« Previous | Main | Next »

Unpredictable weather: why the climate is not a model citizen

Post categories:

Richard Cable | 15:13 UK time, Tuesday, 9 June 2009

One of the awkward things about global warming is that there are no absolutes. No one can say definitively what the climate will do next. Anyone who thinks they can will probably end up looking like one of those TV scientists from the 1950s who said we'd all be holidaying in space and flying around in hover cars by now.

But why is it so very difficult to state anything with complete confidence about the behaviour of our climate? Even the Intergovernmental Panel on Climate Change (IPCC), which periodically publishes 'the largest and most detailed summary of the climate change situation ever undertaken' is only prepared to say that human beings are 'very likely' to be the source of the problem. They are hedging for a reason.

Admittedly, it's a little firmer about the temperature itself, stating: 'Warming of the climate system is unequivocal.' But then that's a bit like saying that, today, it is hot. It doesn't tell you very much about tomorrow.

The weather is chaotic. Chaotic systems are infinitely complex and inherently unpredictable, (although not, as some suppose, random). The climate is simply 'big, long weather' - the atmospheric conditions of a region charted over a period of time - and is therefore also infinitely complex and inherently unpredictable.

The IPCC attempts to predict this unpredictability by using climate models - fiendishly complex computer simulations of the Earth's climate that explore 'emissions scenarios'. Each of these scenarios looks at different levels of emissions, and from them the IPCC draws conclusions about where we might be heading.

The models are not without their critics. In order to accurately model a chaotic system, you arguably have to be able to describe the starting conditions of the system and understand pretty perfectly how each of the elements in that system will act upon every other element in that system.

But we don't yet fully understand key issues, such as to what degree carbon dioxide warms the atmosphere or how clouds form and disperse, and can't yet accurately predict even complex human systems that themselves act on climate, like population growth and economic development.

With this in mind it's hard to see how a computer model with so much potential error in its starting conditions can accurately extrapolate what the climate will be doing in 100 years. That's not to say they never will, although anyone who has ever relied on a British weather forecast for the next 24 hours will instinctively take any predictions with a pinch of salt.

Comments

  • 1. At 9:59pm on 09 Jun 2009, GreenGoblinKing wrote:

    Sure, there will never be a way to produce a perfect model to predict weather. Too many factors. But your article negates the speed at which technology advances and how much more processing power these models will have in 10 years, 20 years, 50 years...just think of where supercomputers where in 1999...

    Complain about this comment

  • 2. At 10:33pm on 09 Jun 2009, Bishop Hill wrote:

    Heresy!!

    Complain about this comment

  • 3. At 02:04am on 10 Jun 2009, Jack Hughes wrote:

    The computer models are hopeless.

    They cannot predict the weather next week, next month, or next year.

    It's crazy to try and forecast the weather 10, 20, 50 years into the future before you can predict tomorrow's and next week's. Crazy.

    Complain about this comment

  • 4. At 04:42am on 10 Jun 2009, Jack Hughes wrote:

    Here is an example of just how bad the computer models are:

    September 2008: Milder and drier winter predicted

    "This winter will be milder than average and drier than last year in the UK, Met Office forecasters have predicted.

    ... the winter months will see relatively high temperatures."

    But then wait until February 2009...
    Heavy snow hits much of Britain

    "Heavy snow has fallen across large parts of the UK, disrupting travel and closing thousands of schools.

    South-east England has the worst snow it has seen for 18 years, causing all London buses to be pulled from service and the closure of Heathrow's runways.

    The Met Office has issued an extreme weather warning for England, Wales and parts of eastern Scotland.

    By late Monday, the South East could be under a foot (30cm) of snow and the North East under 20 inches (50cm).

    Thousands of school children across England and Wales woke up to the news that their school was closed for the day.

    Thousands more are closed across Wales, London, north-east England, East and West Sussex, Kent, Norfolk Cambridgeshire, Lincolnshire and Suffolk, where snow-ploughs were out on the county's roads for the first time in more than 15 years.

    Hospitals in London have put out an emergency call to staff to come into work because of a big increase in 999 calls, while many non-urgent operations have been cancelled and ambulance response times have slowed.


    Southampton Airport closed
    BAA
    London City and Luton airports closed
    London City Airport
    Luton Airport
    London buses withdrawn
    Transport for London
    Train services cancelled
    National Rail Enquiries
    Hundreds of schools closed"

    Were they using the wrong type of models ?



    Complain about this comment

  • 5. At 07:16am on 10 Jun 2009, MangoChutney wrote:

    jack

    #3

    don't mix up weather and climate, especially as i know you know the difference

    the big problems with the computer models, as the IPCC themselves admit, is the models ability to deal with clouds and how climate sensitivity is treated.

    clouds are very difficult to deal with, so the models essentially lump them together and climate sensiticity is still an unknown variable

    the models themselves could actually work, if we knew everything we need to know to programme them properly - unfortunately we don't

    Complain about this comment

  • 6. At 09:00am on 10 Jun 2009, Jack Hughes wrote:

    Here's the money quote:

    "the models themselves could actually work ... if only we could get them to work"

    Complain about this comment

  • 7. At 09:23am on 10 Jun 2009, fjpickett wrote:

    Deep joy! Some sanity on the subject from the BBC at last!

    So much for the 'science is settled', then.

    When will this new realism permeate the rest of the Beeb? When you stop sending Roger Harrabin jetting to exotic locations to tell us that the oceans are turning to acid (it's still alkaline), sea levels rising (they aren't) or the arctic melting faster than ever (it isn't), I suppose...

    Complain about this comment

  • 8. At 10:30am on 10 Jun 2009, Jack Hughes wrote:

    Predicting a mild winter looks like a climate prediction to me.

    If they cannot forecast the climate 5 months ahead of time then 50 years is a joke.

    Complain about this comment

  • 9. At 11:29am on 10 Jun 2009, AlantheBrit wrote:

    Climate is merely weather over a prolonged timescale. The article even says so!

    Also, no matter how powerful a computer you possess, even one as powerful as "Deep Thought" from HhgttG, if the basic data you start with is vague or wrong, &/or the assumptions about inputs you make are incorrect, the output will be no better than one provided by the most primative computer you started with previuosly, but with more decimal points! A computer model can be right or wrong, yet still irrelevant! GIGO.

    Complain about this comment

  • 10. At 11:50am on 10 Jun 2009, Gates wrote:

    Well this is the Real Big Question onclimate change. An accurate measure of how our actions will effect our climate and in what way. Unfortunately I think will be finding out first hand before we get a computer to figure it out.

    Complain about this comment

  • 11. At 1:02pm on 10 Jun 2009, FourDJones wrote:

    Climate isn't "big, long weather" nor "weather charted over a period of time". Climate is weather averaged over a period of time as in "Britain has a climate that is mild in winter and warm and wet in summer". These are statements about average weather and say nothing about a particular day. If I throw a dice I have a poor chance of predicting the outcome; if I throw it 100 times my chances of predicting the 100th outcome are no better but I can predict the average outcome with some precision.
    Whether climate models can predict climate in 100 years I have no idea though I seriously doubt it but they don't claim to predict weather but climate as in "Britain will experience hotter, drier Summers and much wetter Winters".
    If you're still not convinced ask yourself, "Was the climate different in Britain during the early 19th century?" then ask "Was the weather different then?" I think you'll find we have the same sorts of weather but on average it's warmer.
    Climate isn't weather any more than the total is the same as the individual items.


    Complain about this comment

  • 12. At 1:25pm on 10 Jun 2009, Neil Hyde wrote:

    This can't be a BBC blog I'm reading, but I wont believe it until I see it on the front page , and every news bulletin on both radio and TV .

    I heard that the science department was cutting numbers , does Richard Cable realise he has just volounteered ? Or has he been told already and come out with the truth ?

    Complain about this comment

  • 13. At 1:53pm on 10 Jun 2009, SheffTim wrote:

    Climate models function in a very different way to those used in weather forecasting where real world data is fed into a computer, which then projects the likely trends for a few days ahead.

    For example the arrival of a cold-front/depression/high etc can be forecast by using its current speed, direction and distance to estimate its time and place of arrival.

    I find most weather forecasts reasonably accurate (it clearly helps to make the effort to get a full forecast each day); to put it another way I rarely get caught out by bad weather, and I'm a outdoor activity enthusiast. Clearly there can be local differences (one hill can get rained on whilst another close-by remains bathed in sunlight) but for the region the day's forecast is generally accurate overall.
    Most people I know that complain I find dont actually know what the forecast for that day is. I use Metcheck.com, various meteorology blogs as well as the forecasts on TV and radio.

    Climate models are systems derived from the basic laws of physics, fluid motion, and chemistry etc. In other words theyre based on well established physical laws that define the behaviour of the weather and climate.

    Can computers model really complex systems? Well, theyre used in modelling nuclear explosions, hence the lack of underground testing by the superpowers in recent years; theyre also increasingly used in developing new cars, engines, aircraft, missiles and composite materials (thus saving the cost of building prototypes) and skyscrapers and increasingly in medicine etc so they cant be that bad at it

    For all the imperfections (and there is year-on-year improvement) these models are still the best means at our disposal of forecasting future climate. For all the diatribes aimed at these models (usually by the most clueless that simply parrot what they've read elsewhere) no-one has a better alternative to offer.

    Complain about this comment

  • 14. At 2:30pm on 10 Jun 2009, jayfurneaux wrote:

    "This can't be a BBC blog I'm reading, but I wont believe it until I see it on the front page , and every news bulletin on both radio and TV ." ToughNeilHyde.

    Naaah, those writing these blogs realise, as all blog writers do, that they have to put a juicy topic up to attract comments. That's how blog success is judged; otherwise they might as well Twitter. Like HYS it's just another form of opium for the masses.

    Complain about this comment

  • 15. At 2:33pm on 10 Jun 2009, CarryOnRegardless wrote:

    RE: 13
    Because a computer can model nuclear explosions does not mean that by extrapolation they can therefore model everything and anything. Blog #5 hit the nail on the head when he wrote "the models themselves could actually work, if we knew everything we need to know to programme them properly - unfortunately we don't."
    At risk of presenting myself as another one of those "clueless that simply parrot what they've read", I feel compelled to add that the biggest problem is the lack of understanding of how marginal warming effects clouds. It is ASSUMED by the models that warming leads to less clouds that leads to more warming and so on in an ever increasing cycle - the so called positive feedback. This assumption is in every one of the 23 models that the IPCC use in the storylines. This is despite the fact that the IPCC acknowledge that cloud behaviour is not understood. Does this make the models useful? If they model on faulty assumptions then no. For example, the only reason that the decreasing cloud model is assumed is because observations have shown that during the modern warming, cloud cover has somewhat diminished and the inference from this is that warming is the cause of the diminished cloud. However, this in logically fallacious since it is inferring cause from a correlation. It is equally plausible that something caused diminishing cloud cover and this has resulted in the warming. If this is in fact true, as it may well be, then the output from all 23 models will be erroneous. The storylines that result in predictions of 2.5to 6 degrees C of warming, may turn out to be less than 2 degrees. Therefore, the cloud sensitivity issue is of paramount importance. Garbage In, Garbage Out.

    Complain about this comment

  • 16. At 3:11pm on 10 Jun 2009, agwbsdotcom wrote:

    Well, well, well.

    I could take book on how long this page will last. Or its' author's continued tenure at "Auntie Beeb" ;-)

    When the Met office have, at most, 55%, John Nelson had 90% regularly and current solar/lunar forecasters reach 85% success rates,

    When the IPCC themselves admit "In climate research and modeling, we should recognize that we are dealing with a coupled non-linear chaotic system, and therefore that the long-term prediction of future climate states is not possible." (IPCC TAR, p.774.)

    When CO2 is v i t a l for nearly all life on Earth, is at extremely low levels historically and any "extra" we "create" is soon wolfed down by flora and biota that have blossomed(pun intended) as a result,

    Will the politicians, activists and "believers" admit that their using the gas of life to drive policy and garner tax and power is fraudulent?

    The question remains, as always:

    Where is the evidence that CO2 has ever driven the global temperature?

    Models are not evidence.

    Proof of warming is not evidence that CO2 did it.

    Ancillary questions, such as; Why are we not concentrated on deforestation? Why are we not concentrated on over fishing? Why does over 1/3 of the world have no reliable source of clean drinking water? Why are we allowing real, measurable pollution at the same time as we are pretending that a component of the carbon cycle is, somehow, a pollutant?

    Why are the majority of scientists, academies and institutions studying warming? Especially if the science is "settled"? The billions that have, so far, failed to provide evidence that "CO2 dunnit guvnor" have been diverted away from real concerns.

    Environmentalists, and others, are guilty of distraction and all the while the "alternatives" and "renewables" bubble - make no mistake, that is what it is, a bubble funded and pushed forward by the exact same people who brought you the "biggest depression since the nineteen-thirties" - will soon be perceived as unworkable exactly because they have forced those technologies out of their needed decades of continued research into premature production. Joe Public is already asking why wind power is so costly when it doesn't work. The rest of the house of cards is soon to follow.

    This is not an accident.

    Complain about this comment

  • 17. At 3:48pm on 10 Jun 2009, CarryOnRegardless wrote:

    agwbsdotcom:
    Wind power - it's too cheap to meter init? Wind is free after all (lol).

    Complain about this comment

  • 18. At 4:54pm on 10 Jun 2009, Maurizio Morabito wrote:

    The models cannot work. Even if we knew the physics perfectly, and we had written great programming codes about it, still there are so many unpredictable things that can happen on the climatic/multi-decadal timescale (volcanoes, Sun, changes in human activities), it makes no sense to predict anything.

    Complain about this comment

  • 19. At 5:03pm on 10 Jun 2009, SamuelPickwick wrote:

    I continue to be impressed by this blog. At last a BBC journalist who seems to know something about science, chaos, and the difficulties of computer modelling, and the fact that we don't know what the key driving factors are. When will we get to hear Richard Cable on the Today Programme or Newsnight?

    The climate is indeed behaving unpredictably, with no warming this decade, contrary to the IPCC projections.

    The quality of the blog comments is also very good.

    I hope everyone is enjoying the Met Office's 'Barbecue summer'!

    Complain about this comment

  • 20. At 5:20pm on 10 Jun 2009, Lucy_Skywalker wrote:

    Crikey, does this mean the Beeb is (a) actually waking up to the real science; or (b) up for reform and return of its lamented lost integrity and reputation for good science and listening to all points of view; or (c) sucking up to what commentators have been saying for a while; or (d) playing to the audience while it's obviously an unseasonably cold spell?

    My, my, I wish it was (a) and (b). I wish Beeb would investigate fairly, own up and apologize. Same goes for Royal Society, New Scientist, Nature, NAS, and the Met Office. It would be lovely to have real science back.

    Complain about this comment

  • 21. At 5:40pm on 10 Jun 2009, MangoChutney wrote:

    Lucy!

    (And others)

    How dare you assume the BBC has actually started to realise there are 2 views to this story. The BBC have clearly been infiltrated by those nasty deniers and, as soon as they are routed out and disposed of, normal service will be resumed.

    Complain about this comment

  • 22. At 5:53pm on 10 Jun 2009, xyz999 wrote:

    The blog says:
    "With this in mind it's hard to see how a computer model with so much potential error in its starting conditions can accurately extrapolate what the climate will be doing in 100 years."

    However:
    Current climate projections are NOT like weather forecasts. In weather forecasts the starting conditions matter a great deal. The climate (average weather) in 2100 does not care about what the weather is doing today. So, the weather may not be predictable more than a couple of weeks in advance, but that does NOT mean that the climate is unpredictable as the article implies, as the climate response is dominated by the response to much larger factors such as greenhouse gases, volcanoes, solar changes etc.

    There are undoubtedly large uncertainties in how the climate will respond (as the IPCC reports make very clear), but what the weather is doing today is not one of those uncertainties.

    Complain about this comment

  • 23. At 8:48pm on 10 Jun 2009, RiHo08 wrote:

    We all should be concerned with predictions made by mathematical models. The spread of the H1N1 virus within the USA as predicted by 2 super computers, one at Indiana University and the other at Northwestern University were in complete agreement. The only problem was, both were off target by a log of 3. The current financial meltdown is in no small way attributable to the failure of mathematical models to predict the behavior of financial derivatives. The influence of atmospheric CO2, its origins and dispersal, upon climate science is presently modeled using computer mathematical models. These predictive models require a great leap of faith to accept the premise that small changes in this gas's concentration will catastrophically alter the world. More likely, there are many more complex influences upon present and future climate yet to be identified and modeled. My prediction is that atmospheric CO2 levels will be found as following climate events in the bigger climate picture, likely led by of all things, our sun.

    Complain about this comment

  • 24. At 9:18pm on 10 Jun 2009, SheffTim wrote:

    #16 Where is the evidence that CO2 has ever driven the global temperature?

    Paleocene-Eocene Thermal Maximum is one.
    http://en.wikipedia.org/wiki/Paleocene-Eocene_Thermal_Maximum

    Complain about this comment

  • 25. At 10:52pm on 10 Jun 2009, Maurizio Morabito wrote:

    #22 Let's not be disingenuous. The reason why "climate projections" do not care about "starting conditions" is because they are projections, not predictions...all they can tell you is how the climate would evolve with the change of this or that forcing, everything else left equal.

    In the real world, of course, no forcing changes in isolation. Thus the predictive abilities of "climate projections" are zero. By definition.

    #24: Please expand. Where is the evidence that CO2 was driving temperature and not vice-versa?

    Complain about this comment

  • 26. At 11:58pm on 10 Jun 2009, agwbsdotcom wrote:

    #24 - interesting page there SheffTim. Thank you. I have just spent a while going through it and its links. It lead me to some wonderful places.

    The page doesn't specifically attribute the warmth to a CO2 increase, or, to be fair, vice-versa. The page's conclusions generally rely upon.. wait for it... models. Methane gets some prominence.

    The lowering of oceanic alkalinity due to heightened atmospheric CO2 levels is treated impartially for a wiki page that falls within this (amongst many others) very busy guy's remit-

    http://en.wikipedia.org/w/index.php?title=Special:Contributions&limit=500&target=William+M.+Connolley

    The "editing" of the climate change related pages at wikipedia is systemic, enduring and perpetually one-way-

    http://network.nationalpost.com/np/blogs/fpcomment/archive/2008/04/12/wikipedia-s-zealots-solomon.aspx

    I live in hope that, soon, wikipedia, whose pages appear near the top of nearly every search nowadays, will suffer a seismic shift along with other founts, such as this very one.

    Complain about this comment

  • 27. At 02:41am on 11 Jun 2009, Jack Hughes wrote:

    @AGW-BS: Wow !

    Great debating style from "economart" in the comments. He's got Kim-from-wikipedia bang to rights.

    I still fear that the BBC and much of our ruling elite are still in the grip of a Post-Modernist worldview.
    In this Po-Mo world, objective truth does not exist and all ideas have the same intrinsic merit.
    You don't have to base your views on personal experience, on rational thought, or reason - no you can pick views that make you feel good.
    It's enough just to feel good about saving the planet - you don't need any coherent explanation of how sitting in the dark for an hour is going to stop the polar bears from drowning.

    This is great if you are a counsellor or an aromatherapist - but it starts to unravel if you try to engineer the National Power Grid using Po-Mo principles...

    Complain about this comment

  • 28. At 07:44am on 11 Jun 2009, SheffTim wrote:

    #26. I put a link simply to save people time and trouble; Wikipedia is just one site/page. Use Google / Bing / or whatever search engine you prefer on the same topic:
    Paleocene Eocene Thermal Maximum

    Can it be demonstrated that additional CO2 can affect temperature in the atmosphere?
    A small experiment any high school lab can do.

    Take two airtight glass tanks, each with a thermometer inside, and a means of introducing a gas into one without breaking the airtight seal. (An intake pipe etc.)
    Each box to contain ordinary air. Seal both tanks and place both in sunlight so that both are receiving approx. equal amounts. The thermometer readings of both should match at this point.
    Introduce a small amount of additional CO2 into one tank; its thermometer reading will rise above that of the other box.

    For those that read books and want to become more informed I suggest: (Try your library service.)
    'Ice, Mud and Blood' by Chris Turney.
    A summary of key discoveries by scientists about past climate change going back deep in time and the implications for the present.
    'An Ocean of Air: A Natural History of the Atmosphere' by Gabrielle Walker.
    A history of some of the major discoveries about air, gasses and the atmosphere from Galileo to the present day together with explanation as to their importance for life on Earth.
    'Earth: The Power of the Planet' by Iain Stewart & John Lynch.
    An accessible introduction to earth systems and earth's history.
    'Earth's Climate Past And Future' by William F. Ruddiman.
    An account of known factors that have influenced climate change over earth's history.

    Complain about this comment

  • 29. At 11:57am on 11 Jun 2009, xyz999 wrote:

    Re #25 - the difference between climate predictions and climate projections is a subtle one. The blog seemed to imply that IPCC style climate projections are like weather forecasts and thus suffer from not knowing the initial conditions. This is not correct.

    I disagree that 'projections' have no predictive ability. If they predict some of the evolution of the climate due to changing radiative forcings, then this IS predictive ability. Of course they will not get the year to year fluctuations, but that is not what they are trying to predict.

    Also - what do you mean by 'no forcing changes in isolation'? By the IPCC definition this is not true. Changes in the solar forcing are independent from anthropogenic greenhouse gas emissions. If a change in the solar forcing causes a change in, e.g. water vapour in the atmosphere, then this is a feedback rather than a forcing.

    Complain about this comment

  • 30. At 1:04pm on 11 Jun 2009, tulbobroke wrote:

    Re comment 7, I notice that Arctic sea ice extent dipped below the 2007 level a couple of days ago http://nsidc.org/data/seaice_index/images/daily_images/N_timeseries.png

    Although 2007 had a record low level, this does not necessarily mean that 2009 will be a record low year. In any case, short term fluctuations are not as important as the long term trend.

    Complain about this comment

  • 31. At 1:46pm on 11 Jun 2009, SamuelPickwick wrote:

    xyz999, if you know what is meant by a 'climate projection' and how it differs from a prediction, perhaps you would be kind enough to explain.

    The climate is just as unpredictable as the weather. It also involves complicated nonlinear chaotic interactions, but different ones. It's just that the time-scales are longer. Richard is quite right about this.

    Complain about this comment

  • 32. At 2:52pm on 11 Jun 2009, CarryOnRegardless wrote:

    Sheffim,
    Nobody (I think) is arguing that carbon dioxide does not lead to a warming. The question is to do with the temperature sensitivity. Your experiment would provide the sensitivity to say a CO2 doubling without feedback. Even this figure is not known exactly. For example, Lindzen gives 1 degree C for a doubling, and I believe the high end is around 1.5 degrees. No, the contention is over the way the models impute positive feedbacks.
    The first feedback we can consider is low level cloud cover. All 23 models used by IPCC utilise a positive feedback - increased warming from CO2 leads to reduced cloud cover, which allows increased insolation which leads to more warming which leads to a further reduction of cloud cover which leads to . . . There is no proof that the world works in this way. The assumption (and that is what it is) comes from an observation that cloud cover is less when the climate is warmer. However, correlation does not imply cause and effect. It is equally plausible that the cloud cover decrease has another driver and which then causes the observed warming. If this is the case, then the models would have got cause and effect the wrong way round.
    The other main feedback is increased water vapour which leads to more warming which leads to more water vapour etc. This sounds intuitive, and I suppose that's why it was included in the models. However, there is not any empirical evidence that this is what actually happens. Why do I say that? The Aquasat system, which is designed to detect just this increase in water vapour has so far not detected any, so this assumption is suspect. Further more, GHG warming predicts a decrease in outgoing long wave radiation emitted into space, since the greenhouse gases are effectively trapping heat. This also has not been observed since 2003.
    If the models are programmed with the wrong assumptions, then their predictions will be seriously wrong. Whether climate is like weather or not (it isn't) doesn't matter a jot if the physics is wrong.

    Complain about this comment

  • 33. At 2:58pm on 11 Jun 2009, EdBhoy wrote:

    Re no 30
    I thought I would get a second opinion on the extent of Arctic Sea ice so I checked the Nansen Environmental and Remote Sensing Centre in Bergen. Seems like the Arctic sea ice extent is still above 2007 and 2008 although this hasn't been updated since last Saturday.

    http://arctic-roos.org/observations/satellite-data/sea-ice/ice-area-and-extent-in-arctic



    Complain about this comment

  • 34. At 3:21pm on 11 Jun 2009, xyz999 wrote:

    Re #31: Right... this is the difference...

    Both climate 'projections' and 'predictions' use the same climate models with the same future greenhouse gases, solar output etc. The only difference is their starting point.

    A 'prediction (or forecast)' starts from our best estimate of today's state of the oceans, atmosphere, ice etc (in principle). This is equivalent to an extended weather forecast. So... if (say) the Atlantic Ocean is in a relatively hot or cold phase then the effect of this will be included in the prediction.

    A 'projection' starts from a random state of the model and doesn't worry about whether the Atlantic is currently relatively hot or cold. [In fact the models are generally started in 1850 and run forwards to the present day only using information about historical greenhouse gas levels, volcanoes, solar output etc, and continued into the future with projected changes in GHGs etc. And, the fact that they follow the trajectory of the historical climate since 1850 gives us confidence that they have some chance of projecting the future.]

    Currently, the IPCC only uses PROJECTIONS. The reason being that the difference between the two (i.e. the starting conditions) only matters for a few years at most, as the (chaotic) weather slowly destroys the extra information contained in the current phase of (say) the Atlantic. As the next few years is not what the IPCC is generally interested in, they have used PROJECTIONS. PREDICTIONS are much harder to do, but will be included in the next IPCC assessment as the next few years are becoming more relevant.

    This may be as clear as mud, but it is why Richard is wrong. The future climate (beyond 10 years or so) does not care about today's exact conditions. Likening a climate PROJECTION to an extended weather forecast is inaccurate and is why people here are saying things like 'but they can't even predict the weather 2 weeks ahead'. It doesn't matter for the climate in 50 years time, which IS predictable, as we are trying to predict the average of the (unpredictable) weather.

    Complain about this comment

  • 35. At 5:40pm on 11 Jun 2009, CarryOnRegardless wrote:

    xyz999, Richard is not wrong. Your reasoning appears flawed. You say that the future climate does not care about today's exact conditions, which is true. However, you cannot conclude from this that the climate in 50 years time is predictable. Your logic seems to be: weather is unpredictable, but climate is not weather so therefore it is predictable.

    The evidence you cite for their predictive skill comes solely from the historical runs from 1850. However, these runs, when we look a bit more closely, are no more than curve fitting exercises. A GCM has a large number of tunable parameters, and if you play around with the tuning long enough you can get a pretty good match to some dataset. This is a classic trap for backtesters of trading systems. They start with one sample of data and tune and tune the parameters until they acheive an extreemly optimistic outcome expectancy. If they don't then apply the run to a set of other samples they are in trouble as soon as they start trading with real data. The problem with the GCM's is the same problem as the backtester who uses only one sample. The backtester can correct by adding samples, but unfortunately for the modelers, there is only 1 sample available. The models are failing as soon as they try and project the future. None predicted the current decade of static then falling temperatures. And this is going to happen again and again because they have no predictive skill.

    Roger Pielke sr has shown that the GCM predictions of stored heat in oceans have failed to materialize.

    Complain about this comment

  • 36. At 6:47pm on 11 Jun 2009, xyz999 wrote:

    CarryOnRegardless, you seem to agree that the climate in 100 years time does not depend on today's exact conditions, therefore Richard's statement:

    "With this in mind it's hard to see how a computer model with so much potential error in its starting conditions can accurately extrapolate what the climate will be doing in 100 years."

    IS clearly wrong. The starting conditions DO NOT matter for climate in 100 years time.

    As for the GCMs - I did not mean to imply that because weather is unpredictable therefore climate is predictable. That's ridiculous! The problem with your arguments over curve fitting is that the models are NOT tuned to reproduce the 20th century - they are tuned to get the pre-industrial or present day climate as accurately as possible. Implying that they keep tuning the models and rerunning them for the 20th century until they get the 'right' answer is just not true.

    As for the current decade - what evidence do you have that GCMs don't project the lack of warming in the current decade? The average of all the models, by design, will not predict any of the natural fluctuations in the climate - have you examined each individual projection to see whether any have decades when there is no warming, or even a cooling? Others have, and surprise, surprise, they do project decades when there is a cooling:
    http://www.agu.org/pubs/crossref/2009/2009GL037810.shtml
    There are simply periods where the natural variations in the climate system temporarily offset the anthropogenic signals, and the models DO show this.

    And, the Met Office have a climate PREDICTION system, which DID predict an offset of anthropogenic warming for the current period:
    http://www.sciencemag.org/cgi/content/abstract/317/5839/796

    Complain about this comment

  • 37. At 8:29pm on 11 Jun 2009, CarryOnRegardless wrote:

    xyz999,
    the first link refers to a publication this year. Of course they will now say that models predict periods of cooling since they have to deal with that reality. My point is that I cannot find any such statements made before 2003.
    The second link is actually a prediction, that half of the future years will be warmer than the warmest so far. That is itself is a cop out imo because they have a 50% chance of being right. The Hadley centre I believe have predicted a decade of cooling due to the PDO entering a negative phase, so if I were a betting man, I would say that the Met Office will miss even this prediction.
    Climate models predict a continue storage of heat in the global system. Natural variations cannot alter this, they can only hide it. In other words a cooling of surface temperatures may occur due to a natual variation that sequesters heat energy. This has to be so because the radiative forcing mandates that more energy is currently entering the system than is leaving. Hansen himself has written:
    "The decadal mean planetary energy imbalance, 0.75 W/m2, includes heat storage in the deeper ocean and energy used to melt ice and warm the air and land. 0.85 W/m2 is the imbalance at the end of the decade.
    So we would expect to find this heat turn up in the deep oceans. Hansen wrote this in 2003 I believe. According to Pielke, heat should be accumulated at the following rate of joules:
    2003 ~0.98 * 10** 22 Joules
    2004 ~1.96 * 10** 22 Joules
    2005 ~2.94 * 10** 22 Joules
    2006 ~3.92 * 10** 22 Joules
    2007 ~4.90 * 10** 22 Joules
    2008 ~5.88 * 10** 22 Joules
    But his observed estimation of accumulation of joules over the period 2003 to 2008 is zero. In other words the predicted sequestration of heat as mandated by the models has not materialised. The models are wrong.

    Complain about this comment

  • 38. At 10:01am on 12 Jun 2009, agwbsdotcom wrote:

    #27 Jack_Hughes_NZ - yes. Danger. Very sly, clever foxes at work. I agree with your take on the madness of modern crowds also.
    #numbers - you know who you are ;-)
    Where are the papers prior to 2003, or even 2006 that showed the models capable of predicting the current temperature plateau.
    #28 SheffTim - if there were a link more sympathetic to your view and with a clearer message would you have produced it? Similarly, offering a reading list is unhelpful to inhabitants of the electronic soup. If anything worked it would be broadcast on Internet instantly and loudly enough that no-one would be allowed to miss it. If the climateprogress, realclimate, tamino, desmogblog circular circus could, amidst their delegated coercion, personal attcks, furious hand waving and denigration, highlight just one overwhelming pointer to empirical evidence for their case they most assuredly would.
    Watching their antics over Ferenc Miskolczi should show anyone that enters this fray where investigation, theory and truth stand in this debate. The aforementioned could care less for any of those.
    That the other side cannot produce a definitive counter is what empowers the entire agenda.
    I think it beyond most colleges, let alone schools, to introduce precisely, for example, an extra 50ppmv of CO2 into a captured sample of atmosphere, to Scripps standards, for subsequent testing whilst maintaining equal pressure in both enclosures. Then we have a thermometer problem. The additional CO2/T would be immeasurable without specialist equipment. Your reference to Iain Stewart reminds me of his interpretation, and subsequent "proof", that filling a container with CO2 will snuff out visible light. Laughable propaganda that merely added to the public mistrust of the BBC.
    We all know it is far more complicated than this. Adiabatic lapse rates, just one component of our dynamic envelope that elicits two opposing theories, one claiming more CO2 cools the world and one saying the opposite, will interfere with the hypothesis that CO2 alone can affect climate in any measurable form.
    This is why CO2/climate/weather is utterly perfect for policy makers. It is an extremely complicated, young and wide field. Its main premise fulfils many needs for the power hungry whilst diverting attention from truly pressing problems for which they will offer no solution. Why is anyone on this planet EVER thirsty? We have the technology. Why is anyone EVER hungry? We have the technology. Why is anyone undereducated? We have a world body that should address all of these dreadful realities instead of .. well .. everything they fail at.
    The, so called, "green" environmentalists pushing this load are intent on control. Nothing else interests them. Amongst many other devices they will use your suppressed guilt over owning myriad "stuff" at the expense of others, mostly unseen, who have nothing but dirt and pain. They suffer no such guilt, especially the demagogues at the UN who would redistribute our supposed wealth to dictators, and worse, while laughing at their cleverness and our soft weaknesses and flying tens of thousands of "interested" parties to far-flung interesting locations. They could care less that their power is passed to them by those who, for example, tear down trees because hemp is too natural and universal for an industrialist to contemplate competing with. Where would centralized profit come from if a rampant weed were utilised in place of a constrained and expensive resource in just that one example?

    Complain about this comment

  • 39. At 1:03pm on 12 Jun 2009, xyz999 wrote:

    CarryOnRegardless, agwbsdotcom,

    Here is one example of a pre-2003 projection showing both no warming and cooling decades in a projection:
    http://www.metoffice.gov.uk/climatechange/science/projections/air_time.html

    This is the HadCM3 projection for the Third IPCC Assessment in 2001. [You can tell it's the 2001 version because they use the IS92a scenario. For the 2007 report they used SRES scenarios.]

    It has been obvious to climate scientists from the moment they started running these complex models, and you could argue that HadCM3 was leading the world in 2001, that you can get decades where there is no warming even with increasing CO2. This issue is a red herring. If there were another 10 years without warming [assuming no large volcano goes off in the tropics], then you might start to worry.

    Complain about this comment

  • 40. At 2:26pm on 14 Jun 2009, CarryOnRegardless wrote:

    XYZ999,
    Thanks for the link. This does not look to me like a decade long period of non warming. What this shows is an upward trend from 2000 to 2010, with a series of higher highs at 2000, 2005 and 2010. This suggest that the models were written to reflect the well known at the time ENSO events of a year or 2, and definately not the longer term cycles such as PDO and AMO. Look, we've had a decade of non warming, and several years of cooling. Hadley are predicting another decade of cooling. This is nothing like the graph you posted.
    The point I was trying to make in my last post is that projection of surface temps is a red herring itself, since this is a poor metric for measuring climate change. Let's get to the core issue. The fundamental core of the climate models is that there is an ongoing radiative imbalance. Hansen has stated (in 2003) that this should reach +0.85watts/sq. meter in 2010. This imbalance MUST result in accumulated joules of heat in the climate system. All I was trying to say in my last post was that this should be found in the oceans, but it isn't. Initially the argo network reported an apparent cooling anomaly since 2003 but this was later found to be the result of pressure sensor errors. This was corrected and the cooling disappeared. The argo readings now agree with gravimetric and altimetry measurements but still does not show the expected warming, as reported by Willis, Chambers & Nerem (2008) "Assessing the globally averaged sea level budget on seasonal to interannual timescales."
    Thus since 2003, there has been little if any global warming as diagnosed using ocean heat content. Since ocean heat content change has been shown to be a robust metric of global warmng and cooling, then in this sense, the models are not in agreement with empirical data. And, as I've already said, the previously predicted ups and downs of a couple of years are nothing like the decadal divergences we are actually witnessing.

    Complain about this comment

  • 41. At 09:24am on 16 Jun 2009, xyz999 wrote:

    CarryOnRegardless,

    I think you're missing the point I'm trying to make about these temperature PROJECTIONS - they're not trying to predict the evolution of certain decades, they are examples of how the climate might evolve.

    So.... if you were asking for examples of projections which showed that even on a global scale climate models could show decades where there was a cooling, then I have demonstrated that they do (e.g. the 2080s in the last link), and hence the current decade is not unexpected. If you wanted a PREDICTION of the current decade, then no-one has even tried to do that until the last couple of years, so there are no examples before then.

    As for the ocean heat content, there is clearly something to understand better there.

    No more from me on this topic.

    Complain about this comment

  • 42. At 4:41pm on 21 Jun 2009, LarryKealey wrote:

    Jack_NZ, you are close in your distiction between weather and climate. Weather is the short term manifestation of the current state of the climate. You are right on in stating that if they can't predict the weather 5 months (or 5 weeks) from now, they have no hope of making accurate predictions nor projections of any validity for the climate 50 or 100 years hence. While they are different problems, in theory, making weather predictions is a much simplier problem than predicting (or projecting) climate change.

    To xyz999, you are quite wrong in your distinction between projections and predictions. You state that the projections do not require accurate initial conditions. Quite False.

    The earth's climate is a very complex Chaotic Dynamic System. One of the key characteristics of Chaotic Dynamic Systems is "sensitivity to initial conditions". Tweek the initial conditions only slightly and you can see dramatic changes in the output of the system. Also consider that the earth and its climate system is an incredibly complex system. The models used to day have hundreds or even thousands of interrealted variables. The interrelations of these variables are very poorly understood today. Additionally, modellers of today really focus only on CO2 as the primary driver; hence these models all use a positive feedback with large sensitivity to CO2. That is to say that small changes in CO2 (a trace gas) cause large feedbacks in the atmospheric water cycle - a cycle which is very poorly understood. We may yet find that the feedback associated with CO2 is negative, indicating a very low sensitivity to CO2.

    While today, it may not seem intuitive, increased CO2 could acutally trigger other mechanisms in earth's climate system which could usher in the next ice age. Or, CO2 may not be the "primary driver" for climate change at all, and doubling of CO2 could have a negligable effect on temperatures and climate as a whole.

    The whole point is that we do not know enough today to create models which produce accurate predictions nor projections. Basically, they are all meaningless at this point. Until the hard science (not tweeking models, but real world experimentation and study) is done, we have no hope of understanding the earth's climate processes in any meaningful way.

    I applaud the author for presenting the reality of the current situation with climate modelling. He clearly points out that most models are used to explore possible reprecutions of various "emissions scenariors". The implication (or assumption, if you prefer) here is that CO2 and GHG emissions will be the primary driver for future climate change. In fact, from what I can read, all the "doomsday" models are predicated upon CO2 being the ONLY first order forcing agent for climate change. An assumption which I find highly doubtful.

    While keep in mind that proxies are just that - effects of climate impacts on various things - such as tree rings and growth. Much of our understand of past climate is based upon proxies. Perhaps one of the best proxies we have today is the Vostok Ice Cores from Antartica. Now, before we look at the cores and jump to any conclusions, consider the uncertainties in the source of the data. First, while atmospheric CO2 is trapped in the ice at the time the ice is formed, it also can move through the core, contaminating the evidence (CO2 concentrations) at other (nearby) levels within the core. One must also consider that ice is not constantly being formed, during warm periods, we have some melt, re-freezing, even during cold periods, we can have sublimation. So, it is not an exact record. It is just a proxy - but probably one of the best we have available today.

    In analysing the ice core, one will find two things that stand out. First, that CO2 levels generally lag increases in temperatures by about 800 years, rather than preceeding (i.e. driving) temperatures. The second thing the stands out (to me) is that atmospheric CO2 levels reached their peak, just before each of the last several ice ages. CO2 levels (and temperatures) were greater during the last interglacial period, 125,000 years ago, than they are today. This did not prevent the rapid onset of the last ice age. Simple fact: we just don't know enough today to make any meaningful predictions nor projections. The author is quite correct in stating we should take all these models with "a grain of salt" - perhaps we should take them with a pound of salt...lol.

    One of the earliest climate modellers was one of the pioneers of Chaos Theory, a gentleman by the name of Lorenz. In 1961, while crunching a series of runs of a climate model on a computer, he decided to take a short cut - and start the model in the middle of the time series run. He used values for the initial conditions for the new run from the state of the model during a previous run. Copied them exactly from the previous run into the new run and went to get a cup of coffee. When he returned, he found something completely unexpected - the new run had very quicky diverged from the previous run - and was compelete different - no similarity in output. The reason, quite simply: Sensitivity to initial conditions (aka, the butterfly effect) - the numbers he punched in were "calculated values", i.e. rather than exact values (digital vs analog). The changes in initial conditions were extremely small, yet had a dramatic affect on the output. Prime example of "sensitivity to initial conditions". An assoiate of his, Dr. Robert White (who later became director of NOAA) believed that because of this "butterfly effect", we could control the weather and the climate by taking very small actions (i.e. a butterfly flapping its wings in china sets in motion a series of events which causes an outbreak of tornadoes in the Midwest US). A belief that turned out to be quite wrong. Lorenz saw the falicy of this belief in that, while we could "tweek" the conditions only slightly and change the results dramatically, we could not predict the results of the "tweeking".

    Another characteristic of chaotic dynamic systems is that while they are inheriently unpredictable, they are highly deterministic. Basically, take a look at the Lorenz Attractor (you can google a picture quite easily), tweeking the initial conditions and parameters associated with the equation yeilds unpredicitiblity - you can't predict the path of the plot on the attractor, yet the plot will still lay with the confines of the "attractor". Note that even today, the dynamics of the system (a model) used the generate the Lorenz Attractor are still very poorly understood. And this system had only three equations with three variables. Climate models today have thousands of variables - to think we can "predict" the results of this system is simply laughable. Put simply, we can induce changes in the initial conditions of the system (i.e. CO2 concentrations and emissions scenarios), we cannot predict the results of those changes with ANY degree of accuracy. The mechanisms, processes and interractions within the system far from being understood in any meaningful way.

    We will probably never be able to accurately predict the weather nor the climate changes that will come. Consider this - just by measuring a thing, we affect it (a well known tenat of physics - the Heisenberg Uncertainty Principle). Thus, we can never accurately measure the initial conditions of the system to any degree of accuracy, thus, given inherent sesitivity to intial conditions, we have no hope. Make all the predictions or projections you want - but they still have no bearing on reality.

    In today's world, most every natural disaster is blamed on "Global Warming" or "Climate Change". In my lifetime, which unfortunately is almost over, I have seen a dramatic increase in the impacts of severe weather; however, I do not believe there has been a dramatic increase in the occurance of severe weather. Take New Orleans and Hurricane Katrina - often used as a "poster child" by the "Climate Change Gang". New Orleans was (and still is) a disaster waiting to happen. First, much of the city is currently 15 ft UNDER SEA LEVEL. Hello?? It is protected by a series of levies and dikes. The city was built upon a peat bog. As the bog dried out, it compacted. What was once 10 ft above sea level is now 15 ft below sea level. Katrina was a powerful hurricane, but still only a 3 on the Saphir-Simpson scale. The 1900 storm that hit Galveston (and killed 6,000 people, living on an island 5 ft above sea level), the 1919 Keys storm, Hurricane Allen and Hurricane Carla were all more powerful - as have been many others - but none caused as much damage.

    In 1969, I lived in Florida at Patrick AFB, next to Cape Kennedy. I watched Apollo 11 lift off for the moon from my back yard. At that time, one could drive down A-1-A along the coast and there were very few buildings and structures in sight. People did not have the arrogance to believe they could withstand the forces nature would occationally release. Take the same drive today (as I did last fall) and you will find just about every piece of land near the beach (low lying land) has been developed. Common sense tells us that hurricanes will continue to strike the area - and will be "more destructive", but not because the hurricances are more powerful, but because there is more there to be destoryed.

    Even the ancient egyptians were smart enough to not build permanant structures within the flood plain of the Nile river. They knew that floods would be variable and they had no hope of "controlling" them. In fact, they prayed for the great floods, because it meant dramatic improvements in soil conditions and the yeild of food from the plain would be significantly greater following the great floods. It astounds me that we here in the US are arrogant enough to believe that we can "control" the mighty Mississippi river with dikes and levies, channeling all that water into smaller and smaller channels. Building in the flood plain is a bad idea - as is building structures on the beach at levels 4-10 ft above sea level. One thing is certain, as long as we continue to battle mother nature in this arrogant fashion, we will lose. Every 20 years or so, there will be a big hurricane or big floods on the Mississippi. Nothing we can do or build can withstand the power of the rising waters. And by rising water here, I am referring to floods caused by snow melt in the areas drained by the Mississippi as well as the storm surge and dramatic rainfall caused by hurricanes - not "rising sea levels". In my view, and I have spent my whole life on the water, I see no rise in sea levels. I can take you to tide markers on Galveston Bay that have been in place for 50 to 100 years. They are not underwater today - ergo, the sea level has had no significant rise in the last 100 years.

    When I hear of the poor plight of the people of the Island Country of Tuvalu, often used another poster child for the "Climate Change Gang", I have to wonder: How stupid do they ("The Climate Change Gang") think I am? Tuvalu is a small string if islands, mostly atolls. The highest elevation on the islands is about 4.5 M above sea level. They are considered to be on the front line of climate change and supposedly have already felt the impacts of rising sea levels. Like all islands in the Pacific Ocean, the islands are subject to a variety of mechanisms which have lead to erosion and sinking of the land. Put very simply, it is not rising sea levels that pose the danger to Tuvalu - it is subsidence and erosion - both natural forces at work. take a look at the Hawaiian Islands. The only island that is growing is the (currently volcanically active) "Big Island". The Hawaiian islands actually stretch 1600 miles, from the Big Island all the way to Midway. Millions of years ago, Midway was "the Big Island", but the volcanism that created the Hawaiian Islands migrated (with the tectonic plates) eastward, leaving Midway to succumb to the ravages of erosion (from both wind, rain and sea). To blame the plight of the people of Tuvalu on sea level rise caused by "climate change" is downright misleading. Because of erosion, the islands will probably disappear. I say "probably" because, when the next ice age comes, we can expect sea levels to drop dramatically as great ice sheets form over North America and Europe. Perhaps the islands will re-emerge from the ocean for a time.

    No, because of the nature of Chaotic Dynamic Systems, the inability to make precise measurements and the complexities and interreactions involved in the "Earth's Climate System" - we will never have the ability to make accurate predictions (nor projections). Let us focus on spending our money and resources on real environmental issues such as better management of land use, forests, fisheries, pollution control (no, I am not talking about CO2 - which is not a pollutant, but an essential trace gas required for life as we know it, to exist on earth) and adaptation.

    While I will not delude myself into believing that I can change the (religious) beliefs of the "Climate Change Gang", I hope that my words will invoke thought and some modicum of a return to common sense (not so common anymore) in the "common man".

    Personally, I think at best, the "Climate Change Gang" are deluding themselves. I am astounded by their arrogance to believe that the "science is settled" or that the "skeptics are wrong". When there is no room for skeptics in science, it is no longer science, but religion. Before we spend trillions of dollars (or pounds) on a percieved problem (CO2 emissions), should we not develop a better understanding of the climate mechanisms involved? In my view, it is sheer, unfounded arrogance to believe we can "control climate change" and that CO2 is the only first order climate driver. We understand so little and we have so many other pressing environmental and humanistic issues that are much more urgent that we need to spend our money and resources on.

    The older I get and the more I learn, the more I realize how little I really know and how much more there is to learn...At least I am aware of my ignorance.

    Cheers for the Article.

    L Kealey
    Sugar Land TX, USA

    Complain about this comment

  • 43. At 5:27pm on 22 Jun 2009, U13900240 wrote:

    "But then wait until February 2009...
    Heavy snow hits much of Britain

    "Heavy snow has fallen across large parts of the UK, disrupting travel and closing thousands of schools."

    You think that because February 2nd is in winter, winter is 2nd February?

    Weird.

    Obviously skipped school when it came to geography at the secondary school...

    Complain about this comment

  • 44. At 5:29pm on 22 Jun 2009, U13900240 wrote:

    "CO2 is the only first order climate driver. We understand so little and we have so many other pressing environmental and humanistic issues that are much more urgent that we need to spend our money and resources on."

    Uh, the first order driver for car related deaths is speed.

    It's kind of an important one too.

    So where do you come to the conclusion that so little is understood?

    "At least I am aware of my ignorance."

    Doesn't mean that what you're ignorant of cannot be known, nor that it is unknown by others.

    This is why we have specialists.

    And why are you pushing your ignorance on everyone else? I don't know how to fix a carburettor but that doesn't mean I can't take it to the mechanic to be fixed...

    Complain about this comment

  • 45. At 5:33pm on 22 Jun 2009, U13900240 wrote:

    "20. At 5:20pm on 10 Jun 2009, Lucy_Skywalker wrote:

    It would be lovely to have real science back."

    I take it that the ONLY real science is one that supports your preconceived notion, young skywalker. You have much to learn.

    Will you try?

    Complain about this comment

  • 46. At 4:23pm on 23 Jun 2009, U13900240 wrote:

    " 32. At 2:52pm on 11 Jun 2009, CarryOnRegardless wrote:

    Sheffim,
    Nobody (I think) is arguing that carbon dioxide does not lead to a warming"

    Well, there's a change.

    LOTS of people (including senators and so on) were saying that CO2 doesn't lead to warming.

    There have been postings of Gerlich's paper that says it doesn't.

    People seemed to agree with it.

    Didn't seem that long ago either.

    Complain about this comment

  • 47. At 4:26pm on 23 Jun 2009, U13900240 wrote:

    Larry, when you sau "No, because of the nature of Chaotic Dynamic Systems, the inability to make precise measurements "

    You are right.

    A chaotic system is not arbitrarily accurate and errors in starting conditions can make a huge difference to the progression.

    When you say

    "the complexities and interreactions involved in the "Earth's Climate System" - we will never have the ability to make accurate predictions (nor projections)."

    you're wrong.

    A rugby ball rolled downhill will go down hill.

    That the ball's path down hill will depends HIGHLY on how you let it go and all sorts of uncontrolled changes in forces upon it doesn't make it roll uphill.

    Complain about this comment

  • 48. At 04:18am on 24 Jun 2009, LarryKealey wrote:


    @yeah_whatever: "A rugby ball rolled downhill will go down hill."

    Probably true, unless it hits an obstruction which deflects it uphill, in which case, the energy of the ball (as determined by its inertia upon striking the obstruction and the energy lost in the impact) will determine how far uphill the ball will travel...at which point, it may reach an "equilibrium point" and rest upon the ground, have additional energy, bounce and continue uphill, or it may begin to roll back down the hill. There are a lot of factors which will affect the behavior of the ball. And without a very RIGID set of assumptions, the behavior of the ball is in fact unpredictable. Of course, I make the assumption here that by a "hill" you mean a real hill, the surface of which is best described as fractal in nature, as opposed to a concrete "perfect" hill in which, from any point, there is always an unobstructed path (path of least resistance) downwards (like a "smooth" concrete hump - or something you might put in a computer model). Well, I have seen a few hills in England and Wales, and from my observations,most are much like hills we have in Texas and California...they are not perfectly smooth domes, if you roll a ball down, it will bounce, its path is unpredictable - it might even travel up-hill. Ah, but the devil is always in the details, particularly in the real world.

    We can only really say that a ball launched from a hill will follow the laws of physics and if unobstructed, will exchange potential energy for kinetic energy and accelerate. However, we can also state that eventually the ball will reach an obstruction and due to friction and exchange of energy through impacts, it will eventually come to rest (equilibrium). We really can't say whether that will be at the top of the hill, somewhere on the slopes, or at the bottom, can't really say with any certainty. Depends upon the structure of the hill, the kinetic energy we impart upon the ball as we launch it, various aspects of the rugby ball (as well as many other factors) and every hill is different...as is every toss, every ball.

    Now, please point out the falicies of my argument regarding the "ball and the hill".

    Now consider that we have just been talking about a very simple problem. One that should easily be modelled. Yet the best we could really do is to say that "under certain conditions - i.e., on THIS HILL, under these wind conditions and humidity (blah blah blah), if we launch THIS BALL with this trajectory and this much force it has a X% chance of following this path and reaching equilibrium at this point" (X being a very small number).

    Now think about that for a moment, and then consider the Earth's Climate System - and just how complicated a system we are talking about. Most researches would agree with the following statement, regardless of their postion on "climate change": The earth's climate is a very complex non-linear dynamic system containing a lot of very poorly understood processes and mechanisms.

    So, if we can't solve the dynamics of a simple problem such as rolling a rugby ball down a hill with any degree of certainty, what hope have we of predicting, much less controlling the earth's climate system?



    Complain about this comment

  • 49. At 05:36am on 24 Jun 2009, LarryKealey wrote:


    xyz999 writes:

    "IS clearly wrong. The starting conditions DO NOT matter for climate in 100 years time."

    And:(with regards to the models, those developed by the Met Office in particular)

    "surprise, surprise, they do project decades when there is a cooling:
    http://www.agu.org/pubs/crossref/2009/2009GL037810.shtml "

    First, the starting conditions DO MATTER A GREAT DEAL. This is one of the premises of all the GCM's as well as one of the cornerstones of the science: sensitivity to initial conditions. The models all treat CO2 as the primary forcing agent, and assign a large positive feedback to CO2 - so small changes in initial conditions produce dramatic changes in the results, thus initial conditions are VERY IMPORTANT.

    Regarding the link you provided - yes of course, after a decade of cooling, they (the Met office) developed a "model" to "account for this". That does not mean that they have the ability to accurately predict future climate change - and they will be the first to admit to this. Just read all the caveats in their report associated with the models and projections.

    Complain about this comment

  • 50. At 3:59pm on 24 Jun 2009, U13900240 wrote:

    "First, the starting conditions DO MATTER A GREAT DEAL. "

    Not in climate.

    Just like the initial conditions of rolling a ball down a hill even when the ball is oddly shaped and the ground uneven doesn't have a great deal to say about the fact that the ball rolls down the hill.

    "This is one of the premises of all the GCM's"

    Wrong.

    "as well as one of the cornerstones of the science"

    Wrong too. The predictability of chaotic systems is a cornerstone of modern science.

    Have you not generated F=ma from the Schroedinger equations? You have to treat the F=ma as the aggregate average result, but the Schroedinger equations give no concrete value for position. But it happens.

    Heck, the National Lottery uses chaos to make a random number, but the results are predictable enough that the rewards for a result are set before the numbers are rolled and yet the company can STILL make profits.

    Predictable randomness.

    Same for the Ideal Gas Law. Random energy distribution, random collisions, random EVERYTHING. Yet from that, we get

    PV=nRT

    A surprising result if chaos means we can't predict things.

    What about Great Attractors? The Lorentz attractor shows INCREDIBLE predictions about the random chaotic system it is a display of. See if you can think of some.

    The cornerstone of quantum mechanics is that randomness can be predictable if you think about what you can predict. Chaos' breakthrough was what you CAN predict from a chaotic system.

    "That does not mean that they have the ability to accurately predict future climate change"

    Yup, they call them projections. Just like they can predict that if saturated fats are removed from a diet, you get a healthier populace and that you can save X pounds from your healthcare system, EVEN THOUGH they cannot predict who will be better off and who worse or unchanged.

    Complain about this comment

  • 51. At 4:01pm on 24 Jun 2009, U13900240 wrote:

    "Probably true, unless it hits an obstruction which deflects it uphill, in which case, the energy of the ball (as determined by its inertia upon striking the obstruction and the energy lost in the impact)"

    What? Like reducing CO2 emissions or removing CO2 from the atmosphere could reverse global warming? Even though the specific path isn't known?

    Thank you for making my point for me.

    Complain about this comment

  • 52. At 4:03pm on 24 Jun 2009, U13900240 wrote:

    "However, we can also state that eventually the ball will reach an obstruction and due to friction and exchange of energy through impacts, it will eventually come to rest (equilibrium)."

    We can't if there is no such available obstruction on the path taken.

    You did want me to point out the holes in your argument, didn't you?

    And it seems you are saying that you CAN predict things about what happens in a chaotic system. Which isn't what you say earlier when you proclaim that climate is chaotic and you can't predict things because of that.

    Complain about this comment

  • 53. At 3:08pm on 28 Jun 2009, LarryKealey wrote:

    yeah_whatever: It appears that you "carryonregardless", just like the Knight in Monty Python who carries on after losing limb after limb...

    Actually, with regards to Chaos Theory, you are quite mistaken. They are deterministic, yet unpredictable. We can state that the trajectory will lay within the attractor, we cannot predict its path - and it never touches itself. The path followed is determined by the initial conditions of the system. Please read page 775 of IPCC Report Number 3 - they make my point quite nicely for me.

    Take the Lorenz Attractor (as you have already, so ineptly done) - highly deterministic, yet unpredictable behavior. Sure, we can predict that the trajectory of the line will lie within the confines of the attractor - we cannot predict where it will go next - we can't say that it will stay on one "wing" of the attractor or move to the other, nor when, nor what path it will follow - only that it will not intersect.

    Complain about this comment

  • 54. At 08:54am on 29 Jun 2009, U13900240 wrote:

    " Sure, we can predict that the trajectory of the line will lie within the confines of the attractor - we cannot predict where it will go next "

    Yes we can.

    It won't go backwards. It won't jump instantaneously to another distant spot.

    If we didn't know where it would go, how would a computer draw the pretty graph?

    My god, you're a blind fool, aren't you.

    Complain about this comment

  • 55. At 7:59pm on 01 Jul 2009, LarryKealey wrote:


    @yeah_whatever:

    ""20. At 5:20pm on 10 Jun 2009, Lucy_Skywalker wrote:

    It would be lovely to have real science back."

    I take it that the ONLY real science is one that supports your preconceived notion, young skywalker. You have much to learn.

    Will you try?"

    --------end of yeah_whatever post---------------

    Lucy, please carry on. It really would be lovely to have real science back - perhaps then, we can direct our (limited) resources toward real issues which will have a real impact on our environment - as opposed to schemes which have no hope of even solving the (supposed) problems they preport to solve...or unrealistic political agendas which have no hopes of success.

    Lucy, the Jedi council has decided to elevate you to the rank of "Master" as you have demonstrated your ability to think rationally and post intelligently. Please take your seat in the council chamber.

    Complain about this comment

  • 56. At 8:16pm on 01 Jul 2009, LarryKealey wrote:

    @yeah_whatever writes: (trying to redicule me)

    "" Sure, we can predict that the trajectory of the line will lie within the confines of the attractor - we cannot predict where it will go next "

    Yes we can.

    It won't go backwards. It won't jump instantaneously to another distant spot.

    If we didn't know where it would go, how would a computer draw the pretty graph?

    My god, you're a blind fool, aren't you."

    ------------end of yeah_whatever rant--------------------------

    Perhaps I am a fool, but no, I am not blind.

    Before you state that you can "accurately predict" the trajectory of the line on the Lorenz Attractor - change the initial conditions just a wee tad - and then tell me (before you calculate the trajectory) what changes your "tweeking" of initial conditions will cause. You can't.

    Saying that "it won't go backwards" or "It won't do this or that" is not a prediction of "what it WILL do". Can you tell me whether the trajectory will stay on the current "wing" of the attractor during its next iteration, or migrate to the other "wing" of the attractor? No, you can't even make such a simple prediction with any accuracy. If you believe you can, please demonstrate.

    Do you even understand the difference?

    Complain about this comment

  • 57. At 8:21pm on 01 Jul 2009, LarryKealey wrote:


    yeah_whatever writes:

    "My god..." - should we infer from that, that you believe in god?

    Wasn't it you who would dismiss scientists because they "believe in god".

    My god...

    Complain about this comment

  • 58. At 8:30pm on 01 Jul 2009, LarryKealey wrote:


    @yeah_whatever writes: (in yet another pathetic attempt to ridicule me...)

    ""However, we can also state that eventually the ball will reach an obstruction and due to friction and exchange of energy through impacts, it will eventually come to rest (equilibrium)."

    We can't if there is no such available obstruction on the path taken.

    You did want me to point out the holes in your argument, didn't you?

    And it seems you are saying that you CAN predict things about what happens in a chaotic system. Which isn't what you say earlier when you proclaim that climate is chaotic and you can't predict things because of that.

    -------------end of yeah_whatever rant------------------------

    I really don't know what world you live in, but it is certainly not the world I live on. On the world I live on (it's called planet earth) - at some point, the ball will strike an obstruction (or many) and will reach a state of equilibrium - i.e. come to rest. Please show me a hill on planet earth from which you can launch a ball and have it not come to rest.

    Please, just think about it - you will be famous - people have been trying to create the "perpetual motion machine" for a very long time - with no success. In one fell swoop, you could overturn much of our "current scientific beliefs" - Theories which we ascribe "laws of nature" to. Please don't be shy - show us all...

    Oh, you can't...hmmmmm

    Complain about this comment

  • 59. At 8:46pm on 01 Jul 2009, LarryKealey wrote:


    @yeah_whatever

    "And it seems you are saying that you CAN predict things about what happens in a chaotic system. Which isn't what you say earlier when you proclaim that climate is chaotic and you can't predict things because of that."

    Certainly we can "predict" things about what happens in a chaotic dynamic system - we can predict that the trajectory of the line on the Lorenz attractor will lie within the current confines of the attractor (or expand slightly, the bounds of the attractor). Chaotic systems are deterministic - that is not contrary to the statement that "they (chaotic systems) are inheriently unpredictable". We have no hope of saying that "after orbiting this wing of the attractor four more times, the trajectory will move to the other wing of the attractor" with any accuracy whatsoever. Nor what path it will actually follow - other than stating that the trajectory will lie within the attractor (probably).

    Complain about this comment

  • 60. At 00:46am on 08 Jul 2009, Mick wrote:

    Some interesting dialogue here, coincidently it looks like James Hansen's right hand person is now expressing similar caveats about the GCMs and their predictive powers. A section from the article.
    ----
    Another Government Scientist Admits Climate Model Shortcomings

    Another government scientist -- NASA climate modeler Gavin Schmidt -- admitted last week that the "chaotic component of climate system...is not predictable beyond two weeks, even theoretically."

    Schmidt made his admission during a June 29, 2009 interview about the shortcomings of climate models. Schmidt noted that some climate models suggest very strongly that the American Southwest will dry in a warming world. But Schmidt also noted that other models suggest the exact opposite.

    With these two models, you have two estimates one says it's going to get wetter and one says it's going to get drier. What do you do? Is there anything that you can say at all? That is a really difficult question, Schmidt conceded.

    The problem with climate prediction and projections going out to 2030 and 2050 is that we don't anticipate that they can be tested in the way you can test a weather forecast. It takes about 20 years to evaluate because there is so much unforced variability in the system which we can't predict the chaotic component of the climate system which is not predictable beyond two weeks, even theoretically. That is something that we can't really get a handle on, Schmidt lamented.
    ----

    The report is here http://www.climatedepot.com/a/1813/US-Government-Scientists-Shock-Admission-Climate-Model-Software-Doesnt-Meet-the-Best-Standards-Available

    "

    Complain about this comment

  • 61. At 02:44am on 08 Jul 2009, LarryKealey wrote:


    @Mick

    Yes, very interesting, isn't it? Its funny, a climate researcher, Dr. Edward Lorenz, at MIT predicted, in 1961, that we would never be able to accurately make climate forecasts because of the inherent nature of Chaotic Dynamic Systems - Like Earth's Climate System. Here, 48 years later - we finally hear the same thing. Schmidt also made a statement that we have "no hope" of making accurate predictions beyond two weeks due to the chaotic nature of the Earth's Climate System.

    But, back to your comment: one can hardly consider Dr. Schmidt a "denier" - to borrow the term from the "mantra". (I personally believe that there are "deniers" on both sides of the debate). If you read my posts, you will find I have been saying the same thing. We understand so little of the Earth's climate system - and due to its very nature, we have no hope of creating accurate models.

    I would assert that we certainly need to do the "hard" research if we want to ever develop a better understanding of Earth's climate system. You know - old science - before scientists spent all their time in their office, running computer simulations...out in the field, doing the hard science. Even if we have no hope of accurate predictions - this is still so much worth learning.


    Cheers.

    Complain about this comment

  • 62. At 10:28am on 08 Jul 2009, U13900240 wrote:

    "climate researcher, Dr. Edward Lorenz, at MIT predicted, in 1961, that we would never be able to accurately make climate forecasts"

    Define "accurate"?

    I cannot accurately predict the trajectory of a ball I throw. Nor can anyone. Yet there are people who make their LIVING from throwing balls accurately.

    Nobody can accurately predict the trajectory of an artillery or aircraft bomb shell. Yet computers calculate trajectories and targets get hit.

    Complain about this comment

  • 63. At 10:29am on 08 Jul 2009, U13900240 wrote:

    "We have no hope of saying that "after orbiting this wing of the attractor four more times, the trajectory will move to the other wing of the attractor" with any accuracy whatsoever."

    Now, is a 1000 year climate forecast on the SAME WING?

    Yes, it is.

    Complain about this comment

  • 64. At 10:31am on 08 Jul 2009, U13900240 wrote:

    "Another government scientist -- NASA climate modeler Gavin Schmidt -- admitted last week that the "chaotic component of climate system...is not predictable beyond two weeks, even theoretically.""

    I'll just link over to another moose who said the same thing:

    http://www.bbc.co.uk/blogs/thereporters/richardblack/2009/06/climate_meltdown_yet_fusion_la.html#P82584576

    And the response I would give is the same:

    http://www.bbc.co.uk/blogs/thereporters/richardblack/2009/06/climate_meltdown_yet_fusion_la.html#P82588307

    Short is: your reading comprehension stinks. You're reading what you want it to say, not what it says.

    Complain about this comment

  • 65. At 10:32am on 08 Jul 2009, U13900240 wrote:

    "On the world I live on (it's called planet earth) - at some point, the ball will strike an obstruction (or many) and will reach a state of equilibrium - i.e. come to rest."

    Prove it.

    Prove it 100%.

    It will not happen.

    Not unless the slope stops.

    And where did I say the slope stopped?

    Complain about this comment

  • 66. At 10:35am on 08 Jul 2009, U13900240 wrote:

    "Wasn't it you who would dismiss scientists because they "believe in god"."

    No.

    Again your reading comprehension takes a back seat to what you want words to mean. A common problem with denialists: if they read and understood they'd be wrong. And THAT CANNOT HAPPEN!!!

    If someone's belief in god makes them abandon science, then their science is nonexistent where it comes in conflict with their religion.

    Or, alteraatively, show us where I said "dismiss scientists because[/when] they "believe in god".

    Complain about this comment

  • 67. At 10:36am on 08 Jul 2009, U13900240 wrote:

    "Perhaps I am a fool, but no, I am not blind."

    The evidence says otherwise...

    "Saying that "it won't go backwards" or "It won't do this or that" is not a prediction of "what it WILL do"."

    But you say that it cannot be predicted. Saying "it won't do this or that" IS A PREDICTION.

    Complain about this comment

  • 68. At 10:38am on 08 Jul 2009, U13900240 wrote:

    "I take it that the ONLY real science is one that supports your preconceived notion, young skywalker. You have much to learn."

    Nope, Yidish.

    The ONLY real science is the real science.

    And that real science says that AGW is correct and is persued by thousands of climate scientists and presented in the many thousands of papers, many of which are collected and summarised in the IPCC report.

    If you think that you can dismiss that with saying "Well, since you don't believe MY preconcieved notions, you must be just accepting only of science that supports yours".

    This is projection.

    http://www.alleydog.com/glossary/definition.cfm?term=Projection

    Complain about this comment

  • 69. At 10:40am on 08 Jul 2009, U13900240 wrote:

    "It really would be lovely to have real science back"

    Then why not start to research it?

    You've displayed in many cases absolutely no understanding of what the climate scientists are saying unless they are ones denying AGW.

    Understanding the real science is a pre-requisite of knowing when real science is available...

    Complain about this comment

  • 70. At 10:40am on 08 Jul 2009, U13900240 wrote:

    "Lucy, the Jedi council has decided to elevate you to the rank of "Master" as you have demonstrated your ability to think rationally and post intelligently."

    Do you two want to get a room somewhere before we hurl?

    Complain about this comment

  • 71. At 10:42am on 08 Jul 2009, U13900240 wrote:

    "They are deterministic, yet unpredictable. "

    If they are deterministic, you can determine what will happen.

    This is otherwise known to the public as "prediction".

    Complain about this comment

  • 72. At 1:19pm on 08 Jul 2009, MangoChutney wrote:

    This comment was removed because the moderators found it broke the House Rules.

  • 73. At 3:05pm on 08 Jul 2009, MangoChutney wrote:

    This comment was removed because the moderators found it broke the House Rules.

  • 74. At 4:15pm on 08 Jul 2009, MangoChutney wrote:

    This comment was removed because the moderators found it broke the House Rules.

  • 75. At 6:16pm on 08 Jul 2009, Mick wrote:

    This comment was removed because the moderators found it broke the House Rules.

  • 76. At 6:27pm on 08 Jul 2009, Mick wrote:


    " "Another government scientist -- NASA climate modeler Gavin Schmidt -- admitted last week that the "chaotic component of climate system...is not predictable beyond two weeks, even theoretically.""

    yeah_whatever wrote:
    "I'll just link over to another moose who said the same thing:"

    Thank you for your kind welcome.

    yeah_whatever wrote:
    "Short is: your reading comprehension stinks. You're reading what you want it to say, not what it says."

    I did not write it, it was a quote but your response is understandable...

    As this blog topic is about climate model accuracy reviewing input from a significant player might just be relevant especially as Gavin Schmidt is also drawing attention to the vagaries of climate drivers and modelling them. There is also another recent document examining the strengths and weaknesses of the various models that could also be of interest unless the debate is over :), However, I found it of interest in how it outlines the foci of the various models and the question as to whether averaging them together is the most effective way of using the results, a point as they mention that the IPCC also deliberate on.
    I provided a link but as it is a PDF it is not allowed under BBC rules so for those interested it will have to be a search I am afraid. :)

    Complain about this comment

  • 77. At 6:31pm on 08 Jul 2009, Mick wrote:

    Found the citation for the Climate Model review that I mentioned above.
    Reifen, C., and R. Toumi (2009), Climate projections: Past performance no guarantee of future skill?, Geophys. Res. Lett., 36, L13704, doi:10.1029/2009GL038082.

    Complain about this comment

  • 78. At 7:21pm on 08 Jul 2009, U13900240 wrote:

    How about the IPCC reports.

    Or the thousands of papers that say differently, Mick?

    Or is it one paper is worth more than 10,000?

    In which case Mann's single work in 1998 must beat all the other papers that say he's wrong, since his one paper is less than their half-dozen.

    Complain about this comment

  • 79. At 10:52pm on 13 Jul 2009, Mick wrote:

    yeah_whatever wrote:
    How about the IPCC reports.

    Or the thousands of papers that say differently, Mick?

    Or is it one paper is worth more than 10,000?
    --
    I did mention that it was the IPCC that raised the question of strengths and weaknesses in their documents and I also mentioned follow up studies where those questions were further researched and published!
    Also Gavin Schmitt is one of the 50 or so scientists contributing to the IPCC reports so your questions seems unrelated.
    Science will revisit, revise and test and the quote and paper appears to be such examples of that.

    As for one paper being worth more than a 1000, Einstein is quoted as famously saying that it would only take one evidential paper to prove his theory wrong but that is a digression in this case.

    yeah_whatever wrote:
    In which case Mann's single work in 1998 must beat all the other papers that say he's wrong, since his one paper is less than their half-dozen.
    --
    Surreal logic. :) MBH98? Ever wondered why the Hickey Stick disappeared from the front covers?

    Complain about this comment

  • 80. At 08:17am on 14 Jul 2009, U13900240 wrote:

    "Surreal logic. :) MBH98? Ever wondered why the Hickey Stick disappeared from the front covers? "

    It didn't disappear from the report.

    Complain about this comment

  • 81. At 6:56pm on 19 Jul 2009, eumenydes wrote:

    "There is also another recent document examining the strengths and weaknesses of the various models that could also be of interest unless the debate is over :)"
    "Found the citation for the Climate Model review that I mentioned above. Reifen, C., and R. Toumi (2009), Climate projections: Past performance no guarantee of future skill?, Geophys. Res. Lett., 36, L13704, doi:10.1029/2009GL038082."
    This paper is only evaluating the strengths and weaknesses of the models *relative to each other*. It has nothing to do with the models' absolute predicitive abilities or validation thereof. It's studying whether we should treat some models as more reliable than others, nothing more.
    Nor does Gavin Schmidt's quote mean that the climate is unpredictable after two weeks - he's talking about the chaotic component only, which is why you need to look at long-term averages over 20 years. At that kind of timescale you're no longer looking at the chaotic component, but at the more stable component of the climate, which is predictable.
    To explain it simplistically: weather forecasts answer the question "will I need my umbrella tomorrow?", whereas climate predictions answer the question "how many times will I need my umbrella between 2040 and 2060?". You can't produce reliable weather forecasts for the year 2050 (obviously!), but you can produce reliable climate predictions for the decades around 2050. That's what Gavin Schmidt was saying.
    It seems various blogs and commenters are taking both the GRL paper you cited and Dr Schmidt's quote out of context (or just misunderstanding their meaning), which is a shame.

    Complain about this comment

  • 82. At 11:42pm on 26 Jul 2009, LarryKealey wrote:


    @eumenydes

    What I got from seeing excerpts from Gavin Schmidts interview as well as reading Mr. Schmidt's words where quite different than the meaning you imply. Might I provide one direct quote from Dr. Schmidt: (if you feel this is taken out of context, feel free to repost whatever additional paragraphs should be included before or after these two - but I think they make his points accurately and stand on their own).


    [Gavin Schmidt]
    "How do we come up with new expectations when our models are no longer valid? We need new information on which to base decisions being made now that will affect how we deal with the climate in 10, 20, 30, 50 years time because we are now building infrastructure for with these kinds of timeframes.

    We have to ask questions about what expectations we may have for the future based on the physics that we presently know, on processes we can measure, and that resonate with our ability to understand the current climate. Among the questions we need to ask are: Why are there seasonal cycles? Why are there storms? What controls the frequency of such events over a winter, over a longer period? What controls the frequency of, say, El Nino events in the tropical Pacific that impacts on rainfall in California or in Peru or in Indonesia?"

    -----End of Gavid Schmidt quote-------------------

    It seems to me that he is really saying two imprtant things here - first that we cannot "produce reliable climate predictions for 2050". Please read the first paragraph and if you feel I have mis-interpreted it, please provide you reasoning.

    The second thing I got from this is the earth's climate system (and weather) are very poorly understood at this point in time. I took his words (mind you these were his words, not those of a commentator or pundit) to mean that we have barely begun to scratch the surface of all we need to learn and all there is to learn. Science will never be settled.

    I also must disagree with your assertion that "At that kind of timescale (20 years), you're not longer looking at the chaotic component, but at the more stable component of the climate which is predictable" [@eumenydes].

    It just doesn't work that way with chaotic dynamic systems. Small changes in initial conditions of a chaotic dynamic system can produce significant changes in long term output. This is referred to as "Sensitivied to initial conditions".

    In working with a chaotic dynamic system (such as a model of earth's climate) - you are usually dealing with a set of non-linear equations - in the case of climate models - a vary large set of non-linear equations - requiring supercomputer to "iterate" to solutions of all the variables involved for each run. Ideally, one wants to "reduce" the order (size or number of equations and variables) and thus simplify the system (figure out what is really important).

    This can be achieved by applying the "Center Manifold Theorem". This theorem basicaly states that most dynamic systems are composed of three manifolds (stable, unstable, chaotic) and that if one can reduce the equation to the center manifold (chaotic) - we can reduce the complexity somewhat - and still preserve the dynamics of the system. [I wrote a package in Mathematics (R) as part of my Senior Thesis at the University to do just that - in symbolic form].

    It is the chaotic component of the system which is most important and the reason why we cannot make reliable, accurate "predictions" of future climate states.

    I don't wish to come across as being too arrogant here, but it was my field of study and I have made my career in modeling and to a certain extent - chaos theory. If not for the chaotic component, we would be able to really model the climate and make meaningful predictions. His point was that because of the chaotic nature, we will never be able to make meaningful "predictions" nor even projections.

    That is why I am a skeptic of the science. I am also very skeptical of the the suggested policies and programs currently being investigated (only one really: cap and trade).

    There are other solutions. There are other answers. We have much left to learn, but why create a whole economy around "carbon trading" when we are pretty sure it ain't gonna work? There are other approaches - even if you are not skeptical of the AGW/CACC theories out there today. Why not focus on developing the third world with cheap energy - which will make it feasable for them to preserve and restore natural environments (like rainforsts and peat bogs..) as well as reduce suffering and strife?

    I am a skeptic, but so what - even if CO2 is the "great evil" - I have yet to see a real plan to "solve" the problem. Cap and Trade is not going to do it - its just going to make those whe area already really rich, even richer and the rest of us even poorer.

    Ok - I'll get off my soapbox - sorry that was so long.

    Cheers Mate.

    Cheers.

    Complain about this comment

  • 83. At 11:48pm on 26 Jul 2009, LarryKealey wrote:



    @eumenydes

    I would also add that Gavin Schmidt discusses output from two models: one says the southestern US will be dryer with a rise in temperature - while another - fed the same data - projects that it will be wetter - they both can't be right.

    Cheers.

    Complain about this comment

  • 84. At 11:13am on 28 Jul 2009, eumenydes wrote:

    Hi Larry - I'll do my best to respond point by point, hopefully the various interspersed quotes won't make it too jumbled!

    > It seems to me that he is really saying two imprtant things here -
    > first that we cannot "produce reliable climate predictions for 2050".
    > Please read the first paragraph and if you feel I have mis-
    > interpreted it, please provide you reasoning.

    In the first paragraph you provided he isn't talking about climate models not being valid, he's talking about the existing engineering models (used for building infrastructure) which are based on historical climatology, which will no longer be valid in 10, 20, 30 years time because of climate change. Here is the full context:

    "Everywhere you go you see societies based around certain expectations for what their climate is. How far do you build away from the shore? How do you design your agriculture? What kind of air conditioning system do you put in a building? All of these things depend on the expectations you have for what temperature it is going to be during the summer time or how high a storm surge reaches when you have a Northeasterly storm. All of these things require an expectation that has been built over hundreds of years but that now is changing.

    When you have expectations based on past information that aren't any longer going to be valid expectations or you have a suspicion that they are no longer going to be valid, how do you come up with new expectations? How do you inform decisions that are being made now that will affect how people react to climate in 10, 20, 30, 50 years time?"

    He then goes on to explain how we have constructed climate models in order to answer that question.

    In the second paragraph you quoted, the questions he is asking are rhetorical ones - he's not saying we don't understand why there are seasonal cycles or why are there storms, he giving examples of the kind of questions that any climate model has to know the answer to if it's going to provide a valid simulation of the climate system.

    If we really didn't know why there are seasonal cycles, then you're quite right we would just be scratching the surface of climate science and wouldn't really have a clue how the climate works! Fortunately we do understand most of these things, which is what Gavin Schmidt was saying - he was just listing the kinds of issues we have to build in to the models, as he goes on to say later:

    "We have been quite successful at building these models on the basis of small-scale processes to produce large-scale simulation of the emerging properties of the climate system. We understand why we have a seasonal cycle; we understand why we have storms in the mid-latitudes; we understand what controls the ebb and flow of the seasonal sea ice distribution in the Arctic. We have good estimates in this regard."


    > Science will never be settled.

    I quite agree - in many areas of science every question you answer throws up two more. But that doesn't mean we never know anything!

    As it stands, theoretical physicists can't reconcile Einstein's Theory of General Relativity with the Standard Model of particle physics (ie they can't reconcile gravity with the other three fundamental forces). I fully expect that in the future we will have a theoretically self-consistent theory of quantum gravity, but until then it's a mystery. That doesn't mean gravity doesn't exist, or that it's not caused by an object's mass, or that we're actually being held on to the Earth's surface by invisible fairies. The fact that we don't know absolutely everything about gravity doesn't mean we know nothing about it.

    Likewise you're absolutely right to point out that on smaller regional scales climate models sometimes give contradictory predictions for things like rainfall. There is very high confidence in model predictions of global mean temperature, high confidence in their predictions of regional temperature changes, and reasonable high confidence in larger scale precipitation predictions - eg on zonal average or continental scales - but no, we can't say for sure what's going to happen to rainfall in the American Southwest. After more work has been done on this question (and many others!) and more improvements made to the models, I'm sure we will be able to answer that - but we're not quite there yet. That doesn't mean the models' other predictions are worthless.


    Hmm, this post is getting a bit long, sorry! I'll reply to your points about chaotic systems in a new post...

    Complain about this comment

  • 85. At 12:50pm on 28 Jul 2009, eumenydes wrote:

    > If not for the chaotic component, we would be able to really model the
    > climate and make meaningful predictions. His point was that because of
    > the chaotic nature, we will never be able to make
    > meaningful "predictions" nor even projections.

    With respect, I'm not doubting your own knowledge of chaos theory, but on the timescales we are talking about the climate is *not* a chaotic system. It has a chaotic component which makes it impossible to make *deterministic weather forecasts* after a week or so because of the initial condition sensitivity you mentioned. That is why no climate scientist (much less the IPCC) would ever say "on 28 July 2035 the temperature in London will be 22 degrees and there will be intermittent rain showers". That is weather forecasting, and is absolutely not what the climate models do. Climate predictions are over 20-year averages, which is a long enough time period for the stable, predictable component to become dominant over the chaotic component.

    Here is the full quote from Gavin Schmidt's interview:

    "The problem with climate prediction and projections going out to 2030 and 2050 is that we don't anticipate that they can be tested in the way you can test a weather forecast. It takes about 20 years to evaluate because there is so much unforced variability in the system the chaotic component of the climate system that is not predictable beyond two weeks, even theoretically. This is something we can't really get a handle on. We can only look at the climate problem once we have had a long enough time for that chaotic noise to be washed out, so that we can see that there is a full signal that is significantly larger than the inter-annual or the inter-decadal variability."

    In other words, we can't predict climate from year to year or even from decade to decade because the chaotic component is too strong. On periods of 20 years or longer the chaotic noise is cancelled out and we can look at the predictable component. That's exactly what climate models are designed for, and exactly what climate predictions do.

    That's not to say there aren't many uncertainties, but the climate is most definitely not inherently unpredictable on these timescales.

    For a better explanation than my (probably rather botched!) one, have a read of this article by Tim Palmer who is one of the leading experts on chaos theory in relation to climate and weather prediction: http://www.clivar.org/science/magnets.php

    Complain about this comment

  • 86. At 1:01pm on 28 Jul 2009, eumenydes wrote:

    > I am a skeptic, but so what - even if CO2 is the "great evil"
    > - I have yet to see a real plan to "solve" the problem. Cap and
    > Trade is not going to do it - its just going to make those whe
    > area already really rich, even richer and the rest of us even poorer.

    Firstly, I have never heard any scientist say that CO2 is a "great evil" or even any kind of evil. CO2=bad is a straw man argument that various cranks (such as in that Ch 4 "swindle" programme) have come up with to try and discredit climate scientists. I'm sure you didn't mean it that way so I'm sorry to harp on about a throw-away comment, but I've heard this so many times that it's starting to bug me, lol ;-)

    CO2 is a necessary gas for the continuation of almost all existing ecosystems. It's just that too much of it in the lower atmosphere causes an increase in global temperature and other effects. Likewise water is vital for life on Earth - that doesn't change the fact that if you get too much in your lungs you'll drown. A pollutant isn't an evil, it's just too much of something in the wrong place.

    As to the policy aspect - I'm not going to try and discuss policy or cap and trade as I don't know enough about it to offer an informed opinion I'm afraid. I too have reservations about its effectiveness and you may well be right that it won't help reduce CO2 emissions. However what I would say is that the scientific questions and the policy ones are completely separate and I wish people would keep them that way. Scientists answer the questions "what is happening, and what will happen if we continue emitting CO2?". Politicians et al answer the question "what shall we do about it?". Be skeptical of policies by all means (although I expect the relevant economists / policy experts might argue with you), but please don't allow your skepticism of the proposed policies to spill over into skepticism of the science itself.

    I would use the analogy of a medical diagnosis: A patient comes in who is overweight, drinks too much and smokes regularly, has high blood pressure and cholesterol and never does any exercise. The doctor will likely tell him that unless he changes his lifestyle he will almost certainly end up with heart disease, liver damage, increasingly immobile, with trouble breathing and at vastly increased risk of premature death from cancer or a heart attack. (And of course, each of those predictions has a different level of confidence attached!)

    But whether or not he does change his lifestyle is up to him - he may decide that he likes his bacon, beer and cigarettes so much that he can live with the increased risk of cancer and if he has to have expensive and painful surgery for liver failure, clogged arteries or heart failure then that's a price he's willing to pay. He may even make a legitimate value judgment that the pain he will have to undergo in trying to lose weight and give up smoking and alcohol will outweigh the pain of the resultant health problems. That's his decision and he may even be right.

    And even if he does decide to deal with the problem now rather than await the consequences, he still needs to decide how best to tackle it. Should he ditch the beer first, or concentrate on trying to quit smoking? Should he go on a strict diet, or just do more exercise? All are legitimate questions, and it's up to the patient to decide on the best answers.

    But what he shouldn't do is say "Eh, what does the doctor know - he's just pretending things are wrong with me in order to make more money. And anyway, I'm not overweight I just have heavy bones." Well, ok - I'm sure many people do say exactly that, and I can even see the appeal of that kind of delusion (which I'm sure is why "classic denial" is one of psychiatry's stock phrases) but that doesn't mean it's accurate or sensible!

    Heh - I promise I'll shut up now...

    Complain about this comment

  • 87. At 1:05pm on 28 Jul 2009, MangoChutney wrote:

    @eumenydes

    #84

    In the first paragraph you provided he isn't talking about climate models not being valid, he's talking about the existing engineering models (used for building infrastructure) which are based on historical climatology, which will no longer be valid in 10, 20, 30 years time because of climate change.

    Basically he is asking us to make changes to our building infrastructure, because of climate change, which, in his own words:

    It takes about 20 years to evaluate because there is so much unforced variability in the system the chaotic component of the climate system that is not predictable beyond two weeks, even theoretically.

    I'm sorry, but that is not good enough. He and others like him, should stick to reporting the facts and let society decide if there is a problem and how to deal with it

    Complain about this comment

  • 88. At 1:29pm on 28 Jul 2009, eumenydes wrote:

    "Basically he is asking us to make changes to our building infrastructure, because of climate change, which, in his own words:

    It takes about 20 years to evaluate because there is so much unforced variability in the system the chaotic component of the climate system that is not predictable beyond two weeks, even theoretically."

    See my reply to Larry above for the full version of this quote and its meaning. This is why we make predictions as 20-year averages, because on that timescale the climate system is stable enough to predict.

    "He and others like him, should stick to reporting the facts and let society decide if there is a problem and how to deal with it"

    I completely agree. And evidently so does Gavin since that's exactly what he's doing.

    Complain about this comment

  • 89. At 1:47pm on 28 Jul 2009, MangoChutney wrote:

    @eumenydes

    #86

    I've heard this so many times that it's starting to bug me, lol ;-)

    and i heard the term "denier" used many, many times (not by you) and it bugs and offends me, because of the holocaust connotations, but the BBC moderators continue to allow it's use.

    It's just that too much of it (CO2) in the lower atmosphere causes an increase in global temperature

    it's a question of how much of an increase and sensitivity. The IPCC choose a high sensitivity despite many calculations being very low.

    Likewise water is vital for life on Earth - that doesn't change the fact that if you get too much in your lungs you'll drown.

    Agreed, however, you can drown in only 30 mm of water if you are face down - it would take around 50000 ppmv before it even starts to become dangerous. All the emissions in the world won't get us to the point where CO2 will kill us

    Complain about this comment

  • 90. At 2:23pm on 28 Jul 2009, eumenydes wrote:

    and i heard the term "denier" used many, many times (not by you) and it bugs and offends me, because of the holocaust connotations, but the BBC moderators continue to allow it's use.

    Fair enough. I also dislike the term "denier" because of the connotations, however I'm also not keen on the term "skeptic" as skepticism is a healthy part of science. It seems inappropriate to describe people who fiddle data and distort graphs as "skeptics" (I don't mean you, but I've seen this from many so-called "skeptic" sources, e.g. Monkton).

    it's a question of how much of an increase and sensitivity. The IPCC choose a high sensitivity despite many calculations being very low.

    The IPCC sensitivity range of about 3 degrees for a doubling of CO2 is actually pretty middle-of-the-road, and is well-constrained by observational and paleoclimate records as well as by theoretical and modelling calculations. There are some calculations that put it lower or higher, but the very low and very high ends are usually based on faulty assumptions.

    Agreed, however, you can drown in only 30 mm of water if you are face down - it would take around 50000 ppmv before it even starts to become dangerous. All the emissions in the world won't get us to the point where CO2 will kill us

    Heh, sorry - I was being a bit flippant there. I only meant that because a substance has benefits doesn't preclude the possibility it can have negative attributes under some circumstances, like water causing drowning or CO2 causing temperature increase. I didn't mean to imply that the concentration of CO2 in the atmosphere will pose direct health risks!

    PS - thanks to your post I've figured out how to use italics!

    Complain about this comment

  • 91. At 4:20pm on 28 Jul 2009, MangoChutney wrote:

    @eumenydes

    It seems inappropriate to describe people who fiddle data and distort graphs as "skeptics"

    pot, kettle

    in my humble opinion, it seems inappropriate to describe people who fiddle data and distort graphs as "scientists" - i would have thought that was a politicians job ;)

    There are some calculations that put it (sensitivity) lower or higher, but the very low and very high ends are usually based on faulty assumptions.

    I disagree, see the work by Shaviv here:

    http://www.sciencebits.com/CO2orSolar

    and you are welcome for the italics ;)

    Complain about this comment

  • 92. At 8:00pm on 28 Jul 2009, eumenydes wrote:

    in my humble opinion, it seems inappropriate to describe people who fiddle data and distort graphs as "scientists"

    Indeed. Show me one who does so and I'll fully agree they don't deserve to be called a scientist.

    i would have thought that was a politicians job ;)

    Well now there's something we can probably all agree on ;-)

    I don't want to launch into a long critique of Nir Shaviv's work (I'd need another 3 pages to do it justice!), but in brief his calculation of climate sensitivity to CO2 (which falls within, albeit at the lower end of, the IPCC range) is based on the assumption that cloud ionisation feedback from cosmic ray flux accounts for a large part of both historical and paleo-temperature changes.

    Not only has there been no trend in solar activity or cosmic ray flux since 1952 (see for example http://www.agu.org/pubs/crossref/2005.../2005GL023621.shtml), the hypothesis that cosmic ray flux accounts for a large part of the recent warming is inconsistent with the observed decrease in diurnal temperature range (nights warm faster than days) and the cooling trend in the stratosphere, both of which are consistent with CO2 being the dominant forcing mode.

    Even Shaviv's work on multi-million year timescales has been criticised by fellow astrophysicists, see for example Knud Jahnke's paper concluding that Shaviv's trends were indistinguishable from a random distribution: http://adsabs.harvard.edu/abs/2005astro.ph..4155J

    Complain about this comment

View these comments in RSS

BBC iD

Sign in

BBC navigation

BBC © 2014 The BBC is not responsible for the content of external sites. Read more.

This page is best viewed in an up-to-date web browser with style sheets (CSS) enabled. While you will be able to view the content of this page in your current browser, you will not be able to get the full visual experience. Please consider upgrading your browser software or enabling style sheets (CSS) if you are able to do so.