[Bored chemist]

bc says
Well, there is more CO2 and there is warming but there is no cooling

henry says
who says it is warming? you have simply not been paying attention to me in the thread of this post?
earlier on this thread we noted that it has been cooling for at least one whole solar cycle
[link snipped because it messes up the formatting of the page]
Note that this result from various data sets is confirmed by my own results which JP says I may not quote...here? JP and your  team has some peculiar rules when it comes to global warming....

If you want to go nit picking you could also ask: how long has is not been warming? That leaves those poor souls whose miserable lives depend on this global warming scam some stay of execution.
e.g.:  this post contains graphs of running trends in global surface temperature anomalies for periods of 12+
 and 16 years. They indicate that we have not seen a warming hiatus this long since the 1970s.
http://wattsupwiththat.com/2013/06/21/may-2013-global-surface-landocean-temperature-anomaly-update/

However, my own results for the drop in global maxima will show you that the current cooling trend will continue, until at least 2040....or there about.

So, there is no warming trend, and there has not been any, for at least 16 years. You, get on with that. I ask you: why do you keep referring  back to it as if it (i.e. the "global" warming) were truly still happening?

It seems that you were too busy asking if I was paying attention to pay attention to what I already said.
"But nobody takes such a short snapshot seriously so your conclusions are not valid. (and the data might be questionable too, but that's not the real issue)"

So the reply to "So, there is no warming trend, and there has not been any, for at least 16 years. You, get on with that. I ask you: why do you keep referring  back to it as if it (i.e. the "global" warming) were truly still happening?
" is, that it is warming.
Your assertion is like saying it's cooler this evening than it was at lunchtime, so the world is cooling.


Here's the real data
http://www.skepticalscience.com/pics/RFC12_Fig1.jpg

Why are you pretending that the right hand half of that graph is lower than the left hand half?

[MoreCarbonOK]

bc says
Here's the real data
henry says
this is the problem.
people like Cook who have an agenda to "save" their jobs

in this respect I can only trust my own data
(I know that I have no ulterior motive but finding the truth)
which I took from www.tutiempo.net
particularly the drop in maximum temps. (that nobody is looking at)
see here
http://www.thenakedscientists.com/forum/index.php?topic=47181.0


Don't forget: automatic recording with thermo couples only began since the seventies
so how are you going to compare older data with automatically recorded data?

[Bored chemist]

bc says
Here's the real data
henry says
this is the problem.
people like Cook who have an agenda to "save" their jobs

in this respect I can only trust my own data
(I know that I have no ulterior motive but finding the truth)
which I took from www.tutiempo.net
particularly the drop in maximum temps. (that nobody is looking at)
see here
http://www.thenakedscientists.com/forum/index.php?topic=47181.0


Don't forget: automatic recording with thermo couples only began since the seventies
so how are you going to compare older data with automatically recorded data?

OK, so, in order to save his job, Cook tells the truth. There's nothing gets a scientist sacked faster than getting caught faking data.
Glad we have got that sorted otherwise it might look like you were making a libellous statement about someone and that would be a breach of the rules and would get you kicked off the site (in very much the same way that lying would get a scientist sacked).


"Don't forget: automatic recording with thermo couples only began since the seventies
so how are you going to compare older data with automatically recorded data?"
WTF?
Do you think thermocouples are magic?
It's like asking how you can compare peoples heights when they used to be measured in inches, but now they use centimetres.

There's nothing special about a thermocouple, or  the length of a mercury column. Just for the record, neither is the "official" means for establishing temperature.
But you can't use a constant volume gas thermometer (extrapolated to zero pressure) to measure the weather because it takes too long.
So you use some sort of surrogate. As long as you can calibrate the surrogate, it doesn't matter what you choose to use.
A thermocouple is handy if you want to log data automatically, but a mercury in glass thermometer is just fine for most places (you would need an alcohol thermometer for some very cold places).
Or, if you wanted, you could use the 16O: 18O isotope ratio in calcium carbonate.
That last one would be a bit weird- but it has the advantage of recording the data for you in seashells.

So, the biggest data sets all agree that the world is warming. The sea levels and shrinking ice caps support that.
Why should I listen to you?

[MoreCarbonOK]

BC says
So, the biggest data sets all agree that the world is warming.

Henry says
no, the biggest data sets show it has been cooling for the past 11 years (which is the equivalent of 1 solar cycle)
which you (&cook&co) say is not significant
They also show that it has not been warming for about 15 or almost 16 years now,
which you (&cook&co) say is also not significant
So, the biggest data sets all agree that the world is not warming.


Now I (we)  know from my own data set that it has been warming (naturally) since around 1951,
until 1998, when we changed the signal
If you want to claim that it has also been warming before that time, i.e. before 1951
you must show me a calibration certificate of a thermometer from, say 1925?


My results show that we started cooling down from the new millennium and we will continue to cool,
until 2039 or 2040\,despite of what Cook and co or anybody says.
But I am happy if you could come with another interpretation (fit) of my results as reported in my tables?

[Bored chemist]

OK, let's look at your "model": it's a sine wave.
http://blogs.24.com/henryp/2012/10/02/best-sine-wave-fit-for-the-drop-in-global-maximum-temperatures/
It's based on a dozen years or so of data.
There are two ways to test it.
We can look at the predictions and wait to see if they are true or, we can look further back at the real historical data and see if it works.

Guess what! it fails miserably.

Now, your next question was
"If you want to claim that it has also been warming before that time, i.e. before 1951
you must show me a calibration certificate of a thermometer from, say 1925?"
Why?
This guy is on record as calibrating thermometers about two hundred years earlier.
http://en.wikipedia.org/wiki/Anders_Celsius
He's kind of famous for it.
Anyway, here's one from before 1951 (what had 1925 got to do with anything?
http://www.ebay.co.uk/itm/Vintage-EX-MOD-Thermometer-Cert-Signed-C-G-Darwin-Whirling-Meteor-/321024094115

And then you go back to an invalid model mapping the temperature to a sine wave.

"But I am happy if you could come with another interpretation (fit) of my results as reported in my tables?"
If I get bored, I might.

[MoreCarbonOK]

BC says
we can look further back at the real historical data and see if it works.

Guess what! it fails miserably.

henry says

you have global data on maxima?

my data for 47 weather stations is complete from 1974?

there is no other fit than the sine wave for those data on maxima,
unless you want to freeze us to death in the next few decades...?

note that below that global graph there is one from a weather station in Alaska,
which I found had good data on maxima from 1942, which confirmed the sine wave.

In terms of comparison, I am saying that you cannot possibly compare all
results from 1900-1935 and bring this into the equation,
which is why I asked you to provide me with a calibration certificate from 1925.
(they never started calibrating thermometers before the time of the certificate that you quoted. Once manufactured, they were considered "good")
In addition, you must consider the way of recording i.e automatic versus human, at intervals (especially if you want to use means instead of maxima)

[Bored chemist]

I wonder if you can prove this slur on the competency of the scientists of the day
"(they never started calibrating thermometers before the time of the certificate that you quoted. Once manufactured, they were considered "good")"
If not, perhaps it would be polite of you to apologise and withdraw it.
The reason that the certificate I showed has been kept is not that calibrations were unusual at that time, but because of the signatory. Calibrations were, of course, commonplace.
Of course, in reality there were, at and before that time, systems in place to ensure the correct calibration of thermometers.
http://www.bipm.org/en/si/history-si/temp_scales/its-27.html

This "there is no other fit than the sine wave for those data on maxima, unless you want to freeze us to death in the next few decades...?" is just silly.
What I want  doesn't affect the reality that you can plot a quadratic or other function through that data. Given how little data you actually have, it would be difficult to say that the change wasn't linear.

"you have global data on maxima?"
Nope, and I don't need it.
I have got data for averages (and I already cited it).
If the average is slowly rising, but the maxima are following a sine wave, it would follow that the deviations from the mean were, in the past, much much bigger in order to keep the average on track.
But such data would have been noted and commented on In particular, it would mean the weather in the past was more variable than it is now- but that's the opposite of what the historical record tells us.

[evan_au]

The only graph here showing a decline in maximum temperatures seems to come from just 1 site, not 47 sites. The measurements from 47 sites show no such downturn.

Naturally, it is best to have regularly calibrated thermometers. Knowing how vital these temperature measurements become for predicting weather, I am sure that all Navy vessels and Post Office weather stations would have a standard procedure for regularly calibrating all their equipment (including thermometers). Commercial vessels may not be so rigidly disciplined - but there you have the benefit that some devices that might read a bit high will be balanced by those that read a bit low - and they will both be updated by new ones when they drop or are otherwise broken.

There are certainly some blips in the ocean temperature records - a significant one is during the 2nd World War. But after the war, they returned to previous methods & locations for measuring ocean temperatures, and the pre-war upward trend continues.

[MoreCarbonOK]

The only graph here showing a decline in maximum temperatures seems to come from just 1 site, not 47 sites. The measurements from 47 sites show no such downturn.

henry says
the blue in the top graph there is the result (average) of the drop in maximum temperatures coming from 47 weather stations.
(you have to go to my tables to see the individual results of each of those  47 weather stations)

the blue in the bottom graph is the result of one station that had good data on maxima going back to 1942
 
the red in both graphs is my best fit for the available data.
I am open for other proposals from you if you think we can make another best fit?

[MoreCarbonOK]

BC says
The reason that the certificate I showed has been kept

henry says
but that one is from 1948?

I asked you to prove to me that thermometers in weather stations were calibrated regularly before 1930?

[Bored chemist]

Do you think it's the glass tube that changes, or the mercury in it?
Surely you realise that saying "all the thermometers were wrong" is an absurd bit of clutching at straws.
Some might have read high, some low. But what we are looking at is a change so calibration errors are pretty much self cancelling anyway.
However, I'd still like to see the evidence on which you base your claim that all the people measuring temperatures were incompetent.

[MoreCarbonOK]

BC says
Some might have read high, some low. But what we are looking at is a change so calibration errors are pretty much self cancelling anyway.
However, I'd still like to see the evidence on which you base your claim that all the people measuring temperatures were incompetent.
Henry says
well...  we are only looking at a few tenths of a degree C (or K)  of warming from 1900-1950
which might be completely wrong if the thermometers were wrong?
You keep returning the question to me, but I asked you first:
from what time onward in history did we start calibrating thermometers at regular intervals?
I think it was only after the war. ./.as proven by the certificate you provided....

[Bored chemist]

Well, I guess I can't help you. I pointed out that they have been calibrating them since 1889, yet you say they weren't  doing it until WWII.
 And, I am returning the question to you for a reason.
You made an assertion that a group of people who are no longer in a position to defend themselves were not competent.
You should be able to back that up.
So, did you just make it up?

Also, how plausible is it that all of the thermometers are exactly as out of true as eachother. And they are all drifting in the same way?
Because that's what it would take to explain away the change in temperatures.
Can you think of a plausible explanation?

[MoreCarbonOK]

BC says
 I pointed out that they have been calibrating them since 1889, yet you say they weren't  doing it until WWII.
Henry says
I asked: from what time onward in history did we start re-calibrating thermometers at regular intervals?
I think it was only after the war. ./.as proven by the certificate you provided....
But I asked you to produce a certificate, any certificate, from before 1930.
I am sure they must have been calibrated, once perhaps, when they left the factory,
(if they were made in factories)

However, not only am I interested to see at what intervals the thermometers were re-calibrated before the war,
(which indeed was probably not done regularly at all until after 1945, going by the certificate you produced,
unless you prove me otherwise),
I am also interested to know from those certificates what the accuracy was.
I remember that even in my time, a mercury thermometer with one degree C divided in ten portions,
with an accuracy of 0.05 was quite rare and expensive.
So I am asking you again to provide me the certificates, any certificate,  to show me what the accuracy was before 1930.
If the accuracy was in fact only about 0.5, as I suspect you will find, then the the current warming rate of 0,6 or 0.7K per century,
can be considered questionable....and seriously challenged.
Curious, that you refuse to consider that better accuracy and excluding human involvement with the reading,
would or could  lead to different results?
You will find that in the end it could be exactly as I said: we only warmed some 0.3 or 0.4K  from 1950 - 2000
and we will cool by about 0.3 or 0.4K until 2040,
as we follow the global Gleissberg solar/weather cycle.....
So by 2040, we will all be back to where we were in 1950.....
Blaming natural forces all makes more sense, than blaming the poor 0.01% CO2, does it not?

[Bored chemist]

"I think it was only after the war. ./.as proven by the certificate you provided...."
That doesn't prove what you say it does.
Why are you pretending that it does? Do you not understand that if I show you today's paper it doesn't prove that there was no newspaper yesterday?

"I asked: from what time onward in history did we start re-calibrating thermometers at regular intervals?"
You also haven't explained why you think thermometers need to be recalibrated- I asked (and you ignored it) whether you think it's the glass that changes or the mercury.
Well, which is it?
While I'm at it,
"I remember that even in my time, a mercury thermometer with one degree C divided in ten portions, with an accuracy of 0.05 was quite rare and expensive."
So what?
Before the times you are speaking of, they had thermometers that could read to a thousandth of a degree, albeit that you wouldn't use one for measuring the weather.
http://en.wikipedia.org/wiki/Beckmann_thermometer
They were not cheap- but, because people were prepared to put a fair bit of time and effort into this sort of thing, they would be prepared to spend the money too.

"If the accuracy was in fact only about 0.5, as I suspect you will find, then the the current warming rate of 0,6 or 0.7K per century,
can be considered questionable....and seriously challenged."
No, not really.
because the accuracy isn't the issue here, it's the repeatability that matters.
Also, not all the thermometers would misread in the same direction. So the errors would cancel out.
If you have a hundred measurements -say ten people and ten readings, then the average of those reading will be about 10 times better than the error on each thermometer.


"Curious, that you refuse to consider that better accuracy and excluding human involvement with the reading, would or could  lead to different results?"
It's curious, no- let's face it- dishonest that you claim that I didn't consider it. I did. That's why what I said was that the average of a large number of results will not be affected much if the individual results are made more accurate.
I'm allowed to say that because it follows from a proven theorem in maths.
http://en.wikipedia.org/wiki/Central_limit_theorem

"Blaming natural forces all makes more sense, than blaming the poor 0.01% CO2, does it not?"
No.
It's the equivalent (as I have said before) of having 3 blankets on the bed, adding a forth, and saying that you don't expect it to make any difference.

[MoreCarbonOK]

BC says
blaming natural forces all makes more sense, than blaming the poor 0.01% CO2, does it not?"
No.
henry says
Yes, it is exactly like you said:
let us have a planet, let us add some CO2, let us see if the temperature goes up,
IT DID NOT, for at least 15 years and counting, so that is NOT it....
Pity you could not shed me some light on the accuracy of thermometers used 75-100 years ago,
in weather stations, but I will still carry on looking for that elusive certificate from those days...
In the meantime, here is a good evaluation done by one of your countrymen,

http://climate.arm.ac.uk/publications/global-warming-man-or-nature.pdf

He got it right, mostly, just by looking at the available data, he even mentioned the Gleissberg cycle!

I think he just went wrong in the end with the mechanisms that he mentioned  for the warming/cooling, I find it is only partly correct, or it is actually a combination, but very few people have that figured out right yet......

For the rest, this whole paper sums up nicely what I have been trying to tell you.
Rgrds.
H

[JP]

This is a nice article on how prediction and scientific modelling works, and links to peer-reviewed work offering a possible explanation for the past decade of temperature readings.

http://blogs.scientificamerican.com/the-curious-wavefunction/2013/05/15/global-warming-slowdown-retrospectively-predicted/

[alancalverd]

Historic data on global temperature is at best "suspect". Most of the earth's surface is covered by the oceans, for which we have virtually no reliable surface temperature data before the 1950s. Nobody had measured any temperatures in Antarctica before 1900, and the arctic ice cap had not been greatly explored until the 20th century either.

Whilst plenty of thermometers existed in the 18th and 19th centuries, manufacturing variations and the lack of national and international standards laboratories meant that only 0 and 100 deg C measurements can be regarded as equivalent - any intermediate temperature report, and certainly any report below 0 deg C, must be regarded as having a tolerance of at least +/- 0.1 degree and a difference of 0.2 degree between two adjacent thermometers at a nominal 10 deg C would not be remarkable.

Until 1910, accurate (better than 0.5 degree) air temperature measurements were of little interest, and until the adoption of the Stevenson screen around 1870, there was no agreed method of making such measurements. Surface air temperature is principally of interest to aviators, and we can consider the second world war as the beginning of consistent temperature recording of sufficient accuracy, frequency, and precise location, to detect global changes. The most complete data sets therefore date from 1945, when the world's air forces were making 4-hourly readings with credible instruments at thousands of airfields and a few ships. The number and frequency of manual reports decreased  thereafter, and their nature changed: most wartime airfields were rural grass strips, nearly all of which closed in the following 15 years, and most of the remainder acquired extensive tarmac and concrete surfaces, with canvas tents being replaced by steel and concrete buildings. Temperature reporting from rural fields nowadays is far less frequent as they are mainly used by light aircraft, well below any critical loading: the best modern data comes from large urban airports whose microclimates (thanks, inter alia, to several tons of fuel being burned every hour inside the boundary fence) are far from typical of the "natural" area even a few miles away.

It is thus fair to say that any non-satellite data is at best sparse, poorly representative of any meaningful "mean global temperature", and seriously biassed by microclimates that become more atypical with time. If we add in the probable technical uncertainties of ambient temperature measurement before, say, 1900, the scientific conclusion must be that there is no credible data on mean global surface temperature before 1940 or after 1960.       

Satellite data should be more reliable, but there have been several major corrections and recalibrations since 1970, and the magnitude of these step changes is commensurate with any supposed real change in the underlying measurand.

In a word, we have no actual data sufficient to support an assertion of a consistent global warming or cooling trend  within the last 150 years or so. We do have good proxy data suggestive of massive climate shifts over geological history, but nothing of sufficient reliability to implicate recent human activity as causal.   

[alancalverd]

"I asked: from what time onward in history did we start re-calibrating thermometers at regular intervals?"
You also haven't explained why you think thermometers need to be recalibrated- I asked (and you ignored it) whether you think it's the glass that changes or the mercury.
Well, which is it?

The glass, as it happens. There are three major problems: change in volume of the bulb (it is subject to thermal expansion, creep, and distortion by atmospheric compression), variations of stem bore diameter (which determines the precision of intermediate readings between the fixed points) , and the fact that mercury freezes around -40C, making our favourite meteorological thermometer useless over a good 10% of the earth's surface and most of the atmosphere above it!

"I remember that even in my time, a mercury thermometer with one degree C divided in ten portions, with an accuracy of 0.05 was quite rare and expensive."
So what?
Before the times you are speaking of, they had thermometers that could read to a thousandth of a degree, albeit that you wouldn't use one for measuring the weather.

There's the rub. Beckmann and other high-sensitivity thermometers generally measure changes over a small range of temperature, where the absolute value of the starting point is unimportant. I spent a long time (13 years!) working on devices to detect a microdegree change in local temperature but with no absolute reference point. To detect a meteorological trend even in the UK, where the climate is far from extreme, you need to be able to measure absolute temperature to better than 0.01 degree over a range of -30 to +40 degrees, for about 100 years.  I don't know of any device that has actually done this.

[Bored chemist]

The glass tubing does, indeed, creep. But properly made thermometers are annealed carefully to minimise this. Also, it's a known effect and would have been one of the reasons why they would have know that they needed to check the calibrations from time to time.
The freezing point of mercury doesn't change over time- in fact you can use it as a fixed point for calibration.
The fact that people have to use alcohol, or pentane in low temperature thermometers isn't really an issue.

It's trivially simple to check the constancy of the bore of a piece of tubing of the sort used in thermometers, all you need is a drop of coloured water.
It would be an interesting exercise to consult the old chemistry journals and look up the reported melting points of pure chemicals when they were first reported and compare those values with modern measurements. That would let you check the accuracy of the thermometers they used.In the meantime, I think it's fair to assume that they knew what they were doing and, at least, checked the ice and steam points from time to time.

It's also true that good records of temperature only exist for populated areas, but don't forget that, those are the only areas where we have a direct interest in knowing what the climate is doing.
In principle, you can detect a trend in the climate with just one thermometer in one place.
There is, of course, the fact that it the climate warms in one place, it might cool in another but that doesn't stop you being able to measure the local trend. We had no idea what the temperature was in the Antarctic until recently because nobody was there- but we didn't care for exactly the same reason.
You need global temperature measurements if you want to model the system.
But, you can detect changes with just local measurements.



It's obviously a lot easier to sort out the signal from the noise if you have many measurements.
Well, we have lots.
The large number of measurements reduces the scatter of the average of those readings.

As you say, the Beckmann thermometer and those like it only measure changes in temperature.

Fine, that's exactly what we are trying to measure.

And the current rate of change means that over the last 30 years the temperature has risen by about half a degree.
Why would you say that you need to measure to 0.01 degrees over a century to track this trend?