Prat watch #5: Ignorance is bliss

What happens when you deny things? Well, if you deny the reality of global warming, and if you are to be in any way self-consistent, then you have to deny every bit of evidence that it might be happening. Here’s a classic example, drawn from New Zealand’s very own little corner of the climate crank echo chamber, Richard Treadgold’s “Climate Conversation Group” blog. Treadgold concludes a recent post thus:

Once more: let’s stop accepting this palpable nonsense that climate change is responsible for anything.

Climate change means global warming. Global warming has not happened for about 15 years, unless you take a micrometer to the thermometer. And if you have to do that just to detect warming, then it’s hardly dangerous, is it?

Oh – if it didn’t happen, then it didn’t cause anything! No droughts, no wildfires, no floods, no storms. No ice melt.

Look at the bit I’ve emphasised. No warming for 15 years? Tell that to the planet, Richard. Here’s what the World Meteorological Organisation says about the first decade of the 21st century:

…climate change accelerated in 2001-2010, which was the warmest decade ever recorded in all continents of the globe.

No warming for 15 years? After we’ve had the warmest decade ever recorded in all continents of the globe?

The decade 2001-2010 was the warmest since records began in 1850, with global land and sea surface temperatures estimated at 0.46°C above the long-term average (1961-1990) of 14.0°C. Nine of these years were among the ten warmest on record. The warmest year on record was 2010, closely followed by 2005, with a mean temperature estimated at 0.53°C above the long-term average. It was the warmest decade ever recorded for global land surface, sea surface and for every continent.

No warming for 15 years? Not to put too fine a point on it: balderdash, piffle, stuff and nonsense.

And because there has been warming, it can be linked to extreme events1. The IPCC has just released its full SREX report, full title Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation (PDF – 44MB), which finds, amongst other things, that not only is there a clear signal of warming’s effect on extremes such as heatwaves, but that large parts of the world — especially coastal megacities such as Mumbai or countries like Bangladesh and Vietnam — are vulnerable to increasing extremes and sea level rise over the course of the coming century. [AP/NZ Herald, AFP.]

The certainty of Treadgold’s denial is only possible because he creates a carefully cultivated cocoon of ignorance around himself and around the true believers who worship at his blog. The world where the rest of us live is a much more uncomfortable place. We have to work with whatever reality throws at us. Retreating into a fantasy world where warming hasn’t happened for 15 years is a luxury only the deluded can afford.

[Jellyfish]

  1. I’ll have post examining recent work on the attribution of extreme weather events soon. []

104 thoughts on “Prat watch #5: Ignorance is bliss”

  1. The SREX WGII summary video ‘In Harm’s Way‘ is worth watching.

    I doubt that there’ll be any impact on the ineducable beyond increased frothing events, but the world just stubbornly refuses to follow their script, and they can only be swept along by the deadweight of their ideology into an ocean of irrelevance…

  2. What Gareth ? Are things so slow on the doom and gloom front that you’ve had to resort to reviewing sceptical blogs ? Well I never…

    Heard you on Pat Brittenden’s show by the way.

    Mmmm. Not your forte Gareth.

      1. I must admit I had never heard of Mr Brittenden, and checking out his recent guests on his Twitter feed – Ian Wishart, Cameron Slater, etc, I thought, hmm probably not

  3. If you want to keep being syndicated at Sciblogs I’d spend a bit of time working at understanding the difference between the first and second differential with time.

    1. It is statistically quite likely to be “the hottest year/decade on record” whether you are in a warming trend, a plateau, or even just beginning a hypothetical cooling period.

      It’s the likes of the WMO etc who don’t qualify these statements that are the real culprits though.

      1. Statistically quite likely? I don’t know what kind of statistics you use, Andy, but I’d really like to know how you work that out.

        To test how likely it is to be in the hottest decade in 100 years, I ran a Monte Carlo simulation. I computed 100 numbers with a mean of 14.0, with variability of ±0.5, plus an annual trend, then calculated averages for each of the 10 decades. I then ran that 10,000 for a variety of annual trends, and calculated the proportion of time that the last decade was the warmest. The results were as follows:

        Trend: 0.000; Last decade warmest: 525/10000 (5.25%)

        Trend: -0.005; Last decade warmest: 0/10000 (0.00%)
        Trend: -0.010; Last decade warmest: 0/10000 (0.00%)
        Trend: -0.015; Last decade warmest: 0/10000 (0.00%)
        Trend: -0.020; Last decade warmest: 0/10000 (0.00%)

        Trend: 0.005; Last decade warmest: 3537/10000 (35.37%)
        Trend: 0.010; Last decade warmest: 4848/10000 (48.48%)
        Trend: 0.015; Last decade warmest: 7144/10000 (71.44%)
        Trend: 0.020; Last decade warmest: 7877/10000 (78.77%)
        Trend: 0.030; Last decade warmest: 9185/10000 (91.85%)

        So, with no trend, only 5% of the time was the last decade the warmest. That’s what you call “statistically quite likely”, is it Andy? I wonder how likely it is for the last four decades to be progressively the warmest in the absence of a warming trend – but maybe with your vast statistical knowledge, you can tell us that, eh?

        1. Of course, Andy will claim that he’s not saying that there has been no warming trend, just that it stopped or reversed in the last 15 years, so I also ran the simulations with two more scenarios, with no trend for the last 15 years, and then a negative trend for the last 15 years:

          No trend last 15 years:
          Trend: 0.000; Last decade warmest: 450/10000 (4.50%)
          Trend: 0.005; Last decade warmest: 1935/10000 (19.35%)
          Trend: 0.010; Last decade warmest: 3970/10000 (39.70%)
          Trend: 0.015; Last decade warmest: 4219/10000 (42.19%)
          Trend: 0.020; Last decade warmest: 4250/10000 (42.50%)
          Trend: 0.030; Last decade warmest: 5226/10000 (52.26%)

          Negative trend last 15 years:
          Trend: 0.000; Last decade warmest: 552/10000 (5.52%)
          Trend: 0.005; Last decade warmest: 806/10000 (8.06%)
          Trend: 0.010; Last decade warmest: 1283/10000 (12.83%)
          Trend: 0.015; Last decade warmest: 1764/10000 (17.64%)
          Trend: 0.020; Last decade warmest: 161/10000 (1.61%)
          Trend: 0.030; Last decade warmest: 58/10000 (0.58%)

          So, looking at the three scenarios with a trend of 0.015, we get:

          Warming: Last decade warmest: 7144/10000 (71.44%)
          Plateau: Last decade warmest: 4219/10000 (42.19%)
          Cooling: Last decade warmest: 1764/10000 (17.64%)

          You certainly have an odd idea of statistics if you think 71% is just as likely as 18%, Andy.

          1. I appreciate you taking the time to run these simulations, and correcting your initial assumptions to fit the actual scenario.

            The issue is that it is still quite likely (42%) that last decade was the warmest on record, even with a plateau in temperatures.

            Matt Briggs wrote on this issue sometime back.

            1. “The issue is that it is still quite likely (42%) that last decade was the warmest on record, even with a plateau in temperatures. ”

              I thought you had an MA in maths andy? 42% is not a probability you would be betting on is it andy. I’d be betting on a 71% chance, but would be holding my money on 42%, or have you forgotten the Chevalier de Méré?
              And is the description “quite likely” an honest description when the chances are less than half? I think that description is indicative of your rose coloured lenses andy.

            2. Macro:
              Lest you think ill of a fine school…
              You might want to look up how one gets a BA, MA Math @ Cambridge, because the MA does not have the same meaning as it does at many schools.

            3. John Mashey March 30, 2012 at 3:57 pm

              You get an MA by living for 3 years beyond getting a BA.
              What relevance does this have to anything?

              In fact, what is the point of this thread at all?

            4. Gees, now we know andyS is not as bright as he thinks he is since it took him 3 years to do a one year Master of Mathematics degree post a BA.

              No wonder he has trouble with statistics and numbers. In case he accuses me of lying or being wrong here is a link to the Cambridge Masters of Mathematics description:

              http://www.maths.cam.ac.uk/postgrad/mathiii/

            5. Ian: to clarify, a Cambridge BA/MA doesn’t work that way, see also this. Basically, students get both BA+MA for a 3-year program, but the MA is awarded later.

              Of course, this is a good math degree, but, at least decades ago, learning about Navier-Stokes in undergrad math @ Cambridge really wouldn’t teach much about practical issues of computational fluid dynamics. For very few real problems is it possible to “solve” N-S, hence one must use wind tunnels or far more likely, computer simulation. Some problems are mostly CFD, for others (like climate models) there are numerous other factors.
              [This is why I said N-S was a red herring.]

              Serious problems:
              – Aerospace design at NASA Ames & elsewhere
              – Auto design, both of exterior and of engines, etc.
              – Reservoir modeling at petroleum companies
              – Climate models at NCAR, GFDL, NASA, etc.

              Some fun problems:
              – NZ’s Black magic boat was designed with SGI workstations, then tuned up over time in the San Diego boathouse. They’d come back from practice, run simulations of proposed changes over dinner, come back and see results, use computer-based machine tools to fab new parts.
              Great sailors, but also ability to simulate and adapt the design much quicker than the other teams that used supercomputers for original design.
              – Barbie dolls have sometimes been designed with combined CFD + heat codes, i.e., plastic injection molding versus too-small waist can cause breaks.
              – My favorite visit: UniLever’s chocolate ice cream bars.
              [Hot chocolate melts ice-cream, which eventually gums up the works, requiring shut-down and cleaning. Simulation allowed the shapes to be optimized to lessen the frequency.]

              Serious or fun, depending on view
              – There is a a great deal of computational physics in movie animation and games. i.e., in NZ, Weta Digital.

            6. John Mashey March 31, 2012 at 6:28 pm

              So why are you explaining this to Ian when I have already explained it in this thread?

              Is there some special code where you smug, sneering, supercilious pricks need to talk over this rest of us?

            7. As opposed to the “smug pricks” like you Andy, playing politics with the future of the planet?

            8. If I didn’t know better, andy, I’d rather suspect that you felt threatened by John M, who’s, after all, something of a giant in the very field where you’re… something of a minnow, perhaps? 😉

          2. Macro March 30, 2012 at 3:04 pm

            Let me give you another example. I drive my car out of the garage. I head towards a motorway, where I accelerate to 100kph.

            Since I am a good citizen, I don’t speed, but occasionally I might hit 101 kph, 102 kph, sometimes 98 kph.

            At any point on that journey, the 102 kph could be described as the fastest speed on record for that journey even thought the trend is zero.

            This is the point I am trying to make. Do I need more examples?

            Oh and by the way 42% is quite likely in my books. 71% is more likely, but it is completely irrelevant to the discussion.

            1. Your original statement was:

              It is statistically quite likely to be “the hottest year/decade on record” whether you are in a warming trend, a plateau, or even just beginning a hypothetical cooling period.

              You were implying that it was equally likely that the warmest decade on record could happen in the the three scenarios. I have shown that it is not equally likely, in fact by far the most likely explanation is that the warming trend has continued to date.

              You then castigated the WMO for saying that the most likely explanation was the most likely explanation. Silly of them really. Of course, they should have given equal consideration to the equally likely possibility that the warming trend is due to the effects of leprechauns.

            2. CTG

              “It is statistically quite likely to be “the hottest year/decade on record” whether you are in a warming trend, a plateau, or even just beginning a hypothetical cooling period.”

              Just on the logic I’d say it was a safe bet on the first and the last and 50:50 on the plateau in the absence of any further info.

              If however this is an empirical statement relating to where we are today one does have to say you need to make the assumptions you have made to reach the conclusions you do. I could assume leprechauns as you have assumed trends and get to the opposite conclusion as it were.

            3. “a mean of 14.0, with variability of ±0.5, plus an annual [linear] trend”

              I assume leprechauns

            4. Oh, I see, you don’t understand Monte Carlo. I was testing the assertion that it is just as likely that the warmest decade on record would occur if the warming trend is continuing, flat or reversed. I wanted to produce data that had similar characteristics to the temperature record, I wasn’t trying to replicate it.

              Nowhere did I claim that the actual temperature record has a linear trend with exactly the same slope over 100 years.

              So the only person making assumptions is you. Oh, and andyS.

            5. You didn’t claim it, it you assumed it.

              I don’t claim leprechauns, I assume them.

              I understand Monet Carlo, contrary to your assumption.

            6. Actually there is a point to all this. As I said in a comment a little while back you can just look at the words andyS used and he’s probably right if you separate them from any empirical information. You however wanted to make the point that if you fitted a model to the proposition you could show something useful about whether the assertions were true of false. You picked (assumed) a model and ran a simulation and achieved some results (just as climate modellers do).

              The point is that this simply tells us something about the model you chose, not the real world. My point was that I could choose a different model (my leprechauns) and without even needing to use Monte Carlo could get whatever I wanted.

              You assumed linear trends and got a result. That is neither here nor there in having a discussion about what might happen in the real world, particularly since we are pretty certain that your model doesn’t represent the real world system under discussion (the jury is still out on my leprechauns – so my model is probably better).

              I had hope that you might have challenged me on the basis that your model is more likely than leprechauns to describe what is happening in the real world, but instead you have trumped my leprechauns with trolls.

            7. Sorry, Simon, but andyS is still wrong. He offered three scenarios, in a warming trend, in a plateau or in a cooling trend. It is not obvious from his words whether this only refers to the last decade of a period of warming, or just as a general comment about time series.

              So, the Monte Carlo simulations present some synthetic data that let me make some observations about time series in general, and about those three scenarios in particular.

              In my first set of runs, I used three models, none of which show the slightest resemblance to the real world, by intention. The three models are: continuous positive trend, no trend and continuous negative trend. These represent the general cases of andyS’ statement. The first runs showed quite clearly that andy’s statement is false for the general case – in the no trend and negative trend cases, it is highly unliikely that the last decade will be the warmest.

              Then there is the specific case, that following a period of warming, you would still be equally likely to end up with the last decade being the warmest if the last 15 years of the trend continued warming, entered a plateau or started cooling. Now, to investigate this, I used the linear trend model – again, not because it resembles the real world, but because it produces synthetic data that we can make interesting observations about. Once again, we saw that andy’s comment does not hold up – the most likely case is that the warming is continuing – it is nearly twice as likely as that the trend has stopped.

              Now, and only now, can we start to look at the real world situation. The real world temperatures have, undeniably, been in an approximately linear warming trend for the last 40 or 50 years or so. Or do you deny that? If you deny that, we’re done, because if you don’t accept the observational evidence, you are not doing science.

              Given that the original post is about the statement “there has been no global warming for 15 years”, we can use the simulation data to make some observations about that statement. The WMO has said that the last decade is the warmest on record (as were the three decades before, in their turn). But perhaps global warming really did stop in 1998, and we are now in a period of no trend, or even sliding towards a deadly ice age, as Bryan Leyland claims? So of those three scenarios, which is the most likely? Are all three equally likely, as andyS says?

              Based on the simulations, I say that by far the most likely explanation is that the warming trend is continuing. It is certainly possible that the trend has stopped, but only half as likely. As far as the ice age goes, that is a pretty remote possibility.

              See? No assumptions, just observations. That’s how science works, Simon and andyS.

            8. “See? No assumptions, just observations. That’s how science works, Simon and andyS.”

              A couple of observations. In the first half you make it it clear that the model you chose had nothing to do with the real world. As I said before all your models tells us is how your models perform. I trust we can agree on that.

              Second you move to claim that .. well .. really they do represent the real world. You go to ask: “Or do you deny that? If you deny that, we’re done, because if you don’t accept the observational evidence, you are not doing science.”

              Now I’m interested in the observational evidence and as I said above we basically know that the temperature record doesn’t conform to your model.

              So all you did was produce a model that probably wasn’t modelling what andyS had in mind, nor the real world systems (if they were something different), and demonstrate something about how that model performs.

              Arcane and not empirical. The missing step is validating your model against the real world before drawing inferences from it. Exactly the missing step that allows the use of leprechauns or trolls to make the same explanation, and that bedevils climate modellers..

            9. Still way off, Simon. First off, this was just a quick calculation to check the claim by andyS, not a scientific paper, so a close approximation is good enough. If anything, I probably overestimated the natural variability term, which would tend to lowball the probability estimates. Second, I tested over a range of different trends to see what the general effects were. The most interesting results were for a trend of 0.015, because guess what the real world trend has been for the last 30 years.

              I notice that neither you nor andyS have come up with any actual statistical evidence to back up his original claim. You know, where he said it was “quite likely” that temperatures that had been in a plateau for 15 years would have the warmest decade last. I don’t know how andyS defines “quite likely”, but here is how the IPCC defines different likelihood terms. Likely is a probability of at least 66%, very likely is >90%. It doesn’t have “quite likely”, but normal English usage would suggest “quite likely” ought to be somewhere between likely and very likely, so maybe >75%?

              Unfortunately for andyS, 42% is in the “About as likely as not” zone.

      1. Perhaps the most obvious sign of all the heat that the oceans have absorbed over the last 15 years, is the intensification of ENSO (El Nino/La Nina) that seems to be taking place.

        As the oceans grow warmer, evaporation from the oceans grows stronger, which intensifies the rainfall and drought experienced in many parts of the world. Climate modeling suggests that: a) ENSO should intensify through the 21st century. And b), that this process should already be underway.

        This is consistent with the paleodata going back 1100 years – the tropical Pacific Ocean was cooler during Medieval times, which led to a much weaker ENSO compared to modern day. ENSO intensified in the late 19th century and has been more intense ever since – although it does have an internal cycle, of waxing and waning, of around 60-80 years.

        In other words, expect weather to become more extreme in the future. Not just heatwaves, but ENSO-driven floods and droughts too.

  4. Just seemed to me that the headline was about being a prat, and then proceeded to argue against the statement “no trend in temp over last 10 years” by arguing there had been trend in the decades before (and ignoring if there was +ve feedback from CO2 then temp should have been accelerating).

    Has the globe warmed significantly over the last decade? Perhaps read http://wmbriggs.com/blog/?p=5154 for amusement. Hardly a daft idea then that it hasn’t.

    Whether there is positive feedback from CO2 and if so is it linear with temp is also a useful question to debate – it makes a lot of difference..

    1. Treadgold did not say there was “no trend in temp over last 10 years”, he said there had been no global warming for 15 years. Global warming = heat accumulating in climate system, and that this continues can be seen across a range of metrics, not just global average temperature (see DW’s reply above).

      1. This is a bit pedantic really. He did talk about thermometers in the next breath. DW does tell us what climate models say on the subject, but given they don’t do ENSO well its all a bit like CTG and his efforts. You pay your money and take your chances.

        1. Nowhere did Treadgold make any kind of statement about trends or measurements. His was a content-free rant, and it made a good peg for the WMO and IPCC stories – both of which directly contradict his argument.

          Of course, if you want to support the man, that’s your business. Don’t expect anyone to take you seriously if you do.

          1. Just for the record I haven’t got the faintest idea who Treadgold is or you for that matter. I’m just judging what I read here.

            I don’t think saying “Global warming has not happened for about 15 years” is content free (and rather trivially involves statements about both trends and measurement). It is obviously exercising a good many minds well beyond just the punters here.

            I sort of support working out what’s going on – don’t you?

            1. I think, Simon, you should perhaps visit Mr Treadgold’s blog and read the material there, before making any judgements about who is the more interested in working out what’s actually going on.

              For the record: the climate system is complex and fascinating, but pretending that heat is not accumulating in it is a straightforward denial of the evidence. I’ll concentrate on that, wherever it takes us, rather than indulge in wishful thinking.

            2. It is, as we say, obvious that heat is accumulating in the atmosphere. The interesting issue is what is the atmosphere (and oceans) doing about it?

            3. Don’t you just marvel at people who can speak such blatant tosh with such extraordinary confidence?

              SREX Summary for Policy Makers Section B P9:

              Observations of Exposure, Vulnerability,
              Climate Extremes, Impacts, and Disaster Losses

              Followed by several pages on what it says on the label. Lots of obs with rough odds on attribution.

              On page 12 there’s a nice chart captioned as follows –

              Figure SPM.4B | Projected return periods for a daily precipitation event that was exceeded in the late 20th century on average once during a 20-year period (1981–2000)… The box plots show results for regionally averaged projections for two time horizons, 2046 to 2065 and 2081 to 2100, as compared to the late 20th
              century, and for three different SRES emissions scenarios (B1, A1B, A2) (see legend).

              On page 13 there’s another pretty chart that shows ‘Change in consecutive dry days (CDD)’ for the same periods (2046 – 2065 and 2081-2100), then right next to it there’s the same for ‘Soil moisture anomalies (SMA)’.

              Oh, I know, it’s all too far in the future, even if it’s not quite so distant as the casual wave of your hand seeks to imply; what’s much more important is to be self-important now, isn’t it? 😉

              If only you were as competent as you are confident, eh? Or were you merely misled by blogs? Happens a lot, you know…

            4. Gareth and bill

              I can see no option but to bow to your superior wisdom on this matter – the SREX report has nothing to with the Climate in 2100 as projected by climate models (although bill your comment did seem to contain a number of references to the latter).

              I should add I’m actually not particularly confident about much looking at climate science. In particular I’m not as confident as the IPCC about the performance of GCMs outside being useful labs in which to test assumptions (i.e. they have limited value as a medium-term forecasting tool), nor am I particularly confident about the feedback mechanisms around CO2.

              Are you two confident about these things?

              And that’s why I’m more interested in what’s happening now rather than 2100.

            5. The point I was making, and which you are avoiding, Simon, is that the SREX specifically reports evidence of anthropogenic influence on current extreme weather events, and is not solely concerned with future projections.

            6. Didn’t see that, can you point me to one or two – I suspect that what it reported was the impact of recent warming on extreme events, but I’d be really keen look at literature that established recent increases in human CO2 emissions, for example, were the cause.

              It’s all to do with understanding the feedback mechanisms and attribution.

            7. There seems to be this remarkable ongoing phenomenon with ‘Skeptics’ that they cannot parse any statements, at least where those statements contain information they do not wish to understand.

              You:

              It deals with what climate models say might happen in 2100, not what is happen[ing] right now.

              Me:

              SREX Summary for Policy Makers Section B P9:

              Observations of Exposure, Vulnerability, Climate Extremes, Impacts, and Disaster Losses

              Followed by several pages on what it says on the label.

              (And I demonstrated it dealt with a range of projections based on time periods starting from mid-Century.)

              You [sarc]:

              I can see no option but to bow to your superior wisdom on this matter – the SREX report has nothing to with the Climate in 2100 as projected by climate models (although bill your comment did seem to contain a number of references to the latter).

              [/sarc]

              Do you think logic is your thing at all, Simon?

        2. Actually the climate models merely affirm what basic physics tells us – a warming ocean will lead to greater evaporation and consequently greater extremes between La Nina & El Nino. Circulatory changes do little to ameliorate this trend – according to the models which do a better job of simulating ENSO well. The paleoclimate proxies support this too.

          Over 90% of global warming goes into the oceans – they cover over 70% of the Earth’s surface, are dark-coloured (greater heat absorption) and have a heat capacity a thousand times that of the atmosphere. If all the energy that went into the oceans in the last 50 years, went instead into heating the atmosphere, global temperatures would have risen by 42°C!!.

          It takes a heady dose of denial to claim that the Earth hasn’t warmed in the last 15 years.

          1. A couple of genuine questions:

            Which climate models do a better job of simulating ENSO well (and rather how well do they do it)?

            On what basis do you say we know that the Earth (I assume atmosphere and oceans) has warmed in the last 15 years. Is this on a flux basis (more coming in than going out) or absolute measurement of the energy content of each of the components. I had though both were matters of intense debate.

    2. Another way to look at this, if you insist on global temperature as your metric, is to consider running averages:

      The 5yr mean allows (roughly) for the ENSO cycle, and the 11 year mean for the solar cycle. In no way does either support the contention that there has been no warming for 15 years.

      1. All you do with running means is reduce the information available with which to draw conclusions. If you didn’t notice a 132 month running mean just eliminates 11 years from the 15 year period. I’m sure that’s not what you meant to do.

          1. How many points on the new series don’t contain information from before the 15 year period? To be clearer – great if you want to cut off one end and contaminate the other with the previous trend. As I said I’m sure that’s not what you meant to do.

            1. Since the last point on the 11 year curve is based on all the last 11 years of data, it’s not “contaminating” anything – it’s including one full solar cycle’s worth of data and therefore removing that contamination.

            2. You see the funny thing is that using a moving average to take out the solar cycle assumes all solar cycles are the same. They aren’t. Using moving averages is a fudge that neither deals well with the cycle or improves the quality of the time series (as you can see it hides the flattening off over the last 15 years). If you want to deal with the solar cycle we know what the values were – deal with it directly (if you can deduce what the effect is).

            3. Simon The 132 month running mean is done for a very specific reason – specifically to remove the “contamination” of solar cycles, ENSO’s etc. That is why it is a very revealing graph.
              (Incidentally it is not a graph generated by Gareth – it is taken from Dr James Hansen et al’s recent “GISS Surface Temperature Analysis Global Temperature in 2011, Trends, and Prospects”) http://data.giss.nasa.gov/gistemp/2011/

            4. Macro, I don’t think Dr Hansen would use this in any serious piece trying to deal with attribution, just for stuff written for popular consumption.

            5. What you think of Hansen’s motivations is of no consequence. You might instead read what he has written.

            6. Gareth, I do I do.

              The word “contamination” was introduced into this conversation by me when you started using moving averages into a discussion about whether there had been a change in temperature trends 15 years ago.

              Hansen arrived much later, perhaps by way of justification, introduced by Macro. He incorrectly suggests moving averages are a way to remove “contamination” – as Hansen notes it is to reduce the variability – and it does this at the cost of removing data from the end of the series and the inflection point (if there is one). Thus obscuring the discussion that started all this. Even Hansen begrudgingly acknowledges the possibility of less warming in the 21st century even eye balling these graphs.

              To just round off, the discussion has now moved from “was the trend flat for the last 15 years”, to “was this caused by the combined effects of ENSO, solar et al”. This is the point at which you stop using crude tools like smoothing and do more sophisticated attribution studies.

              OK?

            7. a discussion about whether there had been a change in temperature trends 15 years ago

              That is not what was being discussed! Nice try on the goalpost move, though. The subject of this discussion is Treadgold’s assertion that there has been no global warming for the last 15 years.

              You seem hell bent to make it seem as though there is some basis for supporting that contention. Either provide that evidence, or stop trolling this discussion. (See comment policy, specifically the “good faith” provision).

            8. Gareth

              Sorry that was an abbreviation. The discussion is about as you correctly say whether the trend has become flat in the last 15 years. I don’t think anyone was suggesting it was flat before then, so it is ipso facto about whether it has leveled off 15 years ago.

              With that clarification my comment remain.I am a bit surprised about you feeling a need to draw attention to the “good faith” policy, as if a robust debate about what’s been happening to temperature, various statistical techniques, climate and other models were bad faith.

              Perhaps I had the wrong use of the world “faith”. Perhaps we’ll leave it here for now.

            9. No, Simon, it’s not about “the trend has become flat in the last 15 years”, it’s about “no global warming for 15 years”. As I pointed out in an earlier reply, global average temperature is only one metric of the heat accumulating in the system. There are many others. Ocean heat content, as DW points out, is the single most important factor because that’s where the vast majority of the heat goes. Measurements of that are improving all the time, and show plenty of evidence of continued increase, both currently and over the last 15 years.

            10. Simon March 30, 2012 at 10:30 pm
              Just to make it clear – it was YOU who introduced that extremely “technical” term “contamination” to the discussion, not me. I merely used, it as it seemed to be the word you were most familiar with.
              I’m also well aware of what the use of the running mean implies and the effects – but what you fail to notice is that the graph is in fact two graphs a 132 month running mean, and superimposed a 60 moth running mean. They are almost identical (or were you trying to avoid that observation) so all your huffing and puffing regards the end points etc with regard the 132 month running mean, and “contamination” (to use your term) is all just that – so much huffing and puffing! Any distortion is minimal. Furthermore – had you looked further in the article – you would have seen a graph of solar Irradiance for the past 50 years. This clearly shows that solar Irradiance is now increasing (having just past a minimum), and ENSO is moving towards El Nino, and that is why those who have been following this story for some time, are well aware that global temperatures are set to increase above previous records within the next few years, and all your huffing and puffing won’t stop it!
              It may be of interest to you (but probably not – noting your rather supercilious attitude) that the article to which I referred by Dr Hansen et al has been linked in a number of papers. But then – as it doesn’t support your wacky perception of reality – it can’t be up to much.

          1. You need to help me Rob. This is a graph that shows the smoothed temperature series. What’s the compelling take out – that if you smooth the average annual temperatures on 60 and 132 month basis you lose the bit at the end, and you end up with a smoother graph?

            I didn’t know sockpuppets used there own names – learn something every day.

            1. The compelling take-out, Simon, is obvious at a first glance at the graph. Only the ideologically blind and the intellectually dead could fail to see it.

              The 60 and 132-month averaging removes much of the short-term variation, as I’m sure you are aware.

            2. Goodness Simon, it really strikes me how people like yourself can spend SO much energy and enthusiasm trying to fight off something so obvious as the long term trend of the warming climate.

              The chorus of the “Global warming has stopped” brigade is nothing but hilarious by now.

              The very same chorus was heard at other times before, when unsurprisingly the sum of the underlying trend and shorter term variations produced the appearance of plateaus or even reversals of warming if you happened to pick a suitable period.

              Each and every time though those who allowed themselves to be fooled by the appearance of short term variations without comprehending the bigger picture got it very wrong indeed:

              http://www.skepticalscience.com/graphics.php?g=47

              And for the record Simon (and the other members of your blind marching band chorus): I would be well happy if you were right and global warming had actually stopped and so would be every sane person on this planet!

              I am however realistic as there is in fact zero evidence that the trend has stopped or reversed and rather argue for sincere preparations to limit CO2 emissions and to prepare for a very different planet for my grand children’s generation….

            3. Oh Simon, as you are using your real name and as there is a very well known New Zealander with this name who’s stance on Climate change science and politics might be of interest to the public, perhaps you will be so kind as to disambiguate the situation and confirm or deny that you are, or are not, the person below:
              http://www.arnold.co.nz/about.htm

              Thanks.

            4. Thomas March 31, 2012 at 12:12 pm

              I see Thomas is now using the we know who you are and we know where you live line that so many of his disposition love to pull out when they tire of the conversation.

            5. Hmm, andy, could you please point out the reference to where Mr. Arnold lives, because I’m not seeing it.

              I’m sure Thomas regrets any such ‘menacing’ implication as may be construed and bears Mr. Arnold no malice whatsoever.

              Surely the public interest is served if we are made aware that this person is, or is not, this prominent citizen of your country?

            6. No Andy, I have no interest in anybodies personal details whatsoever.

              However if the well known Simon Arnold from Wellington, member of the Royal Society and ex strategy and policy advisor to the Prime Minister of New Zealand and Chief Executive of the NZ Manufacturers Federation (Business New Zealand) would offer his opinion on matter of climate change here in this forum, that would indeed be note worthy don’t you think?

            7. A disambiguation is important, in light of this published statement from the Royal Society of New Zealand. The Society does invite contact from anyone with enquiries, and it would be of interest to know the range of members’ views.

            8. George April 1, 2012 at 8:17 pm

              A disambiguation is important, in light of this published statement from the Royal Society of New Zealand. The Society does invite contact from anyone with enquiries, and it would be of interest to know the range of members’ views.

              Yes, come on Simon, the Brown Shirts are on your case.

    1. A different Andy S, I’m afraid. The Andy S commenting here might change his mind in due course, but I’m sure he’ll understand if I say I’m not going to hold my breath while he considers the issue.

    2. It is statistically quite likely that two people called “Andy S” are the same person, whether you are at a scientific conference, a weblog, or even just beginning a Republican Convention.

      It’s the absence of standards for identity in online forums that is the real culprit though.

      (apologies, AndyS)

  5. I had decided that having been warned by our host of the “good faith” rules I’d not bother to keep helping the debate progress, but coming back and having a look I felt a couple of observations might be helpful.

    To remind I have made a rather limited number of observations here about climate science and the matter in hand – the apparent levelling of the temperature record over the last few years. I’ve suggested that the measured temp has levelled; the temperature record isn’t well model by simple linear models; if you are interested in whether the temperature series has changed 15 years ago it isn’t particularly useful to stick a 12 year moving average through it (even in the name of dealing with solar cycles); GCMs aren’t good at modelling ENSO or for giving a reliable picture of the climate in 2100; and the jury is out on exactly the feedback sensitivity to CO2 in the atmosphere, and hence the contribution that can be attributed to human activity. I’ve also suggested that the scientifically interesting stuff is around what is actually happening now.

    None of which is particularly controversial in the body of literature dealing with climate science.

    In doing this I don’t think I’ve impugned motives or used ad hominem. I did refer to our host as being pedantic, but in my defence not a pedant.

    This is in sharp distinction to what a number here have said about me. I perhaps should have judged from the title of this thread that evenhandness in the treatment of different views wasn’t going to be de rigueur, but my first point is that you are a pretty abusive lot.

    CTG thought I was a nice troll and very mature; our host offered me the option of defending someone I knew nothing about or being a smartarse (tempted to do the former under the circumstances, particular when he later suggested that by doing that no one would take me seriously); bill thought I was a marvel for talking such blatant tosh with confidence, that I was indulging in self-importance, he wished that I was as competent as I was confident, and counselled me not to be misled by blogs (one piece of wise advice); bill then went on to accuse a lack of ability to parse statements, and asked if logic was my thing (I can answer that one – those in NZ better qualified probably count in 100s, not 1000s); and then called me a Zombie and told me that my answers would not be impress at another blog – I’m not sure which advice to take on blogs; our host tried trolling; Macro huffing and puffing and superciliousness; to Rob Taylor I was a sock puppet (without even checking his link to see that I couldn’t be), ideologically blind and intellectually dead; Thomas had a number of things to say including that I had a blind marching band chorus (and got fascinated with my identity, along with bill and George).

    All very grown up.

    Anyway my second point is that there is some serious science to discuss in all this, and as bill says rather than relying on advocacy blogs like sceptical science you are better off looking at some of the literature.

    CTG you are interested in the temperature time series and the use of Monte Carlo techniques for assessing the probability that it fits various kinds of models. In fact the series looks autoregressive (it’s the autoregressive nature of the time series that means simple linear models are inappropriate). You might be interested to read “Influence of Choice of Time Period on Global Surface Temperature Trend Estimates” Brant Liebmann, Randall M. Dole, Charles Jones, Ileana Bladé, Dave Allured. (AMS 2010) where AR models are tested using Monte Carlo techniques – see particularly fig 5(c) that suggests you need to be over 20 years 2008 to get a >90% significant +ve trend (one sided) and I suspect more recent years will be extending this.

    A number of you (Gareth, Macro, Rob Taylor) think that moving average graphs add value to this debate. You should perhaps do a little experiment with a time series that increases by 100 p.a. for the first 15 years, and then goes flat for the next 15. Get a feel for the numbers and what these various transformations are doing. Basically this particular data manipulation doesn’t help elucidate the issue of whether there has been a change in temperature trend or its attribution (Gareth this could, of course, be because the oceans are a sink).

    bill you seem interested in the adequacy of climate models (and with Gareth seem content to use their output in SREX, if not to defend them). Rather than rely on the blogs have a look at http://www.newton.ac.uk/programmes/CLP/clpw01.html and read through the various presentations there. You’ll see the concerns about bias in models, integration of phenomena at low spatial resolutions, uncertainty and something that John Mashey touches on – like Team NZ would simpler stochastic models be better suited for long-term climate forecasting?

    The spatial resolution issue touches on the ability of GCMs to model ENSO well, and Dappledwater you might be interested in “ENSO Feedbacks and Associated Time Scales of Variability in a Multimodel Ensemble” Ali Belmadani and Boris Dewitte AMS (2010), if you aren’t already aware of it.

    The fact that GCMs neither systematically incorporate uncertainty nor report it in a systematically is in my view a significant concern. Unfortunately it is all to common to see papers claiming worrying consequences that on closer examination are based on sticking numbers into climate models and letting them run for a century or two.

    I didn’t mention CO2 sensitivity did I?

    Ho hum

    1. Simon, I pointed out several times that I was not modelling the actual temperature series, nor was I attempting attribution. I created a very small and simplistic model with one specific purpose: to assess what happens to decadal averages under a variety of trend conditions.

      I note with interest that you continue to castigate me for attempting to shed some light on the issue, and yet you have no hard words for andyS, whose original unsubstantiated allegations you have not even remotely challenged. Nor have you offered any alternative statistical interpretation of the matter. All you did were make some offensive comments about me. Very helpful.

      1. nd yet you have no hard words for andyS, whose original unsubstantiated allegations you have not even remotely challenged.

        My unsubstantiated allegations were a general statement about the phrase “warmest year on record” that I backed up with a link to an article by Matt Briggs.

        You then claimed that I had put some kind of statistical weight to different scenarios. I agree with your conclusion, but it isn’t what I said.

          1. So what exactly are they culprits of, then? Please be specific, rather than just throwing around nebulous accusations of general malfeasance, and provide some evidence to back up your accusations.

            1. They are culprits of making a true but misleading statement about the state of the temperature record.

    2. BTW, all the things you talk about here are precisely why climate scientists say that you shouldn’t draw inferences about long-term trends in the temperature data from only 10 or 15 years, so you are actually arguing against yourself, there.

    3. Simon, if I attempt to paraphrase your lengthy comment I would conclude that you express the opinion that:

      Uncertainties exist in the predictions of climate models going forward and there is a possibility that short term observations (decadal scale) MIGHT be an indication of the begin of changes in underlying trends.

      What I am unsure about is what you are actually trying to imply we make of that statement?

      I also need to remind you that uncertainties inherent in our climate modeling and in the interpretation of our observations go BOTH ways and matters of GW could indeed develop faster and more serious than current observations suggest.

      Unless you have some VERY compelling new scientific observations, original research, hypotheses or conclusions to offer, there is no reason to assume that the observations of the last decade suggest that the GW trend is broken. In fact even a cursory look at the temp graphs of the last century suggests that plateaus or downwards periods within the temp series are a common observation on decadal scales and that so far the trend seen in long term running averages reflected the development of our climate best.

      Now some people argue based on economic interests and mechanisms and then maintain that humanity should do nothing to mitigate CO2 emissions. Often the people who argue from this corner leave the economic debate and instead enter the debate on matters of climate science in an attempt to spin doctor the interpretation of the work of scientist in order to rationalize denial of the body of work of science and rationalize the denial of the consequences of a warming planet.

      Perhaps you can enlighten us on your position on this?

    4. Simon, dismissive handwaving does little to further one’s understanding of the changes that are taking place on Earth as it continues to warm.

      The beauty of the climate modeling (excluding CSIRO:MK3, NASA:GISS-AOM, CONS:ECHO-G, NCAR:PCM1 and UKMO:HADCM3 which don’t simulate ENSO well) is that it is consistent with both the basic physics and precipitation anomalies apparent in paleoproxies- i.e ‘global’ warming leads to intensification of ENSO-driven rainfall and drought.

      There’s actually a recent paper in Nature Climate Change (Fowler [2012]) which uses the annual growth rings in Kauri to demonstrate the variability in ENSO. They also expect ENSO-related variability (in New Zealand) to increase in a warming world.

    5. All very grown up.

      Given that you began your “contributions” to this discussion with a not very veiled threat, you should not be surprised at the reception you received. As for your reply: I suggest you Google “tone troll”, then read Foster and Rahmstorf 2012 (discussed here, where the first author notes: “All five data sets show statistically significant warming since 2000.”)

  6. “You should perhaps do a little experiment with a time series that increases by 100 p.a. for the first 15 years, and then goes flat for the next 15.”
    As I said – I’m well aware of what the use of the running mean implies and its effects. But your little experiment would bear no resemblance to actuality and is pointless. As has already been pointed out:
    a. 15 years is far too short a time frame to assess Climatic trends.
    b. The simple fact remains that Global temperatures have NOT plateaued over the past 15 years – and anyone with have a grain of honesty would not claim that it has. See fig 2 http://data.giss.nasa.gov/gistemp/2011/
    c. As previously stated Solar Irradiance is increasing, and with an impending El Nino, global temperatures will see another step change upwards in the near future and you can nit pick all you like, and try explaining things away with ridiculously contrived “statistical” examples, but it’s still going to happen.

  7. Did anyone spot an opinion about AGW in Simon Arnold’s posts? I saw a lot of scepticism as to scientific methods, e.g. appropriateness of models used, but scrolling through the comments I didn’t see a statement of Simon’s own views on AGW. I know this kind of scepticism is almost universally exhibited by those who deny AGW. Yet if I was writing a paper or a thesis on this topic I would doubtless ask similarly searching questions. One must always critique one’s data and methods. It’s just interesting that someone so knowledgeable about scientific method would not venture an opinion as to what they think is actually happening. For the record, Simon, what do you think about Anthropogenic Global Warming? And, in light of the scientific uncertainties, what should be our public policy?

Leave a Reply