Heartland’s International Conference on Climate Change is on again. I can’t help but be impressed by the number of Australian organizations co-sponsoring the conference. Sponsors don’t pay any money — instead they get free admission to all meals and sessions for up to 20 people. And with 58 sponsors and 800 people registered to attend, that means they are giving away more admissions than people registered to attend. It’s likely that almost everyone attending got free admission.

There are seven Australian organizations signed up as sponsors. As well as the obvious ones like Lavoisier and the IPA there are some unfamiliar ones, so let’s look at the whole list.

The Lavoisier Group. John Quiggin on Lavoisier:

This body is devoted to the proposition that basic principles of physics, discovered by among others, the famous French scientist Antoine Lavoisier, cease to apply when they come into conflict with the interests of the Australian coal industry.

The Institute of Public Affairs. Australia’s leading anti-science think tank, with staff including Alan Moran, Sinclair Davidson, Jennifer Marohasy and Tom Switzer (opinion editor at The Australian for much of their war on science).

The Institute for Private Enterprise. This seems to be a one-man operation by Des Moore.

Mannkal Economic Education Foundation. Their postal address is at “Hayek House”, so you can probably guess where they are coming from. Their global warming denial page is a bunch of links to other people’s stuff — they don’t seem to produce much on those lines themselves.

Climate Sceptics Party. Only launched last month. According to their platform:

We are ordinary but proud Australians who are gravely concerned with the unfounded environmental alarmism infiltrating all forms of Australian Government (Federal, State & Local), threatening our way of life and hard fought freedoms

I believe proud and ordinary cohenite is a member.

The Carbon Sense Coalition

is a voluntary group of people concerned about the extent to which carbon is wrongly vilified in Western societies, particularly in government, the media, and in business circles.

they seem to be focused on opposition to policies that reduce net emissions from agriculture (in Australia that refers mainly to land clearing in Queensland).

Australian Libertarian Society. Basically this is John Humphreys, whose response to any disagreement is to accuse you of lying. He announced his sponsorship with whoppers like this

In 2008 we have seen the coldest year since 1994 and the current temperature is nearly exactly the same as the average over the 1970s.

i-80019a093fdc112d2a043fd6f1393550-06.13.08.globalairtemp.png

True to form, when I commented on his post, Humphreys accused me of lying, though this time he also deleted most of my comments as well.

Update: Kevin Grandia:

what really strikes me so far is that it’s the same people attending and talking about the same things they did last year.

Comments

  1. #1 dhogaza
    March 17, 2009

    . . . said Wegman to Mann . . .
    (Not so well received by the latter.)

    Professionals quibbling over obscure technicalities which, as it turns out, have zero impact on the result of the statistical analysis are in no way comparable to what you’re doing, which is to suggest we toss everything ever learned by anyone about statistical analysis in the toilet.

    And to then replace all statistical analysis with cherry-picking.

    You’re even stupider than I thought.

  2. #2 Evan Jones
    March 17, 2009

    This is the very definition of cherry-picking.

    Only if you pick one cherry. Not if you consider all, including 1979 – 1998.

    It makes little sense to study a given trend in any other fashion other then, well, the beginning and end point of the trend.

  3. #3 dhogaza
    March 17, 2009

    I am the sum total of my arguments…

    In other words, you’re a zero.

  4. #4 Ean Jones
    March 17, 2009

    dhogaza: Coming from you, that is a mild compliment.

  5. #5 dhogaza
    March 17, 2009

    Only if you pick one cherry. Not if you consider all, including 1979 – 1998.

    Wrong, wrong, wrong.

    dhogaza: Coming from you, that is a mild compliment.

    You’re right, f***wad is more accurate.

  6. #6 Evan Jones
    March 17, 2009

    which is to suggest we toss everything ever learned by anyone about statistical analysis in the toilet.

    Such as the “getting rid of” the MWP? #B^1

  7. #7 Evan Jones
    March 17, 2009

    You’re right, f*wad is more accurate.

    See Anon., et al, (trad.) Rubber and Glue)

  8. #8 Evan Jones
    March 17, 2009

    In other words, you’re a zero.

    . . . he says as he goes negative.

  9. #9 Lee
    March 17, 2009

    Evan opines:
    “This is the very definition of cherry-picking.

    Only if you pick one cherry. Not if you consider all, including 1979 – 1998.

    It makes little sense to study a given trend in any other fashion other then, well, the beginning and end point of the trend.”

    Evan, conduct this little thought experiment. Go back through the surface temperature data for the last 150 years, start at the beginning,a nd step through it in order one year at a time. Fro each year, imagining yo are actually in that year and don’t have any subsequent data – How many of those years will be ‘declining trends’ based on your “its ok to cherry pick’ methodology.

    for the simplest “method” – that of simply comparing the current year to the previous high year, the answer will be “almost all of the years are in declining trends” The only yeas in which you would admit an increasing trend are the years in which there is actually a new record set.

    Even if you do the somewhat more acceptable strategy o9f using trend lines, yo will still get a strong majority of ‘declining trend” years, simply because you will always be cherry picking a high year to start from.

    And this despite the fact that the data overall shows substantial increase over that time period.

    THIS is what is wrong with your refusal to follow even basic statistical rules about data analysis.

    BTW, Evan – have you asked Watts yet when he’s going to get around to following up on his “GISS is skewed” and “covariance of ordered sets” posts? Given that he promised those followups a full year ago now?

  10. #10 dhogaza
    March 17, 2009

    The sick thing is that this Watts seems to think Evan is a valuable contributor. But we know Watts is as ignorant of statistics as Evan so I guess it’s a marriage made in heaven, eh?

  11. #11 Lee
    March 17, 2009

    Evan: “He notes that urban areas have a significantly steeper trend (at Tmax and esp. at Tmin) when comparing urban vs. rural stations.”

    Evan, why one earth do you think that the surface trend analyses – GISS and HADCRUT – correct urban trends to those of neighboring non-urban stations?
    They are there specifically because the study authors recognize a potential contaminant from urban effects, and try to correct for it.

    This is why JohnV’s analysis mattered. It showed the the corrections work, at least based on preliminary analysis.

    Watts has JohnV’s code. He could have – trivially easy – automated a process that would have given real-time updates in the analysis as more stations were added. He refused to do so. Watts has met his (preciously arbitrary) 75% threshold now. He could, trivially easily, repeat the JohnV analysis with all the extant stations. I suspect it would take about an hour, given that he already has all the station data in hand.

    Instead, he has specifically asked people to leave it alone “until he publishes.”

    Evan, Watts could have his analysis up this evening, available for comment, starting work on it right now, if he really meant to do so.

    But he has a history of hiding from issues and breaking promises to consider his analyses, when it doesn’t favor him to do so. Until I see otherwise, I’ll continue to believe that is what he is doing now that his convenient (didn’t reach it until after the last denialist confab) 75% rule has been met.

  12. #12 Evan Jones
    March 17, 2009

    The honor Anthony accorded me is far from singular. I am far from the only person banned from WattSoup.

    Very few have (and are subject to redemption). However, I believe you are the only case of retroactive deletion.

  13. #13 Evan Jones
    March 17, 2009

    Instead, he has specifically asked people to leave it alone “until he publishes.”

    But as the data is publicly archived, you are under no actual compulsion to comply.

  14. #14 Evan Jones
    March 17, 2009

    Even if you do the somewhat more acceptable strategy o9f using trend lines, yo will still get a strong majority of ‘declining trend” years, simply because you will always be cherry picking a high year to start from.

    And the reverse applies if looking at it from 1979 – 1998. To which I have no objection, whatsoever. Looking at each trend separately is valid. I would go so far as to say conflating them (without also eyeballing them separately) is less valid.

  15. #15 Evan Jones
    March 17, 2009

    The only yeas in which you would admit an increasing trend are the years in which there is actually a new record set.

    No, I merely think that examining trends from beginning to end is the best way to interpret a trend. If you look at them all–and the overall–I don’t think that can be characterized as cherrypicking.

  16. #16 Evan Jones
    March 17, 2009

    Evan, why one earth do you think that the surface trend analyses – GISS and HADCRUT – correct urban trends to those of neighboring non-urban stations? They are there specifically because the study authors recognize a potential contaminant from urban effects, and try to correct for it.

    They do, but the correction is very trivial. If you look at the specifics, almost as many cities have an “urban cooling” adjustment as a “global warming” adjustment.

    What we would like is the full set of adjustment procedures to be made publicly available so they can be replicated.

  17. #17 Evan Jones
    March 17, 2009

    Until I see otherwise, I’ll continue to believe that is what he is doing now that his convenient (didn’t reach it until after the last denialist confab) 75% rule has been met.

    Before.

    A nit-pick, true, but since my own work pushed him over the top before the conference I will, with some satisfaction, point that out.

  18. #18 Lee
    March 17, 2009

    gaaaahhhh!!!!!!

    I freaking know better.

    “Looking at each trend separately is valid. I would go so far as to say conflating them (without also eyeballing them separately) is less valid.”
    “No, I merely think that examining trends from beginning to end is the best way to interpret a trend. If you look at them all”

    Even, look at them all WHAT?!?!?!?!?!?! What are your criteria for identifying the ‘beginning’ and ‘end’ of a ‘trend?’ “Eyeballing” them? What statistical analysis do you use to test whether that is actually a unique inflection point in the data, a change from what was before to what was after? What do you have other than pulling periods out of your posterior?

    And do you realize that your comments display an appalling ignorance of basic statistical concepts?


    “What we would like is the full set of adjustment procedures to be made publicly available so they can be replicated.”

    Oh, good god. Dude, the papers are out there, the code is available for download.

  19. #19 dhogaza
    March 17, 2009

    Dude, the papers are out there, the code is available for download.

    In fact, JohnV’s Open Temps code is a rewrite that replicates the functionality but in a more readable, better structured way.

    That’s what JohnV does – replicates the GISS adjustments to raw data.

  20. #20 dhogaza
    March 17, 2009

    Handy reference to JohnV’s description of what he was up to.

    Note, Evan, where he states “then NASA released the code …”, which is what triggered his rewrite.

  21. #21 Gaz
    March 17, 2009

    Evan Jones (#186) : “..a heat sink may well exaggerate a heating trend..”

    Point is, not enough to make a significant jot of difference to the big picture, not even in the 1.6% of the Earth’s surface covered by continental USA.

    By the way, I admire you use of the term “trend” to decribe the change in temperature beween day and night. Excellent stuff. And here was I thinking 11 years was a tad short. 12 hours. Way to go Evan.

    Now please stop torturing me and go back to torturing the temperature data.

  22. #22 Evan Jones
    March 17, 2009

    If you bothered to address the Yilmaz study you might get somewhere.

    Furthermore, I have said (repeatedly) that more study needs to be done under surface station conditions. This seems to have raised a lot of objections. One wonders why.

    “then NASA released the code …”

    That was FILNET, only. And it was a non-working FORTRAN dump with no operating manuals. Mac & Co. finally got it running, but they can’t reproduce NASA’s results. We are still waiting for the ability to reproduce NASA’s full adjustment.

    What are your criteria for identifying the ‘beginning’ and ‘end’ of a ‘trend?’

    Well, the 20th century trends are pretty obvious, either for the US or the ROW. You should have very little trouble seeing them.

    And do you realize that your comments display an appalling ignorance of basic statistical concepts?

    Suggesting that to understand a trend, one should judge it from beginning to end does all that? Okay.

  23. #23 Evan Jones
    March 17, 2009

    Can we all just conclude that Evan Jones. like Anthony Watts, is an uneducated idiot and just ignore him?

    Hmm. If only everyone who thinks I am an uneducated idiot would ignore me. If only . . .

    What can I say other than that I heartily endorse this suggestion?

    (Well, I can dream, can’t I?)

  24. #24 Evan Jones
    March 18, 2009

    Can we all just conclude that Evan Jones. like Anthony Watts, is an uneducated idiot and just ignore him?

    Hmm. If only everyone who thinks I am an uneducated idiot would ignore me. If only . . .

    What can I say other than that I heartily endorse this suggestion?

    (Well, I can dream, can’t I?)

  25. #25 Evan Jones
    March 18, 2009

    BTW, if you want to “recreate” GISS data, starting with NOAA raw data won’t do.

    In case you weren’t aware, GISS starts with NOAA data, applies a “deadjustment” algorithm (which does not yield the same result as NOAA raw data), and then goes on from there.

  26. #26 dhogaza
    March 18, 2009

    And it was a non-working FORTRAN dump with no operating manuals.

    Bullshit. It was their FORTRAN code which denialists bitched about because “we can’t figure out how to compile it”.

    That’s totally different than saying “it’s non-working FORTRAN code”. Obviously, the code works at GISS.

    You’re such a fucking liar.

  27. #27 dhogaza
    March 18, 2009

    Hmm. If only everyone who thinks I am an uneducated idiot would ignore me. If only . .

    Note that nothing you have done has had any impact on actual science. The real kind, the peer-reviewed, published kind.

    You get attention in the blogosphere because people hope to educate your double-digit IQ mind.

    Not because anyone thinks that anything you’re doing is interesting from a scientific POV.

  28. #28 dhogaza
    March 18, 2009

    Suggesting that to understand a trend, one should judge it from beginning to end does all that? Okay.

    Stupid, because in science “trend” means something of statistical significance. Obviously, a priori choosing of start and end points is arbitrary, which is the definition of cherry picking, and totally in opposition to everything statistics tells us about analyzing of data.

  29. #29 dhogaza
    March 18, 2009

    Of course, if Evan Jones is right, not only will climate science be overturned, but *all* of science will be overturned.

    My MacBook pro will be turned to dust, because the mighty Evan has proven that everything that modern science depends on has been overturned …

    Fat chance of that, dude.

  30. #30 dhogaza
    March 18, 2009

    What can I say other than that I heartily endorse this suggestion?

    No worries, serious science ignores Watts, and even more so, ignores you.

    Your fame is limited to a very small slice of the blogosphere.

    I actually see no evidence that anyone outside Watts thinks you’re even worth listening to in the denialsphere.

    I’d say you don’t have to worry too much about exceeding Warhol’s 15 minutes of fame.

    But you’ve certainly exceeded 15 minutes of ridicule and insult.

    Because you’re deserving of both.

    Science will continue to move forward long after your anti-science, denialist crap has been forgotten.

  31. #31 Evan Jones
    March 18, 2009

    Note that nothing you have done has had any impact on actual science. The real kind, the peer-reviewed, published kind.

    Well, I have developed certain methods and techniques for virtual surveys which may have some future utility. And I expect that the data I have assembled will be used in at least one peer-reviewed paper. As for its lasting impact, if any, I wouldn’t know.

    It was their FORTRAN code which denialists bitched about because “we can’t figure out how to compile it”.

    Well, they did figure it out eventually, of course. (Something I think Dr. Hansen believed they would be unable or unwilling to do.) They also made the most amazing discoveries about it. But I assume you would be aware of all that.

    On the other hand, GISS has a disconcerting way of changing their metadata without announcement, so I suppose that whatever code they were using in June 2007 may have been altered by now.

    I was most suspicious of SHAP, at first. How they got a positive adjustment trend out of that is something that eludes me. (I have never gotten an answer to that one. Perhaps someone will care to explain it to me some day.) And FILNET has proven to be most amusing. But I am holding out hope that the homogenization procedure (which would seem to involve a spot of pasteurization) may yet provide the best entertainment of all.

    And who knows what may be lurking in those TOBS calculations? It is certainly a valid and necessary adjustment. yet I find myself becoming more and more interested in the actual execution thereof.

    Your fame is limited to a very small slice of the blogosphere.

    My word. I had no idea I had any fame in any slice of the blogosphere, whatever. Why, thank you for your kind words!

    As for the rest, I can only ardently wish that as many people as possible read what both of us have to say and come to their own conclusions.

  32. #32 Evan Jones
    March 18, 2009

    but there were lots of sites that were close enough to tree lines and bushed that they were obviously going to be in the shadows part of the day, and obviously were in some of the photos.

    COOP still uses a min-max/2 measurement (we are SO stone-age). So what the NOAA COOP requires is that the sensor be unshaded to the south (which would shade it even at noon). Otherwise, if the sensor is unshaded at Tmax time (usually not too long after noon), the question is entirely moot.

    This, of course, does not even take into account those near enough significant transpiring veg that they would be cooled.

    Yes, but there are not too many with that problem. And those that are usually have warming biases, as well. Check it out. Also, at night, a canopy can trap heat, affecting Tmin. That is GISS’s excuse for applying an “urban cooling” [sic] adjustment in parts of South America.

    Now the whole thing misses the point that what is being measured is an anomaly, not a trend, so what counts is not whether a site is warmer or colder than neighboring sites, but whether the site has changed over the years in a way that would bias the trend, not the absolute temperature. Evan Jones is another idiot who needs to understand the concept, but would rather expose himself.

    Actually I have been strictly differentiating offset and trend, throughout. Had you not noticed this? After all, I would be “another idiot” had I not been very careful to do so. #B^1

    Actually, when one calls someone an idiot for making a basic error, one should check more carefully to make sure he is actually committing the error–as opposed to punctiliously and explicitly avoiding the error. (Sheesh.) Not to do so would be, er, um, careless, as it were. However I remain confident that, as a professional, you meant no offense, whatever.

    Yes, we know that a heat sink may seriously affect the offset.

    The Big Question is if/how much it affects the trend.

    The Yilmaz data very clearly indicates that it does. It certainly requires further study. There seems to be much objection to this, for an unknown (or at least unstated) reason.

    However, having said all that, undocumented station moves (as exemplified by the MMTS debacle) or even nearby construction may well have resulted in offsets getting mixed into the NOAA data and affecting the measured trend, creating a spurious warming effect. Sometimes even documented site moves create serious step-changes, obvious to any observer, yet that somehow manage to appear in the post-adjustment data. Take the USHCN Champas, NM, site, for example (or Lampasas, TX).

    (BTW, did you know that a Nimbus unit can seriously dud out on you if you touch it and create a static-electric shock? I have had more than one complaint about that. You should probably alert your brethren on that score.)

  33. #33 Barton Paul Levenson
    March 18, 2009

    Evan Jones writes:

    It makes little sense to study a given trend in any other fashion other then, well, the beginning and end point of the trend.

    It’s not a trend. A trend has to be statistically significant. You don’t have enough data points for that.

    Read it again: “Climate is defined as mean regional or global weather over a period of 30 years or more.”

    You are taking a small jog in the data that seems to lead to the conclusion you want and isolating it to prove your point. That’s not legitimate. You have to use all the points, not just the ones that support the conclusion you want. And you have to have enough to draw statistically valid conclusions.

    There is no cooling trend. There is no flat trend. A trend has to be statistically significant. If it isn’t it is not a trend.

  34. #34 dhogaza
    March 18, 2009

    Well, they did figure it out eventually, of course. (Something I think Dr. Hansen believed they would be unable or unwilling to do.

    Evan, at times you are unwittingly funny as hell.

  35. #35 dhogaza
    March 18, 2009

    On the other hand, GISS has a disconcerting way of changing their metadata without announcement, so I suppose that whatever code they were using in June 2007 may have been altered by now.

    This is unwittingly funny as well …

    Let’s see …

    Evan believes that the GISS algorithms designed to generate a useful global temperature product aren’t capable of doing so because well, he’s got photos to prove it.

    Then he bitches because perhaps GISS is improving those algorithms and the code which executes them …

  36. #36 Evan Jones
    March 18, 2009

    I do object to making an unannounced change without explaining why. If an adequate explanation is made, then fine.

    (And of course, the “improvements” have always been to cool the past and leave the present as is, thus increasing the warming trend.)

    And I have also made it clear that we do not know if the photos affect the trend or if the data is invalid. Only that the bad siting means that the issue requires study. As does the station move issue.

    You realize that no one is going to pay any attention to what you say if you so badly distort what others are saying?

  37. #37 Evan Jones
    March 18, 2009

    I do object to making an unannounced change without explaining why. If an adequate explanation is made, then fine.

    (And of course, the “improvements” have always been to cool the past and leave the present as is, thus increasing the warming trend.)

    And I have also made it clear that we do not know if the photos affect the trend or if the data is invalid. Only that the bad siting means that the issue requires study. As does the station move issue.

    You realize that no one is going to pay any attention to what you say if you so badly distort what others are saying?

  38. #38 Evan Jones
    March 18, 2009

    BPL: I understand your objection.

    The last three phases have all been less than 30 years. The current phase is only a decade old, the previous phase was only around twenty, and the one before that around twenty-five.

    In 1988, Dr. Hansen did not see fit to restrict himself to the 30-year rule, nor in my opinion should he have. You have to call them as you see them, as indeed Dr. Hanson did. I think he was probably wrong, but I don’t object to his timeframe.

    I suggest the 30-year rule is arbitrary and likely to yield a poorer understanding of climate. For example, I think one learns far more by isolating the c. 20-year warming from the late 1970s to 1998 from the previous cooling and subsequent flat periods. We care, primarily, about what made the temperatures rise during that period.

    (As the meaning of “flat trend” is quite clear and everyone, including you, know what I mean by it, I will continue to employ the term and beg your indulgence.)

  39. #39 Evan Jones
    March 18, 2009

    Evan, at times you are unwittingly funny as hell.

    You are one up on me, then. You never once fail to amuse.

  40. #40 Evan Jones
    March 18, 2009

    he’s got photos to prove it.

    One wonders what you might be saying in the absence of those inconvenient photos, of course . . .

  41. #41 dhogaza
    March 18, 2009

    I have also made it clear that we do not know if the photos affect the trend

    Well, I’ve never seen a photo warm enough to affect even the weather …

  42. #42 dhogaza
    March 18, 2009

    In 1988, Dr. Hansen did not see fit to restrict himself to the 30-year rule, nor in my opinion should he have

    And, please, quit misrepresenting Hansen’s 1988 testimony.

  43. #43 Former Skeptic
    March 18, 2009

    Evan Jones:

    I am sure you have seen geographic graphs running over urban areas that show temperatures lower as they approach a UHI bubble, rise ‘way up as they go over it, then decline once they are past it. What i think is that this has to be done on a year-round basis on a city (or cities) that are windy vs. cities that are not, and on a year-round basis and taking prevailing wind direction into account.

    And as I have said, check the references listed in my previous posts – these near-surface UHI surveys have been done for numerous cities world-wide under different wind conditions, and have shown that there is a noticeable decrease in UHI intensity with higher wind-speeds. The empirical evidence shows that:

    (a.) Pielke and Matsui’s 2005 paper has no empirical proof in urban stations sited in the roughness sub-layer/urban canopy layer, and;

    (b.) Parker’s 2004 and 2006 papers (showing that UHI from “urban” siting and from urban growth/encroachment has, for all intents, no impact on the global surface T record since there is no statistical difference in trends in calm vs. windy conditions) make theoretical sense.

    Look, not to be as snarky like the others are here (even though I can clearly see why they are), but your repeated calls that “more study is required” would hold more water if you do actually read the literature out there and understand it.

    Thank you for your answer about my source areas question, but I think you misunderstood me. I was asking whether there is information in surfacestations.org showing whether anyone estimated the areal size around the temperature sensor which it measures. Yes, it’s certainly important to document the micro-site conditions, but has anyone examined the range of influence of thermometers under different stability conditions? Perhaps this point, rather than how “badly-sited” the sensors are, is the key issue here.

    And yes, that’s the LaDochy study I was referring to (the one published in Climate Research). And as I said, where does it explicitly dispute Peterson’s conclusions? Steve’s group analyzed unadjusted data while Peterson looked at data adjusted for elevation, MMTS vs. CRS, rooftop siting, latitude etc. Clear case of apples vs. oranges here – there’s a big difference in method – did you actually notice it when reading the paper? While the LaDochy findings are somewhat interesting, especially their correlation with PDO, it lacks the spatial scope of Peterson’s study, and it does not examine trends with respect to wind-speed. IMO Steve’s paper could have made a better argument if they had looked at the dT/dt for windy vs. calm conditions, but that’s a moot point.

  44. #44 Evan Jones
    March 18, 2009

    there is a noticeable decrease in UHI intensity with higher wind-speeds. The empirical evidence shows that.

    I am not arguing that it does not. It is purely a matter of degree.

    Yes, it’s certainly important to document the micro-site conditions, but has anyone examined the range of influence of thermometers under different stability conditions? Perhaps this point, rather than how “badly-sited” the sensors are, is the key issue here.

    No I don’t know. The actual science is being done by others. And yes, perhaps stability issues might prove to be far more key that microsite conditions. Or might station moves (either undocumented or documented yet ignored). Or faulty adjustment procedures. Or (etc.)

    Look, not to be as snarky like the others are here (even though I can clearly see why they are),

    Music to my ears. (Eyes?) The effect is acute self-immolation, though they don’t seem to catch on.

    but your repeated calls that “more study is required” would hold more water if you do actually read the literature out there and understand it.

    I am trying to get a rough handle an many aspects. I can’t study every aspect. The “basic literature” on any small aspect of climate is worth an MS all by it’s lonesome. And there are a huge number of “small aspects”. So I am doing what I can to juggle.

    But I outline what I mean by a practical approach to “more study” for UHI, below, and it simply can’t be answered by the basic literature.

    And as I said, where does it explicitly dispute Peterson’s conclusions?

    Starkly, but quite indirectly: i.e., via its conclusions.

    Peterson maintains the UHI effect is relatively insignificant. To the best of my recollections, LaDochy’s observations indicate that the delta-Tmax is double and the delta-Tmin is quintuple for urban areas as it is for rural areas.

    I am not saying who is right and who is wrong. Which leads to my next point.

    And I did rough out the “required study” that would clear up the issue, empirical though it may be.

    To flesh out the detail:

    –Select a sample of cities with a variety of conditions. Windy, non-windy, desert, jungle, tundra, agricultural, sheltered, open, whatever. Mix and matich. You’d know what to pick better than I.

    –Use surface level sensors, as we are primarily interested in UHIE on the USHCN network. (This is important.)

    –Sensors would be placed at varying points in the cities, and outside (upwind, downwind, and crosswind).

    –You’d need maybe a hundred sensors. Maybe 200, depending on your sample size. Standard-issue MMTS units would be preferable, but it might be more practical to use the self-contained, fully automated (and much cheaper) stuff. Those
    In quantity you could probably snag those for maybe $100 to $300 per unit.

    Also, cheaper, better equipment would allow both hourly measurements and a Tmax/Tmin calculation. And they store the data automatically, so no need for daily readings and (joy of joys) no missing records. (25% or more of USHCN data is AWOL and has to be handed over to the tender mercies of FILNET, which, for an unknown reason, adds a whoppingly large warming adjustment using USHCN-1. No mas!) So long as the equipment is compatible, you’d probably get more experimental bang for the buck than by going the non-MMTS/Nimbus route.

    –Run the experiment for at least a full year (for obvious reasons), making sure the stations were properly maintained and that the microsite remains constant.

    This begs theory, atmospheric layers, what have you, and measures UHI effect the same way the GHCN does. In your face, down and dirty, and strictly empirical.

    Then we will find out how UHI affects temperature offsets and trends, and what the UHI adjustment should be. Nothing else will tell really us. Not Peterson. Not LaDochy.

    That’s the “more study” I think needs to be done.

    Funding, anyone? #B^1

  45. #45 dhogaza
    March 18, 2009

    but your repeated calls that “more study is required” would hold more water if you do actually read the literature out there and understand it.

    Another of Evan’s charming attributes is a historical tendency to read a paper then totally misunderstand and/or misrepresent the results. A bit like those creationists who cite research into evolutionary biology as evidence that evolution is impossible.

  46. #46 Evan Jones
    March 18, 2009

    Evan’s charming attributes

    Awwww, you old softie, you. Still trying to butter me up!

    Keep adding. (Maybe you’ll even make hominem someday.)

  47. #47 Ian Forrester
    March 18, 2009

    Evan Jones said: “As the meaning of “flat trend” is quite clear and everyone, including you, know what I mean by it, I will continue to employ the term.”

    Your middle name wouldn’t happen to be “Dunning” or “Kruger” would it?

    Your posts are ideal examples for any high school student wishing to write an easy on the D-K syndrome.

  48. #48 Evan Jones
    March 18, 2009

    Former Skeptic: Have you seen the Ren et al. (2008) paper on UHI in Northern China? Based on urban vs. rural station trends, it estimates UHIE to be c. 1.2C/century.

  49. #49 Evan Jones
    March 18, 2009

    Come to think of it, Ren jibes pretty well with LaDochy.

    (I believe they both use GHCN-adjusted rather than raw data, so the difference might be on top of UHI adjustment.)

  50. #50 dhogaza
    March 18, 2009

    Yes, they suggest that the true warming in this part of china has been 0.8C since 1950.

    Real proof that AGW isn’t happening, eh, Evan?

  51. #51 Lee
    March 18, 2009

    Evan suggests that the ‘trends’ i teh record are obviosu by “eyeballing”

    Let me give it a shot, using Evan’s suggestion. I’ll do just 1990 – now, to make it easier.

    1990 – 1992 – down trend
    1992 – 1995 – up trend
    1995 – 1996 – down trend
    1996 – 1998 – up trend
    1998 – 1999 – down trend
    1999 – 2000 – gee, this must be what Evan means when he ays everyone know what he means by a flat trend.
    2000 – 2002 – up trend
    2002 – 2003 – down rend
    2003 – 2004 – up trend
    2004 – 2005 – down trend
    2005 – 2006 – up trend, but almost ‘flat’
    2006 – 2007 – down trend.

    Clearly, every up trend is followed by a down trend, and every down trend is followed by an up trend, so there cant be any warming. Or cooling. QED.

    Gee, this is fun, Evan. Why did I waste all that time trying to understand the basics of time series analysis, when this is so much easier and more productive.

    http://data.giss.nasa.gov/gistemp/graphs/Fig.A2.lrg.gif

  52. #52 Evan Jones
    March 18, 2009

    Clearly, every up trend is followed by a down trend, and every down trend is followed by an up trend, so there cant be any warming. Or cooling. QED.

    I never suggested that, of course. Obviously not all trends are created equal. I do think there has been a mild warming. (Probably not as much as the IPCC, NOAA, or GISS suggest.)

  53. #53 Gator
    March 18, 2009

    Evan Jones (#186) : “..a heat sink may well exaggerate a heating trend..”

    Nope. A heat sink will either act to throw away heat, thus lowering temperatures in your system. I think you mean a system with a large heat capacity. This will act to dampen the overall temperature swings. As heat is input, the extra heat capacity absorbs the energy and warms slowly. When the heat is removed, the large heat capacity will slowly give up heat and slow cooling.

    So your statement is completely wrong.

  54. #54 Evan Jones
    March 18, 2009

    Yes, they suggest that the true warming in this part of china has been 0.8C since 1950.

    Ren suggests that UHI effect alone has increased the warming tend measurements by 0.1C/decade since 1960.

    Even Phil Jones agrees, in the case of Northern China.

  55. #55 Evan Jones
    March 18, 2009

    Nope. A heat sink will either act to throw away heat, thus lowering temperatures in your system. I think you mean a system with a large heat capacity. This will act to dampen the overall temperature swings. As heat is input, the extra heat capacity absorbs the energy and warms slowly. When the heat is removed, the large heat capacity will slowly give up heat and slow cooling.

    So your statement is completely wrong.

    You are confusing the tactical with the strategic.

    Take a black tar driveway. It does not produce heat. What it does is soak up joules during the day as well as reflecting.

    The reflection increases Tmax. As night falls, heat is released, just as you say. And at Tmin it is still being released, thus exaggerating Tmin.

    The greater the warming, the greater the disparity. Therefore, quite apart from the offset effect, a heat sink exaggerates a warming trend. But there has to be a warming trend in the first place to exaggerate.

    It also exaggerates a cooling trend, as the effect undoes itself. Thus, I suspect the cooling over the last couple of years has been exaggerated by heat sink effect.

    It works in both directions.

  56. #56 Evan Jones
    March 18, 2009

    I mean “thus reducing Tmin”.

  57. #57 Evan Jones
    March 18, 2009

    Increasing Tmin?

    I mean making it hotter.

  58. #58 Lee
    March 18, 2009

    baaaahahahahaaahaha…

    Evan disputes my tongue in cheek ‘conclusion’ from parsing of an 18 year time series into 12 distinct ‘trends,’ but has nothing to say about the parsing into trends in itself..

    Folks, this is D-K committed at a level that can only be called genius.

  59. #59 Barton Paul Levenson
    March 19, 2009

    Evan Jones writes:

    The last three phases have all been less than 30 years. The current phase is only a decade old, the previous phase was only around twenty, and the one before that around twenty-five.

    In 1988, Dr. Hansen did not see fit to restrict himself to the 30-year rule, nor in my opinion should he have. You have to call them as you see them, as indeed Dr. Hanson did. I think he was probably wrong, but I don’t object to his timeframe.

    I suggest the 30-year rule is arbitrary and likely to yield a poorer understanding of climate. For example, I think one learns far more by isolating the c. 20-year warming from the late 1970s to 1998 from the previous cooling and subsequent flat periods. We care, primarily, about what made the temperatures rise during that period.

    You just don’t get it, do you? It’s not a question of choosing whatever period appeals to you. Whether a relationship is statistically significant or not is something that can be MEASURED. You do it with ten years of annual climate data and you get “INSIGNIFICANT.” Capiche?

    Allow me to show you the actual tests:

    Year Anom Slope p
    1988 0.180 0.020 0.000
    1989 0.103 0.021 0.000
    1990 0.254 0.020 0.000
    1991 0.212 0.023 0.000
    1992 0.061 0.025 0.000
    1993 0.105 0.022 0.002
    1994 0.171 0.019 0.011
    1995 0.275 0.016 0.044
    1996 0.137 0.016 0.092
    1997 0.351 0.007 0.424
    1998 0.546 0.005 0.643
    1999 0.296 0.017 0.084
    2000 0.270 0.012 0.279
    2001 0.409 -0.003 0.618
    2002 0.464 -0.012 0.095
    2003 0.473 -0.017 0.116
    2004 0.447 -0.020 0.270
    2005 0.482 -0.040 0.179
    2006 0.422 -0.020 0.000
    2007 0.402

    The third column is the slope of the relationship between temperature anomaly and elapsed time in years. The fourth column is the significance, there is a chance of less than the figure indicated that the relationship is due to chance. The usual criterion statisticians use is p < 0.05 or “95% confidence.” By that criterion, 1998-2007 is no better than flipping a coin. You don’t even start to get significance until you start in 1995 (N = 13), and when you do that all the slopes are significant positive (warming). 2006-2007 is “trivially significant” because there are only two points and any linear fit will always be perfect.

    1998-2008 tells you nothing meaningful about climate. Nothing. Nada. Zip. It’s not enough information. I’m not saying this to be mean or because I don’t want it to be enough. I’m saying it because I ran the regression and it came out statistically insignificant. You can do it too. The anomalies are Hadley Centre CRU, not NASA GISS. Plug the numbers into Excel and run the regression yourself.

  60. #60 Gator
    March 19, 2009

    Evan at 256: Nope, you are now confusing “heat sink” with absorber. Look at the situation when you add a tar driveway near a thermometer. During the day, the black surface will absorb visible light, turn it into heat, heat air nearby and radiate in IR. It is not “reflecting” heat to make you feel hot. It probably will increase Tmax because it is physically absorbing energy right there and heating up. It is not a significant heat capacity though, and as the sun goes down, it will efficiently cool. The same things that make it good at heating the local area (i.e. black, large surface area) also make it good at cooling to the ambient temperature. It probably won’t affect Tmin. If the black surface were over a swimming pool, which has a large heat capacity, then you might affect Tmin.

    In any case, how would this create a trend? If I put in a driveway, that takes a day. The next day, my thermometer might read 2 deg hotter than it would without the driveway. But it’s not going to read 3 degrees hotter 10 years from now. It will be the same driveway, with the same affect, with the same offset.

    It sounds like you are misusing the term “heat sink” and “trend”.

  61. #61 Evan Jones
    March 19, 2009

    Okay, call it “absorber”, The effect is what it is.

    In any case, how would this create a trend?

    A fair question. The greater the difference between Tmax and Tmin, the greater the effect. So when temperatures actually do get warmer, the effect is magnified and added to the total. The greater the actual warming, the greater the spurious warming effect.

    The reverse, of course, applies to a cooling trend.

    So both warming and cooling trends are magnified.

  62. #62 Evan Jones
    March 19, 2009

    1998-2008 tells you nothing meaningful about climate. Nothing. Nada. Zip. It’s not enough information. I’m not saying this to be mean or because I don’t want it to be enough. I’m saying it because I ran the regression and it came out statistically insignificant.

    I know you are not. You have been civil and reasonable, throughout.

    Perhaps I am disputing that a decade of “weather” is statistically insignificant. And I may well be wrong.

    I certainly do not consider the 20 year warming period from 1979 – 1998 to be insignificant (nor do most AGW advocates). Perhaps I am wrong there, also.

    But there appear to be statistically significant drivers with half-cycles shorter than 30 years. So if the 30-year rule is indeed correct, it leaves us in a very awkward position when we try to evaluate it.

    Would you suggest looking at it from high point to high point from the 1930s to 1998 as compared with low point to low point from the mid-1950s to 2009? The trouble being that the current cooling–whatever–may not have spun out its course yet; it could end tomorrow or go on for decades.

    If you take the last 30 years (discounting for Pinatubo), there has been a warming trend, but a rather mild one.

  63. #63 Evan Jones
    March 19, 2009

    BTW, if it turns out that poor siting creates uncertainties that create an error bar of, say 2C, what would that mean for the significance of the 20th century (post-adjustment) warming trend, as estimated (a trend of 0.006C – 0.008C per year)?

    Bearing in mind that the adjustment is NOT negative, but +0.3 to +0.4 degrees per century?

  64. #64 Chris O'Neill
    March 19, 2009

    Evan Jones :

    If you take the last 30 years (discounting for Pinatubo), there has been a warming trend, but a rather mild one.

    Pull the other one.

  65. #65 dhogaza
    March 20, 2009

    Perhaps I am disputing that a decade of “weather” is statistically insignificant. And I may well be wrong.

    In other words … I have a definition of trend that I think scientist should adopt, but I don’t know enough statistics to determine if a decade of weather is statisticallly significant or not.

    In other words, I’m an idiot, but I’m still going to overturn all of climate science.

    Evan Jones is a tard.

  66. #66 Evan Jones
    March 20, 2009

    dhogaza: You do realize how much you are embarrassing to those who agree with you on AGW?

    (Don’t worry, though, you others, I’ll overlook it. I won’t allow him to rub off on you. Besides, I find him almost endearing.)

    Chris O’Neill: You, on the other hand, merit a civilized response.

    First, please bear in mind that I do believe there has been warming. I also believe there is a CO2 fingerprint. It’s all a mater of degree.

    GISS has made itself a bit of an outlier out of itself, as of late. (We can spitball over their recent readjustments if you’d like.) But going by the other three (Hadley, UAH, RSS), using adjusted data, what we’ve had is around +0.6C (on the high side) from 1979 – 1998 and a giveback of maybe 0.1C since then. Pretty much the same trend as 1915 to 1945, which was without much CO2 increase, depending on your proxy.

    These are comparable curves because they both begin at the start of warming trends. We are also stipulating that the adjustments are correct (the raw data + TOBS works even better for my argument).

    Definitely warming trends, but not exactly setting off any fire alarms. And the curve is currently humping towards the cool side, not headed the other way at present.

    If we are going to hit that +3.5C by 2100, we’d better get a move on. We have only 90 years to gain that 3.6C. We’ll need +0.4C/decade for that, and we’ve been doing well under half that for the last three decades.

    It all comes down to whether that CO2-positive feedback loops theory actually works out. (So far the Aqua Satellite sees the opposite effect, but those results are still preliminary. Dare I say that more study is required?)

  67. #67 Former Skeptic
    March 20, 2009

    Evan Jones:

    No I don’t know. The actual science is being done by others.

    Thanks for clarifying. I have my doubts that this is truly the case – but the fact that nothing has been seen in the literature to confirm this issue is telling. I still fail to see why so much effort is spent by Watts and co. on documenting microsites when this issue of instrument source area under different stabilities is, IMO, a more important issue. For one thing, it’ll tell you the extent (10 yard radius? 100 yards? 500 yards?) to which you should document the site conditions.

    I am trying to get a rough handle an many aspects. I can’t study every aspect. The “basic literature” on any small aspect of climate is worth an MS all by it’s lonesome. And there are a huge number of “small aspects”. So I am doing what I can to juggle.

    Doesn’t it feel tiring to juggle so much? Why not get an MS on climate/micro-met then? It’ll help you to learn some of the theory that you appear not to read up on. Or try to publish something on this in a reputable peer-reviewed journal rather than on a website or blog? For one thing, these steps will make you realize some of the errors that you are making.

    And as I said, where does it explicitly dispute Peterson’s conclusions?

    Starkly, but quite indirectly: i.e., via its conclusions.

    OK, unless my English is screwed up…how does this sentence make sense? Starkly but indirectly?

    There is nothing in Steve’s conclusions that directly or explicitly contradicts Peterson’s paper. Why? see below. Once again, have you read the paper?

    Peterson maintains the UHI effect is relatively insignificant. To the best of my recollections, LaDochy’s observations indicate that the delta-Tmax is double and the delta-Tmin is quintuple for urban areas as it is for rural areas.

    There’s the rub. As I’ve said above – the two papers are comparing apples with oranges. If Steve’s group had applied the T adjustments that Peterson made, they would not have such a large dTmin/dt.

    For the last time, have you read the paper and tried to understand it? Or did you just read a cherry-picked summary that I’ve seen posted on blogs?

    In any case, there is a chance I’ll meet Steve LaDochy next week. I’ll definitely raise this issue with him.

    But I outline what I mean by a practical approach to “more study” for UHI, below, and it simply can’t be answered by the basic literature…And I did rough out the “required study” that would clear up the issue, empirical though it may be…To flesh out the detail:(edit)

    And as I said, there are studies in the urban climate literature that have (largely) done what you suggest. E.g. Jerome Fast in Phoenix, Kathy Runnalls in Vancouver, Ingegärd Eliasson in Sweden, Winston Chow in Singapore etc. It helps to read the stuff out there.

    Then we will find out how UHI affects temperature offsets and trends

    And as I said – these folks found that near-surface UHI does “go with the wind.” David Parker used this to show that globally, the UHI has insignificant effect on SAT record trend since 1950.

  68. #68 Former Skeptic
    March 20, 2009

    Evan Jones:

    Have you seen the Ren et al. (2008) paper on UHI in Northern China? Based on urban vs. rural station trends, it estimates UHIE to be c. 1.2C/century.

    I have. And sadly, it is fundamentally flawed for one reason:

    They do not analyze minimum temperatures.

    If you want to study UHI, you have to examine nocturnal T – not mean T, which is what they did. On top of the conflation of larger-scale daytime/nightime T forcings that would affect the “urban” signal (the Asian brown cloud being one example off the top of my head), there is also strong evidence of urban “cool islands”/rural “heat islands” during daytime periods. This is caused by higher thermal inertia (i.e. concrete takes a longer time to heat up than bare soil) and greater shading in urban areas which will obviously affect the data quality. And, as several of their “rural sites” are arid deserts, which generally have higher daytime T than in urban areas, this would strongly contaminate their background T signal.

    I feel very sad when I read this paper. Ren and co. obviously did a lot of work as documented in their methods section, but they chose the wrong metric to analyze, which makes their conclusions fatally flawed. What a bummer.

    Come to think of it, Ren jibes pretty well with LaDochy…(I believe they both use GHCN-adjusted rather than raw data, so the difference might be on top of UHI adjustment.)

    No. Both Ren and LaDochy use unadjusted raw station data. Please read the papers – especially the respective methods sections – before making such assumptions.

  69. #69 Evan Jones
    March 20, 2009

    The anomalies are Hadley Centre CRU, not NASA GISS.

    Yes, one can tell by all those minuses at the end . . .

    I have to admit that if I have to pick between Jim Hansen and Phil Jones, I’ll go with HadCRU. (At least they don’t rely on the “Siberian Thought Criminals” to grid them to the north pole.)#B^1

  70. #70 Evan Jones
    March 20, 2009

    Former Skeptic.

    Hmmm. If it’s flawed, then why is Dr. Jones buying it?

    And are not all effects (both cool park and hot spot deals) all meat for the average?

    I am concerned with UHI effects on temperature data (Tmax & Tmin) more than the inner zen of UHI, itself. So why would one not use the mean to determine that? (For that matter, how would one not use it?)

    I am assuming the brown cloud is a cooling offset overhead (increased albedo)–until it gunks up the arctic ice and increases melt, that is. Albedo up, albedo down. Four legs good, two legs bad. But I’ve heard a lot of disagreement as to its overall effect.

    Ren simply compares urban (by size) data with rural. The data is standard max-min How do you get away from that?

    Please read the papers – especially the respective methods sections – before making such assumptions.

    I read LaDochy a year back, but only the abstract for Ren, so far. I’ll get to it. (Do you recall offhand if they include a TOBS adjustment?) At any rate the UHI adjustment would have been minuscule–about 5% of the observed difference for Ren.

  71. #71 Bernard J.
    March 20, 2009

    BTW, if it turns out that poor siting creates uncertainties that create an error bar of, say 2C, what would that mean for the significance of the 20th century (post-adjustment) warming trend, as estimated (a trend of 0.006C – 0.008C per year)?

    Quite possibly nothing at all.

    Have you ever sat through a first year ‘accuracy’ versus ‘precision’ lecture?

  72. #72 Barton Paul Levenson
    March 20, 2009

    I give up. Evan Jones is unteachable.

  73. #73 Gator
    March 20, 2009

    Evan Jones at 262. If you start out by not understanding what you are talking about, I doubt any of the rest of your argument will be worth anything. Saying “it’s an absorber, it’s a heat sink, whatever” is the equivalent of someone confusing power and ground or a resistor and a capacitor in an electrical circuit. If they then go on to tell me about their magical circuit, I’m not going to waste much time on them.

    I doubt a driveway is going to drive down Tmin. In any case, people observe Tmin increasing, right, so this goes against your simple argument.

    By trend, I mean specifically changing temperatures over a long time. Like what is observed. I can see how adding a driveway near a thermometer could increase Tmax, but why would that increase change over a long time? You will see a step change when the driveway is put in (which should be easy to spot in the data and correct for.) But this would not create an increasing trend, which is what you need to argue for if the thermometer signal we see that indicates global warming is simply installed driveways.

  74. #74 Chris O'Neill
    March 20, 2009

    Evan jones:

    Pretty much the same trend as 1915 to 1945, which was without much CO2 increase, depending on your proxy.

    So you think this is proof of same cause? If you do you’re wrong. BTW, why do you ignore the fact that there has been more recent warming than from 1910-1940. Some intellectual honesty would be much appreciated.

    And the curve is currently humping towards the cool side, not headed the other way at present.

    No it’s not.

    If we are going to hit that +3.5C by 2100, we’d better get a move on.

    We don’t need 3.5C to have serious problems.

    We have only 90 years to gain that 3.6C.

    We already have 0.75C.

    We’ll need +0.4C/decade for that,

    More like 0.3C/decade.

    and we’ve been doing well under half that for the last three decades.

    More than half what’s required for a massive disaster actually.

    It all comes down to whether that CO2-positive feedback loops theory actually works out.

    Not just a theory actually. Observations confirm it.

    (So far the Aqua Satellite sees the opposite effect,

    Don’t believe everything corporate shills like Marohasy tell you about results from this satellite. Contrary to what Marohasy says, water vapor is increasing with temperature.

    but those results are still preliminary. Dare I say that more study is required?

    You can certainly do more study to try to determine if releasing yet more CO2 is going to be safe for the earth’s surface temperature. However, it’s not looking too promising at the moment that releasing more CO2 will be safe. You’ve got the right idea that we shouldn’t take unknown risks with the earth before more study is done.

  75. #75 Former Skeptic
    March 20, 2009

    Evan Jones:

    And are not all effects (both cool park and hot spot deals) all meat for the average?…I am concerned with UHI effects on temperature data (Tmax & Tmin) more than the inner zen of UHI, itself. So why would one not use the mean to determine that?

    You are contradicting yourself. You claim to be concerned with UHI effects on temp data without wanting to know what causes, modifies or influences UHI in the first place.

    Watts up with that?

    It’s like wanting to examine statistical significance without knowing which test to use or how to apply the test. Not understanding the technique will make you look like an idiot (e.g. Pielke Jr.).

    If you understand near-surface UHI theory, you would know exactly why Tmax, DTR and Tmean are poor metrics to measure the urban effect on T. Even Pielke Sr. knows this. There are way too many factors that would distort and possibly increase the urban signal from these metrics, as I’ve stated in my previous post.

    (For that matter, how would one not use it?)…Ren simply compares urban (by size) data with rural. The data is standard max-min How do you get away from that?

    For someone who has admitted to not reading the Ren paper, you seem to know a lot about their methods. Ren and co. did not use standard max-min because, in China, “ordinary stations do not take measurements during nighttime. (p. 1335)” So they had to use mean T, which as I said, is the wrong metric to use. Remember what I said? Please read the papers – especially the respective methods sections – before making such assumptions.

    I read LaDochy a year back, but only the abstract for Ren, so far. I’ll get to it. (Do you recall offhand if they include a TOBS adjustment?)

    Offhand? No.

    At any rate the UHI adjustment would have been minuscule–about 5% of the observed difference for Ren

    Is that fact, or your uneducated guess? What was that I said about not making assumptions prior to reading a paper and not just the abstract? But hey, if you persist in making these guesses, go ahead, no one’s bothering to listen.

    Hmmm. If it’s flawed, then why is Dr. Jones buying it?

    I presume you are talking about Jones et al. (2009)? Forgive me if it’s the wrong paper.

    You have to ask Phil. I don’t buy it at all – especially when Phil’s group does not explicitly explain – despite their promise to do so in his intro paragraph – why he ditched minimum T records when he should know that it is the best metric of urban influence. Para. 18 of his article just merely talks about DTR trends, with no direct mention of Tmin trends, which is sloppy work IMO. While that’s not dishonest, that’s being economical with the truth.

  76. #76 Former Skeptic
    March 20, 2009

    Sigh. Chris O’Neill, dhogoza, BPL, Bernard J., Gator…I share your pain. It’s one thing to try to discuss and to educate; it’s another thing to try to discuss and to educate with someone who is intransigently happy in one’s ignorance.

    Oh well, I tried.

  77. #77 Gaz
    March 20, 2009

    …and your efforts are much apprecaited, Former.

  78. #78 Evan Jones
    March 21, 2009

    Well, you were doing fine until that last post. I can only read so fast, and I have a lot on my plate. I will get to it.

    If you understand near-surface UHI theory, you would know exactly why Tmax, DTR and Tmean are poor metrics to measure the urban effect on T.

    Okay, temperatures are taken at Tmax and Tmin. So that’s when UHI would affect the temperature readings. What happens in between, by definition, can’t. That much seems obvious.

    Looking at the charts in the Ren study, he is simply comparing the readings from urban areas of various sizes with each other and with rural data. This produces the difference. I still don’t see why that doesn’t work. (You’d want TOBS adjustments, if any.)

    And, yes, that’s the right Jones paper. (Note that he does not concede that the China conclusions apply elsewhere.)