Climate change: what do we know for sure, and what is less certain?

In another post inspired by my current first year physics course, The Physics of Sustainable Energy (PHY123), I suggest how a physicist might think about climate change.

The question of climate change is going up the political agenda again; in the UK recent floods have once again raised the question of whether recent extreme weather can be directly attributed to human-created climate change, or whether such events are likely to be more frequent in the future as a result of continuing human induced global warming. One UK Energy Minister – Michael Fallon – described the climate change argument as “theology” in this interview. Of course, theology is exactly what it’s not. It’s science, based on theory, observation and modelling; some of the issues are very well understood, and some remain more uncertain. There’s an enormous amount of material in the 1536 pages of the IPCC’s 5th assessment report (available here). But how should we navigate these very complex arguments in a way which makes clear what we know for sure, and what remains uncertain? Here’s my suggestion for a route-map.

My last post talked about how, after 1750 or so, we became dependent on fossil fuels. Since that time we have collectively burned about 375 gigatonnes of carbon – what has the effect of burning all that carbon been on the environment? The straightforward answer to that is that there is now a lot more carbon dioxide in the atmosphere than there was in pre-industrial times. For the thousand years before the industrial revolution, the carbon dioxide content of the atmosphere was roughly constant at around 280 parts per million. Since the 19th century it has been significantly increasing; it’s currently just a couple of ppm short of 400, and is still increasing by about 2 ppm per year.

This 40% increase in carbon dioxide concentration is not in doubt. But how can we be sure it’s associated with burning fossil fuels? The isotopic signature of the carbon is one important piece of independent evidence. The ratio of the two stable isotopes of carbon – carbon-13 to carbon-12 – has been declining, as would be expected from the different isotopic ratios in fossil fuels. The decline in the atmospheric concentration of the radioactive isotope carbon 14 would be an even clearer signature, as fossil fuel is completely free of carbon-14, if the time series hadn’t been perturbed by the injection of carbon-14 into the atmosphere by nuclear testing.

But it is the correlation of the rise in concentration with the total amounts of human emissions that remains the strongest evidence. To the 375 gigatonnes of carbon released by burning fuel, we need to add another 180 gigatonnes estimated to result from changes of land-use – primarily chopping down forests. When we do the detailed accounting of all this carbon, we find that less than half of the human-emitted carbon has ended up in the atmosphere. The rest has either been absorbed by the ocean or taken up by plants and deposited on land as part of the carbon cycle. While our understanding of these flows of carbon historically is well developed, we can’t be certain of what will happen in the future. For example, we can’t be sure how ecosystems will change in response to increasing temperatures and carbon dioxide levels. The oceans and terrestrial ecosystems are buffers which mitigate, to some extent, the effects of releasing more carbon dioxide, and this uncertainty means we can’t be sure in the future whether a higher or lower proportion of the carbon dioxide we release will end up in the atmosphere.

What has this increase in the carbon dioxide concentration in the atmosphere done to our climate? it’s got hotter – average land surface temperatures have increased by just over a degree since 1850. I’ll discuss temperature trends in more detail below, but first, why should we think that an increase in temperature is caused by the increase in carbon dioxide concentration? To be sure of the causality, we would need to understand the mechanism. Luckily, that mechanism is well understood, and we now have compelling independent experimental evidence for it.

The sun shines on the earth, providing a continuous input of energy. The earth also radiates out into deep cold space, taking energy away; if the earth was in a steady state the amount of energy coming in would be in balance with the energy going out. All bodies radiate according to their temperature; hot bodies radiate more than cold bodies (Stefan’s law). The radiation from hot bodies is distributed over a wide range of wavelengths, but the peak wavelength depends on the body’s temperature (Wien displacement law) – a human being radiates entirely in the infra-red, but for a red-hot poker the peak wavelength is shorter and there’s significant radiation in the visible part of the spectrum. The sun is 5800 K, so it looks white hot, with a lot of radiation in the visible and ultra-violet. The earth is about 290 K on average, so it radiates energy in the infra-red into deep space, whose temperature is 3 K. If the earth was a simple lump of black rock with no atmosphere we could calculate its steady state temperature by calculating how much energy arrived from the sun, and then finding the temperature for which the outward radiation energy would balance the incoming energy from the sun. If we do this calculation, we find an answer that is about 30 K lower than the earth’s actual temperature. The reason for the discrepancy is that the earth has an atmosphere, which leads to a greenhouse effect.

The greenhouse effect arises because the earth’s atmosphere is transparent to the predominantly visible and ultra-violet radiation coming in from the sun, but it is partially opaque to certain wavelengths in the infra-red, corresponding to absorption bands in the optical spectra of gases like carbon dioxide and water vapour. If we change the composition of the atmosphere so it contains more greenhouse gases, then less radiation can leave the earth, while the same amount of energy, in the form of incoming visible and ultraviolet radiation from the sun, continues to enter. If more energy is arriving at the earth than is leaving, then the earth must heat up. As it heats up, it emits more infra-red radiation, until the energy emitted comes into balance with the energy arriving from the sun once again. A new steady state is achieved, with the earth at a higher temperature.

We know the greenhouse effect is important for the earth; if it wasn’t for the greenhouse effect, the earth would be a frozen snowball. But we have more direct evidence than that – we can use satellites to look back at the earth, and measure the wavelength distribution of the radiation it is emitting. What these measurements show are bands of opacity corresponding to ranges of wavelengths in which greenhouse gases absorb infra-red radiation. It is these bands of opacity that reduce the amount of energy that the earth radiates away; as the amount of greenhouse gas increases we expect these bands to become both wider and more opaque. This article from Physics Today, by Raymond Pierrehumbert – Infrared Radiation and Planetary Temperature (PDF) – shows the data and explains the physics very clearly.

The satellite data shows that increasing amounts of carbon dioxide make the atmosphere more opaque to outgoing infra-red radiation. But it also shows that carbon dioxide isn’t the only greenhouse gas. The most important additional greenhouse gas is water vapour. The effect of water vapour is to amplify the effect of carbon dioxide. if the atmosphere heats up, it can take up more water, which absorbs even more infra-red, which amplifies the heating effect. if the atmosphere were to cool down, water vapour would condense and fall out of the sky as rain, amplifying the cooling effect. This is an example of a positive feedback.

Other greenhouse gases are important too, particularly methane. Methane is present in lower concentrations in the atmosphere than carbon dioxide, but it is a much stronger greenhouse gas; its net effect is less than carbon dioxide, but is still significant. Methane concentrations have also been rising since the industrial revolution, though our understanding of where the methane has come from is less well developed than for carbon dioxide. Some comes from land use changes and particularly from farming – rice paddies generate methane, as do the emissions of ruminants. Some comes directly from fossil fuel extraction, in the form of leakage from gas pipelines and escape from wells. There will be feedbacks from a warmer world affecting the amount of methane in the atmosphere, too, though those these will be less direct than for water vapour. A warming arctic will lead to the release of more methane from melting permafrost and from the melting of methane hydrates in arctic seas, for example.

The law of conservation of energy tells us that if more energy is arriving from the sun, in the form of short wavelength radiation, than is leaving, in the form of long wavelength radiation, the earth must be heating up. Of course, this assumes that the amount of energy arriving from the sun is a constant – it isn’t, but we can measure the changing brightness of the sun directly from satellites, and we find the variation is small compared to the effects we expect from changing concentrations of greenhouse gases.

Where is this extra heat going? The three possibilities are into the oceans, into the atmosphere, and into the land surface. One should expect that most of the extra heat energy will end up in the oceans; water has a much higher heat capacity than air, and, unlike the land surface, it’s made of a liquid which can transfer heat effectively by mass transfer. Good worldwide data on the temperature of the oceans doesn’t go back very far in time, so we can only be certain that the upper ocean has been warming since 1971. But it’s very likely that average ocean temperatures have been rising since the mid 19th century, and that the deeper levels of the oceans have been warming too. Significant amounts of heat are absorbed by melting ice, and the combination of melting ice and the thermal expansion of the hotter water of the oceans results in the currently observed rate of sea level rise of about 3.2 mm per year.

It is the average surface air temperatures over land that get the most attention; since 1850 these have increased by a little more than a degree. The rate of increase has not been constant, though; 1910 to 1940 saw a relatively rapid rise, followed by a hiatus until about 1970. There was then an even more rapid warming period until 2000 or so, since when there has been another hiatus. We probably shouldn’t be surprised that the rate of warming at the land surface isn’t steady, given that the majority of the heat is going into the oceans. The study of climate is, in effect, the study of how heat is partitioned between atmosphere, land and oceans, and how that heat moves around from one part of the earth to another, and through the circulation of ocean currents, how heat moves from the surface to the deep ocean. We know quite a lot about our climate system, but much remains uncertain, particularly as our climate system isn’t static but is being driven by an increasing input of heat energy.

So what is certain, and what remains uncertain? We can be certain that burning all those fossil fuels resulted in a 40% increase in the carbon dioxide content of the atmosphere, we can be certain that the resulting increase in the strength of the greenhouse effect has resulted in a warming world, and we can be certain that that warming will continue as we continue to put more carbon dioxide into the atmosphere (barring a significant and unexpected decrease in the total amount of energy emitted by the sun). Three areas of uncertainty remain, though, and it is these that make it difficult to predict exactly what will happen in the future.

Firstly, we don’t know for certain how much more carbon dioxide we will put into the atmosphere. How much carbon we burn in the future will depend on how much the world population grows, on how much the world economy grows and on whether we manage to reduce the amount of energy we need to generate a given amount of economic output. The energy mix we are using will depend, on the one hand, on the availability and cost of fossil fuels, and, on the other, the degree to which we have found cost-effective and scalable alternatives to fossil fuels, and the degree to which we are successful in using policy measures to discourage the use of high carbon fossil fuels. What proportion of the carbon we burn ends up in the atmosphere will depend on how the carbon cycle evolves. Will the uptake of carbon by plants increase, because of the increasing efficiency of photosynthesis at higher carbon dioxide concentrations, or will it decrease due to changes in land use and an inability of ecosystems rapidly to adapt to changing temperatures? How will the ability of the oceans to absorb carbon dioxide change as the water becomes hotter and more acidic?

Secondly, we don’t know for certain how much extra heating effect will be produced by a given rise in atmospheric carbon dioxide concentration. This climate sensitivity is affected by feedbacks, both positive and negative. As we’ve seen, higher temperatures amplify the greenhouse effect of carbon dioxide alone by leading to more atmospheric water vapour and methane. A changing climate will lead to loss of ice cover, decreasing the amount of sunlight directly reflected from the earth and amplifying the warming effect, while an increase in cloud cover would have the opposite effect. We may see attempts to directly manipulate such effects by, for example, releasing aerosols into the upper atmosphere – geoengineering – with effects that themselves are likely to be very uncertain.

Thirdly, we don’t know what effect will a warmer world have on local climate, local weather, and ecology. We understand a great deal about the earth’s climate system, but our knowledge is based on theory and modelling validated by comparison with our experience. As the amount of energy in the climate system increases, we should expect the uncertainties of our predictions to increase. There are still big uncertainties around the possibility of major changes in the circulation of ocean currents, for example. The effects of climate change will not be uniformly distributed across the globe, but we can’t know with any certainty how particular regions will be affected. There is a lot of emphasis in climate modelling in producing consensus predictions for likely scenarios. Perhaps more emphasis should be placed on evaluating less likely, but more extreme outcomes, and then we can decide on how much it is worth to us to insure against those outcomes.

But now I am beginning to address the biggest uncertainty of all – how we should respond to what we know about climate change, in all its certainties and uncertainties. I will return to that in my next post.

On sources. The most comprehensive summary of the data and modelling is to be found in the first section of the IPCC’s 5th assessment report – The Physical Science Basis. For a much longer term perspective, Revolutions that Made the Earth, by Tim Lenton and Andrew Watson, is excellent. Some people attempt to minimise the significance of human induced climate change by saying the earth’s climate has always been changing. Indeed it has; this book tells us why, and sets the current, very rapid, human induced period of climate change in that deep history, allowing us to understand the earth as a system, full of feedbacks, both positive and negative.