A Theory to Explain how Ice Ages can Result from Global Warming
by Graeme Harrison BE(Syd), BSc(VUW), MEngSc(inc-UNSW),
MBA(Harvard), FAIM, MIEAust, MNIA, MIEEE,
The biggest objection to acceptance of anthropogenic global warming (AGW) by skeptics has been the evidence that, at prior times in the earth's history, when carbon dioxide levels in the atmosphere have been high, what has followed such events have been long periods of very low temperatures ('ice ages') - the very antithesis of the sweltering temperatures that are predicted by our current climate models.
This enigma has baffled everyone - on both sides of the climate debate. Importantly, no other source (that I have seen) has answered the problem, providing a credible explanation as to how our current climate models can be so wrong. This enigma has been 'the elephant in the room' (the large problem, obvious to all, that no-one talks about), in the areas of climate science and environmental policy.
The 'do nothing' proponents claim that either:
(a) the world may be heating up, but that this is not man-made in nature, being simply a natural cycle the earth goes through; or
(b) we could undertake expensive/risky geo-interventions, but in doing so, we may actually make the problem worse (if we seek to cool the planet, but the risk is that the planet will cool significantly by itself).
The 'do something' majority believe that we need to act to at least reduce our profligate use of fossil fuels, as such burning of fossil fuels in such a short period is unprecedented in earth's history, and must cause climate change (one way or the other). In just a few hundred years of industrialisation, we have managed to extract a good fraction of all major reserves of hydrocarbons (laid down over hundreds of millions of years) burning it, turning it into carbon dioxide in the atmosphere.
With no valid explanation for the enigma, the debate necessarily centres on the 'Precautionary Principle', arguing that, as earth is the only biosphere that supports life as we know it, we should not be conducting experiments upon it, so it is better to immediately minimise man's impact, given we are so ill-informed of the potential consequences.
My entirely novel theory, not triggered by the reading or hearing of theories of any other person or body, is that there is an entirely valid explanation for the enigma... and indeed why our current climate models fail in this important respect.
FAILURES IN CURRENT MODELLING
The problem with our current climate models is that modelling the whole world's climate, including seasonal atmospheric and oceanic currents is a massive exercise. So some things have to be assumed to remain 'fixed' in terms of developing such models. The earth's surface is just one of those things which is assumed to be fixed for the models. So, as the carbon dioxide levels climb from pre-Industrial levels in the model (due to fossil fuel use), and other greenhouse gas emissions increase (methane from grazing animals etc), the well-known 'greenhouse effect' is predicted by most models. This means that more of earth's heat is trapped, and does not radiate out into space. This in turn (like a blanket) keeps heat in the atmosphere, and the models show higher average temperatures into the future (as well as more extreme weather events). The most worrying feedback loops appear to be positive (reinforcing). The higher temperatures will lead to a loss of ice-cover in the higher latitudes, and this in turn will lessen the amount of the sun's rays reflected back into space by such ice (lower albedo). Hence the models suggest we will probably pass 'tipping points' where we could get 'runaway' (or at least 'hard to reverse') temperature increases - where the planet's systems take over, and even if we cut-back our greenhouse emissions, it may be impossible to restore the earth to a reasonable temperature range..
All of this is perfectly valid, given the assumptions. So how do we explain the models' contradiction with the geological record? Many claim there are just "natural cycles" that we can't change. Some claim that the climate is inherently unstable, and hence we can't predict what will happen. But nowhere else in science do we accept that a large system should move in one direction (global temperature increases), yet it may turn out to go significantly in the other direction (extensive glaciation - going from 'greenhouse' to 'icehouse'). In any other area of science, we would demand a mechanism which explained how it was possible to not just mis-estimate the amount of change, but how it was possible to predict the wrong direction of change (warming not cooling)!
The problem with our current climate models is that they model ONLY the climate (atmosphere and oceans) assuming the earth's solid bits will remain precisely as they were. But will they? This assumption runs entirely contrary to our existing knowledge of the earth, and the other planets. NASA's 'Voyager' mission proved that there are planets and moons much farther from the sun (well out in the dim parts of our solar system) which have geysers, radiate heat into a cooler space, and generally still show signs of being 'geologically active'. We know the earth is geologically active as well, from volcanic activity, earthquakes and prior evidence of tectonic plate movement (shifting of the continents with respect to each other over geological periods). Indeed, an interesting analogy (used by others) to consider is that the earth's crust is like 'a postage stamp stuck to a soccer ball', in that the crust averages around half a percent of the radius of the earth. [The crust is only 5-10km thick under the oceans, and 30-50km thick under the continents.] As much as we like to think we are standing on "solid ground" or "terra firma", most of our 'rock in space' is actually a deformable ball (molten core surrounded by a malleable mantle), held in a spherical shape by its own gravity force, slowly emitting heat out into space from its interior. The crust looks solid to us (as we chisel granite so slowly), but to geo-scale forces, the solidity of the crust is easily overcome, as when massive meteors hit, or when an earthquake adjusts a plate boundary by a few metres, or when a volcano's magma chamber has a little too much pressure to be withheld.
The crust should only be a 'fixed' assumption for any climate model in the short term (geologically speaking). How arrogant of us humans to assume, because the crust movements seem slow in our short lifetimes, that crust movements will continue to be slow, or will not be affected by the scale of climate changes we may trigger? If as predicted, we heat the overall atmosphere by three or five degrees, why would we assume this will heat only the 20-40km column of air above the land, and not also increase the temperature of the land beneath the first millimetre? The UN Intergovernmental Panel on Climate Change (IPCC) acknowledges that all of its models stop right at the surface of the solid bits of our planet (http://www.ipcc.ch/pdf/technical-papers/paper-II-en.pdf). Clearly a warmer atmosphere will lead to warmer surface temperatures, and (over considerable time) a generally warmer crust. And, in the case of the oceans, if we have a 10km-thick column of water sitting above a portion of crust only 5km-thick, and if the ocean is heated, why would we not expect that warmer water to transfer heat to that thin crust under that section of ocean? Basic physics would dictate that such heat transfer will occur, but the IPCC models do not allow for it.
If we are to incorporate how the crust behaves into our climate models, we need to consider how the crust works at present. It looks quite stable, but it is only a hard "skin" on a plastic (deformable) mantle sitting atop the molten core. The thin skin on any such object is clearly highly dependent upon the temperature and volume of the contents within the skin. The skin does not in any way 'encase' (provide the forces to physically contain) the molten/plastic sphere, but rather the crust is just the cooled surface of the hotter deeper layers. It is best to consider the tectonic plates (that comprise the crust) as being in a state of 'dynamic equilibrium'. A simple example of a dynamic equilibrium is a boat anchored in a smooth-flowing river. While everything remains as it was, the boat appears almost motionless, as the forces are in balance. But if one changed the rate of flow of the river, or the elasticity of the anchor line, one must expect the boat's position to change in response. With earth's crust, while things remain as they have been, the non-locked continents move at only decimetres/year with respect to each other, and new crust forms where molten lava is exposed, and at the other end of such movement, crust is subducted under other crust, where it 'doubles up' (eg along earthquake fault line subduction zones). Importantly the tectonic plates have a 'momentum' with respect to each other, where the velocity is low, but the mass is huge. The rate of movement is limited by the 'drag' (resistance) at the fault lines. But if one changed the drag co-efficient of the fault lines, then clearly the rate of tectonic movement would change.
The largest tectonic movements are along the 'Ring of Fire' around the Pacific Rim, from New Zealand, to Indonesia, Philippines, Japan, Alaska, California, Mexico, down to Chile. A single major volcanic event (eg Mount Pinatubo in the Philippines in 1991) can throw so much ash high into the atmosphere that it will cause the planet to cool (half a degree for a number of years in the case of Pinatubo per http://en.wikipedia.org/wiki/Timetable_of_major_worldwide_volcanic_eruptions). Indeed, scientists could not explain the earth's annual temperature data for 1950-2010 until they took into account the cooling effect of putting ash clouds high into the upper atmosphere (as the ash reflects the sun's rays). The 'normalising' of the earth's temperature timeline (to remove the effects of specific major volcanic eruptions) was required to finally determine that, but for this cooling effect from major known eruptions, the earth's atmosphere had indeed been warming quite consistently over the period 1950-2010. The apparently anomalous ups and downs in the raw temperature timeline could be explained (to a significant extent) only by taking into account the effects of known eruptions. Indeed, volcanic ash plumes are the most reliable way to cool the earth.
Most volcanic activity is presently sub-ocean, due to the slow pulling apart of the tectonic plates causing the crust to be thin under the oceans (except for mid-ocean ridges). There seems to be a melting of the mantle at that point where the crust thins (http://en.wikipedia.org/wiki/Volcano). So, there is persistent tectonic and volcanic activity happening on the earth, but in places not easily observed by most of us.
A NOVEL HYPOTHESIS
The primary thrust of my novel contribution is that any heating of the atmosphere must affect crust temperature, and this in turn must upset the dynamic equilibrium of tectonic movement and hence the overall level of volcanism. Arguably, one can't reliably say whether a heating of the crust would cause an increase or a decrease in the earth's overall level of volcanic activity. When one disturbs a complex dynamic equilibrium, all one can say with complete certainty is that the rate of a given (dependent) activity will change. However, if the overall heat retention by our molten/malleable ball is increased, and it leads to an increase in temperature of the ball within, or even just its outermost layers, then basic physics tells us that the sphere's outer surface must expand (due to the positive heat-expansion coefficient of rock - almost all natural materials expand with higher temperature). And if the ball expands, its solid skin must crack, exposing more molten material to the cooler surface temperature, thereby creating the additional skin (crust) surface area needed to cover the expanded sphere. This is what we would observe as large-scale volcanic activity on earth, associated with a far higher rate of earthquakes. So it is reasonable to assume that any heat transferred from warmer oceans and warmer air to the crust, or any general increase in heat-trapping caused by a 'greenhouse effect' atmosphere, will likely cause additional (higher rates) of volcanic activity on earth, resulting from that temperature perturbation.
Even small additional amounts of volcanic activity on earth would be devastating for life as we know it. Volcanoes are not always limited to the smaller ones we've seen in recent times. In historical times, multiple years of widespread crop failures and cooler temperatures have been observed after medium eruptions. And we have not seen any truly-large supervolcanoes (http://en.wikipedia.org/wiki/Supervolcano) in action since homo sapiens evolved. The size of the caldera of the supervolcanoes at Yellowstone, Java, New Zealand and elsewhere would bring cataclysm if they were to be 'awoken' by a slight adjustment of the pressures (tension) within the earth's crust.
My theory relies upon the existing known connection between major ash plumes causing cooling, but can for the first time explain how HEATING of the planet can cause major cooling events (solving the enigma). It is through the conduction of the heat into the crust, and the required 'adjustment' of the crust, as the heat transmissions cause a minute expansion of the earth's skin. A further related effect is that even localised heating of major points of traction on fault lines (already under severe pressure) could also cause these faults to 'give' at a faster rate - meaning that even localised heat (not heating of whole outer portion of our sphere) could lessen the existing friction that currently binds blocks of a fault line locked against each other, to limit the rate of tectonic plate movement. However, the biggest effect would be that, if the skin is pulled tighter over an expanded sphere, clearly the pressures at the current sticking points will be reduced, and that pressure release will act like a 'lubrication' of the tectonic joints, allowing more earthquakes and more volcanic activity.
There is also a second independent source of 'crustal pressure changes' likely to lead to actual adjustment of the crust. This is the change in crust downward pressure caused by specific elements of climate change. If the Greenland ice-sheet melts, this will represent a great 'lessening of the downward load' on the crust under Greenland. Antarctica is a larger example. Even Iceland (already on a fault line) will be affected. We know that the crust is thicker under mountain ranges, and thinnest under deep oceans. Adjusting the weight above should trigger changes in tension within the crust to rebalance the loads, given the tectonic plates are 'floating' with adjustments possible over time. Lessening of ice-loads should cause upwards crustal movements. And we know from the Rim of Fire that volcanic activity goes hand-in-hand with crust changes. So, even if overall crustal heat expansion does not cause increased volcanism, changing of ice-loads on portions of the existing crust should cause its own increased volcanism.
Others have already shown a correlation between periods of major volcanic activity and mass-extinction events (well-summarised at http://en.wikipedia.org/wiki/Timetable_of_major_worldwide_volcanic_eruptions). A rise in volcanism is proposed as a 'cause' for such mass-extinctions. However, other geo-scale events, such as meteor strikes are also closely associated with the major mass-extinction events. Once one accepts that the crust is a highly-sensitive 'dynamic equilibrium', it becomes possible to perceive high rates of volcanism as a consequence of such other events, like a major meteor strike. So a major meteor strike may have put significant material into the atmosphere by virtue of the strike itself (a nuclear winter for a year or two), but it may also have triggered a series of pressure waves pulsing around the crust, like mega-earthquakes, which could have caused dramatic fluctuations in fault line crust tensions, thereby triggering significant changes in the tectonic plate dynamic equilibrium, thereby causing a prolonged period of increased volcanism, before slowly returning to a new equilibrium. This prolonged period of increased volcanism could have had significantly greater impact on the earth's climate that the initial meteor strike. Eventually the dynamic equilibrium will settle again, perhaps at a changed rate of tectonic plate movement, or with a new set of continents shifting towards or away from each other, like icebergs floating around bumping into each other, after perturbation by a series of waves. At the extreme end of the scale, increased volcanism (for even a short period) could put so much ash into the atmosphere, as to cause an ice-age. And we already know that ice-ages can then be self-perpetuating, as the extensive ice-cover reflects so much of the sun's rays, that an ice-age can last only a few millennia, or much much longer. Perhaps it is the gradual re-establishment of the normal level of 'leakage' of heat from the molten core which slowly changes the dynamic equilibrium back to a warmer climate, over a much longer period than it took to start the ice-age.
EVIDENCE IN SUPPORT
There is no explicit "proof" of this theory, as it is only surmising what might happen in the future, as we add a perturbation to a historically-stable 'dynamic equilibrium'. However, it is based on integrating known tectonic plate theory into climate model theory. And importantly, it does provide a credible explanation of the known enigma in the earth's geological history - that high rates of atmospheric carbon dioxide immediately pre-dated ice ages, whereas our current climate models uniformly predict the exact opposite - only a warming of the earth's atmosphere ad infinitum.
Some forms of evidence that supports the hypothesis include:
(a) The IPCC model assumptions assume the earth remains precisely the same sub-surface, even as huge above-surface climate change occurs - yet we know from the basics of heat transfer, that heat will be transferred into the crust over time.
(b) Also implicit in the IPCC models is that the rate of volcanism from all possible sources remains precisely as it has been annually, while the climate changes significantly - yet even the loss of ice-loads could cause an increase in volcanism.
(c) Any amount of heating of long solids (even just at the very surface), must result in expansion. With higher temperatures, the crust ought behave somewhat like railway tracks buckling due to expansion in a heatwave. If we change the existing conditions (temperature, etc) in any way, the physical properties of the solid materials must change. If we make all tectonic plates some metres longer, it must affect the fault-lines where they join.
(d).All tectonic plate changes must result in a change to the dynamic equilibrium. Historically plate changes are associated with volcanic activity (eg edges of the Pacific plate causing Pacific Rim volcanoes).
(e) Some may argue that any sub-surface changes in the earth must take eons, but the counter-proof is that the geological record indicates that the magnetic field of the earth has reversed numerous times, and the momentous re-alignment took only a few years. Similarly, ice-ages seem to have commenced relatively quickly, compared to the timeframe required for their expiration.
(f) In assuming something is fixed in a model, it needs to move at two or more orders of magnitude more slowly than the activity under consideration. So if one was modelling two cars crashing at a road intersection, it is perfectly fine to assume that the earth's surface is not moving while calculating the relative velocities of the two masses that collided. However, the same assumption cannot be used for climate modelling. The climate normally takes thousands of years or tens of thousands of years for significant change, with only the smallest change measurable in a century. Similarly, tectonic plate movements take tens of thousands of years or hundreds of thousands of years for significant change, with only small changes apparent over centuries. In fact, both weather and crust activity can cause localised violent activity in just minutes (major hurricanes/storms and earthquakes/eruptions respectively), but such events are separate to measuring significant change in the overall system. However, the point is that both climate and crust can change at a somewhat similar rate (and may indeed be linked as hypothesised here). Though it is highly unusual for climate to be changed in only a few centuries (as we are now doing), perhaps the crust may also change in an equivalently short timeframe, after some lag. The point is that the crust cannot be assumed to remain a fixed/extraneous item, while we model significant climate change going forward a century or two.
(g) Volcanism is the major known earth-cooling effect. Arguably it is the only one with the capability to reverse a significant warming turning such warming into an ice-age, as the other potential cooling effects are minor in comparison, or very long term. [Every few hundred million years a major meteor strike has a somewhat similar effect as intensive volcanic activity, but as this is a truly random external effect, we can leave it aside for the purpose of climate modelling.] Accordingly, volcanism must be the prime suspect in terms of how to explain the geological record's history of high carbon dioxide periods resulting in global cooling.
Further proof of this thesis could be provided by volcanologists seeking to obtain data on a global scale as to any change in overall volcanic activity since before the Industrial Revolution. Most volcanologists study a particular volcano, or class of volcano, in the hope of developing methods to better predict eruptions. But perhaps individual eruptions are not what we need to fear! Perhaps more volcanologists need to work on world-scale predictive measures, rather than worrying about particular volcanoes. We need to track overall rates of volcanism by decade, in much the same way as we now actively track changes in global atmospheric and ocean temperatures by decade. However, as there is likely to be a very significant lag time between changes in atmospheric temperatures, and resultant changes in crust temperature and ice-loads - and hence crust tensions, it may well be that no measurable change in volcanism is to be expected yet - so soon after we have started our surface warming activities. However, if crust tension changes and levels of volcanism are to be expected (upon further proving of this theory), then even the potential that this theory could be true, is one further reason to attempt to minimise man's effects on our biosphere, before any flow-on effects can't be halted.
The major implication of this new thesis is that the current climate models need to be expanded (if practical) to take into account crust changes resulting from atmospheric and oceanic model changes. This may prove simply too difficult, as there are already many uncertainties inherent in the climate models, and there are many more uncertainties in modelling how the 'rocky bits' of earth might respond to such temperature changes of the fluids (air and water) on the surface, and changed ice-loads. It may be sufficient to simply 'explain' as this paper does, just how an initial heating of the atmosphere could lead to an ice-age, over an uncertain subsequent timeframe. Such an understanding will remove the doubts raised by those who cite the prior geological record as being in direct conflict with the projected results of all of the IPCC's current models of the future climate.
Tectonic plate disturbance (upsetting the dynamic equilibrium) and a consequent rise in overall volcanic activity, should hitherto be considered as another (separate) 'tipping point' in climate change theory. We've long accepted that volcanic activity influences climate, and now we need to see the interaction as two-way: that climate changes can influence the overall level of volcanic activity. As with most areas of science, when we look closer, everything influences everything else... even if it takes us some time to understand the linkage mechanism.
While the above seeks to add a new understanding to the very complex issue of anthropogenic input to earth's climate, it should not be cited as supporting a 'denialist' stance. It may be that geo-forces take over and can turn a climate heating event into an ice age, but that is no justification for trying to generate such a reversal scenario. The 'Precautionary Principle' still over-rides, even if the threat is that the world may lose its crops due to ice-cover, rather than extremely hot weather. Whether we swelter or freeze (or as this theory suggests, we 'swelter for a relatively shorter time then freeze for a very long time'), the issue remains that climate is something which, though naturally volatile, ought not be 'played with'.
For clarity, it would be helpful to refer to the concepts above as Harrison's "Theory of Tectonic Expansion", as the general public would understand this as an extension from "Global (atmospheric) Warming" as this makes it clear that the warming may also affect the earth's crust. A fuller explanatory summary is "Harrison's theory that global warming will cause heat expansion of the earth's crust, and the changes to tectonic plate fault-line tensions will likely cause increased volcanic activity, with the increased ash then ejected into the upper atmosphere potentially triggering an extended ice-age".
The above paper was written based solely on my own original thoughts over 2006-9 on how to address the enigma. In fact I did not want to read any other person's potential discussion of the topics, until my own explanation was fully documented, lest my document become contaminated by thoughts picked up from reading the ideas of others. However, having completed the full documentation in the above paper in 2010, I have since performed a survey of what is available on the internet, dealing with global warming, volcanic activity, glaciation and ice-ages. There is an excellent article by Paul Link on "Icehouse (Cold) Climates" available at:
That 2009 article deals with the differences between warm periods in the earth's history ('greenhouse' conditions) compared to cold periods in the earth's history ('icehouse' conditions). It presents evidence that icehouse periods were associated with much lower levels of volcanic activity than 'greenhouse' periods. The article does not offer any hypothesis as to how the earth might change from a greenhouse state to an icehouse one. That article does suggest "tectonic factors" play a role, but uses such term to refer to only whether super-continents blocked a free flow of ocean water from equatorial regions to polar regions, compared to situations as at present, where the alignment of the continents allows North-South oceanic thermal currents - warm water near the equator flowing towards the poles, with cooler water returning via a deep current.
The evidence that the overall level of volcanic activity is much lower when the crust is subsequently chilled supports my hypothesis: that warming causes higher volcanic activity, causing large ash plumes which in turn cool the earth, causing the change from greenhouse to icehouse. As an icehouse period is self-sustaining for a long period (as the extensive ice cover reflects more of the sun's heat back into space), the fact that volcanic activity is then reduced is consistent with my hypothesis - volcanism should reduce as a result of the crust then being chilled (contracting with lower temperature). Indeed, the isotopic evidence (presented in that paper) links changes in volcanic activity to changes in climate, and this clearly supports my hypothesis. Arguably it is this proven linkage between volcanism and 'climate states' that is the ultimate proof that the UN IPCC is wrong in assuming that the crust remains static during a climate change process.
The crust and climate are inextricably linked, but not in the way Paul Link proposed in that 2009 paper. Link suggests "Icehouse climates are favored during times of relatively slow sea-floor spreading" whereas I propose that sea-floor spreading is slowed by the icehouse conditions. In general, Link and other recent papers he cites perceive crustal changes as purely 'extraneous inputs' to climate, whereas my hypothesis proposes a dual-direction linkage between them - that geology affects climate (well-known for decades) AND that climate affects geology (my new hypothesis). Link states "A decrease in worldwide volcanic activity [during icehouse periods], with reduction in CO2 emission, is controlled by the rate of sea-floor spreading. Such a decrease is favorable for a long-term icehouse event." Link notes that "The presence or absence of deterministic chronotectonic supercontinent cycles is a recurring point of discussion in the study of Earth history." My hypothesis allows acknowledgement of the effect of large supercontinents and consequent blocking of longitudinal ocean currents as factors, but suggests that these are less important than the greenhouse state triggering crustal activity, which in turn increases volcanism, and that a high level of volcanic ash plumes then cause an icehouse, which in turn slows down (through crust shrinking) the rate of sea-floor spreading and resultant volcanism, to allow an icehouse to persist for long periods.
In his summary, Link states "CO2 cycling through the biosphere is internal to the climate system, and thus cannot drive a secular change through time, but does affect feedbacks (see Carbon cycle). Net volcanism, related to sea-floor spreading rate, is an external forcing factor to the climate system, and had a major effect on the levels of CO2 during the Mesozoic Greenhouse period (Kump, 2002)." I disagree with Link's view that CO2 cycling is "internal to the climate system", as I propose that the resultant heating does change the crust and hence volcanism. Accordingly, I also disagree with his opinion that net volcanism is "an external forcing factor", as my hypothesis proposes that volcanism is an intricate part of the cycle. And Link's own data shows that volcanism cycles with climate - though Link suggests volcanism is an extraneous input, while I suggest volcanism is a negative (indeed reversing) feedback mechanism - affected by climate and influencing climate, and is thus not 'external'. In his closing sentence, Link claims "Even though the burning of fossil fuels has caused atmospheric CO2 concentration to rise steadily since observations on Mauna Kea began in 1958 (Chen and Drake, 1986), changes in climate owing to human activity are geologically ephemeral. In taking a long-time view of the Earth’s history, they represent but a quick excursion." That conclusion might be argued only IF you accept Link's theory (the existing orthodoxy) that geological factors are 'external' to climate. Under my hypothesis, humans' rapid increase in CO2 (and other greenhouse gases) may trigger geological changes in a relatively short period (given CO2 levels have never increased at this rate previously) - until a tipping-point is reached whereby crustal pressure changes may cause increased volcanism and hence trigger a prolonged icehouse event. In this context, we should not take comfort from the historically long periods for pre-human climate changes to occur, as we may trigger the necessary conditions for a tipping-point far sooner, with our current rate of fossil fuel burning. Moreover, Link's use of the word "excursion" implies a return after the perturbation, whereas with my hypothesis, the return could take 10,000 to 100,000 years (ie after a long icehouse period).
My theory is not based on some 'Mother Earth' new-age/alternative approach - that 'she' would have restorative measures - and that we humans, as the 'virus' (causing damage to the planet), will be eradicated by some retaliative action by the earth. But, interestingly, if we do trigger additional volcanic activity, the outcome could seem like such a scenario coming true, though I do not ascribe such anthropomorphic attributes (retribution) to the planet. My proposing a 'two-way linkage' between climate and crust is however, the type of 'everything is linked' approach that would be welcomed by new-age thinkers. Paul Link noted "The Earth’s temperature has remained relatively constant for 3.8 by [billion years], within a range where life could exist, even though solar luminosity has increased and atmospheric CO2 levels have steadily decreased, since Archean time. In the Phanerozoic, the atmospheric CO2 concentration has varied drastically between icehouse and greenhouse times". Link claims that "These concentrations are buffered by feedback loops involving water vapor within the hydrosphere and complex relations in the biosphere", I propose that volcanism provides an even stronger negative feedback to 'bound' such temperature deviations. Indeed having a further 'temperature range control process' in the form I have proposed (that when the planet might otherwise get hotter and hotter, instead increased volcanism cools it down) would explain the unusual stability in the earth's temperature record over these 3.8 billions of years. This 'volcanism cooling response to hot climates' (caused by having only a thin crust over a geologically-active interior) may indeed turn out to be another necessary attribute when we are looking for earth-like planets that could have had a prolonged period of evolution in a water-as-liquid environment ('Goldilocks' or habitable zone), to allow long-term evolution, with regular fluctuations (but only moderate ones) which seems to be needed to develop truly advanced intelligence.
Separately, a former Harvard colleague, after reading this paper, asked "Does the model [ie IPCC with my revisions] say that we should try to prevent emissions, etc, or should we not care and believe everything will correct itself over time?" In reply I noted "I'm saying, it won't keep warming... but that the forecasts should be even scarier. The US needs to think of the potential for a return of a 1km thick ice-cover over much of the continent, before deciding that measures to get our economy off carbon are a waste of time. So, yes, even though there is a likely 'boomerang effect' (you heat things up till you get an ice age), I don't think the prospect of such reversal is a reason to 'do nothing'. Indeed on the way back down, with large plumes of ash from new volcanoes, we may pass 20 degrees C again very quickly on our way to Minus 40C or cooler."
In summary, I believe it would be helpful if future IPCC predictions went beyond a likely temperature increase to the year 2100 and also noted that, sometime after that we may also affect tectonic plate fault-lines and overall levels of volcanic activity, and could trigger a prolonged ice age. That would get everyone back "singing from the same song sheet", in that the paleo-climatologists would then agree with the IPCC forecasts, suggesting that what may follow is not a long-term warming, but something in the opposite direction. The IPCC assumption that the net rate of volcanic activity will remain the same (as we heat the earth up) was always a nonsense, and should now be dropped.
The critical work now required is to analyse the historical paleo-climate relationship with volcanic activity (eg ash layers etc) on a more granular basis. Link noted that volcanic activity was on average lower during icehouse periods than on average during greenhouse periods. If, upon closer examination, it was found that volcanic activity climbed towards the end of a greenhouse period, and/or at the very commencement of an icehouse period, that would provide the necessary proof of my hypothesis. Of course, this will not be trivial, as it will be necessary to separate out meteor-strike events from earth-bound increases in volcanic activity, to seek such correlation.
March 2011: The above paper was published prior to the Christchurch NZ 7.1 earthquake of February 2011 and the Japan 8.8 earthquake of March 2011. A number of reviewers have since commented that perhaps these more recent major seismic events are evidence that AGW is already having an effect on releasing crustal pressures on fault-lines. Of course the higher incidence of late does make one wonder. However, that question lacks sufficient data to answer - just as the recent increase in hurricane activity/severity in Queensland Australia makes one wonder if the force of ocean storms is increasing as a result of AGW. We may need decades of such data to have statistical reliability on any such answer. However, the 'precautionary principle' should still prevail, even if we can't answer the question definitively. However, the disaster in Japan shows how even relatively minor adjustment of crustal forces can be devastating to our civilisation.
In response to the issue of earthquakes (which I am suggesting is not as bad an effect as the likely volcanic activity to follow), I have done further searching for earlier authors on the issue of earthquakes and AGW. It seems that following the 9.1 Boxing Day 2004 Indian Ocean tsunami a Letter to 'New Scientist' queried if the tsunami was related to global warming. The thought was generally lambasted by other writers.
Other searches within the New Scientist proprietary database yielded extracts on some other interesting articles on related topics, which I am happy to recognise:
(a) New Scientist on climate link to earthquakes: http://www.newscientist.com/article/mg20327273.800-climate-change-may-trigger-earthquakes-and-volcanoes.html with link back to http://www.newscientist.com./article/mg19025531.300-climate-change-tearing-the-earth-apart.html
(b) New Scientist on ice-lead causing volcanism: http://www.newscientist.com/article/dn13583-melting-ice-caps-may-trigger-more-volcanic-eruptions.html
(c) New Scientist on prior period of high CO2 and high volcanism: http://www.newscientist.com/article/dn11726-did-the-north-atlantics-birth-warm-the-world.html
The last one is of great interest in that it proposed the opposite to my hypothesis - that a period of significant tectonic plate movement may have heated the earth. Of course this opens the question of whether gradual sea-floor spreading and the associated magma exposure might heat the earth, without the effect of the large ash plumes delivered into the higher atmosphere (which cools the planet). I suggest that very gradual sea-floor spreading may well lead to magma becoming exposed on the ocean floor, and that this may chill quickly without large explosive venting of ash. And such gas (not steam) as is emitted may well become absorbed in the ocean, leading to some acidification of the ocean, but with little effect on levels of particulates in the higher levels of the atmosphere, or the albedo of the earth's surface. So gradual sea-floor spreading may have looked more like the images we see of Hawaii's big island having lava slide into the ocean, giving off only steam, as compared to Mt Pinitubo's explosive eruption.of ash particles into the upper atmosphere.
By May 2011, I thought I should provide further explanation to two elements not adequately covered in the original paper above:
(a) Cycles without explanation - Some people say "But others have postulated for a long time that there are longer-term cycles in the earth's climate". Of course longer-term cycles could arise from changes within the sun, within the earth, or within the earth's atmosphere. However, proposing a cycle without any plausible explanation of the science behind such cycle is not equivalent to my paper which provides a 'testable hypothesis' for the greenhouse->icehouse state-change trigger. The best documented cycles are Milanovitch cycles (http://en.wikipedia.org/wiki/Milankovitch_cycles) which are the tidal effect of sun and moon pulling on the non-spheroidal parts (equatorial bulge) of the earth; the effect of Jupiter and Saturn's gravity causing slight adjustments to earth's orbit; and the slow oscillations in the earth's axis of rotation changing the amount of summer sun at high latitudes. However, what is amazing is the extent to which the 100ky and 400ky cycles explain variations in levels of sunlight at higher latitudes but do not explain state-changes such as a transition from greenhouse to icehouse conditions. Some other tipping point is indicated, and crustal changes due to thermal expansion (ie my hypothesis) could provide that explanation.
(b) Reverse Heat Flows - No-one has raised this as an objection to me, but with further pondering, I had to challenge whether surface warming could cause an effect at any deeper/hotter level of our planet. The rationale is that the earth's molten core is losing heat out through the mantle and then finally through the crust into space, so the heat movement (decreasing temperature gradient) is always from inner to outer layers. So could any moderate change of the temperature of the outermost layer run 'contrary to the direction of heat flow' to affect the temperature of any deeper layer? As an example, consider the temperature gradient in the wall of a simple dwelling as an example. Let's assume a one-room dwelling has walls, floor and roof all made of 100mm thick (circa 4-inch) polystyrene solid foam (ie 'coolroom' insulation). As the wall is of uniform density, the temperature gradient will be smooth across the whole width. So, if on a winter's day the outside temperature is zero degrees C and the inside is 20 degrees C (room temperature), then if you stuck a temperature sensor 10mm into the wall from the outside, the temperature would be 2 degrees C (one-tenth of the way from outside to inside temperature). With a warmer outside temperature, of 15 degrees C, and still 20 degrees C inside, that probe at 10mm depth into the wall would register 15.5 degrees C (again one-tenth of way from outside to inside temperature). So, even though the heat-flow is always from the hot interior of the wall to the cooler outside of the wall, a warmer outside does result in the temperature of the outermost layer being higher. This is analogous to the situation I postulated in the above paper that a warmer atmosphere will lead to a warmer crust, even though heat is always flowing outwards from the earth's core.
2012 Update: Prof Paul Link reviewed the above paper and provided two items of feedback. Firstly that, in his opinion, it should be pointed out to the average reader that, in paleo times, icehouse periods followed greenhouse periods but text should be added to stress that such earlier cycles were over incredibly long periods. I agree with this input, and I changed the wording to express this long timeframe. This does not diminish my theory, as in paleo times, the input perturbations were slow/gradual, so the transition between greenhouse and icehouse states were slower. In post-Industrial times, the climate perturbation caused by man's burning of fossil fuels, deforestation etc is major/rapid, and hence one might expect a quicker transition between greenhouse and icehouse in such circumstances. This would not be contradictory to the paleo evidence.
Secondly, Prof Link suggested that my theory would require someone to use advanced physics to model the heat-transfers and mantle activity to take my thesis forward. A number of other leading scientists in the area have (independently) suggested that my thesis would need the physics to be explored to validate my thesis. The problem is that the physics involved ranges from the straight-forward to incredibly complex. The transfer of heat from the atmosphere/ocean to the crust is simple and not disputed. The expansion of solid matter (eg sections of tectonic plates) is also obvious and not-disputed. However, the tectonic plates are driven by eddies within the mantle, and this is 'fluid dynamics'. The issue about fluid dynamics is that the movement of each molecule follows Newtonian physics, but the fluid as a whole acts as a result of billions of billions of such known movements, so the field is complex. The difference is comparable to modelling billiard balls striking each other, versus pouring liquid dye into a glass of water. The outcome of the billiard balls can be simply calculated using Newtonian physics, whereas the path of the dye spread will vary every time it is observed. We do not even know the proportion of the earth's outwards-from-core heat transfer that is achieved through conduction versus convection, making such modelling almost impossible.
Moreover, we have no direct measurements of the mantle eddies, and only very rough estimates of past movements, as seen in tectonic plate history. So my theory simply notes that the forces at the plate boundaries will change, and to understand what catastrophes might result, one does not need to know with any great certainty which plates will react most, which volcanoes will activate, or precisely how much ash might be ejected. Indeed volcanologists cannot predict the ejection timetable of a single volcano with any precision, so it is arguably futile to seek to model what global changes might result from activation of tectonic plate joints. It should be enough for us to treat such potential disasters as a 'call to action' to mitigate our current carbon+heat pollution, to avoid such catastrophes. The downside of such caution is best summarised by the comedic response "What if climate change isn't real and we are making the world a better place for nothing?"
Instead of seeking to model the physics, where too many of the variables have too wide a latitude of possible values, resulting in too many 'degrees of freedom' for any solution, it seemed to me that the answer might lie in the statistics of any possible early warning signs. In the original paper (above) I proposed that there may be some considerable lag between heating of the air+water and that heat being transferred into the crust. At that time, I had assumed that 'crustal effects' may not become apparent for centuries.
However, newer research confirms that earthquake activity has been increasing (indeed accelerating) during the post-Industrial period.
DATES FROM & TO PERIOD # EARTHQUAKES (Mag>6.99)
------------------- --------- ------------------------------
1863 to 1900 incl 38 yrs 12
1901 to 1938 incl 38 yrs 53
1939 to 1976 incl 38 yrs 71
1977 to 2014 incl * 38 yrs 164 (to Mar 2011) predict >190 in total.
The reason only 'major' (Richter Scale of 7 or above) earthquakes are counted is because the ability to detect minor earthquakes has improved over time. But knowledge of 'major' earthquakes is quite complete since about 1880. The source of this schedule is www.earth.webrecs.co.uk and that article claims to have used both all USGS and Wiki listings for major earthquakes. Moreover, the full schedules of such earthquakes are listed on that site, so that the source material can be checked by anyone. So, in summary, the pre-1900 information is not that 'solid' (free of potential measurement errors), but the post-1900 numbers are solid and do show a significant growth in major earthquake number. And even if one queried whether pre-WW2 reporting of all 'major disaster earthquakes' could have missed one in a remote area of the world, even the 1939-onwards figures show significant growth.
Also of note from the same paper is evidence that the eleven year period 1986-96 (inclusive) had only 15 earthquakes of magnitude 7 or greater (per USGS data), yet the next eleven year period 1997-2007 (inclusive) had 99 such earthquakes. The traditional view that earthquakes are individual random events cannot explain such a decadal variation. However, as the Pinatubo volcanic ash event of 1991 cooled the world for only the first half of the 1990s, this significant decadal variation in seismic event frequency may well point to a 'quicker connection' (lesser lag) between global temperature and crustal activity.
Another source expresses the 'major' earthquake data in terms of periodicity:
- From 1 A.D. to 1800 there were approximately 28 major earthquakes recorded in history. This results in an average of one major earthquake approximately every 60 years.
- From 1801-1900 there were approximately 31 earthquakes 7.0 or higher. This results in one major earthquake approximately every 3.2 years.
- From 1901 to 2000 there were 222 major earthquakes 7.0 or higher. This results in an average of one major earthquake every 6 months.
- From 2000 to 2003 there were approximately 59 earthquakes of 7.0 or higher. This results in approximately one major earthquake every 24 days.
- From 2004 to 2007, there were 56 major earthquakes 7.0 or higher. This results in an average of one major earthquake every 25 days.
- In 2008, there were 12 major earthquakes 7.0 or higher. This results in an average of one major earthquake every 30 days.
- In 2009, there were 17 major earthquakes 7.0 or higher. This results in an average of one major earthquake every 20 days.
- In 2010, there were 22 major earthquakes 7.0 or higher. This results in an average of one major earthquake every 15 days.
Again, it should be noted that pre-1900 data is 'subject to' recording errors, but that does not invalidate all of the above data.
Now, the USGS data for magnitude 6 and above does not support the huge disparity in decadal incidence as reported by www.earth.webrecs.co.uk as shown above. However, it does show a near-doubling in 'greater than 6' earthquake activity since the 1980s. The correct way to compare decades is to add the total energy in such bands, so a 8 earthquake is worth ten of Richter 7. The correct 'energy summation' of the USGS data is thus 100x the 8+, plus 10x the 7-to-8, plus the 6-to-7 earthquake count. On this basis the USGS data does show a significant increase in seismic energy over recent decades, though not as high as the earlier-cited web pages. The source material for the USGS counts are at:
So, it would seem that on this small rock floating in space, 'everything IS connected', and my proposal that global warming will result in crustal changes is highly likely. Moreover, the supporting evidence is already available.
By 2011-12 media stories were starting to report crustal changes being likely as a result of global warming. However, in my opinion, these have been 'somewhat conservative' in that most of such reports deal with only vertical adjustments to the crust, that have been occurring and are likely to occur as a result of the loss of glacial ice-cover. I agree that the loss of thick ice-cover will result in an upwards adjustment (rebounding) of the crust at such points, and that the evidence of this happening historically is very strong. In short, the crust 'floats' on the mantle, and any adjustment in weight causes a 'buoyancy' response.. For examples of such articles see
The reason I suggest such reports are conservative is that they ignore the idea of lateral temperature expansion of the crust, and consequent changes in forces at tectonic joints. In my opinion, the risk is global, and it is wrong to see it as affecting only Greenland, Iceland and Antarctica (ie glacial regions). As the overall earthquake data shows, it is more likely that we will see increased crustal activity at all tectonic plate boundaries.
And it needs to be remembered that one does not need the whole 'Ring of Fire' invigorated to make life unpleasant on earth. Just one volcano - the Mt Tambora explosion in Indonesia in 1815 caused 200,000 fatalities in distant Europe during 'the year without a summer' through starvation (http://en.wikipedia.org/wiki/Year_Without_a_Summer). Also, the Toba catastrophe of 70,000 to 75,000 years ago almost wiped out humanity So just one or two major volcanic responses to global warming could be enough to halve earth's population, or even challenge civilisation as we know it.
In the year since this paper was submitted to the UN IPCC, there has been no response. I suspect that the atmospheric scientists involved believe that the geology and mantle physics involved is too far from their field of knowledge to comment. However, I remain of the opinion that the UN IPCC forecasts should note the potential for geological effects in their forecasts for future climate outcomes.
The consequences of my proposal are:
(i) There is a new very dangerous 'tipping point' being 'crustal upset', which is non-reversible in any reasonable time frame;
(ii) The negative impacts of global warming are likely to be far worse than UN IPCC forecasts made thus far;
(iii) The impacts are far far harder to predict than UN IPCC reports (thus far) would lead us to believe;
(iv) There is even greater impetus for the world to address greenhouse gas emissions, deforestation etc;
(v) There is little future for civilisation if we simply 'plan to adapt to climate change', as a 'nuclear winter' from major volcanic eruptions and the possible ice-age to follow, are events beyond any possible adaptation;
(vi) Geo-engineering solutions are far riskier than first thought (as earth may cool not warm);
(vii) Earth-wide seismic and volcanic data need to be accurately measured and tabulated on an annual basis, as temperature data is, to track precisely how we are going on crustal changes.
(viii) The 'carrying capacity' of the earth may be far lower than the earth's present population, even if we shift to a far greater reliance on renewable energy sources;
(ix) The general public needs to be alerted to the "now greatly-increased risk" of the 'do nothing' approach.
ABOUT THE AUTHOR
Graeme Harrison has degrees in a number of disciplines; served as a Harvard Consultant to The White House on IT; was on the team of five (doing Design Review) for the development of Visicalc - the very first electronic spreadsheet (developed at Harvard Business School in 1978-9 - from which both 'Lotus 123' and 'Microsoft Excel' were derived); served as a Deloitte/Touche expert to the US Department of Energy on alternative energies (including large-scale geothermal); co-authored a major study with UC Davis on the future of Californian agribusiness; co-founded Australia's most successful health informatics business; has served on various Australian Government committees, has taught at graduate level, and has been called upon as an independent expert in many large Federal and Supreme Court cases. He studied Earth Sciences at Victoria University of Wellington, New Zealand. He won 'The Thesis Award' at Sydney University where he studied Engineering. He topped his class in Quantitative Methods at Harvard Business School. His first published paper, for the Australian National Power Conference of 1977 was on Stochastic Forecasting Methods. He has 35 years experience in advanced forecasting methodologies.
To contact the author, send an email directly to him at 'prof at-symbol post.harvard.edu', replacing the 'at-symbol' with a "@"
Or you can simply use the form below.
To add a link to this paper, simply use www.warming.weebly.com
(c) Graeme Harrison 2007-12; Reproduction and/or extraction is authorised, provided only that authorship and source url (link) are also disclosed in any copy/extract.
Concept First Divulged: November 2009 (to a number of colleagues for review). Published on internet January 2011. Submitted to the UN IPCC Secretariat on 3 February 2011. Last updated 24 September 2012.