Guest post by Clive Best
Perhaps like me you have wondered why “global warming” is always measured using temperature “anomalies” rather than by directly measuring the absolute temperatures ?
Why can’t we simply average the surface station data together to get one global temperature for the Earth each year ? The main argument to work with anomalies (quoting from the CRU website) is: ”Stations on land are at different elevations, and different countries estimate average monthly temperatures using different methods and formulae. To avoid biases that could result from these problems, monthly average temperatures are reduced to anomalies from the period with best coverage (1961-90)….” In other words although measuring an average temperature is “biased”, measuring an average anomaly (deltaT) is not. Each monthly station anomaly is actually the difference between the measured monthly temperature and so-called “normal” monthly values. In the case of Hadley Cru the normal values are the 12 monthly averages from 1961 to 1990.
The basic assumption is that global warming is a universal, location independent phenomenon which can be measured by averaging all station anomalies wherever they might be distributed. Underlying all this of course is the belief that CO2 forcing and hence warming is everywhere the same. In principal this also implies that global warming could be measured by just one station alone. How reasonable is this assumption and could the anomalies themselves depend on the way the monthly “normals” are derived?
Despite temperatures in Tibet being far lower than say the Canary Islands at similar latitudes, local average temperatures for each place on Earth must exist. The temperature anomalies are themselves calculated using an area-weighted yearly average over a 5×5 degree (lat,lon) grid. Exactly the same calculation can be made for the temperature measurements in the same 5×5 grid which then reflect the average surface temperature over the Earth’s topography. In fact the assumption that it is possible to measure a globally averaged temperature “anomaly” or DT also implies that there must be a globally averaged surface temperature relative to which this anomaly refers. The result calculated in this way for the CRUTEM3 data is shown below:
Fig1: Globally averaged temperatures based on CRUTEM3 Station Data
So why is this never shown ?
The main reason for this I believe is that averaged temperatures highlight something different about the station data. They instead reflect an evolving bias in the geographic sampling of the station data used over the last 160 years. To look into this I have been working with all station data available here and adapting the PERL programs kindly included. The two figures below show the location of stations used dating from 1860 compared to all stations.
Fig 2: Location of all stations in the Hadley Cru set. Stations with long time series are marked with slightly larger red dots.
Fig 3: Stations with data back before 1860
Note how in Figure 1 there is a step rise in temperatures for both hemispheres around 1952. This coincides with a sudden expansion in included land station data as shown below. Only after this time does the data properly cover the warmer tropical regions, although there still remain gaps in some areas. The average temperature rises because gaps for grid points in tropical areas are now filled. (There is no allowance made in the averaging for empty grid points neither for average anomalies nor temperatures). The conclusion is that systematic problems due to poor geographic coverage of stations affects average temperature measurements prior to around 1950.
Fig 4: Percentage of points on a 5×5 degree grid with at least one station. 30 % is roughly the land surface of Earth
Can empty grid points similarly affect the anomalies? The argument against this, as discussed above, is that we measure just the changes in temperature and these should be independent of any location bias i.e. CO2 concentrations rise the same everywhere ! However it is still possible that the monthly averaging itself introduces biases. To look into this I calculated a new set of monthly normals and then recalculated all the global anomalies. The new monthly normals are calculated by taking the monthly averages of all the stations within the same (lat,lon) grid point. These represent the local means of monthly temperatures over the full period, and each station then contributes to its near neighbours. The anomalies are area-weighted and averaged in the same way as before. The new results are shown below and compared to the standard CRUTEM3 result.
Fig5: Comparison of standard CRUTEM3 anomalies(BLACK) and anomalies calculated using monthly normals averaged per grid point rather than averaged per station (BLUE).
The anomalies are significantly warmer for early years (before about 1920), changing the apparent trend. Therefore systematic errors due to the normalisation method for temperature anomalies are of the order of 0.4 degrees in the 19th century. The origin of these errors is due to the poor geographic coverage in early station data and the method used to normalise the monthly dependences. Using monthly normals averaged per lat,lon grid point instead of per station causes the resultant temperature anomalies to be warmer before 1920. Early stations are concentrated in Europe and North America, with poor coverage in Africa and the tropics. After about 1920 these systematic effects disappear. My conclusion is that anomaly measurements before 1920 are unreliable, while those after 1920 are reliable and independent of normalisation method. This reduces evidence of AGW since 1850 from a quoted 0.8 +- 0.1 degrees to about 0.4 +- 0.2 degrees
Note: You can view all the station data through a single interface here or in 3 time slices starting here. Click on a station to see the data. Drag a rectangle to zoom in.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Hi, I’d love to interview you on how mean global temperatures are recorded/determined (and by whom). Also global cooling – which I’m starting read about. It would be an in ‘layman’s terms’ chat. Karyn Wood ABC Coast FM (in Australia)
I take it that once CAGW raises the temperatures of the Northern Hemisphere by, say, 4 deg C,
(which would be possible given the high end of the IPCC climate sensitivity estimates for a doubling of CO2) we will have the same apocalyptic temperatures in Germany as, say, Rio de Janeiro?
Oh the horrors! Quick, somebody erect me a wind turbine.
So the argument is that “the great dying of the thermometers” led to the rise in temperature anomalies.
This is quite believable. I think if you care to look closely, the current temperature plateau coincides with the cessation of the thermometers dying off.
They will have to start kicking out more stations (the lower reading ones, of course) sometime soon, to get the curve to continue its upwards trend.
But fig.1 shows no warming at all for the SH, at least from 1952 onwards.
How can this be?
Is ‘global warming’ a phenomenon that only affects the Northern hemisphere?
Cool!
Dissecting the IPCC-report – one station at a time.
Well Done.
Mike
I appreciate the effort you put into this article BUT note the lack of standard error/deviation bars. You should know better. Provide some sort of statistical measure of sampling error, at least, no matter how bogus it’s basis. We need not delve into the other sources of error in this data since data reliability is quite tenuous as it is.
The fact remains that a non-random sample of 1 (N=1) will NEVER provide a “reliable and independent” sample from ANY sort of population, let alone something as complex as temperature.
GIGO – no “scientist” with any background in sampling and basic stats would ever rely on this kind of crap data to draw any sort of accurate conclusion. Certainly reporting results to tenths and hundreths of a degree are totally absurd.
I have been saying this for years. The reportage of anomalies is a way of not reporting actual temperatures. It not only accentuates by a factor of 100 the so called anomalies, but they are impossible to verify. There is no substantiating data.
If current climatology were a science that needed detailed data, it would be a disaster. If it were a weather report, it would be off the air as an incompetent charade.
A couple of questions:
When deriving the global average temperatures above, how is the ocean area taken into account? Or is the temperature a land surface temperature only?
Also, how does this sort of data compare to the sattelite data?
Another question which comes to mind relates to the averaging calculation. How are the different stations weighted? Are all stations within a grid (5×5 degree) cell given the same weighting? This can be important as the data looks highly clustered and could be biasing the calculation (either up or down) due to more stations in certain areas. In geostatistics (I’m a geologist invloved in mineral resource calculations) we typically employ Kriging algorithms to provide an unbiased (or as unbiased as possible!) estimate of metal concentrations based on spatial data….I think the concept could be adapted for climate analysis.
@Jos Verhulst
That was my first thought too, but if GW only affects NH, then the cause of GW has to be something else. If CO2 was the real culprit, then the warming should be more or less uniform over the globe (presumably subject to lag from the redistribution of CO2 from the NH to the SH).
Also, if the alarmists are correct, shouldn’t the rise in temperature be easily predictable from the individual temperature record and the known warming caused by CO2?
For instance, take the CET and measure the CO2 level, then calculate the difference in temperature between 1750 (280ppm CO2) and 2010 (390ppm CO2), and compare against what the rise in CO2 should have caused.
Actual temperature is not used because you cannot assume that the reading of one thermometer in one location is a valid temperature representative of a large area. Not even in the surrounding kilometer. Local conditions affect the temperature too much, so the absolute data is quite meaningless. However, the temperature anomaly CAN be assumed to be about the same in a large area. Because local conditions change in space, but remain quite the same in time. So if one spot is 2C higher than normal, it is quite reasonable to assume that, whatever the temperature is 1 kilometer further, it will be about 2C higher than normal as well for the corresponding location.
In other words, it is not possible to recreate a Global Temperature with a few thousands of thermometers (what you would get is an average temperature of the specific places where you have thermometers, and not of the Globe), but it IS possible to do the same with the global temperature anomaly. It may not be perfectly accurate, but it is a very good aproximation (well, at least it would be, if they didn’t move the thermometers, and their surroundings didn’t change, and temperature registers were long…)
And more to the point, we don’t care what the global temperature is, but how it is changing. And for that, the anomaly is all the information you need.
So, how do you compute the averages? Linear or using the fourth power?
(15 + 25)/2 = 20 C
or
(((15+273)^4 + (25+273)^4)/2)^0.25 – 273 = 20.13 C
This makes a big difference. For example, there are several ways to increase the linear average by only 1 C. But the computed average using the fourth power is different.
if 15 goes to 17, then dT = 0.95
if 25 goes to 27, then dT = 1.06
If they both increase by 1, then dT = 1.00
This may look insignificant, but when averaged over the whole planet for a full year, the two methods produce significantly different values.
Nylo nailed it. An interesting exercise to ‘construct’ a Global temperature but not very usefull. However, the upwards trend in the 20’s and 50’s caused by an increase in measuring stations IS an interesting observation. Is there, as some commenters have asked above, a different temperature trend between the Southern and the Northern hemisphere? Is this because because the Southern Hemisphere has significantly more ocean and much less land?
Water heats up and cools down more slowly than land but would this effect temperatures in the long term?
A fundamental question is whether the data series suffers from aliasing. Perhaps not so much in the time domain, but surely in the spatial domain with sparse or even non-existent coverage of large parts of the surface.
Until this question is fully addressed, I would recommend these series are treated as not reliable. Error bars produced from aliased data would be similarly unreliable.
Just give me 30 years. I will have a preliminary result in about 10.
The absolute value is meaningless but it does show if it’s gone up or down.
Just click on the Dave A above
Straight average from the same set of 2586 stations taken hourly, end of, no gridded extrapolation. No waiting for days while it is “worked out” it just is what it is. As they “die off” I just strip them from the data and re-run every hour from day 1 again. 27,253,962 individual temperature measurements so far from (alphabetically) Afghanistan to Yeman and most places in between. Being NOAA a heavy bias towards North America. If the Globe really is going to be 6C warmer by the end of the century, on account of a trace gas responsible for the entire food chain of the planet, you will see it reflected in the graphs.
As sure as eggs are eggs
Watch an experiment in progress. The website will be developed further as the tabs allude to.
I promise never to let the dog eat my raw data.
Around 12.8C on our last trip around our local Star. What do you think the chances of it being 18.8C by 2100?
Nylo
It is possible, if you have enough nicely spaced thermometers, to get a pretty good average ofthe temperature of a defined region and the trend in that average. You have to correct or normalise for regional altitudinal and latitudinal effects but that is not too hard to do.
To get the bigger picture you can then combine regional averages, weighted by area if necessary, to get a global land temperature and trend. This does not require one to first anomalise the data. It alows one to use abundant data from numerous relatvely short lived climate stations in addition to the normally preferred long-lived stations. Naturally there would be plenty of issues to work through but it would not be impossible.
Note that there is more data out there in regional and national databases than is present in GHCN so any thought that the data are too sparse is basically untested.
Interesting and provocative piece.
Thank you for writing it.
local average temperatures for each place on Earth must exist
Global temperatures display a 1/f type noise. That is to say, the longer you wait, the higher the noise level. In principle, the noise level is infinite for an infinite period.
That means that there is not an average temperature for any place on the earth.
In other words, it means that no matter how long the period, the next equal length period will have a significant difference in temperature. Unlike gaussian noise where the noise decreases if you take a long enough sample period, 1/f noise just doesn’t go away, averaging doesn’t get rid of it … indeed in a real sense, the longer the period, the higher the noise!!
@BioBob I agree with you – all such plots should have error bars. However – none of them ever do! The statistical errors on the annual averaged values change each year as more stations are added and subtracted. The CRU people quote a statistical error of 0.05C after 1951 and 4 times that in 1850(0.2C). However it is the systematic error that really matters. The systematic error depends on the way you calculate the anomalies. By calculating the normals in a different way I got up to a 0.4C shift in values in the 19th century. Furthermore the annual values are derived from a grid with 2592 cells. Many of these cells have no points at all, others have just one station. So the error per (lat,lon) cell is large.
As an interesting aside GCMs of course work in absolute temperatures (and anomalies are then calculated from them). The absolute temps show bias between models as shown in Slide 20 of the ppt off http://www.newton.ac.uk/programmes/CLP/seminars/082310001.html that using the anomalies tends to hide.
Clive,
An interesting observation and one I’m not qualified to comment on. Your first responder, Karyn Wood, would like an interview with you. I don’t know her work but I do know the ABC is a fervent flagwaver for AGW. I have never heard a sceptical view expressed on the aBC without some follow up from a believer who strives to debunk the words of the sceptic. If you do respond be aware that you may be set up. The ABC in Australia is the least trusted media by anyone who wants the truth.
Is this a fair summary: half of the suggested 20th Century `warming’ may be an artefact of trying to construct a long term record from data taken from incomplete and incommeasurable measuring systems?
And if the actual 20th Century temp rise is, in fact, just 0.4C, that the IPCC forcing assumptions are totally undermined?
So, the whole thing is (more likely than not) utter and total crap? The greatest make-work science-off-the-rails episode in all history?
Where are the hordes of indigant scientists, banging on the doors of their national governments, demanding the reputation of science be no longer trashed with this s%&t?
Can’t you guys get your act together? Trillions of dollars (as a proxy for wasted human effort) are at stake. Get it together, guys, and take these shysters down, please!!!!
@JosVerhulst, Mango Chutney.
Your assumption concerning CO2 might be correct.
After WW2 the progress in the NH was substantial, changing the albedo and increasing the H2O emission. Especially the H2O emissions above the lager NH landmass by coolingtowers , riversurfacewatercoolers, irrigations (for example the famous Soviet example, Egypt lake Nasser and subsequent irrigation), combustion(change from coal to natural gas), population increase. More cloud cover might be expected in autumn and early winter.
So it will be interesting what the coming 5 to 10 years will bring in the expected cooling.
If it outweighs the increase or even goes lower in temperature, this will put the CO2 problem ad acta.
FTA: “The basic assumption is that global warming is a universal, location independent phenomenon which can be measured by averaging all station anomalies wherever they might be distributed.”
I see no basis for such an assumption. The change in temperature should be proportional to the local temperature, which in turn is dependent on latitude, wind currents, and local heat capacity, at the very least.
I have never seen a satisfactory explanation of what information is contained in the average of this intensive variable. What does it even mean?
Jos Verhulst says:
February 7, 2012 at 10:22 pm
“Is ‘global warming’ a phenomenon that only affects the Northern hemisphere?”
Sure looks that way, doesn’t it? Which, of course, appears immediately to falsify the assumption above.
Nylo says:
February 7, 2012 at 11:22 pm
“It may not be perfectly accurate, but it is a
very goodaproximation.”If I got the “strike” tag to work through the “very good” portion, I will have fixed that for you. How do we know how good it really is?
Jordan says:
February 8, 2012 at 12:04 am
“A fundamental question is whether the data series suffers from aliasing…Error bars produced from aliased data would be similarly unreliable.”
Or, at least, tentative. This is a big question with me, too. What is the spatial frequency distribution? Does anyone know? Maybe satellite data could determine this.
I have recommended doing a fit to a spherical harmonic expansion instead of all this ad-hoc area averaging. At least doing that, there is a little relief from the sampling theorem (Shannon-Nyquist is a sufficient condition for being able to retrieve a signal from sampled data, but not strictly necessary). Nick Stokes claims to have done it and seen little difference, but he was not forthcoming with the details of precisely what he did.
Scottish Sceptic says:
February 8, 2012 at 12:58 am
“That is to say, the longer you wait, the higher the noise level.”
True, that can happen. The question is, though, how much is signal, and how much is noise? Music tends to produce a 1/f frequency spread, too. But, aside from some of the grunge bands popular when I was in grad school, it’s not generally all noise.
@John.
1. The data I am using are just the land station data. This is called CRUTEM3 in the jargon. The combined data set HadCruT3 includes sea surface temperature data provided by Hadley centre. However, including the sea surface data changes the anomalies only slightly ( they are reduced). It is the land data which is the main driver for warming. You can see that here
2. The station data are first sorted into (lat,lon) grid points. The data may be either the anomalies or the actual temperatures. Then a simple average is made over any grid points with 2 or more stations in them. The result is a grid file with 72×36 points for each month from 1850 to 2010. Remember that many of these grid points are empty.
The monthly grid time series is then converted to an annual series by averaging the grid points over each 12 month period. The result is a grid series (36,72,160) ie. 160 years of data.
Finally the yearly global temperatures/anomalies are calculated by taking an area weighted average of only the populated grid points in each year. The formula for this is $Weight = cos( $Lat * PI/ 180 ) where $Lat is the value in degrees of the middle of each grid point. All empty grid points are excluded from this average.
@Nylo.
A good argument for using anomalies. But note the implicit assumption you are making : Temperatures can’t rise (or fall) locally – they can only be part of a global phenomenon (caused by rising CO2) – now I will go and measure it under this assumption.
One other problem with your argument is that the anomaly is also only measured at one place. In many cases we have a 5×5 degree grid point spanning 90,000 square miles containing just one station ! So a change of 0.2 +- 0.2 degrees at one station in one year is interpreted as applying to a huge surrounding area. Perhaps that station is also in the middle of town which grows over many decades.
I think it must be possible to define a single global temperature since otherwise Energy balance arguments become circular. There must be an effective average surface temperature of the Earth.
@Robert Clemenzi
I am using exactly the same procedure as the Hadley CRU team. I am actually using and extending their software. The averaging they use is very simple. The monthly temperature for each cell in the 5×5 degree grid is the simple average of all stations lying within the cell. They use your first equation.
However I like your second equation better – so I will try and use that one and see what the result is !
@Dave
Good work – I like it very much.
The ‘average’ temperature of Earth is an impossible concept even using satellites though these give a better result that the surface system in use today. Taken over a few million years the average could be higher or lower than today’s depending on the distance back in time chosen.
In fact there is no correct ‘average’ temperature for the planet. The one you have to suffer is the correct one for that time but not the average.