Geochemistry
The Silicate Weathering Thermostat That Controls Earth's Climate
Roughly 0.3 gigatons of carbon a year — a rounding error against the 10 Gt humans emit, but the difference between a habitable planet and Venus over geological time — is quietly pulled out of the air by rain dissolving rock. When a raindrop laced with dissolved CO₂ strikes a grain of feldspar or basalt, it launches a reaction whose rate rises with temperature. Heat the planet, and the reaction speeds up and draws down more CO₂; cool it, and the reaction stalls and CO₂ rebuilds. That temperature-sensitive feedback is the silicate weathering thermostat, and it has kept Earth's surface within the liquid-water window for over 4 billion years despite the Sun brightening by ~30%.
The chemistry is deceptively simple: carbonic acid + silicate rock → dissolved bicarbonate + clay, followed by carbonate burial in the sea. But the feedback loop connecting it to global temperature is one of the most important control systems in planetary science.
- Net reactionCaSiO₃ + CO₂ → CaCO₃ + SiO₂
- Main agentH₂CO₃ (carbonic acid), pH ≈ 5.6
- CO₂ drawdown~0.3 Gt C/yr net
- Timescale10⁵–10⁶ years
- WhereSoils, floodplains, mid-ocean ridges
- FeedbackRate ↑ with T (Q₁₀ ≈ 2)
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
The reaction: carbonic acid quietly eating rock
Rain is never pure water. As it falls, it equilibrates with atmospheric CO₂ to make a weak acid:
- Carbonation: CO₂(g) + H₂O ⇌ H₂CO₃ ⇌ H⁺ + HCO₃⁻
At a preindustrial 280 ppm CO₂, this sets rainwater pH at about 5.6 — mildly acidic, and enough to attack silicate minerals over centuries. Soil water is far more aggressive still: root respiration and microbial decay drive soil-air CO₂ to 10–100× atmospheric levels, pushing pore-water H₂CO₃ high enough that most real weathering happens meters below the surface, not in the falling drop.
Take wollastonite (CaSiO₃), the textbook stand-in for all Ca- and Mg-bearing silicates:
- Weathering: CaSiO₃ + 2CO₂ + 3H₂O → Ca²⁺ + 2HCO₃⁻ + H₄SiO₄
Two molecules of atmospheric CO₂ are consumed per calcium ion released. The products — dissolved bicarbonate (HCO₃⁻), calcium, and silicic acid — wash down rivers to the sea. Real rocks are messier: anorthite feldspar weathers to kaolinite clay, CaAl₂Si₂O₈ + 2CO₂ + 3H₂O → Al₂Si₂O₅(OH)₄ + Ca²⁺ + 2HCO₃⁻, and olivine (forsterite) is fastest of all: Mg₂SiO₄ + 4CO₂ + 4H₂O → 2Mg²⁺ + 4HCO₃⁻ + H₄SiO₄.
Closing the loop in the ocean: the Urey reaction
Weathering alone doesn't bury carbon — it just converts atmospheric CO₂ into dissolved bicarbonate. The carbon is only locked away when marine organisms (coccolithophores, foraminifera, corals) precipitate that bicarbonate as calcium carbonate:
- Carbonate precipitation: Ca²⁺ + 2HCO₃⁻ → CaCO₃(s) + CO₂ + H₂O
Notice this releases one CO₂ back to the atmosphere. Add the weathering and precipitation half-cycles together and one net CO₂ is buried per calcium — the reaction Harold Urey wrote down in 1952, read left-to-right as the long-term sink and right-to-left as its return:
- The Urey reaction: CaSiO₃ + CO₂ ⇌ CaCO₃ + SiO₂
The reverse leg runs on geological timescales inside subduction zones: buried carbonate is dragged into the mantle, heated, and metamorphically decarbonated (CaCO₃ + SiO₂ → CaSiO₃ + CO₂), venting CO₂ back through volcanoes. So the full carbonate–silicate cycle is a slow conveyor: volcanoes and mid-ocean ridges exhale CO₂, weathering inhales it, and the two are balanced by temperature.
Why it's a thermostat, not just a sink
A sink that ran at a fixed rate would eventually strip all the CO₂ from the air. What makes weathering a thermostat is that its rate is a rising function of surface temperature — a classic negative feedback first quantified by James Walker, Paul Hays, and James Kasting in 1981.
- Temperature (kinetic): mineral dissolution follows the Arrhenius law, k = A·e−Eₐ/RT, with activation energies of ~50–70 kJ/mol. That gives a Q₁₀ near 2 — a 10 °C rise roughly doubles the dissolution rate.
- The hydrological cycle: a warmer atmosphere holds more water (Clausius–Clapeyron, ~7%/°C) and rains harder, delivering more carbonic acid to more rock.
- CO₂ fertilization: higher CO₂ raises soil-air pCO₂ and lowers pore-water pH, accelerating attack.
Chain it together: rising CO₂ → warming → faster, wetter weathering → more CO₂ drawn down → cooling. The loop closes on itself and pulls temperature back toward equilibrium. Crucially, it acts on a 10⁵–10⁶-year timescale — far too slow to blunt a century of fossil-fuel emissions, but decisive over the age of continents.
The faint young Sun paradox it solves
The Sun is a slowly brightening star. Four billion years ago its luminosity was roughly 70–75% of today's — cold enough that, with a modern atmosphere, Earth's oceans should have been frozen solid. Yet the geological record shows liquid water and life back to at least 3.8 Ga. This is the faint young Sun paradox, and the weathering thermostat is its leading resolution.
On an ice-covered Archean Earth, rain and rivers largely stopped, so silicate weathering nearly halted — but volcanoes kept exhaling CO₂. With the main sink switched off, atmospheric CO₂ climbed for millions of years until the greenhouse effect (possibly aided by CH₄) overwhelmed the dim Sun and thawed the planet. As the Sun brightened over eons, weathering ran a little faster on average, drawing CO₂ steadily down to compensate. The thermostat's set-point drifts, but the water stays liquid. The same mechanism is credited with ending the Snowball Earth glaciations of ~700 Ma: with continents entombed in ice, volcanic CO₂ built to an estimated ~0.1 bar (~300× today) before a runaway greenhouse melted the ice in a geological instant.
What controls the dial: rock type, mountains, and life
Temperature sets the feedback's sensitivity, but the overall weatherability of the planet — how much drawdown you get per degree — depends on the raw material and its exposure:
- Mineral reactivity: weathering rate roughly tracks the strength of the cation–oxygen bond. Olivine and volcanic glass weather orders of magnitude faster than quartz. Basalt (Ca-, Mg-, Fe-rich) is a far better CO₂ sink than granite; the eruption of the Deccan and other flood basalts likely drew down CO₂ and cooled climate.
- Tectonic uplift: mountains expose fresh, unweathered rock and steep slopes that strip weathered rinds before they armor the surface. The rise of the Himalaya ~40 Ma exposed vast reactive silicate and is a leading suspect for the Cenozoic slide from hothouse to icehouse.
- Kinetic vs. supply limits: in flat, stable terrain, weathering becomes transport-limited — a thick clay soil (kaolinite, gibbsite) shields the parent rock, so more reaction shifts to whichever fresh mineral is exposed. Uplift and erosion keep the reactor fed.
- Biology: plant roots and mycorrhizal fungi secrete organic acids and pump CO₂ into soils; the spread of land plants in the Devonian is thought to have amplified weathering enough to help draw CO₂ down by an order of magnitude.
The scale, and where it falls short today
Rivers carry the thermostat's fingerprint: the world's rivers deliver on the order of 0.3 Gt of carbon per year as bicarbonate sourced from silicate weathering — the net long-term drawdown. That is genuinely large on geological timescales, but it is ~1/30th of the ~10 Gt C/yr humans now emit from fossil fuels. The natural thermostat simply cannot respond fast enough: it will eventually neutralize the anthropogenic CO₂ spike, but over hundreds of thousands of years, long after the climate consequences have played out.
That timescale mismatch is exactly why some scientists want to speed the thermostat up on purpose. Enhanced weathering spreads finely milled fast-weathering silicate — typically basalt or olivine dust — across farmland and coastlines. Grinding rock to micron scale multiplies its reactive surface area by orders of magnitude, so decades of natural weathering can be compressed into years:
- Mg₂SiO₄ + 4CO₂ + 4H₂O → 2Mg²⁺ + 4HCO₃⁻ + H₄SiO₄ (each ton of olivine can capture on the order of ~1 ton of CO₂ as durable ocean bicarbonate)
The bicarbonate that reaches the sea also counteracts ocean acidification by adding alkalinity — a rare two-for-one intervention. The catch is scale, energy for grinding and transport, and trace-metal (Ni, Cr) release from olivine. It is the same reaction that has run Earth's climate for eons, just dialed up by human hands.
| Property | Silicate weathering | Carbonate weathering |
|---|---|---|
| Example mineral | Wollastonite CaSiO₃, feldspar, olivine | Calcite / limestone CaCO₃ |
| Weathering reaction | CaSiO₃ + 2CO₂ + 3H₂O → Ca²⁺ + 2HCO₃⁻ + H₄SiO₄ | CaCO₃ + CO₂ + H₂O → Ca²⁺ + 2HCO₃⁻ |
| CO₂ consumed per Ca²⁺ | 2 mol (from atmosphere) | 1 atmospheric + 1 from rock |
| After marine burial as CaCO₃ | Net −1 CO₂ (long-term sink) | Net 0 CO₂ (no long-term change) |
| Climate role | The thermostat | Fast buffer, climate-neutral over Myr |
Frequently asked questions
Why doesn't carbonate (limestone) weathering count as a climate thermostat too?
Carbonate weathering, CaCO₃ + CO₂ + H₂O → Ca²⁺ + 2HCO₃⁻, uses one atmospheric CO₂ and one from the rock. When that Ca²⁺ later reprecipitates as CaCO₃ in the ocean, it releases one CO₂ back — the books balance to net zero over the full cycle. Only silicate weathering pulls the carbon that was originally atmospheric permanently into buried rock, so only it acts as a long-term sink.
Is the reaction reversible?
Yes, but on very different timescales. The forward Urey reaction (weathering plus marine carbonate burial) draws CO₂ down over 10⁵–10⁶ years. The reverse runs in subduction zones, where buried CaCO₃ is heated and metamorphically decarbonated (CaCO₃ + SiO₂ → CaSiO₃ + CO₂), venting CO₂ back through volcanoes over tens of millions of years. Balance between the two is what the thermostat regulates.
How fast is the thermostat — can it fix human-caused warming?
Not on any timescale that helps us. The silicate weathering feedback operates over 100,000 to 1,000,000 years. It will eventually neutralize today's fossil-fuel CO₂, but the current ~0.3 Gt C/yr natural drawdown is roughly 30 times slower than human emissions. The thermostat is a geological safeguard, not a policy tool.
What is the faint young Sun paradox and how does weathering solve it?
Early in Earth's history the Sun was 25–30% dimmer, so the planet should have been frozen — yet liquid water existed 3.8 billion years ago. The weathering thermostat resolves this: whenever Earth cooled, weathering slowed, volcanic CO₂ accumulated, and the stronger greenhouse compensated for the faint Sun. As the Sun brightened, weathering ran faster and drew CO₂ back down, keeping surface temperatures in the liquid-water range throughout.
Does the type of rock matter?
Enormously. Weathering rate scales with mineral reactivity, and Ca/Mg-rich volcanic rocks weather far faster than quartz-rich granite. Basalt and especially olivine are the strongest natural CO₂ sinks, which is why the emplacement of large basalt provinces and the exposure of fresh rock by mountain uplift (like the Himalaya) can shift global climate over millions of years.
Can we deliberately accelerate silicate weathering to capture CO₂?
Yes — that's enhanced weathering. Spreading finely ground basalt or olivine on fields and beaches multiplies reactive surface area so the natural reaction (e.g. Mg₂SiO₄ + 4CO₂ + 4H₂O → 2Mg²⁺ + 4HCO₃⁻ + H₄SiO₄) runs in years instead of centuries, capturing roughly a ton of CO₂ per ton of olivine while adding alkalinity that fights ocean acidification. Limits are the energy to grind rock, transport logistics, and trace-metal release.