more recent era saw some gains, as did 1895 and 1898 oddly enough. 

These are quite often changes to ties,  either a loss of a share or a gain. You can see how many are of that origin just by checking the counts at 0.5 and -0.5 although that is not perfect because a year could have more than one. In fact both 1980 and 1990 ended up neutral after three record changes for each year. They lost and gained equally. Just as with the mean daily records, 2005 was the overall net winner, going from 5.5 legacy records to 8.0 v2.0 records. Its victims were 1908, 1927 and pushing into a 2020 record for a tie. The two outright takeovers were in the late October to early November warm spell in 2005. 1895 gained two days at either end of its already impressive late September run, improving that from four to six days, at the expense of 2000 and 2011. A few years were extinguished from the records such as 1884, 1887, 1903 and 1912 (they had legacy records or at least a tie, and lost them to v2.0, usually to recent years). Some other outright swaps are neither here nor there as to climate change issues, for example, 1959 grabbed a 1965 record in October. The colder outcome in late 2000 shows up also with a loss of 1.5 legacy records for that year. 

Anyway, that's how it looks and I would describe it as a modest shift, I said earlier that the midpoint of the 366 daily records was only 37 years ago, in the legacy set it was 43 years ago. So I don't find the v2.0 records to be a big shift, although the rich got richer and the poor got poorer, generally speaking. 

[LetItSnow!]

I’m interested to know which years overall got warmest compared to their original value and which years got coolest. What I mean for example, 2000 used to show it averaging at 10.3something and now it’s 10.19. I’m interested if there were any overall years that gained or lost overall temperature in a notable way. 

[Midlands Ice Age]

10 hours ago, Roger J Smith said:

So here's a graph of the differentials. The positive numbers in the graph are gains from legacy to v2.0, the negative values are losses (to legacy totals). The basic analysis is that the middle portion of the record, from about 1900 to 1940, saw the greatest attrition, while the more recent era saw some gains, as did 1895 and 1898 oddly enough. 

These are quite often changes to ties,  either a loss of a share or a gain. You can see how many are of that origin just by checking the counts at 0.5 and -0.5 although that is not perfect because a year could have more than one. In fact both 1980 and 1990 ended up neutral after three record changes for each year. They lost and gained equally. Just as with the mean daily records, 2005 was the overall net winner, going from 5.5 legacy records to 8.0 v2.0 records. Its victims were 1908, 1927 and pushing into a 2020 record for a tie. The two outright takeovers were in the late October to early November warm spell in 2005. 1895 gained two days at either end of its already impressive late September run, improving that from four to six days, at the expense of 2000 and 2011. A few years were extinguished from the records such as 1884, 1887, 1903 and 1912 (they had legacy records or at least a tie, and lost them to v2.0, usually to recent years). Some other outright swaps are neither here nor there as to climate change issues, for example, 1959 grabbed a 1965 record in October. The colder outcome in late 2000 shows up also with a loss of 1.5 legacy records for that year. 

Anyway, that's how it looks and I would describe it as a modest shift, I said earlier that the midpoint of the 366 daily records was only 37 years ago, in the legacy set it was 43 years ago. So I don't find the v2.0 records to be a big shift, although the rich got richer and the poor got poorer, generally speaking. 

Thanks Roger..

It does show that there were  more reductions in the Legacy in the earlier years. whereas there are more new records set in the new version in the later (50?) years. 

The new minimum record anomalies will also be interesting.

MIA

[Roger J Smith]

 

Here's an answer to your questions about overall temperature changes. This is not precisely the answer to changes in data base annual means because I don't know if the legacy and v2.0 methods are precisely the same. There are three ways you can derive an annual mean temperature. One is to average the twelve months regardless of their different lengths. Or you can average the annual average max and min (which might involve the same question). Or you can take an average of the 365 or 366 days. 

The differentials should be very similar for any method used since all the variables introduced will tend to average out to zero. 

I averaged the mean daily temperatures of the days of the year to get these differentials. Since legacy has two decimal annual means and v2.0 has one decimal, I don't see much point in taking differentials of those, who knows if the 10.24 of one set and the 10.3 of the other set are .06 different or anywhere from .01 to .11 different (which they could be in theory). 

As mentioned there are no changes to the data base daily values before 1853 (other than the introduction of daily data for Dec 1786 which did not change the average cited before for that month). So this comparison begins with 1853. I have arranged this graphically so you can just scroll up and down and see the trends. Years that have warmed up from legacy to v2.0 are further to the right in the column, and those that have lost a tenth or two of a degree on average in v2.0 are to the left. The actual values of the differentials are shown. Now to cut down on the size of this graphic, I noted that from 1853 to 1973, the changes are all very small (all within the range -.02 to +.02, most -.01 or +.01 if not .00 which is very common by the decades 1931-70) except that 1860 drops -.04 and that is because a larger downward correction was applied to the month of December 1860. So there is little point in displaying this very uniform set of very small changes before the interval where the changes become significant. Three lines of the graphic therefore cover 1853-59, 1860, and 1861-1973 and are not to scale. The overall average for the earliest interval 1853-59 is very slightly negative (-.01) meaning that there are a few more days adjusted down than up. If you look at the daily differentials you find that about two-thirds of the days are left alone and a slight majority of the other third go down some small amount (mostly) and the rest go up slightly. There is even less change applied from 1861 to 1973 and many of those years averaged .00 change. But starting in 1974 the changes become visually quite a bit larger as the graphic shows. 

Although the years before 1974 have small average changes, there are intervals here and there with larger changes but they always seem to balance out (except in 1860). For example, the warm spell in late September 1895 was adjusted somewhat higher and as it was competing with later years that received downward adjustments, it managed to extend a run of four record high maxima to six. So there was a case opposite what might be perceived as a trend towards boosting recent warmth at the expense of "legacy" warmth. However, despite the overall slight changes in years around 1901-30 (on average for the years) there seemed to be a tendency to downgrade some of the warmer means and upgrade some of the colder ones, both of which cut into the number of extreme records set in those years. So it's a bit complicated overall.

 

GRAPHIC PRESENTATION of CHANGES from LEGACY mean annual to v2.0 mean annual

<< V2.0 colder _ neutral _ V2.0 warmer >>

_______________ 1853-59 (-.01) __________________________________________

________ 1860 (-.04) ______________________________________________________

_________________ 1861-1973 (.00) ______________________________________

____ 1974 (-.08) ____________________________________________________________

___ 1975 (-.09) _____________________________________________________________

____ 1976 (-.09) ____________________________________________________________

__ 1977 (-.10) ______________________________________________________________

_ 1978 (-.11) _______________________________________________________________

_ 1979 (-.11) _______________________________________________________________

____ 1980 (-.08) _____________________________________________________________

____ 1981 (-.08) _____________________________________________________________

______ 1982 (-.06) ___________________________________________________________

________ 1983 (-.04) _________________________________________________________

________ 1984 (-.04) _________________________________________________________

________ 1985 (-.04) _________________________________________________________

______ 1986 (-.06) ___________________________________________________________

_______ 1987 (-.05) _________________________________________________________

_____ 1988 (-.07) ____________________________________________________________

____ 1989 (-.08) ______________________________________________________________

____ 1990 (-.08) ______________________________________________________________

____ 1991 (-.08) ______________________________________________________________

____ 1992 (-.08) ______________________________________________________________

______ 1993 (-.06) ______________________________________________________________

____ 1994 (-.08) ______________________________________________________________

____) 1995 (-.07) ______________________________________________________________

____ 1996 (-.07) ______________________________________________________________

____ 1997 (-.08) ______________________________________________________________

____ 1998 (-.09) ______________________________________________________________

____ 1999 (-.08) ______________________________________________________________

** 2000 (-.11) ______________________________________________________________ (Dec 2000 -0.5)**

________ 2001 (-.04) __________________________________________________________

_____ 2002 (-.07) _____________________________________________________________

_____ 2003 (-.07) _____________________________________________________________

______ 2004 (-.06) ____________________________________________________________

__________ 2005 (-.02) ________________________________________________________

_______________ 2006 (+.03) ___________________________________________________

_______________ 2007 (+.03) ___________________________________________________

_______________ 2008 (+.03) ___________________________________________________

_______________ 2009 (+.03) ___________________________________________________

________________ 2010 (+.04) __________________________________________________

________________ 2011 (+.04) __________________________________________________

________________ 2012 (+.03) __________________________________________________

_________________ 2013 (+.03) _________________________________________________

__________________ 2014 (+.04) ________________________________________________

__________________ 2015 (+.03) ________________________________________________

__________________ 2016 (+.03) ________________________________________________

__________________ 2017 (+.03) _________________________________________________

__________________ 2018 (+.03) _________________________________________________

___________________ 2019 (+.03) ________________________________________________

__________________ 2020 (+.02) _________________________________________________

_______________________ 2021 (+.06) ___________________________________________

______________________________ 2022 (+.12) ____________________________________

As noted for 2000, the annual average changes sometimes mask brief intervals of a different trend. Dec 2000 saw a more substantial drop of about 0.5 C and the cold days at the end of that month dropped even more. So the slightly larger annual drop was due to the larger decreases made in December. The change in trend came around 2005-06, when the years began to warm slightly relative to the interval 1974-2004. As it turned out some fairly substantial warmings were applied to several record warm intervals in 2005 but I didn't note the same changes after 2005 ended, the records established in most other years 2006-22 were left about as they were with a few going up by 0.1 or 0.2, some staying level. The larger change for 2022 so far may be a function of the months we have averaged, maybe the annual average would be back down closer to 2021, but as legacy readings will probably not exist from 1st May on, we won't have any way to measure that. 

In overall terms these are slight changes to the data base, the rationale behind the larger changes made in two or three short intervals would have to come directly from source, I don't have any insight into what made Dec 1860, Dec 2000 and June as well as late Oct-early Nov 2005 worthy of larger changes that affected records, or why the Nov 1904 and Feb 1917 cold spells were more significantly warmed than the rest of those years. All of the decisions were probably "automatic" in that they came from the decisions made about how to treat data from various stations, I rather doubt that it was as direct as a committee saying "we need to change these record values." Indeed, it's our fascination with records that exposed these changes and very likely other people working with CET data would not necessarily care about them or note them, for some, the overall average changes are probably all that would matter. And those are by and large insignificant. 

(For those longer intervals that I blended, these are the years that actually have two-decimal change values, from 1853 to 1973, if a year in that interval is not mentioned here, its average change was .00) ... noting as before that 1860 lost .04 ...

Years that lost -.02 _____1856,64,67

Years that lost -.01 _____1854,57,61,63,66,74,78,96 _ 1924,53

Years that gained +.01 _1853,55,58,62,70,73,77,87,98_1902,03,05,12,14,46,47,55,60,65

Years that gained +.02 _ (none)

 

 

Edited May 28, 2022 by Roger J Smith

[Roger J Smith]

Making gradual progress on the low minimum file, should be posting that on Monday at the rate things are going (just doing quality control on what's already published in this thread). 

Edited May 29, 2022 by Roger J Smith

[Roger J Smith]

Now able to summarize the record low minima. The top years were 1892 (16.5 records), 1895 (11.5), 1881 and 2010 (tied with 10.0), 1879, 1888 and 1947 (all at 9.0). Looking at the graph, you can see the peak for 1947 after the low-count interval of about 1932 to 1946, then a rather constant moderate stream of records around the 1960s, a marked absence of them 1992 to 2009, then the spike for 2010. 

This graph shows the count in v2.0 year by year 1878 to 2021 (none have occurred yet in 2022) ... will post the differentials next.

[Roger J Smith]

This is the graph of the differentials from legacy CET to v2.0, for the record low minima. Positive values here mean that the v2.0 has more of these records in a given year, and negative values mean that the legacy set had more (and lost however many the negative value shows). 

Clearly there was a draw down on records from the 1901-1930 era, in particular 1915 and 1917 lost several each. Most of these lost records ended up in the already record-rich interval of 1878 to 1890. It looks like the more recent period, while not racking up a lot of records, didn't lose very many and gained a handful of the earlier years' losses. April 2021 had set three records in legacy but lost one of them (to 1911). As with the record highs, a lot of this change involves ties broken or created, rather than outright losses. 

Including all the other analyses, I've come to the conclusion that the period 1931 to 1973 is regarded as well-instrumented and changes are small and infrequent in the data. Larger changes seem to be more frequent on either side of that core interval, although more recently there hasn't been much change either. 

[Earthshine]

9 minutes ago, Roger J Smith said:

Now able to summarize the record low minima. The top years were 1892 (16.5 records), 1895 (11.5), 1881 and 2010 (tied with 10.0), 1879, 1888 and 1947 (all at 9.0). Looking at the graph, you can see the peak for 1947 after the low-count interval of about 1932 to 1946, then a rather constant moderate stream of records around the 1960s, a marked absence of them 1992 to 2009, then the spike for 2010. 

This graph shows the count in v2.0 year by year 1878 to 2021 (none have occurred yet in 2022) ... will post the differentials next.

Thanks for this analysis.  Just checking, is the x-axis years since the start of the min CET series?  Might be a bit clearer to have year markers (e.g. every 20 years or so) but just a suggestion!

[Roger J Smith]

Yes, the x-axis is years after 1877 (1878 = 1 etc, so 1901 would be 24, 1951 would be 74, and 2001 would be 124).

That also applies to the two graphs for the max CET changes. 

[Roger J Smith]

A bit early to define these as "definitive answers" to speculations about new procedures, but, as of afternoon of 1st of June, May 2022 numbers are updated, none of the first half of the month changed from what I had downloaded then, can't say for sure that any numbers ever seen by anybody from then to end of month are same as first published. Also don't know for sure if these are "final final" numbers or like EWP, final numbers that can change slightly up to 3-4 months later before really settling in forever. 

Will just have to keep an eye on what happens going forward. 

[Midlands Ice Age]

Roger thanks for all your work on evaluating the differences.

My own summary is that the minimum temperatures seem to be balanced about the mean (as one should expect if the work is truly genuine), but the figures produced for the maximums seem to not fit the same  agenda, especially as  Minima values do not show any ongoing adjustment bias.

For maximum values the fact that no changes outside +/-0,05c are to be found (with exception of specific known events), until the last 10 years and then the figures just seem to escalate and reach over 0.1C over the last 2 years just doesn't make much sense to me.

0.1C doesn't sound much but actually is a further 20% warming on the last 40 years.

To me it indicates that we trust the figures more from 150 years ago, more so than the ones taken in the last few years. 

This goes against every other scientific reason for changes being made that i have heard. I find that hard to accept.

MIA

Edited June 2, 2022 by Midlands Ice Age

[Roger J Smith]

Well following their logic all changes are due to different blends of station selections than was used for legacy CET, so I don't think it's a case of trusting data in some past years or decades, but selecting data, if that is not what lawyers call a difference without a distinction? 

I would have to dive deep into their document, but not living in Britain I only have a vague idea where some of the locations they mention are located, relative to each other, and no idea of their "aspect" differences. I drew attention to larger changes made to December 1860 than all months around then, similar to the larger changes to December 2000. Not sure what lies behind those two anomalies, possibly one change in location to a place with greater snow cover at those two times, hence considerably lower temperatures. All things being equal, two locations in a similar air mass can have 4-8 C deg temperature differences if one has snow cover and the other does not (the snow radiates heat into space more efficiently than frozen bare ground). 

Anyway, it is what it is, after all, it is their data base, and we can take any approach we want, for example, we could ignore legacy altogether now, or we could stick with legacy for all historical purposes, or we could always quote both where different. I am willing to bet that after a few years, legacy will disappear from the website and then will live on only in some of these hoary threads and the memories of old weather enthusiasts (hands up here if counting). I now have the unenviable task of updating a lot of material I created both here on threads, and in house in my data files, so that they reflect v2.0 and not legacy CET. I have finished the main records in this thread, still have to go over high min and low max records that I compiled (in historical weather section thread). Then I have files of running CET (at least will not have to recalculate 1772 to 1852) and average 1981-2010 and 1991-2020 daily CET. Those are all going to change very marginally within 1 decimal place, some will not change at all. 

Another consideration is that all temperature data bases tend to be quite approximate, as differences from location to location often vary within a range of 1 to 2 C deg, so the problem for climate researchers is to assess whether a data set actually reflect a consistent measurement of the average climate, as site location characteristics change. For example, with both Toronto and NYC (Central Park) data bases, the locations may not have changed much at all, but the surrounding landscape changes, within a changing large city urban heat island. New York City weather enthusiasts claim that Central Park has seen a net cooling effect on warm-season temperatures brought about by an increase in the coverage of the "urban forest canopy" but it is very difficult to assess that sort of thing from data without relating it to nearby locations that have seen no changes in their site aspect. The Toronto location seems to have been in very similar immediate surroundings (on a university campus) since it began operations but its climate reliability has to be filtered through urban heat island assumptions and very few locations exist for comparison back in the earlier part of its record. 

I don't know if anything like this exists, but something like a tree foliation record (for trees in similar aspects over decades or centuries) might be a more robust climate index than temperature series. There's no fooling mother nature about climate change, if the climate warms, trees will foliate earlier in spring and retain leaves longer into autumn. I have a pretty good memory and can recall some variations (back in Ontario where I lived then) that certainly correlate with the temperature record. I noted for example that in warm springs of 1986 to 1988, as well as a very warm April 1976, trees reached full foliation by April 15 to 20, which exposed them to risks of being snow loaded in full leaf as happened in the 1976 case. I also recall that in a very warm autumn of 1971 the fall colours stayed up on the trees in Toronto well into late November. In a colder year, which is also likely to be windier, they get blown down in mid-October and the trees are usually about 50% defoliated even in sheltered downtown Toronto by early November. Another thing that is a robust climate indicator is snow cover. Anomalies of snow cover at either end of winter seasons will correlate very well with temperature anomalies because they are amplifying them. Hubert Lamb had a table in his famous master opus on climate showing dates when ice had formed on a pond in Tokyo, Japan, that went back a long way and provided the only clue we might have to weather variations in that distant past and location. The problem with all such indicators is whether or not they have been maintained continuously and by observers using comparable techniques. 

[Midlands Ice Age]

I accept your comments 100% Roger...

But why the change in the last 5 years to max temps?

I am not aware of any changes of locations or stations during that period.

Surely with the automatic weather stations you would have thought they would be more accurate, and so  with lesser reasons for adjustments.

 I am puzzled why they need to change the latest data?

MIA

[Roger J Smith]

You're right, according to their document the three stations used in the legacy CET were maintained in v2.0 after 2006 (according to their document). So to change that data must involve something like recalibration, perhaps they tested the three sites and found one with a wonky thermometer, so they applied a general 0.1 increase which seems to be quite frequently the change seen since 2006. However it isn't applied every day, there could be such factors as a realization that climatological day reporting practices were inconsistent (a check of hourly obs might clear up a few errors thus introduced). Let's say for the sake of argument, one of the three observers was on occasion found to have allowed an overnight temperature rise to be reported in "yesterday's" high, either through lack of training or sloppy habits. So perhaps that tendency was noted and eliminated from the data sets. (not trying to make any insinuations here, just trying to figure out a reason for any changes even being possible if the stations used are not changed). Or perhaps they applied a new protocol about urban heat island changes, an overall increase 2006-22 would imply that a less strenuous reduction for urban heat island had been decided upon (it might have been differentially applied, with cloudy wet days being treated differently from clear days). 

They probably figure that the changes are generally so slight that nobody would be interested in the details, but of course we are. Somebody should see if Tim Legg or any other appropriate person could maybe answer some of our questions either in a series of e-mailed questions or by appearing on a certain agreed day on line here. Then maybe some of these puzzles could be solved. 

[BLAST FROM THE PAST]

On 02/06/2022 at 23:47, Midlands Ice Age said:

I accept your comments 100% Roger...

But why the change in the last 5 years to max temps?

I am not aware of any changes of locations or stations during that period.

Surely with the automatic weather stations you would have thought they would be more accurate, and so  with lesser reasons for adjustments.

 I am puzzled why they need to change the latest data?

MIA

I’m not puzzled in the slightest.   
There was absolutely no legitimate reason to scrap the Redhill Aerodrome ‘weather station’ except it showed temperatures lower than they liked.

Anyway, nowt we can do about it

 

 BFTP

Edited June 3, 2022 by BLAST FROM THE PAST

[Greyhound81]

On 03/06/2022 at 22:49, BLAST FROM THE PAST said:

I’m not puzzled in the slightest.   
There was absolutely no legitimate reason to scrap the Redhill Aerodrome ‘weather station’ except it showed temperatures lower than they liked.

Anyway, nowt we can do about it

 

 BFTP

Typical denier BS

[Midlands Ice Age]

3 hours ago, Greyhound81 said:

Typical denier BS

Typical activist BS...

Please give a reason why you think it is bullshit..

MIA

[Greyhound81]

2 hours ago, Midlands Ice Age said:

Typical activist BS...

Please give a reason why you think it is bullshit..

MIA

I am not an activist, I have never been to a protest in my life.  Anyone is entitled to thier views on AGW, but what is bullshit is the idea that the only reason the CET data has been changed is that it is part of a big conspiracy, which is what BFTP was suggesting.  It is an insult to the hard working people at the Met Office.