SCIENCE AT THE SHINE DOME canberra 7 - 9 may 2008

Symposium: Dangerous Climate Change: Is it inevitable?

Friday, 9 May 2008

Dr Michael Raupach
CSIRO Marine and Atmospheric Research

Michael Raupach has a PhD in micrometeorology from Flinders University in South Australia. After a postdoctoral position at the University of Edinburgh, he joined CSIRO in 1978. His interests include land-air interactions, micrometeorology, the fluid mechanics of turbulent flows, particle transport and soil erosion by wind, global and continental carbon and water cycles, and carbon-climate-human interactions. He is a co-chair of the Global Carbon Project of the Earth System Science Partnership and a Fellow of the Australian Academy of Technological Sciences and Engineering.

The carbon cycle at the climate crossroads

(Click on image for a larger version)

> Thank you, Amanda, and thank you for the honour of being present and speaking at this meeting.

This talk is going to focus not only on the natural carbon cycle – I will say something about that – but also on the human dimensions of the carbon cycle.

(Click on image for a larger version)

In particular, I am going to look at what is driving current carbon dioxide emissions, and the implications of those emissions.

We will first have a very general look at some of the basic facts about climate and the carbon cycle, including the way that carbon dioxide is behaving in the atmosphere, and the reasons for that behaviour. Then we will look at trends in emissions from fossil fuels; and finally we will take a look at the crossroads where we stand now.

(Click on image for a larger version)

The three factors that together drive climate change are: anthropogenic emissions of greenhouse gases, the response of the atmosphere to those emissions, and the response of climate to atmospheric greenhouse gas concentrations. These are pictured in this slide by the curve of CO₂ emissions for the last 150 years, shown at the top, the curve of atmospheric CO₂ concentration in the middle, and the global temperature curve at the bottom. You don't have to look very hard to see that there is a case to be answered that there is a relationship here.

There are more radiative forcings on global temperature than CO₂. We also have solar fluctuations, volcanic forcing, and forcing from other greenhouse gases. Beyond that, there is more to the climate system than radiative forcing: there are feedbacks involving clouds, ice, and atmospheric and oceanic circulations. (Neil Hamilton has already touched on some of the most important feedbacks in Arctic regions.)

(Click on image for a larger version)

However, if we look at the components of radiative forcing, using this very well-known picture from the Fourth Assessment Report of the IPCC [Intergovernmental Panel on Climate Change], we see a couple of interesting things. At the top is the radiative forcing from CO₂, then a series of forcings from other greenhouse gases, and then (mainly negative, or blue, in this picture) the non-gaseous forcings from aerosols and albedo changes. Finally at the bottom we have the sum of all of these forcings. As a reasonable approximation, the sum of all of the forcings is presently about equal to the forcing from CO₂. The rest of the forcings trade off each other off, summing to nearly zero. That is not to say that they will continue to trade off in the future – just that this is the current situation.

There is one other important point about this diagram: the forcing we see there from CO₂ is the net forcing, not the total forcing due to emissions. I will show you in a moment that the forcing due to emissions is about double what you see on that picture.

(Click on image for a larger version)

What are the consequences? Here is another well-known picture from the IPCC Fourth Assessment, showing the projected evolution of global temperature to the end of the 21st century, under several different emissions scenarios ranging from low to high. The orange line in the centre is a scenario in which gas concentrations are held constant at their present level, about 380 ppm of CO₂, and you see that there is still some entrained warming. This continuing warming demonstrates that the climate system has inertia, as mentioned by Neil.

The most important point about this graph is that, based on the global emissions trajectory that we are on at the moment, the lower-emission scenarios are almost off the agenda. I will show you more detail about that soon. But the reality is that it is going to be very difficult to achieve those low-emissions scenarios.

(Click on image for a larger version)

Here is another picture about climate impacts, this time showing patterns of precipitation changes. Neville Nicholls is going to talk a great deal more about that. I would simply note that there is much more to climate than temperature. The distribution of precipitation is going to be particularly important, and of course one of the projected impacts is mid-latitude drying – you see the orange belts in both hemispheres in the mid latitudes. We happen to be in a mid-latitude country and location, so these projected trends are of great importance for us.

(Click on image for a larger version)

Are these trends actually happening already? Well, this graph shows the observed flow in the River Murray over the last 50 years, from a gauge just downstream of the confluence of the Murray and Darling Rivers. You see the decline since about 2000. The flow in the Murray since 2002 has been about 24 per cent of what it was in the preceding 50 years.

Is there a connection between this and the previous picture of predictions? I don't think many people would say that the connection is proved at the moment, because it is hard to separate a climate change signal from natural varaibility. However, the evidence for a connection gets stronger with each year that we see this kind of pattern, and the trend is continuing through this current season, as Neville may indicate in a succeeding talk.

(Click on image for a larger version)

I will turn to the carbon cycle now.

(Click on image for a larger version)

Let's look at the factors which have affected the concentration of atmospheric CO₂ over the last 150 years or so.

As a background, let's note that atmospheric CO₂ has been stable in the atmosphere for about the last several thousand years at about 280 ppm. Since then, as the earlier graph showed, it has risen to about 380 ppm, and what we are going to do now is walk through the factors which affect that rise.

First we have a series of emissions, which appear above the horizontal line. The first contributor is the CO₂ emission from land use change, mainly deforestation, which now contributes about 1.5 petagrams of carbon per year – that is gigatonnes of carbon, in carbon dioxide, per year – to the input of CO₂ into the atmosphere. Back in the 19th century most of that land clearing was extra-tropical, but it is now almost exclusively tropical.

(Click on image for a larger version)

The next and largest contributor (the 'elephant in the room', if you will) is emissions from fossil fuels, contributing about 7.6 petagrams per year, averaged over the last six or seven years. There is also a small contribution which you can see at the top there which comes from non-fossil fuel industrial sources such as cement manufacture.

(Click on image for a larger version)

All of these emissions have to be accounted for somehow. They have to go somewhere. Where do they go?

(Click on image for a larger version)

About 45 per cent of the total emissions accumulate in the atmosphere. That is why I said a moment ago that the net forcing from CO₂ is only about half of the gross forcing.

(Click on image for a larger version)

Of the part that does not accumulate in the atmosphere, some – about one quarter of the total emission – goes into the oceans. That is a reasonably steady uptake. The oceans have been a reliable sink of carbon dioxide for the last 150 years or so.

(Click on image for a larger version)

The remainder goes to land systems, and you will notice that this terrestrial sink of carbon dioxide is quite variable. It oscillates up and down very sharply. This happens because land, far more than oceans, is subject to perturbation by a couple of major influences – the ENSO cycle, which affects precipitation distribution and thereby the biological activity of land surfaces, and also volcanoes.

(Click on image for a larger version)

Here is another look at that same budget – same data but slightly amalgamated: the red line at the top is the total emissions, land use change plus the industrial emissions and fossil fuel emissions; the blue line is what accumulates in the atmosphere; and the green line below is the sum of the land and ocean sinks.

(Click on image for a larger version)

We can look at the ratio between the accumulation in the atmosphere and the total amount going in, the total of all of the emissions. This ratio is known as the airborne fraction (AF). It is the fraction of emissions that accumulate each year in the atmosphere. We said a moment ago that it was about 45 per cent, but there is more to it than that.

If we look at the AF over the last 50 years –the duration of the high-quality CO₂ records from stations such as the Moana Loa station in Hawaii – then we see a slow but detectable increase.

You see here two curves to demonstrate that. The one on the left is the ratio that you get if you simply take the monthly observations. As you see, it is quite noisy and it is impossible to see a statistically significant trend.

If we clean up those data by removing the fluctuation component which is associated with both ENSO and volcanic activity, then we get a much cleaner signal; we can see a significant trend. It is about 0.24 per cent per year, which means that over the last 50 years or so the airborne fraction has increased by somewhat more than 10 per cent, from around 0.43 to 0.48.

What is the implication of that? It is that each year the land and ocean sinks, which are 'forgiving' us for some of the anthropogenic emissions of CO₂ that enter the atmosphere, are taking up a progressively smaller fraction of those emissions.

(Click on image for a larger version)

This slide shows how the total uptake is distributed between the land and ocean sinks. The top curve is the AF, the fraction of emissions that accumulates in the atmosphere. The middle and lower curves are the corresponding fractions that go to the land and ocean sinks. And, of course, the three curves have to add up to 1. The important point is that the land sink, according to these data, has been reasonably stable over the last 50-odd years, but the ocean sink is taking up a progressively smaller fraction of emissions.

This doesn't mean that the ocean sink is declining in absolute terms, because the emissions are rising very rapidly. It does mean that the ocean sink is weakening relative to the rate of growth of emissions. The sinks, and particularly the ocean sink, are 'losing the race' with these rapidly increasing emissions.

(Click on image for a larger version)

We come now to the question of the future vulnerability of the land and ocean sinks. Clearly, if the sinks are changing, then the future trajectory of climate change will depend critically on whether we continue to have this 'forgiveness' by the Earth system, through the land and ocean sinks, for about half of our emissions. Will this continue?

This is a major question. It has been investigated in a large-scale intercomparison experiment called C4MIP, the Coupled Current Carbon Cycle Model Intercomparison Project experiment, where a number of feedbacks were included in global climate models with carbon components. These feedbacks encompassed ocean CO₂ uptake, the CO₂ fertilisation effect on land net primary production, and climate effects on carbon releases from land pools by some processes, such as temperature and moisture effects on soil respiration.

The conclusion of that experiment was that all of these feedbacks together yield increased CO₂ and therefore increased warming, relative to climate predictions without the feedbacks included. The additional temperature increase in 2100 from including the feedbacks is between 0.1° and 1.5°, depending on the particular model.

These are significant conclusions. One of the most significant is that some of the models, and one in particular – the black one in the land graph, which is the Hadley Centre global climate model coupled with a carbon model – produce an actual turning around of the terrestrial sink. That turnaround, if it occurs, would imply that instead of a terrestrial sink we have a terrestrial source of CO₂. And that would induce a significant acceleration of CO₂ in the atmosphere.

A first impression is that the carbon components of all of these models are rather scattered, that they don't agree very well. This is true, but it's not the main reason for the turnaround in the Hadley Centre prediction of the land sink. The turnaround is primarily to do with the climate-carbon coupling, because the atmospheric circulation in that model largely dries out the Amazon, producing a large flux of carbon into the atmosphere.

(Click on image for a larger version)

We can compare the observed increase in airborne fraction, roughly 0.24 per cent per year over the last 50 years, with what the C4MIP models predict would have happened to the airborne fraction over that period. Here is that comparison. The models were run in two different modes, coupled and uncoupled – the difference doesn't matter for now. The important point is that most of the models predicted that the airborne fraction decreased over the last 50 years. It has actually increased. This suggests that we still have some way to go in the science of coupled carbon-climate modelling, and that the predictions of the C4MIP family of models fall into yet another area where current modelling is somewhat conservative: it is underestimating likely trends.

(Click on image for a larger version)

We turn now to the coupled carbon-climate system, to examine the drivers of both CO₂ emissions and the CO₂ growth rate in the atmosphere. Let's begin by looking at the drivers of emissions.

Looking at the upper graph here, we can consider changes in emissions to be the consequence of changes in population, changes in wealth (affluence) and changes in the carbon intensity of wealth generation. Those are represented here by the black line for the trajectory of emissions over the last 50-odd years, the red line for the trajectory of population, the green line for the trajectory of wealth per person or affluence, and the blue line for the trajectory of the carbon intensity of wealth generation – how much carbon we emit to generate a dollar of wealth. All lines represent averages across the whole world.

You see that the product of the three coloured lines, red, green and blue, is the black line. There are roughly equal contributions to the increase in the black line from population and from wealth generation. It is often said that in climate science we ignore the population problem. That is not true. Here is the population problem, and it is half of the story of increasing emissions.

If we go to the lower graph, we can now trace the consequence of those increasing emissions, using the airborne fraction. We write a slightly different identity in which the growth rate of CO₂ includes the same three terms – population, affluence and carbon intensity – together with an additional term, precisely this airborne fraction we have been talking about.

In black is the curve over the last 50 years for the CO₂ growth rate. It is very noisy, because of the ENSO and volcanic factors that we have already mentioned. The main reason for the noise is the airborne fraction; the other curves are the same as before. What we see is that relative to the other contributions – from population, affluence and carbon intensity – over the last 50 years, the change in airborne fraction has not been a major factor in increasing the atmospheric CO₂ growth rate. The atmospheric CO₂ growth rate has accelerated through that period by something like 1.9 per cent; about 0.2 per cent of that acceleration is due to the airborne fraction, and the remaining 1.7 per cent is due to the other three factors: population, affluence and growth rate.

(Click on image for a larger version)

I will pass over this slide because we are challenged for time.

(Click on image for a larger version)

I come now to emissions.

(Click on image for a larger version)

Global emissions are accelerating and the curves fall above IPCC scenarios. We published this result in 2007. We received a little push-back from some people in the scenario community, for reasons that I won't go into, but we have examined the result in great detail since then and it is absolutely true.

You see on the bottom left a curve showing recent actual, observed emissions, together the family of IPCC scenarios for the same time period. The current growth rate exceeds almost all the scenarios.

(Click on image for a larger version)

Here is how it emissions break out among regions. For this analysis we used a set of nine regions, some of which were individual countries such as the USA and China. You see the dominant contributions from the USA, you see Europe, you see China rapidly growing – look at its rapid acceleration in recent years. The growth rate of emissions in China is of the order of 10 per cent per year.

(Click on image for a larger version)

To look at this data in a slightly different way, we can look at development trajectories. On the horizontal axis I have plotted the per capita income for these nine different regions, plus a few individual countries like Australia, and on the vertical axis the per capita fossil fuel emissions. The first impression we gain is that there is a linear relationship. It is changing a little from time to time; the points joined by straight lines represent the situation in 1980, 1992 and 2004, from left to right. But on the whole that relationship is nearly linear.

If we are to achieve the sorts of emissions reductions we need to stabilise climate, then that relationship has to change, not in a minor way but grossly. To achieve emissions reductions of the order of 60 to 80 per cent by 2050, these smeared points indicate roughly where we need to be. To get there, the trajectories have to change dramatically.

(Click on image for a larger version)

I come now to the last point.

(Click on image for a larger version)

The carbon-climate-human system now involves a close coupling between carbon emissions, energy use and affluence. I want to show you something about trends in affluence which I believe is striking. The upper picture shows the global development of population and gross world product (GWP) for the last 2000 years, from a data set put together by Angus Maddison. You see in the lower picture the per capita wealth, formed by dividing GWP by population. That per capita wealth remained more or less constant through the whole of those 2000 years until – we can almost pick the decade – about 1830. In 1830 the human system underwent a phase change, and our per capita wealth started doubling every 45 years.

Why is that important for emissions? It is because wealth generation at the moment is pulling energy and CO₂ emissions with it, through more or less linear relationships. These are the relationships we have to break.

(Click on image for a larger version)

Let me leave you with three conclusions. First, we have an increasing trend in the airborne fraction; it is one of the signs that the carbon cycle is changing. It is about 0.24 per cent per year, signalling that sinks – both land and ocean, but particularly ocean – are 'losing the race' with emissions. And our current models for the carbon-climate system don't capture this.

Next, emissions have been accelerating at over 3 per cent per year since 2000. This trajectory is continuing; there is no sign that it is weakening. Consequently, the Garnaut review is using as its base case a kind of super emissions trajectory that is more emissions intensive than any of the SRES [Special Report on Emissions Scenarios] scenarios used in the Fourth Assessment Report.

What do we need to do to manage the atmospheric global commons? There are four essential components: the setting of a cap, the sharing of that cap, the timing of emissions trajectories within the cap, and a compliance mechanism. The technology is available and affordable – it will be discussed later today. There are many estimates of the cost of getting Australia from the emissions trajectory we are on now to the one we need to play our part in solving the climate problem. The estimates are of the order of, or less than, the Australian defence budget. So this is not something that would cripple the country. I will leave you with that thought.

(Click on image for a larger version)

Thank you.

Discussion

Question: One relationship which you didn't show was that between CO₂ concentration in the atmosphere and temperature. Does that go linear all the way up? Does it flatten off? And if so what are the analytical coefficients that control that relationship?

Michael Raupach: It is much more complicated than going linear, because there are time lags involved in that relationship, of the sort indicated by the fact that if we stabilised CO₂ now, temperature would continue to increase. If we simply plot, regress, CO₂ in that first picture that I showed against temperature, that is not a particularly sensible thing to do. Nevertheless, it is possible to study, both analytically and with models, the relationship between CO₂ change in an equilibrium world – that is, if we allow the whole system to come to stability – and temperature. This number is called the climate sensitivity. The sorts of numbers that emerge have a wide scatter but the mean quoted by the Fourth Assessment Report was about 3° of temperature change for a doubling of CO₂. At this stage those numbers, particularly for the sorts of reasons that Neil brought forward, are looking somewhat conservative.

Question: There is one thing you didn't mention which I think relates to the carbon cycle, and that is the issue of methane, and particularly the release of methane. How will that, effectively, influence the scenarios that you look at in terms of the warming? After all, it is somewhat more greenhouse-gas efficient.

Michael Raupach: There are many things I didn't mention. But methane is important. In fact, I skipped over my slide headed 'Vulnerable land and ocean carbon pools (2000–2100), which had some information about methane.

The reserves of methane in the methane hydrates that Neil mentioned amount to hundreds of petagrams of carbon in total. If all of those were released, they would provide a perturbation on atmospheric CO₂ which is of the same order as the projected fossil fuel emission over the next, say, three to five decades.

There is one difference, that methane has an atmospheric lifetime of the order of 12 years, at present. So that release does cause a rapid warming effect, because methane has a high global warming potential, but if the release were rapid the consequence would be in the nature of a pulse.

Question (Peter Pockley): You use modelling a great deal in your work, and I just want to ask whether you and others, who depend very heavily for the kinds of conclusions that you are presenting publicly, appreciate that the so-called critics or deniers of climate change very frequently attack the models for being faulty in some sort of way. How do you convince people about the reliability of the modelling techniques that you use?

Michael Raupach: It's a good question. In fact, I think this was a pretty observationally oriented talk. Most of the time axes of the graphs ran from 1950 to present, so most of this talk was looking back at observed trends.

With regard to how we convince people that the models are true, I would make a couple of points. First, the current conviction by the scientific community that climate change is real and dangerous does not only rest on models, it also rests on observations and on basic physics – all of these. Second, despite many faults that all of us in this room would be conscious of, the models in fact do a reasonable job now of hindcasting the behaviour of climate over the 20th century, and therefore their predictions for the 21st century have some credibility.

Those would be a couple of reasons why it is not logical to say that the whole edifice of climate change rests on models, models are faulty, and therefore you should not believe it. In fact, if that is your argument against climate change then you would never board an aircraft, because these days aircraft are designed using computational fluid dynamics models built on the same flow equations used in climate models.

Question: What is the cause of the increase in the airborne fraction? Is it a rise in the surface temperature of the ocean, or is it just a reflection of the long lifetime of the CO₂ in the atmosphere?

Michael Raupach: It is a combination of the rise of surface temperature and the change in ocean pH. As CO₂ is accumulating in the ocean, the pH is lowering – it has gone in the mean from something of the order of 8.2 to 8.1 in the 50 years we are talking about – and the consequence of that is a reduction of the buffering capacity of the ocean.

Question: Thank you very much for introducing population. We have 200 'elephants' in this room. We are the final, common cause of the phenomena that you are so beautifully describing.

In your list of things to do you did not mention to contain human population growth, which I would say is the first thing we must do. And China has done it. China's one-child family policy, introduced 15 or so years ago, has been remarkably effective in containing China's population growth rate. So I think all of us need to look at human population growth and think, 'How can we stop the current 6.7 billion people rising to 9.2 billion' – the UN accepted estimate – 'by 2050?' That is almost unsustainable.

If I could, I would leave us all with a rendition of a quote of 400 years ago from John Donne's famous sermon in St Paul's Cathedral: 'Ask not for whom the bell tolls; it tolleth, man, for thee'. How true is that? But he also said, 'I do nothing upon my self, and yet I am mine own Executioner.'

Michael Raupach: I absolutely agree with the thrust of those comments. With regard to quotes, I didn't have a chance to discuss my last slide, which had a quote from Joni Mitchell, 'We are starstuff, billion-year-old carbon. We are golden, caught in the devil's bargain' – which I think would echo, perhaps, three centuries later, John Donne's sentiments.

With regard to population, it is clearly a critical issue. I would argue that the change we need to make in the trajectory of emissions is of the order of 80 to 90 per cent, and that change is the product of population, affluence and technology. We could perhaps find 20 per cent of that change in population. That is critically important, but it means that we have to look for most of the change in the other areas.

Amanda Lynch: Thanks very much for a very interesting talk.

Michael Raupach: Thank you.