I would presume that when making seismic measurements and "adjustments"
are made to the data that these "adjustments" have been empirically
confirmed in some manner. That is, these "adjustments" have been found
to be reliable (e.g., they found oil).
I can't say that I understand this "homogenization" process, except that
it is an attempt to make the temperature measurements coherent. I
suppose we need to be clear what measurements we are speaking of. Rich
appears to be saying that since it is gradients that we are interested
in, and not absolute temperatures, systematic errors at each site can be
eliminated (assuming that the systematic errors are independent and
produce constant effects, or at least that the systematic gradients are
much less than the temperature gradients).
The question would be whether such "homogenization" has proven to be
reliable. No data set can be said to be measuring a single parameter
(e.g., temperature). There will likely be many other influences that
affect individual measurements. Assuming that these unmeasured and
uncontrolled for influences are independent helps, as does that they are
linear, and have a zero mean affect on the measured quantity. Whatever
we are measuring, it appears essential to have some handle on the kinds
of unmeasured and uncontrolled affects. This understanding (or at least
presumption) will guide us in knowing how to massage the data and what
kinds of useful data can be gleaned from the data set. So that perhaps
absolute temperatures would not be unambiguous, but perhaps "trends" (as
Rich) says might be. But in saying this I have either presumed
something about the environment in which the data is taken or know from
experience something about it.
I am by no means an expert regarding this data set. My point here is
only to emphasize that an explicit discussion of the various
environments in which the data was taken and a characterization of the
unmeasured and uncontrolled influences might be useful. Glenn has
emphasized that some of the data was taken near "air conditioners" and
similar "biasing" environments. Saying this is not sufficient to know
how to treat the data. An attempt must be made to characterize what
influence air conditioners will have on the data. Having taken a shot
at this, one might be able to find "trends" reliable, but this is far
from certain. Given that such biased data is consistent with
independently biased data, we might be led to find the data and our
manipulations more reliable.
Well, I've said nothing that everyone doesn't already know. So I'll
stop and go do something else.
On Tue, 15 Dec 2009, Don Winterstein wrote:
> I appreciate your efforts here.
> You wrote: "I learned that if you want to measure some quantity X you don't mess with the observed values when you average up all your measurments."
> In at least two previous posts on this list I explained that there's often a need to apply adjustments to geoscience data in order to make a satisfactory interpretation. This need generally arises from unknown variations encountered during recording, and as a rule the recording can't be repeated to get a handle on the reason(s) for the variations. Data that you know from physical principles should be coherent frequently turn out instead to be incoherent, so you can't interpret without making adjustments. An example of a common kind of adjustment: in reflection seismology we always apply static corrections to land data; this process is done automatically with special-purpose computer programs. The programmer has built into these programs assumptions about many things, and there are usually several parameters the user can vary, depending on the level of sophistication. (I'm not telling you anything you don't know.) So oil company geophysicists also must "mess with the observed values" before they can interpret, and the way they do the messing can affect the interpretation.
> The need for such adjustments of geoscience data sometimes gives the interpreter flexibility to make the outcome go in a desired direction without being unethical about it. So, although I've not "engaged with the climate data," I was sure that climate scientists were making such adjustments, and I have felt that, because their biases are so obvious, some of the adjustments they've made are sure to have moved their interpretations in the direction of their biases. You're effectively arguing that much of the effect they want to see is in fact not in the data but comes from their "homogeneity adjustments." Partly because of this expectation I made "a modest proposal" a week or two ago that a different set of geoscientists, such as oil company geophysicists, with a different set of biases, should completely reinterpret all the climate data to see whether different conclusions might be plausible from the same data. Before the world spends trillions on a possible goose chase it might be good to get a second opinion.
> ----- Original Message -----
> From: Glenn Morton<mailto:firstname.lastname@example.org>
> To: email@example.com<mailto:firstname.lastname@example.org>
> Sent: Tuesday, December 15, 2009 7:12 PM
> Subject: Re: [asa] Data doesn't support global warming
> Replies for Rich Blinne, Michael Roberts, Rich Blinne again, Murray Hogg,
> Randy Isaac, Murray Hogg again, Michael Roberts again, and Gordon Brown.
> A comment on the 4 post limit. What difference does it make if one posts 29
> small emails or 1 big one? Answer, doing the former gets you chastised on
> the ASA, doing the latter doesn't.
> Rich Blinne wrote:
> Yeah, yeah, yeah, the Arctic isn't warming. This just shows you have zero
> credibility. <<<<
> I would like everyone sane on this list to note that Rich just ducked
> dealing with the monthly raw data and refused to engage with the data. His
> "scientific" reason for not paying attention to the discrepancy is "yeah
> yeah yeah the arctic isn't warming". This is so laughably not engaging with
> the data.
> >>>>TIKSI, Russia - Freed by warming, waters once locked beneath ice are
> >>>>gnawing at coastal settlements around the Arctic Circle.
> In Bykovsky, a village of 457 on Russia's northeast coast, the shoreline is
> collapsing, creeping closer and closer to houses and tanks of heating oil,
> at a rate of 15 to 18 feet a year. Eventually, homes will be lost, and maybe
> all of Bykovsky, too, under ever-longer periods of assault by open water.
> "It is eating up the land," said Innokenty Koryakin, a member of the Evenk
> tribe and the captain of a fishing boat. "You cannot do anything about
> And Have you considered that part of this might be, gasp, SUBSIDENCE? Of the
> Timan-Pechora basin in which Tiksi lies, it is written
> "The active rift stage of the West Siberian Basin terminated in the Late
> Triassic and was followed by tectonic subsidence and deposition of
> predominantly siliciclastic sedimentary rocks of Jurassic to Quaternary
> age." Mann et al, "Tectonic Setting of the World's Giant Oil and Gas
> Fields," in Michael T. Halbouty, Giant Oil and Gas Fields of the Decade,
> 1990-1999, Tulsa: AAPG, p. 48
> Not all subsidence is caused by Global WARMING, Rich. Are you going to try
> to prove that this subsidence is, or are you like the above going to say,
> Yeah yeah yeah Tiksi isn't sinking because of subsidence... blah blah
> >>>Coastal erosion is a problem in Alaska as well, forcing the United States
> >>>to prepare to relocate several Inuit villages at a projected cost of $100
> >>>million or more for each one.
> Across the Arctic, indigenous tribes with traditions shaped by centuries of
> living in extremes of cold and ice are noticing changes in weather and
> wildlife. They are trying to adapt, but it can be confounding.<<<<
> I want everyone to note how Rich has ducked, dodged and evaded the questions
> I asked him about why the monthly raw data doesn't show siberian warmth, how
> one can correct the output of a thermometer next to an air conditioner, why
> the satellite data has a step function in the middle of it, and why are
> modern thermometers made to read warmer than those of 1900. So far in this
> 'reply' of his, he is ignoring those questions. I will keep asking him to
> Rich claims to be the expert on AGW here on this list but he can't or won't
> answer a simple question about how a particular station is to be corrected
> if it has a heat source next to it.
> Now, since I will deal with issues, Building modern energy intensive
> cities, even small ones, on top of permafrost, will melt the permafrost
> below and cause them to sink into the ice. I used to work for ARCO and we
> had this problem at Prudhoe Bay. So, once again you can't prove that your
> sinking is due to warming from CO2. it might be because the huts now are not
> igloos but homes with heaters emitting too much heat for the permafrost
> below to remain solid.
> >>>Take the Inuit word for June, qiqsuqqaqtuq. It refers to snow conditions,
> >>>a strong crust at night. Only those traits now appear in May. Shari
> >>>Gearheard, a climate researcher from Harvard, recalled the appeal of an
> >>>Inuit hunter, James Qillaq, for a new word at a recent meeting in
> Let's see, most AGW folk would now claim 'anecdotal'. Is this the best you
> can do??
> >>>One sentence stayed in her mind: "June isn't really June any more."<<<
> MICHAEL ROBERTS WROTE:
> >>>Even in this funny little island we have lost our winters in the last 15
> >>>years . That I know from winter climbing. Before then you could guarantee
> >>>some good snow and ice climbing in the mountains of England and Wales,
> >>>and get many good days in. Today you are lucky to get one or two.<<<
> Maybe you should look out the window at your country, Michael. Scotland's
> Cairngorm ski resort has opened early. The article says it is an early
> Rich Blinne wrote again:
> >>>>Not only that but what about "cooling" Darwin? The following is from the
> >>>>October 2009 NCDC global report:
> Darwin, Northern Territory, Australia experienced warmer-than-average
> temperatures during October 2009. The capital city recorded an average
> maximum temperature of 34.8Â°C (94.6Â°F) on October 2009, the highest on
> record for any month (Source: Australia's Bureau of Meteorology).
> I guess the people at the Darwin Airport were firing up their air
> conditioners. :-)<<<
> Actually Rich, that wasn't what I asked you. I asked you to explain how one
> could correct the bias in the temperature record caused by heat sources next
> to air conditioners. Here is another picture.
> Rich, both the house and the air conditioner are heat sources. Siting
> requirements say that the mmts should be 10 m away from a house. Do you
> think that above picture fulfills that requirement? Don't duck weave and
> evade this question.
> And yes, if it was warm, and the Australians have air conditioners next to
> thermometers, that will warm them. Do you deny that blowing hot air
> conditioner exhaust on a thermometer will actually warm the thermometer?
> Murray Hogg wrote:
> >>>I suggest you relocate them away from the air-conditioner.<<<
> I want everyone to note how cavalier these people are when there are
> problems with measurments of the data upon which they depend. This is
> certainly not a scientific attitude--it is an attitude of a politician who
> doesn't really care about truth. Scientists are supposed to care about
> truth--in the case of the thermometers, what is the TRUE temperature.
> Frankly I want to cry when I see such abysmal sloppiness excused and
> tolerated by people who are interested in science.
> Christine Smith, I have a question for you. Do you think Murray's cavalier
> uncaring attitude is what scientists should do?
> Randy Isaac wrote:
> >>>>I think you're on the right path, doing what good scientists would and
> should do. Whenever a complex primary analysis leads to a result with
> significant impact, the first thing to do is a sanity check, or a "sniff
> test" to see if the result holds true. In essence, you've done that for the
> global warming case. Instead of taking all the data for all the earth and
> all the time periods, you've done a sanity check on specific areas for
> certain periods and found it didn't hold up.<<<<
> Thank you Randy, you don't know how much I appreciate your attitude. It
> seems that few on this list care to do a sniff test on the data (look at the
> shameful answers and evasions above).
> >>>When a sanity check of data isn't consistent with the primary analysis,
> there is one more critical step that must be taken before conclusions can be
> drawn. There are four logical possibilities:
> a. The primary analysis is wrong
> b. The sanity check is wrong
> c. Both are wrong
> d. Inconclusive--more investigation is needed.<<<
> Agreed. I would make one observation, Randy, I have probably downloaded more
> raw data here than anyone else on this list, yet almost none of them care to
> follow your sense checks. Of course my data could be wrong, but one isn't
> going to show my data wrong by saying " Yeah, yeah, yeah, the Arctic isn't
> warming. This just shows you have zero credibility" as Rich says. To show
> how I might be wrong, one must repeat my work and show where I made the
> error. Unthinking parroting of GISS sites is merely an excersise in mantra
> spouting. Thus I really really appreciate your attitude. At least you are
> willing to look at the data and possible problems
> >>>The key point is that until the source of the discrepancy between the two
> points is identified and verified, the answer is d. I have experienced many
> situations of this kind and the story isn't over until the understanding of
> why the two approaches differ has been confirmed. It doesn't matter how each
> approach can have its strong advocates. There must be an explanation of the
> That is fine, more investigation is always needed. But I will tell you from
> my experience in debating YECs that they too will say more investigation is
> needed and then ignore the problems presented. Some go so far as to say we
> should draw no conclusions until all knowledge is known. So, since I don't
> think you are a person like that, what exactly would convince you that
> global warming is false. Please be specific about what steps one should
> take. And then assure me that if I did that herculean task (which I am not
> sure I can do whatever it is) that it would actually change your mind. I
> have met far too many people in this world who say they will change in the
> face of evidence but who never will. Which kind are you Randy?
> >>>You provide eloquent descriptions of why and how various subsets ought to
> a good substitute for the full blown analysis and I'm not arguing with any
> of it. We still have a discrepancy to resolve. To do that, we have to try
> all sorts of different approaches before the matter is settled. Until we
> know the difference between the two techniques that leads to the different
> results, we will remain inconclusive.
> May I suggest an approach that could test some of the possible sources of
> discrepancy? First, instead of tackling the full global trend, one might
> focus on a more regional observation analyzed with the same techniques. So,
> for example, the GISS data which was used for global trend analysis, was
> also used for assessing the geographical distribution. Chart number 11 in
> this presentation http://www.columbia.edu/~jeh1/2009/Trinity_20090605.ppt<http://www.columbia.edu/~jeh1/2009/Trinity_20090605.ppt> is
> an example. Here we can see that their data indicates a 0.5 to 1.0C
> difference (except for Texas) in temperature for the US during the period
> 2001-2007 compared with 1951-1980. This is a specific result for a specific
> region. (By the way, that chart shows what I was referring to about
> variations in geographical regions.<<<<
> Let me ask this. If that trend shown in the above slide 11 is to be
> believed, don't you think that the monthly temperature record should show
> it? Do you think the RAW temperature should show SOME rise in temperature?
> Answer this question before we continue along this path.
> >>>>As you point out, the variations in CO2
> concentration are small, though not zero, and the physics is the same,
> though the radiation intensity is different. <<<<
> Given that the atmosphere is so thoroughly mixed that in every breath we
> take we inhale molecules that had been in the lungs of every famous
> historical person, I would challenge the idea that a stationary deficit of
> CO2 can hang around Texas or any other location for a long time (or hover
> over Brookhaven MS). So please provide a reference to a scientific paper
> saying that the CO2 content can vary for decades over a given spot. I would
> be very interested in learning that I am wrong about my view in that regard.
> >>>>However, several ocean and wind
> circulation patterns with decadal oscillations can influence the analysis
> for specific regions) For your analysis, to avoid any possibility of skewing
> data by month, it should be possible to take all 12 months for the period
> from 2001 to 2007 and derive the average anomaly, taking care to blend the
> data from the various states by accounting for the geographical area
> represented. If the error bars from that work have an overlap with the error
> bars of the GISS data, then the discrepancy disappears for this particular
> situation. <<<<
> How about I do it for the 18 cities I have spread from one end of Siberia to
> the other since you all seem to be fixated on Siberia and I see no warming
> Murray Hogg wrote:
> >>>>That the homogeneity adjustments (i.e. "editing") are responsible for
> >>>>the difference between the raw and adjusted data is quite clear. The
> >>>>real question is whether those adjustments are reliable.<<<<
> I would agree with you. I would point out that when I learned my physics
> (something Michael Roberts did not learn and isn't qualified to comment on),
> I learned that if you want to measure some quantity X you don't mess with
> the observed values when you average up all your measurments. If I am
> measuring the speed of light, I must measure the distance, and I must
> measure the time, and I might also have to measure rotation of a toothed
> wheel etc. Now, all those measurements have error in them. If I know that
> the previously measured speed of light is 299,792,458 m per second, and I
> adjust my distance measurement to make my result come out to match the
> previous value, you could rightly accuse me of cheating, of scientific
> But, somehow that doesn't seem to apply when we are trying to measure the
> TREND of the temperature. Remember, we want to observe the TREND. The
> homogeneity filter CHANGES THE TREND, THE VERY THING WE ARE TRYING TO
> MEASURE. And when you change the trend at hundreds of stations, the average
> trend output by this process is not an observed trend but something else.
> That is why the homogeneity filter is so bad.
> Here for the benefit of the lurkers are some examples, all taken from data
> downloaded at
> Note that the above is a NOAA site. I am merely reporting what they have
> Here are some examples of changing the trend, the very thing we are trying
> to MEASURE.
> I have verified these but a friend did the work and he deserves the credit.
> These pictures are from
> >>>>As you are no doubt aware, there is an extensive literature addressing
> >>>>that question, and the Peterson article from which you derived the graph
> >>>>is a part of that literature.
> Now, I regret that this particular subject is so far outside my area of
> expertise such that there is doubtless an enormous amount of background
> knowledge one needs in order to make sense of Peterson's article. But I will
> venture on the following brief observation; <<<
> Yes, but there are lots of people who think the homogeneity issue is where
> the fraud is perpetrated. I am one of those. Scientifically one should not
> alter the measurement of what one wants to determine.
> >>>Peterson's methodology is essentially to compare the trends of adjusted
> >>>data (for stations with known siting issues) and non-adjusted data (for
> >>>stations with no known siting issues) by way of running a cross-check on
> >>>the adjustments.<<<
> It is a subjective judgement that the cooling stations in his study are the
> badly sited stations. I looked into those stations and the cooling stations
> looked better sited according to the rules but he thought they needed to be
> >>>I short, am I not correct in thinking that Peterson's evaluation of the
> >>>reliability of homogenity adjustments on data from poorly sited stations
> >>>is somewhat at odds with your own?<<<
> Darn tootin' I differ from Peterson. I have tried to engage climatologists
> on the issue of stations next to air conditioners and they are all quite
> blaise about them. That is not consistent with a group of people who care
> about getting the data right, so I am not ashamed to go against Peterson on
> his changing the trend. Truth is determined by data, nothing else.
> >>>I wonder if you have taken the time to bring your analysis to Peterson's
> >>>attention, and what his response (if any) may have been?<<
> I have brought it to the attention of others. they all say what I am doing
> is well known and not newsworthy!!! Gordon Simons was witness to two of
> these instances if anyone wants to ask him for verification.
> MICHAEL ROBERTS said:
> >>>There are more than two options. There are those who don't want GW to be
> >>>true and throw up red herrings, those who are just bloody-minded for a
> How scientific of you Michael. Yep, that is a grand scientific reason.
> Anyone who has an ounce of skepticism is just bloody minded. Christine
> Smith, is this the kind of answer you think is a good one?
> GORDON BROWN WROTE:
> >>>It is not true that trees don't live in permafrost. Much of the boreal
> is located on permafrost. The lower elevations of Alaska's interior are
> not tundra. However the trees' roots cannot go deep because they encounter
> permafrost. I recall once that a tour guide pointed out an area of forest
> he called a "drunken forest" in which the trees did not stand up straight
> to the permafrost underneath them. The existence of forest in these
> is apparently determined by whether the average summer temperatures are high
> Gordon Brown (ASA member)<<<
> Not so fast Gordon. Tree roots in permanently (as in year round) frozen
> ground can't take up water. Permanentely frozen ground is the definition of
> permafrost--the perma short for permanent. Trees live ABOVE permafrost, not
> IN permafrost. Please acknowledge that I am correct on this.
> To unsubscribe, send a message to email@example.com<mailto:firstname.lastname@example.org> with
> "unsubscribe asa" (no quotes) as the body of the message.
To unsubscribe, send a message to email@example.com with
"unsubscribe asa" (no quotes) as the body of the message.
Received on Wed Dec 16 09:45:43 2009
This archive was generated by hypermail 2.1.8 : Wed Dec 16 2009 - 09:45:44 EST