Guest essay by Neil Catto
The CET record started in 1659 close to the minimum of the little ice age. As such, it is with no surprise that last year (2014) was the warmest on record. It would appear to be a natural recovery. The monthly mean temperature of 8.87 Deg C in 1659 has increased to 10.95 Deg C in 2014; which equates to 0.06 Deg C/decade.
I used the CET mean monthly data 1659-2014: Downloaded 6th Jan 2015 http://www.metoffice.gov.uk/hadobs/hadcet/data/download.html
My main interest in this data set is to gain better understanding between natural variation and AGW. I consider the CET as a reasonable representation of Northern Hemisphere trends. In 1739 Mount Tarumae in Japan erupted with a VEI force 5. The mean monthly CET temperature in 1739 was 9.21 Deg C, in 1740 there was a significant drop to 6.84 Deg C à and in 1741 a recovery to 9.32 Deg C. This natural occurrence had the equivalent drop in temperature of -23.5 Deg C/decade and recovery of 24.6 Deg C/decade. With a natural variation of this magnitude I never understand the alarm about 2.0 deg C/decade, human life survived and exponentially grew in numbers.
The last time I downloaded CET data was 22nd May 2013. Out of interest I thought I would compare the two data sets. The results were interesting to say the least.
Fig 1 anomalies between CET downloaded in May 2013 with CET downloaded in Jan 2015 (data to Dec 2014)
It is noticeable that nearly every adjustment is positive, with no negative changes. The whole data set shows an average increase of 0.03 Deg C in 20 months or equivalent to 0.18 Deg C/decade.
Discussion:
What is the reason for these data adjustments?
How often and by how much are these data adjusted?
Is this anthropogenic warming caused by man-made adjustments?
It seems the entire record went up, from far to near past. How could this affect trends?
Neil Catto ==> Had you seen this record from Greenwich?
http://wattsupwiththat.com/2014/02/23/historical-note-greenwich-england-mean-temperature-35-yr-daily-averages-1815-1849/
Don’t know how they relate to the CET — but they are “original” and published by the man you kept the record during that period.
It might be interesting to figure the adjustments made during these 34 years alone — as we have a guaranteed set of actual recorded temperatures that have never been touched.
– kh
I don’t know how often the following needs to be said before crimatologists take note but here goes :
In no other science is it acceptable to alter, modify, adjust or change in any way shape or form, past data. It is what it is and must be left untouched. You may choose to critique the data but you may not change it.
Only in climate novels is it seen as normal practice to alter both past and present data and, of course, future data.
E.g.: the air pressure reading at a given location is 1290hPa. Surrounding stations show an average of 1025hPa. In this case, it seems to me that action is required. There has never been an air pressure of 1290hPa, but averaging that out by mixing this value with other stations values would be a gross distortion. So the only way not to distort the results is -to me- to omit that 1290hPa completely with a remark “data not plausible – typo”.
Would you agree that there are situations where it becomes necessary to “manipulate” data?
Imagine there’s no thermometers.
It’s easy if you try
No adjustments possible
Above us only sky
Imagine all the people
Living as they do
Some years they skate the Thames
It isn’t hard to do
While once vinters worked
and we had wine not brew.
Imagine all the people
Living life in peace
You may say I’m a denier
But I’m not the only one
I hope someday for science
And the world will be as one
Imagine no adjustments
I wonder if you can
No hockey sticks or models
where true science isn’t banned
Imagine all the sunshine
and all the climate too
You, you may say I’m a dreamer
But I’m not the only one
I hope someday there will be science
And the models are all gone.
……
And the world can continue do do what it damn well pleases!
Good one! Lol.
The fastest warming over a 50-year period was from Oct. 1688 to Sept. 1738 with a warming of 1.90059 degrees. The warming of the second half of the twentieth century was 0.82415 degrees, less than half as fast.
I live in central England, if this is the warmest year ever, I must have totally imagined 1976 (massive enduring heatwave). I think I may need some revisionist history training, I believe there’s a course available at the UEA?
Your confusing a hot summer with a hot year.
2014 hasn’t been particularly hot, just not particularly cold.
Only 1 month was below normal, whereas in 1976 4 months were, in spring and autumn, so you wouldn’t really notice. OTOH the summer was an average of 2.5 degrees above normal
Massive heatwave indeed – but for a few months. What about the rest of 1976 ?
The lack of cold weather in the early months has heavily skewed the averages for CET.
That lack of cold weather last winter was due to a stronger jet strean across western europe which in turn was caused by very cold plunges across north america.
Perversely,our warm winter last year was actually a sign of a cooling globe with more meridional jets and greater global cloudiness.
Ah that explains it, Steve. The “lack of cold weather” is the reason it was warmer.
Cheers, mate.
I think the issue is that “warmest” seems to be that the average temperature is higher, which does not necessarily mean higher summer temperatures. For example 2012 was the “warmest” year in the US, but actually due to a mild winter/spring, see https://climatebeagle.wordpress.com/2014/12/31/digging-into-the-definition-of-warmer/
I wonder if there are better definitions for comparing temperature across two time periods, reducing the entire range of a year to a single value loses a lot of information.
Looks more like a bulk shift to me. You need to run a line-producing algorithm.
Although CET mean was a record, funnily enough neither CET max nor CET min was a record. The former, which I was watching closely for the whole of December (using a good proxy to CET) came second to 2003, and the latter came second to 2006.
Re Sophie and 1976, none of 2014’s seasons was a record on CET max. Winter (counting December 2013 rather than 2014) came 6th, spring came 8th, summer came 28th, autumn came 4th. See http://hadobs.metoffice.com/hadcet/ssn_HadCET_max_sort.txt for these.
Rich.
I hate abréviations and soon will stop reading your blog
Do you think that people knows what CET means ? You should at least explain when you use it the first time in a post
Fritz,
There is a glossary in the mast head.
CET = Central England Temp.
“The CET record started in 1659 close to the minimum of the little ice age. As such, it is with no surprise that last year (2014) was the warmest on record”
And the link between those two events is ?
If we get a record warm year in 2020, or any other year, would the author say the same ? Presumably he would as he just thinks that current warming is part of an endless rebound from the LIA.
Its another version of string theory – how long is a piece of it.
James Abbott says:
And the link between those two events is ?
I can hardly believe that is a serious question. The LIA was one of the coldest episodes of the entire Holocene. What does Abbott expect? That when temperatures go down, that they stay down forever?
dbstealey
The fact that you find my question so unbelievable says a lot. Yes the LIA was a cold episode, well documented. But the article does not explain, and nor do any of those who rely on the LIA rebound as a cause of current warming, as to WHEN the LIA rebound ends. So my point is, we are now over 300 years since the lowest point of the LIA – how long is this piece of string ?
The CET shows a notable warming from about 1900 but which then flattens out until about 1980 – then it clearly jumps upwards. The ten year running mean is clearly at a higher level post the 1980s than the C20th recovery period or the preceding LIA period.
How can that upwards jump from 1980 onwards still be pinned to the LIA recovery ? How do you define the end of the LIA ?
http://www.metoffice.gov.uk/hadobs/hadcet/
The peak to trough periodicity is historically about 500 years.
The lowest level of solar activity was about 350 years ago.
There could be another solar induced warmer period before we reach the warmth of the Mediaeval Warm Period again and then start a decline towards the next colder spell.
It is possible that the latest solar quietness is the beginning of an earlier decline but not certain..
Regarding Central England and London to be just one large Urban Heat Island makes a lot of sense. There is precious little open space left in Central England and population has increased from to 8,900,000 in 1800 to 53,000,000 now.
My opinion is that the CET is an adjusted data set that shows very little upward trend till the 1980’s.
In the 1980’s it became the established scientific consensus that increasing CO2 and other man made gas emissions would cause the surface temperature to warm. I am not surprised that the temperature data measurements match the consensus theory, any other result would not be allowed.
However I would have expected that the effects of man made emissions would also be noticeable to a similar extent in the precipitation data record but my understanding is that so far the record shows no trend.
Adjustments after 1975 includes reducing temperatures after 1975. Adjustments for UHI. Raw unadjusted temperatures are higher.
Must be wrong. We want raw unadjusted data. They show more temperature increase. That does not match the concensus theory.
Very sad.
Hitherto I had more confidence in CET than other data sets.
All gone.
You loose confidence in CET because there is an difference between the yearly values and the yearly mean of the monthly values in the latest hadcet file?
A difference that does not make any difference for the temperature increase.
Easily unconvinced. Unconvincingly so.
vukcevic said:
January 9, 2015 at 1:52 am
“There is nothing suspicious about this:”
Vuk,
Does the feature you spotted account for ALL the anomalies in the head post?
Hi Mr Wilde
As far as I can see graph I got is identical with corresponding section of what is shown by Mr.Catto. In my view it doesn’t change anything significant. I did think if the annual CET contains two decimal points (which may or may not be justifiable) then Met Office should recalculate data based on a year of 365/6 days and not on 12 equal months.
In the email to the Met Office I said:
Since monthly data is made of the daily numbers and months are of different length, as a test (1900-2013 data, see graph attached)), I recalculated the annual numbers, using weighting for each month’s data, within the annual composite, according to number of days in the month concerned.
This method gives annual data which is fractionally higher, mainly due to short February (28 or 29 days), see the attachment.
Differences are minor, but still important, maximum difference is ~ 0.07 and minimum 0.01 degrees C.
I am of the view that the month-weighted data calculation is the correct method.
This is copy of the actual graph I attached to the email:
http://www.vukcevic.talktalk.net/CET-MO.gif
Thanks, Vuk, very helpful.
Re the “rebounding theory”:
Why was there no rebounding 1820 – 1920?
And no rebounding 1940 – 1980 (actually the opposite of a rebouding)?
Does the rebounding theory imply that the LIA ended in 1980?
Data for models keeps getting revised after the fact
Factor key to the models is admittedly unknown (how Clouds work)
All Models failed… after 18 years….
DNA evidence would never have been allowed in at trial with the type of ‘scientific’ foundation seen in the AGW area…….any clear thinking juror or judge would be skeptical. After that miserable testimony on the everchanging historical data and the failed models, jurors would be particularly skeptical when the scientist on the stand was unable to even understand why anyone else would be skeptical… screaming like an old time preacher, calling everyone in the room ‘deniers’ of the ‘obvious’ truth to his ‘science’
Conclusions based on Astrology would have a better chance at trial based on that type of proof.
B
Right and ….
No, Neil, it’s not because the ‘adjustments’ haven’t produced any warming.
The question that needs an answer here is how can the yearly average be so high while the max.temps and min temps are not exceeded,why does the atmosphere not accumulate heat(the green house effect).My answer is that the clouds(storminess) does not increase the greenhouse effect ,yes the surface temperature is a little higher but clouds reflect more sunlight back into space and lose heat.It is false logic to believe that if the surface temperature anomaly rises then it must have been caused by the green house effect.
The head posting arose because 2014 had the highest mean CET on record (though only 2nd for max and for min).
Today, 10 days after 2014 was ushered out, I believe I have just witnessed the warmest English January day evah! That is, my estimate of CET max for January 9th is 14.8degC, smashing 1930’s record of 13.6degC. (There are 17 previous years breaking the 13 degree mark.)
It’s been a bit freakish. Oddly enough, my thermometer registered only 14.6, but Pershore’s 15 miles away apparently made 16.0. Last night I drove north to Worcester (beyond Pershore) and noticed it getting warmer as I drove north.
One final point: there was statistically significant cooling in CET max from 2002-2013 inclusive. That cooling now seems to be over, unless 2014 and January 2015 turn out to be a mere blip.
Rich.
Amazing science! We have temp data, precise to the hundredth of a degree, from all over the Earth, recorded in the thirteenth century. This information just blows my mind as I thought that temperature data of that quality has only been available in the latter part of the satellite era. Silly me.
No you aren’t silly (you were being ironic I know).
There was nothing to see all along.
The data remains the data and is unaltered.
And the respected posters on here got righteously indignant.
Which was of course all that was wanted.
You are right, completely unaltered. Just like the axe that has been in my family for 350 years. Over time we had to replace the head three times and the handle five times. Same axe.
The Met Office made an elementary error in averaging months of different lengths to produce the annual temperature. They have corrected this by averaging the 365 (or 366) daily temperatures, which leaves all the daily readings unchanged and increases the annual temperature very slightly because the shortest month is one of the coldest of the year.
To suggest that there is some subterfuge involved would be a pretty dumb isolated comment. How does it acquire the status of “Guest Essay”?
Well done Vuk for getting them to correct it. I’d noticed a slight discrepancy in my own spreadsheets which also use the appropriate number of days for annual, seasonal and monthlyaverages, but had just assumed I must be missing the odd significant digit in the monthly figures here and there.
Sloppy work is not acceptable. It leads to thoughts of incompetence. How many years have those ‘learned’ folk been operating? They had to have an outsider point out something fundamentally wrong with their ability to average something correctly.
A bit harsh vs Catto there Alex.
rooter
I was referring to the MET office. I have no problem with Catto
Perhaps fiddling is warranted then Alex. The previous yearly averaging was “sloppy work” according to you Alex. The new method is more correct.
I take it that you don’t prefer the sloppy work.
I also did a recalculation of yearly means adjustet for monthly length. And got results very close to the newest hadcet for yearly values. One reason for the minimal discrepancy is of course that I did not bother to account for leap years. Perhaps they use daily recordings.
http://i.imgur.com/L3kZOAo.png
The reason for the series being shifted up is of course that we get a decrease in winter recordings vs summer recordings. And as most of us know: summer is warmer than winter. This series have absolute temperatures and when you get more summer recordings vs winter recordings the series will give higher temperatures on average.
The difference is small, but the new calculation for yearly averages is obviously more correct. Some will of course maintain that it must be wrong because it is different from the previous calculation.
So what happens? Some loose their “confidence” in the series. Some call it rape. Etc. The usual stuff.
When we know the reason for the adjustment for yearly averages: Why not update the post? Would Catto answer his own questions? And perhaps he could correct this:
“The whole data set shows an average increase of 0.03 Deg C in 20 months or equivalent to 0.18 Deg C/decade.”
Which is just s….. wrong.
My comments are about my dislike of data being fiddled with. I have absolutely no idea if Catto has got it wrong with his calculations. Around the world you have stations moved and data adjusted to reflect what would have been if the station had been at a particular place for 200 years. So you have this ‘floating average’ that is in fact not an average of anything, its just -ish. You can’t have a floating ‘earth’ in electrical circuits because you get screwy results. The whole climate thing is so screwy now that I have no idea on whether the climate is getting warmer or cooler. I’m better off licking my finger and testing the wind or asking grandma if her bones are aching.
It’s ‘climate something’ not ‘climate science’
It does seem strange that the producers of this record would consider the creation of monthly averages, (fine) then using this intermediate average, with its rounding, to be averaged again to produce the annual result.
It seems obvious that you should use 365 days worth of readings to create the annual figure, which will minimise any unwanted mathematical artefacts being introduced.
I wonder if this process has common usage in the climate industry?
Using absolute temperature will give this effect, small as it is. Using anomalies will make this effect even more neglible.
The accuracy of early thermometers has been questioned by some, but others in this thread suggest that because the early experimenters were ‘gentlemen scientists’ they would take great care at using well made thermometers.
CET records temperature from 1659.
The Royal Society set up a committee in 1777 to find how to reliably fix, and find out, what the temperature of boiling water was.
http://www.hps.cam.ac.uk/people/chang/boiling/
Modern experiments show that at sea level, boiling point of water varies from 100 C to 103 C depending upon how you go about the boiling process, as documented in the above link.
How can we rely upon pre 1777 temperatures when we could not ‘fix’ the boiling point of water to enable the calibration of thermometers?
Also, when did it become common knowledge and practice to adjust your boiling point calibration setting to account for your altitude?
Were many thermometers of the day made in the Midlands (Birmingham, UK) is 99m above sea level, changing the boiling point of water by nearly 0.4 C or were they manufactured on the banks of the river Thames at sea level?
With so many factors creating genuine unknowns in early temperature metrology, perhaps we should stop trying to reduce the size or error bars and increase them instead.
Or you could of course not use the recordings before ca 1750. Many issues. Including measurements taken indoor, not actual temperature readings and not from England. Etc
http://en.wikipedia.org/wiki/Central_England_temperature
So at 10.3 C it is actually 10 C. How accurately did they read ‘better quality’ thermometers up to 100 years ago. Have you heard about parallax error? Rounding up and rounding down?