NASA spokesmen Gavin Schmidt and James Hansen have observed that the US accounts for only 2% of the earth’s surface and conclude that problems in the US network don’t matter. There are only 6 Brazil stations in GHCN which have records extending through the 1930s and as late as 2004, which is an indication of the radical difference in coverage between two nations with large land areas. One of these stations is identified in the GHCN meta-data as “rural”. Can you identify which one?

Plot of 6 Brazil GHCN stations (unadjusted) with values through the 1930s and at least 2004 (in all cases 2007). All in deg C anomaly 1961-90.

I’ve shown GHCN unadjusted data, which, as far as I can tell, is the version used in the NOAA composite. I have no idea what version is used by CRU. I’ll look at GISS on another occasion. Recall that Jones et al and other studies cited by the IPCC have said that urbanization impact is under 0.1 deg C.


  1. vincent Guerrini
    Posted Aug 17, 2007 at 8:56 PM | Permalink

    This should be the beginning of a worlwide/continent by continent analysis of rural v urban sites (unless this has been done). Answer QUIXERAMOBIM. Have a look at Puerto Varas or Punta Arenas, Chile. You’d think that by now ONLY rural data should be used for world surface data analysis. BTW does anybody know why UH satellite data for July 2007 not posted yet? (CRU3 etc has)
    good work keep it up.

  2. TCO
    Posted Aug 17, 2007 at 8:59 PM | Permalink

    [snip] no f-words allowed

  3. Paul S
    Posted Aug 17, 2007 at 9:00 PM | Permalink

    I think one would have to commandeer use of a satellite to determine if a site is truly “rural” 😉 but I’d guess Quixeramobim. But only 6 Brazilian stations in GHCN? Wow.

  4. Frank K.
    Posted Aug 17, 2007 at 9:07 PM | Permalink

    “I’ve shown GHCN unadjusted data, which, as far as I can tell, is the version used in the NOAA composite.”

    Just so I understand this correctly…NASA/NOAA uses UHI corrections for the US (presumably because there are usually rural stations close by the compare with) but not for the rest of the world??

    And there are only ** six ** stations in Brazil with historical data dating back to the 30s?? What about 1880 – 1930? All of the global average temperature charts date back to 1880. Maybe Sounth America, like the US, isn’t important…

    And these (uncorrected) six stations are somehow representative of the entire country of Brazil (which is a significant fraction of the South American continent)? Look at Sao Paulo! UHI run amok…

  5. Larry
    Posted Aug 17, 2007 at 9:17 PM | Permalink

    It would be interesting to plot the populations of these respective towns/cities on the charts, and see how well the knee in the temperature curve lines up with a knee in the population curve.

  6. bernie
    Posted Aug 17, 2007 at 9:33 PM | Permalink

    I have been to Sao Paolo and it is an amazing place with incredibly rapid growth. It is similar to LA in that the city grows everywhere with apparently minimal urban planning.
    That Sao Paolo under any circumstances should represent a significant portion of the historical data for a country the size of Brazil is crazy.

  7. Harold Vance
    Posted Aug 17, 2007 at 9:38 PM | Permalink

    Urban heat in, global warming out.

  8. bernie
    Posted Aug 17, 2007 at 9:47 PM | Permalink

    I forgot to mention that Sao Paolo’s metro population is close to 20 millions!! And I have to tell you without air conditioning the place is kind of tough. A quick google search produced this set of data on Brazil’s urban growth and select city population and Brazil.

  9. windansea
    Posted Aug 17, 2007 at 9:48 PM | Permalink

    James Hansen replies to the “jesters”

    Click to access realdeal.16aug20074.pdf

  10. Doug F
    Posted Aug 17, 2007 at 9:50 PM | Permalink

    Interesting. I don’t see anomolously high temperatures in the 30’s. The sample is small, but suggests that the temperatures in Brazil did not follow North American temperatures in the 30’s. Perhaps proxies are required because the number of sites is so tiny.

  11. Steve McIntyre
    Posted Aug 17, 2007 at 9:51 PM | Permalink

    These were sites that had values in the 1930s AND extended to 2004 or later. There might be a few sites that go back to the 1930s but end before 2004. I’ll check some time.

  12. Paul S
    Posted Aug 17, 2007 at 10:05 PM | Permalink

    Re Post # 9 by windansea

    I read that post. Remarkable. Hansen is a politician, no doubt about it. And he gets in the Goreian shibboleth about Exxon too.

  13. Posted Aug 17, 2007 at 10:06 PM | Permalink

    &#9835 One of these things is not like the others… &#9835

  14. bernie
    Posted Aug 17, 2007 at 10:09 PM | Permalink

    Interesting statement but hardly revelatory. Citing Jefferson, the slave owner, on the Bill of Rights I think says it all.
    The defense seems to be – OK, AGW appears to be minimal in the US but just look at the ROW. Brazil does not appear to be a great case example.

  15. Posted Aug 17, 2007 at 10:10 PM | Permalink

    PS. Those numbers are the code to double-eighth notes. Strange that they worked in the preview plain but not when posted. Where is John when we need him??? 🙂

  16. Steve Moore
    Posted Aug 17, 2007 at 10:16 PM | Permalink

    RE #9:

    Remarkable reading. From the Al Gore School of Climatology:

    “…large climate impacts become inevitable, including the loss of all Arctic sea ice, destabilization of the West Antarctic ice sheet with disastrous sea level rise later this century, and extermination of a large fraction of animal and plant species…”
    “There is still time to avert the most dramatic climate effects, if we promptly begin to address both CO2 and non-CO2 climate forcings. But just barely.”


  17. Posted Aug 17, 2007 at 10:23 PM | Permalink

    Doug said:

    Interesting. I don’t see anomolously high temperatures in the 30’s.

    That is one of the foibles of the “hottest day / month / year on record” for any small area. Sure, it may be breaking records there, but there is certainly somewhere else where the temps are boringly average, or annoyingly below average, like this summer in Fresno for example. And of coarse these average / below average measures will never make headlines… unless it happens to snow on the same day Al Gore shows up to give a panicked sermon on AGW.

  18. Mike H.
    Posted Aug 17, 2007 at 10:43 PM | Permalink

    sonicfrog, semicolon at the end of the number. ♫
    Steve, the current post of the Patriot Post has a synopsis of the current discussion with Hansen. Would you care to have it forwarded?

  19. Posted Aug 17, 2007 at 10:44 PM | Permalink

    A simple question: If Hansen is a scientist and genuinely concerned at our future you would think he would accept the error and make the decesion himself to ensure all the other data was checked. Does he want the real story or not? An assured and confident man would ask Steve for assistance not attack his findings.

  20. Dave Dardinger
    Posted Aug 17, 2007 at 10:46 PM | Permalink

    re: #9

    Good Gravy! The man is insane. Quick, how old is he? Perhaps AGS is causing premature senility in scientists who work to close to it. I’m over 60 now. Perhaps I should stay away from here and take up a safer hobby.

  21. steven mosher
    Posted Aug 17, 2007 at 10:58 PM | Permalink


    In my “discussion” with Gavin over at RC with regards to the “culling”
    of Norcal stations, he has remarked that these five stations were culled because the
    data was Weird. ( his word. That’s a students W test ) anywho. CUIABA shows a cooling regime that
    is Weird. Weird at like 3sigma weird, like at totally weird. Cull that data!

    San Paulo, on the other hand shows a trend that is in consensus with the ROW and
    so it is the reference station.

  22. SidViscous
    Posted Aug 17, 2007 at 11:00 PM | Permalink

    Just for comparison. Brazil is similar in size as the U.S. (Lower 48).

    Actually before we purchased Alaska (For a case of Vodka and a cartoon of Marlboro’s) Brazil was actually larger than the U.S.

  23. Joel McDade
    Posted Aug 17, 2007 at 11:05 PM | Permalink

    Looking quickly at the GISS site, using “after homogeneity adjustment”, Curitiba has no warming to speak of. Sao Paulo and the others are absent for some reason after this final adjustment, but not before. Why?

    The Rio de Janeiro record goes back to 1880. Why isn’t it in GHCN?

    Brasilia, the capitol, should be looked at closely. As I understand, it was built from the ground up starting in the ’50s. This might make it unique from a temperature POV.

  24. Joel McDade
    Posted Aug 17, 2007 at 11:26 PM | Permalink

    Hmmm… many of the GISS charts in Brazil go for years without data, then start again. How they can adjust – ah never mind.

    I love Brazil. Anthony, I volunteer, but need a little funding.

  25. Posted Aug 17, 2007 at 11:29 PM | Permalink

    I have purchased and am willing to host and donate the name to the investigation of stations in Canada. If anyone want to help identify and catalog Canadian Stations give me a holler at . I’ll be setting up the website in the next few days.

  26. Jim Edwards
    Posted Aug 17, 2007 at 11:30 PM | Permalink

    Steve M:

    Most of the ROW will likely have been too poor to give a *$#* about temperature readings before 1960. What about British Commonwealth countries. The Royal Society have been gathering science data in the empire for centuries. I’d guess you’d find more spatially dispersed ‘old’ sites for urban / rural comparison in Commonwealth countries.

  27. paul graham
    Posted Aug 18, 2007 at 12:12 AM | Permalink

    Do we have the exact location and the metadata for these sites, other wise i would be hard to check exactly
    Below you’ll find a google map links; Any complains about the link can go to hell ‘TCO’.

    What i find interesting is Salvador and Manaus are by the sea and a lake; as if the water is at the same tempeture. Quixeramobim is the most rural and show a cooling effect and finally cuiaba is in the heart of the amazon and shows a massive drops possible cause by station moving and Deforestation.
    In general all site are to different degree effected by Urbanation and Quixeramobim having the least tarmax/asphalt.

    Now i willing to admit that this isnt the best audit; but unless i get tickets to brazil and more info on the stations its a start.

    Sao Paulo
    a very large city

    A very denesly pack town of two story buildings.

    A large city with many building above five floors

    A large town with many home Surrounded by the sea.

    A large town with a mixiture houses and warehouses.

    Quixeramobim is the smallest town small town. This must be the rural station.

  28. Shawn
    Posted Aug 18, 2007 at 12:33 AM | Permalink

    I’m confused how your wrote about the US only making up 2% of the earth surface, adding that it doesn’t matter, I didn’t understand where that was going, but this is what was I think.

    If you think of the smoke stack capital of the world; it’s probably not far from you. However, in a scale of a month, our world atmosphere quickly flows from one place to the next, and it moves all around the globe, wouldn’t you agree? So isn’t it safe to assume that if North America were to immediately stop emitting GHGs that the effect would merely act as a hole which closes, but has already set in emotion GHGs that will float around the globe for centuries. I don’t think it all hangs in one place. Maybe over cities more because they are essentially GHG producing super producers (one collective smoke stack).

    But I refuse to believe, I haven’t been breathing some of your city’s Carbon Dioxide output, and that our industrialized nations aren’t spreading a thickening blanket of millions of years of once absorbed CO2, over the oceans and lands which are greatly sensitive to their climate.

    The climate is always in motion. The 1930s may have been a period as a result of sort of an anomaly that lasts only a few years , which climate forcings did not move much, or converged in such a way, that the US was very, very warm for a period then, not necessarily because of CO2 but say, because of the lack of Artic Air, who knows… But we all acknowledge that we are currently experiencing equally warm periods now, as the 30s, which we consider really warm, every year in fact. Surely, the entire decade of the 00s will prove to be warmer than the 90s, the 90s more than the 80s and so on. But now also, we are able to observe and understand the climate significantly better.

    It’s not entirely far fetched to consider we are experiencing warmer temps in the US is because we’re out of control with C02 now, not only are we smothering in our own gasses but now more from the world around us, including even our past, all joining in the mix.

    The cat is out of the bag. The WORLD’s warmest years on human and geological records dating back over 400,000 years have all occurred… recently (and I don’t have to go back as early as 1934 A.D.). What say you of that?

  29. paul graham
    Posted Aug 18, 2007 at 1:12 AM | Permalink

    Shawn# Greenhouse gases have increase since the 1930; yet temperatures haven’t; the two are incompatible. The point is that if climate alarmist models are correct then temperature would have increased. In addition, these theories rely on the climate being incredibly stable however when we look into the past see don’t see this; in fact we see temperatures that are often hotter that today. All this adds up to a very doggy theory.

  30. Brooks Hurd
    Posted Aug 18, 2007 at 1:48 AM | Permalink

    Re: 9, Windsea,

    Hansen’s temperature graphs which he uses to compare the US to the ROW are amazing. The y axis for the US is 3 C whereas the y axis for the ROW is 1 C. This is on the same page where he accuses others of misrepresenting data to prove their points.

  31. chrisl
    Posted Aug 18, 2007 at 1:57 AM | Permalink

    Brooks: Well spotted, indeed if you re-scaled the y-axis you could get flat lines out of both graphs

  32. Louis Hissink
    Posted Aug 18, 2007 at 2:56 AM | Permalink


    Just occurred to me but if the argument is about climate and changes in climate, then should we not be using statistics of climate data rather than weather to determine if the climate is changing?

    Climate is defined as weather averaged over 30 years, so this definition would produce a rather small number of data points (unless one uses a moving average and then the last data point would be 2007 – 15 = 1992) that might pose some difficulties in modelling. Much like trying to convince the folks that a trend discerned from two data points is important.


  33. paul graham
    Posted Aug 18, 2007 at 3:34 AM | Permalink

    Louis Hissink# this an excecllent point, and i belive that this is covered else where ie historical hurricane activity. Here’s some links.,2933,277164,00.html

    Click to access Chapter%205.pdf

  34. KevinUK
    Posted Aug 18, 2007 at 3:34 AM | Permalink

    IMO James Hansen is a very dangerous man.

    Why is he dangerous? Because since his 1988 alarmist AGW testimony to the Senate, he has caused billions of dollars to be diverted to the ‘AGW industry’ rather than for that same money to be spent on real global issues like poverty, disease, famine, poor education etc. A a consequence IMO many hundreds of thousands, probably millions, of people have died who otherwise might not have done.

    He no doubt sincerely believes that he is almost single handedly saving the planet (to the benefit of the affluent western world) but at what expense. Answer? At the expense of the un-developed third world.

  35. Posted Aug 18, 2007 at 4:00 AM | Permalink

    Quixeramobim is rural. An accident that it is the most constant graph? A somewhat funny detail: this rural (R) location may be found by Google at

    That’s quite an interesting program and list of data. One of the things I did was to look for “Czech” on that page. The first “Czech” label you find is Arvavaralja. Doesn’t sound too Czech to me. It’s a Hungarian name for a Slovak village. The next one is Schossl and Tetschen, German names that could be outside Czechia. It’s interesting because there are many Czech cities around that are not labeled “Czech”.

  36. Paul
    Posted Aug 18, 2007 at 4:09 AM | Permalink

    #9. How can anyone take Hansen seriously??

    The real deal is this: the ‘royalty’ controlling the court, the ones with the power, the ones with the ability to make a difference, with the ability to change our course, the ones who will live in infamy if we pass the tipping points, are the captains of industry, CEOs in fossil fuel companies such as EXXON/Mobil, automobile manufacturers, utilities, all of the leaders who have placed short-term profit above the fate of the planet and the well-being of our children. The court jesters are their jesters, occasionally paid for services, and more substantively supported by the captains’ disinformation campaigns.

    There seems to be a dramatic rancour among people like Hansen resulting from the fact that general populous find the case so underwhelming that they aren’t selling their cars, ripping out central heating, cutting off their electricity.

    Get over it mate.

  37. JS
    Posted Aug 18, 2007 at 4:11 AM | Permalink

    Shawn # 28

    You read that where, Newsweek?

  38. Paul
    Posted Aug 18, 2007 at 4:15 AM | Permalink

    #22 Just for comparison. Brazil is similar in size as the U.S. (Lower 48).

    Yeah, but its only 2% of the world surface.

    ..and Australia? Well its only 2% of the world’s surface…

    …And China? well its only…

    For supposed smart guys with PhDs they sure are stoopid. It about the treatment of the data dumbies…..

  39. Michael Smith
    Posted Aug 18, 2007 at 4:46 AM | Permalink

    Regarding #9, I love the way Hansen closes:

    Criticisms, as always, are welcome.

    Right, Hansen, you welcome criticisms. When NASA administrator Michael Griffin made the perfectly reasonable observation that we in the west have no right to determine some sort of optimal temperature for the globe and then force that on the third world and future generations, you “welcomed” that comment.

  40. dover_beach
    Posted Aug 18, 2007 at 5:34 AM | Permalink

    Hansen’s agitprop is more disturbing then the recent amendments to GISS data. On topic, I simply cannot believe that UHI is still discounted considering the palpable difference in temperature trends between rural and urban locations as indicated above.

  41. windansea
    Posted Aug 18, 2007 at 5:49 AM | Permalink

    Hansen comes off as as immature clown who does not like his data reviewed

    I found it on InstaPundit so hopefully it will get some exposure

    Hopefully Mr McIntyre will do a post on it so we can link it far and wide

  42. EW
    Posted Aug 18, 2007 at 5:53 AM | Permalink


    One of the things I did was to look for “Czech” on that page. The first “Czech” label you find is Arvavaralja. Doesn’t sound too Czech to me. It’s a Hungarian name for a Slovak village. The next one is Schossl and Tetschen, German names that could be outside Czechia. It’s interesting because there are many Czech cities around that are not labeled “Czech”.

    Luboši, I worked on location of those on W. Hughes’ request. Both are Czech – Schossl is Střelná, seems to be a local small airport station and Tetschen is Děčín. Also there was a German-sounding place Neu-somewhat, attributed to Czechoslovakia, which we finally located in Poland, near Klodzko. IMO, in that list, the stations have their pre-WW II and maybe pre-WW I names!

  43. Douglas Hoyt
    Posted Aug 18, 2007 at 5:54 AM | Permalink

    Southern Brazil is one region where there are big differences in trends between MSU and the surface stations. There was a paper in JGR many years ago (by Jones, I think) showing the differences.

  44. John F. Pittman
    Posted Aug 18, 2007 at 6:11 AM | Permalink

    I wonder if Hansen would appreciate the practical error in his posturing. As #34 KevinUK has pointed out, real world economies are not that simple. Quoting from the publication without peers, the IPCC WGII, we see

    Even the most stringent mitigation efforts cannot avoid further
    impacts of climate change in the next few decades (Working
    Group I Fourth Assessment Report, Working Group III Fourth
    Assessment Report), which makes adaptation unavoidable.
    However, without mitigation, a magnitude of climate change is
    likely to be reached that makes adaptation impossible for some
    natural systems, while for most human systems it would involve
    very high social and economic costs (see Chapter 4, Section
    4.6.1 and Chapter 17, Section 17.4.2)

    It is not yet possible to answer the question as to whether or
    not investment in adaptation would buy time for mitigation
    (high confidence).
    Understanding the specific economic trade-offs between the
    immediate localised benefits of adaptation and the longer-term
    global benefits of mitigation requires information on the actions’
    costs and benefits over time.

    First, there are or have been effects. Secondly, the economic trade-offs are unknown in terms of buying time. Somewhere in the AR4 isthe acknowledgement that rich economies can more easily afford mitigation than poor economies. From IPCC the costs of carbon:

    peer-reviewed estimates have a mean
    value of US$43/tC

    Lets see what a ton of carbon is. We will use methane as the basic carbon unit. Carbon is .75 by weight. So we will see what we get for 1.333 tons of methane. Let’s just use gasoline as our estimate. We get about 400 gallons to burn for that $43 dollars. The question is what starving 3rd world person would not want to pay 11 cents a gallon for gasoline to have food shelter and medical care? I wish I only had to pay 11 cents a gallon. I believe that the hypocricy of Hansen et al is properly thought as hubris. “Pride cometh before the fall” or more classically, Hansen, et al, believe they are on par with the gods (greek Tradegy).

  45. Douglas Hoyt
    Posted Aug 18, 2007 at 6:32 AM | Permalink

    Re43. After thinking some about it some more, the paper was probably by Christy and Spencer. The MSU trend variations were very smooth geographically whereas the surface trends had lots of spatial variation. The paper had nice color plots of the globe with color coded trends in the grid boxes. Anyway, the analysis should be done again. I think Steve could do it rather easily in R.

  46. Stan Palmer
    Posted Aug 18, 2007 at 6:44 AM | Permalink

    Hansen thinks that he is insulting skeptics when he calls them “court jesters.” Court jesters were the people who were allowed and expected to make the king face the truth. They could force the king to reconsider his most cherished ideas.

    The famous scene of King Lear on the moor with his jester is a good example of that point. Lear was railing against the force of a hurricane. The hurricane is a metaphor for the indifference of reality to Lear’s delusion about his own importance.

    So if the skeptics are court jesters for forcing Hansen to admit his error; then they are court jesters in the finest tradition of that role.

    King Lear could be a very apt play to represent this whole AGW debate. Jim Hansen as Lear

  47. kim
    Posted Aug 18, 2007 at 7:03 AM | Permalink

    Everytime the Gorebellied Fool opens his mouth, somewhere something sentient freezes. Not good Karma, me boy.

  48. Posted Aug 18, 2007 at 7:05 AM | Permalink

    Dear EW #42, thanks for your info. “Strelna” was new for me. But is it OK to say that “Czech” was no good for “Orava” neither before the first war nor before the second war? 😉

  49. N Leaton
    Posted Aug 18, 2007 at 7:25 AM | Permalink

    Clearly a major issue is the weight of US data to the world average.

    Now its 2% of the area, but 1200 stations out of some 3000 odd.

    I can’t quite believe that its weight is just 2%.

    Does anyone know what the real figure is?

    This also shows another problem illustrated by the Brazil and other countries data.

    If you reduce the weight of the US to 2%, then you end up increasing the weight of a small number of stations else where that might not be accurate, and so end up having an undue influence.

    The result that feels right, is then something in the middle.


  50. bernie
    Posted Aug 18, 2007 at 7:35 AM | Permalink

    #28 Paul
    The link, I gave in #8 provides population estimates and projections for these cities from 1915 through 2015.
    #30 Brooks
    Steve provided these graphs rescaled to the same scale. When you do that you see that the 30s appears to be somewhat of an US anomaly. This, I believe, is why Steve is looking for Brazil data for this period.

    As you probably know, Brazil is very much of a melting pot, with many German, Italian and Japanese immigrants. In fact Brazil has the largest ethnic Japanese population outside of Japan. Brazil also has some strange climates with a arid region in NE Brazil due to the cold current off the coast.

  51. DocMartyn
    Posted Aug 18, 2007 at 7:45 AM | Permalink

    The Brazilians had a mini-civil war from 1918 to 1928. There was a breakdown in central government control, starting at the end of WWI, during this period. I find it odd that the cuiaba and curitiba records have, what appears to be gaps, here.

  52. Gunnar
    Posted Aug 18, 2007 at 7:49 AM | Permalink

    #28 >> If you think of the smoke stack capital of the world (US);

    Shawn, you really need to look into this more. In reality, the communist (china) and the former communist areas of eastern europe/russia are far, far bigger producers of raw pollution. Since the free world is inherently cleaner, based on private property rights, people who want a clean world should be clamouring for freedom, not green communistic state control.

    >> has already set in emotion GHGs that will float around the globe for centuries.

    Freudian slip? That’s exactly what AGW is doing, setting GHGs into our emotions. For certainly, there is no data or scientific law that can support the idea that C02 increasing from .035% to .038% can have any deleterious effect. C02 cannot be pollution, since plants need it like we need oxygen. You’re not a plant bigot, are you? Even the AGWers have shifted their always slippery argument from “greenhouse warming” to C02 absorbing radiation, and re-radiating.

    I’ll be the first one on the AGW bandwagon, as soon as someone comes up with a coherent scientific hypothesis (does not violate other known laws), and then repeatedly fails to falsify it with experiments.

    In the meantime, I’ll let you in on a little secret. C02 is soluable, which you may have noticed when you drink a soda pop. This means that as C02 floats around, it is absorbed into the colder parts of the ocean. The average residence time for C02 in atmosphere is about 5 years. Therefore, based on AGW assumptions (man is the only C02 source & C02 causes temp rise), if we stopped all C02 output in 2090, all C02 would out of the atmosphere by 2100 and the temperature would plummet. No need for any action.

    >> Maybe over cities more because they are essentially GHG producing super producers (one collective smoke stack).

    Actually, a 2003-2005 study failed to find a C02 dome over a city. The average persons C02 output is negligible in his immediate environment. When you multiply this by the number of people on earth, it becomes even more negligible, since a lot of land in sparsely populated, and all the ocean is.

  53. Gunnar
    Posted Aug 18, 2007 at 7:52 AM | Permalink

    >> the 30s appears to be somewhat of an US anomaly

    I guess people really do over extrapolate a lot. For me, it never occurred to me that the US 30s heat wave was global, so I’m not at all surprised by this. I am very surprised that other people are surprised by this.

  54. bernie
    Posted Aug 18, 2007 at 8:10 AM | Permalink

    The issue is one of what the cause of the 30s heatwave was. If it was local “weather” you are absolutely right. However, if it was a combination of sun and ocean driven forces then I would expect to see similar patterns elsewhere. The point of raising the question is that the temperature record in the 30s is partial for much of the world, which raises the question of how this record was actually constructed, i.e., much of the Asian land mass was in turmoil and much of Africa had little if any record. Until Steve helped reorder the “hottest” year data, I hadn’t even thought about the likely gaps in the 30s record but I had simply taken the record on faith. In addition, the temperature blip in the 30s appears to be a North American phenomenon not just the lower 48, which changes the weight from 2% to 6% of the surface or from 6% to 16% of the land record. So from a data point of view, if the ROW does not show the same peak, the greater the weight for area with a peak, the greater the need for a flat or downward trend in the ROW. All I tthat is being done is a double checking of the data given known likely sources of error or discontinuities.

  55. Posted Aug 18, 2007 at 8:15 AM | Permalink

    I am not scientist, but it sounds pretty wierd to me: you guys cannot obtain from NASA or from CRu list of station they use for constructing their graphs and reports on global mean temperature? How its possible?
    I live in Serbia. Is there any Serbian station in GISS or CRU network? China, Indya? You ahve no way of knowing becaiuse they will not give you the data? Crazy..

  56. Murray Duffin
    Posted Aug 18, 2007 at 8:16 AM | Permalink

    Hansen should be ashamed of himself. To write a polemic about distortion, and support his position with a major distortion of the Y axis scale that most people won’t notice is hypocrisy squared.
    Interestingly Sao Paolo has been growing like mad since the 1950s, while Curitiba is to a large degree a planned city that experienced major growth since the early ’70s. Clearly shows in the curves. Sao Paolo shows almost 3.5 degrees C of warming. Wow!! Murray

  57. Steve McIntyre
    Posted Aug 18, 2007 at 8:27 AM | Permalink

    #56. The GISS station list is available, not the CRU.

  58. Posted Aug 18, 2007 at 8:41 AM | Permalink

    I know you requested from Jones CRU data and he refused, by saying that he will not give the data because you will try to find something wrong with them. Is it the end of story? IS there any other way to identify the stations?

  59. steven mosher
    Posted Aug 18, 2007 at 8:45 AM | Permalink

    anybody know porteguese

  60. Bob Meyer
    Posted Aug 18, 2007 at 8:55 AM | Permalink

    Re 54

    If the high temperatures of the 1930’s were a strictly US phenomenon then if follows that local effects can be as great, if not greater than global ones. This is in contradiction to the “CO2 uber alles” theory of GW.


    A – the temperatures measured in the US are valid, which means that there is some unknown process of incredible power that raised the temperature of an entire continent seventy five years ago, or

    B – the temperatures measured in the US are invalid which means that conclusions drawn from that data are useless.

    Surely a local phenomenon that raised the temperature of North America as much as CO2 raised the temperature of the planet is worth investigating. A real scientist would chase down that effect, especially if it called his own theories into question. Hansen shows no interest in pursuing this, he is more interested in calling people jesters and clowns.

    This North American temperature anomaly makes the station and data audits more important than ever.

  61. steven mosher
    Posted Aug 18, 2007 at 9:45 AM | Permalink


    What do the grids look like for Brazil?


    Slide 6. looks like there are more stations….. Might be an interesting

  62. Gunnar
    Posted Aug 18, 2007 at 9:55 AM | Permalink

    >> 2% in the atmosphere (mainly as carbon dioxide).

    Not sure what you mean, atmospheric C02 is .038%. Are you saying that 2% of the world’s carbon is in the atmosphere? Can’t be that.

    >> Increasing CO2-levels in the atmosphere today is consistent with the medieval warm period

    Yet, they seem lower now than in 1940.

  63. Joel McDade
    Posted Aug 18, 2007 at 9:55 AM | Permalink

    Looking further at GISTemp, here the rural sites with data to the ’30s that I could find:

    – Quixeramobin, 1896-2007, no trend
    – Turiacu, 1911-1963, positive trend since the ’30s
    – Caetite, 1931-1981, no trend(?), lots of missing data
    – Iguape, 1895-1988, positive trend
    – Igarapava, 1901-1930

    However, Quixeramobin is the only one with current data. ALL rural sites with current data include

    – Quixeramobin, 1896-2007, no trend
    – Caravelas, 1949-2007, mod. positive trend since 1950
    – Conceicao do (Araguala), 1960-2007, strong trend since 1975
    – Carolina, 1973-2007, strong trend since 1975
    – Barra do Corda, 1960-2007, strong trend since 1975

    At least I think I caught them all.

    So, there are only five rural sites currently in operation in Brazil. All are in the northern half of Brazil, except maybe Caravelas (mid way, on coast). Cities like Curitiba to the south pick up a rural station or two in Paraguay and maybe Uruguay.

  64. bernie
    Posted Aug 18, 2007 at 10:06 AM | Permalink

    That is real helpful – can you add links? I was looking at the National Brazilian site mentioend above (I have zero Portuguese or Spanish but pretty good schoolboy Latin!) and came across this set of recent (2/2007)
    conference presentations. I looked at 4 or 5 of them. A lot are reworking of IPCC reports – but what seemed to be missing is any detailed historical chart of temperatures for Brazil. I found that very surprising. There is one chart that showed trends for South America but in a scale that made it very difficult to read. What it showed was a very slight upward trend = much less that the supposed global trend.

  65. Posted Aug 18, 2007 at 10:28 AM | Permalink

    re#9 Hansen’s reply

    The maps shown in Hansen’s figures 2 and 3 are not equal area world maps, and therefore distort the data they contain.

  66. Sudha Shenoy
    Posted Aug 18, 2007 at 10:37 AM | Permalink

    We’ve been learning about the state of numerous American weather stations, & also about the problems with long-term records: eg, change in equipment but not noted in the records, etc. The US is the world’s richest country. What are these 6 Brazilian stations like? Can it be that they are better maintained? Do we have any info on procedures, who’s doing the measuring, maintenance, what sort of equipment, etc, etc? And what about the other LDCs?

  67. Joel McDade
    Posted Aug 18, 2007 at 10:38 AM | Permalink


    Damn, I’m on a Mac and can’t readily view the presentations that you linked to. I know only enough Portuguese to get into trouble, and can’t find a T chart anywhere on the site.

    I’m just looking at the data and graphics at the standard GISS site:

    Miraculously enough, Sao Paulo appears to get tossed by GISS for (I guess) not having enough nearby rural data for the UHI correction. Curitiba and to a lesser extent Salvador get flattened out by the last correction, apparently using rural data in Paraguay.

    A quick eyeball look at Manaus, Cuiaba, etc, suggests they don’t get corrected much, presumably becuase the rural sites in N Brazil also have a strong warming trend.

  68. Paul Linsay
    Posted Aug 18, 2007 at 10:42 AM | Permalink

    #77, Ken

    I love the 2% argument.

    The Hockey Team: half a dozen trees in the Sierra Nevadas are sufficient to determine the global temperature for the entire twelfth century.

    Hansen, Schmidt, and the Hockey Team: 1,200 weather stations covering the entire US are insufficient to determine the global temperature for the twentieth century.

    Houston, we have a problem.

  69. Larry
    Posted Aug 18, 2007 at 10:50 AM | Permalink

    82, What’s really funny is that the hockey stick is based on a smal area of the SW US, and they have no problem extrapolating that to the world, but when you point out that the surface temperature record for the entire continental US is less than alarming, their answer is that it’s “only 2% of the earth’s land area”. What percentage of the earth’s land area is the bristlecone proxy indicative of? 0.1%, maybe?

  70. Gunnar
    Posted Aug 18, 2007 at 10:55 AM | Permalink

    >> What percentage of the earth’s land area is the bristlecone proxy indicative of?

    What percentage of the earth’s surface is Mauna Loa?

  71. MikeW
    Posted Aug 18, 2007 at 11:20 AM | Permalink

    Re #64

    Am in agreement with most of your post, except I would suggest a third possibility:

    C – the temperatures measured/proxied in the ROW are incomplete/invalid which means that conclusions drawn from that data are useless.

    As for the Brazilian graphs…..I have always found the fact that the apparent current warming seems to be higher in the Northern Hemisphere than the Southern a little curious, if the problem is actually caused by a well-mixed gas in the atmosphere.
    If we accept that difference as accurate, I would fully expect the 1930’s anomaly to not show (or, at least, show to a lesser degree) in the Southern record, but mostly/entirely in the Northern.

    It seems to me that this should be easily (relatively) testable: check the Southern stations and compare them against the US historical, check the northern ROW and compare them against the US historical, compare Northern historical with Southern historical.

    Has anybody done such a study?
    If so, have they tested for a lessening of impact as you approach the equator (something else that I’ve always been curious about)?

  72. Wolfgang Flamme
    Posted Aug 18, 2007 at 11:22 AM | Permalink

    Germany: has some pictures of stations maintained by the Deutscher Wetterdienst (DWD) and other weather services.

  73. Wolfgang Flamme
    Posted Aug 18, 2007 at 11:26 AM | Permalink

    Sorry, the first link got messed up. Search Google for:

    ‘Deutscher Wetterdienst’ for DWD stations.

  74. bubba
    Posted Aug 18, 2007 at 11:40 AM | Permalink

    I’ll play along. QUIXERAMOBIM?

  75. EW
    Posted Aug 18, 2007 at 11:52 AM | Permalink

    Lubos, when talking about the station list , just look at the Deutschbrod, which is called almost a century Havlickuv Brod ([probably they haven’t noticed yet…). In the list, I suppose that the station environment is given in the second to last column. Deutschbrod has “crops”, although the station is situated in the Potato Research Institute (No.1) inside the town, in a suburb-type neighborhood (family houses and gardens)…

  76. steven mosher
    Posted Aug 18, 2007 at 1:09 PM | Permalink

    Another two thing that bugs me.

    1.hansen uses nighlights photos to measure brightness. How recent are these photos?

    2. He uses 1980 census data. WTF?

  77. Philip_B
    Posted Aug 18, 2007 at 1:16 PM | Permalink

    Re: Northern versus southern hemisphere.

    Because there is much less land mass in the southern hemisphere weather is less variable. I don’t have a metric for this, but anecdotally it is certainly the case. If there is a global process driving climate then the signal should be clearer in the southern hemisphere. The fact that there is significantly less observed warming in the southern hemisphere, makes me think that most (or at least a large proportion) of the observed warming is due to local effects in the northern hemisphere.

    If anything the northern – southern hemisphere divergence is increasing. July 2007 was the 5th warmest for the NH land but only the 26th warmest for SH land. Resulting in the global land overall as the 5th warmest July. Weighting is clearly heavily to NH.

    Also of note is that the southern hemisphere has much less divergence between land and ocean temperatures compared to the NH, which indicates most warming is occuring over land in the NH and presumably results from local effects on land.

  78. bernie
    Posted Aug 18, 2007 at 1:27 PM | Permalink

    Back to Brazil:
    I pulled down all the Stations listed for South Americ and started looking at the rural stations. The first one seemed to be legitimately rural – though I can see some issues with the actual location of the station and changes ion land use. The second one I looked Coari is listed as rural. Here the population is 117, while here it is over 84,000! Now of course the actual station could be located outside of the city — but the articles also mention that Coari has grown rapidly in the last few years. Bottom-line the GISS data needs to be double-checked.

  79. steven mosher
    Posted Aug 18, 2007 at 1:29 PM | Permalink

    Another two thing that bugs me.

    1.hansen uses nighlights photos to measure brightness. How recent are these photos?

    Found it… Interesting

    2. He uses 1980 census data. WTF?

  80. Joel McDade
    Posted Aug 18, 2007 at 1:46 PM | Permalink

    hi bernie:

    Google maps has Coari as a pretty decent sized town W of Manaus, Amazonas Province. 84K seems likely. GISS has data from 1961 to 1991. We can only hope, lol, that population growth is why they quit using it.

    Lots of records with sparse data, too.

    For your 1st language lesson, look up “Bunda” 🙂 It’s all you really need to know.

  81. Magnus
    Posted Aug 18, 2007 at 2:16 PM | Permalink

    paul graham #29: Agree, that greenhose warming don’t seem to be a good theory/idea, especially since the atmosphere on the altitudes of 3000 to 10000 feet has not warmed as the surface has warmed. Not much heat radiation from greenhouse gases here…

    Approximately temp change between 1950 and 2005, city name and population 1950, 1965, 1980, 1995 and 2010.
    How the station is located is of cource crucial, but I guess UHI quite much influence the temp figures.

    -0.5 Quixeramobim ? ? ? ? 60(?) (In the geographical centre of Brazil)
    +0.5 Cuiaba 27 71 264 606 857 (Close to Bolivia, about 1500 km from the Atlantic Ocean)
    +0.5 Manaus 90 206 604 1159 1898 (At the Amazon river, about 1500 km from the Atlantic Ocean)
    +2.0 Curatiba 158 498 1310 2138 3320 (About 7(?) km from the Atlantic Ocean)
    +1.0 Salvador 403 847 1683 2644 3695 (At the Atlantic Ocean ->> more stable temp???)
    +2.5 Sao Paulo 2334 5494 12089 15948 19582 (About 7(?) km from the Atlantic Ocean)

    The largest cities has the largest temp increase.

  82. Magnus
    Posted Aug 18, 2007 at 2:21 PM | Permalink

    Did my “research” with

    and the bernie #51 url ( bernie #51


  83. Posted Aug 18, 2007 at 4:01 PM | Permalink

    Being bored I checked out the gistemp plots. What’s up with Scandinavia?
    Every thing around Turuk, Finland has a spike at 1990 then they all level off on average after a 2 degree C approximate spike. This is less that 0.5% of the earth’s surface, so it is meaningless of course. I was just wondering if I missed something between 1987 and 1990.

    South Carolina, BTW seems to be a refuge for those concerned with warming.

  84. Posted Aug 18, 2007 at 4:13 PM | Permalink


    Very nice work. In honor of your work I have published an article on Newsvine about “Cargo Cult” science that reprints Dr. Richard Feynman’s famous 1974 Caltech Commencement Speech.

  85. Posted Aug 18, 2007 at 5:07 PM | Permalink

    I have a question related to this entire subject of warming. If it has been answered in the past please point me to the right place as I have not posted in this forum before.

    I have extensively reviewed the arguments at and the core of the discussion seems to relate to the phenomenom of the “pressure broadening” of CO2 as an absorber. The other major effect is doppler broadening, (which be definition is depdendent upon temperature, which makes it a feedback not a forcing). I have my quantum mechanics related to pressure broadening (Collision broadening) and if you look at the argument, and the equations that govern how the emission line changes, they are both dependent upon temperature (and pressure). Also, the total energy does not change. The waveform of CO2 emission shifts from a purely gaussian distribution, to a Lorentzian distribution but there is NO NET GAIN OF ENERGY, except that input from the collsion itself, which of course is temperature dependent.

    Why has this not been investigated more (if it has then please point me to the papers and calcuations).

    My data comes from the textbook: The Quantum Theory of Light by Loudo, 1978, Oxford Press


    Steve- please repost on another thread.

  86. Ianric Ivarsson
    Posted Aug 18, 2007 at 5:09 PM | Permalink

    First I must to apologise for carrying the CO2 thread along,
    it doesn’t belong here.
    Secondly I want to express my sorrow concerning the academic debate.
    It should not be in media – neither the traditional, nor the Internet.
    But after my long experience in the academia,
    I find peer review is not sufficient.
    There is so much bad science published.
    I’m the co-author of one publication from witch
    I demanded my name stricken out.
    Thus I know how some of the “consensus 2500 scientists” feel.
    From my own personally biased point of view,
    the three most flawed fields of science today are:
    1. Gender
    2. Medicine
    3. Climate
    In that order.
    But I may be wrong.
    If I knew what I was doing, I wouldn’t be doing science.
    The climatology orthodoxy scares me.
    It makes me think of the Spanish inquisition.
    Mr. McIntyre is a heretic.
    He questions the dogma. I like that.
    Keep up the good work Steve.

    Posted Aug 18, 2007 at 5:55 PM | Permalink

    #122 Magnus…Sao Paolo is some 60 km from the Atlantic.
    Take a look with Google Earth and you can see why it is
    heating up… It perhaps still is Sweden’s “second”
    industrial city after Gothemburg. In twenty years or so
    Sao Paolo and Santos are one big urban area!?
    #128 captdallas …I suppose you mean “Turku” a former
    mostly Swedish city on the south western coast of Finland
    I think most of NW Europe was very warm 1989-90 Mean temp
    in Stockholm Feb 23 1990 was +9.6 degrees C That’s close
    to summer temps for us! But 1975 was warmer in Stockholm
    unadjusted and IMO should stay the warmest thanks to
    the heatwave first 10 days of August Our equvalent of
    POR-ESP-FRA-ITA-SCH-GER-UK-NED Heatwave of 2003! Aug 6-8
    35C all three, 8 consecutive days 30C or more. Extended
    Sirocco I call it! No Nordic Tropical nights(least 20.0C) at Bromma
    airport just outside the inner city though..(19C max Tmin)
    My hypothesis about more frequent heatwaves in S and central
    Europe is that increasing rainfalls in S Sahara is making this
    desert smaller and pushing the heat northwards towards Europe…
    But now Mr Lindstroem you’re far from Brazil..Good movie
    though… LOL PS. As it is about Brazil I don’t change my
    family name’s corrupted form…second LOL in this post!DS.

  88. Posted Aug 18, 2007 at 6:22 PM | Permalink

    Ref 89, yup, I type lisdexic at times, Turku. This is a danged odd jump at 1990.

  89. Larry
    Posted Aug 18, 2007 at 7:42 PM | Permalink

    87, Look at the sidebar, and click the “Lubos Motl” link. That’s a physics blog that gets into those issues in a fair amount of detail, though you’ll have to search for the appropriate thread.

  90. Magnus
    Posted Aug 18, 2007 at 9:04 PM | Permalink

    STAFFAN LINDSTRÖM: “Sao Paolo is some 60 km from the Atlantic.”

    You’re right of course, and I meant about 70 km (not 7 km). The same for Curatiba.
    In comment #27 Paul Graham posted links to google maps with arrows pointing out weather stations…

  91. MarkR
    Posted Aug 18, 2007 at 10:18 PM | Permalink

    #50 If NASA had a calculated degree of confidence in their version of the global temperature record, then surely that confidence level should be changed now.

    Firstly, shouldn’t the confidence level be different for different regions depending on the number of samples being taken for a given area. A uniform or average confidence level is meaningless.

    Secondly, as Mr Watt, SteveM and others are showing that the quality of the US data is suspect, and indeed also the Chinese data, shouldn’t a value for that element of doubt be included in the confidence level.

    Going back to basic statistical theory, isn’t the confidence level dependent on the number of samples, also dependent on the quality of the samples. eg If the confidence were 95% but depended on data that was only 70% accurate, then the true confidence level would be some function of 95% and 70%?

  92. doubleg
    Posted Aug 18, 2007 at 10:41 PM | Permalink

    Don’t know if this comment is appropriate here, but a non-climate scientist is wondering about the latency involved with CO2 and actual atmospheric warming. Just a dumb observation- driving from coastal CA inland today just north of the Vaca fire (burning since July 4). There is a HUGE plume of smoke in the Kern County area (can be seen very well from satellite). About 5 minutes before I entered the smoke area the temp was 91. The further I went into the cloud, the cooler it got, down to 79 within 30 min. As I left the smoky area towards the town of Taft (on the outskirts of the smoke at the time) the temp had climbed back to 90 (about an hour start to finish). So, obviously the smoke cover has a very dramatic effect on cooling/blocking incoming heat. There must also be a tremendous amount of CO2 release from the fire as well. Does soot/ash have a much larger role in cooling than CO2 in warming? Shouldn’t there be warming in that area or somewhere nearby with that release of CO2? If the lower level soot/ash zone prevents any effect from CO2, wouldn’t clouds do the same? Finally, is the net effect of CO2 release into the atmosphere immediate or what is the time frame?

  93. Papertiger
    Posted Aug 18, 2007 at 11:52 PM | Permalink

    Captdallas #85

    Being bored I checked out the gistemp plots. What’s up with Scandinavia?
    Every thing around Turuk, Finland has a spike at 1990 then they all level off on average after a 2 degree C approximate spike. This is less that 0.5% of the earth’s surface, so it is meaningless of course. I was just wondering if I missed something between 1987 and 1990

    Steve McKitrick has the answer for you. It has to do with station culling, in the wake of the collapse of the Soviet Union. Why it would matter in Finland?
    Maybe the climate guys were trying to level the playing field?

  94. Posted Aug 19, 2007 at 12:08 AM | Permalink


    If NASA had a calculated degree of confidence in their version of the global temperature record, then surely that confidence level should be changed now.

    Good luck finding one, thats the thing that bugged me about Hansen’s 2001 paper concerning the US temp record, no clear description of the uncertainty/confidence. His graphs have a few oddly spaced error bars at various intervals and his greater than 0.1C line is in the text. Can’t he just say that he believes its +/- 0.2 95% or something like that? Maybe I missed it somewhere, but it is not as easy to find as I think it should be.

  95. paul graham
    Posted Aug 19, 2007 at 12:18 AM | Permalink

    Papertiger# A though accorded to me on the station culling, why after the soviet union collapsed. If these temperature station also contained listening devices/radar their removal would make sense as cold war funds where cut back. However with this theory! and I may be breathing in some heady GG; you would expect more rural sites to be culled. This theory, and I admit it’s only a theory I have just seen Bourne Ultimatum, would create an increases Urban heat tend.

  96. Magnus
    Posted Aug 19, 2007 at 12:26 AM | Permalink

    The mathematical confidence level at 95 or 99 percent don’t refer to lack of relevance or potential problems in collecting data.

  97. Posted Aug 19, 2007 at 12:44 AM | Permalink

    93: Upon further reading, Gavin at RC claims the uncertainty of the global temp record is 0.1, which he says would be better than a regional temperature record. I don’t know if there is a citation for that, the comments are 400+ long.

  98. Papertiger
    Posted Aug 19, 2007 at 12:52 AM | Permalink

    Steve McKitrick has the answer for you. It has to do with station culling, in the wake of the collapse of the Soviet Union.

    That should be Ross McKitrick.

    My bad.

  99. Posted Aug 19, 2007 at 2:06 AM | Permalink

    I think you can easily find large areas with similar temperature pattern as USA in 20th Century. Those are e.g. Greeenland and Artcic, ironically same areas AGW should treat as a proof of GW. In both Greenland and Arctic temperatures were warmer in 30s and 40s than now. See and and

  100. hillrj
    Posted Aug 19, 2007 at 2:07 AM | Permalink

    re 51, 61 >> the 30’s a US phenomenon..
    If you look at the Australian data
    (select type: temperature, region: Australia)
    you will see an interesting bump in the 1930’s
    so, maybe it wasnt just a US phenomenon.
    The Aussie data shows a much more pronounced anomaly than the US in recent decades. I didnt check if the temps are “corrected” or “raw”.

    Puzzling is a century long regular decline in the max-min spread. Is that a UHI effect?

  101. Brooks Hurd
    Posted Aug 19, 2007 at 6:51 AM | Permalink

    Re: 93 & 96

    The propagation of errors seems to be a course missed by many of the people who Steve refers to as “the team.”

    If your data has errors, “adjusting” it does not remove the errors so that they may be ignored.

    If your data has an error of x over a day, it does not that your error per minute is x/720.

  102. Brooks Hurd
    Posted Aug 19, 2007 at 7:13 AM | Permalink

    Re: 102

    Puzzling is a century long regular decline in the max-min spread. Is that a UHI effect?

    I noticed this in Parker’s Tmin and Tmax plots which he sent me. Tmin is increasing faster in his data than is Tmax, leading to a reduction in diurnal temperature variation. I have not read a good explanation for this phenomenon.

    As whether it is a UHI effect or not would require comparing Tmin and Tmax at stations which are truly rural with those effected by varying degrees of UHI. The issue which Watts and Steve have been raising is that many stations which we thought were rural are being effected by manmade heat sources.

  103. bernie
    Posted Aug 19, 2007 at 7:17 AM | Permalink

    Wouldn’t increased cloud cover at night do the trick – but then do we have the records to show this?

  104. bernie
    Posted Aug 19, 2007 at 7:59 AM | Permalink

    The S. American issue seems to have been picked up elsewhere, in part prompted by this article
    Alas I do not speak the language. Can someone paraphrase the main argument?
    There is an ongoing discussion on this UK site.

  105. Erik
    Posted Aug 19, 2007 at 8:20 AM | Permalink

    I do not understand the data presented in the IPPC report at

    Click to access AR4WG1_Pub_SPM-v2.pdf

    Since the number of reporting stations has increased dramatically why has the data not tightened? There is the same noise in 1900 on the data as today. How is this possible? Based on the thermal mass of the system (the earth) I do not see how there can be this much variability in the data. Does the energy from the sun change that much?

  106. Robert Wood
    Posted Aug 19, 2007 at 8:25 AM | Permalink

    #9 I notice fig 1 has different y-axes scales for US and world, which helps reduce the “significane” of the US numbers.

  107. steven mosher
    Posted Aug 19, 2007 at 8:45 AM | Permalink

    RE 9 I brught this to Gavins attention on RC and he says
    the “noise is different” so the axis are different

  108. Larry
    Posted Aug 19, 2007 at 8:45 AM | Permalink

    109, others noticed that, too. Which raises the question, is it sloppiness, or is it a deliberate attempt to deceive? Neither one makes me feel warm and fuzzy about NASA, but it’s like some of the media errors that we’ve been witnessing over the past 10 years; there’s an odd tendency for their errors to always end up going one way. They never err in a way that undercuts AGW.

  109. steven mosher
    Posted Aug 19, 2007 at 8:48 AM | Permalink

    RE 107.

    See my link to a brazlian page. there are more than 6 brazlian sites

  110. Larry
    Posted Aug 19, 2007 at 8:49 AM | Permalink

    110, more specifically, they fit the scales on both to the mins and the maxes. That’s fine for a single graph, but it’s deceptive to do that and then put them side-by-side. I have a hard time believing that they don’t see the problem with that.

  111. Larry
    Posted Aug 19, 2007 at 8:52 AM | Permalink

    And why would the “noise” be so different between the US and the ROW? More specifically, why would the US data have a lot more “noise” in it? Are we doing something wrong, or is the ROW data cherry-picked?

  112. steven mosher
    Posted Aug 19, 2007 at 8:54 AM | Permalink

    Here is a link I thught I put up .. has a little info on brazlian data

    Click to access Garcia.pdf

  113. Robert Wood
    Posted Aug 19, 2007 at 9:16 AM | Permalink

    #61 Bob Meyer, you have neglected option C. That the US re3cords are valid and the ROWs aren’t.

  114. steven mosher
    Posted Aug 19, 2007 at 9:37 AM | Permalink

    The OTHER thing is that he is comparing a LAND OCEAN index.. to a LAND Record

    So, the question is what does a land versus land comparision look like.

    So US LAND record versus Global Land. The accuracy issue is the land record nt SST

  115. Howard
    Posted Aug 19, 2007 at 10:24 AM | Permalink

    Quixeramobim shows warming from ~1929-`1959 with cooling to about 1977 with latest warm period about 0.5-deg C cooler.
    Not wildly different from rural US stations indicating there is a global signal buried in the noise.

    If there is no significant temperature increases; yet if, for the sake of argument, global climate change is increasing in it’s extremes the cause is likely other than CO2.

    The current push for carbon credits is beginning to look like radical surgery with chemo and radiation as a cure for heart disease.

  116. aurbo
    Posted Aug 19, 2007 at 10:52 AM | Permalink

    Re #104;

    There are two reasons for the narrowing of the diurnal range. One has to do with the actual temperature, the other with the instrumentation.

    Quayle, Karl et al in 1991 did a study of the differences between the liquid in glass readings from the Stepehenson Screen, (AKA cotton region shelter)and the electronic MMTS (Maximum Minimum Temperature Sensor) equipment which replaced the liquid in glass thermometry circa 1990. There study showed that the MMTS equipment tended to produce lower maxes (-0.4°C) and higher minima by about (+0.3°C). This change alone reduced the diurnal difference by a significant (0.7°C). But since the differences in maxima and minima nearly balance each other out in the daily means, the changes were considered acceptable.

    The other more significant difference comes from the nature of rural vs urban temperature environments. Most properly located rural stations tend to be “good radiators” that is that the nature of their siting allows for efficient surface cooling under clear skies and calm or nearly calm conditions at night so that their minimum temperatures can be as much as 10°F or more cooler than their urban neighbors. This is partially a result of the effectiveness of radiational cooling at the ground surface under the instrumentation and the lack of advection of warmer temps either laterally from nearby warmer areas or vertically by convection since the cold air forming at the ground from radiationaal cooling and in the airmass immediately above by conduction is quite stable. Under typical radiating conditions there is a steep temperature gradient from the surface to about 10 meters so that the precise height of the temperature sensors is significant. The standard height for thermometers in Stephenson Screen shelters was 4.5 feet or about 1.5 meters. (Photos on Watts’ website of some MMTS sensors adjacent to innactive Stepehenson Screens shows height differences which could create significant differences in temperature under ideal radiational cooling conditions.)

    The reason that maximum temperatures are less affected is that heating of the air in contact with the ground crerates a super-adiabatic lapse rate which creates auto-convective overturning of the unheated air above and the result is a well mixed areas where temperatures are essentially uniform up to the boundary layer. This means that a considerably greater volume of air is heated during the daytime hours than is cooled under nighttime radtiaonal conditions. Thus, maximum temperature observations reflect a greater volume of air than the compressed layers of cool air under nighttime radiational cooling. During windy conditions creating a well mixed deep layer of air near the surface, all bets are off.

    BTW, air is a lousy conductor of heat and a poor recipient or emitter of thermal radiation. The temperature of of the lower atmosphere over land areas is controlled principally by conduction with the solid surfaces of the ground, and to a usually much lesser extent by solid particulates suspended in the airmass itself. The amount of temperature increase from solar or surface radiation passing through the air is neglibile. Things are slightly different over water as heat is transferred to the air not simply by conduction, but by the introduction of latent heat from water vapor evaporated from the water surface. At the same temperature and pressure, water vapor is much lighter, hence more bouyant, than air and has no problem rising through the lower atmosphere carrying its latent heat upward.

    I think this answers the #104,s question. Perhaps when people think more deeply about the physics and the T-word ending in dynamics, one might wonder how much the radiative properties of CO2 has to do with climate.

  117. aurbo
    Posted Aug 19, 2007 at 10:57 AM | Permalink

    Re my #119:

    The reference should have been primarily to #102 and only periperally to #104.

    My goof.

  118. Dano
    Posted Aug 19, 2007 at 11:22 AM | Permalink

    Since we are a little off topic, there is an interesting post in John A, I Quit post about peer-review. I think it is #101.

  119. BarryW
    Posted Aug 19, 2007 at 11:29 AM | Permalink

    Re #119

    With respect to the T word, there is a site that Steve and John A have set up:

    Auditblogs has been created for people who want to examine issues in science, math and economics in general, not just climate science (although that should be a popular subject). So if you want to go through the Stern Review or the IPCC 4AR line-by-line, or make comment upon the latest scientific announcements or synthesize the latest news in solar physics, then this is the resource for you.

    John A :: Dec.09.2006

    Why don’t one or two of the regular posters on the physics of atmosphere set up a blog or two over there where this discussion could be delved into in more depth? I for one would be interested in that sort of discussion.

    Steve, could you put a reference to the Auditblogs site and some information on this site (in the sidebar maybe) so new people would be aware that this is available?

  120. DocMartyn
    Posted Aug 19, 2007 at 11:48 AM | Permalink

    “BTW, air is a lousy conductor of heat and a poor recipient or emitter of thermal radiation. The temperature of of the lower atmosphere over land areas is controlled principally by conduction with the solid surfaces of the ground, and to a usually much lesser extent by solid particulates suspended in the airmass itself. The amount of temperature increase from solar or surface radiation passing through the air is neglibile. Things are slightly different over water as heat is transferred to the air not simply by conduction, but by the introduction of latent heat from water vapor evaporated from the water surface. At the same temperature and pressure, water vapor is much lighter, hence more bouyant, than air and has no problem rising through the lower atmosphere carrying its latent heat upward.”

    What would be the difference in the max/min temperature of a sensor in a proper shelter on
    1) a dry lawn
    2) a watered lawn.

    Water usage has changed hugely in the last few decades. I am sure that we have more water present in our green spaces (like parks) than ever before. My guess is that Tmin would jump up and Tmax would decline a little. However, I don’t know what a sensor would read.

  121. Magnus
    Posted Aug 19, 2007 at 11:58 AM | Permalink

    Erik #108: “Since the number of reporting stations has increased dramatically why has the data not tightened? There is the same noise in 1900 on the data as today.”

    I don’t know which data set in that document you refer to. I’m an amateur myself, but “noise” measured e.g. as highest and lowest values (“data spread”) should not be fewer if you have more stations. Rather the opposite, don’t you think? (The spread relatively constant.) But you shall of course get a somewhat smaller confidence intervall.

    By the way I’m not sure that the number of stations has constantly increased. Due to McKitrick (se link in comment #95) the number of stations decreased 1990, while the temperatur instantly shifted upward! This (if it’s true) clearly indicates a bias due to the exclusion of stations. Here’s the link about that:

    Dr. Robert J Balling in chapter 3 of the book “Shattered Consensus” writes about this. I’ve not yet read (just started reading the book), but in the introduction Patrick J Michaels quote Balling; regarding “instrumental changes and adjustments … ‘… all biases are positive [in the warm direction] implies that the indicated [global temperature] rise of 0.7 C is nonetheless an overestimate.'” (Brackets not added by me, but in Michaels quotation.)

    So what we have since 1970 is (officially) a surface temperature increase of 0.4 C (GISS data), an 0.2 C increase in the troposphere according to satelite data (since 1979), but almost no increase in the balloon data temperature record at 5000 to 30000 feet height (since satellite measure started well following satellite troposphere temperature record at the same height, which indicating accuracy [1]). A balloon data correction by Mears and Wentz a few years ago changed the temperature increase from the balloon data record with only 0.09 C/decade [1].

    So a fundamental thing in the theory of greenhouse global warming isn’t there. How does heat in the atmosphere reflecting down to the earth’s surface?

    The problem discussed in this thread is the relevance of urban station data. The GISS temperature increase of 0,4 C do *not* correspond to +0.2 C from satellite data and seems — due to McIntyres collected weather station data — have an urban heat island bias …oh!: …as well as possible biases in data and instrumental adjustments and also weather station selection. Hm… Am I a denial? :-/


  122. Jerry
    Posted Aug 19, 2007 at 12:00 PM | Permalink


    The Brazil data is mind-boggling. My sister-in-law was born there and she said the urban growth in her lifetime has been incredible and out of control. Same with a colleague who worked there in the 70’s.

    Is it true that there are only 3,000 surface stations being used by the IPCC? If so, the U.S., with only 6% of the world land mass, accounts for 40% of the weather stations. This would give me a lot more confidence in the US records rather than records from the ROW. At the least, if we cull out the UHI-vulnerable sites, our data base will be on a superior position with the ROW, even assuming the others are purely rural. For the U.S., it would be a move towards quality of data. Maybe not so with the ROW.

    U of Delaware website showing the evolution over time of the number of surface stations is a shocker. Question is – did the number of weather stations get reduced by the IPCC in attempt to enforce some quality control (i.e., culling out the urban stations)? I notice that Canada and Australia also showed a collapse in the number of surface stations around 1990, so I would believe that the culling factor might apply here. However, with the collapse of the Soviet Union (geographically, economically and demographically), did the IPCC try to salvage a rapid dropoff in rural databases due to inattention by the monitors in Russia? It seems we’re going to be left with massive ROW data gaps in any grid system selected. Given the collapse in the internal geographic Russian (and Chinese) station distribution, I’m guessing the problem is local volunteers abandoning the interior rural stations and not quality control. Not good for the accuracy of the database.

    As for the Chinese, we have an authoritarian political class trying to that diverts political discontent by riding a Wild West economic quasi-capitalist machine into earning a renmibi coupled with a rapacious destruction of the environment – both within China and transported internationally through its wind-driven “brown cloud” pollution. China is now the biggest CO2 emitter in the world. If the surface station monitors are government-paid, do you think that maybe, just maybe, there will be pressure on the monitors to down-play the temperature readings by, say, a tenth of a degree, especially in the urban centers? If Exxon was in this position of power, the Left would justifiably scream about a predatory conflict of interest and throw out any Exxon-tainted data. Just a thought.

    On another thread, Steve previously connected us to the Australian station photo and data survey. I remember that A) the general designs appeared to be uniformly pretty decent, B) they did show an increase in temperature over time, and C) they appeared to be located near airports. Looking at the latest BOM data, using an 11 year moving average, an increase in temperature is apparent. But I think if a 5 year moving average was used, the trend line would have collapsed last year (though they seem to be having a high-temperature year in 2007).

    Wolfgange Flamme,

    Thank you for your connection to the German station photos. Scrolling through them shows what appear to be excellent “rural” station designs. But it also shows a lot of ridiculous roof-mounted and locally corrupted installations. Anyone know of which stations were used by the IPCC?

  123. Magnus
    Posted Aug 19, 2007 at 12:11 PM | Permalink

    BTW: Equally 0.2 C increase in the troposphere and at the earth’s surface since 1970 can of course (but must not) be partly caused by a human greenhouse effect. (Data makes me more convinced that sun activity- and cosmic rays, in the theory presented in “The Chilling Stars”, by Svensmark/Calder, has …an important role in climate change.)

  124. Posted Aug 19, 2007 at 12:32 PM | Permalink

    “an important role in climate change”I’m thinkin’ that we need to change the name (again) to: “Climate Diversity”. That way we can celebrate and embrace it knowing that is makes us stronger.

  125. Posted Aug 19, 2007 at 12:36 PM | Permalink

    “an important role in climate change”
    I’m thinkin’ that we need to change the name (again) to: “Climate Diversity”. That way we can celebrate and embrace it knowing that it makes us stronger. (Attempting a reformatting)

  126. John F. Pittman
    Posted Aug 19, 2007 at 12:44 PM | Permalink

    Jerry #125

    Why would the Chinese even care. Their impetusus is for economic growth. The IPCC has a section that if you run the numbers indicates, despite their Summary, that the best way to have a leg up on climate change is to be able to afford it. I posted this somewhere on CA, but can’t remember where. It had the numbers.

  127. Posted Aug 19, 2007 at 12:46 PM | Permalink

    This issue with Brazil temperature stations is common to many parts of the world. Southern Africa shows similar differences between rural and urban (see the regional summary at – lower left under regional summaries). The regional summary for South America shows a cooling trend for the west coast. South America shows greater warming in the equatorial parts than in the southern part of the continent – not consistent with the CO2 theory.

  128. Jerry
    Posted Aug 19, 2007 at 1:15 PM | Permalink

    [snip -offtopic]

  129. Erik
    Posted Aug 19, 2007 at 2:33 PM | Permalink


    Increasing the number of samples MUST decrease the noise. You can see this on any scope. If you crank up the number of averages the noise reduces. If you do not see this either the number of samples has not changed or there is something wrong with the data. On a scope if you do not see the noise reduce you do not know how to use the scope.

    If the number samples has not changed. I would question if these are biased towards population centers.

  130. bernie
    Posted Aug 19, 2007 at 2:44 PM | Permalink

    Nice site and useful summary of South American data. I have a feeling though that there are actually few if any real rural stations at least in Brazil. I am just compiling that lsit now. For example, Quixeramobin is in fact Urban assuming the NOAA classification scheme with a population > 50000. Rural is defined as

  131. aurbo
    Posted Aug 19, 2007 at 2:54 PM | Permalink

    Re #123:

    A watered lawn makes it slightly more complicated, but not because of radiation, but because of thermal mass and conductivity. Water is a reasonably good radiator, but has a high thermal mass compared to dry grass on dry soil. The dry grass and soil also have relatively low thermal conductivity. Thus, radiative cooling over these elements is effective in lowering the surface temperature because not much heat energy needs to be removed. Water with its higher thermal mass and also its ability to make the top few inches of the soil more thermally conductive means that more heat energy has to be removed through radiation for the surface air to reach the same temperature as over dry soil. The effect of water vapor, created by evaporation, in carrying heat away, will serve to narrow this difference, but the rate of evaporation is dependent on the actual temperature of the water so its importance can vary considerably.

  132. Magnus
    Posted Aug 19, 2007 at 3:30 PM | Permalink

    #133 Erik: If a sampel consists of a number of “measurment points”/raw data these samples becomes more homogeneous, yes. And it also results in a more limited range for a confidence interval. But individual stations added on the same area will have about the same distribution as the original stations. That’s what I mean. The range “stationwise” will be approxemately the same.

    Did you see the McKintrick document which shows a decrease in number of stations 1990? Does that make sence to the data you see (I guess you mean the graph on page 6). But more stations not necessarily reduce the interval that much. To get half the accuracy one often need …maybe a 5-10 times increase of data (depending on original size; I guess it was a large increase in the beginning of the 20th century). Similarly to reduce data by half mustn’t reduce quality that much. The UHI effect since 1950 may results in disparity in the data?!? I really don’t know… (…what these dots are either).

    (The definition of avarage temperature is another interresting subject. If we cut the forrests in Scandinavia the changed albedo makes winters -30 degrees C instead of -20 degrees C, but to what extent will more extreme winters save anyone anywhere? 😛 )

  133. Jerry
    Posted Aug 19, 2007 at 5:19 PM | Permalink


    I didn’t think it was offtopic, but I didn’t realize I strayed off the theme of the thread. My apologies.


  134. Erik
    Posted Aug 19, 2007 at 5:31 PM | Permalink

    The point is that the more samples act as a low pass filter. High frequencies (noise) stuff should not be able to appear on the chart. In statistics this would be more homogeneous. The underlying math is the same.

    I do see that there is something funky going on with then number of stations around 1990. But the amount of noise on the chart I referenced from the IPPC is constant. Why do I not see changes in the amount of noise on the chart that correlate with the number of stations. There is no change around 1990.

    The forest where I grew up (Minnesota) is about the same as yours in Scandinavia. It seems that swedes are drawn to these place. I am not sure why. It is really cold.

  135. steven mosher
    Posted Aug 19, 2007 at 5:43 PM | Permalink

    RE 129.

    If I were a Chicom. I would do TWO things.

    1. Spew as much C02 as possible and call upon the ROW to do something!
    a. fund leftnutjobs, etc etc etc
    2. Bias my land recrd recordings up as much as possible.

    In short. Spew C02 and ask others to cut emmissions.
    Bias my reporting up such that clitmatolgists ( opps) get excited and upset.

  136. Sandy McIntyre
    Posted Aug 19, 2007 at 8:17 PM | Permalink

    Why would any rational individual not adjust for changing temperature as land use changes? The Dunlop Observatory at University of Toronto was the primary source of temperature readings in Toronto up until the 1950s when it moved to Aurora in the suburbs of Toronto. When the observatory was bult in the early 20th century the location was undeveloped. By the 1950s this location was in a highly congested portion of the city and was relocated to farm land in a green belt north of the city. The current location is now in an area experiencing rapid increases in density. This issue was examined in a Scientific American article in the 1990s examining historical temperature readings as recording sites “migrated” into urban centers. The same article examined over 200 years of British Admiralty readings of sea temperatures.

    Steve: Hi, Sandy. To others: in Canada right now, I’d be much more likely to be identified as Sandy’s older brother.

  137. Stephen Mattison
    Posted Aug 19, 2007 at 8:19 PM | Permalink

    Linked to your blog from a newspaper article. I’ve always said that global temperature data needed a statistical audit. Now you have gone out and given one! Excellent job. I would be curious to know how your donations spike with the surge of publicity you are sure to receive.

  138. Frank Ewing
    Posted Aug 19, 2007 at 9:21 PM | Permalink

    How many of you have begun to let your legislative representatives know that we know there are huge flaws in climate change research? The money for the vast majority of the climate change research is government funded i.e. YOU are paying for it out of your own pockets. Make the phone call; write the letter; send the email. Keep your elected representatives aware of what this site and others are bringing to light.

  139. bernie
    Posted Aug 19, 2007 at 9:51 PM | Permalink

    I finished tabulating 47 GHCN/GISS stations (including only those GHCN stations in the GISS calculations, i.e, 47 of 57). The picture leaves a number of open questions. First, there have been cut-offs established for designating a station as Rural, Suburban and Urban. 10000 but 50000 is Urban. There are apparently 47 GISS stations. For GISS stations, 16 of these 47 as designated Rural, 5 as Suburban and 26 as Urban. However, it is very unclear as to when this designation was made. I haven’t checked but it looks like prior to 1990(!!). When we use 2006 estimated data from Brazilian Census the picture changes dramatically. Using 2006 population data there is 1 Rural station, 12 Suburban stations and 34 Urban stations. Of the 16 previously designated Rural stations 11 should now be classified as Suburban (Average size 32000) and 4 should be classified as Urban (Average size 78000). Of the 5 previously designate Suburban stations 1 remains Suburban (Size 47000) and 4 are now Urban (Average size 75000+).

    Moreover the single rural stations is apparently on an island far off the coast of Brazil!

    Finally, the rate of growth of these locations is truly amazing. Compared to the officially recorded population figure, 70% or 33 have doubled in size and 51% or 24 have tripled in size. Clearly the impact of these growth rates on UHI or microsite factors could be significant. More telling is why the population data is so out of date. What happens to temperature trend data if the original designations are amended to reflect actual population numbers? Are adjustments themselves modified to reflect population growth?

  140. VG
    Posted Aug 20, 2007 at 12:51 AM | Permalink

    What has happened to Satellite/UAH July 2007 data. None of the sites (even Nasa GISS) seem to have updated July 2007 data. Quite keen to see surface/troposphere data re cold temps Australia, South America, Tropics, Oceans etc. Does anybody know?

  141. Derek Tipp
    Posted Aug 20, 2007 at 2:44 AM | Permalink

    The clip below is from the British Met Office website. Unfortunately the graph didn’t come out, but to view the whole thing go to I would be most inerested to hear any arguments against this conclusion.

    “Isn’t the apparent warming due to urbanisation?
    No. The below image shows the results of a recent analysis at the Hadley Centre of temperature trends over the last 50 years deduced separately from measurements on the most windy nights and the least windy nights, from the same stations. If urbanisation were causing the observed warming, one would expect calmer nights to have warmed more, as it is in these conditions that the heat island effect would act to warm the station compared to its rural surroundings. In fact, the results showed no difference between trends on windier and calmer nights, confirming that urbanisation is not to blame. “

  142. Derek Tipp
    Posted Aug 20, 2007 at 2:50 AM | Permalink

    Below is an extract from the UK Met Office – any comments please?

    “Isn’t the apparent warming due to urbanisation?
    No. The below image shows the results of a recent analysis at the Hadley Centre of temperature trends over the last 50 years deduced separately from measurements on the most windy nights and the least windy nights, from the same stations. If urbanisation were causing the observed warming, one would expect calmer nights to have warmed more, as it is in these conditions that the heat island effect would act to warm the station compared to its rural surroundings. In fact, the results showed no difference between trends on windier and calmer nights, confirming that urbanisation is not to blame. “

  143. Posted Aug 20, 2007 at 3:07 AM | Permalink

    If urbanisation were causing the observed warming, one would expect calmer nights to have warmed more,

    Of course “one would expect” is quite different than “It has been shown by ….”.


  144. Thomas Bolger
    Posted Aug 20, 2007 at 3:40 AM | Permalink

    Re 145
    I have measured temperatures under many conditions and have come to the conclusion that rural cooling is caused by the emission of water from plants.
    I have measured temperatures on rainy days and found that the difference between “rural” and “urban” is minimised. However I have seen no such reduction of temperature range on windy days
    Has anyone investigated the difference in warming for rainy conditions vs dry conditions?

  145. Derek Tipp
    Posted Aug 20, 2007 at 3:57 AM | Permalink

    Sorry for the double post – but after the first one I got a message saying it had been rejected

  146. Sean Houlihane
    Posted Aug 20, 2007 at 4:56 AM | Permalink

    Back to Brazil… Is there anything useful that can be done with these data? The number of sites is small enough to make some additional measurement possible, but how is it possible to critisize any adjustments? Steve says we have one good station, they say they have 5 that agree. Is there any more good data hidden anywhere?

  147. Vernon
    Posted Aug 20, 2007 at 5:04 AM | Permalink

    RE:145 I have failed to understand why alarmist climatologists are against doing the simple work of siting stations in an urban environment and then then siting stations is adjacent rural environments, then actually measure the differences. But no, that is to easy, it is better to take stations that no one has checked to see if they are in compliance, are actually urban or rural, and then make up windy studies, rainy studies, when they could just do a real scientific test that actually does address UHI.

    But then I guess if the urban stations were identified, the warming trend might not be a trend.

  148. John F. Pittman
    Posted Aug 20, 2007 at 5:05 AM | Permalink

    #147 Look at 134.

    I think you are right. The rain would have to be quantified, as in small versus large in mm of rainfall. You would expect to see the rural drop with light, but the urban to only drop with heavy rains.

  149. MarkW
    Posted Aug 20, 2007 at 5:25 AM | Permalink

    The US is 2% of the world’s surface, but you have to remember that around 70% is virtually unmeasured.

  150. Posted Aug 20, 2007 at 5:42 AM | Permalink


    Might not the diminution of the temperature difference between rural and urban in rainy conditions be due to the rain cooling the hard infrastructure which is causing the UHI? Latent heat of evaporation would make the rain very efficient at cooling tarmac and buildings.

    Consequently, wouldn’t it be more appropriate to compare temperatures in dry conditions and make an adjustment for plant transpiration in the rural setting?

    You would have to be alert to micro conditions e.g. the urban forest versus the urban concrete jungle. For example, a sensor located in Central Park would be affected by transpiration just as much as a truly rural site.

  151. bernie
    Posted Aug 20, 2007 at 5:57 AM | Permalink

    There are 47 locations in Brazil that are used as part of the GISS assessment. The quality of those sites is unclear. There have been up to 5 changes in site for some of these locations. This is not surprising given the tremendous growth rate in Brazil – but it means that the data has to be very carefully reconstructed in order to minimize micro-site effects.

    I think the questions regarding the quality of the Brazil data can be applied to many other countries.

  152. Antero
    Posted Aug 20, 2007 at 5:57 AM | Permalink

    In Finland there was really an heatwave in 1930’s and i found info that in Britain too. Even before 1970’s they estimated that UHI was playing some 1 C role in Helsinki, Kaisaniemi station. As i’ve written before Mariehamn has been warmer than Helsinki until 1950’s. Nowadays Helsinki is some 0.5 C warmer than Mariehamn or Jomala in island of Åland.

  153. Posted Aug 20, 2007 at 6:29 AM | Permalink

    Re: 54 and 62

    The Central England temperature index shows a warming in the 1930s.

    Chart at

    Generally speaking sustained hot temperatures for North America point to a more northerly track of the jet stream, with little lateral variation and strong precession. Because the jet streams are circumpolar, these conditions would also apply to Eurasia.

    And the Central England series bears this out.

  154. Posted Aug 20, 2007 at 7:04 AM | Permalink


    There was a loss of data for 10 days in July at UAH. They are trying to reconstruct data now and will published it as soon as possible. In short, this is what John Christy answered me.

  155. MopTop
    Posted Aug 20, 2007 at 7:31 AM | Permalink

    Conservatives keep claiming that Hansen is being secretive about his “algorithms,” although the methods for NASA’s analysis are all laid out in painstaking detail


  156. Brooks Hurd
    Posted Aug 20, 2007 at 7:49 AM | Permalink


    Those are analyses of temperature change, not algorithms for adjusting temperature data.

  157. Frank K.
    Posted Aug 20, 2007 at 10:07 AM | Permalink

    Re: 158, 159

    Moreover, a ** description ** of the process (which indeed can be found in the papers cited as well as at several NASA/NOAA web sites) is NOT the same as the actual code or computer algorithm itself. Anyone who has tried to replicate numerical results generated by computational fluid dynamics algorithms (which is my professional background) can atest to this! There are usually many undocumented corrections and additions which reflect bug fixes and enhancements that have been implemented at various stages of the code’s life, and accordingly are not reflected in the original papers. This is why we would like to see the NASA/NOAA source code used to generate the “corrected” temperature data. I’ve worked with many government codes in the past, and since there is nothing proprietary about weather/climate algorithms, I fail to see why they are so reluctant to release the source code.

  158. windansea
    Posted Aug 20, 2007 at 10:17 AM | Permalink

    I asked this question at RealClimate but no answer so far

    US land mass is 2% of total earth surface, and 8% of total dry land on earth right?

    How many temp stations are in the world used by GISS? What percentage are US stations of the total?

    How many temp guages are in the oceans?

  159. Tim Ball
    Posted Aug 20, 2007 at 10:25 AM | Permalink

    1. The decline in surface weather stations in 1990 was part of a general policy that assumed satellite data would make them redundant, a pattern begun in the late 1960s. GISS analysis of the stations they use showed a decline in numbers, but also a decline in the area covered. Fully 20% of the Southern and 10% of the Northern Hemispere lack coverage. Satellite data was rejected for many years for not being a long enough, statistically significant, record. In addition, many shortcomings of early satellite data were exposed. I suspect the 1990 anomaly was a part of the extensive introduction of AWOS surface stations.

    2. Attempts to show the difference betweeen rural and urban climates is most remarkably underscored by the significant difference in temperatures of even a park one block square within the city. Oke showed this in Hamilton and Vancouver, while we found a similar effect in Winnipeg. The urban heat island efect is most dramatic in colder regions.

  160. bernie
    Posted Aug 20, 2007 at 10:29 AM | Permalink

    #161: Windansea
    These are readily answered by downloading the lists from here
    However, the more difficultquestion to answer is the number of high quality sites. The rural designation I think is very suspect.

  161. Steve McIntyre
    Posted Aug 20, 2007 at 10:35 AM | Permalink

    #162. there’s an issue with the GHCN collection of surface station data that’s not been fully explore. Many of the GHCN records that end around 1990 do not end in 1990; it’s just that GHCN hasn’t collected them and since GHCN didn’t collect them, CRU and GISS didn’t use them. As a result the station population post 1990 has become more airport oriented with whatever biases that involves.

  162. windansea
    Posted Aug 20, 2007 at 10:46 AM | Permalink

    Thanks for the answers, I think I saw a post or comment here in another thread stating that 40% of the stations used by GISS are US based, just wanted to check this.

    PS In the latest thread at RealClimate I find Gavin’s repeated responses to questions about lack of transparency (of methods and formulas etc) Hansen’s modeling quite ludicrous.

    “it’s all in his papers”

    “do it yourself”

    “it’s too complicated for you to understand”

  163. Sam Urbinto
    Posted Aug 20, 2007 at 10:47 AM | Permalink

    There’s only a limited number of valid reasons to refuse to provide working code, and neither “you don’t need it” or “it’s more valuable to construct a method yourself and compare the results” are in that list of valid reasons.

  164. Demesure
    Posted Aug 20, 2007 at 10:47 AM | Permalink

    “I suspect the 1990 anomaly was a part of the extensive introduction of AWOS surface stations.”

    #162, Dr Tim Ball (if you are you).
    On the reverse, in France, all rural stations used by the GISS have their data ended in the 1990s, which is unexplicable (apart from the desire to reject no-warming trends) because high quality, long record rural stations like Mont Aigoual ou Pic du Midi are professionnaly manned by Meteo France and haven’t interrupted their records (I know them two for having visited them). Urban stations on the other side are mostly situated near airports, so if you want a “manmade” upward trend in temperatures, you can’t dream of better.

    That’s the situation for France, with about less than 40 stations for the GISS, “un autre cadavre dans le placard” ? I don’t have the brain of Steve to check out the junk but it looks like a fish and smells like a fish.

  165. BarryW
    Posted Aug 20, 2007 at 10:53 AM | Permalink

    Re # 164 & #162

    In the BBC thread a reference was made to a Hadley Centre graph showing a divergence between the land and sea records about that time. They’re showing about a half degree difference.

    Land-Sea Plot

    Does Hadley give an explanation?

  166. PaulM
    Posted Aug 20, 2007 at 11:22 AM | Permalink

    # 145 Derek Tipp. This has been discussed at great length already on other threads. Search for ‘Parker’ and ‘UHI’. Parker made a speculative assumption about wind that has no justification. The way to answer the question is simply to look at the data for urban and rural sites and compare them. The GW alarmists don’t do this, because they know it would give the ‘wrong’ answer. Just look at the graphs at the top that Steve has put up. Pretty spectacular. The extent of warming seems to be proportional to the size of the town / city! Has anyone checked this?

  167. steven mosher
    Posted Aug 20, 2007 at 11:30 AM | Permalink

    #158. Moptopper

    Hansen 1999:

    “All flagged data were graphically displayed along with neighboring stations that contained
    data during the period in question, and a subjective decision was made as to whether the apparent
    discontinuity was flawed data or a potentially real climate anomaly. The philosophy was that if the
    data were not quite obviously flawed, it was retained. Only a very small portion of the original data
    was deleted: approximately 20 station records were deleted entirely, in approximately 90 cases the
    early part of the record was deleted, in five cases a segment of 2-10 years was deleted from the
    record, and approximately 20 individual station months were deleted.”

    Gavin on RC has made this repeated claim that the algorithms are described in painstaking detail.
    Then, you read a passage like the one above.

    1. A subjective decision was made. Who made it, what criteria did they use?
    how was this decision checked for author bias?
    2. Which stations were deleted, which 90 stations had early records deleted?

    Bottom line you can’t even check the subjective assesment with out this.

    THEN we have this:
    “We also modified the records of two stations that had obvious discontinuities. These
    stations, St. Helena in the tropical Atlantic Ocean and Lihue, Kauai, in Hawaii are both located on
    islands with few if any neighbors, so they have a noticeable influence on analyzed regional
    temperature change. The St. Helena station, based on metadata provided with MCDW records, was
    moved from 604 m to 436 m elevation between August 1976 and September 1976. Therefore
    assuming a lapse rate of about 6°C/km, we added 1°C to the St. Helena temperatures before
    September 1976. Lihue had an apparent discontinuity in its temperature record around 1950. On
    the basis of minimization of the discrepancy with its few neighboring stations, we added 0.8°C to
    Lihue temperatures prior to 1950.”

    1. define obvious discontinuity. Your QA proceedures failed to detect the last
    input error, after all.
    2. The standard enviromental lapse rate is 6.5C per km. However sme have argued that it is closer
    to 6C/ km. Which ever value you select, hwever, there is another issue. Lapse rate varies
    with location and other factors. SO, did hansen add the adjustment DETERMINISTICALLY into
    the record, or did he add a lapse rate adjustment with an attendant estimation error. The
    first preserves variance ( falsely) the latter would maginify the variance due to additinal
    error inherent in the adjustment estimate.

    3.The proceedure in the Lihue case looks to be suspect. but not enough detail is given.

    MopTop, EVEN if, the process were adequately described in the paper one still needs to check
    the code to see if it conforms to the description

  168. bernie
    Posted Aug 20, 2007 at 11:36 AM | Permalink

    One type of correction for UHI is in fact primarily the size of the immediate population. However, there can be numerous confounding factors such as energy us. Sao Paulo for example stretches for miles and there are undoubtedly microclimates within this amazing city. What seems odd is that the approach to corrections appears to be extensive (lit vs unlit)rather than intensive (see the earlier cited papers in the recent Tuscon detective thread. It is a bit like understanding consumer behavior by simply tracking what they buy, without understanding why they buy what they buy.

  169. steven mosher
    Posted Aug 20, 2007 at 11:37 AM | Permalink

    RE 149.

    See my links above. Appears to be some more data. perhaps enough to call the
    6 stations into question

  170. Frank K.
    Posted Aug 20, 2007 at 11:59 AM | Permalink

    Re: 170

    There’s a word for this kind of data manipulation by NASA:


    All the more reason to publish the source code…

  171. windansea
    Posted Aug 20, 2007 at 12:46 PM | Permalink

    not sure if anyone here has ever seen this before but GISS NASA is located in same building in NYC that Seinfield show used for exterior shots of the corner restaurant where they always hung out.

    I was wondering what sort of forms GISS employees sign concerning the work they do, intellectual property they create etc

    Most private companies I have worked for require you to sign a form giving company the rights to anything you create or write using their resources.

  172. Michael Jankowski
    Posted Aug 20, 2007 at 1:03 PM | Permalink

    Re#174, that Seinfeld/GISS relationship has always been amusing to me. If I recall correctly, once you turn the corner down 112th, GISS is the first door you hit to (looks like there are the restaurant windows under the sign, then two more individual windows with a/c units, and then the entrance to GISS). FWIW, it was also the basis for the song “Tom’s Diner,” by Barnard (all-women’s school of Columbia) grad Suzanne Vega.

    I actually took a class at GISS one semester…I wonder how many name past, current, and future, climate folks, RC and otherwise, I might have bumped into there without realizing.

  173. Jeff
    Posted Aug 20, 2007 at 1:04 PM | Permalink

    I grew up in Northern VA, the suburbs of Washington DC. Whenever the local news gave temps it was ALWAYS, without fail, several degrees difference from Downtown DC as compared to the surrounding suburbs. Anecdotal, I realize, but should be easily quantifiable. Even the temp difference between Dulles Airport (in the NoVa suburbs) was different from DC.

  174. windansea
    Posted Aug 20, 2007 at 1:12 PM | Permalink

    well, found answer to questions about NASA software policy in the Townhall thread

  175. windansea
    Posted Aug 20, 2007 at 1:17 PM | Permalink

    RE #175

    a show about nothing rings a bell somewhere 🙂

  176. Steve Moore
    Posted Aug 20, 2007 at 1:57 PM | Permalink

    Is anyone sure that Terry Gilliam isn’t behind this?

  177. Ian McLeod
    Posted Aug 20, 2007 at 2:39 PM | Permalink

    Barry W #168

    The global temperature profile from the Hadley Centre does not show the hot years during the 1930s in the US, there is a blip, but that is all. I appreciate the difference between the US and the rest of the world (ROW). However, by what weighting is the US data compared to the ROW. If there are roughly 2000 locations in the US, how many are there in the ROW and do they comply with the US standards? Does the Hadley Centre explain this in their methodology anywhere?


  178. BarryW
    Posted Aug 20, 2007 at 3:29 PM | Permalink

    Re #180

    I didn’t follow your answer. My question is the difference between Hadley’s land temperatures and sea temperaturesm not between US and ROW. Somewhere around 1990 the anomalies diverge. Before that they closely track each other. Unless there is a good explanation for the divergence I would say this makes the land record more than a little suspect. According to Steve this is when the number of sites was significantly reduced.

  179. steven mosher
    Posted Aug 20, 2007 at 3:34 PM | Permalink

    In a thread labeled Brazil, I fully expect TCO to enter stage left and in a lame
    robert deNiro impression start expounding on HVAC.

    Best Movie ever… next to Repo Man and Fisher King.. and After Hours…

  180. Posted Aug 20, 2007 at 4:45 PM | Permalink

    Regarding Hadley vs GHCN, the following link compares the two sets of data for a rectangle approximating Brazil:

  181. Posted Aug 20, 2007 at 5:17 PM | Permalink

    Speaking of GISS and South American temperatures, Chile is having its coldest year in at least 4 decades. The other day it snowed in capital city Santiago for the 3rd time this winter. Only this time it was even harder than the previous ones and it covered basically the whole Central Valley, ruining avocado plantations and including many locations with no similar recorded events.

    Considering that snowfall in Santiago is more or less a one-in-20 years occurrence, this can surely be rated as a several-sigma episode. I guess it’s more or less like having the California Central Valley covered in snow…..3 times since December to early February.

    Neighboring countries like Peru, Argentina, Paraguay, Uruguay are also having an unusually cold winter. But GISS has just informed us in its famous July report that the Southern Hemisphere has experienced a +0.38 positive anomaly (??).

    Having been checking the (equally eccentric) NOAA anomaly website quite regularly , I don’t know where they get that warning signal from.

    As far as I know, neither Australia-NZ nor Southern Africa are having any warm winter at all. The Antarctic sea-ice extent has remained over the average for several months in a row and the Chilean met office attributes the abnormal weather to unusually cold water in the Pacific. The only SH regions with apparently less frigid temperatures than normal have been some locations in the interior of Antarctica but, considering the lack of coverage in that area, there must be a lot of gross extrapolation, not to mention that those areas are too small to offset the general hemispheric trend (much smaller than they appear in the above link due to the Mercator projection).

    Nothing new, really. As usual, GISS temperature reports look more than suspicious. How can the one person in the world whose career depends on GW becoming a reality be also the one who does the calculations to assess how much GW we’re observing? And how can he also be allowed to keep his software secret?

  182. Posted Aug 20, 2007 at 5:36 PM | Permalink

    Similar to Brazil, the Rio de la Plata estuary in Argentina-Uruguay shows a remarkable example of UHI-contaminated data. Buenos Aires and Colonia share basically the same latitude, longitude and altitude. They’re just opposite each other across the estuary so there can’t be any long-term natural climatic trend affecting one but not the other.

    However, BA has expanded tremendously in the last century to become one of the largest urban areas of the world. Its temperature history, unsurprisingly, shows a dramatic warming:
    Colonia, on the other hand, has remained a small town and shows, if anything, a cooling trend:

    In fairness to the GISS records, it must be said that eastern Argentina is one of the few regions in the world largely unaffected by GW during the past century, so their secret code must have somehow done the trick here, but a quick look at all the stations of the area shows that very few are rural and the only ones with a decent, uninterrupted temporal record are in urban areas:

    UHI just 0.006°C per decade (IPCC)??

  183. CWells
    Posted Aug 20, 2007 at 5:44 PM | Permalink

    Did anyone notice that TNR had an online article (8-13) on The Plank by Bradford Plumer? Plumer backhandedly calls Steve M a liar; he includes a link in pargraph four which goes to the NASA publications page for Hansen. Plumer aludes that supposedly the ‘algorithms’ for Hansen’s adjustments are included in one of his listed and pdf accessible articles. Damned if I could find anything beyond Hansen’s SD averaging in Hansen et al, ’98.
    I assume Steve has previously perused the articles, but could someone confirm for me that the formulae aren’t there? Did Plumer the journalist blatantly LIE!!!

  184. Posted Aug 20, 2007 at 6:25 PM | Permalink

    Re: #185 – Often the GISS temperature graphs come from multiple close proximity stations. If you change the parameters you can get different views. For example, changing the data_set=1 to data_set=0 will display the multiple stations usng different line types. Also changing the num_neighbors to 2 can give you a comparative plot. In the Colonia example, this will plot the Colonia and Buenos Aires together:

  185. Posted Aug 20, 2007 at 8:57 PM | Permalink

    CWells : TNR.. do you mean The New Republic, the magazine recently embroiled in the second or third scandal where it published reports which were fabricated and they didn’t bother to check? Sounds like a real quality publication…

  186. bernie
    Posted Aug 21, 2007 at 5:41 AM | Permalink

    Do you speak Spanish and/or Portuguese? If so, it would be great if you could scan the Brazilian sites and check to what data they are focusing on. There are over 300 weather stations in Brazil, but only 47 made the GISS list.

  187. Don Keiller
    Posted Aug 21, 2007 at 6:22 AM | Permalink

    re #185 & #187, lloks like a UHI of about 0.6C in 40 years that’s about 25 times the IPCC estimate.
    I did some comparisons like this about 3 years ago, cross comparing obvious urban sites (e.g Pretoria and Tokyo) with the nearest supposed “rural” site (Jan Smuts and Ajiro, respectively). Over every continent the same pattern- obvious by Mark 1 eyeball. No fancy statistics, or “adjustments” needed. Now the IPCC tell me that the UHI effect is 0.006°C per decade??!!

  188. Posted Aug 21, 2007 at 8:01 AM | Permalink

    Hi Bernie, I do understand Portuguese but I haven’t been able to follow the discussions very closely these last days. Where exactly could I find the list of Brazilian sites filtered by GISS? Also, did GISS provide any explanation for their site-filtering?

  189. bernie
    Posted Aug 21, 2007 at 8:30 AM | Permalink

    If you read the top of this page yo will see two “here”
    links. The first takes you to GISS sites, the second takes you to all the GHCN sites. But if you check this site you will see many more weather stations. I cannot vouch for the quality of the list and I have not compared it in detail to the GISS list. Perhaps you can find an official Brazilian site that lists the stations and the data, especially raw data.

  190. Posted Aug 22, 2007 at 4:03 AM | Permalink

    Nope. South American met offices will sell some of their data and processed products but I’ve never managed to get anything detailed for free, hard though I’ve tried.

    In any case, here’s a map of all currently active (conventional) meteorological stations in Brazil:
    Left pane, link titled “Superfície Convencional INMET”. Very nice flash, but not too friendly for the purpose at hand. It does however provide the WMO number and lat/long data for each station, thus allowing a comparison with GHCN.

    A list of all currently active Argentinian stations is here:

    I believe it should be possible to get the data and perhaps some metadata for these stations from this NOAA ftp site, but I haven’t downloaded the files to check:

  191. MC
    Posted Aug 23, 2007 at 8:31 AM | Permalink

    Today (23/08/2007) in New Scientist there is a link to a paper about the lack of correlations between solar activity and mean global temperature record i.e. GISS. The authors Lockwood and Frolich only use direct satelite observations and make the statement that based on solar activity the average temperature should be getting cooler but that’s not what the GISS says.

    Looking at the initial trend of rural to urban US and Brazil stations, it is becoming more urgent that a proper analysis of ground station data over the whole world is made because at this moment, to me anyway, the suspicious data looks like the GISS data and the implication is that the sun really is a principal driving effect in GW.

    If the GISS is correct then the most probable conclusion is that recently there is another driver and honestly, a possibility is CO2.

    But for now that is just speculation. I am neither a denier or an alarmist. Clear methods and data analysis first.

  192. PaulM
    Posted Aug 23, 2007 at 10:56 AM | Permalink

    MC, the Lockwood and Frohlich paper has already been discussed at length on other threads here and other blogs for example Warwick Hughes. The paper is nonsense – there has been no decrease in the suns output. Google ‘richard willson lockwood’ to get an expert rebuttal.
    Anyway that’s a bit off topic. The main point is as you say to look carefully at data and adjustments all over the world, now that it is clear that the US is no warmer than it was in the 1930s.
    Would it not make sense to look first at the areas where, according to GISS and the IPCC, warming is greatest?
    As I understand it, this is the arctic, right?
    And does anyone know why the arctic is warming but the antarctic is not?

  193. bernie
    Posted Aug 23, 2007 at 2:04 PM | Permalink

    One of the issues that needs to be addressed are microsite impacts especially in the Arctic. A while back I looked at Cambridge Bay in Canada and tracked the actual growth of the town. Granted it is a very small settlement, but relative to its size it has grown significantly over the last two decades and there was some correlation with the warning trend. Moreover, many of these locations are military installations so the local infrastructure has continued to be developed. The net result is that these stations could have been subject to UHI/micorsite warming trends. At the same time Arctic warming does not appear to be uniform with some clear hotspots and coldspots. Perhaps we will return to some of those earlier threads once a new, data rich article covering the Arctic appears.

  194. DeWitt Payne
    Posted Aug 23, 2007 at 2:37 PM | Permalink

    Re: #195

    And does anyone know why the arctic is warming but the antarctic is not?

    Taking my cue from William Gray, IMHO it’s to a large if not primary extent a function of ocean currents (or thermo-haline circulation) which transfer heat from south to north. The AMO ramped up in 1995, not coincidentally the year the Arctic started it’s rapid warming trend. The Pacific equivalent has also been strong, but both should now be ramping down and I predict that Arctic temperatures have peaked and will soon be declining. See my post here. The same currents deliver chilled water to the Antarctic which helps to keep it cool.

  195. MC
    Posted Aug 23, 2007 at 2:59 PM | Permalink

    Apologies is this has been covered in another post/blog:

    The UK Met Office station data can be found here. I’ve only looked a a few stations but there is an interesting increase of around 0.8 degrees in the years since 1987. From looking at other parts of the website this is probably due to the processing technique they use to weight stations and create a temperature grid across the UK. I don’t know how much contribution the UK adds to the global temperature estimate but it looks like something similar to the US data is going on in the UK. I suspect the data provided is not the raw data measurements. I can maybe see how some people would jump to the conclusion that there is unprecedented heating in the last 20 years but it looks more like something is up with the processing technique. I always try to apply the Principle of least astonishment and there’s something not quite right with the way the curves I have looked at have common artefacts.

  196. bernie
    Posted Aug 23, 2007 at 4:24 PM | Permalink

    The map is great and it clearly shows many more sites than the GHCN/GISS 47/57. While many may have shorter data records, it certainly would be interesting to find some more inequivocal rural sites for Brazil. One on the island in the mddle of the S. Atlantic seems a bit sparse!!

  197. PaddikJ
    Posted Aug 25, 2007 at 8:10 PM | Permalink

    DeWitt Payne 197:

    Thanks for the post. I’ve been pursuing “The Great Melt Down” as a pet project the last several weeks. Since you seem up on the basics, maybe you could answer a few questions, or point to some sources.

    I agree w/ Dr. Gray because my own (very)tentative suspicion is that there is simply not enough energy in either the atmosphere, or from insolation (or some combi thereof) to account for the loss of ice extent we’ve been seeing. Odd as it may seem, I have not been able to find a thing for basic heat transfer between atmosphere-ice (real world vs lab or course: temps, RH, wind factors, etc). Same for increased/decreased insolation & cloud cover. NSIDC has some general comments on summer cloud cover, which leads me to suspect that soot, etc, on the ice (increased absorption) also has not much to do w/it, but I can’t find tabularized monthly cloud cover summaries for the region, so at the moment, that’s just a guess as well.


    Surely, someone has run the basic heat transfer calcs – how many joules req’d to melt how many cubic meters of ice, after accounting for wind, waves, etc, and then compare to atmospheric heat capacity and insolation? That would narrow the search. Also, I’d guess that the peculiar composition of sea ice would have to be factored, althought I don’t recall seeing that over at NSIDC either; must check again. The data must be out there somewhere; I must be missing it.

    In fact, how many cubic metres are we talking? Does anyone know, or is this all about area? Can those fancy satellites “see” ice depth?

    Cloud cover stats – per-cent monthly daytime coverage, type and radiant transparency (& at what wavelengths?), ie: how much insolation is actually reaching the ice? In solar heating systems design, the cloudiness factor – “K” – cannot be ignored. Or maybe someone has stuck some sensors in the ice on a semi-regular grid and actually measured? Tough to do on sea-ice, I would guess.

    Lastly, with all those fancy “we can measure the elevation of the Greenland ice cap to w/in 2cm” satellites, surely we can measure incoming vs outgoing radiance for very small regions. That at least would establish how out-of-balance the Arctic is, if not why. Has this been done?

    If our suspicions are correct, and the melt-down is related to more energetic ocean currents – perhaps the NAO? – the next few years should tell the tale.

  198. DeWitt Payne
    Posted Aug 26, 2007 at 12:29 AM | Permalink

    PaddikJ August 25th, 2007 at 8:10 pm

    Steve Sadlov seems to be the ice expert around here. Needless to say, I think you’re correct about needing heat from ocean currents to melt a lot of ice in a hurry, but this probably isn’t the right thread to discuss the details of your question. Try asking on the current Unthreaded #18 (at the moment). I need to construct some sort of chart to see if I can show some evidence of a change in trend without having to wait another five years.

  199. Posted Jul 16, 2010 at 11:45 PM | Permalink

    Very interesting… Does anyone have the population charts of those citys?

3 Trackbacks

  1. By Brazil » Blog Archive » Brazil on Aug 18, 2007 at 6:07 AM

    […] NASA spokesmen Gavin Schmidt and James Hansen have observed that the US accounts for only 2% of the earth’s surface and conclude that problems in the US network don’t matter. There are only 6 Brazil stations in GHCN which have records … …more […]

  2. By » McIntyre / NASA Follow-up on Aug 18, 2007 at 10:49 AM

    […] now back up. Last night he addressed the quality of non-US direct temperature measurements. He uses Brazil as an example. Brazil is the 5th largest country in the world, 12% smaller than the US and next in […]

  3. […] NASA spokesmen Gavin Schmidt and James Hansen have observed that the US accounts for only 2% of the earth’s surface and conclude that problems in the US network don’t matter. […]

%d bloggers like this: