The Carbonate Buffer That Steadies the Sea

The four-species equilibrium that runs the ocean

Dissolve carbon dioxide in seawater and it does not simply sit there. It enters a linked chain of acid–base equilibria that together are called the carbonate system. Reading left to right, the reactions are:

• Gas dissolution (Henry's law): CO₂(g) ⇌ CO₂(aq)
• Hydration: CO₂(aq) + H₂O ⇌ H₂CO₃  — slow, and only ~0.3% of dissolved CO₂ ever becomes true carbonic acid
• First dissociation (K₁): H₂CO₃ ⇌ H⁺ + HCO₃⁻, pK₁ ≈ 6.0 in seawater
• Second dissociation (K₂): HCO₃⁻ ⇌ H⁺ + CO₃²⁻, pK₂ ≈ 9.1 in seawater

Chemists lump the dissolved CO₂ and the tiny H₂CO₃ pool together as H₂CO₃*. The sum of all three carbon-bearing species — CO₂(aq)+H₂CO₃, HCO₃⁻, and CO₃²⁻ — is the dissolved inorganic carbon (DIC), typically ≈ 2000–2200 µmol/kg in surface water. Because surface pH (~8.1) sits neatly between pK₁ and pK₂, the system parks almost all of its carbon in the middle species: about 90% HCO₃⁻, 9% CO₃²⁻, and only ~1% dissolved CO₂. That lopsided distribution is the whole secret. The ocean is a bicarbonate solution.

Why bicarbonate is a shock absorber for protons

A buffer resists pH change because it holds a large reservoir of both a weak acid and its conjugate base. When you add acid, the base mops up the extra H⁺; when you add base, the acid releases H⁺. Seawater does this with the HCO₃⁻/CO₃²⁻ pair. Add fossil-fuel CO₂ and the incoming protons are largely neutralised by carbonate ion:

• The master reaction: CO₂ + H₂O + CO₃²⁻ ⇌ 2 HCO₃⁻

Notice what this does: one molecule of CO₂ is converted almost entirely into two bicarbonate ions by consuming a carbonate ion, and only a sliver of free H⁺ is released. The pH barely moves because the reaction destroys a strong proton-acceptor (CO₃²⁻) and makes a weak one (HCO₃⁻). This is exactly the Henderson–Hasselbalch logic: pH = pK₂ + log([CO₃²⁻]/[HCO₃⁻]). With ~90% of DIC as HCO₃⁻, the ratio is large and stiff, so it takes an enormous proton load to shift the log term.

The price of that stability is carbonate depletion. Every CO₂ absorbed eats a CO₃²⁻ ion. Since preindustrial times the surface-ocean carbonate concentration has already fallen by roughly 10–20%. The buffer keeps pH steady precisely by spending down the very ion that shell-builders need.

Alkalinity: the conserved quantity that anchors everything

The reason the buffer is so robust — and so precisely predictable — is a conserved bookkeeping quantity called total alkalinity (TA). Alkalinity is the ocean's capacity to neutralise acid, defined as the excess of proton acceptors over donors. To a good approximation:

• TA ≈ [HCO₃⁻] + 2[CO₃²⁻] + [B(OH)₄⁻] + [OH⁻] − [H⁺]

Carbonate counts twice because it can accept two protons. In the open ocean TA ≈ 2300 µmol/kg, set over millions of years by the balance between silicate and carbonate rock weathering on land (which delivers Ca²⁺, Mg²⁺, and HCO₃⁻ down rivers) and carbonate burial in sediments. Critically, when CO₂ dissolves it does not change alkalinity — CO₂ is a neutral molecule that adds no net charge. It changes DIC but leaves TA fixed. Knowing any two of the four master variables (pH, DIC, TA, pCO₂) lets you solve for the other two exactly, which is how oceanographers reconstruct pH from ships that only measured alkalinity and carbon. The fixed alkalinity is the fulcrum the whole buffer pivots on.

The Revelle factor: why the sponge is stiffening

Here is the counterintuitive heart of the problem. You might expect that a bigger ocean carbon reservoir means more capacity to soak up CO₂. In fact the buffer's efficiency is governed by the Revelle factor (β), the ratio of the fractional change in seawater pCO₂ to the fractional change in DIC:

• Revelle factor = (ΔpCO₂/pCO₂) / (ΔDIC/DIC) ≈ 10–15

A Revelle factor of ~10 means that to raise DIC by just 1%, the water's CO₂ partial pressure must rise by ~10%. In blunt terms: the ocean stores far less added CO₂ than its huge carbon content suggests, because most incoming CO₂ is immediately re-emitted to equilibrium rather than locked away. Roger Revelle and Hans Suess pointed this out in 1957, and it is the reason the ocean cannot simply swallow all our emissions. Worse, the Revelle factor rises as CO₂ climbs — as carbonate ion is consumed, the buffer stiffens and takes up proportionally less. Cold high-latitude waters (Southern Ocean, subpolar North Atlantic) already have Revelle factors near 15, making them the first regions to lose buffering capacity and the first to see corrosive undersaturated conditions.

Temperature, pressure, and the deep-sea saturation horizon

The buffer is not uniform. Its equilibrium constants shift with temperature, pressure, and salinity, which is why the deep ocean behaves differently from the sunlit surface.

• Temperature: CO₂ is more soluble in cold water (Henry's law), so polar seas hold more DIC and are naturally lower in carbonate and less buffered.
• Pressure: The solubility product of calcium carbonate, K_sp, increases with pressure. Below a depth called the calcite (or aragonite) saturation horizon — roughly 4–5 km for calcite in the Atlantic, shallower in the Pacific — CaCO₃ becomes undersaturated and dissolves.
• Biology: Sinking organic matter is respired back to CO₂ at depth (CH₂O + O₂ → CO₂ + H₂O), lowering pH and dissolving carbonate — the engine of the biological pump.

The relevant saturation is expressed as Ω = [Ca²⁺][CO₃²⁻]/K_sp. Where Ω > 1 shells form; where Ω < 1 they dissolve. Surface waters today sit at Ω ≈ 3–4 for aragonite in the tropics but already dip below 1 in parts of the Arctic and Southern Ocean. The reaction that dissolves a coral or a pteropod shell is simply the master buffer reaction run in reverse: CaCO₃ + CO₂ + H₂O → Ca²⁺ + 2 HCO₃⁻. On the timescale of millennia this seafloor carbonate dissolution is the ocean's ultimate defence, neutralising excess acid by consuming limestone — but it operates far too slowly to save this century's reefs.

The quantitative scale, and what it means for life

The numbers frame the stakes. The ocean holds about 38,000 gigatonnes of carbon as DIC — roughly 45 times the atmosphere's ~880 Gt. It has absorbed on the order of 170 Gt of anthropogenic carbon since 1850, which is why atmospheric CO₂ (≈ 420 ppm) is lower than it would otherwise be. In doing so, surface pH fell ~0.1, hydrogen-ion activity rose ~30%, and carbonate ion dropped measurably.

For the biosphere, the casualty is calcification. Corals, coccolithophores, foraminifera, pteropods, oysters, and mussels all build skeletons from CaCO₃, and their construction reaction — Ca²⁺ + 2 HCO₃⁻ → CaCO₃ + CO₂ + H₂O — runs harder and costs more energy as carbonate becomes scarce. Larval oysters in the U.S. Pacific Northwest already suffered mass die-offs in hatcheries when upwelled, low-Ω water reached the coast; growers now monitor aragonite saturation in real time and buffer their intake water with soda ash. That is a single industry adapting to a buffer that, at the planetary scale, is slowly running out of room.

Is the change reversible, and can we help the buffer?

On human timescales, largely no. The chemistry itself is reversible — stop emitting CO₂ and surface pH would recover over decades to centuries as the ocean re-equilibrates with a lower-CO₂ atmosphere. But the CO₂ already dissolved will keep penetrating into the deep ocean for centuries, and full neutralisation by seafloor carbonate dissolution takes ~5,000–10,000 years, with silicate weathering cleanup running to hundreds of thousands. The paleo-record confirms the sluggishness: the Paleocene–Eocene Thermal Maximum (PETM) ~56 million years ago dissolved deep-sea carbonate worldwide and took roughly 100,000 years to recover.

Proposed interventions all work by adding alkalinity — the one lever that shifts the buffer toward stability without simply removing carbon:

• Ocean alkalinity enhancement: grinding and spreading olivine or basalt (enhanced weathering), which reacts as Mg₂SiO₄ + 4 CO₂ + 4 H₂O → 2 Mg²⁺ + 4 HCO₃⁻ + H₄SiO₄, raising TA and drawing down CO₂.
• Electrochemical or lime-based additions of hydroxide/carbonate to coastal water.

These would push the ocean the way natural rock weathering does — just fast enough to matter. Whether they can scale without new side effects is the open question, but the underlying chemistry is exactly the buffer's own recovery mechanism, sped up.

Speciation of dissolved inorganic carbon in surface seawater vs. a hypothetical unbuffered CO₂ solution

Property	Surface seawater (pH 8.1)	Pure water + CO₂ (unbuffered)
Dominant carbon species	HCO₃⁻ (~90%)	H₂CO₃ / dissolved CO₂
CO₃²⁻ fraction	~9%	negligible
Dissolved CO₂ fraction	~1%	~100%
Total alkalinity	≈ 2300 µmol/kg	≈ 0
pH resulting from ~420 ppm CO₂	≈ 8.1	≈ 5.6
Response to added acid	Strongly resisted (buffered)	pH drops immediately