Guest post by David R.B. Stockwell PhD
I read with interest GHCN’s Dodgy Adjustments In Iceland by Paul Homewood on distortions of the mean temperature plots for Stykkisholmur, a small town in the west of Iceland by GHCN homogenization adjustments.
The validity of the homogenization process is also being challenged in a talk I am giving shortly in Sydney, at the annual conference of the Australian Environment Foundation on the 30th of October 2012, based on a manuscript uploaded to the viXra archive, called “Is Temperature or the Temperature Record Rising?”
The proposition is that commonly used homogenization techniques are circular — a logical fallacy in which “the reasoner begins with what he or she is trying to end up with.” Results derived from a circularity are essentially just restatements of the assumptions. Because the assumption is not tested, the conclusion (in this case the global temperature record) is not supported.
I present a number of arguments to support this view.
First, a little proof. If S is the target temperature series, and R is the regional climatology, then most algorithms that detect abrupt shifts in the mean level of temperature readings, also known as inhomogeneities, come down to testing for changes in the difference between R and S, i.e. D=S-R. The homogenization of S, or H(S), is the adjustment of S by the magnitude of the change in the difference series D.
When this homogenization process is written out as an equation, it is clear that homogenization of S is simply the replacement of S with the regional climatology R.
H(S) = S-D = S-(S-R) = R
While homogenization algorithms do not apply D to S exactly, they do apply the shifts in baseline to S, and so coerce the trend in S to the trend in the regional climatology.
The coercion to the regional trend is strongest in series that differ most from the regional trend, and happens irrespective of any contrary evidence. That is why “the reasoner ends up with what they began with”.
Second, I show bad adjustments like Stykkisholmur, from the Riverina region of Australia. This area has good, long temperature records, and has also been heavily irrigated, and so might be expected to show less warming than other areas. With a nonhomogenized method called AWAP, a surface fit of temperature trend last century shows cooling in the Riverina (circle on map 1. below). A surface fit with the recently-developed, homogenized, ACORN temperature network (2.) shows warming in the same region!
Below are the raw minimum temperature records for four towns in the Riverina (in blue). The temperatures are largely constant or falling over the last century, as are their neighbors (in gray). The red line tracks the adjustments in the homogenized dataset, some over a degree, that have coerced the cooling trend in these towns to warming.
It is not doubted that raw data contains errors. But independent estimates of the false alarm rate (FARs) using simulated data show regional homogenization methods can exceed 50%, an unacceptable high rate that far exceeds the generally accepted 5% or 1% errors rates typically accepted in scientific methods. Homogenization techniques are adding more errors than they remove.
The problem of latent circularity is a theme I developed on the hockey-stick, in Reconstruction of past climate using series with red noise. The flaw common to the hockey-stick and homogenization is “data peeking” which produces high rates of false positives, thus generating the desired result with implausibly high levels of significance.
Data peeking allows one to delete the data you need to achieve significance, use random noise proxies to produce a hockey-stick shape, or in the case of homogenization, adjust a deviant target series into the overall trend.
To avoid the pitfall of circularity, I would think the determination of adjustments would need to be completely independent of the larger trends, which would rule out most commonly used homogenization methods. The adjustments would also need to be far fewer, and individually significant, as errors no larger than noise cannot be detected reliably.
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
I have yet to see any justification for ‘homogenization’.
A working and supposedly well sited autonated observation system reports the local temperature as XdegC and 30 miles away to the East another says YdegC and to the West another says YdegC. This does not mean that the local XdegC tremperature is incorrect. Similarly, the fact that I should not really get frost in early summer and the system reports frost – does not mean that it is incorrect. This type of homogenization algorithm should only be used as a flag for an observation to be checked. But the use of undergraduate style software in a batch job just smoothing the data towards each other and the expected climatology worldwide must always be incorrect. If a site is different to the others the reason needs to be checked rather than the value massaged by a parameter to something the programmer thinks is more likely.
This comes back down to a total lack of quality management by the operators of these climate systems NASA/NOAA/Met Office and UEA and the other international bodies. The poor quality management is due to their laziness. Each reporting site needs to be individually assessed and any alteration logged and signed off and countersigned as valid with the reason for that alteration. There are not that many observation sites. Banks don’t give up and homogenize client accounts because there are a lot of them.
Paul Homewood: “Then add in the smaller urban sites (even comparatively small towns will have UHI effect, particularly a growing town). And you get a significant number.”
I spent some time working my way down the list with Google Earth. I had a couple pages of stations examined before I found my very first station that was more then a mile from the nearest US Post Office.
The ratio “Percent of the US landmass that’s farther than one mile from a Post Office” to “Percent of the USHCN/CRN/etc. weatherstations farther than one mile from a Post Office” would be … outrageous.
The urban/rural determinations have issues.
Victor Venema:
In your post at October 15, 2012 at 10:57 am which is apparently aimed at – but not addressed to – me you say
Unfortunately, I was beaten to it by Paul Homewood at October 15, 2012 at 10:58 am where he says
However, of itself that does not fit your criterion of “most of the data”. Hence, I add to that by pointing out the effect would be spread to at least half of the rest of the data by homogenisation. Thus, my addition fulfills your “most” criterion.
Can I now have your assurance that you will nominate Richard Muller and me for whichever Nobel Prize you consider to be appropriate? And, importantly, will you please tell us how to get the “millions from the Koch brothers”.
Richard
PS I have not been a “hero” before and I can’t say I notice feeling any difference now you say I am one.
REPLY: Since there are people who are trying to actively discredit it (such as yourself), I decided not to update the work page regularly until we had our full revision completed. … – Anthony
I thought you had requested readers of WUWT to review your study. My suggestion on your work page was constructive and would make the study stronger. Sorry, if I misunderstood your request.
If you have a solid study, it can not be discredited by arguments.
REPLY: LOL! The fact that the Climate Reference Network exists at all is proof of the fact that NCDC takes the issue of UHI and siting seriously. It has four years of complete data (since 2008) and you call it “the same” as the old network, yet in your other argument you claim I don’t have enough years of metadata to establish siting trends. You can’t have it both ways. Make up your mind, because your bias is laughable. – Anthony
The USCRN started in 2004, not 2008. “The data for the period 2004–2008 are extremely well aligned with those derived from the USHCN version 2 temperature data. For these five years, the r2 between the 60 monthly USCRN and USHCN version 2 anomalies is 0.998 and 0.996.” (Menne et al. On the reliability of the U.S. surface temperature record. J Geophys. Res., VOL. 115, D11108, doi:10.1029/2009JD013094, 2010.)
I never claimed it was a proof, but at least the periods of the two dataset are the same. It does make it less likely that there are problems. Together with all the other evidence, I do not see it as productive to study urbanization. You have to pick your fights.
You should be happy, that scientists always try to make their work more accurate. If your blog produced the pressure to get funding for such improvements: thank you very very much. And please keep up the good work. Without your efforts climate science would be seen as a solved problem in Europe and we would mainly be doing applied climate impact studies. Due to your industrious efforts, there is more money to study fundamental questions, which is what scientists like the most.
Paul Homewood says: “Urban areas are heavily overrepresented in the siting of temperature stations: less than 1% of the globe is urban but 27% of the Global Historical Climatology Network Monthly (GHCN-M) stations are located in cities with a population greater than 50,000. Then add in the smaller urban sites (even comparatively small towns will have UHI effect, particularly a growing town). And you get a significant number.”
Thank you for giving a number: 27% of the stations. The period of urbanization is typically in the order of 30 years. After this there is no longer a bias in the trend, the temperature just has a fixed constant bias, which does not affect trend estimates. The the amount of data affected will be much less than 27%.
50 thousand inhabitants is not what I would call an urban region. If you would have a proof for a growing urban heat island effect in such cities, you would have a case that urbanization is a problem. Especially if you can proof this for even smaller town you have a good case. Then you could collect your price with the Koch brothers and live in luxury. You would get a guest post on WUWT and get more than your 15 minutes of fame.
But please do not start about noticing a perceived temperature drop when coming out of a small town. Perceived temperature is influence by radiation, wind and humidity, not just temperature.
Luther Wu says: “VV said: “ Life is too short to follow every piece of misinformation spread on WUWT. Cite one example, please- just one…”
1) The Watts et al (2012) manuscript falsely claiming half of the warming to be due climatologists.
2) The post again falsely claiming half of the warming to be due climatologists on the “peer reviewed paper” by two Greek hydrologists.
http://variable-variability.blogspot.com/2012/07/investigation-of-methods-for.html
3) Calling Fritz Vahrenholt a former environmental minister without mentioning that he is currently manager of a large utility.
DirkH says: “(This is not an attack on scientists, as the IPCC AR4 has mostly not been written by scientists.)”
The report is written by scientists. Only the summary for policy makers is written together with government officials. If you read the report itself, you should be getting good information and you can always go back to the citations.
@outtheback. NOAA only homogenizes the annual means. In most other countries and studies also the seasonal and monthly means are homogenized. This can help you to find more breaks, in case the sign changes from season to season. And you need this if you would like to study trend changes as a function of season to understand the physical reasons for the changes better. To compute the global mean temperature, which is a summary statistic that attracts most public attention, the annual means are sufficient.
If the changes in the magnitude of the urban heat island (UHI) go in clear jumps (corresponding to large developments), it is relatively easy to correct for them using homogenization. Then you could still see the jumps even if the reference stations contain some jumps as well. The difficult situation is when there is a gradual increase in the UHI and most of the reference stations have a simultaneous similar gradual increase during the same period. Then relative homogenization would not notice this data problem and keep the artificial heating.
Urban readings in itself are not a problem. The problem is urbanization. For stations in the center of major the trend is about the same as for the rural stations. In this case the UHI just causes a constant bias. Studies comparing the trend for rural and urban stations (defined in many different ways) show the same trend for both types of stations.
Ian W says: “I have yet to see any justification for ‘homogenization’. A working and supposedly well sited autonated observation system reports the local temperature as XdegC and 30 miles away to the East another says YdegC and to the West another says YdegC. …”
It is always better not to need homogenization, but that is difficult. For one, this would only help you in future (and is now tried in the US reference network). You would still need homogenization to study the climate of the past. And you will have a hard time to keep your measurement and the surrounding of the station constant over a period of centuries. Climatologists are not the infinitely powerful elites, they are portrayed to be at “sceptics” blogs.
richardscourtney, maybe work on the proof a little more, until it is water tight. 🙂
@Victor Venema: Thank you for your response, but I think each of your rebuttals are incorrect.
You said “The equation: H(S) = S-D = S-(S-R) = R is wrong. You do use the difference time series (D) to determine the size of the jump, but you do not replace all values by the ones in the regional climate signal.” I said “While homogenization algorithms do not apply D to S exactly, they do apply the shifts in baseline to S, and so coerce the trend in S to the trend in the regional climatology.” You have simply restated exactly the same thing that I said with a couple of extra equations.
You said: “Homogenisation is used to be able study large scale climate variability in a more accurate way. Removing the too low trend for an irrigated region, is what homogenisation is supposed to do.” Imagine a perfectly correct temperature record in the irrigated region that shows a falling temperature. By your standards a perfectly correct record should be adjusted to show a rising temperature! That is rewriting history, revisionism, and just plain wrong IMHO.
You say: “why don’t you submit an abstract at the General Assembly of the European Geophysical Union? There you would get more qualified feedback on the quality of your work.” I have published a lot of papers in peer-reviewed journals, but when it comes to critiques of climate science, one of two things happens, the editors cannot find anyone to review it, or it gets through and is never cited. You find better qualified feedback on the technical blogs.
@Victor Venema: Thank you for your response, but I think each of your rebuttals are incorrect.
You said “The equation: H(S) = S-D = S-(S-R) = R is wrong. You do use the difference time series (D) to determine the size of the jump, but you do not replace all values by the ones in the regional climate signal.” I said “While homogenization algorithms do not apply D to S exactly, they do apply the shifts in baseline to S, and so coerce the trend in S to the trend in the regional climatology.” You have simply restated exactly the same thing that I said with a couple of extra equations.
You said: “Homogenisation is used to be able study large scale climate variability in a more accurate way. Removing the too low trend for an irrigated region, is what homogenisation is supposed to do.” Imagine a perfectly correct temperature record in the irrigated region that shows a falling temperature. By your standards a perfectly correct record should be adjusted to show a rising temperature! That is rewriting history, revisionism, and just plain wrong IMHO.
You say: “why don’t you submit an abstract at the General Assembly of the European Geophysical Union? There you would get more qualified feedback on the quality of your work.” I have published a lot of papers in peer-reviewed journals, but when it comes to critiques of climate science, one of two things happens, the editors cannot find anyone to review it, or it gets through and is never cited. You can find better qualified technical feedback on the technical blogs.
(David Stockwell)
@Leopold Danze Morgan: You said: “Much as I’d like to be able to conclude that you have disclosed a fundamental error in calculating temperatures, and therefore there is no need to worry about ‘thermageddon’, all I can realistically conclude is that you have not communicated clearly.”
The problem of circular reasoning is not that the conclusion is wrong. It could well be right if the assumptions are right. The problem is that it is an incorrect inference, dressed up as a tested result. So it is quite easy to get errors with high significance values, i.e. to fool yourself.
A number of studies now have shown that the contribution of homogenization to the overall warming trends last century is on the order of 0.3 to 0.5 degrees. This would be grounds for concern, I think.
Thanks for writing suggestions.
@Victor “The period of urbanization is typically in the order of 30 years. After this there is no longer a bias in the trend, the temperature just has a fixed constant bias, which does not affect trend estimates.”
And this is EXACTLY why temperatures have leveled off for the last 15 or so years. Thanks you !!!
Once “adjusted”, it is no longer data.
data – any fact assumed to be a matter of direct observation
Here’s a little exercise in basic maths.
Suppose in a hypothetical region of size 20000 sq miles there are 3 urban areas of size 250, 500, and 250 sq miles. In this region there are 5 weather stations , one in each of the urban areas and 2 rural stations.
Now, over the past 50 years these stations have seen the following trends:
Urban1 = 2.1oC, Urban 2 = 2.0oC, Urban3 = 1.5oC, Rural1 = 0.1oC, Rural2 = -0.3oC
If we want to calculate the Average temperature change, and we just apply equal areas to each station, and we get an average rise of …………………………..
If, however, we apply the urban stations ONLY to their respective urban areas, and split the rural area equally, we get an average temperature rise of ……………………………….
——————————————————————–
Now of course, if you homogenise the rural data first, so it matches the trend of Urban areas, you have truly stuffed up the whole thing and now have a massive trend where, in reality, none exists.
darn degree signs didn’t work, sorry.
And Victor.. if you come here mentioning Skeptical Science.. expect to get laughed at.
Victor Venema:
You admit to having visited SkS and it seems that you are used to posting on similar warmist ‘echo chamber’ web sites. So, I write offer you some genuine and sincere advice.
My “proof” is much, much more “watertight” than your armwaving about “no more than a few percent of the data are affected by urbanization”. Clearly, you are not aware of the nature of WUWT but you mistakenly think WUWT is similar to warmist blogs although WUWT has a different ’cause’ from them.
WUWT differs from the warmist sites in that WUWT is not censored so all viewpoints are available for scrutiny here, and people of sincerity can obtain much credibility whatever their view. For example, on WUWT Leif Svalgaard is clearly the most respected solar expert who frequents WUWT and he champions the IPCC view of solar (non)influence on climate. Hence, if you want to gain credibilty here then you need to provide clear and logical arguments which are supported by referenced information which can be challenged. Armwaving assertions don’t ‘cut it’: such assertions are usually flatly rejected or are ridiculed (as I ridiculed your silly Nobel Prize assertion).
You can be a respected champion of your views on WUWT if you give clear, logical arguments supported by referenced information. Alternatively, if you make the kinds of posts you have made so far then you will – rightly – be treated in the same manner of condescending tolerance as the existing WUWT resident trolls (i.e. John Brookes, LazyTeanager, Izen, etc.).
Please note that I provide this advice with complete and genuine sincerity.
Richard
REPLY: I echo Richard’s sentiment. SkS has proven themselves to be nothing more than a club of angry petulant children hell bent on smearing anyone, no matter what their standing in science, if they disagree with the premises they hold dear. By their own actions, they have proven themselves to be liars, conspirators, bullies, and post facto revisionists. Citing them is akin to citing Al Gore. -Anthony
laterite: “Imagine a perfectly correct temperature record in the irrigated region that shows a falling temperature. By your standards a perfectly correct record should be adjusted to show a rising temperature! That is rewriting history, revisionism, and just plain wrong IMHO.”
Just as with the urban heat island effect, it depends on what you are interested in. Just as for cities, there will likely be more stations in irrigated areas as in surrounding.
Thus if you are interested in the large scale regional climate this is a local artefact, which you would like to remove. Just as in the case of the urban heat island. If you have sufficient stations (I just read that you had 4 stations in the irrigated region) and the network density is similar outside the irrigated area there is likely no problem. In this case the stations are representative of their surrounding and can be used to compute the large scale climate signal.
I understand your side, if you are interested in biodiversity, agriculture (or city climates), you would like to keep this part of the signal. In that case you should only compare the stations in the irrigated area with each other. (And by the way, comparing stations with the Australian mean temperature as reference is completely wrong, it is no wonder you produce false positives that way. The reference should be the best estimate of the regional climate at the station you are testing.)
I also like working on a variety of topics and thus often work on topics on which I am not an expert. Then I always make sure, that I collaborate with an expert to avoid making errors. In case of your extended abstract, an expert may have convinced you not to cite Steirou and Koutsoyiannis (2012). Especially not as a study on the global temperature, as most stations were from the USA, the station density outside the US was insufficient, and no effort was made to account for differences in station density.
It might not be fair, but using as temperature unit “C” and not “°C” or calling SNHT Standard Normal Homogenization Test (and not rightly Homogeneity Test) makes an unprofessional impression, which makes reviewers more critical. If you use an equation, it should be right, maybe you can be a bit lax in a blog post, but at least in a scientific work it should be right. The Standard Normal Homogeneity Test is based on hypothesis testing. There are homogenization methods using information theoretical measures, but SNHT is not one of them.
If an editor has problems finding a reviewer for a paper of yours on homogenization, feel free to mention my name. To get cited, people need to know that your study exists. Another good reason to visit scientific climate conferences. I am not sure, whether the conference for which you wrote this extended abstract qualifies as scientific, with invited speakers talking on “Demonising carbon dioxide: Science of the absurd”.
AndyG55 says: “And Victor.. if you come here mentioning Skeptical Science.. expect to get laughed at.”
🙂 No problem. To be honest, I do not expect to be able to convince you, Anthony and most of the commenters here. I just hope that some of the readers of WUWT will start questioning the content of this blog.
By their own actions, they [Skeptical Science] have proven themselves to be liars, conspirators, bullies, and post facto revisionists. … -Anthony
I am sure, they feel the same way.
As a scientist, I can only note, that they are better at scientific argumentation and know the scientific literature much better.
Victor Venema:
At October 15, 2012 at 2:58 pm you say of SkS
Well, that has blown any credibility you may have had as a scientist.
Richard
Have you done my little maths exercise yet, V ?
They “feel” the same way? No, they cannot. Their approach is fundamentally different.
SkS DOES revise and rewrite entire threads. They also completely remove comment after comment after comment when the comment is in any way inconvenient to their story.
WUWT DOES NOT revise and rewrite entire threads. And WUWT does not remove any comments for reasons of agreement, disagreement or inconvenience (over the top rudeness is not tolerated, however).
listening to: Guided By Voices– I Am A Scientist
______________
Am I perpetrating a logical fallacy?
Maybe I should cite the Dandy Warhols?
Tell you what, try this in court to defend a speeding fine: I have homogenised my speed with that of the surrounding traffic and this shows that I was not actually speeding.
Paul Homewood says:
October 15, 2012 at 11:09 am
Paul, who are you going to believe, data or a “scientist” who’s been told stuff by his “scientist” friends?
For the record, temperature trends since 1979:
GISS: 0.20/decade
RSS: 0.13/decade
UAH: 0.14/decade
http://www.woodfortrees.org/plot/uah/plot/uah/trend/plot/gistemp-dts/from:1979/plot/gistemp-dts/from:1979/trend/plot/rss/plot/rss/trend
Of course, given uncertainties around measuring temperature anomalies, maybe our “scientist” friend thinks (or has been told by his friends) that those are all the same within the margin of error.
Victor Venema ,
I usually don`t make personal comments here , just the odd random observations from time to time , but on this occaision I shall make such a comment ,
proselytism and dogma doesn`t cut it , raw data and falsifiable hypotheses do (at least untill they`re falsified )
– – – – – – – –
Victor Venema,
Your claim of superior science argument at Cook’s blog can be taken in several ways.
One way your claim can be taken is it makes false presumptions about what constitutes normal modes of scientific argumentation. Your claim is false if you mean that proper scientific argument consists of the following activities at Cook’s blog: broadly applied censoring skeptical comments / unauthorized revisionism of comments / main post alteration without documenting what the revisions were and sometimes when made / appeals to the IPCC as the sole authorities in climate science / uncivil and unequal moderation of comments / myopic causation bias.
Another way your claim can be taken is as just being innocently naive; as possibly being made by a person who is unaccustomed to successful and productive participation in the dynamics of an open and independent climate science blog (WUWT) which has very light handed moderation. If it is innocent naivety that explains your claim, then welcome to the real world of climate science in the 21st century.
Yet another way your claim can be taken is as just a knee jerk emotional repartee; as being of the nature of a child saying, “my mother is better than your mother, because my mother doesn’t wear army boots”, or something like that. : )
But I think the most advantageous way your claim can be taken is that it is a challenge to a debate of WUWT’s best denizens versus Cook’s site best denizens. Is that the essence of your claim? I hope so.
John