PUBLIC LECTURE
Climate change, human aspiration and the finite capacity of planet Earth
The Shine Dome, Canberra, Wednesday, 19 August 2009
Dr Michael Raupach, FAA
CSIRO Marine and Atmospheric Research
Michael Raupach is a research scientist in CSIRO Marine and Atmospheric Research. He is a Fellow of the Australian Academy of Science and a Fellow of the Australian Academy of Technological Sciences and Engineering. From 2000 to 2008 he was an inaugural co-chair of the Global Carbon Project of the Earth System Science Partnership. His research encompasses global and continental carbon and water cycles, carbon-climate-human interactions, land-air interactions, fluid mechanics and particle transport. He is a frequent contributor to the policy and public debate on climate change.
Click here to view Question time
Click here to view video presentation
Chair (Kurt Lambeck): Well, good evening everybody. I am Kurt Lambeck. I am the President of the Academy of Science. It gives me a great pleasure to welcome you here this evening. It is great to see such a large crowd here, and an enthusiastic crowd.
Welcome to everybody and welcome to the Academy's 2009 National Science Week events.
It now gives me great pleasure to introduce tonight's speaker, Dr Michael Raupach. Mike is a research scientist in CSIRO Marine and Atmospheric Research. He is a fellow, both of the Australian Academy of Science and of the Australian Academy of Technological Sciences and Engineering. This gave me, in fact, great pleasure earlier this year to welcome Mike to this Academy.
The election to both the academies, of course, is testimony to the high quality of his research, but also to the relevance of his research to present day society. His work encompasses global and continental carbon in water cycles, carbon climate, human interactions, land/air interactions, fluid mechanics and particle transport. And in all of these areas he has made important contributions to our understanding.
From 2000 to 2008 he was the inaugural co–chair of the Global Carbon Project of the Earth System Science Partnership. And he and Pep Canadell were joint finalists in the Environmental Research category of the Eureka Prizes. Mike's going to speak to us tonight on the topic of climate change, human aspirations and the finite capacity of planet Earth. I invite you to welcome Mike to address us.
Michael Raupach: Thank you very much, Kurt.
[SLIDE: Climate change, human aspiration and the finite capacity of Planet Earth]
I would like to talk to you tonight about a subject which clearly has many controversies, both scientific and non–scientific. I am going to touch on some of these controversies as we pass through, but the main thing I want to do is try to frame the subject of climate change in the context of the finite capacity of our planet for supporting human societies and aspirations. In the context of climate change, our concern is the capacity of the planet to recycle carbon dioxide (CO2) and other greenhouse gases that we, as human societies, produce in the course of our activities.
This talk owes much to many people. Kurt has already mentioned the Global Carbon Project, and I am deeply indebted to Pep Canadell and many other colleagues in that project and in CSIRO. I would just like to mention particularly John Finnigan and Peter Briggs, who have helped me a great deal in the preparation of this talk and in many other ways.
[SLIDE: Plan]
What I would like to do is to go through the evidence of human induced climate change, looking at this evidence from four points of view. First, the simple physics; second, what we learned from the deep past; third, the recent past; and last, what we learn from climate models.
Then I would like to just touch briefly on climate impacts, which determine what all this means for societies including Australian society. Then I will take a deeper look at atmospheric CO2 (which as you know is the main anthropogenic greenhouse gas) and the human drivers of CO2 emissions. And finally, I’ll look at the question of targets: how we set targets and what risks we face as targets are being set.
[SLIDE: Svante Arrhenius]
Let's begin with the physics. A way to do this is to return to the work of some of the pioneers of this subject, which extends well back into the 19th century. Perhaps the first paper to raise the possibility of anthropogenic or human-induced climate change was by Arrhenius, written in 1896. It was a paper – very widely quoted these days – called On the influence of carbonic acid in the air upon the temperature of the ground. ‘Carbonic acid’ was his name for CO2, carbon dioxide.
The reason he undertook the work, it should be said, was not because he was interested in anthropogenic climate change: it was because he was trying to understand what caused the ice ages, or at that time the ‘ice age’, as only one was known.
He concluded, towards the end of his paper: 'A simple calculation shows that the temperature in the Arctic regions would rise about 8 or 9 degrees, if the carbonic acid [the CO2] increased to 2.5 or three times its present value.'
That result is very close to what we understand now, and is about in accord with current estimates.
Also, in the context of understanding the ‘ice age’, he made a prediction for the concentration of CO2 during glacial periods which was very close to modern estimates. The real answer, as we now know, is about 180 parts per million (ppm) relative to a CO2 concentration at the time Arrhenius was writing of about 280 ppm. The estimate that he obtained was 160 ppm: not bad, given the tools at his disposal.
[SLIDE: Guy Stewart Callendar]
Further very important contributions were made by Guy Stewart Callendar. I would like to thank Frank Bradley for bringing this to my attention and for lending me and subsequently giving me a biography of Callendar, which helped me to understand his work.
Callendar produced a series of papers on climate change beginning in 1938 and continuing through to his death in 1964, He was the first to seriously attribute climate change to human causes, specifically the emission of CO2 and other gases by industrial activities.
He built the first radiative transfer model which formally accounted in much the way that it is now done for both water vapour and carbon dioxide and for multiple pathways of transmission through the atmosphere. He made an estimate of 2 degrees for warming for a doubling of CO2, which is close to, perhaps a little lower, than the estimates that we now make.
He also said: ‘how easy it is to criticise and how difficult to produce constructive theories of climate change!’ And in doing so he outlined a list of the criticisms that had been levelled at his theories in 1938 and through the forties and fifties. It is remarkable how similar that list is to the criticisms that are still being levelled against the theory of anthropogenic climate change today. They included, for instance, temperatures being measured in cities or at airports and not in remote country locations, the possibility that this was an accidental association rather than a causal effect, the possibility that water vapour was completely saturating infrared absorption bands, and so forth. All of these arguments are still around. They have been rebutted many times. But it is interesting how the subject returns to such arguments, despite the rebuttals.
[SLIDE: Greenhouse gases in the Earth System]
A brief summary of the physics is that greenhouse gases in the Earth system act in the following way: the Sun shines solar radiation upon our planet. The planet absorbs most of that radiation and re–emits it as thermal radiation in an almost exact energy balance: the energy coming in very nearly equals the energy going out. And we can use that simple sum to do a calculation of what the Earth's surface temperature would be if the Earth had no atmosphere. The answer is about minus 15 degrees Centigrade, which is clearly quite wrong. The Earth's actual surface temperature, averaged over the whole planet, is roughly plus 15 degrees Centigrade, so there is an extra 30 degrees of warming coming from somewhere.
The extra warmth comes from the greenhouse or radiatively active gases in the atmosphere, including water vapour, carbon dioxide, methane and a number of others. These gases absorb and re–emit heat radiation, which means that as heat radiation passes up through the atmosphere, some of it is absorbed and re–emitted, potentially many times. The result is that the atmosphere acts as a sort of dam, holding back part of the heat radiation returning to outer space. Eventually the dam overflows and all of the radiation returns, but in the process energy is held up in the system and that energy accounts for the surface warming. Thus, we have a sort of thermal blanket which keeps the Earth habitable.
Now, water vapour is the most important greenhouse gas in one sense but not in another. It is the most important gas in the sense that somewhere between 85 and 90 per cent of this trapping of outgoing radiation is due to water vapour. However, it is not the most important greenhouse gas in the sense that water vapour is not the controlling gas. Water vapour is completely tied to the temperature of the Earth. That means that as the Earth warms, more water vapour comes into the atmosphere from evaporation, mainly from the oceans, so that more trapping occurs. This is an example of a ‘reinforcing feedback’, or in technical terms a positive feedback. I will show you a few more examples in a moment.
CO2, on the other hand, is a controlling gas in the sense that CO2 can be altered (for example by human activities) independently of the state of the physical climate system, thereby having a controlling influence on the physical climate system. This contrasts with the situation for water vapour, which is part of the physical climate system.
[SLIDE: Climate over 800,000 years.]
Turning now to the climates of the past: there exists a remarkable record from the Vostok site in Antarctica, which now extends back for over 800,000 years. Using ice cores drilled at that site it is possible to reconstruct from the tiny bubbles of air trapped in the ice the temperature and the concentrations of many gases over the whole of that period. What you see here is the temperature record. In fact, look only at the red line of the two, which is the observed temperature record; we will come back to the other one in a moment.
The observed temperature record shows a couple of very remarkable things. The first is that it is enormously variable. It varies by about 5 degrees, perhaps a little more, between inter-glacials – the state that we are in at the moment – and deep glacial cycles where the temperature of the Earth, planet–wide, was around 5 degrees cooler than it is now. This curve has been corrected, by the way, for the fact that the temperature at the Vostok site itself varies by a good deal more than that.
[SLIDE: Feedbacks in ice–age climate change]
What causes this? A very simple sketch of the current theory is that the major changes are ultimately driven from outside the Earth by orbital fluctuations, that is, by wobbles in the orbit of the Earth around the Sun. There are three important kinds of wobbles, caused by the eccentricity of the Earth's orbit around the Sun, its axial tilt, and precession of the Earth's axis of rotation. These three together produce a variations in the radiation falling on the Earth. The amount of incoming radiation doesn't vary by very much when you average across the Earth as a whole, changing by only tenths of a per cent, but it changes more in its distribution across the hemispheres. In other words, the distribution of radiation between the northern and the southern hemispheres changes far more than the total.
These relatively small orbital fluctuations are not enough to explain that very large temperature change of about 5 degrees that we see between glacial and interglacial periods. What causes that? Well, the temperature change itself produces some further changes. One, obviously, is that there is a change in the amount of ice on the planet. And that means that the albedo, the reflectivity of the planet, changes. When we have more ice, the planet is lighter in colour, and more of the Sun's radiation gets reflected back from the Earth to space.
This provides a reinforcing feedback to a cooling temperature because as ice accumulates more radiation is reflected, less energy is absorbed bv the planet and so the temperature cools further. This is a classic example of a reinforcing feedback.
A second important reinforcing feedback occurs because carbon dioxide is also responsive to temperature. The level of carbon dioxide in the atmosphere depends on the temperature of the Earth, in part through the amount of CO2 that is dissolved in the oceans. This increases as temperature cools, so the CO2 in the atmosphere is pulled down, resulting in less greenhouse effect from the carbon dioxide, and hence further cooling. The combination of these two reinforcing feedbacks, together with a few smaller feedbacks, is sufficient to explain the large difference of 5 degrees between glacial and interglacial periods.
[SLIDE: Climate over 800,000 years]
Returning to the temperature record over the last 800,000 years, let us look at a few other records over the same period. The top graph shows the changes over that period in CO2 and methane (another significant greenhouse gas), and in sea level. Sea level, of course, changes because of the formation and breakup of ice sheets in both hemispheres. Sea level is a proxy for the amount of ice that we have on the planet and, therefore, the strength of the lower reinforcing feedback loop from ice in the previous diagram. The CO2 record is telling us how the upper feedback behaves.
[SLIDE: Climate over 800,000 years (2) (Three graphs)]
Put these together and we obtain a pair of series (in the middle panel of the three shown here) which describe the radiative forcing, the amount by which these processes are changing the Earth's energy balance. Radiative forcing is measured in a quantity called watts per square metre. It is a measure of power. The forcing varies through a range of about 3 or 4 watts per square metre as we go through the glacial cycles.
Just for illustration, the power in the Sun on a bright sunny day is a little more than 1,000 watts per square metre. And one watt per square metre would be something like a couple of radiator bars in this large auditorium.
These two series can be put together to obtain a total radiative forcing, and the Earth's temperature response to this forcing can be calculated. Returning to our temperature record at in the bottom panel, this calculation yields the blue curve. Thus the sum of the ice and CO2 forcings, which are about equal in magnitude, accounts very well for the observed temperature record through the glacial cycles with a ‘climate sensitivity’ of about 3 or 4 degrees per watt per square metre. That's the amount by which the temperature of the Earth changes in response to a change in radiative forcing.
And this is in very good accord with current estimates from computer models, but of course obtained by a completely different route, namely by observation of the deep past.
The argument I've just summarised is from work by James Hansen and colleagues, published last year in a paper which is now receiving a lot of attention, called Target atmospheric CO2. They go on to examine the implications for modern climate change, as we will do in a different way.
[SLIDE: Human–induced climate change]
Now we come to human–induced climate change, and in doing so we will move to the recent past. This simple figure shows the three major factors involved. The first is the driver, consisting of emissions of CO2 and other greenhouse gases by humans (only CO2 emissions are shown here). The second is the atmospheric CO2 concentration, which rises as the atmosphere accumulates some of the CO2. And the third, of course, is the response of surface temperature. These three represent a chain from cause to immediate effect to further effect. Beyond surface temperature, of course, there are yet more effects, but these three are enough for the current discussion.
[SLIDE: Feedbacks in human–induced climate change]
We can now look at this in the context of our earlier diagram of feedbacks in the Earth system. CO2 emissions cause a change in CO2. That, with a climate sensitivity of roughly 3 to 4 watts per square metre, causes a temperature change. The temperature change then initiates similar feedbacks to those that act in ice age cycles, producing changes in ice and albedo and further changes in CO2 which reinforce the original temperature change and create the possibility that temperature change and climate change will be larger than what our 3 to 4 degree per watt per square metre climate sensitivity is telling us. This is one example of what is meant by ‘tipping points’.
There are a couple of other aspects to this argument. The first one is that you may have heard climate-change skeptics say that temperature changes during glacial cycles occur ahead of CO2 changes, which is alleged to prove that temperature drives CO2, not the other way around. The first of those two statements is absolutely true: temperature changes do occur before CO2 changes in the glacial record by roughly 700 years. It is precisely what you would expect from this feedback system because if the temperature is being changed by something external, namely orbital forcing in that case, it takes a while for the CO2 changes to take effect and the feedback system to get going. However, it is not correct to infer that CO2 responds passively and has nothing to do with the temperature, because CO2 influences temperature just as much as temperature influences CO2. Therefore, if the system is turned around and CO2 is forced, we have the opposite situation. That's the one we now have with anthropogenic climate change.
[SLIDE: Temperature 1950–2008]
We come now to the temperature record over the last 60–odd years. The top trace with the different coloured dots – representing different seasons – shows global average temperatures inferred by one research institution, the Goddard Institute for Space Studies. The trace shows an increase of temperature of the order of 0.7 to 0.8 degrees over the last 100 years, if we extend back slightly further than the time range of this graph.
We also see a couple of other interesting things. You see that there are strong associations between high global temperature anomalies and El Nino events [the upper red part of the bottom panel on the slide]. These are warming events for sea surface temperature in the tropical Pacific Ocean. Likewise there are associations between low temperature anomalies and La Nina or cooling events in the tropical Pacific Ocean. Hence there is more going on in this record than the rise of CO2; there is also a great deal of inter-annual variability. While there are many contributors to this inter-annual variability, the temperature of the Pacific Ocean as reflected in the so–called ENSO (El Nino southern oscillation) cycle is one of the main drivers of variability. There are also contributions to variability from other ocean-atmosphere oscillations, volcanic eruptions and more.
You might also note that very recently we have been in a La Nina period. If you follow these things you will know that we are coming out of this La Nina, meaning that there is a strong chance of a return to higher temperature anomalies in the next year or two.
[SLIDE: Ocean heat content: The Earth's thermal flywheel]
We have other measures of the increasing temperature of the Earth. This one is a measure of the ocean heat content, which has been available for only a few years. Its importance is that the oceans are the thermal flywheel of the Earth's system. The temperature in the atmosphere oscillates a great deal, partly for reasons to do with the ENSO and similar cycles, as we said. But the oceans accumulate heat much more slowly, and keep the temperature of the Earth rising in a fairly systematic way. This is a more robust measure of the trend in planetary heat content than the air temperature, and it clearly shows the same sort of rise.
These measurements are made by a large number of probes which are lowered to substantial depths in the ocean and return their data by satellite. Those data give temperature profiles down to many hundreds of metres in the ocean, which can be used to obtain a curve like this.
[SLIDE: Global and continental temperature change]
Now we will come to models. This picture from the IPCC, the Inter–governmental Panel on Climate Change, was published in 2007. It shows the temperature predictions over the last century from a number of different climate models.
There are two curves here for the model predictions. One is in pink. This indicates the predictions of all radiative forcings applied to the models, including both the anthropogenic greenhouse forcing from CO2 and other gases, and also the non-anthropogenic forcing from volcanic eruptions, solar changes and so on. The other blue curve is the model prediction with the non–anthropogenic forcings only. The difference between the two curves is a test of the extent to which the models are distinguishing between forcing the humans are putting into the system and the other forcings such as volcanoes that exist in the system independently of human action.
What we see, of course, is that the observed temperature trends (the black lines) are strongly consistent with the class of predictions which include human actions. This applies globally, for land as a whole, for the ocean as a whole, and for a number of continental regions.
[SLIDE: Emissions and temperatures from 2000 to 2100]
If similar climate models are now used to project forward, we obtain a set of trajectories for average global temperature out to 2100. The temperature origin in these graphs is set at roughly 1990. These graphs show temperature rises of up to 4 degrees for the end of this century, depending on what we assume about anthropogenic emissions. The assumptions for anthropogenic emissions are shown on the left.
[SLIDE: Temperature changes through 21st century]
Of course the temperature trends are not uniform across the Earth as Arrhenius found way back in 1896. The Arctic and Antarctic are likely to warm more, relative to the rest of the globe. This, of course, is one contributor to that reinforcing feedback involving ice. Because there is more heating there, the ice will melt faster.
[SLIDE: Plan]
We come now to a little on climate impacts (it is going to have to be a very little because of time).
[SLIDE: Climate impacts]
There are a number of systems that are affected by climate change. They include our water supplies, our ecosystems, our food systems, coasts through sea level rise and, of course, human health.
[SLIDE: Sea level rise tracking near upper limit of projections]
There are impacts of climate change in all of these areas and there is only time to highlight a couple. The first is sea level rise, which is currently tracking near the upper end of the projections of the Inter-governmental Panel on Climate Change at about 3 millimetres, or slightly more, per year.
[SLIDE: Projected patterns of precipitation changes in 2095]
The second major impact is water and precipitation. Precipitation is notoriously difficult to predict from climate models. The maps in the upper panels show the projections from an ensemble of climate models where the orange colours indicate decreases in precipitation below current mean values, and the blues indicate increases. Overall the globe gets wetter, but there is a belt of orange in the mid–latitudes in both hemispheres demonstrating that in some areas we are going to see decreases in precipitation.
[SLIDE: Soil moisture anomalies for 20–26 April 2009]
We happen to live in one of those areas. This is an example of work from our group which is monitoring soil moisture across Australia (results of this monitoring are available on the web at www.csiro.au/awap/). Shown here is the soil moisture distribution in shallow and deep soil layers, for one week a couple of months ago. We could have picked any week, all the way up to last week. You see on the right that south–eastern Australia has a very severe and ongoing soil moisture deficit in the deep soil layers extending from about 0.2 to 1 metre.
[SLIDE: River Murray flow has declined by over 75% since 2002]
This is reflected in flows in the Murray River. These data, which I find very striking, show the Murray River flow over the last 60 or so years. They are real data, including a few periods with missing data. You see the enormous step change in Murray flow which occurred around 2002, amounting to a decline of roughly 75 per cent. The flow has fallen to a quarter of its previous value.
It's possible to trace the reasons for that change. About a fifth of the decrease is directly due to the fact that the rainfall has decreased. Three–fifths is due to decreases in the fraction of the rainfall which ends up in the river, because the vegetation takes the lion's share of what falls on the land. The remaining fifth is due to a decrease in the fraction of this decreased flow down the river that is left after people extract water from the river.
One result of this decreased flow is what we see in the Coorong and Lower Murray Lakes right now. These beautiful wetlands are in severe ecological stress.
[SLIDE: Subtropical ridge]
The reason for the decreased flow, and the reason why it can be associated with climate change, has been worked out very recently by colleagues of ours in a program called SEACI, the South–East Australian Climate Initiative. It has to do with a persistent weather feature called the subtropical ridge. This is today's weather map. You see the typical ridge of high pressure that exists in the southern part of the continent. We know that high pressure systems tend to push the cold fronts further south. There are a couple of cold fronts coming through today, and the forecast predicts that the one in the centre of the screen is not likely to have much effect for us tomorrow. Part of the reason is the blocking high sitting here in the eastern part of the continent. If these high pressure systems become systematically more intense, the fronts will be pushed further south and we will experience reduced rainfall.
[SLIDE: It is increasingly likely that current drought in Southern Australia is exacerbated by climate change]
The intensity of the high pressure belt is measured by an index called the subtropical ridge index. This measure shows that the subtropical ridge is getting more intense. Here is the subtropical ridge index is plotted in red over the last 100 years together with global annual temperature in black. You see a very strong association. That association can be verified with climate models.
So we have a smoking gun, if you will. There is a link between climate change and the current trends in rainfall in the southern part of Australia and, in particular, in the Murray Darling Basin. Our conclusion is that Australia is subject to climate impacts, and it is perhaps more subject to climate impacts than many other developed nations.
[SLIDE: Plan]
So now we come to atmospheric CO2.
[SLIDE: Atmospheric CO2 from 1000AD to present]
This is the record of atmospheric CO2 over the last 1,000 years. It comes from several sources, including air bubbles from various ice cores, which are in different colours. The blue curve on the right comes from the observations at the Mauna Loa observatory in Hawaii, which has been measuring CO2 daily, with remarkable precision, since 1958.
You see the near-perfect overlap of these curves, obtained without adjustment. It is a testimony to the precision of the independent sets of data that such an overlap can be obtained. I think it is a remarkable achievement.
[SLIDE: Global CO2 budget 1850 – 2007]
What causes CO2 to increase like this? To understand the factors at work we can look at the budget of atmospheric CO2, that is, how much CO2 comes into the atmosphere, how much remains there, and how much goes out.
In the budget for atmospheric CO2 over the last 150 years, the first thing that goes in is the CO2 coming from deforestation. As forests are felled their carbon dioxide is emitted and it accumulates in the atmosphere.
[SLIDE: Global CO2 budget 1850 – 2007 (2nd)]
Next, and much larger, is the contribution from the burning of fossil fuels. The small line at the top is other industrial emissions, from cement production and so forth.
[SLIDE: Global CO2 budget 1850 – 2007 (3rd)]
All of that CO2 somehow has to be accounted for, because this is a budget. So we have to account, on the negative side of the budget, for exactly that amount of CO2 that goes in on the positive side.
[SLIDE: Global CO2 budget 1850 – 2007 (4th)]
Some accumulates in the atmosphere: a little less than a half.
[SLIDE: Global CO2 budget 1850 – 2007 (5th)]
Some goes into the oceans, roughly a quarter;
[SLIDE: Global CO2 budget 1850 – 2007 (6th)]
and some goes into land systems, to make up the other quarter.
[SLIDE: Sinks discount emissions by 55%, but not forever]
Putting all of that together we see trends in the behaviour of the major sinks of CO2 in land and oceans. Over the last 50–odd years, the fraction of all the incoming CO2 remaining in the atmosphere is increasing slightly [the top plot]. The fraction that is going to the ocean [in the bottom plot] is decreasing. The fraction that is going to land systems is roughly steady, to within the scatter of the data. Thus, the combined land and ocean sink is weakening, relative to rapidly rising emissions.
[Accelerating global CO2 emissions]
We now have a closer look at global CO2 emissions from fossil fuels and other industrial processes, which account for about 85 per cent of all CO2 emissions (the other 15 per cent coming from land-use change). At the moment, these emissions [shown in black] are tracking well ahead of most scenario projections published around 2000. The lower graph shows this for the period 1990 to 2010, as a cut-out from the upper graph which runs from 1850 to 2100. An interesting part of the lower graph is a projection [the open circles] for what the global financial crisis is likely to do to emissions. It has sometimes been said that the global financial crisis will cause a decrease in emissions and therefore, at least for a while, our problems will be over. Our calculation is that the decrease will be worth roughly six weeks' of total emissions. In other words, if we shut down the emissions for six weeks and then resumed, that is the same as the two–to–three year effect of the global financial crisis. Alternatively, the decrease is equivalent to three–and–a–half years of growth in emissions, which at the moment are growing at 3.5 per cent per year. This tells us that the global financial crisis is not a long-term solution to the problem of rapidly rising CO2 emissions.
[SLIDE: Drivers of global emissions.]
Let’s look at the drivers of CO2 emissions. This graph characterises the emission drivers by using a very simple piece of algebra which equates emissions to the product of three factors: population, income per person, and emissions per dollar of income. You now start to see why I included the phrase ‘human aspiration’ in the title of this talk, because the second factor, income per person, is a rough economic indicator of the results of human aspiration. World CO2 emissions – the black line – is shown as the product of these three factors; red is population, green is income per person and blue is emissions per dollar of income.
These factors together drive the emissions we are seeing now. In particular, the strong increase in emissions over the last five to seven years is accounted for mainly by an increase in income per person, and to some degree a weakening of the rate of decrease of emissions per dollar of income. Emissions per dollar of income have decreased historically over a century or more, because people and industries naturally become more efficient. This is not a greenhouse mitigation measure: it is rather a natural seeking for efficiency to improve productivity. It is what causes the blue curve to trend downwards in the long-term average. However, in the early 2000s, mainly for reasons associated with rapid economic and technological transitions in China, that decrease had a temporary halt.
[SLIDE: USA, Australia, India and China]
We can do the same breakout for any region or country in the world. You see here the curves for the USA and Australia. I have stretched India and China so that the relativities are the same. You see the enormous growth in emissions [in black] of India and for China, and the slower growth in the USA, and a growth in Australia which is somewhat faster than the USA, You also see the strong role of growth in income per person [the green curve] in causing emissions to accelerate, particularly in China.
[SLIDE: A phase transition in human ecology]
This is a symptom of what could be called a ‘phase transition’ in human ecology. If we look back over the last 2000 years we can examine the Earth's population and its gross world product, from data assembled by Angus Maddison. These together give the global average income per person, which is the green curve in the previous diagrams. Income per person was nearly steady for almost all of that period of 2000 years, but between about 1800 and 1820 it started to accelerate. Since that transition, per capita income has doubled every 45 years. Human ingenuity has given us the ability to do that and, of course, there is a great desire from the point of human aspiration for that to continue.
[SLIDE: Plan]
Finally, we come to targets for future CO2 emissions. What is the implication of these trends for the targets that we need to set to avoid dangerous climate change, and the risks around those targets?
[SLIDE: Relating emissions targets to climate danger: Cap on cumulative emissions]
There has recently been a change in the way that the carbon community is thinking of the problem of defining targets to avoid dangerous climate change. Dangerous climate change is often defined as a change of two degrees relative to preindustrial temperatures, or about 1.3 degrees relative to present temperatures. I will use that definition here, recognising that it is a value judgement rather than a scientifically based figure.
People often talk in terms of emissions reductions, for example ‘we should reduce our emissions by 60 per cent by 2050’, or ‘10 per cent by 2020’. It has recently emerged that a robust way of examining this problem is to look not at the emissions reductions by particular dates but to look instead at the total amount of CO2 that we can emit to the atmosphere before we have to stop.
This translates the target-setting problem from one of characterising the detailed shape of an emissions trajectory to placing a cap on the total that can be emitted, which simply means limiting the area under the curve.
Several emissions trajectories are shown here. Historical emissions are to the left of the vertical dotted line, and of course we can do nothing about them. There have been 550 billion tonnes of carbon emitted as CO2 since the start of the industrial revolution up to last year, including emissions from both fossil fuels and land use change.
A recent estimate in a paper published in Nature by Allen et al was that we can emit about a trillion tonnes of carbon to the atmosphere in total, before we exceed the 2 degree target with 50 per cent probability. Of that 1 trillion tones, about 550 billion, or just over half, have already been emitted.
That implies that we are now at, or just past, ‘peak CO2’. We must regard CO2 emissions, like oil, as a non–renewable resource. The problem of sharing these remaining CO2 emissions is the problem facing us at the December 2009 Climate Conference in Copenhagen.
[SLIDE: Future CO2 and warming pathways]
Let’s look at the implications of placing a cap on cumulative emissions. Here, we have the warming pathways [bottom panel] and the associated CO2 trajectories [top panel] corresponding to the set of capped emissions trajectories in the last slide. These results come from a simple model (it runs on my laptop). The traffic lights on the right indicate how the outcomes of various cumulative amounts of CO2 emitted into the atmosphere from last year onwards translate into climate danger; green indicating relative climate safety, orange indicating concern (around the 2 degree warming mark), and red being territory into which we definitely do not want to go.
[SLIDE: Cumulative emission targets and climate risk]
We can now look at how the cumulative emission, plotted on the horizontal axis, relates to the peak warming that would be brought about if that amount of CO2 is put into the atmosphere. This is my last picture; it is a little complicated, so please bear with me as we walk through it.
The vertical axis is a measure of climate risk. It is the peak warming above pre–industrial levels, which is an indicator of most of the impacts that I spoke about earlier. Again, we can put traffic lights on it: at the orange light, for two degrees of warming, there is a roughly 50 per cent chance either way for dangerous versus non-dangerous climate change. Warming of three degrees or higher is definitely in the danger category, indicated by the red light.
Of course these projections are not absolute: there is uncertainty involved. The dashed lines indicate a probability distribution for this warming, accounting for the uncertainty. The coloured lines indicate uncertainty in a different way, by giving the peak warming that can be avoided with different levels of probability. So, for example, the red line indicates that with the injection of a trillion tonnes (a thousand billion tonnes) of carbon into the atmosphere from the start of the industrial era, we avoid two degrees of warming with the probability of 0.5. If we want to avoid two degrees of warming with a higher probability of 0.9, we would need to follow the blue curve, which requires a much lower injection of CO2 into the atmosphere in order to stay safe; in fact, just over 600 billion tonnes of carbon. Of these totals, 550 billion tonnes have been emitted already.
So you see that in setting an emissions target, defined as a cap on cumulative emissions, two decisions are involved. The first one is what we consider to be dangerous climate change, and the second is how much risk we are prepared to incur that we might cross into dangerous climate change because our predictions are not perfect. Imagine you are driving along a road and a sign says ‘danger ahead’; depending on your knowledge of what that danger was, you might go to quite some lengths to avoid it. You might want a 95 to 99 percent probability of avoiding the danger before you would travel further along the road.
So we have two choices to make: how much warming we want to avoid and how certain we want to be of avoiding that warming.
We can compare this with past emissions. Past emissions take us up to the point indicated on the diagram by the grey ‘past emissions’ bar. If we use up all conventional fossil fuels as currently inventoried by a number of agencies, then our cumulative emissions would take us to the end point of the ‘conventional fossil fuel reserves’ bar. But of course we have a very large supply of unconventional fossil fuels, including tar sands, shale oils, methane hydrates and a number of others. By using these unconventional reserves we can achieve cumulative emissions which go beyond the right hand edge of this diagram, as indicated by the grey arrow. This would carry us far into the zone of climate danger indicated by the red traffic light on the vertical axis.
[SLIDE: Summary].
To summarise: Climate on Earth has always been variable. We have excellent records for the last million years, showing that small changes in the Earth’s orbit have triggered large changes in temperature. We know, therefore, that climate is a system which has reinforcing feedbacks in it. It is a system which, if pushed, can respond more violently than a small initial push would suggest.
Current climate change, however, is different from what has happened over this million-year period, because it is driven not by orbital variations but mainly by human–induced greenhouse gas emissions. We know the current drivers of the rising emissions of CO2 and other greenhouse gases, which are population and growth in income, offset by improvements in the efficiency with which carbon emissions are used for wealth generation. There are persistent links between wealth and energy and emissions, and of these, we have to break the link between energy and emissions to bring this problem under control.
The mitigation challenge is to put a cap on cumulative CO2 emissions, that is, to treat CO2 emissions as a non–renewable resource. CO2 emissions have helped us to get rich as a society, and as a globe, to the extent we are now. However, we have reached the point where further emissions of CO2 very quickly will become destructive for the planet. Peak warming of less than 3 centigrade requires a mitigation rate of about 2 per cent per year. This is a very large rate. Peak warming less than 2 centigrade requires a yet larger mitigation rate of about 5 per cent per year.
These are very, very challenging targets but I believe they are targets that it is essential that we, as a community and as a planet, undertake.
Thank you very much.
Discussion
Chair (Kurt Lambeck): Thanks very much, Mike. I am sure there are going to be quite a few questions here. I hope you will entertain them.
Question: Thank you very much for that very clear exposition of the CO2 story. I would like you to expand a bit on what effect humans are having on water fluxes on the planet. You pointed out that water vapour is probably the most potent of the greenhouse gases, about 90 per cent in the uncondensed form. But when it is condensed in thick cloud cover it is actually cooling the planet. Is the human activity that is leading to global warming something to do with increased water vapours from vehicle emissions or burning coal and reduced cloud cover by cutting down forests and replacing it with grassland? Are data available, for example, for the Amazon, from satellite observations, over what is happening there in the last few years?
Michael Raupach: That is a very good question with several parts to the answer. The first would be that we know that the energy balance of the Earth's surface includes terms, in addition to evaporation – you mentioned changes in albedo – and these changes are having just as significant effect on local energy balances as are evaporation changes.
The question of forestation or deforestation and the effects of forestry on the temperature balance of the Earth comes down to a competition between whether the CO2 saved by a forestry project, for example, is going to be offset by the change in albedo, given that forests tend to be darker than many of the surfaces they replace, particularly in northern latitudes. It turns out that the answer, broadly speaking, is that forests in northern latitudes have a warming effect because of this change in albedo. Forests in mid-latitudes and tropics have a net cooling effect.
As to the role of clouds, clearly they are crucial in the climate system. One of the remarkable features of the climate system over the last century, for as long as observations have been around, is the stability of the planetary albedo, indicating that there are some powerful feedbacks that control clouds in a similar way to the feedbacks that control water vapour in the clear atmosphere.
Other roles for aerosols are now becoming better understood, particularly as they interfere with the set of feedback mechanisms. For example, the aerosols contributing to the so-called Asian brown cloud are having a net warming effect in the lower atmosphere in that part of the world.
Question: Thanks very much, Mike. I enjoyed that. There was one part though that I didn't buy, and that was the part where you said we happen to live in one of those places that's going to dry out. If you look at the outputs from the Fourth Assessment Report, or indeed the model intercomparison that was made just after the Fourth Assessment report, which includes all of those, plus a couple of others that came in late, like the CSIRO Mark 3.5, if you look at those outputs for the Murray Darling Basin then the average precipitation across all models in the Fourth Assessment Report plus the one that came in late, is actually for a slight increase in precipitation.
I was talking today to Leo Rotstayn, whom you know well, who is looking at further predictions. We have a couple of unpublished scenarios for Australia. And it turns out that it is very sensitive to the starting conditions that those models were run in. Unfortunately there was only one run for the future for Mark 3.0, and there was only one run for 3.5 at the time it went in the Fourth Assessment Report. But runs that have been done at CSIRO on both those models since that time show that some of them have drying and some of them have wetting. So it is about 50 per cent of the models show the Murray Darling Basin drying and 50 per cent of the models show wetting, if you just look at models. But then within models, 50 per cent of the runs show drying and 50 per cent of the runs show wetting.
I believe everything else that you said, but I think that the water vapour story and how it turns into precipitation is such a dynamic and difficult issue that I think it is very dangerous for us to assume that the present drought we are seeing is necessarily anything to do with anthropogenic climate change, because the models upon which we rely tell us that it could actually be getting wetter with CO2 forcing.
Michael Raupach: I take that caution very seriously, Graham. The reason that I think it is possible, at least indicatively, to go a little bit further than that now is because of the evidence coming from the Subtropical Ridge Index and the Indian Ocean dipole. We now have not just model runs but a mechanistic association between observable indices over the past and trends in rainfall. The likely future trajectory, particularly of the subtropical ridge, is for further intensification. That, of course, is an easier problem to predict than the precipitation.
So absolutely, it is very difficult for the models to make a prediction about precipitation. But given an intermediary, which the models find easier, I think the evidence is starting to accumulate that we are looking at a microcosm of the future in the last 10 years.
Question: Mike, you have already answered my question about forest cover and albedo. But I have another question about atmospheric CO2. I noticed in one of your graphs that the rate of increase of atmospheric CO2 flattened off in the middle of last century. What caused that?
Michael Raupach: No–one knows. It is thought that it was an ocean overturning which caused the uptake of a significant amount of CO2 into the ocean. Isotopic evidence indicates that. But it is not possible to sheet that home to a specific mechanism at this stage.
Question: Thanks Mike. You spoke about the reinforcing mechanisms that act on Earth and it is certainly very worrying. Given that the Earth over long time scales is roughly in equilibrium, it doesn't go off and on, could you discuss what mechanisms are in place to stop the reinforcing mechanisms?
Michael Raupach: Essentially your question is what mechanisms prevent runaway climate change into states like a snowball Earth or a completely ice free Earth. I guess nothing in principle prevents that. There is certainly evidence that the Earth has been ice free in the geological past, much earlier than the records we were looking at.
There is some evidence, controversial, that the Earth has been a snowball in the past. So these are possible states of the Earth's system. The factors that prevent the Earth from heading into those states at the moment include a number of changes in the processes that we were talking about that together bring the positive feedbacks under control. The positive feedbacks keep acting until the system changes enough for some other negative feedback is turned on. For example, as CO2 drops to levels of around 180 parts per million in glacial periods, there is the possibility that we have reached an ecosystem compensation point below which CO2 cannot go, so further decline in CO2 does not occur. At the other end, the uptake of CO2 by both terrestrial and marine systems keeps CO2, in the long–term record, to about 280 parts per million. So that uptake becomes sufficiently strong that the concentration doesn't exceed that level.
It's a very good question, and it is not fully known what guard rails hold the Earth System between those limits. But the fact that there are guard rails is clear from the record. Also clear from the record is the fact that the system is sufficiently prone to reinforcing behaviour that changing the rules by which those guard rails are set carries the risk of unpredictable consequences.
Question: Thanks for your talk. I just wanted to hear a few more of your thoughts on risk. Do you think humans are perhaps too willing to take risks? Are we by nature risk takers? And how do you see society working together to achieve these mitigation rates when different groups in society are willing to take different levels of risk? So Pacific islands would not want to take very much risk at all, whereas say the coal industry might be willing to take higher risks because they might think there will be technological solutions.
Michael Raupach: That is a really interesting question and one that I can only give a couple of indicative thoughts on.
One is that in circumstances where threats are real and can easily be perceived, and have a definite time frame, humans are usually very risk averse. Most of us would not take a risk that involved possible loss of life with any probability that didn't have a lot of zeros in front of it.
Climate change seems to be a phenomenon which finds its way through many of these mechanisms that humans have evolved over years to calibrate and assess risk. For example, it involves collective action, the threat is diffuse, and it is long term. Action has to be taken by a generation, maybe two generations, before there is demonstrable benefit. This combination of attributes means that we are not able to sense the threat from climate change in the same way that we might sense, say, the threat of a war or something like that. We respond to these things in utterly different ways. There is a field of study on this at the moment, and I believe our perceptions of climate risk represent one of the major dangers that we are facing.
Question: Thank you, Michael. You didn't mention soil carbon. McAlpine and his group in Queensland have calculated that as much CO2 is being released into the atmosphere by deforestation and loss of CO2 from humus in the soil as there has been from fossil fuel combustion.
The answer to that is optimistic, in that if we can restore soil fertility, stop land clearing, this would be one way of actually reducing atmospheric CO2 rather than just slowing its accumulation. In other words, biosequestration is an unexplored area. The government keeps talking about geosequestration, which is untried, very expensive and would probably take 30 years.
Biosequestration, if we really start an agricultural revolution and stop land clearing, that might have a more immediate effect.
Michael Raupach: That's a good comment and I completely agree. One of the things I did not have the opportunity to do at all in this talk was to explore mitigation options, which is another discussion.
Certainly biosequestration is a major mitigation option. It needs to be treated with caution in the following respect: carbon that has entered the atmosphere from fossil fuel burning has come from a pool which has been locked away for hundreds of millions of years. This is now shared between atmosphere, ocean and land pools which exchange their carbon with time scales of the order of tens to hundreds of years, sometimes less than that. Sometimes just years. When we put carbon into biosequestration we are undertaking a management option, or reversing a previous management option, which causes a change in that equilibrium.
There are two things about biosequestration we have to ensure. The first is that that management needs to be propagated for as long as we want the sequestration to last. In other words, the management has to be stable for many human lifetimes. The other is that land carbon, much more so than ocean carbon, is prone to fluctuation because of short–term droughts and other climate variability. Therefore, carbon sequestered in the biosphere flows in large quantities, particularly in this country, to and from the atmosphere.
As an example, the average net primary productivity of Australia is roughly the same as the entire greenhouse gas budget of the country. The net primary productivity fluctuates from year to year by an amount of about one–third of that. Year–to–year variations in biosequestration have enormous short–term accounting implications as well as some dangers in terms of the possibility of setting up systems which might not be stable from the management point of view in the long run. All of that said, every option is really important and, therefore, biosequestration needs to be explored with all of those things in mind.
Question: A very interesting talk, Mike. I wanted to ask you two questions. One relates to the current level of carbon dioxide. According to Copenhagen this is 460 ppm equivalent. This is only about 40 or so ppm below, in terms of the recent history of the Earth, the 500 barrier to the formation of the Antarctic ice age, plus or minus 50 ppm. Maybe we are closer if you are looking at the past history of the Earth.
In the mid Pleiocene levels are as low as 400 parts per million. Already west Antarctica and Greenland have melted from the rise of sea level to 25 metres and only for 400 parts per million and 2 or 3 degrees warming.
Secondly, people keep on talking about 0.7 or 0.8 degree rise. We have to look at the poles. Mean temperature rise in large parts of the poles are 4 and 5 degrees. And in one interpretation it is the poles which are the key to the temperature of the Earth. It is the circulation vortex, the polar winds and so on which cool the temperate zone.
People look a lot at the tropics and also the subtropics where you get water vapour. In the poles there is hardly any water vapour, so the argument falls down. We know that the temperature at the poles have risen 10 degrees or so in the mid-plane of things. So I think somehow the overall picture is lost about global warming when we talk about some regional effects. Thank you.
Michael Raupach: Both of those are good comments, Andrew. With respect to the first, the possibility of changes or instabilities occurring at CO2 equivalent levels, which are comparable to where we stand now: that possibility exists. And, of course, it was the reason why Hansen set 350 ppm as his target CO2.
The other side of the coin is that right now it is almost impossible to meet a target – speaking now in terms of cumulative emissions of 500 gigatonnes – which would be consistent with about a 2 degree temperature rise. When we work through the various equivalent concentration levels it is almost impossible to meet that target, given the rates of mitigation we need to undertake. Humanity faces a choice between bad and worse: that’s what it comes down to. This just adds to the need for rapid action. I will let that answer stand for the second part of your question too.
Question: I may have missed something, Mike, but on the subject of biosequestration, how do you see the role of the incorporation over the longer term of biochar into the soil?
Michael Raupach: Potentially important for exactly the same reasons as we spoke about before. Clearly biochar has proximate benefits in terms of nutrient levels in the soil and so forth. We do need to watch two things: the sustainability of the system in a management sense – can we keep this going for a long time – and the ancillary costs in a carbon sense. We just need to be sure that we are in fact greenhouse positive in the production and so forth of the biochar.
Question: Thanks very much, first, for this wonderful evening. My name is Ben Mathers. I would like to ask, do you see a need for geoengineering, for instance, to artificially create water vapour in order to reflect sunlight?
Michael Raupach: Geoengineering, I think, is attractive because we are in a dilemma and it would appear to provide a way out. My main reservations about geoengineering are quite generic. We don't understand enough about the Earth system to know what we are managing; that is problem one. The second problem is that any geoengineering solution we undertake, I think we can predict with near certainty, will have winners and loses. In other words, climates will be affected in ways that some people don't like, while at the same time being affected in ways that the majority do like. So the question of how to handle the globe as a whole, condemning some peoples to a very unfavourable climate future, is one that is really hard. For those two reasons I think we should try everything we possibly can before we go to geoengineering.
Kurt Lambeck: Thank you very much, Mike. I think that is a good point at which to leave the discussion this evening. Thank you very much for what I have seen as a very informative talk. What I particularly liked, Mike, is the way that you have emphasised that while there is a high degree of certainty about the science there are still important areas of the science that need to be sorted out. It is in your hands and other peoples' hands to do this.
