Forgot your password?
typodupeerror
Science

Surfacestations: NOAA Has Overestimated Land Surface Temperature Trends 474

Posted by Unknown Lamer
from the climate-change-driven-by-hippies dept.
New submitter BMOC writes "Anthony Watts of Surfacestations project (crowdsourced research) has finally yielded some discussion worthy results (PDF). He uses a siting classification system developed by Michel Leroy for Meteofrance in 1999 that was improved in 2010 to quantify the effect of heat sinks and sources within the thermometer viewshed by calculation of the area- weighted and distance-weighted impact of biasing elements to calculate both raw and gridded 30 year trends for each surveyed station, using temperature data from USHCNv2. His initial claims are that station siting is impacting the surface temperature record significantly, and NOAA adjustments are exacerbating that problem, not helping. Whether you agree with his results or not, recognize that this method of research is modern and worth your participation in the review. Poke holes in publicly sourced and presented research all you can, that's what makes this method useful."
This discussion has been archived. No new comments can be posted.

Surfacestations: NOAA Has Overestimated Land Surface Temperature Trends

Comments Filter:
  • Oh dear... (Score:3, Insightful)

    by Shrike82 (1471633) on Tuesday July 31, 2012 @05:17AM (#40826389)

    He uses a siting classification system developed by Michel Leroy for Meteofrance in 1999 that was improved in 2010 to quantify the effect of heat sinks and sources within the thermometer viewshed by calculation of the area- weighted and distance-weighted impact of biasing elements to calculate both raw and gridded 30 year trends for each surveyed station, using temperature data from USHCNv2.

    Had to read that a couple of times before my internal parser came back with an approximate translation into lay-English.

    I fear that this will be ammunition for the climate change deniers, which if I understand correctly is not the intention here. The gentleman in question is merely pointing out possible bias and error and the open invitation is to critically analyse and see if his theory stands up. You know, like real scientific method! Still, I'll sit back now and watch the fireworks in what promises to be yet another pitched battle between the deeply entrenched sides in a war where actual fact is not nearly as important as name calling and idealogical strength of will.

    And the insults start in 3.....2......1......

    • Re:Oh dear... (Score:5, Informative)

      by Rei (128717) on Tuesday July 31, 2012 @05:25AM (#40826421) Homepage

      No, he's a well known denier (probably *the* most well known among lay people). He's not a climatologist; he's a local meteorologist for a small Fox affiliate in southern California. And this is unpublished, and will almost certainly be ripped to shreds when it gets submitted, like most of the other trash he submits. He's funded by the Heartland Institute (a conservative organization that takes industry money and uses it to push various forms of denial of interest to them, including things like global warming denial (funded by Koch Industries), denial of the links between tobacco and cancer (funded by Philip-Morris), etc.)

      Actual published, peer-reviewed work analyzing his "work" has reached precisely the opposite conclusion [noaa.gov].

      • Re:Oh dear... (Score:5, Informative)

        by Vintermann (400722) on Tuesday July 31, 2012 @05:47AM (#40826493) Homepage

        he's a local meteorologist for a small Fox affiliate in southern California

        A weatherman. In some countries, weather presenters are called meteorologists, but in general you need to have a graduate degree (heavy on math and physics) involving actual meteorological research to be called a meteorologist. Watts' highest completed education is high school, as far as anyone has been able to make out.

        • Re:Oh dear... (Score:5, Insightful)

          by gatzke (2977) on Tuesday July 31, 2012 @06:09AM (#40826563) Homepage Journal

          Attack the work, not the man or his (lack of) credentials.

          AFAIK, the work is relatively simple statistical analysis of time series data. No advanced science required. I have not looked in detail, but they claim that the adjustments made to climate data are biased.

          If this is an erroneous claim, it should be easy to demonstrate.

          • Re:Oh dear... (Score:5, Insightful)

            by Anonymous Coward on Tuesday July 31, 2012 @06:25AM (#40826643)

            But he didn't attack the man. He merely pointed out that he's not a meteorologist. In my country when I see somebody on TV forecasting the weather that person is a meteorologist, they have spent years studying the weather and know a lot more about the climate than I do. In the US they're just an actor reading from a script.

            The US does _have_ meterologists, who might know something about the climate, but this man is not one of them.

            • Credentials aren't everything. Consider the MCSE, for example. But as a general rule, someone who has documented evidence of study and experience in a field is more credible than someone who doesn't.

              Locally, our TV meteorologists are generally worth the title. They carried AMS certifications when it was the exception rather than the rule, and most of them have been here long enough to understand how the local weather patterns behave.

              That doesn't mean that I'd blindly accept statistics from them, however. T

              • Re: (Score:2, Insightful)

                by Anonymous Coward

                Credentials are, however, something.

                Watt's lack of academic credentials are not the primary reasons to dismiss his work. Those are, in rough order:

                1) His results contradict enormous amounts of well-scrutinized other scientific work on both the same and related topics, and explanations offered for that contradiction are grossly inadequate;
                2) His past results appear to have involved deliberate deception;
                3) His past results, and a casual reading of these results, suggest a serious ignorance of science;
                4) His p

            • Re:Oh dear... (Score:4, Insightful)

              by gatzke (2977) on Tuesday July 31, 2012 @08:43AM (#40827283) Homepage Journal

              He was dismissing the paper because one author appears to not be formally trained as a meteorologist.

              Again, this is more about data analysis than weather. You have hundreds (thousands?) of different stations of different quality. Some records are incomplete. Records stop and start at different times. Measurement devices change. Measurement times change.

              There is not much in the way of climate science in this, just putting data together. There are implications for climate science, however.

              The problem is to put all that data together intelligently. The current paper says the traditional methods are biased to the warming side. If this is not true, it should be refuted.

              If you followed the climategate issues, you might realize some established climate scientists appear to have biased their own work to fit an agenda. Some also say they have also worked the peer review system to unduly suppress opposition opinions.

            • by tomhath (637240)

              And this is unpublished, and will almost certainly be ripped to shreds when it gets submitted, like most of the other trash he submits

              Sure looks like an attack on the man to me.

          • And this will be done rather quickly I expect, by people who are much more competent to judge it than me. It's not quite a simple time series analysis - you also need some domain specific information which I haven't, related to physics and even history (e.g. the introduction of electronic thermometers).

            Also: some bloggers do go back to school, and get respected in their field of interest. Just not Watts.

        • Re: (Score:3, Informative)

          by Rei (128717)

          Not that it'd even matter anyway, as meteorologists aren't climatologists, and actually deal with very different phenomena. It's the difference between a biologist and a paleontologist.

        • but in general you need to have a graduate degree (heavy on math and physics) involving actual meteorological research to be called a meteorologist

          Actually it's a sub discipline or specialization of Geology, along with some other atmospheric sciences like climatology. You don't need to be a physicist or have any advanced math degrees to be competant in meteorology (like you might for atmospheric physics and fluid dynamics).

      • Why do we celebrate college dropouts who form spectacularly successful Internet startups then?

        Science is wonderful: not just the science that supports our world view. Articles on Slashdot frequently parade movements to challenge the science status quo (people making rockets in their garage, open source approaches to publication, etc. Perhaps his findings are garbage, but most of your post was nothing more than an ad hominem attack; your only reference to his research was a link to someone else's challenge,

        • by haruchai (17472)

          Spend some time on his site and you'll quickly learn just how much garbage he and his sycophants produce.

  • Let the bunfight begin.
  • The gist of it (Score:4, Informative)

    by Jesrad (716567) on Tuesday July 31, 2012 @05:22AM (#40826407) Journal

    The research has classified all the surface stations into 5 classes of relevance, from "reliably close to environment" to "poorly sited" in order to evaluate whether and how much the location of the thermometer and its proximity from airports, cities and so on would skew its measures over time. The end result is that there is a warming over time, but that warming is +0.155 C / decade using the best surfacestations, and twice that (+0.309C / decade) if you use them all.

    • Re:The gist of it (Score:5, Informative)

      by Jesrad (716567) on Tuesday July 31, 2012 @05:24AM (#40826417) Journal

      No wait, I read that wrong. It says there is a +.155C warming/decade using the best (classes 1 & 2) stations, +.248C using the worse stations (classes 3, 4 and 5), and that, somehow, NOAA managed to get a +.309C / decade result out of them, by adjsuting upwards the bad stations in order to make up for their poor fidelity, and THEN adjusted upwards the good stations so they would match the poor, adjusted ones.

    • by Rei (128717) on Tuesday July 31, 2012 @05:29AM (#40826443) Homepage

      It's not peer-reviewed. Everything else he's submitted for peer review on the matter of climate change has been ripped to shreds; why should this be any different?

      • Re: (Score:3, Insightful)

        by Anonymous Coward

        If it's public, then stop shouting and screaming and pointing fingers like little children and act like a proper scientist and show where it's flawed, if it is.

        In the manner with which you're currently acting, you would have taken the odd FTL results from the Italian physicists to be true instead of pouring over it and finding what errors there were.

        Science has checking and verifying results as a major part of itself. Leave your bias and presumptions at the door.

        • by Rei (128717) on Tuesday July 31, 2012 @05:49AM (#40826499) Homepage

          If it's public, then stop shouting and screaming and pointing fingers like little children and act like a proper scientist and show where it's flawed, if it is.

          Which is exactly what the peer-review process does. Which is why you never trust non-peer-reviewed work. I can write whatever I want about anything, make it look like a paper, and then send it out to the media. Which is precisely what happened here.

          you would have taken the odd FTL results from the Italian physicists to be true

          You're walking down precisely the opposite road. Even one peer-reviewed paper on "remarkable claims" isn't enough - that's just the start of a process that can only be confirmed by a series of followup studies, spawning a process that can lead to dozens or hundreds of papers before one can feel confident in the truth of the matter.

          This here is *zero* published results.

          Science has checking and verifying results as a major part of itself.

          And the scientific process is the peer-review process, which this has not undergone, and will almost certainly fail like Watts' other "work". If he even bothers actually submitting it instead of just saying that he's going to.

        • by Rei (128717) on Tuesday July 31, 2012 @06:15AM (#40826589) Homepage

          To elaborate on the problem, I started reading the "paper" and he's outright misleading right on the first page. He says that siting in peer-reviewed works showed an effect on minimum temperatures but no effect on the mean. The actual papers show a small increase in minimum temperatures, but a much larger *decrease* in maximum temperatures. I'm also noticing in the paper him mixing in peer-reviewed cites with non-peer-reviewed cites without even commenting on the fact that he's doing so, which is a huge no-no.

          Basically, his previous work not having shown what he claimed it showed after the peer-reviewed process got ahold of it, he simply changed his formula until it showed a different result. Which will almost certainly get likewise ripped up.

          Here's the reality of the situation. The many papers published on the subject of the land record and all of their reviewers are not idiots ignorant of Watts' rogue genius. The issues that he "raises" have been discussed and analyzed for ages. Because of these issues, nobody just takes the raw data and submits it as a result. There are all sorts of calculations to detect biases and compensate for them, and all of these adjustments are analyzed with higher-precision real-world data to see how well they work, as well as cross-correlated with totally different lines of measurement. One study, to pick a random example among many, broke the data down between windy days and calm days, as the urban heat island effect dramatically diminishes on windy days. The calm results were then compared with the windy results to see if they reached the same conclusion.

          Of course, it should be obvious that Watts is wrong just by even a rudimentary look at the surface warming trends [nasa.gov]. Notice where they're strongest, generally? Sparsely populated areas. We're supposed to believe that the extreme warming of Siberian or Canadian tundra is due to a "urban heat island effect" not visible in, for example, New York, Tokyo, London or Los Angeles?

          Needless to say, you don't just have to judge based on your eyes; this has been statistically analyzed and published as well.

          • by Troed (102527)

            Of course, it should be obvious that Watts is wrong just by even a rudimentary look at the surface warming trends [nasa.gov]. Notice where they're strongest, generally? Sparsely populated areas. We're supposed to believe that the extreme warming of Siberian or Canadian tundra is due to a "urban heat island effect" not visible in, for example, New York, Tokyo, London or Los Angeles?

            From the press release (didn't you say you read it?):


            Other findings include, but are not limited to:

            * Poorly sited station trends are adjusted sharply upward, and well sited stations are adjusted upward to match the already-adjusted poor stations.
            * Well sited rural stations show a warming nearly three times greater after NOAA adjustment is applied.

            • by Rei (128717) on Tuesday July 31, 2012 @06:50AM (#40826733) Homepage

              Which are, of course, un-reviewed claims, and totally distort the picture (the reason for the adjustments and how they're tested was just discussed in the post right above yours - to sum up, people looking at the GISS dataset aren't idiots and know about the various ways station data can be biased, and have automated algorithms to detect and correct for bias - algorithms which have been rigorously tested by peer-reviewed research, and it should be noted, actually yield a lower warming trend than the raw data, which also shows a greater rural warming trend than urban).

              • by Troed (102527)

                I think the point of the paper is that using new information, Leroy 2010, the way adjustments should be made becomes different. You referencing algorithms that do not reflect the conclusions from Leroy does not in any way disprove that.

                Why are you so hostile towards science?

  • Disclaimer: I did work in atmospheric science for 15 years.

    I understand all this talk about adjusting temperature results for urban sprawl around the measuring stations, but bear in mind that those stations are weather forecast stations, never intended as climatology primary source of inputs. So why don't we simply use a better designed system, such as a thermometer a couple of feet inside the ground: depending on the depth you can average out the daily thermal cycle (a few inches) or even the yearly cycles (a few feet). And there you have your reliable long therm^Hterm trends without any supercomputers or fancy models.
    • by Rei (128717) on Tuesday July 31, 2012 @07:15AM (#40826817) Homepage

      Indeed, that's why there is the (growing) Climate Reference Network [noaa.gov]. The USCRN is a smaller subset of stations which are carefully chosen in terms of siting and instrumentation and carefully monitored in a way that couldn't realistically be done with all stations. The results from the USCRN are then compared with the broader results in both localized and aggregate comparisons and used A) to help refine the adjustment algorithms used to detect and compensate for localization biases, and B) to determine the accuracy of the aggregate results.

      • Indeed, that's why there is the (growing) Climate Reference Network [noaa.gov]. The USCRN is a smaller subset of stations which are carefully chosen in terms of siting and instrumentation and carefully monitored in a way that couldn't realistically be done with all stations. The results from the USCRN are then compared with the broader results in both localized and aggregate comparisons and used A) to help refine the adjustment algorithms used to detect and compensate for localization biases, and B) to determine the accuracy of the aggregate results.

        Mod this one up if you have the points.

    • by Mindcontrolled (1388007) on Tuesday July 31, 2012 @08:00AM (#40827051)
      Borehole reconstructions [noaa.gov] are routinely done and consistent with other proxies as well as with the instrumental surface record.
  • First Problem (Score:4, Informative)

    by Stirling Newberry (848268) on Tuesday July 31, 2012 @06:26AM (#40826651) Homepage Journal
    If siting were the problem, then temperature variance would track suburban sprawl and urbanization closely, as it would be a systematic error. It doesn't [noaa.gov]. Instead it tracks mountain regions with greatest snow cap loss, as would be predicted by AGW.

    AGW wins again on the data.

  • WTF is a Surfacestation ? Is it the workstation version of Microsofts new tablet?

COBOL is for morons. -- E.W. Dijkstra

Working...