RealClimate logo

Technical Note: Sorry for the recent unanticipated down-time, we had to perform some necessary updates. Please let us know if you have any problems.

Short and simple arguments for why climate can be predicted

Filed under: — rasmus @ 12 August 2006

Sometimes, I encounter arguments suggesting that since we cannot predict the weather beyond a couple of weeks, then it must be impossible to predict the climate in 100 years. Such statements tend to present themselves as a kind of revelation, often in social settings and parties after I have revealed for some of the guests that I’m a climatologist (if I say I work for the Meteorological Institute, I almost always get the question “so, what’s the weather going to be like tomorrow?”). Such occasions also tend to be times when I’m not too inclined to indulge in deep scientific or technical explanations. Or when talking to a journalist who wants an easy answer. In those cases I try to provide a short and simple, but convincing, explanation that is easy for most people to understand why climate can be predicted despite the chaotic nature of the weather (a more theoretical discussion is provided in the earlier post Chaos and Climate). One approach is to try to relate the topic to something with which they are familiar, such as to point to empirical observations which most accept (I suppose with hindsight it could be similar to the researchers in the early 20th century trying to convince that nuclear reactions were possible – just look at the Sun, and there is the proof! Or before that, the debate about whether atoms were real or not – just look at the blue sky, and you look at the proof…). I like to emphasised the words ‘weather‘ and ‘climate‘ above, because they mean different things.

Fig. 1: Temperature variation at 10E/60N from ERA40

It is true that we cannot predict the weather indefinetely (or even beyond a couple of weeks), because of the chaotic nature and infinitesimally small uncertainties in the state as we know to day, will affect how the weather evolves in a few weeks (the ‘chaos effect’). But, still I say that I know with certainty that there is a very high probability that the temperature in 6 months will be lower than now – when winter has arrived (it’s summer on the northern hemisphere at the present). In fact, the seasonal variation in temperature and rainfall (wet and dry seasons in the tropics) tends to be highly predictable: the winters at high latitudes are cold and summers mild (if anyone doubts, read on here); the southeast Asian Monsoon usually starts over India in the first days of June. I don’t usually bring with me maps and figures to social events, but it would be nice to show a picture such as the one in Fig. 1 to illustrate. If the person is not convinced, I may continue with other arguments for why the climate is predictable: take the latitude for instance – the poles are cold and tropics warm. Furthermore, maritime climates at higher latitudes with wet and mild (small day-to-day or season-to-season temperature variations) are distinct to continental climates far away from the sea (dry with great temperature variations). It is well-established that high-altitude places tend to have lower temperatures and greater temperature variations. Most hikers and mountaineers have experienced that. These are local climatic properties that we can predict if we know the geography, even if we cannot predict the weather on an exact day far in the future. To convince further, I may add that empirical evidence suggesting that (local) climate is not unpredictable, but rather systematically influenced by external factors (boundary conditions) is that Northern Europe enjoys a mild climate: Oslo is roughly on the same latitude as the southern tip of Greenland. There is a reason for that – Oslo has a considerably warmer climate because of the effects of oceanic heat transport/capacity and prevailing winds. I also remind that people really have known for centuries that there are systematic factors influencing the local climate, it’s just that this fact sometimes gets forgotten by those who claim that we cannot predict climate. Isn’t it silly? I may ask if there is any reason to think that the predictability stops at the seasonal and geographical variations.

Fig. 2: Mean 2m temperature from ERA40

I may continue with in a hand-wavy manner: In a similar fashion as seasonal and geographical effects, changes in Earth’s orbit around the Sun alters the planetary climate by modifying the amount of energy received from our star (but because of terrestrial response, the atmospheric composition is modified as well, enhancing the effect even further), and changes in the atmospheric composition affects the climate because grenhouse gases absorb heat that otherwise would escape into space – greenhouse gases are transparent to sunlight, but opaque to infrared light due to their molecular properties and their ability to absorb energy (if I say it’s quantum physics, people tend to understand it’s getting a bit technical). I stress that the greenhouse effect is also beyond doubt – without it, the energy balance between total energy Earth intercepts from the Sun and the energy lost through black body radiation implies that Earth’s surface on average would be about 30K cooler than we know it. Volcanoes also affect our climate, and we have theories explaining why. Furthermore, looking to other planets, the observation that Venus has higher surface temperature than Mercury, despite being further away from the Sun, can only be explained as a result of different absorbing properties of their respective atmospheres (a strong greenhouse effect at Venus).

So, my question is, do you think people get the message that I try to convey this way? Is it too simple or too complicated? Somebody who knows of every-day examples demonstrating the central principles? Any suggestions on how to explain for laypersons not connected to the Internet?

219 Responses to “Short and simple arguments for why climate can be predicted”

  1. 101
    Lynn Vincentnathan says:

    # 73, the same would apply to insurance companies (climate) v. individual events, such a car accident (weather).

    Re the gambling analogy: I went early in the morning & saw them “fixing” the slot machine at which I had won pretty good the night before. That’s like human GHG emissions. They said they were only fixing the print-out tape (denialists), so hoping they were right I (regular shmucks) poured down money in the same slot & lost all my winnings, plus some.

  2. 102
    pat neuman says:

    re 93. Dan, what you said is true, that many scientists have observed, documented, and published their results showing that rapid global warming has been taking place, with many years of good work (e.g. Lonnie Thomson at Ohio State). I’ll ask my question from 52. again… what action do people here think should be taken regarding the NWS Duluth MIC (Meteorologist In Charge) statement to TV media? He said:

    “Bottom line is we’re not really sure what is causing global warming or if it is even going on,” … (NWS chief in Duluth, May 2006)

  3. 103
    ike solem says:

    With regard to how the government agencies work these days, what we have is a government attack on basic science – a “stick your head in the sand” approach. The demise of the National Biological Survey is one example; the bloodletting at the USGS in the mid 90′s is another. NASA is now refusing to include Earth science study as part of its research program- and minor functionaries within NOAA and probably the NWS have been attempting to censor senior scientists in those organizations. (Thanks to pat neuman for the discussion of the hydrology of the Western US.) This attack on data collection is incredibly short-sighted, but depressingly familiar.

    As far as petroleum formation goes, if you want a good explantion read “Hubbert’s Peak” by Kenneth Deffeyes ( ) he includes a fascinating discussion of oilfield formation and is a highly respected petroleum geologist. Oil is a biotic product, but it is not enough to have marine sediments rich in organic matter; the geological history must be correct. An example of how tricky this can be is the Mukluk, Alaska story – a $2 billion dollar dry hole drilled into a ‘perfect formation’ in the 80′s- except that it was full of salt water, and no oil.

    Why is there no oil being formed today? With a few exceptions (such as the Santa Barbara sill area), anoxic bottom waters don’t exist, so the algal detritus is recycled back to CO2. If you want to read a non-’junk science’ explanation of eastern Meditteranean sapropel formation, take a look at ; you will see that increased nutrient fluxes and bottom water anoxia played important roles in producing the organic-rich sediment layers that are the source of the MidEast oil deposits. My comment was intended to show the drastic changes taking place in the world’s oceans, and as I said, it was ‘purely speculative’. The main sequences for mideast oil were laid down some 90 and 120 million years ago, according to Deffeyes, when climate conditions were quite different. Similarly, the world’s coalfields were created in terrestrial enviroments hundreds of millions of years ago under very different climate conditions. You can read about that here: We are raising CO2 to levels not seen for millions of years; we should expect drastic changes as a result.

    In the modern world, human nitrogen fixation and agricultural runoff produces high nutrient fluxes to areas such as the Gulf of Mexico and off of South Carolina (pig farms are the culprit here); that’s one driver of dead zones. By the way, including nitrogen is very important in any analysis of biological responses to increased CO2 and global warming. If you add a slowdown of ocean circulation (halting bottom water formation around the globe) then deep ocean anoxia is a real concern – along with ocean acidification. The dead zone off the Oregon coast appears to be more due to low oxygen levels in upwelled sub-Arctic water then to high terrestrial nutrient fluxes.

    Finally, I have a hard time accepting climate change skeptics who have a vested interest in continuing all-out fossil fuel use for the obvious reason that economic interests tend to collide head on with scientific accuracy. If the head of Shell (or of Hewitt Minerals) says ‘there’s nothing to worry about’ – well – that’s not the most disinterested source, considering that they are bidding on new oil leases and would hate to see a drop in demand for their product. If they want to stoop to personal insults – well, that’s just a sign of desperation. To be fair, I am very interested in seeing renewable energy companies take over the energy market – but the science supports my position.

  4. 104

    #102, That meteorologist should post his opinions, in a detailed and convincing way, since there was something like “bottom line” sum up with very few bits of science explained. Bottom line for me, statements from people using their title doesn’t convince unless its from some powerful scientist which really know what they are talking about. I found the Russian cooling of last winter an important point which needs to be discussed one day….

  5. 105
    Alexander Harvey says:

    Regarding Chaos:

    I am not sure that this will help but I shall give it a try. I will take some liberties, this is only meant to be illustrative not rigrous. Please feel free correct anything I get really wrong.

    Chaos can be displayed in systems that are totally deterministic, that is one of the things that makes it so intriguing. Systems that are totally described by state equations can be chaotic but are not random.

    Return the system to its initial state and it will evolve in the same way. Return it to its initial state but then change its state only slightly at it will exhibit its chaotic nature by evolving through diverging route in its state space. (This is totaly true of deterministic models but less so for systems that suffer from uncertainty (Quantum Mech.) but it should be undertood that the chaos does not arise from uncertainty it results from completely deterministic behaviour.

    Now we have to be careful how we define the state space. If we take a gas and define the state as a grand Hermitian of the position and momentum of every molecule then the states will diverge very quickly, the behaviour will be highly chaotic at the nanosecond scale.

    On the other hand if we define the state of exactly the same system in terms of temperature, density, pressure, and volume. We get the gas laws and the behaviour will be highly predictable (within certain limits).

    It is tempting to equate the former with weather and the later with climate.

    The weather is chaotic but it never strays too far from its norm. Its state describes a non overlapping path in its state space but it does not often stray into the realms of the highly unlikely. Also on occasion it will very nearly repeat itself (come close to some previous part of its path) and then diverge again. This is chaos folks, but perhaps not quite how we imagined it. Chaos refers to the inability to predict where in state space the weather will be in ten days but whatever weather we have in ten days it will be just plain old regular weather just like many days that have passed and many that will follow. Also a divergence that gives rise to rain instead of blue skies in ten days will be seamless there will be no discontinuity. Rain, shine, or snow we will have weather never, boiling oceans nor vacuums.

    The time it takes for its state to loop around state space to once again approach a prior state is not fixed but has a probability (expectation). In time intervals greater than this expectation the system can be viewed using a set of statistics (based on some type of averaging) of the original state parameters and these will have that layer of chaos smoothed out of them. This is not cheating, it is the very complexity of the systems that make this not only possible but inevitable. See gas laws above. (Strictly speaking it is the timescale before the system returns to a state that is similar, e.g. it does not matter if one exchanges all the particles, one can and must regard all the O2, etc. molecules as indistinguishable.

    So far so good, the ability for the gas laws to exist even though (or because) the underlying system is chaotic does not mean that a gas can not behave chaotically at a macroscopic level. Clearly it does, weather is chaotic, airflow around a body can be chaotic. But again a change in state parameters could once again restore low chaotic order. This is what is believed to be possible for climate modelling. Changing the state parameters and particularly the timescale can make a world of difference to what one can say about a system.

    It does seem that climate has a very low level of chaos. At least on any timescale relevant to the current argument.

    There is however a small fly that might be in the ointment. As long as the climate is not evolving it may have a tightly confined path in its state space. Cause it evolve and it may stray into regions that are more chaotic or worse still contain cusps (remember catastrophe theory?).

    If the first is the case then climate may become increasingly difficult to predict (more chaotic) if the second is the case the climate may suddenly transfer to a different part of its state space move away from the cusp and commenced to behave again in a well order but markedly different way.

    I believe that the current position is that gases are molecularly chaotic (position, velocity) but normally non-chaotic (pressure, density). The atmosphere is chaotic (weather) but non-chaotic averaged over long timescales (climate).

    The climate seems to have tracked in a closely confined non-overlapping path in its space for sometime and there is no good reason to suppose that this will not continue nor that any chaotic climatic behaviour will not become evident from running the models repeatedly for climate just as we would do so for weather to test for its chaotic nature.

    The one thing that we can not so easily know for certain is that there are no cusps in the system.

    Finally, I have to say that although the gas laws arise from the statistical treatment of molecules they have a very diferent form to the basic laws of motion. I do not know if there exists a similar set of climatic laws based on the statistical behaviour of weather. Laws that are above the vaguaries of the weather but totally due to it. Ideally one should be able to model climate with confidence without reference to weather at least in the long term (millenial or greater) perhaps this is the case. We may be unlucky in that there is not a sufficient jump in orders of magnitude in time between weather and near term climate to see the bigger picture and operate at that level. At that level there would be no argument about the predictive power of climatic laws no more than there is for the gas laws. Unfortunately the relevant state parameters might be termperature, pressure, density, humidity as averaged over 10,000 years. It is all a matter of timescales, statistics and convenience.

  6. 106
    Doug Percival says:

    ike solem in #103 wrote: “To be fair, I am very interested in seeing renewable energy companies take over the energy market – but the science supports my position.”

    Well, here’s something hopeful — according to these two energy experts, we are on the verge of a “revolution” in which the availability of cheap photovoltaics will fundamentally change the way we produce and distribute electricity:

    Solar Cells Change Electricity Distribution
    By Dave Freeman and Jim Harding
    The Seattle Post Intelligencer
    Thursday 10 August 2006

    In separate announcements over the past few months, researchers at the University of Johannesburg and at Nanosolar, a private company in Palo Alto, have announced major breakthroughs in reducing the cost of solar electric cells. While trade journals are abuzz with the news, analysis of the potential implications has been sparse.

    We approach this news as current and former public electric utility executives, sympathetic with consumer and environmental concerns. South Africa and California technologies rely on the same alloy – called CIGS (for copper-indium-gallium-selenide) – deposited in an extremely thin layer on a flexible surface. Both companies claim that the technology reduces solar cell production costs by a factor of 4-5. That would bring the cost to or below that of delivered electricity in a large fraction of the world.

    The California team is backed by a powerful team of private investors, including Google’s two founders and the insurance giant Swiss Re, among others. It has announced plans to build a $100 million production facility in the San Francisco Bay area that is slated to be operational at 215 megawatts next year, and soon thereafter capable of producing 430 megawatts of cells annually.

    What makes this particular news stand out? Cost, scale and financial strength. The cost of the facility is about one-tenth that of recently completed silicon cell facilities.

    Second, Nanosolar is scaling up rapidly from pilot production to 430 megawatts, using a technology it equates to printing newspapers. That implies both technical success and development of a highly automated production process that captures important economies of scale. No one builds that sort of industrial production facility in the Bay Area – with expensive labor, real estate and electricity costs – without confidence.

    Similar facilities can be built elsewhere. Half a dozen competitors also are working along the same lines, led by private firms Miasole and Daystar, in Sunnyvale, Calif., and New York.

    But this is really not about who wins in the end. We all do. Thin solar films can be used in building materials, including roofing materials and glass, and built into mortgages, reducing their cost even further. Inexpensive solar electric cells are, fundamentally, a “disruptive technology,” even in Seattle, with below-average electric rates and many cloudy days. Much like cellular phones have changed the way people communicate, cheap solar cells change the way we produce and distribute electric energy. The race is on.

    The announcements are good news for consumers worried about high energy prices and dependence on the Middle East, utility executives worried about the long-term viability of their next investment in central station power plants, transmission, or distribution, and for all of us who worry about climate change. It is also good news for the developing world, where electricity generally is more expensive, mostly because electrification requires long-distance transmission and serves small or irregular loads. Inexpensive solar cells are an ideal solution.

    Meanwhile, the prospect of this technology creates a conundrum for the electric utility industry and Wall Street. Can – or should – any utility, or investor, count on the long-term viability of a coal, nuclear or gas investment? The answer is no. In about a year, we’ll see how well those technologies work. The question is whether federal energy policy can change fast enough to join what appears to be a revolution.


    Dave Freeman has been general manager of multiple utilities, including the Tennessee Valley Authority, Los Angeles Department of Water and Power and New York Power Authority. Jim Harding is an energy and environment consultant in Olympia and formerly director of power planning and forecasting at Seattle City Light. Also contributing was Roger Duncan, assistant general manager of Austin Energy in Austin, Texas.

  7. 107
    Ike Solem says:

    Thanks Alexander for the very infomative comment on chaos as applied to climate and weather.

    There is another reason why global deep ocean anoxia (stratification was probably the wrong word) wouldn’t result in oil formation – plate tectonics and seafloor spreading. The oldest seafloor is some 100 million years old and most is younger, due to the formation of new seafloor at the mid-ocean ridges and seafloor burial at continental margins. The only place that a geological record of ancient ocean anoxia would be preserved is in the scraped-off sediments on continental margins (I believe the term is opheolite sequences). A discussion of modern deep water circulation with images can be seen here:

    The original atmosphere of the Earth is thought to have been high in CO2; photosynthetic fixation of that CO2 and burial of the resulting organic carbon explains the statement that ‘free oxygen in the atmosphere is a debt against buried carbon’. Happily, we don’t have to worry about stripping the O2 out of the atmosphere (I think) due to the fact that only a tiny fraction of buried carbon is in the form of oil and gas deposits.

    Consider oil and gas formation in California’s Central Valley, which used to be a shallow inland sea (similar to the Tethys sea). The continental location allowed for relatively shallow burial of sediments, and heat and pressure ‘cracked’ the organic carbon down to oil and gas. You can look at a pdf image of California oil and gas fields here:

    The northern areas (red) are gas fields, and the southern areas (green) are oil fields; this reflects the different geological history of the two regions (higher temps and pressures produce methane instead of oil due to more complete cracking; too high temps produce black carbon). The coastal zones are somewhat different, a great history of those regions can be found at

    Now, California produces about 30% of the hydrocarbons that are needed for in-state energy demand. By replacing all hydrocarbon energy imports (including foreign oil and out-of-state electrical transmissions from the Southwestern coal-fired power plants) with renewable solar, wind and biofuels, we’d have the ~70% reduction in fossil fuel CO2 emmissions that is needed to slow global warming. There are a number of pending bills in California (AB 32 and proposition 87) that will help meet this goal.

  8. 108
    Jim Dukelow says:

    Re #76

    Not so fast, Bryan. I followed your non-link to the Alaska Climate Research Center,, and find it does not support your suggestions in Comment 76.

    If you look at the trend for 1945 to 2005 for Alaska mean annual temperature,,edu/ClimTrends/Change/TempChange.html, you can clearly see how noisy the annual averages are compared to the 5-year moving average, with deviations 2.5 def F up or down being quite common. If you look at the time lines for individual cities (also available on the site), they are even noisier. If you look at the monthly summaries for January through July of 2006, you find some months warmer than usual and some colder. In the warmer months, usually a couple of the six cities summarized will have been cooler than usual, and vice versa in the cooler months. There is nothing in the data for the first 7 months of 2006 at all inconsistent with the high variability of Alaskan (and indeed all northern high latitude) weather/climate. Seven months of Arctic temperatures do not a trend make.

    Others have noted the inaccuracy of your comment about the sea ice. The ACRC web site also has a link to the U. of Illinois cryosphere snapshots,,The current snapshot, 15 August 2006, shows clearly how far back from the Arctic coastline the sea ice has retreated, from Norway east to Banks Island, Canada.

    Someone in the current thread seemed to be suggesting that the Arctic was becoming more saline because of diversion of the large Siberian rivers that feed the Arctic. I seem to remember a recent paper establishing exactly the opposite, increasing freshening of the Arctic due to increase inflow of fresh water from Siberia. I’ll hunt for the citation.

    Best regards.

    Jim Dukelow

  9. 109
    pat neuman says:

    Re 104.

    Wayne, it’s not people like you that need convincing in any way. Bottom line is that there are 120 NWS offices in the US, each with their own chief (MIC). Maybe just a few of them talk publicly like that but their comments off the record to media and people in other gov agencies are just as harmful to building any kind of public support to reduce GHG emissions as those who are more up front, probably even worse. Thus, unless they crack down on them all, I wouldn’t want to see them go after a guy that spoke out… but I agree that they should at least say what they’re basing their conclusions. Making general statements which are false is very harmful, if they know they’re false and say them anyway I’d call that criminal behavior of the worst kind.

  10. 110

    Pat , Scientists asking for having faith in them, should practice religion. It’s the facts, and interpretation of those facts that persuade, even for the lay, often ignored for their intelligence. Its encouraging that RC has such a high readership, it may just be that people are hungry for correct answers.

  11. 111
    joel Hammer says:

    Today’s climatologists remind me of some historians. After something has happened, they write their histories as though no other outcome was possible, that the actions of individuals counted for nothing, events were driven by large social, geographical, economic, etc. forces.

    If climatologists were confident enuf in their predictions, they could make millions of dollars playing the commodity markets. Afterall, how tough is it to forcast drought? Everytime we have a drought (somewhere) they say: See, we told you so. Climate change.

    Or, how about hurricanes? If you could really predict bad hurricane seasons, you could make millions. How? Wall Street. Go long or go short on insurance companies with Florida exposure, depending on your prediction of the hurricane season coming up. Sounds easy.

    Many pragmatic people, like me, notice that the climatologists are best at predicting the past. Like those historians.

    Climate is what you expect. Weather is what you get.

  12. 112

    Re # 99 – RE: #83 – “Here is something for yet another young, ambitious PhD candidate. A study looking at the changes in salinity in the Arctic between 10 and 100 Deg E Latitude owing to the massive diversions of North flowing rivers by the USSR (and maintained by the CIS). That would be an amazing study.”


    This is a new one on me, since I thought those river projects had been defeated in late Soviet times by a Soviet ecological movement led by Siberians.

    Am I wrong?

  13. 113
    Chuck Booth says:

    RE #108
    Here are a couple of the papers you may be thinking of:

    Nature. 2003 Dec 18;426(6968):826-9.

    A change in the freshwater balance of the Atlantic Ocean over the past four decades.

    Curry R, Dickson B, Yashayaev I.

    Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543, USA.

    The oceans are a global reservoir and redistribution agent for several important constituents of the Earth’s climate system, among them heat, fresh water and carbon dioxide. Whereas these constituents are actively exchanged with the atmosphere, salt is a component that is approximately conserved in the ocean. The distribution of salinity in the ocean is widely measured, and can therefore be used to diagnose rates of surface freshwater fluxes, freshwater transport and local ocean mixing–important components of climate dynamics. Here we present a comparison of salinities on a long transect (50 degrees S to 60 degrees N) through the western basins of the Atlantic Ocean between the 1950s and the 1990s. We find systematic freshening at both poleward ends contrasted with large increases of salinity pervading the upper water column at low latitudes. Our results extend a growing body of evidence indicating that shifts in the oceanic distribution of fresh and saline waters are occurring worldwide in ways that suggest links to global warming and possible changes in the hydrologic cycle of the Earth.

    Science 17 June 2005:
    Vol. 308. no. 5729, pp. 1772 – 1774

    Dilution of the Northern North Atlantic Ocean in Recent Decades

    Ruth Curry, and Cecilie Mauritzen

    Declining salinities signify that large amounts of fresh water have been added to the northern North Atlantic Ocean since the mid-1960s. We estimate that the Nordic Seas and Subpolar Basins were diluted by an extra 19,000 ± 5000 cubic kilometers of freshwater input between 1965 and 1995. Fully half of that additional fresh waterâ??about 10,000 cubic kilometersâ??infiltrated the system in the late 1960s at an approximate rate of 2000 cubic kilometers per year. Patterns of freshwater accumulation observed in the Nordic Seas suggest a century time scale to reach freshening thresholds critical to that portion of the Atlantic meridional overturning circulation.

    However, the story may be more complicated:

    Science 16 September 2005:
    Vol. 309. no. 5742, pp. 1841 – 1844

    Influence of the Atlantic Subpolar Gyre on the Thermohaline Circulation

    Hjálmar Hátún, Anne Britt Sandø, Helge Drange, Bogi Hansen, Heðinn Valdimarsson

    During the past decade, record-high salinities have been observed in the Atlantic Inflow to the Nordic Seas and the Arctic Ocean, which feeds the North Atlantic thermohaline circulation (THC). This may counteract the observed long-term increase in freshwater supply to the area and tend to stabilize the North Atlantic THC. Here we show that the salinity of the Atlantic Inflow is tightly linked to the dynamics of the North Atlantic subpolar gyre circulation. Therefore, when assessing the future of the North Atlantic THC, it is essential that the dynamics of the subpolar gyre and its influence on the salinity are taken into account.

  14. 114
    Jeffrey Davis says:

    The ability of making predictions, or at least plausible scenarios for the future, is in my opinion one of the reasons that we can call ourselves a highly civilised society.

    “Civilized” is only part of it. Passively waiting for the future to push us this way or that is fatalism and destrutive. I don’t think Global Warming skeptics are actually fatalists: they’ve made a choice and simply want their choice to stand unexamined. Or at least unchallenged.

  15. 115
    Steve Sadlov says:

    RE: #113 – According to Russian salinity maps, there is pronounced intrusion of higher salinity waters from the Atlantic, consituting a vast tongue that amazingly corresponds with the seemingly growing ice free area North of Europe. However, by the same token, it would appear that higher precip in Eastern Siberian and far Northwestern North America has resulted in higher runoff and subsequent increases in sea ice extent especially north of Western North America. Looking at the imagery of the annual minima over the past 27 years, one will note that the ice free area in this latter region at the minima has tended to decrease and in fact, this year is almost non existent. Bottom line – salinity matters.

  16. 116
    Joseph O'Sullivan says:

    #113 (Chuck Booth) & #108 (Jim Dukelow)
    RealClimate compared these studies in the “Saltier or not” post.

    #102 (Pat Neuman)
    I would ask who the “we” is when the NWS chief in Duluth says “we’re not really sure.”

    #98 (Bryan Sralla)
    I understand the scientific process. You seemed to be over-interpreting the importance of one paper. You seemed to be trying to use one paper to prove your doubts about the ability to predict climate and to model climate. It may be an important study or it may not be important. It is too early to tell.

    Who are the truly outstanding scientists I referred to when I used the word contrarian? I think you are over-interpreting my comment ;)

  17. 117


    you wrote “You fix the Earth to the space ship and tow it away from the Sun. What do you expect, that the surface mean temperature is constant as the distance between the Sun and the Earth increases? I don’t think so, but I’d like to see an argument to the contrary if somebody has one to offer.”

    In fact it would say under normal circumstances, ie those which have prevailed over the last 4 billion years, it would be constant. During that time the sun has increased its intensity by about a third and global temperatures have remained broadly constant. So if we dragged the earth out to where solar radiation is about a third less, then the temperature would remain constant. Beyond that, at some point, we would suddenly get a Snowball Earth. This “broadly constant temperature” means one suitable for life, but not neccessarily human life.

    Of course, it will be argued that it was high levels of CO2 that kept the earth warm when the sun was cooler, and for my “normal circumstances” I am assuming that the level of CO2 is close to the average over those 4 billion years. That may seem like cheating, but if you take current conditions, then the Milankovitch cycles provide a sort of tow rope making solar conditions oscillate. For the last 5000 years those cycles should have caused the planet to cool but until now the temperature has remained fairly constant.

    Of course we are not in normal conditions, we are in an ice age, and if the planet had been towed away before the Industrial Revolution, then it would have cooled because of the positive feedback from the growth of the ice sheets. But even with that effect, so far during this ice age, the earth has not become a Snowball Earth. So we could move the earth some way away before the whole planet became unihabitable.

    The point to realise is that the climate is a dynamical system, in which catastrophes can happen. There is a tendency to equate dynamical systems with chaotic states, and think that a strange attractor is a dynamical system, but it is not. It is only one state of such a system. A dynamical system can change from one stable state to another, in other words from one strange attractor to another. While this is happening the system is in an unstable chaotic state – a catasrophe. As Woodcock and Davies explain in “Catastrophe Theory” page 42 “the passage from the initial state or pathway [strange attractor] is likely to be brief in comparison to the time spent in stable states.” You can see this in the D-O cycles.

    This is of course related to what Andrew Harvey posted in #105. If we take a volume of gas as a system, then we can see that as we raise the temperature we move montonically from one stable state to another. However, if that gas is a mixture of hydrogen and oxygen, when the temperature gets to 500C there is an explosion (catastrophe) and we then have a new stable state of a gas called water vapour. Knowing the gas laws it is not possible to predict this catastrophe, so knowing the laws of weather does not mean that you can predict a catastrophic change in climate.

    Responding to my post #38 you wrote “One thing that has increased my confidence in the climate models is that features that are observed in the real world, such as El Nino (ENSO – some models do not give as good a representation as others though…), Hadley Cell, the North Atlantic Oscillation Kelvin waves, Rossby waves, and Tropical Instability Waves (TIWs) drop out of these models …” But I am saying that the models cannot reproduce rapid climate change. You appear to be saying that the glass is half full while I am saying it is half empty! However, you seem to be claiming because it is half full it is (kind of) full. I am saying because it is half empty it is not full.

    Your inquisitor at the cocktail party wants to know if unknowns exist with weather how can you be so sure about climate with its unknown unkowns. Honesty is the best policy :-(

    [Response:Right, honesty is the best policy! You're also right that the models are not yet perfect, and they are still evolving. Still, they are the best tools we have for making scenarios for the future. I do not agree with the half-full/half-empty analogy, though. The question of rapid climate change is still not very well understood, at least to my knowledge, and then it's difficult to judge if the models are bad. The reason for this is that we don't know if it was due to changes in the forcings(?). Is it due to unstable small-scale (up-scling) processes not yet resolved (we don't know if the models would predict them, if the spatial resolution was higher)? Or is the interpretation of the empirical data right (are we talking about rapid changes long time ago, or very recently, i.e. 1976/77)? The models still do have certain systematic biases (eg ENSO, sea-ice, sea surface temperature)- could those inhibit the models? My motivation for this post was to argue that I believe from empirical evidence that climate is in principle predictable - at least to some extent (the question of degree of skill is the next step) - when I am confronted with statements saying it's useless. If you look back 50 years, the numerical weather models (NWMs) were not as good as they are today, and they have improved slowly but steadily over time as the computers have become more powerful, numerical algorithms and observational network have improved, and as we have learned more about the Earth's atmosphere. There are still limitations to the NWMs as well as global climate models (GCMs), and there is some debate about weather a higher resolution may improve them. There are approximations and limitations associated with the description of complicated small-scale processes not resolved by the models (eg clouds, which I did mention, but was omitted from the quotation) and the representation of for instance sea-ice and snow cover (I mentioned the Monsoon, didn't I?). Still, the models give a very realistic picture: hence me bringing up examples such as ENSO, the NAO, the Hadley circulation, Kelvin- & Rossby waves, and TIWs. They are not prescribed in the models. You may also examine the annual cycle, which for the local climate at least, represents a case where the forcing is undergoing periodical changes: the models do quite well although not perfect. Hence, the GCMs give a realistic picture, but not a 100% exact solution. I'd say, pretty impressive, and very useful for making scenarios (thus the reason that the IPCC stress the word scenario). But, to wind up, my original purpose was to try to avoid referring to models, but try to find some convincing arguments based on empirical observations (a bit like the blue sky proof for atoms and the sun for nuclear reactions, really). Finally, I'd like to thank everyone for their thoughtful comments! :-) -rasmus]

  18. 118
    pat neuman says:

    The ‘Short and simple arguments for why climate can be predicted have been given. Now I have a short and simple prediction for global temperature in 2100 (upper right corner of yellow box at:

    Oh, and I will not be backing down what I said in 7. that: The increasing Dec-Feb temperature trends are so obvious in northern Minnesota that I can predict right now with a high degree of confidence that the winter of Dec 2006 to Feb 2007 in northern Minnesota will once again be above 1971-2000 temperature averages, and will continue to be above the 1971-2000 temperature averages for hundreds of thousands of years to come. I suppose if I’m wrong in 7. I’ll be wrong about 2100.

  19. 119
    Steve Bloom says:

    Re #115: Steve S., I have to say I get a bit peeved when on the one hand you demand greater precision in climate science, then say things like: “Looking at the imagery of the annual minima over the past 27 years, one will note that the ice free area in this latter region at the minima has tended to decrease and in fact, this year is almost non existent.” This statement is nonsense since we are as yet a month away from this year’s minimum. Quite a bit of ice will go away between now and then. If you want to do a comparison of current sea ice levels, try July of this year versus July of last year.

  20. 120
    Ike Solem says:

    Re #117 and comments – there is always the possibility of a big catastrophic event that is linked to the geological and astronomical systems – meteorite strikes and so on. My favorite area to visit is the Eastern Sierra, and when one drives by the Long Valley Caldera, one can’t help but wonder what the effect of suddenly injecting 500 cubic kilometers of material into the atmosphere would be (as happened ~800,000 years ago). I only mention this because of the original cocktail party theme of this thread.

  21. 121
    Jerry Steffens says:

    Re 109

    Interesting to note that the Monterey Bay NWS page has a link to an EPA page that discusses the potential impacts of global warming.

  22. 122
    Dan Robinson says:

    Maybe this has been mentioned somewhere here. I think the simplest explanation is that there are many more factors affecting specific local weather, producing relative weather chaos, than affect global averages.

  23. 123

    Re Rasmus’ reponse to #117

    Thanks for your long reply. I know how time consuming preparing these can be. Also on behalf of the other poster that you thanked can I say how nice it is to be acknowledged. It is nice to know that our posts are being read even when they do not get a response.

    You asked us for arguments to be used at a cocktail party to persuade a fellow guest that the climate models can predict the future climate. Here’s a rhetorical question – What arguments do you suggest I use to persuade serious scientists that their models contain a fatal flaw, and they should listen to my suggestions to correct it?

    I have tried honesty, and claimed that I have come up with the answer to rapid climate change, but that only makes me vulnerable to the accusation of being arrogant.

    I have argued that the current models do not reproduce rapid climate change, but you argued that it may be caused by an unknown forcing. But all the forcings are known and have been shown to be either not present or too small a scale to produce the effect. The surface temperature of the earth depends on only solar flux, albedo, and the greenhouse effect. We know from geological investigations that the forcing agents for these factors did not change sufficiently to cause the rapid warmings, thus it must have been the feedbacks that did it: water vapour and clouds. The forcing agents were mainly unchanged viz:
    1) Solar flux did not change because the Be isotope remained constant.
    2) Although there may have been changes in albedo due to changes in snow cover, these would not have been abrupt enough to cause rapid climate change.
    3) Changes in sea ice cover could have been abrupt because of the ice albedo effect, but the areas affected were not large enough to produce the effects seen.
    4) The evidence shows that the “fixed” greenhouse gases carbon dioxide, methane, and ozone did not change.

    That leaves water vapour, and it means that the strength of the feedback from water vapour is being underestimated.

    Agreed that the models are performing well at present, just as the gas laws work well for a mixture of hydrogen and oxygen at less than 500C.

    But you say you don’t want to discuss the models. I can see your eyes glazing over in disbelief as I try to get my point across. Oh well, I am not alone. You have the same problem at cocktail parties :-)

    [Response:I think you bring up an important point, but I don't know the answer to why the models cannot reproduce past rapid climate changes. There is an article by Wally Broackner (Was the Younger Dryas Triggered by a Flood?), which questions which mechanisms could be responsible for the initiation of the Younger Dryas episode (a ~1000 year of cold, characerised by abrupt changes, after the last glacial period had ended). If certain episodes such as accumulation of melt water suddenly pours into the ocean causes such events, then climate models with very crude ice representation may not capture these. Nor if such events were triggered by meteorites or volcanoes. If this is the case, would you still think that the models contained a fatal flaw? (by the way, I don't think you can be accused of being arrogant for asking this kind of question). I'd call it a limitation, as the models obviously can describe other climatic aspects with realism. Mind you, we may have different definiutions of 'flaw' and 'limitation', but it really depends on what you want to use the models for. If the climate models are being used to predict rapid climate changes, such as in the past, you're right they are flawed. But if you use them for making scenarios for the future given a slow change in the forcing, then the models are OK as long as no rapid climate change takes place (also, the Thermohaline Circulation could break down, but the oceanic models probably do not have a sufficient resolution to provide an accurate description of the ocean currents - which can be narrow - which play a central role). Sure, there are caveats associated with the climate models, and the future have many surprises in store. Climate researchers are aware about this - see for instance the discussion on thermohaline circulation and preception of risk. -rasmus]

  24. 124
    Alexander Harvey says:

    Is there a governing principle for climate and or weather?

    Something along the lines of:

    Climate evolves (tends) towards a state that maximises/minimises measurable climatic parameter(s) X(s). E.G minimising temperature maximising entropy.

    Or of similar importance:

    A deterministic climatic model evolves (tends) towards a state that maximises/minimises measurable climatic parameter(s) X(s).

    In the later case, I presume, when climate models are allowed to run with constant forcings, etc. some collection of parameters stabilise in the long term. (Is this so?)

    If it is then, it may be possible to construct governing equations that describe climatic tendencies in terms of the controlling parameters (by using the model with incremental changes of parameters to determine the governing equations, and to apply these governing equations to predict stable climates without the boredom of running the model each time. (I presume that something akin to this is done at least in the sense that micro-events like light showers can not be modelled individually but must be handled in terms of their regional significance.)

    I haven’t the time to think hard enough about this just now but I think this is the same as saying that the fabric of climate space is relatively “smooth” e.g. free of cusps, deep holes, etc. and so would always tend to a unique point regardless of how far away the initial conditions are (is this true of the models?). Or do the models contain metastable states or are they multistable?

    This is kind of important to know, I think.

    If they do not tend towards a unique point is it true to say that a similar law has local validity.


    Climate and its model evolve (tend) towards a state that locally maximises/minimises measurable climatic parameter(s) X(s).

    This is I think the same as saying that climate is locally stable in the long run. It might take a complex and chaotic path towards stability but that the tendency is clear and inevitable.

    Can anyone say what is known of the models in these type of terms. Not whether they produce particular results or that they are useful but whether they stabilise in a unique way?

  25. 125

    Re #117 and “if we dragged the earth out to where solar radiation is about a third less, then the temperature would remain constant.”

    No, it would drop precipitously, as some simple radiation calculations show. The time scale for adjusting to changes in Solar radiation is extremely slow, at the very least in the millions of years. You could probably glaciate Earth completely by suddenly dragging it out to where there was a third less sunlight.


  26. 126
    Alexander Harvey says:

    Can Climate be predicted?

    If not, what is the alternative?

    If climate is inherently unpredictable it means that no matter how much we struggle we can not know which of two or more significantly different futures is in store for us. More than that, it implies that the future is indeterminent. That is is impossible to know anything meaningful about future climate and that there is nothing in the current state of affairs that can determine it, and nothing that we can do to influence the outcome in a desired direction.

    I think that this alternative is nonsense (opinion) and even if it was so it is irrelevent (opinion).

    It would imply that increasing CO2 might make the world warmer, or cooler or do nothing. (Not simply that we can not currently predict what it would do, but that it is impossible to predict what it would do, starting from here it could take divergent paths, and that starting from here again it could do something quite different in a re-run. Also that even if it did either of the first two (heat up/cool down) , that decreasing or stabilising CO2 could either remove, enhance or do nothing to that trend.

    Now if the sun went out, would the planets get warmer of cooler?

    Cooler I think.

    So, climate is predicable to some degree at least.

    Someone needs to advise me here. The climate modelling of current interest is presumably for a future where CO2 forcing is varied. Over the short run different initial conditions may lead to different results. When these models are run with stationary forcings do they each converge in a meaningful sense to their individual stable points?

    Is there anything inherent in the process that says that they must or that they won’t uniquely stabilise? If models are inherently divergent (within each one, not in comparison) then they can not predict the future in the often accepted sense.

    This is a smaller but I think relevent consideration. Before we can anwser about the predictability of the real climate one needs to know if the model predictions are predictable (individually provide unique outcomes insensitive to plausible changes in initial conditions).

    Do the models predict anything? In the sense that multiple runs with different initial conditions produce a single cluster with a recognisable probability distribution. I presume the answer is yes, and I believe a mass trial of a weak model (climateprediction) with varying CO2 is in progress, does it confirm that the model is inherently capable or incapable of predictions?

    I should like to know if the big models stabilise if all other things are left unchanged and particularly how quickly they stabilise. The stability time constant might tell me something about the difference in timescales between weather and climate.

    [Response:I think you bring up an important point, but I don't know the answer to why the models cannot reproduce past rapid climate changes. There is an article by Wally Broackner (Was the Younger Dryas Triggered by a Flood?), which questions which mechanisms could be responsible for the initiation of the Younger Dryas episode (a ~1000 year of cold, characerised by abrupt changes, after the last glacial period had ended). If certain episodes such as accumulation of melt water suddenly pours into the ocean causes such events, then climate models with very crude ice representation may not capture these. Nor if such events were triggered by meteorites or volcanoes. If this is the case, would you still think that the models contained a fatal flaw? (by the way, I don't think you can be accused of being arrogant for asking this kind of question). I'd call it a limitation, as the models obsiously can describe other climatic aspects with realism. Mind you, we may have different definiutions of 'flaw' and 'limitation', but it really depends on what you want to use the models for. If the climate models are being used to predict rapid climate changes, such as in the past, you're right they are flawed. But if you use them for making scenarios for the future given a slow change in the forcing, then the models are OK as long as no rapid climate change takes place (also, the Thermohaline Circulation could break down, but the oceanic models probably do not have a sufficient resolution to provide an accurate description of the ocean currents - which can be narrow - which play a central role). Sure, there are caveats associated with the climate models, and the future have many surprises in store. Climate researchers are aware about this - see for instance the discussion on thermohaline circulation and preception of risk. -rasmus]

  27. 127
    Dan Hughes says:

    #123 Alexander. There are numerous papers and reports about modeling the climate, and other characteristics of planets, using a principle of maximum entropy production. Google the phrase “maximum entropy production” with ‘climate’. Bejan, Lorenz, Partridge, and others have papers and reports on this subject.

    [Response: While this theory is aesthetically pleasing, it isn't terribly well thought of in the community. For one, it doesn't consider rotational effects at all, and doesn't match lab experiments when rotation is included. And when it's used to to predict anything, it gives the same answer as was known already (Kleidon et al, 2003). But for another view read Ozawa et al (2003). - gavin]

  28. 128
    Dan Hughes says:

    This paper Constructal Theory of Global Circulation and Climate does in fact consider rotation and the abstract says, ” The earth averaged temperature difference between day and night was found to be approximately 7 K, which matches the observed value.”

    Are you saying that the large literature on application of entropy maximum to analyses of the properties of planets has been dismissed by the climate change community?

    It is my understanding that while GCMs do rotate neither the daily or seasonal variability are considered to be reliable. If this is not correct will you kindly point me to published results for these quantities.

    [Response: As far as I am aware, MEP theory has not produced any useful results, possibly because the climate fluctuations of entropy production are actually a very small fraction of the total entropy production and so are not usefully constrained by the theory (Stephens and OBrien, 1993; O'Brien 1997). Think of it this way - if the uncertainty in the climate sensitivity to doubled CO2 is a function of the micro-physics of clouds, how can a theory which neither contains greenhouse gases, clouds or micro-physics succeed in predicting that quantity? What about regional climate changes from the addition of aerosols? etc.

    With respect to the diurnal or seasonal cycles in GCMs, why do you think they are not well-modelled? They're not perfect, but the main features are well captured. I will however read the paper you point to and report back... - gavin]

  29. 129
    Dan Hughes says:

    #127. Gavin, and why should a model not prdeict what is already known? The GCMs do not predict at various spatial and temporal scales what is already known. This latter characteristic (not predicting) seems to be a major problem in contrast to the former (correct prediction) which many consider to be a major requirement of successful, useful modeling. The lack of the former in the GCMs is the focus of much discussion and debate.

  30. 130
    Dan Hughes says:


    We can take this offline, but I can’t find your e-mail address on RC. You have not correctly characterized some of the MEP models. And I hope you will provide corrections online. Additionally, I asked for publications in which the daily and yearly variability from a GCM is reported. You merely stated that ‘it is good’. Kindly send me reference citations and/or report them on RC. And, kindly send reference citations to predictions of regional climate and comparisons with measured data.


  31. 131
    ike solem says:

    One of the issues that plays into any discussion of climate physics is this: many people unfortunately still have a nineteenth century view of physical theory (this is not directed to anyone on this thread; but it is a fact). The notion that physics is an ‘exact science’ is at odds with 100 years of quantum theory and chaos theory research; that is why the term ‘constrained’ is used in place of ‘determined’. The primary culprit here is likely basic science education, which really should introduce these concepts at an early stage.

    Those constraints can be very, very accurate, but as gavin and alexander note the timescales are of fundamental importance. Imagine that we were coming out of a glacial period, and that vast pools of liquid water existed all over the Northern Hemisphere continents, held back by flimsy dams consisting of ice sheet remnants. At some point the walls would be breached and catastrophic floods would occur, but we should not expect a climate model to accurately predict the timing of such events – perhaps they would be triggered by an earthquake (try predicting those!). This relates to Alastair’s comment in #123, which includes the statement “but all the forcings are known” – but those forcings may be quite random, or be the products of systems with a high order of chaos. I would instead think of a situation in which a system builds up large potential energy, which can then be converted to kinetic energy by a little trigger (for example, Atlantic hurricanes are triggered by easterly waves, or in other words, “a subcritical bifurcation of the radiative-convective equilibrium state”).

    In that case, an important aspect of constraining rapid climate change is to carefully examine the system for such buildups of potential energy (i.e. warm sea surfaces and warm upper ocean profiles in the case of hurricanes). Then ask what mechanisms would result in a rapid and violent dissipation of that energy vs. a slow and gentle dissipation. At the end, you come up with a probabilistic assement, not a deterministic one.

    One other thing is chaos – which alexander brings up in #124. Experimental study of chaos is usually carried out in tiny ‘macroscopic’ systems – chaos in a test tube. This is quite tricky itself and one issue that arises is why reversible but chaotic microscopic processes lead to irreversible behavior – and applying this to something as large and complex as the climate system using the notion of ‘states’ and so on seems a big reach, but is perhaps a good guiding principle.

    Re #127 and the notion of maxium entropy production: Thermodynamics is often more useful in local situations where all the likely mechanisms can be included. Energy is the sum of enthalpy (heat) and entropy (disorder). I’ve been rereading Kerry Emanuel’s article over and over; he says it much better, so here you go: :

    “The relatively high surface temperature also means that atmospheric radiation exports entropy to space. The reason is that the atmosphere is heated at approximately the surface temperature, but it cools at the much lower effective emission temperature of Earth. In equilibrium, the planet must generate entropy, and the vast majority of that entropy is produced in the atmosphere, mainly through the mixing of the moist air inside clouds with the dry air outside them and through frictional dissipation by falling raindrops and snowflakes. Were it not for moisture in the atmosphere, the entropy would have to be produced by frictional dissipation of the kinetic energy of wind. The resulting air motion would be too violent to permit air travel.”

    This concept is then applied to the local situation of hurricanes. The balance of power generation and dissipation is then used to get an expression for the maximum wind speed a hurricane can generate, using the thermodynamic concept of the Carnot heat engine. As gavin points out a global use of entropy (maximum entropy production) would likely hide too many details to be useful. Cheers!

  32. 132

    Re 131 where Ike writes

    “At some point the walls would be breached and catastrophic floods would occur, but we should not expect a climate model to accurately predict the timing of such events – perhaps they would be triggered by an earthquake (try predicting those!). This relates to Alastair’s comment in #123, which includes the statement “but all the forcings are known” – but those forcings may be quite random, or be the products of systems with a high order of chaos.”

    The breaching of a dam wall wall is not a climate forcing. It may trigger a climate forcing by creating sea ice in the North Atlantic which changes the albedo, but all those forcings which could be triggered have been measured. The change in water vapour as a result the sea ice, has not been measured, and I am arguing that the climate modelling of that feedback is wrong. That is why the past cannot be simulated correctly by the models.

    Of course an eruption of a super-volcano could occur, which would invalidate all the models which predict a rise in temperatue of 3C. You are confusing the testing of the models with past data which can be done, with using them to predict the future under all circumstances which can never be done. We can only be sure they are correct when the future arrives!

    But what I am saying is that the models are predicting a montonic rise in temperature as a result of global warming. We know from the ice cores that the climate does not change that way. It switched out of the Last Glacial Maximum to the warm B-A inter-glacial, then switched into the cold Younger Dryas glacial where the temperature was fairly steady for 1000 years, then it switcted out of the Younger Dryas into the Holocene interglacial, and temperature has remained stable for the last 10,000 years. Without Man’s intervention it would have switched back to glacial conditions, but now we are warming the planet the next switch will be similar to two of the last three, ie to a hotter climate.

    A tipping point seems to have been reached where everyone believes that rapid climate change is the Younger Dryas stadial. The only people who still seems to have any sense are Seager and Battisti. See R. Seager, D. S. Battisti, in The General Circulation of the Atmosphere, T. Schneider, A. S. Sobel, Eds. (Princeton Univ. Press, 2005), and “Is the Gulf Stream responsible for Europe’s mild winters?” R. Seager ; D. S. Battisti ; et al, Quarterly Journal of the Royal Meteorological Society 128, 586 Page: 2563 — 2586.

    Even if the THC stops, Europe will still stay warmer than Newfoundland. Its climate is kept warm by winds blowing over the Gulf Stream, which is a western boundary current and will not stop flowing provided the Earth continues to rotate!

    [Response:I believe there are plans for a post on the Seager & Battisti paper in the near future, so I won't comment on it here. -rasmus]

  33. 133
    ike solem says:

    Hi Alastair,

    There seems to be a slight inconsistency in your argument that I’ve seen over and over again when it comes to the notion of an oncoming ice age as a result of a stable glacial-interglacial cycle. On one hand, you state that “You are confusing the testing of the models with past data which can be done, with using them to predict the future under all circumstances which can never be done. We can only be sure they are correct when the future arrives!”.

    This is a common problem in modelling; the preferred scientific approach since the Renaissance has been to conduct experiments under controlled conditions (Joseph Priestly is just one of the originators of this approach – now there was a great scientist!). Unfortunately, we don’t have a parallel Earth in which we could have kept things in a pre-industrial state and observed the resulting effects – we would actually have needed at least three such parallel Earths, and two more industrialized Earths, to make a minimal statistically valid comparison. Thus, the best we can do is detective work, and the only possible model tests are to compare the model output with the actual data (which again is why comprehensive data collection is so important).

    However, you then go on to say that “Without Man’s intervention it would have switched back to glacial conditions”; that’s a pretty definitive statement. You may be right, but the evidence for that is very slim – far less certain then the evidence for global warming. The last 8,000 years of climate have been remarkably stable compared to the previous glacial-interglacial cycles, according to the geologists. Why should this long stable period have dipped back into another round of glacial-interglacials? In fact, the only evidence I’ve seen for that is anecdotal reports of the horrible blizzards of the late nineteenth century, the local “Little Ice Age” that sent the Scandinavians southwards on ocean exploration voyages, and the ‘hockey-stick’ climate record that has come under so much attack in Congress.

    So, how do you resolve that contradiction? Shouldn’t we base our actions on the known data and models, and if so, shouldn’t we take drastic steps to meet that ~70% reduction in fossil fuel CO2 emissions that the models say might stabilize the climate?

  34. 134
    Alexander Harvey says:

    My Thanks goes to Dan Hughes and Gavin.

    MEP is just the sort of governing principle I was grasping for. So thanks again.

    It will take me a while to get up to speed but already I feel that the reservations as to how it can be used predictively (Gavin: “how can a theory which neither contains greenhouse gases, clouds or micro-physics succeed in predicting that quantity?”) could be its great strength.

    By that I mean that the principle (OK, hypothesis) cares not about the mechanism. It will exploit what ever means are available to maximise work in the system. It might however fail to predict the nature of the weather in general or the climate in fine detail.

    This is to say that it might not by itself produce better ways to model climate but it is something that is clear enough to hold in one’s head, and think about in order to see its consequences, which I shall try and do.

  35. 135
    Steve Sadlov says:

    RE: try July of this year versus July of last year.

    Been there done that, both July and current are have a negative anomaly with a lower amplitude than last year. I stand by my prediction.

  36. 136
    Steve Sadlov says:

    FYI …. report from yesterday from the NWS Anchorage ice desk …. highly dependable, the “Deadliest Catch” gamblers depend on it:

    THE MAIN ICE EDGE LIES FROM 71.4N 140W TO 72N 146W TO 70.5N 145.5W TO
    71N 150W TO 72.4N 153W TO 71.4N 160W TO 72.6N 160W TO 71.5N 171W





  37. 137

    Re #133 Re However, you then go on to say that “Without Man’s intervention it would have switched back to glacial conditions”; that’s a pretty definitive statement. You may be right, but the evidence for that is very slim – far less certain then the evidence for global warming.

    Here is the CO2 and Temperature records from the Vostok core that the Russians drilled in Antarctica. You can see that over the last four years there have been short periods, roughly 100,000 years apart, when the temperature was as high as that today. They did not last long, so it is reasonable to suggest that this interglacial would be similar, and that Antarctica would soon descend slowly back into glacial conditions.

    But it is more complicated than that. This interglacial has lasted longer then the two preceding ones but not that 400,000 years ago. Why is that? Is it because the Milankovitch cycles are in a similar state now to those 400,000 years ago, or is it because Man has kept the climate warm as a consequence of the development of farming? We don’t know, but I am assumng if Man was not here then the planet would return to another glacial period sooner or later.

    For the results of a core that goes futher back see: (note this chart has the time flowing from right to left)

    The “Litttle Ice Age” which happened 400 years ago would not show up on the Vostok ice core even if it were there. This Ice Age began 2,000,000 years ago and it contains glacial periods when it is cold, and interglacial periods when it is warm like today and during the Little Ice Age. In othre words the little Ice Age was not an ice age, it was not even a glacial period. It was just slightly cooler.

    Those two pages showed what happened at the south pole, but the vast maajority of people live in the Morthern Hemisphere. The ice cores from Greenland only go back to the last interglacial, (note this chart has the time flowing from right to left)
    But they show much more violent changes in temperture than in the SH. Unfortunately it is not clear what happened at the end of the last interglacial in Greenland because the ice at the bottom of the core may have been distorted.

    It seems that these violent changes were due to the formation and rapid disapearance of sea ice sheets triggering runaway changes in water vapour content, because of the Clausius-Clapeyron effect. The sea ice no longer streches down to the Irish coast as it did during the Younger Dryas, but when the Arctic sea ice disappears, then we will have another rapid warming. It is too late to prevent that happening by reducing CO2 levels. We, or at least the USA, will just have to put a sun shield into space, but they better do it damned quick!

  38. 138
    pat neuman says:

    re 103. A thank you note to ike solem for expressing appreciating. My research on earlier snowmelt runoff dealt with the Red River (ND/MN), the St. Louis River (near Scanlon/Duluth MN) and the St. Croix River (MN/WI). I haven’t done my own research in the Rocky Mtns, but I have read articles as I post them to my yahoo group called ClimateArchive. I thought the series in 2004 by Jim Erickson, Rocky Mountain News was good.

    Part 2: Going, going, gone?
    Front Range glaciers declining; researchers point to a warming world

  39. 139
    Steffen Christensen says:

    Re: Alastair in #117 and Rasmus earlier – towing the Earth.
    If you tow the Earth out today a significant distance, it’ll freeze. Explaining the very long-term climate stability of our little home here is tricky, and not currently well known – save that the answer certainly has to do with CO2 concentration, and quite possibly some of: methane, availability and abundance of continental shelves to evaporate lots of H2O to the air, relative amounts of geological energy going down over geological time as the U235 decays off, placement of continents near the poles or near the equator, and possibly a few others I can’t think of off the top of my head. Google “wikipedia supercontinent cycle” and related links for a plausible theory of how the Earth froze solid a couple of times while staying right in its present orbit. However, past performance of the Earth is no guarantee of future performance. Absent a plausible physical model (towing asteroids around?), about the best you can do is plug the numbers into a GCM and see what happens. As Barton pointed out around #125, certainly you’ll get a proper Ice Age as the drop in incoming solar flux overwhelms the current deglaciation of the planet that our little adventure in carbon burning. In a feedback-free Arrhenius model of the Earth, towing the Earth out to 1.225 AU – where the sunlight is 2/3 of our present sun – drops the global mean temperature by some 6.9 degrees Celsius. Feedbacks should amplify that effect somewhat. In GCM terms, that’s a global longwave forcing of -22.2 W/m^2. I don’t know how easy it is to reconfigure a GCM, but if that doesn’t freeze the oceans, I’m not sure what will. Also you’d have to move the moon, else we lose tides and such. ;)

  40. 140
    Pekka Kostamo says:

    On the issue of ice ages. I remember reading about a year ago a report I am no more able to locate.

    Some French astronomers had re-computed the Earth orbit using latest and best values for the sun, the planets and all their moons. This for the past and the future 25 million years.

    Their conclusion was that there would not be a next ice age, or other major variation driven by celestial mechanics. Earth orbit has assumed a relative stability for some considerable period of time.

    It was just a press release.

    Best regards,

    [Response: Loutre and Berger, I think. See - William]

    [Response: Belgian (U. Catholique de Louvain), not French. -gavin]

  41. 141
    ike solem says:

    Hi Alastair,
    You state, “but when the Arctic sea ice disappears, then we will have another rapid warming. It is too late to prevent that happening by reducing CO2 levels.”

    I fear you are correct, but the magnitude of that rapid warming may be reduced by drastic actions – a massive reforestation program and a massive renewable energy program, and a total ban on fossil fuel combustion (that’s about as drastic as you can get, I suppose).

    Here is a bbc report from 1999:

    And here is one from 2005:

    Warming at the poles is agreed by all to be the most intense, so people who say that undersampling the Arctic ocean is not important are way off base. Shutting down thermohaline circulation seems unlikely to cause a cooling in Britain because of the compensating increased polewards heat transfer. The thing to look at will be the salinity. Increased evaporation vs. freshwater dilution? Hard to say, I suppose.

    If I had to bet a large stretch of coastline on your statement being correct, well – just look at insurance company policy these days on coastal areas. They seem to agree with you as well.

  42. 142
    ike solem says:

    Forgot to include this in that last post:

    Now, shouldn’t NASA be involved in this? If the ‘skeptics’ had honest intentions, wouldn’t they be clamoring for more data and a new satellite? Sigh…

  43. 143
    Dan Hughes says:

    This paper, on calculations of orbits states that chaos limits the accuracy of the calculations. The last sentence in Section 4 states, “The ultimate obstacle to achieving higher accuracy is chaos and not simulation errors, assuming that we have a sufficiently accurate physical model.” I read this to say that the presence of chaos ultimately limits what can be predicted and the accuracy of the predictions.

    All corrections to my understanding are welcomed.

    [Response: This is an initial value problem (just like weather), and just like weather, the accuracy through time is limited by the size of the Lyapunov exponents which describe the divergence of two close initial states. You might want to check what the time-scales for that to be felt in orbital calculations and then report back as to whether you think it's relevant on the ten-to-fifty thousand year time scale..... - gavin]

  44. 144
  45. 145
    Hans Erren says:

    so climate cannot be predicted due to the unknown future development of the boundary values (sun, volcanos, emissions).

    It is always “what if”, and that is not a prediction s.s.

  46. 146
    Dan says:

    re:145. “Oh my goodness.” ;-) There is so much information on this web site about climate modeling that it is astonishing and sad to read comments like that. Unless of course it is simply another in the long line of “drive by” postings with no scientific support. Guess we shall see about that. Meanwhile, since around 1970, global natural forcings can not explain global temperature trends over the same period. In short, your “boundary values” (natural forcings) have been overwhelmed by man-made forcings. Which have been modeled quite well within reason (i.e. there is always some scientific uncertainty). Future modeling is based on various scenarios due to a variety of reasons. The results produce a range of possibilities that are consistent.

  47. 147
    Steve Sadlov says:

    RE: #145 – Do you understand what he meant by boundary values? Why is the mention of them sad? Are not boundary values the essence of getting to solution sets for complex thermal scenarios? This is basic stuff, basic upper division calculus, basic upper division heat flow. Did you ever take such courses?

  48. 148
    Steve Sadlov says:

    Oops, I meant to write “RE: #146 …..”

  49. 149
    Dan Hughes says:

    I’m reporting back re: chaos, initial values, boundary values, and Lyapunov exponents. The question is not what the Lyapunov exponents are for the calculations of orbits. The question is what are the values for the continuous equations for long-range climate calculations. Are they of such values that the accuracy is not limited for climate calculations over the time scales of interest, or of such values that the accuracy is limited. In the sense that climate is chaotic, these are questions that demand rigorously determined answers such as provided in the linked paper. The Lyapunov exponents for orbit calculations are not those obtained for AOLGCM calculations.

    A major part of the paper linked in #143 is devoted to Verification of the accuracy of the numerical integration methods. And because the basic equations are ordinary differential equations ODEs the Verification can be rigorously determined. The calculations of orbits is an initial value ODE problem because the boundary conditions were set by the physical model mentioned in the quoted sentence.

    Climate models are partial differential equations (PDEs) and require initial and boundary values. Unlike the case of ODEs, determination of the accuracy of numerical solution methods for PDEs is more nearly an art; not a mathematical science as for ODEs. Numerical approximations for boundary conditions, by the way, can destroy the most carefully-laid plans of numerical solution methods. So far as I know, the Verification of the numerical solution methods for the PDEs used in AOLGCM models has not been demonstrated. The order of the accuracy of the numerical approximations in the asymptotic range has not been numerically demonstrated. The order of the as-implemented truncation errors for the entire model and its applications procedures is also unknown. And, additionally, the Lyapunov exponents for the continuous equations have not been determined. If this statement is not correct, let me know. Additionally, any papers in which the Lyapunov exponents have been rigorously estimated for the case of numerical integration of a system of non-linear PDEs would be appreciated. The specific case of any AOLGCM is especially of interest.

    One successful method that is usually applied to numerical solution methods for PDEs is demonstration that the calculated results are independent of the sizes of the discrete temporal and spatial scales used to integrate the equations. This exercise will allow the as-implemented-and-applied order of convergence and truncation errors to be estimated. If this exercise has been conducted for any AOLGCM model and code and calculation I have not been successful in locating it. Any assistance along these lines would be appreciated.

    Equally important is this sentence from the conclusions of the paper, “The present results, therefore, need to be checked by researchers who work independently of our group and use different simulation methodologies, if possible.”

    A book by Pat Roache, Verification and Validation in Computational Science and Engineering provides an excellent detailed discussions of these issues.

    This report Verification, Validation, and Predictive Capability in Computational Engineering and Physics has additional information.

  50. 150
    Dan Hughes says:

    Oh, now I get it. Some thought I was questioning the accuracy of the Planet’s orbits as providing BCs for AOLGCMs.

    Nope, that’s not it. It’s the problem of chaos in the AOLGCMs.

Switch to our mobile site