14.1 — The Carbon Cycle

Every atom of carbon in your body has been in the atmosphere, in the ocean, in rock, and in other living things, repeatedly, for four billion years.

Carbon does not get used up. It moves. The whole of climate science is the accounting of where it is and how fast it moves between places.

And the problem this Part is about can be stated in one sentence: human activity is moving carbon from a store where it sat for hundreds of millions of years into a store where it changes the temperature, faster than the natural processes that move it back.

The reservoirs

Where carbon sits, in rough magnitudes.

Reservoir	Approximate carbon held
Sedimentary rock, mostly carbonate	tens of millions of gigatonnes
Deep ocean	~37,000 gigatonnes
Fossil fuel reserves	~4,000–5,000 gigatonnes
Soil and permafrost	~2,500 gigatonnes
Atmosphere	~890 gigatonnes
Living vegetation	~550 gigatonnes

A gigatonne is a billion tonnes.

The atmosphere is one of the smallest reservoirs, which is exactly why moving carbon into it has a large effect. Adding a given quantity to the deep ocean changes it imperceptibly; adding the same quantity to the atmosphere changes it substantially.

The fast cycle

Photosynthesis and respiration, moving carbon between air, life and soil on timescales of days to centuries.

Plants take carbon dioxide from the air and, using sunlight, build sugars. Animals and decomposers eat those and release carbon dioxide back. Roughly 120 gigatonnes moves each way each year between land and atmosphere, and around 80 between ocean and atmosphere.

These enormous flows nearly cancel, which is why the atmospheric concentration was stable for thousands of years before industry.

And it is visible in the data. The Keeling curve — atmospheric carbon dioxide measured continuously at Mauna Loa since 1958 — shows a saw-tooth wobble on top of the rising trend. The wobble is the northern hemisphere's growing season: land plants, which are mostly north of the equator, draw carbon down in northern summer and release it in northern autumn. You can see the planet breathing in the measurement.

The slow cycle

Geology, over hundreds of thousands to hundreds of millions of years, and this is where the fossil fuels come from.

Chapter 1.4 gave the weathering thermostat. Rain, made mildly acidic by dissolved carbon dioxide, reacts with silicate rock; the products wash to the sea; marine organisms build them into carbonate shells; those settle and become limestone. This removes carbon from the atmosphere permanently on human timescales.

And it is temperature-sensitive, which makes it a genuine thermostat: warmer and wetter means faster weathering means faster removal means cooling. It has kept Earth's climate within liveable bounds for billions of years, and it operates over hundreds of thousands of years, which is far too slow to help now.

Volcanoes return carbon from subducted carbonate rock — roughly 0.3 to 0.4 gigatonnes a year, which is a very small figure and is worth remembering because the claim that volcanoes emit more than humans is common and wrong by a factor of about a hundred.

And the fossil fuels are the burial that failed to complete the loop. Chapter 1.10 gave the specifics: most coal from the Carboniferous, when trees evolved lignin before the fungi that decompose it caught up; most oil from marine plankton buried in anoxic sea-floor mud. Carbon that was removed from the atmosphere over tens of millions of years and stored.

Burning it returns that carbon to the atmosphere in a couple of centuries.

That rate ratio is the entire problem. Nothing about the carbon is unnatural. The speed is.

What humans have added

Around 2,500 gigatonnes of carbon dioxide since 1750 from fossil fuels, cement and land-use change.

Where it went, and the accounting is well constrained.

Roughly 45 percent stayed in the atmosphere.

Roughly 25 percent went into the ocean — with the acidification consequence of Chapter 1.11.

And roughly 30 percent into land vegetation and soil, partly because higher carbon dioxide concentrations stimulate plant growth.

So the ocean and the land have absorbed over half of what was emitted, which has substantially slowed the warming and which is not free.

Atmospheric concentration has risen from about 280 parts per million before industrialisation to over 420 now — higher than at any point in the ice core record covering 800,000 years (Chapter 2.8), and, on proxy evidence, higher than at any time in the last several million years.

And the isotopes prove the source. Plants preferentially take up the lighter carbon-12 (Chapter 1.1), so fossil carbon is depleted in carbon-13. Atmospheric carbon-13 has been falling as concentration rises, which is the signature of fossil carbon. And fossil carbon contains no carbon-14, having decayed away over millions of years, so the proportion of carbon-14 in atmospheric carbon dioxide has fallen — the Suess effect. Two independent isotopic fingerprints, both pointing to fossil fuel.

Meanwhile atmospheric oxygen has fallen by exactly the amount expected from burning that quantity of carbon, which is a third independent confirmation.

The other greenhouse gases

Carbon dioxide is not the only one and the comparison requires care.

Methane. Around 80 times more powerful per molecule over 20 years and around 28 times over 100 years, because it breaks down in the atmosphere in about a decade. Sources: agriculture, particularly rice paddies and ruminant digestion; fossil fuel extraction leaks; landfills; and wetlands. Its short lifetime means cutting methane produces fast temperature benefits — which makes it the highest-leverage near-term target.

Nitrous oxide. Around 270 times carbon dioxide over a century, with a lifetime over a hundred years. Principally from nitrogen fertiliser.

Industrial gases — refrigerants and others — with very high potency and small quantities. The Montreal Protocol's Kigali Amendment addresses the most damaging of them.

And water vapour, which is the largest greenhouse gas by contribution and which is a feedback rather than a forcing. Its concentration is set by temperature — warmer air holds more (Chapter 2.3) — so it amplifies whatever change something else causes and does not initiate change. Adding water vapour directly does nothing lasting; it rains out in days.

The feedbacks

The reason a modest direct effect produces a larger total, and this is where the uncertainty lives.

Amplifying feedbacks.

Water vapour, as above — roughly doubling the direct effect of carbon dioxide.

Ice-albedo (Chapter 2.8). Melting ice exposes dark ocean or land, which absorbs more sunlight, which melts more ice. Arctic sea ice extent has declined substantially and the Arctic is warming several times faster than the global average.

Permafrost. Frozen ground across Siberia, Alaska and Canada holds roughly twice the carbon currently in the atmosphere. Thawing releases carbon dioxide and methane. How fast and how much is one of the largest open questions.

And weakening carbon sinks. Warmer oceans absorb less; stressed forests absorb less; parts of the Amazon have been measured as net sources in some years.

Dampening feedbacks.

Increased radiation to space as the surface warms, which is the fundamental stabiliser.

And clouds, which are the largest single uncertainty. Low clouds reflect sunlight and cool; high thin clouds trap heat and warm. How the balance shifts with warming is the biggest remaining source of spread in the models, and recent work has narrowed it somewhat toward the higher end.

Climate sensitivity

The single most important number in the field.

Equilibrium climate sensitivity is how much the global average temperature eventually rises for a doubling of atmospheric carbon dioxide.

The current best estimate is around 3 °C, with a likely range of about 2.5 to 4.

And the striking thing about that number is how little it has moved. Svante Arrhenius calculated it by hand in 1896 and got a figure in the range of 5 to 6 °C. A 1979 American assessment — the Charney report — gave 1.5 to 4.5. Forty-five years and enormous advances in modelling and data have narrowed the range and have not moved the centre.

Multiple independent lines converge on it: physical modelling, the instrumental record since 1850, palaeoclimate reconstructions including the ice ages, and the response to volcanic eruptions (Chapter 1.7's Pinatubo, which the models predicted correctly).

The residence time

The fact most often missed and the one with the largest policy consequence.

Carbon dioxide does not have a single lifetime in the atmosphere.

A pulse of emitted carbon dioxide is absorbed in stages. Roughly half is taken up by ocean and land within a few decades. A substantial fraction remains for centuries. And around 20 to 25 percent remains for many thousands of years, until the slow weathering cycle removes it.

Which means the warming is effectively irreversible on any human timescale unless carbon is actively removed.

And it means that stabilising the temperature requires reaching net zero emissions, not merely reducing them. Cutting emissions by half means the concentration rises more slowly and still rises. This is the single most misunderstood point in public discussion of the subject.

Where this shows up in your life

Every fuel you burn moves carbon from the slow cycle to the fast one, and a share of what you emit today will still be in the atmosphere when your great-great-grandchildren are alive.

And the framing that follows from this chapter is worth holding. The problem is not that carbon dioxide is a pollutant in the ordinary sense — it is a normal, essential part of the biosphere and plants require it. The problem is the rate. A quantity of carbon that took the slow cycle tens of millions of years to bury is being returned in about two hundred, and no natural process operates fast enough to compensate.

What the next page covers

Chapter 14.2 covers how we know — the temperature record and how it is constructed and corrected, the palaeoclimate evidence, the fingerprints that distinguish greenhouse warming from every alternative explanation, what the models can and cannot do, how the scientific arguments were actually settled, and an honest treatment of the remaining genuine uncertainties as distinct from the manufactured ones.