Forced vs Unforced Variability

This could well be another post that illustrates my ignorance, more than anything else, but maybe I’ll learn something from it, even if noone else does. I have been assuming that it was generally agreed/accepted that periods of major climate change were forced rather than unforced. It’s possible that I’m using these terms incorrectly, so what I mean by forced is that the changes are driven by some kind of external influence. Unforced being internal variability. [Addendum : I had initially used the term natural/internal variability but, as Kevin points out in the comments there can be natural external forcings, so I’m trying to stick with internal – to the climate system – variability when referring to unforced changes].

Examples of forced variability would be volcanic outgassing – hundreds of millions of years ago – allowing greenhouse gases to accumulate in the atmosphere, warming the planet and driving us out of a snowball earth. Asteroid impacts have been associated with other periods of climate change. Orbital variations (Milankovitch cycles) are thought to be the driver for climate variability in the last 500000 years. Variations in solar intensity and variations in volcanic activity (in this case associated with aerosols) have been linked with the Medieval Warm Anomaly (MWA) and with the Little Ice Age (LIA). Today, the driver of climate change today is thought (known?) to be anthropogenic emissions of CO2.

A few recent discussions have, however, made me wonder about the issue of forced versus unforced variability. From discussions on this post it seems that even though millenial reconstructions can tell us a lot about our past climate history, there is much we still don’t know and much more (in particular about forcings) that they could still tell us. From discussion on this post it seems that even though the timings of the orbital variations (Milankovitch cycles) coincide with variations in our climate, we still haven’t shown definitively that they are indeed linked. Science of Doom has an interesting recent post that discusses this issue.

So, could this uncertainty about our past climate history indicate that some past climate changes could have been unforced, rather than forced? Our climate is clearly very complex and can be chaotic (on short timescales at least). It certainly seems plausible that some kind of internal variability could change – for example – a major ocean current, send warm water to the poles, melt polar ice, change the planet’s albedo, and produce a phase of warming that could then be amplified by the release of greenhouse gases. It’s my understanding, though, that there’s no actual evidence for this – at this stage at least. Also, although it may be true that we haven’t definitively shown, for example, that Milankovitch cycle are the drivers for major climate change in the last 500000 years, the timings seem so similar that it, at least, seems plausible that this is the driver, even if we haven’t definitively shown how it actually operated. Maybe not, of course, but I’m not aware of a plausible alternative.

So, I guess the two issues are : is there any evidence for unforced variability (long-term rather than short-term at least), and – in terms of global warming/climate change today – does it matter? I’m well aware that many “sceptics” argue that global warming/climate change today could be due to internal variability (unforced). So, if one could find evidence for unforced variability in our past climate, that would – I suspect – lead some to argue that we can’t rule out a significant unforced influence to climate change today. The problem with that is that we “know” that there is a significant external forcing (from anthropogenic greenhouse gasses). If you want global warming/climate change today to be unforced, you then need to show how the anthropogenic forcings have been cancelled while, independently, internal variations have produced a warming that is consistent with that expected from this external forcing. Anything’s possible – I guess – but some things are just very unlikely.

So, as I said at the beginning, maybe this post is more indicative of my ignorance than anything else. An earlier comment by Arthur Smith partly motivated this post and makes me think that – at least – I’m not alone in wondering about this. If others have thoughts or views on this, feel free to make them through the comments (although taking the new Comments Policy into account would be appreciated).

[Addendum : I’ve edited this slightly to try and better illustrate that unforced means internal variability, while forced applies to external influences, some of which can still be natural.]

Advertisements
This entry was posted in Climate change, Climate sensitivity, ENSO, Global warming, Science and tagged , , , , , , , . Bookmark the permalink.

153 Responses to Forced vs Unforced Variability

  1. Kevin says:

    Good observations. A few random thoughts:

    1. Across the spatial and temporal scales we’re talking about here, climate is a combination of forced and internal variability. Observations of the climate have an additional uncertainty, which includes measurement uncertainty but also the network of observations relative to the continuous climate field (e.g. you have weather stations/proxy data at only certain, often non-random, locations) — that uncertainty is typically smaller for instrumental observations compared to paleoclimate proxies.

    2. The forced response is a function of the forcing, which is uncertain even in recent decades. The situation is obviously worse going back through the last millennium when forcing estimates are themselves estimated from other proxy measurements (ice cores, radiogenic isotopes, etc). The change in forcing is larger, for instance, at the LGM vs. the preindustrial last millennium, but still subject to uncertainties.

    Given 1 and 2, the challenge then becomes to separate the forced and unforced response (or, to look at it another way, to Detect and Attribute the climate response to different forcings, given uncertainties)

    3. In general, at the low frequency (~long term) and larger scale the forced response will be more important vs. the local-to-regional scale and the shorter term, where the internal variability will contribute more substantially.

    4. Hegerl and her colleagues have some very nice papers on aspects of this, for instance (2011, Nature Geoscience): ‘Influence of human and natural forcing on European seasonal temperatures’, which finds, amongst other things that 25-35% of the decadal-scale variability in European temperatures of the last 500 years is externally forced, compared to 52 to 70% for the Northern Hemisphere over the last 700 years (from an earlier paper). You could look at this another way, too, which is that internal climate system variability plays an important role in >decadal scale variability.

    5. Another example of what you can do with forcing estimates, GCMs, and proxy estimates, even with uncertain forcing, uncertain observations, and imperfect models, is in the recent paper by Schurer et al. in Nature Geoscience, where they concluded that ‘neither a high magnitude of solar forcing nor a strong climate effect of that forcing agree with the [millennial, hemisphere-scale] temperature reconstructions.’

    6. There are definitely features of the millennial climate reconstruction that are hard to explain as a primarily forced response — for instance, a lot of the reconstructions have a peak in Medieval temperatures around 1000 CE, which can’t readily be explained as a forced response. Now, is this forcing uncertainty? Proxy uncertainty? Internal variability?

    7. The climate system can do some pretty cool stuff on it’s own — in long (unforced) control runs of climate models, you get centennial-scale SST variability (Karnauskas et al), unforced multidecadal megadroughts, ENSO modulation, things that look like the ‘Little Ice Age’; you can get multidecadal AMOC behavior (e.g. Latif et al. 2006), etc. The climate system is pretty awesome, even when just doing its thing.

    8. Yeah, obviously the presence of internal, unforced variability doesn’t mean CO2 isn’t forcing recent trends (all evidence says it is a major factor) – the interesting question then is what is the relative balance of forced (natural+anthropogenic) and unforced contributions (having estimates of internal climate system variability is important for things like decadal prediction, too) at different time scales.

  2. Kevin,
    Thanks. Some very interesting random thoughts 🙂

    In general, at the low frequency (~long term) and larger scale the forced response will be more important vs. the local-to-regional scale and the shorter term, where the internal variability will contribute more substantially.

    This was my general impression.

  3. Rachel says:

    This is a really good post. I think I follow you this time. SOD’s post was good too. If I understand correctly, a possible unforced mechanism could be the cause of an interglacial and you give the example of changing ocean currents melting polar ice and changing the planet’s albedo and so on. I wonder whether another possible mechanism could be via changes in cloud cover due to, for instance, changes in vegetation?

    Even if this has has happened in the past though, I don’t see how that could change the case for anthropogenic warming today.

  4. Rachel,

    I wonder whether another possible mechanism could be via changes in cloud cover due to, for instance, changes in vegetation?

    I believe that this is one idea put forward by Roy Spencer (see this post). He’s arguing that ENSO events can produce a forcing through changing cloud cover. There are a number of issues with this. One is that I don’t think there is any evidence for a sufficiently strong link between ENSO cycles and clouds. Another is that if this is true, why do ENSO cycles produce a long-term warming trend? It can’t only produce warming or we would, presumably, have had some kind of runaway by now (many centuries of ENSO cycles). A third, as I tried to point out in the post, is that if you want what’s happened over the last 50 – 100 years to be primarily unforced (natural) you then need to explain how the anthropogenic forcings were largely cancelled (while – coincidentally – a natural/internal variation produced a warming consistent with what would be expected from the anthropogenic forcings).

    So, yes, another mechanism could – I think – be clouds but I’m not really aware of any evidence for this ever having actually happened.

  5. John Russell (Twitter@JohnRussell40) says:

    A useful post and I’ll be very interested to see what the more informed commenters have to say about it given that, like ATTP, I have only a, hopefully, informed layperson’s understanding and have probably gathered a few misunderstandings along the way.

    Of course, I do have my own ideas (based on the science I’ve read, I hasten to add) of what causes frequent changes in our climate. In fact I had a discussion with my wife last night, having just watched the ‘Horizon’ programme on BBC4 called ‘Global Weirding’. To my great interest, when I described the extreme weather other parts of the world were experiencing at the moment (the Horizon programme was a repeat, made in 2012) she, quite innocently, came out with several denial memes (like, “…but the climate’s changed before, like the Little Ice Age”). Now I know she’s not in denial—and with an Open University degree in science she’s better qualified than me—but she’s just never thought about these things before, even though I’ve always tried to interest her in the subject. I guess she’s just one of the vast majority of people who don’t really think that deeply about climate changes and consequently are a soft touch for the cynical David Rose/Daily-Mail-type propaganda. Anyway I explained some of the forcings that affect climate and she seemed to accept it, but I’m guessing she’s still thinking “but how do they know that this time it’s humans who are creating it?”

    If only there were a simple way to prove conclusively in a few words that this time it’s down to us: especially when scientists appear to be so dithery on the subject, with their ‘maybes’, ‘mights’ and ‘probablys’. And all the time those-in-denial are so effing sure of themselves—which is a sure sign to anyone who has any understanding of the science that they’re the ‘true believers’.

  6. Kevin says:

    Something I forgot to mention, but natural ~= unforced (MATLAB syntax), since ‘natural external forcing’ would include solar, volcanic, etc. –http://www.assessment.ucar.edu/paleo/past_forcings.html

  7. Rachel says:

    TheresPhysics,
    I didn’t really understand that post but perhaps I’ll go and have another look. In any case, I don’t really want to be associated with Roy Spencer if that’s ok 🙂 As far as I’m concerned the climate is warming and humans are almost entirely the cause. Doesn’t Roy Spencer dispute that? But it is interesting to read about why the climate has changed in the past and even more interesting knowing that there is some disagreement about this.

  8. John,

    If only there were a simple way to prove conclusively in a few words that this time it’s down to us: especially when scientists appear to be so dithery on the subject, with their ‘maybes’, ‘mights’ and ‘probablys’. And all the time those-in-denial are so effing sure of themselves—which is a sure sign to anyone who has any understanding of the science that they’re the ‘true believers’.

    This reminds me of something I’ve been thinking about but haven’t written about; which is attribution. Especially when it comes to extreme weather, many seem to argue that even if there is a trend (statistically significant or not) we can’t yet attribute changes to anthropogenic influences. I’m sure this is true since – as far as I’m aware – attribution is extremely difficult. However, given that many who make such claims have backgrounds in statistics, it surprises me that they don’t consider the combined likelihood. Maybe this has been done, and – presumably – it’s also difficult, but we don’t really (I would argue) care if we can attribute individual events (we probably can’t and shouldn’t try) or types of events to anthropogenic influences. What we’re interested in is whether or not the overall change in our climate could be simply natural variation or whether or not it is more likely to be driven by anthropogenic influences.

    I don’t know if I’ve expressed that all that clearly, which is probably why I haven’t tried writing any posts about this 🙂

  9. Rachel,

    Doesn’t Roy Spencer dispute that? But it is interesting to read about why the climate has changed in the past and even more interesting knowing that there is some disagreement about this.

    Yes, I think Roy Spencer is trying to argue that global warming today could be natural. As you say, even though he is likely wrong, it is still interesting to consider what might have influenced changes in out past climate and that there are disagreements rather undermines the argument that it’s all some kind of conspiracy amongst climate scientists.

  10. Kevin,

    Something I forgot to mention, but natural ~= unforced (MATLAB syntax), since ‘natural external forcing’ would include solar, volcanic, etc. –http://www.assessment.ucar.edu/paleo/past_forcings.html

    Indeed. I thought I’d been quite careful when writing this post, but you’re right that there can be natural external forcings and I didn’t quite make that sufficiently clear in the post.

  11. Rachel says:

    John Russell,

    …she, quite innocently, came out with several denial memes (like, “…but the climate’s changed before, like the Little Ice Age”).

    I’ve heard exactly the same thing from friends. More than once too. I usually respond with something like we’re the cause and the pace of change is unprecedented.

  12. Arthur Smith says:

    ATTP – did you mean to link to this previous comment – https://andthentheresphysics.wordpress.com/2013/12/27/forcings-and-feedbacks/#comment-10909 – instead of the other above?

    I think the question of unforced “variability” is a bit subtle, but to me there are two distinct relevant meanings:
    1. regional and short-term “variability” – but “climate” itself remains the same as the long-term average, these are just excursions within fixed statistical distributions that represent the current “climate”.
    2. boundary-changing “variability” – things that change ice sheets or ocean currents or CO2 – so that the Earth system changes its underlying climate state fundamentally, without external forcing.

    Is unforced variability of the second type possible? Certainly yes if long-term feedbacks push the Earth response into the unstable regime. Is it the condition under which Earth currently or in the recent past has behaved? I don’t think so, but it seems hard to be sure…

  13. Arthur,
    I linked to the one I intended to link to, but the one you highlight may have been better.

    Yes, you’ve expressed quite clearly what I was trying to say. It seems clear that the first type of unforced variability happens. I was more referring to your second type of variability. As you seem to be indicating (and as I had thought myself) there isn’t any real evidence of unforced variability changing the underlying climate state without an external forcing. Of course, we haven’t yet definitively shown that all past instances of changes to the underlying climate state were forced, but that – in itself – isn’t evidence of it being unforced.

  14. Kevin says:

    In the vein of (abrupt, multidecadal to centennial-scale) climate change without external forcing, a few articles of potential interest:

    Hall and Stouffer, 2001, Nature, An abrupt climate event in a coupled ocean–atmosphere simulation without external forcing

    Hyde and Crowley, 2002, Paleoceanography, Stochastic forcing of Pleistocene ice sheets: Implications for the origin of millennial-scale climate oscillations

    Drijfhout et al. 2013, PNAS, Spontaneous abrupt climate change due to an atmospheric blocking–sea-ice–ocean feedback in an unforced climate model simulation

    A lot of this will be dependent on what timescales and phenomena you are interested in (for instance, closer to my own interests, it seems increasingly clear that the climate system can generate unforced 30+ year droughts). But certainly, on earth system time scales, to paraphrase Richard Alley’s 2009 AGU talk ‘nothing makes sense without CO2’.

  15. I’ve literally spent most of the day on just this subject so I’m not going to attempt a proper reply.

    Robert Wilson asked a very similar question: “if CO2 didn’t cause it what else did?”.

    This shows a very different approach between us citizen scientists and what do I call you … “less sceptical”? Because that is a very relevant question if you want to understand climate, but it is not if what you want to do is to predict climate. Then relevant question is how much of the the change in climate is predictable and how much is not. Because if your aim is to predict the climate there is no need to understand everything about climate to produce the best prediction that is possible.

  16. ScotScep – I’d really rather you didn’t call me anything if you’re going to be insulting. I don’t really understand the rest of what you’re trying to say.

  17. Kevin says: January 9, 2014 at 4:45 pm

    Thanks for the items, but every single one is behind a paywall.

  18. ScotScep, that’s more the fault of the current publishing industry who make a very large amount of money from publication costs (the UK – I believe – spend £200 million per year on journal fees). The current UK plan to solve this problem is to given them more money.

  19. Rachel says:

    Scottish,
    You can often access scientific papers like those from a University library. You would need to go in to the library though and use their computers.

  20. Kevin and BBD,
    Thanks. So it seems that my suggestion that there is no evidence is not strictly correct but it does still seem to be relatively short time-scales, although longer than the decades I had envisaged as being the norm for unforced variability.

  21. John Russell (Twitter@JohnRussell40) says:

    When I’m explaining climate change to people who are uninformed—rather than arguing with people who are informed by denial—I tend to just refer to two types of cause: ‘natural’ and ‘artificial’. I use the term ‘artificial’ because the dictionary definition is, “made or produced by human beings rather than occurring naturally, especially as a copy of something natural; which I would say is very appropriate. I then go on to say that natural climate change—unless brought about by a violent event such as a major volcanic eruption or a meteor strike—tends to be slow-acting in human terms, taking place over hundreds and thousands of years. In comparison, because of the speed we’re pumping greenhouse gasses into the atmosphere, human-caused climate change is rapid and very difficult to reverse and will have major impacts on our weather.

    This provides a good starting point for those who want to ask further questions. I’d be interested to see whether any of the scientists object to anything about this explanation.

  22. John,
    I think this may be one of the areas where judgements need to be made about how best to communicate and where I find there is often some disagreement. As Kevin pointed out above, forced can include natural (Sun, volcanoes), while unforced would imply internal variability. However, from the perspective of what’s happening now, I would argue that how you describe it to those who are largely unaware of the science is probably fine and may be the best way to get the basic message across.

  23. Rachel says:

    John,
    The other thing I like to add is that when the climate as changed dramatically in the past, it has often been accompanied by mass extinctions. And then I also like to say that humans were not around during times of hot-house Earth. I think some people, especially those who live in cooler climates, are under the impression that they’re going to inherit an outdoorsy, cafe climate.

  24. John Russell (Twitter@JohnRussell40) says:

    I note that Dana has a new post out in the Guardian on this very same subject: http://www.theguardian.com/environment/climate-consensus-97-per-cent/2014/jan/09/global-warming-humans-not-sun

  25. John Mashey says:

    Until ~8,000 years ago, all climate changes were a mixture of natural forced and unforced variability at various timescales from short-term unforced randomness, continental drift on hundred-million year scales.
    Milankovitch is in the middle, and actually they understand that pretty well, as summer solar insulation ~60degN mattered a lot starting a few My ago, although it won’t for the next few 100ky, maybe 500ky.

    But since then, humans have contributed an increasing amount of forcing.
    The most coherent, thorough explanation for the (unnatural) Holocene is Bull Ruddiman’s new book Earth Transformed, the subject of his Tyndall Lecture @ AGU a month ago. For instance, neither the MWP nor LIA were entirely natural, the latter caused in part by the 50Mperson dieoff in the Americas, and consequent unique 9ppm drop in CO2 into 1600CE, just before the Maunder Minimum.
    Ie, part of the LIA was natural: volcanoes and solar minima, and part if it was human forcing.

  26. Arthur Smith says:

    Hey, I thought Scottish Sceptic wasn’t coming back here?

    Anyway, I also cannot parse the comment up above, but if it was supposed to indicate that we seem to be talking about current warming not being due to CO2, Scottish Sceptic must be lacking in some basic background on this topic.

    A fundamental feature of variability in physical systems is that the larger unforced variations are, then so much larger is the forced response. We’re talking about potentially very high unforced variations possibly being associated with some major past change (including glaciations?). All else being equal, if that was true, that indicates the response to the major forcing humans have added through CO2 emissions will be much much worse than currently anticipated.

    That’s not a good thing.

  27. Arthur,
    I think I’ve asked, or hinted, that ScotScep not bother commenting here, but he doesn’t seem to take the hint. Can’t fault his perseverance.

    Anyway, I also cannot parse the comment up above, but if it was supposed to indicate that we seem to be talking about current warming not being due to CO2, Scottish Sceptic must be lacking in some basic background on this topic.

    Indeed, and he thinks Murry Salby’s ideas may have some merit. Need I say more?

  28. BBD says:

    ATTP

    Sorry. I didn’t meant to be curt or gnomic; I posted the link for Scottish without pausing for thought because he clearly hadn’t bothered to look and I knew at least that H&C02 was available because I’ve downloaded it before.

    DO events during the last glacial do fit the bill as cyclic (or at least quasi-periodic?) events arising from internal variability. And they are the only ones I can think of and you don’t get them during the Holocene (note: *Holocene* “Bond Cycles” are somewhat speculative and not the same type or frequency of event as the bundles of DO events during MIS3/last glacial also called Bond Cycles).

  29. John Russell (Twitter@JohnRussell40) says:

    Rachel said, “I also like to say that humans were not around during times of hot-house Earth”.

    Actually, when ‘the hot-house Earth’ is brought up—you know, the time those-in-denial like to enthuse about it being “green, lush and full of life in an atmosphere of 1000+ppm CO2”—my response is usually to say, “oh, you mean the Eocene, when mammals were no larger than cats and our ancestors lived in burrows to escape being eaten by dinosaurs?

  30. andthentheresphysics says: “ScotScep – I’d really rather you didn’t call me anything if you’re going to be insulting.”

    It was a dig that no matter how polite we sceptics try to be that no one of the “please don’t call me anythings” will come up with something to describe themselves as a group.

    It is rather like having a football match between two groups in the office … one has a name … the other group can’t be bothered to think of a way to describe themselves and then feels it has a right to complain when the other side comes up with one that they don’t think is ideal.

    Maybe we could call you “the ones who are without a name” or “TOWAWAN”s?

  31. Rachel says:

    John,

    oh, you mean the Eocene, when mammals were no larger than cats and our ancestors lived in burrows to escape being eaten by dinosaurs?

    And giant 14m snakes weighing more than a tonne.

  32. jsam says:

    Citizen scientist? That sounds a bit vigilante to me. 🙂

  33. [Mod: This isn’t really relevant here, Scotty and I think we’ve had enough discussion already about names in previous threads. Thanks]

  34. ScotScep, the issue is not that you were having a dig. That’s fine. I’m just rather tired of your public claims that somehow all you citizen scientists are the real sceptics and others (who happen to disagree with you) are not. I think Eli Rabett’s comment on your most recent post was pretty spot on. I think it’s time you realised that you don’t have some special perception that academics seem to lack and that maybe academics aren’t quite as blinkered and lacking in real-world experience as you seem to think. Of course, academics don’t know everything and make mistakes like everyone else. Also, there are many areas where citizen scientists make very valuable contributions (including in climate science). However, simply being a citizen scientists doesn’t make you superior, as you seem to think. Maybe it’s time you gave some thought to how your views might be perceived elsewhere.

  35. jsam says:

    I really enjoy the “you can’t do science by consensus” mob attempting “but you can do scepticism by consensus”. There is no-one less sceptical than a “climate sceptic”.

  36. jsam Citizen scientist? That sounds a bit vigilante to me. 🙂

    And what is the big bang theory? A joke about premature publication!

  37. jsam says:

    The Big Bang theory is newer than climate science.

    http://en.wikipedia.org/wiki/History_of_the_Big_Bang_theory

    http://en.wikipedia.org/wiki/John_Tyndall

    You should be more sceptical.

  38. BBD says:

    Rachel

    Titanoboa was a fearsome beastie, and perhaps a direct evolutionary consequence to the peak Eocene hothouse – snake paleothermometry anyone? – (and also my little boy’s favourite extinct snake) but the dinosaurs were long gone by then.

  39. John Russell (Twitter@JohnRussell40) says:

    Actually, Rachel, I meant ‘reptiles’ not ‘dinosaurs’. An error on my part there. I don’t know whether the link to ‘Titanoboa’ was a gentle hint, but it’s taken!

  40. John Russell (Twitter@JohnRussell40) says:

    Arrgh, BBD. Your comment crossed with mine… honest!

  41. John Mashey says:

    Back from coffeeshop:
    Here’s ToC for Earth Transformed.

    Again:
    1) Forcing versus internal variation: timescales really matter.
    2) For ~6,000 years, humans applied slow, modest forcings in terms of CO2 and (later) CH4, slowing the usual slow, jiggly drift downward, so that pre-IR CO@ was ~280ppm instead of the expected 240-250ppm.
    3) From 2000 years ago, there started to be human impact on the multi0-decadal to century timescales in terms of CO2/.CH4 jiggles … and since the IR, we have of course changed CO2 enough to see 100ky-sacle effects and put us back into the state before CO2 was low enough for Milankovitch cycles to start.

  42. Rachel says:

    Thanks, BBD. That link is much better. Matt Huber featured in the film, Thin Ice and I thought he was really good. Great film too. I’m not bothered by snakes but I think I’d probably find the Titanoboa a bit scary. But spiders…..ugh

    John,

    I don’t know whether the link to ‘Titanoboa’ was a gentle hint, but it’s taken!

    I don’t do gentle hints. I’m brash and tactless. And I’m not even sure what my link is supposed to be a hint of?

  43. John Mashey says:

    On a related topic, I recommend consideration of the term pseudoskeptic:
    Richard Cameron Wilson, captured it well in the New Statesman:
    ‘In a sceptical age, even those disseminating wholly bogus ideas – from corporate pseudo-science to 9/11 conspiracy theories – will often seek to appropriate the language of rational inquiry. But there is a meaningful difference between being a “sceptic” and being in denial. The genuine sceptic forms his beliefs through a balanced evaluation of the evidence. The sceptic of the bogus variety cherry-picks evidence on the basis of a pre-existing belief, seizing on data, however tenuous, that supports his position, and yet declaring himself “sceptical” of any evidence, however compelling, that undermines it.’

    People may pride themselves on being true skeptics, and in fact be so on many topics and then act as intense dismissives on others, falling prey to selective use of Morton’s Demon*. In 2007, a straightforward article on global warming in Skeptical Inquirer generated a small firestorm that revealed some readers to be intense pseudoskeptics on that one topic, discussed in followup, after the dust had settled a bit.

    *Morton’s Demon ‘stands at the gateway of a person’s senses and lets in facts that agree with that person’s beliefs while deflecting those that do not. This demon is used to explain the phenomenon of confirmation bias.’

    Some people have really active Morton’s Demons, who roam widely to seek any bit of nonsense that supports their beliefs. For example, think what it takes for anyone to assign Viscount Monckton or the “slayers” any credibility on climate science. For a good example, see the 2 pages of comments on my review of Salby’s new book.. Some comments offer useful data on pseudoskeptical thought patterns and I’ve been WebCIting away lest they disappear.

  44. John,
    Thanks for all the comments.

    Forcing versus internal variation: timescales really matter.

    That’s certainly what I’m getting from this. It really does depend on what timescale one is considering.

  45. John Russell (Twitter@JohnRussell40) says:

    Isn’t that the whole point, jsam? There is absolutely no consensus amongst climate sceptics.

    The irony they don;t seem to appreciate is that the few maverick scientists who turned out to be right, subsequently became part of the consensus. But to convince others they had to publish a unifying theory based on evidence that was overwhelmingly convincing. The climate sceptics as a group are awash with multiple pet theories and favourite denial memes that contradict one another, with hardly a published paper between them. Those that do manage to publish invariably fall at the first hurdle.

  46. John Russell (Twitter@JohnRussell40) says:

    In the light of John Mashey’s comment, I should have referred to “climate pseudosceptics” in my last comment. A good term, which I’ll use from now on.

  47. John and jsam,
    I think there is confusion what is meant by the consensus. As I see it, the consensus is really only that CO2 cause warming (around 1.2 degrees per doubling) and that feedbacks are positive (increasing the warming to at least 2 degrees per doubling, but possibly higher). That’s really all the consensus is. There’s no consensus about a precise amount of warming (is it 2, 3, higher, ESS, ….). There still seems to be quite a lot of disagreement about climate reconstructions (in the details at least), there’s disagreement about attribution of climate events (weather) to anthropogenic influences, and disagreement about quite what we might expect regionally from climate change.

    So, yes there’s a consensus, but it only refers to the basic result that we will warm, globally, if we continue to add CO2 to the atmosphere and that if we follow a BAU scenario (or something similar), we will quite soon have global temperatures higher than for all of human history.

    That’s not to say that there isn’t some level of agreement about some of the other factors, but the consensus (as far as I’m aware) really only refers to the basics of global warming and not to the details of climate change.

  48. John Mashey says:
    On a related topic, I recommend consideration of the term pseudoskeptic:
    Richard Cameron Wilson, captured it well in the New Statesman:
    ‘In a sceptical age, even those disseminating wholly bogus ideas – from corporate pseudo-science to 9/11 conspiracy theories – will often seek to appropriate the language of rational inquiry. But there is a meaningful difference between being a “sceptic” and being in denial. The genuine sceptic forms his beliefs through a balanced evaluation of the evidence. The sceptic of the bogus variety cherry-picks evidence on the basis of a pre-existing belief, seizing on data, however tenuous, that supports his position, and yet declaring himself “sceptical” of any evidence, however compelling, that undermines it.’

    There is a paper by Lewandowsky, which if you look at the details and not the commentary [Mod : removed unfounded accusation] , you will find that sceptics look at a graph and whether or not they are told it is graph of economics or climate, they will come to pretty much the same estimate of the future.

    In constrast, when those believing in global warming see the same graph they completely change their estimate of the future depending on whether they are told it is a graph of global temperature or one of economics.

    This shows that sceptics really are sceptical of what they are told and instead look at the data whereas those who believe in global warming change their interpretation of a graph depending what they are told it should show … and those who write about sceptics take the data showing that sceptics base their conclusions on the data and try to turn the interpretation on its head.

    So when you suggest that sceptics “cherry pick” the data and use pre-existing beliefs you are actually going against the evidence presented by one of the most vile critics of sceptics in his own paper.

    But somehow I doubt you will consider the evidence and instead just look at the fact I’m a sceptic and think what you want.

  49. BBD says:

    But somehow I doubt you will consider the evidence and instead just look at the fact I’m a sceptic and think what you want.

    You aren’t a sceptic, Scottish. You reject information that contradicts your strong prior commitment to non-science.

  50. BBD says: You aren’t a sceptic, Scottish. You reject information that contradicts your strong prior commitment to non-science.

    That’s a possibility but I’d like to see evidence rather than just take your word for it.

  51. ScotScep and BBD,
    Let’s avoid taking this any further.

  52. uknowispeaksense says:

    Just a couple of points from me. One relevant to the topic. The other relevant to a few of the comments.

    On syntax, I personally dislike the term natural/unnatural and internal/external. It could quite easily be simplified with the terms normal/abnormal. Normal forcings encompass natural internal and external forcings like volcanoes, feedbacks, milankovitch cycles and the sun. Abnormal would be anthropogenic emissions and deforestation. Simple. All the normal stuff has been measured and modelled and doesn’t explain the current rapid warming we are currently experiencing which leaves the abnormal. Simple.

    The second thing relates to the complaints about paywalls. Why is it mostly …. [Mod : I’ve removed a term you used here. I know some think it’s fine and appropriate, but I try not to use it and should probably have made that clearer in my moderation page.] who moan about this? My hypothesis is that it gives them an excuse to remain wilfully ignorant. I care about global warming enough to purchase a dozen subscriptions and no I dont have a lot of money.. If you really want to know, you will make the effort to get to a university library and access the paper versions of the journals.

  53. [Mod : I’m not really interested.]

  54. uknowispeaksense,
    I guess I could see issues with normal/abnormal too in the sense that how does one then distinguish between forced and unforced. It is a bit of a minefield though. Easy to get everything confused as I manage to do quite regularly.

  55. andrew adams says:

    Scottish Sceptic

    There is a paper by Lewandowsky, which [smear deleted] you will find that sceptics look at a graph and whether or not they are told it is graph of economics or climate, they will come to pretty much the same estimate of the future.

    In constrast, when those believing in global warming see the same graph they completely change their estimate of the future depending on whether they are told it is a graph of global temperature or one of economics.

    Yes that sounds quite likely to me. When people know what a graph actually represents they make judgements based on their understanding of the underlying subject, not purely on the graph itself.

  56. KR says:

    Imbers et al 2013, which was just discussed over at SkS, examines internal/unforced variability versus greenhouse gas forcing.

    They modeled both short and long persistence variability, the extreme cases, applied multiple regression against variability and forcings – and in both cases found the GHG signal statistically significant.

  57. jyyh says:

    considering that 90% of warming goes into ocean (as we now know it) I find the orbital forcing theory less convincing, but nevertheless it’s the only one with the proper frequencies, thus the mechanisms of glacial initiations and ends are (a bit) in dispute.

  58. Tom Curtis says:

    Scottish Sceptic: “There is a paper by Lewandowsky …”
    Name, Journal and date of publication please. Citing a reference while hiding the information that would actually allow people to look it up and check its claims is a feature of pseudo-skepticism.

    Mod: You have edited ScSc words when quoted by Andrew Adams, but not in the original comment @7:38 pm

  59. Tom, actually Andrew did that himself, but you make a good point.

  60. John Mashey says:

    “There still seems to be quite a lot of disagreement about climate reconstructions (in the details at least)” etc.
    I think I disagree with the general idea, or at least would phrase it along Feynman’s quote:’ I have approximate answers and possible beliefs, in different degrees of certainty, about different things. But I’m not absolutely sure of anything and of many things I don’t know anything about’

    Consensus sounds like a binary choice, but the clear reality of climate science is:
    a) There is a huge set of agreements on many facts, as seen in IPCC WG I, to within quite narrow confidence intervals.
    b) There is a large set of propositions around which there are probability distributions, where the literature and experts end up with a range. The question is: are the ranges narrow enough or not to take appropriate action or not? For instance, regional models sometimes disagree, but I doubt any recent models predict that Hadley Cells will shrink not expand, and the inevitable result is that the SW US gets drier, and the Northern tier gets wetter. They disagree about what happens Northern California. They tall agree there will be less snowpack here.
    The IPCC goes to great lengths to try to specify confidence intervals when possible or express some indication of level of scientific understanding, far more pervasively than most. At least some of this was from our deceased friend Steve Schneider, who campaigned persistently to recognize and express levels of uncertainty, which after all, helps identify problems that need more attention. Scientists are always arguing around the margins, and some effects are new enough there really isn’t good shot-duration paleo evidence to help. For instance, we are clearing getting into a temperature and rate of change domain that has not been usefully seen for Greenland.

    The problem with a binary consensus/noconsensus framing is:
    10 cardiologists say there’s a good chance one will die if you don’t have an operation. One asks how long? and they range from 6-12 months, although most are in the 8-10 range.
    Clearly, there is no consensus, so one can wait until they are sure – more research is needed.

    Finally, as for reconstructions:
    a) Different statistical techniques give different (temperature) curves for the same datasets, but the competent ones tend to have overlapping confidence intervals.’
    b) But often, “spaghetti graphs” of temperature accidentally convey a misimpression, because they often do not show the confidence intervals, and often they measure different geographies, which naturally differ, especially in variability at some timescales. IPCC AR4 Fig..6.10(b) illustrates this well. Some of those are NH (0-90), some are either 30-90 or 23.5-90, (i.e., only 50 or 60% of the NH) and then they vary in concentration of proximity to North Atlantic basis, which matters. One really needs to read the details and sometimes look at the actual papers. Given the different land/ocean mixes and ice/snow-albedo feedback, one expects North-dominated geographies to jiggle more … which is part of the reason that some reconstructions gyrated a bit more than the MBH99 hockeystick which tried to do full NH.
    Anyway, that’s why I like IPCC FIg.6.10(c) better.

  61. John Mashey says:

    Oops, see also my review of Measuring the Universe, which I liked because it considered the history of different sorts of measurements, including errors and changes of mind.

    Necessarily, students must be taught current state of knowledge, but too often, that loses the paths by which the knowledge was established. Of course, the climate classic is Spencer Weart’s fine The Discovery of Global Warming in book or free online.

  62. John Mashey says:

    jyyh: if you haven’t read Ruddiman, Kutzbach, Vavrus (2011), you might want to take a look. Although the main thrust is different, the comparisons of interglacials are usefully shown togetherr, with argument for the “insolation alignment” view. OF course, that just addresses timing, not the exact mechanisms on inception of glaciation.

  63. BBD says:

    jyyh

    considering that 90% of warming goes into ocean (as we now know it) I find the orbital forcing theory less convincing, but nevertheless it’s the only one with the proper frequencies, thus the mechanisms of glacial initiations and ends are (a bit) in dispute.

    Total insolation hardly changes during glacial termination under orbital forcing, rather TSI is regionally and spatially reorganised, with big increases at higher latitudes in the NH. This is mostly land (and of course at the time, ice sheets). So the surface forcing change in W/m^2 is “targeted” right on the ice itself.

  64. John,

    “There still seems to be quite a lot of disagreement about climate reconstructions (in the details at least)” etc.
    I think I disagree with the general idea, or at least would phrase it along Feynman’s quote:’ I have approximate answers and possible beliefs, in different degrees of certainty, about different things. But I’m not absolutely sure of anything and of many things I don’t know anything about’

    Yes, that may be a better way to put it. I was really just meaning that there seems to still be a healthy disagreement within the community about some of the details – at least, that’s my impression.

    KR,
    Thanks. I was aware of that paper, but I thought it was just considering the IPCC attribution statement and hence only looking at recent climate change. I was thinking more of whether or not there was any evidence, ever, for unforced variability producing a long-term change to our climate state.

    jyyh,
    Yes, I get the impression that there isn’t yet a definitive mechanism but, again, is there a chance that the main cycle could be unforced? That would seem quite remarkable given how well it compares to the orbital timescales and also why would it produce a reasonably regular looking cycle with such a long period.

  65. william says:

    hmmm

    “Examples of forced variability would be volcanic outgassing – hundreds of millions of years ago – allowing greenhouse gases to accumulate in the atmosphere, warming the planet and driving us out of a snowball earth”

    are you sure.

    “A drop of volcanic activity in the early 20th century may have had a warming effect. However, volcanoes have had very little impact on the last 40 years of global warming”

  66. William,

    are you sure.

    Never sure about anything, but pretty sure that it’s a reasonable statement. The subtlety here – that maybe you don’t get – is that volcanoes emit aerosols and CO2 and, presumably, hundreds of millions of years ago, the outgassing of CO2 was more significant that it is today. However, that’s not really the point. Aerosols have a short term cooling influence before precipitating out. Volcanic activity hundreds of millions of years ago was sufficient that GHGs could accumulate in the atmosphere to eventually provide significant warming (i.e., the aerosols didn’t accumulate, GHGs did). Today this isn’t the case, as I understand it, and so the influence of volcanoes, today, is mainly through aerosols that have a short-term cooling influence.

  67. Kevin says:

    Hi Anders,

    I think your impression that ‘there seems to still be a healthy disagreement within the community about some of the details [of millennial temperature reconstruction]’ is absolutely correct. I really don’t understand the purpose of efforts to downplay real disagreements and uncertainties present in the community, and I don’t know that it is particularly helpful to claim that all the reconstructions fall in the error bars of one or another other reconstruction, particularly when we know that methodological choices and various flavors of proxy uncertainty are important for influencing the final product. There was a paper in Climate of the Past just the other day (http://www.clim-past.net/10/1/2014/cp-10-1-2014.html) looking at various aspect of this. If all one wants is the MCA-LIA-Modern shape, Lamb’s schematic provided that in 1965.

    John,

    One of the major reasons for the difference between MBH1999 and, for instance Esper et al. 2002 (see Esper et al. 2002, Figure 3), is a methodological issue related to tree-ring detrending.

    William,

    Anders’ volcanic example is correct (CO2 outgassing from volcanoes driving post-Snowball Earth thawing vs modern volcanic eruption summer cooling or not via aerosols)

    cheers,
    Kevin

  68. William,
    I found this website which suggests that, globally, volcanoes emit – today – around 200 miillion tonnes of CO2. Human emissions are around 30 billion tonnes of CO2. Contribution via volcanoes is essentially insignificant.

    If I remember correctly, the conversion from GTCO2 to ppm is 7.81 GTCO2 = 1 ppm. So, without human emissions volcanoes would increase atmospheric concentrations by about 0.01 – 0.02 ppm per year (depending on what fraction goes into the oceans).

  69. verytallguy says:

    I think, but look to others with more expertise to correct this, that in a snowball earth situation there is essentially no carbon sink (oceans frozen over, little or no weathering), so CO2, even if outgassed volcanically at a low rate, builds up indefinitely until the snowball starts to melt.

    Once melting, massive positive albedo feedback kicks in and the climate flips back to non-snowball.

  70. vtg,
    Yes, that is – I believe – correct.

  71. BBD says:

    Kevin

    I really don’t understand the purpose of efforts to downplay real disagreements and uncertainties present in the community

    I suspect that this is a reactive response to the sustained assault on certain individual scientists and their work by pseudosceptics. It is unfortunate, even regrettable, but perhaps understandable in the broader political context of the “climate wars”. I am not suggesting that it should be condoned.

  72. BBD says:

    vtg and ATTP

    That’s my understanding too: no carbon sinks but ongoing volcanism. A good example from paleoclimate of the efficacy of CO2 as a forcing. It can – eventually – overcome the massive albedo of a SE.

  73. Kevin,

    I really don’t understand the purpose of efforts to downplay real disagreements and uncertainties present in the community

    The problem, as I see it, is that there are some who appear to be using these disagreement to claim that MCA-LIA-Modern shape is wrong. If this only happened on blogs then one could obviously completely ignore it. However, I’ve also seen it claimed in mainstream UK newspapers in articles written by people who have actual influence. I agree with you completely that we shouldn’t be downplaying the disagreements that exist. It’s an extremely healthy part of scientific investigation. However, I don’t know how one then addresses the claims made in the mainstream media regarding the basic shape. Maybe nothing. Maybe the scientific evidence will eventually become obvious and whatever journalists may or may not have said will be irrelevant. On the other hand, maybe not.

  74. > If all one wants is the MCA-LIA-Modern shape, Lamb’s schematic provided that in 1965.

    Some might want to exploit the Lamb story (say to reminisce over the Deming Affair) and still downplay this schematic, like Wegman did under oath:

    [E]ssentially a cartoon.

    http://neverendingaudit.tumblr.com/post/1498489564

  75. OPatrick says:

    I suspect that this is a reactive response to the sustained assault on certain individual scientists and their work by pseudosceptics. It is unfortunate, even regrettable, but perhaps understandable in the broader political context of the “climate wars”.

    And I would say the perception that people are downplaying real disagreements and uncertainties is much greater than the reality because, whilst many are spending a disproportionate amount of time focused on these issues, much of it is actually challenging the exaggerated and misleading claims that ‘sceptics’ put forward. I suspect this may be a deliberate, if not necessarily conscious, strategy on the part of those ‘sceptics’.

  76. John Mashey says:

    Kevin: just to clarify, maybe I didn’t say that as precisely as possible.
    Indeed there are real differences from statistical methodology, and arguments on those are important and worthy of arguing about. I don’t mean to downplay that.
    I don’t think all the reasonable reconstructions are always in each other’s error bars… Although I do think most are usually there if they started from ~same set of proxies.

    My problem is that my first decade post-PhD was in a place where statistics thinking was rather influenced by John Tukey and reinforced by later experience in computer performance benchmarking. I’d summarize by saying:
    My experience is that when most non-specialists see a spaghetti temperature chart, their reaction is that the reconstructions disagree strongly and we really don’t know much about what was going on.
    My reaction was more like Fig 6.10(c), ie Tukey-inspired:
    Those graphs told me there were differences and a lot of rough agreement, and that people were struggling hard to extract information from messy, incomplete data sets, and that results would improve over time, and some statistical techniques would fall by the wayside. Also more data would be added, although we’d never get the equivalent of modern coverage. I really, really hate graphs with the standard error bars, except for experienced people: visually, they over-emphasize the extremes, as do spaghetti charts.
    I’ve had this same sort if discussion over at Andrew Gelman’s blog, and people are experimenting with better graphic techniques. But it is fundamentally hard to use static 2d graphs of multiple curves to express how much agreement/disagreement there is.
    Most spaghetti graphs have no error bars, especially not multiples.
    Almost no one outside specialists digs out the papers and checks the geographies, and seeing lines on same graph that claim to significantly different geographies makes it way worse. (Whether or not any given reconstruction does a good job is a real argument, for specialists, and hopefully leads to improvements.)

    Anyway, none of this argument drives from the deliberate efforts of people to confuse things and create doubt about reconstructions, it is for something I was long taught to guard against: over-interpreting graphs, especially line graphs, when one doesn’t know the details and the error bars. Experts usually know, but not the general population, so this is primarily a challenge of communicating results beyond experts.
    (Note: at Bell Labs, I built a group that was half software and half cognitive psychologists who worried about information displays and understanding. Tukey’s warnings urgings for EDA, graphical displays for insight and avoidance if over-interpretation of sparse data were relevant. Of course, having spent some years as a Chief Scientist @ SGI, static 2-d graphs were not my favorite things. I fought a long, mostly successful battle (via SPEC) to improve the meaningfulness and presentation of multiple benchmark sets, but could not quite talk people into incorporating the right uncertainty metrics… impart because of graphical limitations. This whole topic does come up now and then over at Andrew Gelman’s… And I do have some hope, but as- communication mechanism, I certainly like Fig 6.10(c) better than ( b).

  77. William says:

    Volcanoes.

    In the past, not insignificant numbers does the heat have any bearing on the atmosphere.

    On Venus- “Researchers expected the topography data to reveal volcanic features on Venus but they were surprised to learn that at least 90% of the planet’s surface was covered by lava flows and broad shield volcanoes”

    http://geology.com/stories/13/venus-volcanoes/

    so lets assume that the lava flows at around 600 degrees+ – does this have any effect on the atmosphere.

  78. William,

    so lets assume that the lava flows at around 600 degrees+ – does this have any effect on the atmosphere.

    I don’t think so. What makes you think that it would? Volcanoes are thought to be responsible for the creation of the atmosphere through volcanic outgassing. This probably happened when the Earth was very young (4 billion or so years ago). I don’t think that lava flows 500 million years ago would have much impact on the atmosphere. As others have pointed out, the lack of an ocean (or at least having one covered in ice) meant that the was no CO2 sink and so the volcanic activity did not need to be substantial to lead to the accumulation of GHGs in the atmosphere.

  79. william says:

    “I don’t think so. What makes you think that it would? ”

    well 90% of the planet covered with lava flows at a possible 600+ degrees. A lot of heat traveling somewhere.

    What height would this heat have an effect.

  80. William,
    I don’t think the Earth had 90% of it’s surface covered with lava flows 500 million years ago. 4 billion years ago, maybe, but not 500 million years ago.

  81. Joshua says:

    An off-topic observation (hopefully, it wont’ be moderated out):

    One interesting aspect of this blog is that the Physicist Formerly Known as Wotts is, as a rule, very careful to acknowledge uncertainty, the limits of his/her knowledge, and to use caveats when appropriate.

    Yet, instead of engaging him/her in respectful, good-faith dialog, we have a number of “skeptics” (although not all), who insist in engaging him/her in bad faith and on trying to judge him/her by climate scientists that those “skeptics” feel do not acknowledge uncertainty, the limits of their knowledge, or use caveats when appropriate.

    It’s almost enough to make one think that some “skeptics” are actually more interested in confirming their biases than in engaging in open discussion about climate change when the opportunity presents itself.

    That all said, I want to offer the observation that despite the nonsense that does take place here sometimes – nonsense that is indistinguishable from the nonsense that pervades the climate blogosphere more generally – I have seen here what seems to me to be some of the best give-and-take I’ve seen yet between smart and knowledgeable people with contrasting views on climate change. It does seem that some folks here are actually interested in listening to, and learning from, each other despite not agreeing with each other to a large degree.

    I hope the following comment will not be in violation of the spirit of the blog moderation rules, but I am particularly interested that BBD – who I had tended to think of as a climate warrior – has prominently figured in much of the notable good-faith exchange.

    My sense is that this is taking place because the Physicist Formerly Known as Wotts has stated a specific intent w/r/t engaging in dialog – one of learning from others to test his/her own understanding, and in an open and transparent way has struggled with how to maintain that goal and admitted that he/she has, at times, made missteps. I feel that generally, the type of exchange that takes place on a particular blog is largely a reflection of the blogger. Bloggers have more control over the dialog than they often seem to recognize, and moderation policy is only one (and I would argue relatively less significant than people seem to think) factor that influences the quality of exchange that takes place in their blog. IMO, the most influential factor is the mechanism by which a stated goal or intent becomes evident to the blog’s readers.

    As an educator, the dynamic reminds me of when educators hand-wring about the large number of students who cheat, and then struggle with enforcement mechanisms to prevent cheating – without realizing that there is an underlying mechanism that motivates students to cheat, a mechanism in which they, as educators, play a very important role. How can a teacher look at a student that thinks that an empty education is the goal, and not look at what they do as educators that led the student to that fallacious belief? How can a blogger look at commenters who engage in bad-faith dialog, not look at what they do themselves as bloggers that engenders such an approach in those who comment at their site?

    And a nod to SOD, also – as as “skeptic” (I use the quotation marks there to be consistent and not to imply that SOD’s skepticism is faux), whose input has moved the dialog in a positive manner, IMO.

  82. William,
    Maybe I should be more careful in my response. The article you’re referring to uses cratering from impacts on Venus to indicate that the surface is mostly less than 500 million years old. This is probably similar to the Earth in that surface is probably also mostly less than 500 million years old. That doesn’t mean, however, that 500 million years ago 90% of the surface was covered in volcanoes. It simply means that volcanoes and other geological activities (erosion, tectonic activity) have resurfaced the planet on 100 – or so – million year timescales.

  83. Joshua,
    Thanks, I appreciate the feedback and the compliments 🙂

  84. BBD says:

    Thanks Joshua. I like to think I have a range of modes of expression…

    😉

    But more seriously, I do agree with what you say about the Power of the ATTP. It’s just not the place you misbehave in. Wouldn’t feel right. And Then There’s The Squirrel… you just don’t.

  85. Rachel says:

    Joshua,
    I agree with you completely. I feel that our physicist friend here has established a very open and respectful atmosphere and although I know that sometimes people get frustrated/angry, I think this is generally ok and to be expected – I also don’t believe in repressing emotions – just as long as people continue to treat each other with respect. And as for BBD, I don’t think I have moderated any of his comments. Ever.

  86. oarobin says:

    Kevin,
    I don’t know that it is particularly helpful to claim that all the reconstructions fall in the error bars of one or another other reconstruction, particularly when we know that methodological choices and various flavors of proxy uncertainty are important for influencing the final product.

    one of the problems that i find in the public discussions of paleo-climate millennium reconstructions (PMR) is that the space is dominated by discussions about Michael Mann and the methodological choices he makes to solve some of the paleo-climate issues he faces while the underlying issues and other methods of addressing the issues are almost never discussed.

    the most effective one to redress this balance, IMHO, is to just provide more public discussion / content on these issues and other methods so that an interested individual may learn with some guidance from the professionals. One shouldn’t complain too much about the state of the public knowledge when hands down climateaudit has the most PMR blog content and arguable Gavin Schmidt has arguable had the most PMR content from a climate professional, usually in response to these blogs. where are the views of the other PMR experts from a public discussion standpoint?

    in this light and to give a broader perspective can you give us like a top 5 list of issues that you think are important and the range of views and researchers working on them?

    from a lay person perspective some of the issues i find fascinating and i think would move the public discussion forward are:

    I. The biological and physical and statistical motivations for choosing some indicator as climate proxies.
    II. the relative magnitude of the various uncertainties that are in the PMR reconstructions, and how they interact.
    III. what is a broad outline of the mostly settled science, the reasonable debated but inconclusive science and the hot debated cutting edge science.
    IV. The tradeoffs of inverse vs forward modelling especially an evaluation of how good the Bayesian Hierarchical Methods are? and how to compare methods?

  87. John Mashey says:

    Back at desk, where bookmarks are:
    1) See Graphs showing uncertainty using lighter intensities for the lines that go further from the center, to de-emphasize the edges. Gelman is a *very* good statistician, with a lot of interest in graphical display techniques. The comments are worth reading.

    2) that was followed by Graphs showing regression uncertainty: the code!.
    Again, more useful comments, including links.

    3) Simon Hsiang has done much work on different kinds of graphics, including what he calls watercolor regression. Again, comments worth reading,

    4) Felix Schönbrodt pointed at Visually weighted/ Watercolor Plots, new variants: Please vote!, which had more examples, and again, good discussion, at least a few of which were concerned with using such displays for time-series data.

    Anyway, good visual display of quantitative information (thanks Tufte) is hard work, and a weird mixture of science, statistics, art and sometimes cognitive psychology, as it matters very much whether a display is intended for:
    a) Subject-matter experts, who know about the uncertainties or who may have read the base papers
    b) Non-subject experts, but used to such data analysis
    c) The general public.

    It is hard enough to show a single time-series well, in a way that does not mislead the general public. It is far harder to show multiple time-series in ways that are both accurate and convey the level of agreement/disagreement. Human eyes tend to be drawn to extrema and high-frequency changes, somewhat related to the “twinkling” problem of moving shapes on computer graphic displays, for which “anti-aliasing” methods had to be created, once upon a time very expensive. (Early simulators for fighter planes tended to make the edges of planes twinkle as they changed aspect, and pilots unconsciously got used to that, but in real life, they don’t twinkle, and that had to get fixed.) Typical regression lines or plots with error bars visually over-emphasize the outer edges, and the spaghetti graphs tend to visually over-emphasize the edges of the envelope.
    Nick Stokes did a nice interactive, More proxy plots with JavascriptL, which at least lets you rollover and call out individual plots from the spaghetti.

    Of course *not* having some visual indication of uncertainty is even worse.
    I liked the progression from the usual error bars in MBH99 to the gray zone of the TAR, but then sometimes the gray zone disappeared, not good. (That ignores arguments about the size of the bars: I was worried that the pre-1400AD bars weren’t bigger.)

    Anyway, this is not a complaint at IPCC or paleo folks: as seen from the arguments at Gelman’s, a lot of people struggle with good visual expressions of this material suitable for different levels of audience, and those quite fine for a specialist science paper may have unintended effects for general audiences. It would also help if graphing tools got routinely better, although static 2D printed material has inherent limits.

  88. Rob Painting says:

    I see the flavour of posts has veered off toward hockey stick territory. Of course millenial scale proxy-based temperature reconstructions by their very nature have significant uncertainty, but the sea level trend throughout the Holocene does suggest that current sea level rise is anomalous within the context of the last 10,000 years. With change in global land-based ice volume being a very good indicator of global temperature, it therefore follows that current warming is likewise anomalous too.

    Probably a candidate for further discussion at a later time……

  89. John Mashey says:

    I wouldn’t really call this hockey-stick territory, that’s just one example of a more general issue under the topic of the post.
    It is a very important *attribution* question to sort out the causes of variability:
    natural internal variability, which sometimes, with better understanding turn into the others:
    natural forced variability, as from Milankovitch, volcanoes, solar insollation
    human forced variability
    timescales

    The paleo record is one of the major sources of input, and not just for temperature, but for precipitation, sea levlel rise, atmospheric composition, and all the various ocean oscillations and occasional massive state changes (like Younger Dryas). There is no doubt that we’re making a major change from I.R. onward, and I think people are coming around to Ruddiman’s views given all the research in the last 10 years, Try p.14 of Strange Scholarship..:

    ‘The Earth‘s future temperature track will be determined by:
     The current state of the Earth, especially total energy content, glacier masses, vegetation coverage and other factors that affect Earth‘s
    overall albedo, i.e., fraction of energy reflected into space without creating heat. Ice and snow reflect more than oceans.
     Biological, chemical and physical processes.
     Human choices, with ―public policy implications,
    Future natural temperatures simply do not depend on the temperature in 1000AD or on our knowing anything about it. Wegman even said this in testimony, A.3. However, better understanding of the past‘s natural variability helps researchers calibrate climate models, which is why researchers argue fiercely over the shape and jiggles of the shaft. If people somehow got a full set modern-grade temperature measurements from 1000AD onward, nothing would change for the future except our ability to forecast it better.
    The uncertainty limits on each emissions scenario represent huge differences of impacts and costs , so narrowing those limits helps inform human choices.’
    See p.142, where I discussed the spaghetti-graph issue, having had to overly a bunch of graphs.

    Of course, the other issue is And Then There’s Physics :-), which generally make me suspicious of rapid temperature jiggles on decadal/multi-decadal timescales over large areas like hemispheres. On really wants to have explanations for them that work with physics, i.e., conservation of energy is tough to beat. As per Younger Dryas, transitions *can* happen very fast, but usually , they don’t. When there’s a point on a reconstruction, really one should think of it as:
    a) a bounding box for +/- date uncertainty and +/- value uncertainty but even better is:
    b) a 2-D probability distribution function, i.e., likely denser in the middle and least dense at the corners.

    Although there is uncertainty in each of several points, there are additional bounds imposed by physics on the speed of change in the line connecting them. Among the arguments on Marcott, etal last year (see this at Bart Verheggen’s, for example, various people argued that their reconstruction could have missed rapid gyrations. That is absolutely possible, but at various durations, rapid up-and-down jiggles are bounded by physics. If you take a temperature reading approximately hourly, it might be possible for a 30-degree spike to have happened in between measurements … but hard..

    Anyway, I think there are lots of interesting problems, and techniques to be newly applied, as in Cowtan and Way(2013), in some sense, no surprise, but nice application of technique to reduce uncertainty.

    As always, see Tukey quotes:
    ‘The combination of some data and an aching desire for an answer does not ensure that a reasonable answer can be extracted from a given body of data. ‘
    and
    Far better an approximate answer to the right question, which is often vague, than an exact answer to the wrong question, which can always be made precise.

    It’s really hard to say much about forced/unforced variability without characterizing the uncertainties.

  90. > Bull Ruddiman’s new book

    That’ll be the more forceful version of Bill Ruddiman’s?

    As for the topic: people used to be moderately convinced that the D-O oscillations you see during the last ice age were unforced variability. Charge-discharge cycles of the NH ice sheets were a popular mechanism. https://en.wikipedia.org/wiki/Dansgaard%E2%80%93Oeschger_event. I don’t think that ever got sorted out.

  91. William,
    Thanks, interesting. Is there an easy way to explain what is meant by a charge-discharge cycle in the NH ice sheet?

  92. I could wave my hands, I think that’show its done. The wiki page says:

    The events may be caused by an amplification of solar forcings, or by a cause internal to the earth system – either a “binge-purge” cycle of ice sheets accumulating so much mass they become unstable, as postulated for Heinrich events, or an oscillation in deep ocean currents (Maslin et al.. 2001, p25).

    The idea is, kinda, that the ice sheets grow and then – bing – realise that they’re too big. They grow enough for pressure melting underneath that lubricates their discharge fora while, till they run out of pressure and then stick? That sort of thing.

    See-also the “causes” section of https://en.wikipedia.org/wiki/Heinrich_event

  93. Thanks. Hand waving is my personal preference, in many circumstances.

  94. jyyh says:

    -thanks John Mashey for the link to the Ruddiman et al. paper, I think I saw the glacial-interglacial transposition image somewhere before but I haven’t seen the whole article. ATTP, local solar forcing is till the best theory for synchrony but which is the locality? Heat in the ocean diminish the carbonate in area so a strong enough local forcing Anywhere might be enough. This is far from easy.

  95. John Mashey says:

    jyyh:David Archer’s The Long Thaw Chapter 5 is good general discussion.
    Summer insolation ~60-65degN degN is usually considered a good indicator.
    Ruddiman’s Earth Transformed pp.5-13 discusses it.

    See IPCC AR4:
    “Starting with the ice ages that have come and gone in regular cycles for the past nearly three million years, there is strong evidence that these are linked to regular variations in the Earth’s orbit around the Sun, the so-called Milankovitch cycles (Figure 1). These cycles change the amount of solar radiation received at each latitude in each season (but hardly affect the global annual mean), and they can be calculated with astronomical precision. There is still some discussion about how exactly this starts and ends ice ages, but many studies suggest that the amount of summer sunshine on northern continents is crucial: if it drops below a critical value, snow from the past winter does not melt away in summer and an ice sheet starts to grow as more and more snow accumulates. ”

    That’s a bit old, but gives the general idea.
    Think of each himisphere as 3 zones, over last few million years::
    a) Always ice on land: Greenland, Antarctica
    b) Never ice on land
    c) Sometimes ice on land, and that part responds to the orbital state.
    In that zone, summer insolation might be just warm enough to melt the winter’s snow, or it might be more (ice retreats) or less (ice grows).
    Anyway NH at that latitude has plenty of land, and ice-albedo feedback works.
    In the SH, there is almost no land there, and Antarctica stays cold.

    If CO2 were lower than the usual 180ppm minimum, ice sheets would grow nearer to the Equator, but not melt back as far, i.e., the key latitude would be less.

    A local version of this can be seen around ski slopes, where keeping the ground covered with snow keeps it colder than just across the street,

  96. OPatrick says:

    Also, although it may be true that we haven’t definitively shown, for example, that Milankovitch cycle are the drivers for major climate change in the last 500000 years, the timings seem so similar that it, at least, seems plausible that this is the driver, even if we haven’t definitively shown how it actually operated

    Is there any way to quantify the degree of correlation between the (combination of?) cycles and the pattern of glaciation over the past 500 000 years? Has this been done? I’ve been skimming through SoD’s ghost posts but didn’t see anything directly on this quantification – I certainly haven’t read carefully or deeply though, so my not seeing something doesn’t say much.

    My naïve view is that if there is a clear correlation then this is conclusive evidence for causation (forcing) in this case, given that the normal rules that if A and B correlate it cannot be concluded that B causes A as it could be that A causes B or that C causes A and B do not apply – the Milankovitch cycles cannot in any sense be said to be caused by glaciation or anything else that could coincidentally also cause glaciation, other than the laws of physics, I suppose. Is this too naïve?

  97. OPatrick,
    I think one has to be very careful with corellation/causation, but – as you say – given that changes to our climate can’t drive changes in our orbit, it would seem reasonable to conclude that it is as least likely that if they are correlated that it’s the orbital variations that are driving the climate variations. Although, as others have pointed, precisely how this works is not yet agreed (although there are various ideas regards latitudonal variations in insolation).

  98. BBD says:

    The correlation between orbital variability (specifically the 41ka obliquity cycle) and deglaciations goes back 2.8Ma (marine oxygen isotope analysis of ocean sediment cores: Lisiecki & Raymo 2005). There’s a period called the Mid-Pleistocene Transition (MPT) from 1.2Ma – 0.7Ma when after which the frequency of deglaciations changes to a 2x or 3x multiple of obliquity cycle, so they occur at 82 – 123ka intervals (Huybers & Wunsch 2005). Antarctic ice cores provide additional information for the last 750ka (EPICA Community Members 2004).

  99. BBD says:

    There are vast piles of studies bolstering all this, notably those using SSTs derived from Mg/Ca ratios in shelly forams (ocean sediment core analysis) and Chinese loess (windblown dust) deposits, the latter providing continuous terrestrial records going back 2.5Ma.

  100. John Mashey says:

    One more time: there is overpowering evidence that the ice ages are driven primarily by summer solar insolation in the relevant latitude zone of the NH, while the Earth:
    a) Has continents in the right places
    b) Has the right range of levels for CO2.

    The Ruddiman paper I mentioned shows the insolation curves back to Isotope Stage 19, 700K+ years ago. Bill did a much research(and heavily-cited) in paleooceanography, one of the major evidence sources for this.

    All this is well-explained in multiple sources at various levels of detail, from the general-public explanations in The Long Thaw to the large scientific literature to the IPCC assessments.

  101. BBD says:

    Indeed he did. The “Ruddiman belts” being the depositional layers of ice-rafted debris in the N Atlantic that (I think) led the way to the identification of Heinrich events.

  102. Kevin said:

    3. In general, at the low frequency (~long term) and larger scale the forced response will be more important vs. the local-to-regional scale and the shorter term, where the internal variability will contribute more substantially.

    What is the evidence for this? (Evidence for forcing response being larger scale and longer term)?

  103. John Mashey said:

    One more time: there is overpowering evidence that the ice ages are driven primarily by summer solar insolation in the relevant latitude zone of the NH, while the Earth:
    a) Has continents in the right places
    b) Has the right range of levels for CO2.

    What is the overpowering evidence? And when you say ‘the ice ages are driven primarily’ do you mean the 100k time periods or the shorter interval periods corresponding to 40k and 19k/23k?

    1. Why did Termination II start around 135kyrs ago?
    2. Why did the last ice age end 18 kyrs ago (start of Termination I)?
    3. What was the summer solar insolation at these times at the relevant latitude?
    4. And why, when this same – or higher – summer solar insolation occurred prior to the termination didn’t the ice age end?

  104. SoD

    What is the evidence for this? (Evidence for forcing response being larger scale and longer term)?

    Maybe I’ve misunderstood what you’re asking, but I didn’t think that Kevin was suggesting that the forcing response was larger scale and long-term. I thought he was simply suggesting that the climate changes that are large scale and longer term are more likely to be forced than unforced.

    A more general question. You seem to be suggesting that the climate variations that are often associated with Milankovitch cycles could be unforced. Am I right in thinking that? If so, is there a suggested mechanism and would one then simply argue that the timing similarity (with the orbital cycles) is just coincidental? Or are you simply suggesting that we don’t yet have a largely accepted mechanism, but that the climate variations (the long-timescales ones at least) are likely forced.

    Another quick question. If we could show that some previous climate variations were unforced, would – in terms of what we’re undergoing today – actually matter?

  105. OPatrick says:

    …the timing similarity…

    And can I just add as an addendum my earlier question – how strong is this similarity? Can it be quantified? Has it been quantified? (I didn’t see anything in BBD’s links that quantified it, but I may have missed it.)

  106. Andthentheresphysics

    Maybe I’ve misunderstood what you’re asking, but I didn’t think that Kevin was suggesting that the forcing response was larger scale and long-term. I thought he was simply suggesting that the climate changes that are large scale and longer term are more likely to be forced than unforced.

    I thought that was his suggestion as well. Let’s take that as the proposition. What is the evidence?

    A more general question. You seem to be suggesting that the climate variations that are often associated with Milankovitch cycles could be unforced. Am I right in thinking that?

    You are correct that I believe it is plausible.

    If so, is there a suggested mechanism and would one then simply argue that the timing similarity (with the orbital cycles) is just coincidental? Or are you simply suggesting that we don’t yet have a largely accepted mechanism, but that the climate variations (the long-timescales ones at least) are likely forced.

    If there was a plausible mechanism that linked orbital changes to the timing of the start and end of the ice ages, then the answer would be a lot easier. The timing of the major cycles doesn’t correspond to any plausible mechanism so we are all starting from the same point. Even if we don’t realize it.

    And what exactly is the timing of the ice age terminations? It appears that T2 termination – T1 termination is 117 kyrs. What physical mechanism relates to this period?

    I don’t have answers. Accepting what the data tells us and rejecting a theory that fails to link the data to physics is a healthier start than accepting a popular theory because it is popular but works by some mystery mechanism. (Spoiler alert – next commenter will explain “a mechanism” and then I will provide (in vain?) 10 papers that say ‘the mechanism is still not explained’ and another 5 papers with mutually exclusive explanations of this mechanism, all by reputable climate scientists)

    Another quick question. If we could show that some previous climate variations were unforced, would – in terms of what we’re undergoing today – actually matter?

    I don’t know. But I expect that knowing more about the climate of the past will be useful for predicting the future.

  107. Tom Curtis says:

    Huybers and Wunsch, 2005:

    “The 100,000-year timescale in the glacial/interglacial cycles of the late Pleistocene epoch (the past ,700,000 years) is commonly attributed to control by variations in the Earth’s orbit. This hypothesis has inspired models that depend on the Earth’s obliquity (,40,000 yr; ,40 kyr), orbital eccentricity (,100 kyr) and precessional (,20 kyr) fluctuations, with the emphasis usually on eccentricity and precessional forcing. According to a contrasting hypothesis, the glacial cycles arise primarily because of random internal climate variability. Taking these two perspectives together, there are currently more than thirty different models of the seven late-Pleistocene glacial cycles. Here we present a statistical test of the orbital forcing hypothesis, focusing on the rapid deglaciation events known as terminations. According to our analysis, the null hypothesis that glacial terminations are independent of obliquity can be rejected at the 5% significance level, whereas the corresponding
    null hypotheses for eccentricity and precession cannot be rejected. The simplest inference consistent with the test results is that the ice sheets terminated every second or third obliquity cycle at times of high obliquity, similar to the original proposal by Milankovitch. We also present simple stochastic and deterministic models that describe the timing of the late-Pleistocene glacial terminations purely in terms of obliquity forcing.

    (My emphasis)

    Huybers, 2006:

    “An agemodel not relying upon orbital assumptions is estimated over the last 2Ma using depth in marine sediment cores as a proxy for time. Agemodel uncertainty averages !10Ka in the early Pleistocene (“2–1 Ma) and !7Ka in the late Pleistocene (“1Ma to the present). Twelve benthic and five planktic d18 O records are pinned to the agemodel and averaged together to provide a record of glacial variability. Major deglaciation features are identified over the last 2Ma and a remarkable 33 out of 36 occur when Earth’s obliquity is anomalously large. During the early Pleistocene deglaciations occur nearly every obliquity cycle giving a 40 Ka timescale, while late Pleistocene deglaciations more often skip one or two obliquity beats, corresponding to 80 or 120 Ka glacial cycles which, on average, give the “100Ka variability. This continuous obliquity pacing indicates that the glacial theory can be simplified. An explanation for the “100Ka glacial cycles only requires a change in the likelihood of skipping an obliquity cycle, rather than new sources
    of long-period variability. Furthermore, changes in glacial variability are not marked by any single transition so much as they exhibit a steady progression over the entire Pleistocene. The mean, variance, skewness, and timescale associated with the glacial cycles all exhibit an approximately linear trend over the last 2 Ma. A simple model having an obliquity modulated threshold and only
    three adjustable parameters is shown to reproduce the trends, timing, and spectral evolution associated with the Pleistocene glacial variability.”

  108. OPatrick,

    Is there any way to quantify the degree of correlation between the (combination of?) cycles and the pattern of glaciation over the past 500 000 years? Has this been done? I’ve been skimming through SoD’s ghost posts but didn’t see anything directly on this quantification – I certainly haven’t read carefully or deeply though, so my not seeing something doesn’t say much.

    I haven’t covered it yet.

    In brief, over-simplifying the last 40 years of discussion – eccentricity has a 100 kyr period but has a tiny impact on annual insolation.

    When you look at the spectrum of insolation at any given latitude and season 2 frequency bands only stand out – 40 kyr and 20 kyr. No 100 kyr frequency is seen.

    When you look at the spectrum of climate change the 100 kyr band stands out, with 40 kyr and 20 kyr as much less important bands.

    Lots of papers point out that there is no mechanism linking the 100 kyr eccentricity cycle to glaciation/deglaciation. Lots of other papers propose (mutually exclusive) mechanisms for how insolation might be the cause of the ice age inception / ice age terminations.

  109. OPatrick says:

    Apologies to BBD – I thought I had looked at the Huybers and Wunch paper, but may have mistakenly clicked on the wrong tab on my browser! Too many papers on the go at one time. And thanks to Tom for pointing it out so clearly.

  110. OPatrick says:

    I do note, though, in the H&W paper:

    With only seven cycles, it is unclear whether adequate data will ever be available to distinguish between stochastic, simple deterministic and chaotic deterministic models of the glacial variability.

  111. Continuing OP’s point: I think its unlikely that the “100 kyr” periodicity can be pinned down exactly enough from the obs, to enable you to select (from that alone) which theory you like. You won’t get the precision (there’s another thing to be careful of, which is that since the Mtheory is so widely accepted some of the obs have a time-scale tuned to the known periods, so be careful of circularity).

    The wavelet analysis in ftp://ftp.elic.ucl.ac.be/loutre/paper/Dom/Berger_paleoc_05.pdf is quite instructive.

  112. BBD says:

    I seem not to have conveyed the point that there is no “100ky” periodicity to deglaciations.

    Thanks to Tom for emphasising this once again (although I have already posted the abstract from H&W05). The quote Tom provides from Huybers (2006) is, if anything, even clearer on this point.

    Obliquity, obliquity, obliquity.

    😉

  113. BBD says:

    Just in case, may I repeat myself?

    The correlation between orbital variability (specifically the 41ka obliquity cycle) and deglaciations goes back 2.8Ma (marine oxygen isotope analysis of ocean sediment cores: Lisiecki & Raymo 2005). There’s a period called the Mid-Pleistocene Transition (MPT) from 1.2Ma – 0.7Ma when after which the frequency of deglaciations changes to a 2x or 3x multiple of the obliquity cycle, so they occur at 82 – 123ka intervals (Huybers & Wunsch 2005). Antarctic ice cores provide additional information for the last 750ka (EPICA Community Members 2004).

  114. OPatrick says:

    My limited understanding (I know Anders often self-deprecates in this way but I’m on a whole different scale) is that the 40k year obliquity, as BBD has just restated is the key cycle but that the correlation is only with every second or third cycle. This seems an unusual sort of pattern (or is it?) so are there common statistical tests to identify correlation in such cases?

  115. I don’t think that’s right. See for example fig 1 of ftp://ftp.elic.ucl.ac.be/loutre/paper/Dom/Berger_paleoc_05.pdf, which I ref’d above. There is significant power at ~100 kyr, at least over the last ~600 kyr. It bleeds down to ~95, but not to 82. The idea that 100 is a mix of 2 and 3 times 40 (http://www.people.fas.harvard.edu/~phuybers/Doc/pace_nature2005.pdf) is cute, but I don’t see how it can be borne out.

  116. BBD says:

    Well, it’s developed in Huybers (2006) – see Tom’s link above. The obliquity cycle seems to be key (remember also the 41ka pacing pre-MPT back to ~2.8Ma – the one that Milankovitch didn’t predict).

  117. OPatrick says:

    Am I roughly right in thinking that wavelet analysis picks out regular, superimposed cycles in noisy data? But would it identify a cycle where only the second or third oscillation, on an irregular pattern, is apparent?

  118. An interesting discussion has been going on in my absence. Thanks.

    SoD,

    I thought that was his suggestion as well. Let’s take that as the proposition. What is the evidence?

    Okay, so I probably misunderstood. Can we take this a step further. As I understand it there is similarities in the timings and we have latitudonal variations in insolation that one might expect to have some effect on our climate. That doesn’t mean that this is the driver, but is suggestive and – I would argue – that just because we can’t explain precisely how these orbital variations drive climate change doesn’t then imply that unforced variability suddenly becomes viable. I’m not saying it can’t be unforced, simply that the evidence for it being forced is stronger than the evidence for it being unforced. What I guess I’m getting at is that simply because we don’t understand something fully doesn’t then suggest that an alternative becomes equally likely. Do you agree/disagree with this?

    I don’t know. But I expect that knowing more about the climate of the past will be useful for predicting the future.

    Yes, I agree but this is probably always true. What I was getting at was that if we could show that some significant changes to our climate in the past have been unforced, it doesn’t suddenly suggest that current changes to our climate could be unforced. It seems fairly clear that current changes are largely forced (anthropogenic GHGs). If anything, showing that our climate could be sensitive to unforced variability could make the future even more uncertain.

    A final point. Is there not some heuristic argument against our climate being very sensitive to unforced variability. Wouldn’t we expect it to be much less stable than it appears to be if it really is sensitive to unforced variability? I don’t know the answer to that, so just thought I’d throw that out there.

  119. Andthentheresphysics

    On the earlier point:

    Maybe I’ve misunderstood what you’re asking, but I didn’t think that Kevin was suggesting that the forcing response was larger scale and long-term. I thought he was simply suggesting that the climate changes that are large scale and longer term are more likely to be forced than unforced.

    I agreed we both had the same understanding of his point.
    Your response in the comment above doesn’t seem to address any evidence for this at all.

    Maybe you are addressing points related to whether there is evidence that orbital forcing starts and ends ice ages. I will respond on that in a separate comment.

    Can you, or Kevin, write the statement a little more specifically, and provide some argument in support.

    Let me suggest a specific proposition, in the spirit of teasing out the real claim:

    A temperature variation of 1’C over 100 years has the same effect on climate (causes the same climate response) whether it results from an external forcing or internal variation.

  120. Andthentheresphysics

    A final point. Is there not some heuristic argument against our climate being very sensitive to unforced variability. Wouldn’t we expect it to be much less stable than it appears to be if it really is sensitive to unforced variability? I don’t know the answer to that, so just thought I’d throw that out there.

    I think maybe you have a strange idea about the causes, implications and results of variation in complex systems. Lorenz has good examples in his papers, some derived from simplified forms of basic atmospheric air movements.

    I’m a beginner in that subject although I’ve read a few books and played around with some numerical modeling of chaotic systems. Maybe I don’t understand something fundamental.

    But, if you take a complex system with chaotic behavior and apply some forcing you will find lots of modes and cycles appearing that are of different frequencies to the applied forcing. The system can overall be very stable, that depends on the kinds of dissipations and feedbacks in the system.

    These modes and different frequencies that appear are natural and inevitable results of the nonlinear interactions within the system. It implies nothing about the stability of the system. Lots of modes would appear to an observer to clearly have ‘a forcing’, even though they don’t.

    I don’t think I’m doing a good job of explaining it in a comment.

  121. Kevin says:

    SoD,

    The way you’ve phrased your proposition makes me think we’re talking about different things. An observation of 1C variation over a century _is_ the climate response we’re talking about, so I’m not sure what you mean by that temperature change _causing_ a climate response (unless you’re referring to feedbacks? Which wasn’t what my statement was about at all).

    My observation was that, in general, at larger spatial scales and longer time periods, the forced component of the observed climate response is larger/more detectable, vs. shorter term (interannual) and smaller scale (local/regional) where internal variability, weather, etc. play a larger role. As one point of evidence, please see the Hegerl et al. results I described above. Goosse et al. also have a paper that talks about this in the context of the last millennium (http://www.sciencedirect.com/science/article/pii/S027737910500020X), LIA, MCA, etc. This isn’t to say there aren’t long-term, large-scale modes of internal variability — in fact, I just finished making a slide that talks about the centennial-scale internal modes of variability described by Karnauskas et al. That’s one reason I suggest the challenge, particularly at the time scales I work at (interannual to centennial), is separating forced from internal variability.

    unfortunately, I won’t be able to participate much more in this conversation over the next week, as I’m traveling and consumed by conference things (3 talks in the next 3 days).

    cheers,
    Kevin

  122. Tom Curtis replied to my comment.

    From his comments on my integrity In an earlier post, there is no point my responding to him.

  123. Kevin says:

    There is another complication, which is detecting the response of a mode of internal variability (e.g. ENSO) to an external forcing (solar variability, volcanic eruptions) — lots of work on this in my community, lots of remaining uncertainty.

    cheers,
    kevin

  124. John Mashey says:

    1) “A final point. Is there not some heuristic argument against our climate being very sensitive to unforced variability. Wouldn’t we expect it to be much less stable than it appears to be if it really is sensitive to unforced variability?”

    I suggest a different framing of that will lead to more productive discussion, because the sensitivity to unforced variations (as in ocean circulation changes, for instance}, depends on other factors of the state of the planet, Some states are much more stable than others, and for the last few My the Earth has been in a state susceptible to major gyrations from a relatively modest external forcing. (But some of the cycle arguments almost get Scafetta-like :-), since a better view is SummerSoltice65N, I don’t think ice sheets care much about clocks, but once again, the physics matters, they grow or shrink for energy reasons.

    2) See Five Myr Climate Change. We didn’t have ice age oscillations until CO2 fell far enough to allow it over last 40My or so. The continents haven’t moved enough to make much difference, orbital cycles are the same … but over the last few My, we’ve gotten into increasingly-sensitive state, t least at the multi-millennia time-scales.

    3) Positive feedbacks work in both directions, go until they run into limits or negative feedbacks that stop them For ice-age cycles, we have a predictable external forcing, approximated by solar insolation @ 65N as an approximation for the total summer insolation (correlated, but not identical), as a summary for what counts the most, the summer insolation around the margins of the ice sheet. Of course, CO2 levels matter, and as seen in the Ruddiman, etal paper, they vary.

    a) We have (ice/snow)-albedo feedback, which is positive in either direction, although melt naturally happens faster than growth.

    b) We have CO2/temperature feedback, which of course is crucial to the size of gyrations. If CO2 got low enough, one would get permanent cie sheets on Baffin Island, or further South. Of course, no longer any danger of that any time soon.

    c) Glaciers of any size grow if winter snow exceeds summer melt and vice versa, with geography mattering. Mountain glaciers show natural time filters, as per Swiss glacier records, i.e., smaller, shorter ones respond more quickly than longer ones. Multi-km ice-sheets take a while longer.

    d) Unless the planet is really warm, Antarctica stays covered, as does Greenland, mostly, but as the Earth warms, and the North American / Eurasian ice sheets shrink, eventually, they run out of land. When there are big ice sheets in those areas, unforced variations can grow or shrink the ice. When it’s warm enough that there are no big ice sheets on Baffin Island, similar temperature jiggles don’t do much.

    e) At the other end, the ice sheets come some distance South, before local summer warmth stops them, probably with help from colder/drier climate, hence less snow. Exactly how far South they come varies.

    4) I recommend Archer’s “The Long Thaw”, but this time, Chapter 12 on “trigger points”:

    “Figure 19 shows that whenever the northern hemisphere summer sunshine gets dimmer than a particular trigger value, the amount of ice on Earth grows and seal level falls, without exception through the last 800 thousand years. ….
    Summertime sunshine may not be the only factor that determines the nucleation of an ice sheet, but it sure seems like the most important factor. … But the real world is fuzzier than the trigger model…(other factors). Also, even if the trigger mechanism were all-powerful, the atual trigger value for the real world can be diagnosed from the ice volume record, but not without a lot of uncertainty.’ Then, by Fig.21, he looks at the trigger W/m^2 as a function of CO2 concentration. Ruddiman’s Earth Transformed(2013) Chapter 15 is “is the next glaciation overdue?” Back in 2009, Archer mentioned Ruddiman’s view, but wasn’t convinced; I don’t know what he thinks now, and in any case, both of them expressed good scientific discussions of uncertainties around trigger points … excellent examples for states susceptible to jiggles from positive feedbacks. But both of them certainly subscribe to solar insolation as major forcing, and I think this is a topic that has gotten a lot better understood over last 10-15 years.

    5) Of course, the coolest description of Milankovitch cycles has to be that of Richard Alley.

  125. BBD says:

    @ WMC

    I’m afraid I have to admit defeat with B&L – this paper is beyond my feeble powers of comprehension, so I can’t continue in good faith. My apologies.

  126. You don’t have to read all the equations, just look at the pictures. Its all I did. Think of the wavelet stuff as a localised Fourier transform.

  127. SoD,
    I’m on my phone so will try to respond in more detail later, but either I’m not explaining myself very well or you’re trying very hard not to understand what I’m trying to say. I’ll try to explain what I mean more clearly as long as you try to understand what I’m trying to say.

    I’ve been travelling all day, so apologies if that seems a little cranky.

  128. Kevin,

    Perhaps we are talking about different things. I posed the question the way I did to see your response and hopefully understand the essence of your statement.

    Maybe this can be discussed further on this blog when you are back?

    I will read the papers you suggested. What is the Karnauskas et al paper?

  129. BBD says:

    WMC

    Its all I did.

    Thanks, I feel better for knowing that.

    🙂

  130. SoD,

    Your response in the comment above doesn’t seem to address any evidence for this at all.

    Maybe you are addressing points related to whether there is evidence that orbital forcing starts and ends ice ages. I will respond on that in a separate comment.

    Yes, all I’m doing is suggesting that there is evidence that orbital forcing starts and ends ice ages. I appreciate that there is no definitive or agreed mechanism, but there is still evidence to suggest that variations in our orbit drive/trigger variations in our climate. So, that doesn’t prove that the 100000 years associated with Milankovitch cycles are forced, but just because we don’t have a definitive mechanism doesn’t suddenly mean that it is as likely to be unforced as forced.

    I think maybe you have a strange idea about the causes, implications and results of variation in complex systems. Lorenz has good examples in his papers, some derived from simplified forms of basic atmospheric air movements.

    I’ll try not to express an opinion about your ideas if you try not to do the same about mine. All I was trying to suggest was – for example – if we’re sensitive to internal variability (unforced) how do we explain that our climate appears to have been quite stable throughout the Holocene. In other words, how can our climate appear stable for timescales that are much longer than most other climate timescales (day, year) and then suddenly respond to some unforced variability and switch into a new climate state? William, in a comment above, suggests that the D-O oscillations you see during the last ice age were unforced variability, so maybe there are unforced processes that can drive climate variations on reasonably long timescales, but I was just asking the question as to whether or not there might be reasons why one might argue against this in general.

  131. Andthentheresphysics,

    All I was trying to suggest was – for example – if we’re sensitive to internal variability (unforced) how do we explain that our climate appears to have been quite stable throughout the Holocene. In other words, how can our climate appear stable for timescales that are much longer than most other climate timescales (day, year) and then suddenly respond to some unforced variability and switch into a new climate state? William, in a comment above, suggests that the D-O oscillations you see during the last ice age were unforced variability, so maybe there are unforced processes that can drive climate variations on reasonably long timescales, but I was just asking the question as to whether or not there might be reasons why one might argue against this in general.

    Our climate hasn’t been ‘quite stable’ through the Holocene. On many different timescales over that period it demonstrates considerable change.

    And the concept is not ‘suddenly respond to unforced variability’, it is a concept of multiple systems interacting nonlinearly to produce other modes at very different timescales. If the systems are nonlinear then it is expected that interactions are not additive.

    If by ‘quite stable’, you are asking “why hasn’t the climate gone into a new ice age or into a much hotter mode of operation”, this is not an argument against unforced variability, because unforced variability does not imply that.

    In due course, perhaps as part of the continuation of the Ghost of Climates Past series, I will show some simple models to attempt to illustrate these points.

    ..maybe there are unforced processes that can drive climate variations on reasonably long timescales, but I was just asking the question as to whether or not there might be reasons why one might argue against this in general.

    I think that’s my question. In the comments in this article (and also the last one in which I participated) it seems everyone else commenting disagrees with me. I’m asking why. The reason I try to press the point, at the significant risk of wearing out my welcome, is without understanding exactly why someone thinks what they do, how do you come to appreciate another point of view?

  132. BBD says:

    Are we back to a mix of orbital (mainly obliquity) forcing and stochastics (eg instability of large NH ice sheets beyond a threshold involving N latitude extent and ice sheet thickness, etc?

    I could live with this.

    😉

  133. Andthentheresphysics,

    ..all I’m doing is suggesting that there is evidence that orbital forcing starts and ends ice ages. I appreciate that there is no definitive or agreed mechanism, but there is still evidence to suggest that variations in our orbit drive/trigger variations in our climate. So, that doesn’t prove that the 100000 years associated with Milankovitch cycles are forced, but just because we don’t have a definitive mechanism doesn’t suddenly mean that it is as likely to be unforced as forced.

    There are two separate points I think.
    There is plenty of evidence that variations in the earth’s orbit drive variations in our climate. Ice sheet size and temperature move around with the 40k and 20k cycles.
    The evidence that ice ages start and end due to changes in the earth’s orbit is a different story. (Well, specifically whether they end due to orbital changes).

    ..just because we don’t have a definitive mechanism doesn’t suddenly mean that it is as likely to be unforced as forced.

    As a general point that could be correct. And perhaps that’s a question for people who’ve read more than just Thomas Kuhn, Structure of Scientific Revolutions on the history of science. I.e., I have nothing.

    In the case of this topic, there are a limited range of drivers for how orbital changes can affect climate at this particular frequency, and 100s, if not 1000s, of the brightest scientific minds have tried to find the answer. I have about 200 papers on the ice ages that I’ve looked at, at least 100 in some detail, many of them more than once, comparing them with other papers, and all from reputable journals – Nature, Science, Quaternary Science Reviews, Climate Dynamics..

    My current working conclusion from these papers is that the cause of specifically ice age terminations due to orbital variations would be completely obvious by now – if there was such a cause. I’ve plotted the insolation changes in detail in Part Eleven and in Pop Quiz, and it’s clear in those plots that the last ice age termination is not related to insolation.

    This isn’t a radical opinion. (In a later article on my blog I will provide the statements of 20 or 30 papers on this subject as a demonstration of this claim).

    Onto ‘unforced mechanisms’. I don’t think there is such a clear distinction between this and ‘forced’. if we consider the last termination the SH leads the NH, with CO2 closely in step with the large temperature changes in the SH. Parrenin et al 2013 suggest that CO2 doesn’t lag the Antarctic temperature changes at all and therefore it implies that large increase in CO2 might be a forcing rather than just a strong feedback. What does ‘unforced variation’ mean in this context? The temperature rise before the CO2 rise is – at a maximum – quite small. (If there is no lag then it is zero). What, in practical terms, is the difference between a forced variation and an unforced variation?

    The actual questions of interest become completely different ones.

  134. Joshua says:

    “The reason I try to press the point, at the significant risk of wearing out my welcome, is without understanding exactly why someone thinks what they do, how do you come to appreciate another point of view?”

    Based on observing the discourse at this site, I doubt that simply trying to understand the views of others would wear out your welcome. In fact, pressing the goal of understanding other is the expressed philosophy of the blog.

    Wearing out your welcome would be much more likely if behind an ostensible intent of understanding the views of others, lies an intent to prove them wrong (or prove yourself right). It can be a fine line sometimes, but what’s important about that, IMO, it that it is fallacious to try to prove someone else wrong if you don’t understand what they’ve said/can’t accurately represent their views. So in line with that, I’d say press on.

  135. Bobby says:

    Yes, please press on. This has been a really interesting discussion, and I’ve learned a lot already.

  136. Rob Painting says:

    John Mashey – “We didn’t have ice age oscillations until CO2 fell far enough to allow it over last 40My or so. The continents haven’t moved enough to make much difference, orbital cycles are the same … but over the last few My, we’ve gotten into increasingly-sensitive state, t least at the multi-millennia time-scales.

    Mash – don’t forget the Eocene hyperthermals. Some researchers suggest that those too were orbitally-paced. If correct, the ocean circulation – for example moving to an El Nino-like state at times of high obliquity & precession – may also be a trigger.

  137. John Mashey says:

    Rob, yes, but I was in hurry, about to leave for concert and I should have written what I was thinking: : ‘We didn’t have ice age oscillations in the last 40My or so until CO2 fell far enough.’ i.e., after those Eocene events, but I haven’t followed those discussion, so don’t have an opinion.

  138. SoD,

    Our climate hasn’t been ‘quite stable’ through the Holocene. On many different timescales over that period it demonstrates considerable change.

    Yes, I realise it shows change, but – as far as I’m aware – compared to both what we’re undergoing now (in terms of the rate of change) and compared to the changes that have been associated with Milankovitch cycles, it appears quite stable. Maybe I’m wrong, but that’s been impression.

    And the concept is not ‘suddenly respond to unforced variability’, it is a concept of multiple systems interacting nonlinearly to produce other modes at very different timescales. If the systems are nonlinear then it is expected that interactions are not additive.

    Okay, so maybe I’m expressed myself poorly. I was meaning how it appears to respond, rather than the underlying processes that lead to the change. So, this may be one of the fundamental points that I was trying to get at in the post and you’ve probably expressed more fundamentally than I had tried to. So, yes, this appears plausible. However, do we really have any evidence that this happens. I agree that in a non-linear system interactions are not additive and that this could well imply that some climate changes are unforced, but do we have evidence that the system actually behaves like this? This is a very interesting issue and I certainly don’t know the answer or have a strong view either way.

    If by ‘quite stable’, you are asking “why hasn’t the climate gone into a new ice age or into a much hotter mode of operation”, this is not an argument against unforced variability, because unforced variability does not imply that.

    Well, yes, that is kind of what I was asking. I wasn’t using it as an argument against unforced variability, I was essentially asking how unforced variability could appear to operate on timescales that appear to be much longer than all timescales associated with our climate. Orbits the Sun once a year. Rotates on its axis once a day. ENSO cycles appear to produce variability that has decade-like timescales. So, I was really just asking if there was some way to explain how our climate can appear to be reasonably stable for 10000 years or more (and I mean stable in the not switching to another ice age sense, than no variability at all) and still be sufficiently sensitive to unforced variability so that it could actually undergo a major change.

  139. SoD,

    The reason I try to press the point, at the significant risk of wearing out my welcome, is without understanding exactly why someone thinks what they do, how do you come to appreciate another point of view?

    Apologies if my tone last night was more curt than normal. I had been travelling for most of the day. A good night’s sleep and I feel much better today 🙂

  140. SoD,

    What, in practical terms, is the difference between a forced variation and an unforced variation?

    In the context of this post, I was using forced to imply a definitive external change (which would include, for example, outgassing of CO2) and using unforced to refer to some kind of internal change to the climate system. However, if I understand what you’re saying, one could envisage a scenario in which some internal variability in the climate could lead to outgassing of CO2 which then becomes a forcing. Hence, the distinction may not be always obvious.

  141. andthentheresphysics on January 13, 2014 at 7:31 am:

    in response to my question:

    What, in practical terms, is the difference between a forced variation and an unforced variation?

    Suggested:

    In the context of this post, I was using forced to imply a definitive external change (which would include, for example, outgassing of CO2) and using unforced to refer to some kind of internal change to the climate system. However, if I understand what you’re saying, one could envisage a scenario in which some internal variability in the climate could lead to outgassing of CO2 which then becomes a forcing. Hence, the distinction may not be always obvious.

    This is correct. This is what I was trying to get at.

    Laboring the point, perhaps for other readers..

    Take scenario A – an “external forcing” leads to a 0.5’C temperature rise which leads to a significant CO2 increase (I pick this because CO2 increases around the last interglacial are difficult to explain and probably a key determinant of the transition) which leads to the climate moving from the LGM to the holocene, with the associated 120m level increase in sea level.

    Now take scenario B – an “internal variation” (perhaps due to some combination of different internal climate modes) leads to a 0.5’C temperature rise which leads to a significant CO2 increase..

    The climate doesn’t distinguish between the original causes. We, of course, want to. (Well, I do, and I’m sure that everyone visiting this blog does).

    So if at that time on that day/month/year/century a 0.5’C rise would cause “an event”, and the event is much more significant than the original cause, then should we say the event is forced, unforced, or is the distinction just semantics?

    Really the key point is to find the reason why the small change caused the much larger change, why at that time it did, and why a similar small change didn’t cause the much larger change at earlier times. And the reason behind the 0.5’C temperature rise, being extremely common in the scheme of things, is much less significant in the story.

    If climate was linear it would be quite a different story.

  142. andthentheresphysics on January 13, 2014 at 7:25 am

    Apologies if my tone last night was more curt than normal. I had been travelling for most of the day. A good night’s sleep and I feel much better today

    No problem, and your comment is much appreciated.

    Likewise, apologies if any of my comments have caused offense. I often appear to have an adversarial style, it’s not aimed at being adversarial, it’s not the intent, it aims to highlight the real point, the real difference – if in fact there is a difference – “state your real claim” – and often that is misunderstood as trying to pin an idea on someone that that person would not claim.

    Sometimes similar but different language is actually the exact same idea with two people trying to emphasize subtle differences, and sometimes it is two people with a completely different idea hidden by common language. How to separate these two situations without pressing the point? This I understand. How to separate these two situations without also insulting/annoying the other party? On this, I probably do a bad job.

  143. SoD,

    Take scenario A – an “external forcing” leads to a 0.5′C temperature rise which leads to a significant CO2 increase (I pick this because CO2 increases around the last interglacial are difficult to explain and probably a key determinant of the transition) which leads to the climate moving from the LGM to the holocene, with the associated 120m level increase in sea level.

    Now take scenario B – an “internal variation” (perhaps due to some combination of different internal climate modes) leads to a 0.5′C temperature rise which leads to a significant CO2 increase..

    I agree, but with one caveat. It seems that ENSO events can lead to significant surface warming. The 1998 event seems to have produced an increase of around 0.2oC. Hence, it doesn’t seem unreasonable that some internal event could produce an increase of around 0.5oC. However, the land and atmosphere have a relatively small heat content. My quick calculation suggests that a 0.5oC rise in temperature would require around 5 x 1021J. If, however, this temperature rise results in the surface temperature being 0.5oC above equilibrium, then this excess energy would be lost within about a month.

    So, there is still possibly a difference between an external forcing that increases the equilibrium temperature by 0.5oC and some internal event that increases the surface temperature by 0.5oC. (As an aside – this is one of my issues with the kind of ideas that Bob Tisdale keeps presenting.) So, yes, if some internal event can produce a rise in surface temperature that can be sustained for long enough that it can lead to other feedbacks, then I agree. I’m still – however – unsure as to whether or not there is actually any evidence for that, but I can see that it may well be possible.

  144. An aside.

    I did this quickly, so maybe did this wrong. A quick calculation suggests that if we want to change the equilibrium temperature by about 1oC by changing the albedo, then the albedo needs to change from around 0.3 to 0.29 (so about 3%). About 10% of the Earth’s surface is covered by ice, so we would need to remove around 1.5 x 1012m2 of ice coverage. If we assume that it’s all sea ice and 3m thick, that’s 5 x 1015 kg. The latent heat of ice is 334000 J kg-1, so that would require 1.5 x 1021J. This doesn’t sound implausible, but is comparable to the amount of energy associated with the surface warming for a major ENSO event (1998 for example).

  145. andthentheresphysics on January 13, 2014 at 7:23 am
    SoD,

    Yes, I realise it shows change, but – as far as I’m aware – compared to both what we’re undergoing now (in terms of the rate of change) and compared to the changes that have been associated with Milankovitch cycles, it appears quite stable. Maybe I’m wrong, but that’s been impression.

    Here’s a graphic from 800,000 Years of Abrupt Climate Variability, Barker, Knorr, Edwards, Parrenin, Putnam, Skinner, Wolff & Ziegler, Science (2011):

    [perhaps this can be created within the page so everyone can view it?]

    ..So, yes, this appears plausible. However, do we really have any evidence that this happens. I agree that in a non-linear system interactions are not additive and that this could well imply that some climate changes are unforced, but do we have evidence that the system actually behaves like this? This is a very interesting issue and I certainly don’t know the answer or have a strong view either way.

    From the point of view of the equations that determine climate as far as I know there is no evidence either way.

    The essence of complex non-linear systems is that identifying the parameter set that results in chaotic intransitive behavior vs chaotic transitive behavior vs deterministic behavior is a very painstaking exercise. With a typical GCM equation set this would be a massive undertaking.

    To cite EN Lorenz 1968:

    Mathematical theory now tells us, however, that both transitive and intransitive systems exist. Moreover, no simple way has been discovered for examining an arbitrary system of equations and determining whether it is transitive or intransitive

    ..I was essentially asking how unforced variability could appear to operate on timescales that appear to be much longer than all timescales associated with our climate. Orbits the Sun once a year. Rotates on its axis once a day. ENSO cycles appear to produce variability that has decade-like timescales. So, I was really just asking if there was some way to explain how our climate can appear to be reasonably stable for 10000 years or more (and I mean stable in the not switching to another ice age sense, than no variability at all) and still be sufficiently sensitive to unforced variability so that it could actually undergo a major change.

    This is easy to demonstrate with some “simple” nonlinear equations with a periodic forcing. There are some interesting explanations in Irregularity: A fundamental property of the atmosphere, EN Lorenz (1984). A good textbook is Nonlinear dynamics and Chaos Strogatz (1994). It is an essence of nonlinear systems that we find irregular periods longer than the periods of the individual systems that comprise the overall system.

    In essence, complex behaviors arise from nonlinear equations with irregular time periods much longer than any forcings.

  146. The attempted graphic from 800,000 Years of Abrupt Climate Variability, Barker, Knorr, Edwards, Parrenin, Putnam, Skinner, Wolff & Ziegler, Science (2011) – is here

  147. Rachel says:

    John Mashey,

    5) Of course, the coolest description of Milankovitch cycles has to be that of Richard Alley.

    I hadn’t seen that before. It’s excellent.

  148. SoD,

    In essence, complex behaviors arise from nonlinear equations with irregular time periods much longer than any forcings.

    I had not appreciated that, but I can see how that make sense. A question would still be whether or not any of these longer period internal variations are sufficient – by themselves – to trigger significant changes to our climate. I have the impression that we don’t really know if this is possible or not, but can’t rule it out.

    Having thought about this a little more, there is also something that may be related to a point Kevin was making about the influence of external forcings on internal variability. It may well be that one could have a change to an external forcing that is insufficient to trigger a major change to our climate. However, if there is internal variability that operates on a shorter timescale, it then becomes possible that the combination of this change in external forcing plus some internal variability could act to trigger some climatic changes. That wasn’t quite what Kevin was suggesting and maybe I haven’t expressed it all that clearly, but it seems possible – if not likely – that everything is much more complicated than simply being forced versus unforced.

  149. John Mashey says:

    Maybe it would help to separate questions, as I think people have been arguing different ones.
    1) The exact mechanism for inception of deglaciation is still being argued about. It certainly was in 2009 when Archer published the Long Thaw, (p.77). There are plenty of hypotheses.
    I think this is the part that SoD is focused on.

    2) On the other hand, whenever a deglaciation gets started, the mechanism that moves the planet from glacial to interglacial (when I wrote “drives” that’s what I meant, not inception, sorry for any ambiguity) is quite clear is the combination of all 3 of:
    a) Strong summer insolation in NH, external forcing, which goes up ~440 to ~540 W/m^2 during the periods of high swings, in ~half a precesssion cycle. A difference of 100W/m^2 matters.
    b) Ice-albedo feedback
    c) GHG feedback
    (ATTP as addressing some of this).

    But, as clearly seen in Ruddiman etal(2011) Fig 3, although 5 of the deglaciations got to 90% of the 18O change in ~10,000 years (meaning, pretty much done in 13,000 years, half the precession), Stages 11 and 15 took longer, and 11 is generally weird. This kind of figure makes it much easier to see the similarities and differences among the various deglaciations, compared to eyeballing them on the usual chronological charts.There is clearly some random variability not understood, and clearly differing initial conditions in terms of ice extent and CO2 level that matter, and maybe ocean circulation. It matters how far South the NH ice gets, and the rough trend over last few My has been to increase maximum ice volume, but with jiggles.

    This seems like a a system with 2 major states, with strong forcings+feedbacks to change states, but very sensitive conditions for initiating state changes, although sensitive in different ways:
    a) Reglaciations start from relatively similar conditions, and start when the insolation falls below the rigger value set by the CO2 level.
    b) Glacial inceptions start from more varied conditions, depending at very least on ice coverage.

    3) ATTP: Ruddiman provides a very strong set of evidence for a reason for the relative stability of the Holocene: we kept CO2 within a relatively narrow band, 260-280ppm (fig 2). AS noted before, we certainly contributed part of the LIA via the 50Mperson die-off in the Americas and the resulting ~9ppm CO2 drop,

    Whether or not we would have already started a reglaciation may still be arguable, although I think evidence for that is accumulating, too. Of course, it takes longer to accumulate ice sheets than to melt them.

    Of course, going forward, this is a bit mot, given that that it will be a long time before anotther reglacation, despite those who are sure there is another just around the corner.

  150. Pingback: Another Week of Climate Disruption News, January 12, 2014 – A Few Things Ill Considered

  151. In case anyone visits this page long afterwards..

    I cover Termination II in Ghosts of Climates Past – Thirteen – Terminator II with many papers to demonstrate the radiometric dating of TII at about 140 kyrs BP, a time of low and falling NH solar insolation.

    TI and TII plotted on 65’N summer insolation in this image.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s