How Fake Skeptics Fool Themselves, Part 2

Jeff Condon didn’t seem pleased with my last post about his fake skepticism. He protested that all he had done was “plot some data in a reasonable fashion and make no conclusion.” I think his definition of “single-year ice” isn’t single-year ice and that single-year ice isn’t a good metric for the cryosphere anyway. I also think he selected his definition and presented his plot in order to imply an already-made conclusion. Could it be that he already had an opinion about sea ice, one based on faulty reasoning — the kind that fake skeptics use to fool themselves? Let’s take a look at another of his posts about sea ice.


In “Oops, we hit average” he notes that global sea ice area (all of it) has “hit average.” The first sentence characterizes his interpretation of this event: “Apparently sea ice doesn’t agree with the global warming agenda.” He goes on to say “Despite my belief in CO2 global warming’s effect, I really don’t believe it has had any scientifically discernible impact on sea ice. Nutin!!” He closes with “I’m thinking that after I turn in the useless tax-payer sucking 501C we’ve been discusssing, to no useful effect, I’m going to write a letter to god and turn in the planetary poles for not listening to the government.

All this because global sea ice actually “hit average.”

Let’s apply some actual skepticism to the notion that “hitting average” indicates global warming has had no “scientifically discernible impact on sea ice. Nutin!!”

Here’s the latest global sea ice area from Cryosphere Today:

Note that it’s back below average again. But it did rise above zero earlier this year. Does that mean there’s been no scientifically discernable impact on sea ice?

One thing we can do is look for a trend with linear regression:

Yes there is one, an average loss of about 40000 km^2 per year, and yes it’s statistically significant (yes corrected for autocorrelation). Contrary to Condon’s belief, there is most certainly a “scientifically discernable impact.”

We can also smooth the data, which indicates that maybe the trend is significantly nonlinear:

This suggests fitting a nonlinear trend, so let’s try a quadratic. That looks like this:

The quadratic term, however, just misses statistical significance at 95% confidence — it only passes at 93% confidence. Nonetheless, that’s evidence (but not proof) that the decline in global sea ice has actually accelerated during the satellite era.

But the real issue at hand is that “hit average” thing. Does that really mean anything? Let’s make some artificial data. We’ll give it a trend of -0.04 (million square kilometers per year), and add noise with standard deviation 0.6 (million square kilometer). Finally, we’ll set the scale so that the average value is zero. Here ’tis:

Notice that it keeps rising above “average” — briefly — despite the strong (and “scientifically discernable”) trend. That’s due to the fact that just because there’s a real trend, that doesn’t mean it stops fluctuating. Yes, folks, it still jiggles up and down. Those jiggles are big enough to take it “above average” sometimes, but the trend is still there.

We can even go back to the actual data and count how often it has risen above the average line. Here’s the fraction of each year for which global sea ice area was above average:

As the years have gone by, the fraction of the time that global sea ice area is above the current “average” level has declined. “Hit zero” events, of which Condon made such a big deal, are getting rarer. Fast.

But that won’t stop Condon and other fake skeptics from making a big deal every time the noise goes against the trend. In fact it’s one of their favorite ploys. And it illustrates, perfectly, just how much they lack any shred of actual scientific skepticism.

The fact is that as sea ice continues to decline (in scientifically discernable fashion), it will still continue to rise above the “average” line — occasionally — until it has declined so much that the noise is no longer big enough to get it back to “average” — ever. I wonder what Jeff Condon will say when that happens?

About these ads

56 responses to “How Fake Skeptics Fool Themselves, Part 2

  1. Edgar Walsh

    There is a noisy signal. I would have thought even he could understand that. Him being so smart and all.

  2. 6sn7gtabill

    hi. i hated stats at school with a passion, i’ve never found it intutative, normal maths,no probs. can you tell me how you get the stat significance level numbers from the charts? sorry if this is a dumb question

  3. I wrote a short piece on this at the time: A negative year

    This was the longest period on record of a negative global SIA anomaly. According to my calculations based on the CT SIA data the previous longest period lasted 444 days and ended just before the end of 2007. Another period of 333 days took place from mid-2009 to mid-2010, not much later followed by this new record period of 502 days. As the period ended at the beginning of this year we also just witnessed the first full calendar year on record without a positive anomaly..

    • But the whole point is, dear Neven, that you shouldn’t look at that sort of liberal economy wrecking stuff. You should be focussing on the fact that GSIA has been briefly *above* normal again after being 444 days below average (oops!) and that this proves AGW is nothing to worry about.

  4. But clearly the fake ice extent you created is starting to go up again! Just look at the graph http://tamino.files.wordpress.com/2012/03/artif.jpg

    Even the fake data just won’t behave itself…

  5. Was his post tongue-in-cheek? I cannot believe that someone with his smarts got caught up in the noise trap. Any reader of this blog could have pointed out his obvious error. OTOH, it was a treat to read this from OKM in the comments section: “For example prior to the 2009 release of Climategate emails and documents, I had no hint world leaders directed efforts to hide information on the origin, composition and energy source of the Sun (1971-2009).”

  6. Ray Ladbury

    6sn7gtabill,
    Basically, there is no way to ascertain significance merely by eyeballing a plot. The human brain, having evolved to spot predators hiding in the grass or trees, is really good at spotting patterns whether they are there or not. You have to work with the data itself. Then you have to assume some sort of model that you think fits the data (e.g. normal variability about some trend. Then you look at how much the data deviate from the model. Sounds simple, huh? I ain’t. I would recommend perusing some of Tamino’s old posts and perhaps getting out an old stats text and finding a project to work on. Stats is really something you have to learn by applying it. Good luck and don’t be afraid to ask questions. The dumb questions are the 0nes you just assume you know the answer too–as Condon demonstrates so impressively.

    • Susan Anderson

      This is not about significance but the most basic level of probability, rather fun:

      I agree, those of us who don’t know much get a lot of tolerance as long as we clearly define our limits. The real troublemakers are the bigheads who think they know stuff outside their own field without doing their homework.

  7. “I wonder what Jeff Condon will say when [sea ice no longer rises above 'average']?”

    Well, that’s an easy one. All you have to do is change the period you’re using for the average. “The current level is above the average for the last 10 years! So much for ‘global warming’!”

  8. Another issue is that hitting average really doesn’t mean the global sea ice is where it was 20 years ago. The average is based on the entire period of record and clearly the first ten years of data was biased well above average (so if we made an average based solely on those years, it would be substantially higher than the reported average). So, even hitting “average” is still generally lower than we would have been 20+ years ago.

    • And just to clarify my point… I realize this is simply an artifact of the trend + significant noise statistical analysis that you perform. I just point this out to show that even with no background in statistics, a person should be able to eyeball the data and conclude that something is going on with the sea ice (unless they are a “fake” skeptic, as you contend).

  9. Plus, if you want to claim that “Despite my belief in CO2 global warming’s effect, I really don’t believe it has had any scientifically discernible impact on sea ice. Nutin!!” you have to actually understand a little bit of science, so as to know what kind of effects global warming might have. This is very confusing to some skeptics – Goddard’s “lets look at snow cover in the coldest month of the year” metric comes to mind – and with sea ice, it is complicated. Global warming changes melt rate, changes evaporation, changes salinity, and changes weather, all of which have an impact. Certainly, I have seen work suggesting that in Antarctica, increased precip makes for a fresher surface which freezes more easily…

    (now, of course, skeptics then like to claim “if you can make global warming explain both more ice and less ice, then why should we believe you?” But that’s like claiming that because I have figured out why balloons and birds don’t always fall, you shouldn’t believe me when I say that gravity exists…)

    -MMM

  10. I see it is time to move on from the last incorrect argument. Good call.

    “Could it be that he already had an opinion about sea ice, one based on faulty reasoning”

    As though it were my first cryosphere work eh? Yes, Virginia, I do have opinions on sea ice. I also think it is you who has the faulty reasoning. Sorry doc, but your case is not very convincing. You have mixed the presence of a detected trend with the cause of that trend.

    Could it be that you already had an opinion about sea ice based on faulty, over stiff, models and confirmation bias?

    Deech56
    “Was his post tongue-in-cheek?”
    Not quite but close. Meeting average means near-nothing, if you read the thread, my readers know that. I was grumpy about the believer groups getting ridiculous passes on free use of taxpayer money for of alarmist stuff. Sea ice variance, like paleoclimate, is greater than can be realized in a 30 year trend. Tamino appears convinced that it has never been lower in thousands of years.

    I really like the fact that of all the posts I have done on sea ice, Tamino picked the one designed to tease his sort.

    From my reading and study, I think most of the Arctic ice loss was from a weather pattern change after 2000. Plenty of mainstream stuff on that. You can actually see it in the video’s I’ve produced. However, Carrick was unconvinced by the video and sent me back to the data.

    My guess is that in the near future there will be a springback of Arctic ice to some levels but I really don’t know. I have repeatedly made the point that CO2 warming is real but I don’t think it is responsible for the trend in sea ice. I could be wrong, but that is a sensitivity question which is nowhere near as settled as the good Doc posits. The presence of a trend mixed with the claim that it will continue (and that will show me) shows Tamino has already concluded the matter in his head.

    Pot and kettle?

    • in other words you got caught pulling a steve goddard

    • If you believe the warming is real, is it fair to say that at least some of the trend in sea ice is due to that warming?

    • I don’t see that Tamino is claiming that the data proves the trend will continue, just that the data does not show that the trend has ended.

      If you don’t want people to pick on your weakest posts, don’t post them.

      Why do you guess “that in the near future there will be a springback of arctic sea ice”? Given a temperature rise due to CO2 warming, what mechanism would produce a significant springback? Phlogistonial decay due to ocean acidification?

    • Hrmm . . . those pesky weather pattern changes. That’s the whole monkey wrench in the works, eh? Since weather pattern changes are completely disconnected from climate changes, we can’t really say what the Earth will put on next. I mean, the Arctic, right? It just suddenly warms up of its own accord. A strong high occurs . . . just randomly! (polar amplification is a conspiracy!) Weather patterns shift in unusual ways. Europe freezes. Missouri feels like spring all winter. Hadley cells begin to shift latitude — again, just randomly! You say CO2 warms the atmosphere, but you must be insane: there’s no evidence for it. I mean, wouldn’t there have to be changes in ocean and surface temps? Wouldn’t that energy start melting sea and glacier ice? As you’ve shown, global sea ice is actually increasing! (Well, I did have to tilt the graph a little) And glaciers are increasing (according to my study, in which I measured the shadow cast by a block of ice on the north side of a Himalayan glacier last February the 18th from 8:00 PM to 8:37 PM). Wouldn’t we see changes in the biosphere? I mean, seed companies change their planting maps all the fricking time (well, maybe not the lines, but the colors for sure!).

      Nutter is as nutter does.

    • Jeff

      “was grumpy about the believer groups getting ridiculous passes on free use of taxpayer money for of alarmist stuff. ”

      This is just crazy talk. I think it succinctly expresses your true belief. It has nothing to do with science and everything to do with politics.

    • “Was his post tongue-in-cheek?”
      Not quite but close.

      [snip]

      I really like the fact that of all the posts I have done on sea ice, Tamino picked the one designed to tease his sort.

      Ah, the trademark Marohasy ‘socratic irony gambit’. Yeah, that’ll fly.

      Jeff Condon, I asked this of you on the previous thread, but for the sake of capturing your attention I’ll repeat it here. Could you conduct (if you have not already done so) a ‘Condonic’ analysis on Arctice sea ice volume?

      Morbid curiosity compels me to know how you would frame the trajectory of this metric…

      • Ooh, ooh! Let me guess! He’ll say there’s no way we can know sea-ice volume, and it’s all dependent on models and anyway, blah, blah blah. Then he’ll tapdance off stage. Do I win a pony?

  11. Jeff condon:
    Believer groups like the US Coast Guard and the US Navy? What about the negotiators on Arctic EEZ boundaries between the US, Canada, Russia, plus a few other countries? I suppose all these are all environmentalist whackos out to increase their research funding?

  12. “From my reading and study, I think…”
    “My guess is that…”
    “…but I really don’t know”
    “…CO2 warming is real but I don’t think…”

    How sciencey-sounding Jeff is! In his rush to prosecute his agenda, Jeff abandons science and logic for dissembling and hand-waving. And ignores multiple studies showing the Arctic sea ice to be in a death spiral (only a fool would make arguments about Antarctic sea ice…).

    Jeff’s score:
    Rhetoric = 6.5 (10-point scale)
    Logic and science = Fail

  13. From my reading and study, I think most of the Arctic ice loss was from a weather pattern change after 2000.

    Global cooling and the coming ice age has caused arctic ice to steadily disappear?

    • Jeff Condon

      This is mainstream stuff dhog. Read up on it someday.

      • At your blog you told me the opposite!! You said it was just melting and when I showed you a link to NSDIC saying it was a combination of warming and wind, you declared it was just warming!

        You need to get your story straight Jeff, because you’re making a fool of yourself.

      • Nathan,

        Read it again.

  14. See? Problem’s over:

  15. Bryson Brown

    Jeff, that’s a wildly unjustified conclusion– even by your own lights, which includes (you say) warming as part of the recent climate story. How did you come up with this hypothesis, that arctic ice decline has something to do with some unspecified change in weather pattern after 2000? That’s a strong attribution claim– especially since (without any comment or explanation) it excludes warming as a relevant factor. Can you actually tell us anything about this pattern that would explain why you think it’s nothing to do with global warming?

  16. Bryson Brown

    A broader point in the neighbourhood here: The method Jeff is applying seems to be the method of hypothesis– that is, he looked at the evidence (ice cover changes over time) and proposed a hypothesis that, if true, would (or at least might) explain it. (Change in weather patterns since 2000 qualifies by this standard– but it’s a pure stipulation that there has been some such change, that it’s causally impacted ice extent and that it’s independent of warming.) Now we know (since a long time ago) that there are lots of (mutually inconsistent) hypotheses that can be invented to cover a collection of data in this way. Without some kind of testing being applied to winnow the candidates, forming hypotheses post facto and then arguing for whichever is your ‘favourite’ is useless.

  17. Ray Ladbury

    Jeff Condon,
    If you were really interested in doing science, you would publish your own theory to account for ice loss rather than plotting a statistic you yourself say is meaningless and drawing conclusions based on the attainment by that statistic of a meaningless level.

    Put your theory forward and show that it is more consistent with the data than the standard theory–that warming is driving ice melt.

    Until then, this is more about rallying the troops and whistling in the graveyard than it is about science. That is beneath a real scientist.

  18. Daniel Bailey,

    Aristotle defined rhetoric as the ability to see the available means of persuasion in a given situation. He identified the means as ethos, pathos, and logos. Given the situation, the only relevant means of persuasion is logos. Thus, if Jeff fails logic, Jeff fails rhetoric, too. Jeff is a failure: QED.

    • Alas, winnebago, the relevant means in this situation is pathos. Jeff knows that the best way to look at the ice is overall volume, area, and extent, and record lows were set for all three last summer/fall. Jeff’s target audience is comprised of people who don’t understand that. His target audience is people who want a particular message (“It’s not happening”) and don’t really have the time, means, training, and/or motivation to question the methodology that produced the message (and its logical consistency). His job, and Watts’ job, is quite easy: give the message and make it look science-y enough to fool the target audience. Mix the message with warnings of hoaxery and tax theft, and voila! The mob begins to light torches. When an expert provides a critique, the path of action is dissemble, hem-haw, shift goal posts, dance, re-define, and generally obfuscate until the critical dialogue is so messy that the target audience is reduced to depending on pathetic key words for determining a winner and loser. Note that Condon has hit the critical key words early, including that monster: “capitalism” which immediately invokes the absent “communism,” which, of course, rings emotional bells across the heartland. The thing is, it’s not even subtle. Jeff knows what he needs to do, but he sucks at it, and that’s almost as pathetic as his scientific ability. Unfortunately, he’s probably good enough to fool some of the people some of the time.

    • Spot-on. I was trying to be kind, for (as you point out) failing the logic & science part means catastrophic failure on Jeff’s part no matter the empty facade he constructs to hide his paucity of game.

      These deniers are all smoke & mirrors and no substance; ephemeral gossamer arguments that dissipate with the slightest breeze of inspection.

  19. It’s rather obvious that this is a pre-emptive strke before the Arctic melt season starts. It’s become a tradition that some time in Februray or early March a claim that Arctic sea-ice has recovered (along with a graph, which may or may not have been doctored, and a picture of a nuclear sub) is released into the same old echo-chamber. The Arctic is the denier-world’s Achilles heel, and they know it – they’re tiny-minded in there, but not all are stupid.

  20. Jeff, you are so very obviously straining for data that supports your hopes to “win the public policy battle”.

    http://noconsensus.wordpress.com/2009/12/09/sea-ice-copenhagen-update/

    “Sea ice – extent full satellite record, barely significant trend WRT high frequency weather noise.

    Global sea ice area is even closer….

    ….without 1978 the past 30 years have not shown a statistically significant decline in sea ice. What makes it even better is that in 1978 there are only 2 months of data starting from October 26. So the above graph which shows a less than 95% confidence that the decline is not ‘weather’ noise is only missing the first 2 months…”

    Why, it’s nearly almost close to a refutation of something.

    • jasonpettitt

      …Oh boy :(

      “Global Sea Ice…” ~ Jeff Condon

      As almost certainly the least educated and ill-qualified person here, I feel it would be wrong of me to suppose that Jeff might be forgetting Albedo and a little thing called the Solar Radiation Budget, or how they relate to the following in summer (and its polar opposite in winter).

      Also, apart from having no practical merit, what is that graph comparing Arctic and Antarctic trends Jeff? 1 + 1 is the same as 1000 + 1000, right…

  21. Jeff Condon:However, Carrick was unconvinced by the video and sent me back to the data.

    Just out of curiosity, is that the same Carrick who is currently obsessively tossing chicken bones over at James Annan’s site, counting commas and trying to find his way to the conclusion he’s sure is only just beyond his grasp?

    Probably susceptible to subliminal suggestions; might try those in your next flick.

  22. ““Hit zero” events, of which Condon made such a big deal, are getting rarer. Fast.”

    Excellent observation.

  23. I’m just wondering what the noise level is vs the mean anomoly vs both global warming and ice extent, volume, age etc…and then a conversion to how many years are required to define a trend. (not because I agree with the hollow rhetorical approach of Jeff) but because I’m curious, and I don’t think I quite have the statistcal chops to do this right myself.

  24. Cugel’s point made me realise: if someone in the world wasn’t doing what Cugel highlights – shouting “sea ice is recovering!” or “sea ice isn’t changing!” at least once a year – there would be no source for the meme to spread. It’s veracity is, in fact, completely unimportant/irrelevant for the job it needs to do: providing the little spot of grit around which a denier pearl can form for the next year.

  25. Also, apologies for incorrect use of apostrophe…

    • Apostles of Apostrophe
      can omit the data, you see.
      They bandy arguments on ice,
      and disclaim any means of excise:
      How thought doth vacate their premises.

  26. Am I missing something here? The temperature goes up; ice melts. It’s melting virtually everywhere. It’s just one of the physical manifestations of a warming planet. If you’re going to call up changing weather patterns to account for it, better get busy because it’s not just melting arctic and glacial ice but also causing temperatures in lakes and rivers to rise and snowpack to melt earlier and, well, you get the picture…or not.

  27. Kevin MacDonald

    “The fact is that as sea ice continues to decline (in scientifically discernable fashion), it will still continue to rise above the “average” line — occasionally — until it has declined so much that the noise is no longer big enough to get it back to “average” — ever. I wonder what Jeff Condon will say when that happens?”

    If he’s anything like any other “skeptic” I’ve debated with, he’ll complain about the cherry picking of the base line period

  28. This statement by Condon made me laugh out loud– I think he has been taking lessons from Judith Curry:

    “My guess is that in the near future there will be a springback of Arctic ice to some levels but I really don’t know.”

    What pure unadulterated BS.

  29. Condon then chastizes dhogaza for calling BS on his comment that
    “From my reading and study, I think most of the Arctic ice loss was from a weather pattern change after 2000.”

    As it happens there is a new paper out by Stroeve et al. (2012) who provide an overview of the latest understanding of Arctic sea ice, they attribute the loss to a combination of dynamic and thermodynamic processes,

    “The sequence of extreme September sea ice extent minima over the past decade suggests acceleration in the response of the Arctic sea ice cover to external forcing, hastening the ongoing transition towards a seasonally open Arctic Ocean. This reflects several mutually supporting processes. Because of the extensive open water in recent Septembers, ice cover in the following spring is increasingly dominated by thin, first-year ice (ice formed during the previous autumn and winter) that is vulnerable to melting out in summer. Thinner ice in spring in turn fosters a stronger summer ice-albedo feedback through earlier formation of open water areas. A thin ice cover is also more vulnerable to strong summer retreat under anomalous atmospheric forcing. Finally, general warming of the Arctic has reduced the likelihood of cold years that could bring about temporary recovery of the ice cover.”

    Note they suggest that the acceleration of sea ice loss is because external forcing, not weather pattern changes (i.e., internal variability). Specifically,

    “Since about 2000, positive anomalies dominate all months. This warmer Arctic climate has been linked to increased concentrations of atmospheric greenhouse gases, increased concentrations of black carbon aerosols (Shindell and Faluvegi 2009), increased spring cloud cover that has increased downward longwave radiation at the surface (Francis and Hunter 2006), variability in atmospheric circulation and horizontal atmospheric energy transport convergence (Yang et al. 2010), and reductions in ice extent (discussed shortly).”

    Yang et al (2010), however, note that in the last decade has been “fairly neutral in terms of temperature and energy transport change.” The role of the Arctic dipole in the summer months is an important modulator for individual ice melt seasons, but does not explain the long term trend.

    More from Strove et al. (2012),
    “With more thin ice starting the melt season compared to 20 years ago, atmospheric patterns favoring summer ice loss, such as seen in summer 2007, are more effective than they used to be. Conversely, atmospheric patterns that favor ice retention are becoming less effective (e.g. Stroeve et al. 2011). Finally, as the Arctic continues to warm in response to radiative forcing, the probability of a sequence of unusually cold years in the Arctic that could bring about recovery declines.”

    It is also called polar amplification (e.g., Screen and Simmonds 2010) in response to a warmer climate.

    • Condon then chastizes dhogaza for calling BS on his comment that
      “From my reading and study, I think most of the Arctic ice loss was from a weather pattern change after 2000.”

      Note that I’m very aware that anger lies close to the surface of JeffIDs personality. His political rants make that obvious.

      So in all fairness to him, I was tweaking him intentionally. He is, if nothing else, predictable.

  30. JeffId fires back at WUWT. Apparently he’s found all kinds of evidence that the trend in sea ice is an artifact of flaws in the data. This will be fun to watch.

    • He should have no trouble getting that published in a decent scientific journal. He’s wasting his scientific insights and acumen by publishing in blogs.

  31. Oh dear, poor Jeff is throwing a tantrum at his place. And not doing a terribly good job of evading direct questions. Readers are assured that all is clear if one reads his posts…clear as mud that is.

    Jeff declares:
    “Unlike the Air Vent and WUWT, his crowd is comprised primarily of non-technical readers who often jump at any statement they can find with literally zero understanding of why or what they are attacking.”

    Honest to God, I think he actually believes this. But contrary to his belief, the posters’ comments at WFUWT refute Jeff’s belief and make a mockery of his
    ridiculous claim.

    Ah ha, and yes, now readers are led to believe that the data are not to be trusted because Mr. Dunning-Kruger (aka Jeff Condon) claims to have possibly found an error.

    Gosh darn it, but the deniers/lukewarmers/fake skeptics are tying themselves in knots.