Big thanks to regular Talkshop fly-by visitor ‘Vuk’ who pointed me to this page on his site, which contains a graphic which, well, graphically shows where a lot of the Northern Hemisphere land surface warm-up came from in the latter 30 years of the C20th.
click image for fullsize
So, permanent Russian heat-wave, or NASA-GISS data cock-up? It’s interesting to note the ‘heat-wave’ doesn’t extend into Kamchatka, and that Alaska is colder than average by around the same amount as Russia is hotter over the 69-71 period. Or did a big warm-up in the arctic seas off Siberia missed by GISS skew the Russian data?
Hopefully Vuk will call by to tell us more.
[update: image updated 8th April 2012 –Tim]
Tallbloke's Talkshop 
TB there is an update for the graphic.
comment:
It could be a perfectly normal event.
On the other hand, it could be that the Brezhnev propaganda, fearing the US & Canada grain imports embargo in the wake of the invasion of Czechoslovakia, by reporting higher temperatures would be indirectly announcing to the West ‘we are doing fine’.
How to explain the temperature anomaly only within USSR borders was up to +4C (?!), while for the rest of the globe (except for few very minor exceptions) the anomaly was zero or negative?
Temperatures do not respect borders, unless decreed to so by the Politburo.
Speculative , but it was a highly unusual event.
Subject to the data being reliable I see it as a natural consequence of a faster and more zonal air flow from Atlantic across the Eurasian interior.
Such a flow would mitigate midwinter cooling in such regions for a disproportionate temperature response compared to other regions.
Rog, the data only relates to a short period. To validate my suggestion we would need to see a similar pattern throughout the 30 year late 20th century warming period.
The article seems to suggest just that though doesn’t it ?
TB,
Without the wobble or tilting of this planet, the temperature line would be constant and predictable to just the same differences day and night at each latitude of planetary rotation.
[Reply] Ya think? 🙂
Hi Steven
As in the 1960s space race the Soviets led the way and the world followed in the
gloabal warming race
Yes, as you can see from the link Russia has been permanently ‘hot’ place. Temperature and weather reports are highly strategic information as Napoleon and later Hitler found to their cost.
Some caution is advisable.
Has the graphic changed Vuk? I can’t see any difference.
Stephen, there’s always the possibility that Hansen decided he liked the 69-71 jump and decided to keep it… rejigging the baseline so the strong anomaly wasn’t visible as an anomaly later on…
Anyway, the article, such as it is, is pretty non-comittal. Vuk’s graph speaks for itself. Just that different people will hear it saying different things.
GISS dts vs HADSST2gl has a less strong divergence after ~1977
http://woodfortrees.org/plot/gistemp-dts/from:1960/plot/hadsst2gl/from:1960
Tb
re graphics: bottom left
http://www.vukcevic.talktalk.net/69-71.htm
Temps 1880-2010 graph
Thanks Vuk. WordPress caches images unfortunately so the old verion will continue to appear for a while. I recommend people visit Vuk’s site and do a shift-reload if they don’t see the additional info.
Vuk, it’s worth changing the filename when you update images, so people don’t get previously cached copies.
Thanks, Vuk, for showing us the step reduction in SST ~1970 and the 4C ‘USSR hot spot’. None of this could have been caused by the slow increase in CO2, so in both cases it must been caused by politics or some other unnatural mechanism at work.
Do I trust this data? No I don’t.
Mean annual temperatures in SIberia commonly fluctuate by several degrees C over periods of a few years, and between 1969 and 1971 there was an increase of maybe a couple of degrees. Subtracting SSTs, which fell during the 1970 La Niña, magnifies this effect and gives us a big red blob on the map.
However, if we used the preceding 1967-1969 interval Siberia would appear as a big blue blob (and as a red blob between 1987 and 1989, a blue blob between 1986 and 1987 and so on and so forth.)
I don’t think the Russian temperature records are seriously skewed by politics or other unnatural influences. I’ve been through them in some detail and in my opinion they’re more reliable overall than the US records, although that isn’t saying much.
Thanks Roger. That doesn’t explain the ongoing divergence after 71 though. maybe Vuk will tell us more about the specific datasets he used. Can you replicate Vuk’s result Roger? My wood for trees plot is significantly different.
TB: I think we’re back to this again:
So we’re saying that the SAT – SST cycle came to Russia late? Or the ‘great climate shift’ of ’76 came to Russia early?
the chart shows 1969-1971. I imagine vuk looked through all the possible data slices to find this period. If you want to see how temp evolves spatially over time just look at the Berkeley animation.
on short time scales heat moves around. go figure. global warming is a boundary problem.
WRT SST and Land surface temp consider this. It’s 43 minutes. [snip]
http://www.newton.ac.uk/programmes/CLP/seminars/081814001.html
Here are some more
http://www.newton.ac.uk/programmes/CLP/seminars/082310001.html
[Reply] Thanks Mosh. I’m sure you felt as good typing it as I did snipping it. 😉
The great climate shift of 1976 appears in global average temperatures. It gets a lot more complicated when you look at regional air temperatures. In places like Alaska and much of South America the shift arrived on time, but in Siberia it seems to have arrived in the late 1980s, and in other areas such as Australia and Nova Scotia there’s no sign it ever arrived at all. I once tried to make sense of all this by plotting arrival times (or the lack thereof) on a map of the world, but if there was a trend there it escaped me. Maybe I’ll give it another shot one day.
This is for historians of the future when dealing with the chapter: “Climate Change”….
Mosh says: “on short time scales heat moves around. go figure.”
Yes indeed. Figuring is what we are trying to do. Since nobody seems to know what causes these kind of quasi biennial events, it’s no wonder modelers have problems with regional climate. You can’t parameterize what you don’t understand.
“global warming is a boundary problem.”
Well energy is more or less conserved in the system on an multi-annual basis, roughly speaking. So the question is about redistribution of energy at the regional level, and trying to figure what causes the longer term patterns like the Mycean, Roman, Medieval and Modern warm periods. All of which are around a thousand years apart, with cold spells in between. If that pattern holds, then we are around the top of the curve, so ‘global warming’ is just global warm, for which we should be grateful for our allotted time in history.
It is very simple, look second graph in
http://www.vukcevic.talktalk.net/GT-AMO.htm
no good reason or explanation available why the GT and the AMO would permanently separate from 1970 despite previous good agreement between two.
I am reasonably confident that the AMO is OK since 1950s onwards.
It just shouldn’t happen naturally
The datasets have been diddled around with so much it’s much more likely an artifact of someones ‘normalization’ procedure than anything natural.
There must be some reason why the newspapers and some scientists started muttering about ice age in the 70’s though…
What I recall of the Russian/Siberian temps:
At the end of the USSR, some meteorologists/climatologists said that, during the Central Committee control on fuel supplies, it was standard to downsize the Russian temperatures, so as to get more money/fuel. Once the central control came off, the values were what they were – warmer. The complaint was raised regarding GISTemp values being used and ‘warming’ the records and supporting AGW.
The question is: are the earlier values of the USSR reliable, or at they contaminated by consistent COLD bias to generate more money/fuel for the region? The result would be a significant (in terms of 1C temperature rises!) warming of that part of the world.
There must be some reason why the newspapers and some scientists started muttering about ice age in the 70′s though…
Agree, and I remember it well. 1950-1970 drop in the global temperatures wouldn’t justify it since there were similar changes in the beginning of the century, but if you assume that the GTs (1950-1970) were tracking the AMO as in previous decades then the temperature reduction would be reason for a concern.
True or False
H. H. Lamb was the greatest climatologist of his time. He, almost single-handed, alerted the world to the inconstancy of present climate…………….
An irony is that, now the world is acutely aware of global climate change, Lamb had maintained a guarded attitude to the importance of greenhouse gas warming. Although many others have accepted this, he felt that there was too much reluctance to consider the full range of other, natural, causes of change.
Right to the end of his life, he was promoting his “different view”.
http://www.independent.co.uk/news/people/obituary-professor-h-h-lamb-1249739.html
In 1973 and 1975 he arranged for two international conferences which were hosted in Norwich. At first his view was that global cooling would lead within 10,000 years to a future ice age,
but over a period including the UK’s exceptional drought and heat wave of 1975–76 he changed to predicting that global warming could have serious effects within a century. His warnings of damage to agriculture, ice caps melting, and cities being flooded caught widespread attention and helped to shape public opinion.
http://en.wikipedia.org/wiki/Hubert_Lamb
Doug proctor
The timing isn’t right. The upward shift in the Siberian records occurred in 1987-88 but the breakup of the Soviet Union didn’t occur until the end of 1991. There was also an upward shift in the Chinese records in 1987-88. So either the Chinese and Russian Central Committees conspired to distort the Asian temperature records or the shift was natural.
Vuk:
I can’t replicate the shift shown in your graph. Exactly which data sets did you use?
RA;
1988 was “glasnost” (openness) time. The breakup took a while longer.
The match of the ’70-’71 hot Siberia boundaries to the political boundaries is exact. The Null Hypothesis has to be that the temperatures are also political.
Roger Andrews says:
April 4, 2012 at 12:10 am
Doug proctor
The timing isn’t right. The upward shift in the Siberian records occurred in 1987-88 but the breakup of the Soviet Union didn’t occur until the end of 1991.
>>>>>>>>>>
I believe fuel subsidies based on temperature ended in the USSR prior to breakup. Further, I don’t think you could correlate to the ending of fuel subsidies alone. For how long did the “policy” of subtracting a few degrees from the actual reading stay in place for any given weather station? I can gaurantee that there were plenty of weather stations where someone asked “why do we subtract 5 degrees from the temperature readings?” and the answer was “I don’t know, that’s the way we’ve always done it”.
You would have to go to each and every weather station, and try and find someone who worked there, remembers what adjustments were done, when they started, and when they ended. And would be willing to admit to it.
There was also an upward shift in the Chinese records in 1987-88>>>
Same thermometer supplier?
Abstract of paper on Russian soil temperature
Data is available but as usual the provenance is poor. I write that whilst sitting on dire findings about a soil temperature dataset which has only come to light because of unusual access. I’ve partially undone the “adjustments”, then done it properly, but some site problems leave the whole thing as too poor.
Brian H:
“The match of the ’70-’71 hot Siberia boundaries to the political boundaries is exact. The Null Hypothesis has to be that the temperatures are also political.”
The match of the ’68-’69 cold Siberia boundaries to the political boundaries is exact too. Evidently SIberia had political cold temperatures as well as political hot temperatures.
http://data.giss.nasa.gov/work/gistemp/NMAPS/tmp_GHCN_GISS_250km_Trnd0112_1968_1969/GHCN_GISS_250km_Trnd0112_1968_1969.gif.
And the hot and cold political temperatures pretty much canceled each other out between 1959 and 1979. A foretaste of glasnost, perhaps?.
And total warming in Siberia since 1945 isn’t out of line with Canada, so if party functionaries did cook the climate books they didn’t do a very good job of it.
davidmhoffer:
“I believe fuel subsidies based on temperature ended in the USSR prior to breakup. Further, I don’t think you could correlate to the ending of fuel subsidies alone. For how long did the “policy” of subtracting a few degrees from the actual reading stay in place for any given weather station? I can gaurantee that there were plenty of weather stations where someone asked “why do we subtract 5 degrees from the temperature readings?” and the answer was “I don’t know, that’s the way we’ve always done it”.”
People have been making claims like this about the Siberian records for years but so far as I know no one has yet come up with any proof that they’re correct. Do you have any?
[RA, those image links are all bad, can’t fix them, probably they are dynamic content so they would need copying to a static location –Tim]
Roger Andrews says:
April 4, 2012 at 12:10 am
Re: the timing isn’t right. Man, my typing is faster than my thinking at times.
Yer right. I need to keep my day job.
RA;
People have been making claims like this about the Siberian records for years but so far as I know no one has yet come up with any proof that they’re correct. Do you have any?>>>
I didn’t make the claim. I pointed out that if such a thing happened, that it would not end at some specific point in time such as the year of collapse of the Soviet Union, and that in fact it would be an effect, that if it existed, would more likely wither away over a period of time, not end all at once. I further pointed out that to know with any certainty, one would have to find people who worked in thos weather stations over that course of time and find out from them what they did (or didn’t) do and that they would be unlikely to admit to it in any event.
All that said, the Soviet Union did in fact have fuel allocations based on “need” defined by the harshness of the winter. No, I’m not going to take the time to “prove” it to you. But how else would a centrally run economic system determine how much fuel to send where? From there it doesn’t take a leap of logic to figure out that at least some of the communities would have tried to game the system. In fact, it would be completely illogical to suggest that it didn’t happen at all.
Roger A: “I can’t replicate the shift shown in your graph. Exactly which data sets did you use?”
One is the AMO
I’m not sure about the GISS datasets, but it looks consistent with their global stuff.
So, With the r^2 values he offers in his comment at 8.24pm, I think Vuk is implying GISS seem to have pegged their dataset to the AMO but .4C higher. Is that right VUK?
It does seem to be the case that the 1970’s cooling period has been diminished in recent years as the GISS dataset has undergone continual adjustment. As I pointed out earlier, if the GISS dataset with it’s almost constantly upward trajectory was the truth, why were scientists and newspapers muttering about a new ice age in the early seventies?
Every 50-70 years, the Arctic and much of the northern hemisphere’s land mass warms. It’s because cloud albedo falls as extra aerosols are generated in the Arctic summer by biofeedback.
It’s the same process that leads to the end of ice ages. I hope to publish soon.
Roger Andrews says:
April 4, 2012 at 12:10 am
……………
I used
For the AMO
http://www.esrl.noaa.gov/psd/data/correlation/amon.us.long.data
or you can use
http://www.esrl.noaa.gov/psd/data/correlation/amon.us.long.mean.data
and calculate your own anomaly
for the global temperatures
http://data.giss.nasa.gov/gistemp/
in all cases: add monthly numbers, divide by 12 to obtain the annual values.
Full results are shown in here:
http://www.vukcevic.talktalk.net/GT-AMO.htm
tallbloke says:
April 4, 2012 at 7:24 am
……..
See
http://www.vukcevic.talktalk.net/GT-AMO.htm
difference is 0.325C, 3rd graph (to the right)
davidmhoffer wrote to Roger Andrews: “In fact, it would be completely illogical to suggest that [adjusted temperatures within the Soviet Union to game centrally planned economy] didn’t happen at all.” in fact, it did.
A couple years after Soviet communism fell, I took a seminar in the subject at the University of Colorado at Boulder. The late Professor, Edward J. Rozek,
http://m.rockymountainnews.com/news/2009/feb/26/edward-j-rozek-escaped-slave-labor-camp-taught/
conducted the seminar – his lifelong area of expertise. (In truth, the man looked oddly too much like Leonid Brezhnev, leader of the Soviet Union from 1964 to1982. http://en.wikipedia.org/wiki/File:Brezhnev-color.jpg)
Part of the uneconomic distortions of central planning meant that aggregate production targets existed, unconcerned end use. For example, total iron output – regardless of quality needed for manufacturing – or total shoes, regardless of sizes needed. So the deceptiveness of record keeping within the old Soviet Union was a fact that was touched upon several times.
In fact, neglect to record temperatures in Northern Canada by our military during the Cold War has also been admitted, as Anthony Watts reported a few years ago. States one first-hand record keeper: “For numerous reasons many [weather] reports were fabricated.”
Anthony continues: “For example, look at this B91 form provided by the Marysville California observer (PDF format). Note all the missing days. Thanks to NCDC’s FILNET program, those missing days get made up into a complete data set much like the data on the DEW line did. With a “best guess” programmed into a data sorting and analysis program.”
From the comments of that thread, writes an old US Army statistics keep (for intelligence):
“As an aside. In the cold war period the Soviets were constantly practicing various forms of what they call ‘maskirovka,’ which is denial and deception to hide their true intentions/capabilities. Well one area they used to do that in was to deliberately falsify their weather reporting from many of their weather reporting sites in the central Soviet Union and the areas bordering China. They assumed we were collecting their weather data being relayed by HF radioprinter back to Moscow and wanted to deny the US forces access to information that might help us make better weather forecasts in support of strategic strike planning against the Soviet Union or, especially during my era, weather over Southeast Asia in support of our operations in Vietnam. Although never proven, from observations of the amazing number of Soviet aircraft lost to weather during training flights over their own territory we suspect that they were actually using the falsified weather for their own predictions. Perhaps their weather twinkies hadn’t been told that their own synoptic data reports were not reliable.
“Now, forty years and more later I am sure that hordes of unsuspecting worldwide climatologists are actually using those deliberately corrupted datasets to make profound judgements on global weather trends.”
Yet another comment from the above thread at WUWT led to a comment from an even earlier blog of Anthony’s, citing the venerable Australian atmospheric chemist — well known to some AGW skeptics — Vincent Gray, (available from the old (late) John Daly blog:
“‘Examination of the data shows that almost all of the 1901-1996 temperature rise for Russia/Soviet Union took place in one year, 1987 to 1988.’ This was about 3 years before the collapse of the Soviet Union, but two years after Perestroika and Glasnost.”
“Gray says in another paper ‘Although some Russian stations have excellent records over a very long time, the service has deteriorated in recent years, together with the rest of the Russian economy. In 1988 there were 244 temperature stations, but in 1989 135 were closed; mainly the smaller ones, leaving only 109 stations. Most of the 91 5°x5° grids in Russia/Siberia in Figure 2 will be represented by single stations. Recent monthly records from Russian stations show many gaps and doubtful figures.’ In the same paper he notes ‘Monthly temperature records for the Russian stations show an extreme temperature range of around 60°C. Early measurements are likely to have been in primitive or deprived conditions. Stations would have been operated by political prisoners.'”
http://www.norcalblogs.com/watts/2007/08/1998_no_longer_the_hottest_yea.html#comment-125343
So how to assess all of these discontinuities? Has it even been seriously attempted? Even a sample effort? My hunch is that a shout out to Professor Roger A. Pielke (Jr or Sr) at Colorado may find a helpful answer.
Finally, a lively comment from Anthony’s 2008 thread on top to close this post”
“There’s an old Soviet saying: ‘As long as they pretend to pay us, we’ll pretend to work.’ That would almost certainly include ‘we’ll pretend to measure the temperature.'” Indeed, the problem was endemic to their social and political culture. This informations gap caused the system to fail.
UPDATE: are old Soviet temperature records unreliable? A quick search of the foia online Climategate emails indicates that the Hockey Team disputes that it is. But based on only a few early studies.
EMPHASIS MINE
http://di2.nu/foia/foia2011/mail/0715.txt
A same day reply came from one Pasha Groisman, sporting a NOAA email address:
orson2 says:
……
Thank you very much for the info.
Russia and Canada are very important to the global temperature calculations because of their size, since presumably each sq km is given same weighting. If they use longitude-latitude greed it would make matters far worse.
Steven Mosher has somewhere given precise definition, so if Steven is about perhaps it would be useful to remind us.
CONFOUNDING?
In December, 1995, Pat Michaels cites a chart appearing in a paper by Jones and Briffa on the non-changing [!] Soviet growing season temperatures.
Few people recall that Arrhenius predicted greenhouse warming at night and in the winter, says Michaels. But in the chart of seasonal Soviet temperature anomalies, there it is: WARMING IN WINTER (from about 1980 to 1995):
Michaels gives the caption emphasizing the point: Figure 1. Seasonal and annual temperatures over Russia; mean latitude is 60°N. Jones and Briffa have further divided the data and found that all the observed warming is in the cold half of the year.
http://www.worldclimatereport.com/archive/previous_issues/vol1/v1n8/hot.htm
But is this simply the old data picture? Recycling suspect Soviet temperature data?
I’ll let someone else read the Jones and Briffa paper and report back, if interesting:
Jones, PD and Briffa, KR (1995) Growing season temperatures over the former Soviet Union. International Journal of Climatology, 15 (9). pp. 943-959.
ABRSTRACT and eprint LINK:
https://ueaeprints.uea.ac.uk/33835/
Earth’s Land Surface 148,300,000 km²
USSR 22,402,200 km²
Canada 9,984,670 km²
China 9,598,086 km²
USSR +Canada + China = 28 % of Earth’s Land Surface
1C error in the same direction for those three would amount to 0.3 C of the global anomaly, which is well over 1/2 of the land temperature rise since 1940s.
Latest:
More global warming coming your way
History of climate change re-written with release of Russian data
Published Mon 19 Mar 2012 16:41
The history of a changing climate has been officially re-written following the release of new data from Russia and bases within the Arctic Circle.
Scientists have now calculated that 2010 has overtaken 1998 to now be the warmest year on record, followed in second place by 2005 as 1998 is pushed into third place.
The recalculation of the annual global mean temperature records follows the release of weather data from more than 600 locations around the Arctic Circle.
The dataset is compiled by the Met Office and the University of East Anglia’s Climatic Research Unit, who today announced the update.
http://www.clickgreen.org.uk/news/international-news/123326-history-of-climate-change-re-written-with-release-of-russian-data.html
Orson2: Welcome, and thanks for your input.
Vuk: So, a warmer arctic eh? Should just about balance a cooling Antarctic no?
Vuk:
Thanks for the AMO data link. Using the Kaplan AMO data set I can replicate your results quite closely, and I also see an apparent abrupt upward shift of about 0.4C in GISS global SAT relative to the AMO index between 1970 and 1972. However, I think this shift is an apparent effect generated by a “spike” in 1972. Without the 1972 data we see SAT warming gradually relative to the AMO Index between 1955 and 1975 but no sign of an abrupt upward shift.
The 1970-72 shift also isn’t visible in any of the following comparisons:
GISS global SAT vs. AMO Index derived from HadSST2 or ERSST.
GISS global SAT vs. North Atlantic SST (undetrended AMO).
GISS Siberia SAT vs, AMO or North Atlantic SST.
Sorry I can’t produce any backup graphs right now but I’m still trying to figure out how to use some of the software on my new computer. Maybe later.
Thanks Roger.
I am not too bothered too much with the global numbers, the AMO accuracy is far more important for my own research. There is already good indication from the Reykjavik Atmospheric pressure measurements (my very own discovery http://www.vukcevic.talktalk.net/theAMO-NAO.htm ) that AMO is probably OK.
One interesting part of this period is relatively good AMO-SSN synchronisation from 1968 to1992 , but it became nearly perfect during those 2 years in question 1969-1971.
http://www.vukcevic.talktalk.net/AMO-SSN.htm
Maybe a coincidence, maybe not, in which case perhaps a further confirmation that the deep in the AMO did indeed occur.
Vuk
No question that the AMO is real. It controls precipitation on both sides of the Atlantic, it’s closely correlated with hurricanes and it even controls glacier retreat in the Alps – see Figure 3 in http://www2.gi.alaska.edu/~regine/Huss&Hock2010_GRL.pdf – although I claim prior rights here, having plotted this figure before Huss & Hock did 🙂
One interesting thing I can’t explain, though, is why Arctic air temperature anomalies lead the AMO by about 5-10 years. Any ideas?
And I’ve just noticed that Huss & Hock’s Figure 3 shows that glacier mass balance changes in the Alps also lead the AMO by 5-10 years. Curiouser and curiouser.
Roger just read my ‘discovery’ paper
http://www.vukcevic.talktalk.net/theAMO-NAO.htm
you can start on page 7, most of it is in there, but a new project will explain all details, see the last graph in
http://www.vukcevic.talktalk.net/NAOn.htm
Hi Vuk
Obviously I make a good straight man for you.
That looks like a good piece of work. I plan to study it in more detail, but immediately I have two questions:
1. Why do you use July-August Reykjavik pressures? Why not January-February, April-May, or for that matter annual pressures?
2. Where did you get your July-August pressure data? The summer (presumably JJA) Reykjavik pressure data on the KNMI site don’t seem to match it.
Hi again
Not any prior or superior knowledge of the subject. The JA pressure shows highest correlation, reasons I think is simple, Icelandic Low (IL) moves further north with the summer ice retreat, although ice coverage is lowest in September, insolation of the Arctic is considerably reduced and the IL is already on the move back(sun + ice). By the end of November IL is back at the southern tip of Greenland in the Irminger Sea.
It is difficult to say why the far north IL is more important, I am not sure, but I think could be that the IL’s effect on the SPG is at that time weakest, hence negative correlation. Btw, there is no huge difference between summer and winter as measured at Reykjavik (Fig.9). Both can be used with some confidence to forecast the Reykjavik’s winter and annual temperatures (another little discovery of mine, widely ignored):
http://www.vukcevic.talktalk.net/RF.htm
Data: originally I used http://www.esrl.noaa.gov/psd/cgi-bin/data/timeseries/timeseries1.pl, and when correlation proved good, I wrote to Dr. Hurrell and got the anomaly data file used by NCAR.
Vuk:
Have you looked into correlating the AMO with the 60-year Jupiter/Saturn cycle? Seems to me you should get a pretty good match.
Cloud variability alert:
http://hockeyschtick.blogspot.com/2012/04/new-paper-finds-significant-natural.html
Roger Andrews says:
………
It has crossed my mind some time ago:
http://www.vukcevic.talktalk.net/GTC.htm