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.
I’m sure I’m not the first to post this thought, but using an anomaly as a measuring stick allows for an arbitrary baseline to be used and therefore the warming may be exaggerated. Exhibit A: The lowering of historical temperature data makes the late-20th century warming anomaly appear greater.
I think that Phil Jones recently admitted that 30% of all the stations showed a temperature decline. This strongly argues against any kind of global mean. Or to paraphrase a famous political saying, all climate is local.
EM Smith aka Chiefio has also analysed temperature records in depth. He adopted a different method. Along the way he discovered and revealed the changes in station counts and commented on their implications. They can be found here: http://chiefio.wordpress.com/category/dtdt/
Among other things, his approach identified what he called pivot points that marked significant changes in the stations used to record temperatures, notably around the 1990/91 period when the overall number dropped from c6000 to c1200 with only 200 of those stations common to the pre and post 1990/91 periods. He observed that scientists do not change their measuring equipment during the course of an experiment and then expect to get a valid, usable result. Yet that is a problem in the historical data sets used in climate science. It is a particular problem when the 1961-90 baseline includes c6000 stations, and the post 1990 period only includes c1200. As far as I am aware, no one knows whect effect this has on the reported temperature anomalies.
A word of warning to those that visit his site. You need time to spare for he embarked on an epic journey analysing temperature records in every country that reported data, producing graphs for each one.
oldtimer says:
February 8, 2012 at 8:07 am
EM Smith aka Chiefio has also analysed temperature records in depth. He adopted a different method. You need time to spare for he embarked on an epic journey analysing temperature records in every country that reported data, producing graphs for each one.
The best analysis I have seen to date, backed up by others in the Comments who have done similar analysis on individual ares.
@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.”
Exactly. NIWA immediately comes to mind. 11 or so weather stations, all but 2 in the same location for many years. Yet the use of anomalies rather than real data covered up the fact the real data was constantly ‘adjusted.’ NIWA seemed unaware that the historic readings were actually available to the public. A quick review of the data showed that the sharp rise in New Zealand’s temperature, as measured via the anomalies was in fact nonexistent, and individual station increases likely due to UHI.
Nylo says at 11:22 pm
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.
Nylo,
For the anomaly to remain about the same across the 5×5 gridcell, you are accepting an underlying assumption: That neither the weather nor the climate changed here.
You’re right about how bogus measuring ‘temperature’ with a ludicrously low number of thermometers is, but none of these thermometers have actually been ‘calibrated’ into determining gridcell anomalies. They’re used ‘as is’.
The instrumental error bars are not appropriate measures of either actual temperature or anomaly measurements for the gridcell.
The ‘anomaly method’ seems quite reasonable for getting -some- data, and the way the math all ends up with unknowables cancelling out is definitely fortuitous. But the omnipresent assumption that a given thermometer has a constant relationship with the true anomaly even in different weather and climate patterns, and the failure to calculate sane error bars are both still glaring and fatal flaws.
Clive
An interesting article, thank you. For those that prefer real temperatures rather than anomalies clicking on my name will take you to my site where I collect (mostly) pre 1860 temperature data sets from around the world expressed in real terms. Its also here;
http://climatereason.com/LittleIceAgeThermometers/
However, anomalies do help to better compare temperature changes between the various highly disparate data sets so they do perform a worthwhile function. However they have had the side effect of becoming the basis for a meaningless (in my view) single ‘global’ temperature.
A global temperature has a number of problems, not the least of which is that it disguises regional nuances.Around one third of global stations have been cooling for some time as we observed in this article below. Separating out the individual stations from the composite of stations used to create an ‘average global temperature’ shows that warming is by no means global as there are many hundreds of locations around the world that have exhibited a cooling trend for at least 30 years-a statistically meaningful period in climate terms.
http://wattsupwiththat.com/2010/09/04/in-search-of-cooling-trends/
These general figures were confirmed by the recent BEST temperature reconstruction which reckoned that 30% of all the stations they surveyed were cooling. Many of the rest (but by no means all) are in urban areas, which many of us believe do not reflect the full amount of localised warming caused by buildings/roads etc. Add in that many stations are not where they started out and have migrated to often warmer climes such as the local airport, and that many stations have become replaced by others or been deleted, and we start to see an immensely complex picture emerging where we are not comparing like with like.
There is a further complication with lack of historic context. For reasons best known to themselves GISS began their global temperatures at 1880 and as such do not differentiate themselves enough to Hadley which began thirty years earlier. I suspect this date was chosen as this was when many of the US stations started to be established, but as regards a global reach, a start date around 1910 or so would bring in more global stations and have the advantage of greater consistency, as by that time the Stephenson screen was in almost universal use.
The start date of 1880 does not allow the context of the warmer period immediately preceding it to be seen, which means the subsequent decline and upward hockey stick effect is accentuated (the hockey stick uptick commenced with instrumental readings from 1900) . I wrote about the 1880 start date here; where I link three long temperature records along the Hudson river in the USA.
http://noconsensus.wordpress.com/2009/11/25/triplets-on-the-hudson-river/#comment-13064
I think the most we can say with certainty is that we have warmed a little since the depths of the Little Ice age, which would surely come as a relief to most of us, but instead seems to be the source of much angst afflicting most of the Western World, who apparently have stopped learning history and are confused by statistics and context.
I hope you will be continuing your work and develop your ideas.
Tonyb
Clive@Allan MacRae
I agree that there is a tendency in the AGW mainstream to ignore data which doesn’t quite fit the story. I also find the whole paleoclimate debate fascinating. What really causes Ice Ages ? Why have they been coincident with minima of orbital eccentricity for 1 million years ? I have spent a long time thinking about this.
The next Ice age will begin in ~2000 years time. Can global warming save us ?
__________________________________________________
Hi Clive,
Re your question: Can global warming save us from the next Ice Age?
My best guess answer is no, not even close. Even if the mainstream argument is correct (that CO2 drives temperature), this will be no contest – like firing a peashooter into a hurricane.
Geo-engineering might work; don’t know.
Ironic isn’t it, that our society is obsessed with insignificant global warming, just prior to an Ice Age?
I recommend Jan Veizer and Nir Shaviv on the science.
Clive Best: “The way the annual global temperature anomaly is calculated assumes a global phenomenon because it uses a simple annual average af all (area weighted) individual station anomalies.” – actually no, that’s not how it is done. Even if it was, that would not imply anything about the global or otherwise nature of the change measured.
“The geographic spread of stations is such that the averaging will bias the global average to those regions with lots of stations. So if North America rises by 2 degrees while the Sahara falls 2 degrees then the net result will be strong positive. That is my main point. The result assumes global phenomena and rules out local phenomena.”
I’m afraid you’re totally wrong. Biases could indeed be introduced by coverage issues, but this has absolutely no bearing on whether phenomena are local or global. And one can estimate the size of those biases, and this has of course been done. Why do you think GISS temperatures only start at 1880, even though there are temperature records which go back two centuries before that? Have you, in fact, read any of the papers describing the GISS and other ground-based temperature datasets?
Clive,
One recurring question I have that you might be able to answer would involve an actual calibration with the satellite data.
That is:
Pick a gridcell.
Calculate the ‘average’ for that gridcell.
(You’ve done this for all the cells.)
Now:
Do a -calibration- between a single gridcell and the satellite data’s best estimate of surface temperature for that same cell.
There are regular reports that the surface stations ‘agree in general’ with the satellites. But these are -correlation- studies, not intended to be calibrations. And they tend to compare the end results (GMST) instead of comparing gridcell-by-gridcell.
This can arrive at an estimate of an actual error for a gridcell’s temperature measurement.
Even if anomalies are calculated on a station-by-station basis, it seems to me that, in general, 50% of the data should be in the positive range and 50% should be in the negative range (unless the date range for the “mean” is different than the data range). So when I look at Figure 2 of Mann et. al. (2012) it doesn’t seem to fit that impression, as less than 10% of the data is in the positive region.
[IMG]http://i44.tinypic.com/r1zmhf.jpg[/IMG]
Clive Best asked about Fig1: Globally averaged temperatures based on CRUTEM3 Station Data,
“So why is this never shown ?” Well I don’t know, but if I had to guess it’s because it’s not scary looking enough.
“…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…”
The problem with this assumption is that we actually know what “normal” is. The charts are drawn so that it’s eiter above or below “zero”, with no explination of how the “zero” was determined.
I’ve constantly harped on the fact that each database uses a different grouping of stations, adds or drops the Arctic, and uses a different averaging period.
Since we can all agree that the earth is warming, we can state that any past averaging period will be colder than today. As your averaging period moves closer to today, the “zero” will change.
They get around this by stating “well, it’s the TREND that’s more important, not the zero”.
This is a problem.
Let’s say that your accountant told you you’ve seen a 30 dollar rise in your income. Have you:
1. Started at 45 below, and are now 15 below,
2. Started at 30 below and are now at zero,
3. Started at 15 below and now at 15 above…
You can see that the trend is the same, and can continue on this track.
Add to that the fact we’re only tracking trends of 1 degree of rise, makes the use of a “zero” on an anomaly chart useless.
Thanks for everyone for the comments.
@Volker Doorman
I agree with you that averaging over monthly and local data to get just one temperature per year is not that smart. However a cynic might suspect this is exactly what is being done to produce “Hockey Stick” type graphs for the public as hard evidence of AGW.
@climatereason
Also agree that the local effects are important. I once did this back of the envelope calculation for the Urban Heating effect and came up with reasonably large values.
Total Average World Energy consumption Rate ( fossil, nuclear,hydro) = 15 TW (wikipedia, for 2005 and increasing by 2%/year). My guess is 80% is eventually converted to heat (2nd law thermodynamics)
Land Surface Area of the Earth = 150 x10**12 m2 of which urban areas are ~1.5%
If we assume that this energy consumption is concentrated in these urban areas then the human heating effect there works out at 5.5 watts/m2.
Radiant energy from the sun is distributed unevenly on the earth’s surface but the average absorbed energy globally is 288 watts/m2. This energy is then radiated from the surface as heat (infrared) and can be comparable to the human generated heating also at the surface.
Direct heating by Man in urban areas comes out at approximately 2% of direct solar radiant heating.
These are just ball park figures and will depend on the latitude of the urban areas and their locations. Also not all human energy consumption is generated in urban areas but counteracting this is the fact that most large cities are in high northern latitudes. The anthropogenic effect at night and in winter could be even higher.
What temperature increase does this lead to ? We use Stefan Boltzman’s law.
(T+DT)**4 = 1.02T**4
(1 +DT/T)**4 = 1.02
DT = (0.02/4)T ( T = 285)
DT = 1.4 degrees ! (i.e. averaged temperature increase in urban areas)
@Clive Best
Cheers – I am glad that you like it, I’ll keep going. Hopefully one day it will become a useful resource in this debate. I was appalled by the contents of the ClimateGate emails and decided I needed to know the truth. I grew tired of the “It’s warm in Wagga Wagga” jibes. Incidentally this is Wagga Wagga since the beginning of the year http://www.theglobalthermometer.com/igraphs/stations/YSWG.png – there’s plenty of ways to splice the data once collected 😉
@P Solar
“You are using monthly mean and attributing it to the end of the period not the middle. I would guess by that error that you will also be using a running mean. That is one god-awful frequency filter.”
Spot on – it’s the 30 day running mean updated hourly. It probably represents the average temperature of around a couple of weeks earlier. As I say, it is what it is, raw and unadulterated.
Third Reason already alluded to by previous commenter, there is no global warming in the Southern Hemisphere since 1952 as shown by figure 1. The GAT is derived from COMBINING the NH and SH temps, hence through unethical (or incompetent) mathematical trickery the GAT is biased by the NH temp trend. This is a violation of basic mathematical theory and statistical practice which why most engineers and meteorologists reject AGW. The entirety of the AGW hoax should have been unravelled years ago by focusing on this embarrassingly obvious glaring fact. No need to calculate W/m^2, blah, blah, blah as this is an obfuscation OF THE DATA. No matter how flawed the siting of the temp recording system or any of the other multitude of issues, the irreconcilable fact of ZERO SH positive temp trend falsifies the AGW hypothesis, not withstanding the hand waving using water vapor.
AGW is not a theory as it is not universally recognized by the scientific community, it never has been when such notable people were dissenting, e.g. Fred Singer, the father of the US weather satellite system, Bill Gray, hurricane researcher, etc, That’s why the AGW hoaxers were always playing upon the imprimatur of “consensus”.
“Perhaps like me you have wondered why “global warming” is always measured using temperature “anomalies” rather than by directly measuring the absolute temperatures ?”
http://junksciencearchive.com/MSU_Temps/NCDCabs.html
NCDC Global Absolute Monthly Mean Temperatures from 1978 to present.
http://junksciencearchive.com/MSU_Temps/NCDCanom.html
Now the Global Monthly Mean Temperature Anomalies.
Ask yourself what graph would you publicise to promote the AGW cause?
@ur momisugly J.Fischer
I don’t know about GISS – but I do know what the Hadley software does. Each station has 12 monthly normals which are the average temperature for each month (eg. Jan, Feb, Mar…) between 1961 and 1990. Then…..
1.Anomalies are defined for each station by subtracting the monthly values for a particular month from these normal values. Stations without normals for 1961-1990 or where any anomaly is > 5 standard deviations are excluded.
2. The world is divided into a 5×5 degree grid of 2592 points. For each month the grid is populated by averaging the anomalies of any station present within each grid point. Most grid points are actually empty – especially for those early years. Furthermore the distribution of grid points with latitude is highly asymmetric with over 80 percent of all stations outside the tropics.
3. 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.
4. Finally the yearly global temperature anomalies are calculated by taking an area weighted average of all 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 midle of each grid point. All empty grid points are excluded from this average.
I’m afraid you’re totally wrong. Biases could indeed be introduced by coverage issues, but this has absolutely no bearing on whether phenomena are local or global. And one can estimate the size of those biases, and this has of course been done. Why do you think GISS temperatures only start at 1880, even though there are temperature records which go back two centuries before that? Have you, in fact, read any of the papers describing the GISS and other ground-based temperature datasets?
I think that starting at 1880 is better than starting at 1850. But already in 1880 only 8% of all grid points contain at least one station. I think the normalisation study shows that the data are consistent and free from bias only after 1900 – Look at Figure 5. After 1920 the data are a reliable reflection of the station data.
For the local/global issue. There are essentially zero stations in Saudi Arabia and Yemen. How would GISS detect if temperatures were to fall there by 2 degrees ?
This got me thinking…curious if there is a sleight of hand with the relationship of Celsius to Fahrenheit, considering there is a gap without decimal measures. The gap in typical habitable temp zones is about 1.8F from Celsius integers, correct?
Perhaps is nothing, perhaps small, but I’m not just skeptical, I’m a pessimist as I just do not trust people. If truth cannot stand up to scrutiny, then by default, it is not true.
I know Clive is snarc-ing, but this is almost certainly what the warmists do say.
But in jumping from temperatures to CO2 – and equating the two – they are jumping ahead to the conclusion/asserted correlation and using the conclusion to prove the conclusion. They don’t see the contradiction or the flawed logic, but it is as plain as the nose on your face. They would get an ‘F’ in Logic 101 with that kind of reasoning. Since they believe that CO2 concentrations equal increased temperatures, they have no trouble saying this. But that is exactly why there is a skeptical community in the first place – because their conclusion came before the science, so it drives their science. But they simply can’t do that and still maintain it is scientific reasoning.
Joe Bastardi
Thanks Joe.
I along with many others have been saying similar things here for quite some time. If the temperature in low latitude areas were to change by just -0.1°C, it would be possible for many degrees of change in say the high Arctic. This lets them say “Look at x in the Arctic circle, it’s boiling!”.
Without taking humidity into account, temperature is meaningless in terms of energy!
DaveE.
@More Soylent Green! says:
February 8, 2012 at 7:28 am
“I’m sure I’m not the first to post this thought, but using an anomaly as a measuring stick allows for an arbitrary baseline to be used and therefore the warming may be exaggerated. Exhibit A: The lowering of historical temperature data makes the late-20th century warming anomaly appear greater.”
Exactly.
The current WMO base line is 1961-1990, and since a few years back, like around 2007 onwards and more and more my country and EU countries have adopted to compare national averages to it.
But if one compares to the previous WMO base line, not much, if any, warming.
If one compares to the national compiled base lines, not much, if any, warming.
If compared to satellite era, not much if any warming.
The same applies for the supposed 4 inch sea level rise in the baltic sea over the last 30 years. At all the diving spots I’ve been too, the land rise seem to have won though.
Some EU countries, like Bulgaria, has it’s climate statistics compiled by the Hong Kong Observatory, which spells WMO, and apparently, snow is something rare in Bulgaria, just ask the News stations of the northern EU countries who’s amazed at the recent years “extreme weather” in Bulgaria.
Is it the weather or climate that is extreme and exaggerated or is the WMO data and the abuse of statistics, I wonder. For instance how do you check how much of the data is missing if you’re not allowed to see the original unmolested raw data for Swedish data? Apparently Swedes can access everybody else’s data but nobody are allowed to access their data, not even the Swedes themselves. :p
Wow, I’m astonished as to the number of responses my comment has received. A few clarifications:
1) I was explaining why anomalies are used instead of absolute temperatures. I explained why using anomalies is preferable. I never meant that it is perfect. when I say that you can get a very good aproximation to the average anomaly… I mean in comparison with trying to do the same calculation of the average temperature of the planet using temperature values from thermometers. I would not believe any error bars of less than 5 full degrees with respect to the second, with the current available thermometers. Trying to estimate the average temperature of a 500x500km area from the reading of 1 thermometer? Cmon. I may believe, however, that IF that single thermometer shows an average temperature increase of half a degree in the previous 50 years, and IF its surroundings have not changed significantly, then it is very likely that the temperature in all that 500x500km area has also raised, somewhat close to that half of a degree as well. I can believe that the error bars in calculating anomalies for a big area in that way are very probably less than half a degree. Which still is a quite poor performance given the small change that we are trying to measure, but it is about an order of magnitude better than the other alternative.
2) That anomalies do not change much in distances of even a few hundreths of a kilometer is an underlying assumption when calculating global temperature anomalies that I have not yet seen discredited. Please show me two good temperature records from two well placed thermometers not too far from each other in which the resulting anomalies are very different from each other’s.
3) Even if it results, not only that it is a bad assumption, but that it is a completely wrong assumption, to use absolute temperature instead of temperature anomalies, you would need to demonstrate that you get a BETTER result, a better representation of reality. And there’s no way you can demonstrate that.
All of this is common sense. I live in Madrid. I know that the temperature in Retiro Park is 1 or 2 degrees colder than the temperature in the surrounding streets, always. If I have a thermometer in Retiro and another one 1km into the city, I can use NONE of them to represent what is the temperature of the city. It would be wrong in both cases. However, they are both very likely to agree that today’s temperature is X degrees hotter or cooler than normal in the city. So I can look at ANY of them for that information. And not only that. If the Retiro thermometer says that the city is 5 degrees colder than normal, it is very likely colder than normal as well in all cities around Madrid, even in distances of hundreths of kilometers. The farthest, the less likely to agree, of course. But absolute temperature? It could differ in more than 20 degrees due to the different local conditions. So when you have a problem of sampling, and it is not possible not to have it in this big planet, the only reasonalbe thing to use is anomalies. In big areas the assumption may be wrong, but there is no reason to believe that it will be wrong in a particular direction (warm or cold). So you can put bigger error bars if you want, but not claim bias of any kind.
No particular reason.
http://i22.photobucket.com/albums/b331/kevster1346/wolframalpha-20120118164548020.jpg
Nylo says:
February 8, 2012 at 2:01 pm
” So when you have a problem of sampling, and it is not possible not to have it in this big planet, the only reasonalbe thing to use is anomalies. In big areas the assumption may be wrong, but there is no reason to believe that it will be wrong in a particular direction (warm or cold). So you can put bigger error bars if you want, but not claim bias of any kind.”
You are correct that resort to anomalies per se doesn’t introduce bias. However, the common assumption that anomalies at stations within the same region are coherent enough to be interchangeable certainly can. Look at the anomaly discrepancies betweeen Retiro Park and Barajas airport (14km away), as well as those at Valladolid (166km), Zaragosa Aero (266km) and Badajoz (316km). You will find not only appreciable differences in the year-by-year anomalies, but significant differences in multi-decadal “trends.” Bias arises from the UHI-affected station-shuffle that is endemic throughout “climate science” in synthesizing “average anomaly” series over time scales of a century or longer.
Only intact century-long records at relatively UHI-free stations can provide unbiased results. Unfortunately, such records are very much in short supply throughout much of the globe, leaving many grid-boxes totally devoid of unbiased data series. Neither blind anomaly averaging nor the the statistical splicing of fragmented actual temperature series done by BEST overcomes that fundamental deficit of high-quality data.