The Carbon Cycle

The carbon cycle describes the movement of carbon atoms among four principal reservoirs, or pools: the atmosphere, the oceans, the terrestrial biosphere (living organisms plus soil organic matter), and the lithosphere (sedimentary rocks, fossil fuels, and dissolved inorganic carbon in deep sediments). Carbon moves between these pools via fluxes — rates of transfer, typically measured in gigatons of carbon per year (GtC/yr; 1 GtC = 10¹⁵ grams of carbon). The size of a pool and the speed of the fluxes into and out of it determine its residence time — how long, on average, a carbon atom stays there before moving on. Ecologists distinguish two nested cycles operating at very different tempos. The fast (biological) carbon cycle exchanges carbon between the atmosphere, land plants, animals, microbes, and the surface ocean on timescales of days to centuries. Photosynthesis and respiration are its engines. The slow (geological) carbon cycle exchanges carbon between the atmosphere/ocean system and rocks or sediments via volcanism, weathering, and burial, operating over thousands to millions of years. Under natural, pre-industrial conditions these two cycles were in approximate balance: the biological cycle churned enormous amounts of carbon annually, while the geological cycle very slowly bled carbon out of circulation (via burial of organic matter and carbonate rock) and returned it (via volcanic outgassing and weathering) at rates low enough to leave atmospheric CO₂ roughly stable over centuries. The key quantitative fact that organizes everything else: pool sizes vary enormously. Roughly:

The ocean dwarfs the atmosphere as a carbon reservoir, and sedimentary rock dwarfs everything — but the rock pool is nearly inert on human timescales, while the atmosphere-ocean-biosphere system exchanges carbon rapidly. This asymmetry — small, fast pools versus huge, slow ones — is the single most important structural fact for understanding both how the cycle self-regulates and how it can be destabilized.

Step 1 — Photosynthetic fixation. In chloroplasts of plants, algae, and cyanobacteria, the enzyme RuBisCO (ribulose-1,5-bisphosphate carboxylase/oxygenase) catalyzes the fixation of atmospheric CO₂ into organic carbon via the Calvin cycle, using ATP and NADPH generated by light-driven electron transport. Globally, this gross primary production (GPP) fixes approximately 120 GtC/yr on land and roughly 50 GtC/yr in the ocean (mostly via phytoplankton), for a global GPP near 170-200 GtC/yr — the entry point of carbon into the biological cycle, building on the machinery covered under Primary Productivity.

Step 2 — Respiration returns carbon to the atmosphere. Roughly half of terrestrial GPP is immediately released back as CO₂ by the plants' own mitochondria (autotrophic respiration), leaving net primary production (NPP) — the carbon actually available to build new tissue — at about 60 GtC/yr on land. That NPP is then consumed by herbivores, passed through food webs, and ultimately released as CO₂ by heterotrophic respiration in animals and, most importantly in mass terms, by decomposer microbes breaking down Decomposition and Detritus — dead leaves, wood, roots, and animal remains. Aerobic respiration is essentially photosynthesis run in reverse: organic carbon plus O₂ yields CO₂, H₂O, and usable energy (ATP).

Step 3 — Soil carbon storage and slow release. Not all detritus decomposes quickly. Lignin-rich material, carbon bound to soil minerals, and carbon in cold, waterlogged, or anoxic environments (peatlands, permafrost) can persist for centuries to millennia, making soil the largest actively-cycling terrestrial carbon pool — larger than the standing biomass of all plants combined. Permafrost soils alone hold an estimated 1,400-1,600 GtC, more than the atmosphere, currently locked by cold temperatures.

Step 4 — Ocean-atmosphere gas exchange. CO₂ dissolves into and out of the sea surface by simple diffusion, driven by the partial pressure difference between air and water. Once dissolved, CO₂ reacts with water to form carbonic acid, which dissociates into bicarbonate (HCO₃⁻) and carbonate (CO₃²⁻) ions: CO₂ + H₂O ⇌ H₂CO₃ ⇌ H⁺ + HCO₃⁻ ⇌ 2H⁺ + CO₃²⁻ This carbonate equilibrium means the ocean can absorb far more CO₂ than its dissolved-gas content alone would suggest — about 95% of oceanic inorganic carbon exists as bicarbonate, not as dissolved CO₂ gas. This buffering is why the ocean holds ~45 times as much carbon as the atmosphere.

Step 5 — Two ocean "pumps." The solubility pump: CO₂ dissolves more readily in cold water, so at high latitudes cold, CO₂-rich surface water sinks (forming deep water masses), physically transporting dissolved carbon to depth for centuries. The biological pump: phytoplankton fix CO₂ into organic matter at the surface; a fraction of that organic matter (and calcium carbonate shells) sinks as it dies, exporting carbon to the deep ocean and sediments before it can respire back to CO₂ near the surface, where it would quickly re-equilibrate with the atmosphere.

Step 6 — Geological fluxes. Over millions of years, some organic carbon escapes decomposition entirely — buried in anoxic sediments, it is slowly transformed under heat and pressure into kerogen, coal, oil, and natural gas. Marine organisms also draw carbon out of seawater to build calcium carbonate shells and skeletons; when buried, this becomes limestone. These processes are balanced, on geological timescales, by volcanic and metamorphic outgassing of CO₂, and by chemical weathering of silicate rocks, which consumes atmospheric CO₂: CaSiO₃ + CO₂ → CaCO₃ + SiO₂ This silicate weathering reaction is the principal long-term thermostat of Earth's climate: warmer, wetter conditions accelerate weathering, drawing down CO₂ and cooling climate over ~100,000-year timescales — a negative feedback far too slow to buffer the current pulse of fossil-fuel emissions.

Step 7 — Human perturbation. Fossil fuel combustion and cement production move carbon from the essentially closed geological pool into the fast atmospheric pool at roughly 9-10 GtC/yr (as of the 2010s-2020s), with an additional 1-1.5 GtC/yr from land-use change (deforestation reduces standing biomass and soil carbon storage). Natural sinks — land ecosystems and the ocean — currently absorb roughly half of this, each taking up about 25-30% of annual emissions, while the remainder accumulates in the atmosphere, driving the observed rise in CO₂ concentration.

Charles David Keeling and the Mauna Loa record (1958-present). Before Keeling, scientists suspected atmospheric CO₂ might be rising due to fossil fuel combustion but had no continuous, precise measurements to confirm it — earlier 19th- and early 20th-century measurements were scattered, used inconsistent methods, and were plagued by local contamination. Keeling built a highly precise, standardized infrared gas analyzer and began continuous measurements at the Mauna Loa Observatory in Hawaii in 1958, chosen for its remoteness from local pollution sources and high altitude (mixing well with the free atmosphere). The result — the "Keeling Curve" — showed two things unambiguously: a seasonal sawtooth (CO₂ falls each Northern Hemisphere summer as land plant photosynthesis draws it down, and rises each winter as respiration and decomposition dominate, since most of Earth's land and plant biomass is in the Northern Hemisphere), superimposed on a relentless year-over-year upward trend, from about 315 ppm in 1958 to over 420 ppm today. This was the first direct, continuous proof both that the biological carbon cycle actively modulates atmospheric composition seasonally, and that human activity was raising the atmospheric CO₂ baseline. Ice cores later extended this record backward: CO₂ trapped in air bubbles in Antarctic and Greenland ice (Vostok, EPICA Dome C cores) show CO₂ oscillating between roughly 180 ppm (glacial) and 280 ppm (interglacial) over the past 800,000 years — establishing that today's 420+ ppm is unprecedented in that entire record.

Isotopic fingerprinting of the source. How do we know the added carbon comes from fossil fuels specifically, rather than, say, volcanoes or the ocean? Carbon has three isotopes: stable ¹²C and ¹³C, and radioactive ¹⁴C (half-life 5,730 years). Photosynthesis preferentially fixes the lighter ¹²C isotope over ¹³C, so plant-derived organic carbon — and fossil fuels derived from ancient plant and plankton material — is depleted in ¹³C relative to atmospheric CO₂. Since the mid-1800s, measurements show atmospheric CO₂'s ¹³C/¹²C ratio has been steadily declining, exactly as expected if isotopically "light" fossil carbon is being added — a signature volcanic CO₂ does not match. Fossil fuels are also completely devoid of ¹⁴C (having been buried for millions of years, far longer than 10 half-lives), so the measured dilution of atmospheric ¹⁴C over the industrial era (the "Suess effect," first documented by Hans Suess in the 1950s) independently confirms a fossil-derived carbon source. Together these isotopic signatures — not mere correlation with combustion statistics — constitute direct, mechanistic evidence pinpointing fossil fuel combustion as the driver of the CO₂ rise.

Global carbon budget accounting via mass balance. Since the 1990s, international collaborations (culminating in the annual Global Carbon Project "Global Carbon Budget" reports) have combined independent estimates — fossil fuel statistics, land-use change models, atmospheric CO₂ growth rate, and ocean/land carbon uptake models validated against ocean CO₂ partial pressure surveys and terrestrial flux-tower networks (eddy covariance measurements of CO₂ exchange, extending techniques used in Primary Productivity studies) — to close the global carbon budget. The consistent finding across decades of this accounting: roughly half of anthropogenic CO₂ emissions remain in the atmosphere each year, while land and ocean sinks absorb the rest in roughly equal shares — a quantitative confirmation of the fast-cycle fluxes sketched conceptually above, and the basis for all climate projections.

Example 1 — Estimating residence time from pool size and flux. Residence time (τ) is estimated as pool size divided by a flux into or out of that pool: τ = Pool / Flux. Atmospheric CO₂: Pool ≈ 870 GtC. Gross flux out of the atmosphere via photosynthesis (land + ocean) ≈ 120 + 50 = 170 GtC/yr. τ ≈ 870 / 170 ≈ 5.1 years This is the average time an individual CO₂ molecule spends in the atmosphere before being taken up by some photosynthetic organism — a genuinely fast turnover. But this is not the same as how long an "excess" pulse of anthropogenic CO₂ takes to leave the combined atmosphere-ocean-land system, because most of what's fixed by photosynthesis is quickly respired right back (the gross fluxes are nearly balanced, so net removal is much slower). Because carbon cycles rapidly among the fast pools without leaving the fast-cycle system altogether, a pulse of extra CO₂ effectively persists for centuries — some fraction is removed by the ocean within decades, but a residual (~15-20%) remains detectable in the atmosphere for over 1,000 years, since final removal requires slow processes like silicate weathering or deep ocean burial. This is why "residence time of a molecule" (~5 years) and "adjustment time of an emission pulse" (centuries to millennia) are different concepts students often conflate.

Example 2 — Partitioning a national carbon budget. Consider a hypothetical temperate country with the following annual carbon fluxes:

Net anthropogenic carbon flux to the atmosphere = 1.20 + 0.15 − 0.25 + 0.05 = 1.15 GtC/yr Now suppose this country's total land area supports a forest NPP of 0.9 GtC/yr, of which 55% is respired by heterotrophs (decomposers and consumers) within the same year, and the reforestation uptake figure above is already netted against that respiration. The country's fossil emissions (1.20 GtC/yr) exceed its entire net land-based carbon sink capacity (−0.25 + minor soil effects), meaning this country is a strong net carbon source despite active reforestation — illustrating why reforestation efforts, while valuable, typically cannot offset fossil fuel combustion at national policy-relevant scales without also cutting emissions directly. This kind of source/sink partitioning is exactly what national greenhouse gas inventories under the UNFCCC attempt to quantify.

Charles David Keeling (1928-2005) began his measurements almost by accident. As a postdoctoral researcher at Caltech in the mid-1950s, he was investigating how CO₂ concentrations in natural waters related to atmospheric levels, and he needed accurate atmospheric CO₂ data to do it — but found the existing literature values wildly inconsistent, differing by tens of ppm depending on location, time of day, and instrument. Frustrated, he built his own highly precise manometric measurement system and started sampling air at various California sites, quickly noticing a clean, reproducible daily and seasonal cycle that prior sloppy measurements had obscured. This precision-first instinct — insisting on reproducibility before drawing conclusions — is what made his subsequent long-term program scientifically decisive rather than merely suggestive. In 1958, with funding secured by oceanographer Roger Revelle (who had separately warned that humans were conducting a "large scale geophysical experiment" with fossil carbon), Keeling established continuous monitoring stations at Mauna Loa Observatory in Hawaii and at the South Pole. He maintained the Mauna Loa record personally, fighting recurring funding cuts, for the rest of his career — a rare case in modern science of a single, unbroken instrumental time series run essentially by one dedicated investigator and successor (his son, Ralph Keeling, continues the program today at Scripps Institution of Oceanography).

Climate policy and carbon accounting. National greenhouse gas inventories, carbon offset markets, and international agreements (the Paris Agreement's carbon budget framework) all rest on quantifying carbon fluxes exactly as outlined above — separating fossil emissions, land-use change, and biological sink capacity. Understanding that reforestation offsets are finite and reversible (not permanent) directly shapes how offset credits are designed and regulated.

Agriculture and soil management. Practices like no-till farming, cover cropping, and biochar application aim to increase soil organic carbon storage, leveraging the same decomposition and stabilization processes described in Decomposition and Detritus. Because agricultural soils have often lost 30-50% of their original carbon content to cultivation, they represent a genuine (if finite) potential carbon sink, actively studied in "regenerative agriculture" research.

Ocean acidification and marine conservation. As the ocean absorbs roughly a quarter of annual anthropogenic CO₂ emissions, seawater pH has already fallen by about 0.1 units since pre-industrial times (a ~30% increase in hydrogen ion concentration), directly threatening calcifying organisms — corals, shellfish, some plankton — whose shell-building chemistry depends on the carbonate ion concentration described in the mechanism above. This links the carbon cycle directly to the health of the systems covered in Aquatic and Marine Ecosystems.

Permafrost and tipping-point risk. Arctic warming threatens to thaw permafrost, potentially releasing a substantial fraction of its ~1,400-1,600 GtC store as CO₂ and methane — a feedback loop actively monitored via remote sensing and flux towers, since it could add a self-reinforcing amplification to human-caused warming.

Biotechnology and carbon capture. Engineered approaches — enhanced weathering (grinding silicate rock to accelerate the natural CO₂-consuming weathering reaction), direct air capture, and bioenergy with carbon capture and storage (BECCS) — are all attempts to artificially accelerate slow geological carbon removal processes to match the pace of anthropogenic emissions, informed directly by the flux magnitudes and timescales this article has quantified.

Ask the tutor: How does the carbon cycle interact with the nitrogen and phosphorus cycles to limit how much extra CO₂ plants can actually absorb?