Part 1, Central England Temperature timeseries and Manley papers
This large article was composed some time ago, last edit Oct 2013 it seems but left unpublished, one of dozens.
There is a disturbing story behind the current CET dataset as will be revealed in Part 2. The above plot sets the scene, a straight illustration of the whole dataset with timeline.
This blog article is intended as a backgrounder covering a variety of information including links to official copies of the two historic Manley papers which are the basis of the CET dataset. In my opinion CET is misrepresented as more instrument based than it is in reality. More reasonably it is an expert opinion on weather.
The Met Office CET web page mentions the whole data then plays pea and thimble silently showing a plot of the later subset. Reason to be revealed in part 2.
Phillip Eden writing in the Daily Telegraph explained CET is an expert compilation of information snippets
The original work, stretching from 1659 to 1973, was compiled by the late Prof Gordon Manley in a work of immense scholarship which took over 30 years to complete. He painstakingly examined all available historical weather diaries and journals, as well as early thermometer readings which were taken variously in gardens, unheated north-facing rooms, and on shaded walls, and he integrated all these into a homogeneous record.
Manley was an academic geographer who had unrivalled knowledge of the British climate. …
– Phillip Eden, The Daily Telegraph, Saturday November 22nd 2008, p 40, h/t Shirley Wittering in the Thriplow Society journal 2008.
(I seem to have lost the direct link, apologies if the above is not fair usage, will be corrected immediately if necessary)
The Wiley / Royal Meteorological Society paywall the Manley 1953 work, but then make it freely available provided copyright is honoured.
The abstracts of the 1953 and 1974 Manley works follow below and then links to the RMETS archive.
Some of the following is copy typed, errors are mine.
The mean temperature of central England, 1698-1952
by GORDON MANLEY
Bedford College, University of London
(manuscript received 24 November 1952, in revised form 23 February 1953)
A table of monthly mean temperatures representative of the English Midlands has been constructed for the period 1698–1952. From 1815 it is derived from the average of the ‘Radcliffe’ (Oxford) and ‘Lancashire’ monthly means. From 1771–1815 it has been built up by averaging the departures, or anomalies, for each month at a number of inland stations whose records are long enough to be bridged into the years 1815–1840; from 1771–1798 the values thus obtained are reinforced by direct values based on the long Midland record at Lyndon. Before 1771 the existing Edinburgh, Greenwich and Lancashire means are supplemented by further direct values based on Lyndon and Exeter. For all years previous to 1752 the Old Style calendar months have been rectified to the New Style, and the direct values derived from early MS. journals in London, Plymouth, and the Midlands have been carefully collated back to 1728, and extended by means of other short records from Halifax, London and elsewhere back to November 1722. Finally, a direct reduction of Derham’s Upminster record, 1699–1706, has been used as a basis for the earliest years. For the intervening years estimates are provided from a consideration of the Utrecht reductions, supplemented for 1713–1722 by non-instrumental observations of wind and weather in England. Some corroboration of the values so obtained during the eighteenth century is provided; it is hoped that the table can serve as a general standard for studies of English temperature variations until such time as something better can be provided.
Link to free access PDF of the 1953 paper follows the 1974 abstract.
Central England temperatures: monthly means 1659 to 1973
By GORDON MANLEY
Emeritus Professsor of Environmental Sciences, University of Lancaster
(Manuscript received 17 July 1973; in revised form 10 January 1974)
An up-to-date Table is provided of the monthly and seasonal means representative of the air over central England from 1659 onward, incorporating some minor revisions and extensions of the earlier Table (Manley 1953 and 1959). Comments on its construction, and on the prospect of further extension backwards in time, are added.
I have no information on a 1959 work.
Here is the general URI for RMets Classic Papers made free access, select M for Manley.
If you look at the detail in figure 1 you will notice the stripes at whole degrees for the early data, plotted to make that clear.
An email about CET.
Long ago I archived a copy of a 1995 email quoting an email from a senior Met Office person “email@example.com” the old Internet domain they used, gov.uk did not exist back then.
Before anyone assumes, this has nothing whatsoever to do with the so called Climategate emails, was freely provided by an American organisation, attached with a publication of CET data. Came from the archive here.
For context reasons the email is selectively quoted omitting a sentence to be revealed in Part 2. An American scientist wrote
We should make the appropriate corrections
as indicated below. We sould also
put the mail msg in the “Readme” file associated
with this dataset.
Here is part of the 1995 email giving data corrections but also revealing others failed in trying to update Manley’s work.
>Phil Jones has been in the USA for the last week so I haven’t been
>able to find out where their version of the series originates from. I
>will email him before I go home this evening about it.
>1740 2 should be -1.6, not 1.6
>1898 2 should be 4.8, not 5.8 (this error is also in Manleys 1974 paper,
> but not in his original 1953 paper)
>Manleys original monthly series was published in 1953. I guess that
>the UEA series comes from a different source from 1953 and 1974.
>Apparently there were several attempts by others to update the series
>before Manley did so in 1974. Our monthly series is due to Manley
>up to 1974, and since then has been calculated using the same set of
I disagree with the final sentence which should be taken as casual writing and a context is the year 1995. See the UHI and other matters to do with CET for more on this.
The Wayback machine revealed this which is well known.
Central England Temperature (CET) is representative of a roughly triangular area of the United Kingdom enclosed by Bristol, Manchester and London. The series began in 1659, and to date is the longest available instrumental record of temperature in the world. Since 1974 the data have been adjusted by 1-2 tenths °C to allow for urban warming.
The archive also reveals a glimpse of AGW starting to rear it’s head.
A 1999 paper on CET I found in the archive here turns out to be open access today
Central England temperatures: long-term variability and teleconnections
Timothy C. Benner
Past long-term climate variability is important for the prediction of both future climate and human impacts on climate. Teleconnections help to reveal the interactions between the components of the climate system. This research examines the central England temperature record from 1659 to 1997, both for its own variability and for its relationship to other climatic records. Results support a possible warming trend, especially in recent years. Four independent spectral analyses show several common, prominent periods of oscillation in the record, from a few years to nearly two centuries. Wavelet analysis emphasizes the non-stationary nature of this variability. Temperatures may be related to solar irradiance and sunspot numbers over long periods. They show a connection to the North Atlantic Oscillation, especially over periods of 7–8 years. However, they show no apparent relationship to the El Niño–Southern Oscillation. Copyright © 1999 Royal Meteorological Society
Link to open access RMetS paper via Wiley online
(I hope this works, lots of grief here trying to put a working link in an article, strange characters in uri)
Modelling Trends in Central England
DAVID I. HARVEY AND TERENCE C. MILLS*
Loughborough University, UK
Trends are extracted from the central England temperature (CET) data available from 1723, using both annual and seasonal averages. Attention is focused on fitting non-parametric trends and it is found that, while there is no compelling evidence of a trend increase in the CET, there have been three periods of cooling, stability, and warming, roughly associated with the beginning and the end of the Industrial Revolution. There does appear to have been an upward shift in trend spring temperatures, but forecasting of current trends is hazardous because of the statistical uncertainty surrounding them.
To their credit a paper saying ‘not a lot going on here.’
Many more papers exist. A critical one will be linked and quoted in Part 2.
Figure 1 basis Monthly_HadCET_mean.txt, 1659 to date (32 Kb), to Sept 2013
2. The Wayback Machine carries a little content from 1996 through 2003 when the Met Office switched to a new URI. This gives a glimpse of a different era. http://web.archive.org/web/*/meto.govt.uk