Earth’s Geologic Carbon Cycle

Revised and adapted from ‘The Global Carbon Cycle’, David Archer (2010)

In the previous two chapters we learned the fundamental physics that explains the greenhouse effect and why \(\text{CO}_2\) is particularly important: a strong absorption band of \(\text{CO}_2\) happens to sit right in the middle of Earth’s emission spectrum. So changes in \(\text{CO}_2\) can have a big radiative impact on Earth. But how much carbon or \(\text{CO}_2\) is in the Earth and its atmosphere? And how has \(\text{CO}_2\) influenced Earth’s climate in the past?

Earth’s Carbon Reservoirs

\(\text{CO}_2\) is a trace constituent in the atmosphere, comprising only about 0.0415% of all the gas molecules (which works out to 415 ppm, or parts per million). If all the \(\text{CO}_2\) in the atmosphere were to solidify into dry ice, the snowfall would be only about 10 cm deep. The atmosphere currently contains about 780 Gton C (Gton is ‘giga-ton’, where 1 Gton C = \(10^{15}\) g C).

The atmosphere acts as a kind of Grand Central Station or King’s Cross Station with respect to the carbon cycle: the other reservoirs interact with each other primarily by trading carbon through the atmosphere, even though the atmosphere holds only a tiny fraction of Earth’s carbon. Earth and Venus contain approximately the same amount of carbon, yet it’s distributed completely differently: most of the carbon on Venus is in its atmosphere while most of Earth’s carbon is in its rocks. Earth has liquid water that can weather rocks and take \(\text{CO}_2\) out of the atmosphere and into the solid Earth. Venus lost its water early on, however, and its climate has never returned to its likely moderate initial conditions.

The land biosphere is the most visible part of the carbon cycle to us, and it holds much more carbon than the ocean biosphere does (e.g., trees are much larger than single-celled plankton). The land surface stores carbon in organic form in living things and even more abundantly as the organic carbon remains of plants in soils. Grasslands accumulate a lot of organic carbon in their soils, so that the total amount of carbon per acre is about equal to that in forests, even though the carbon is more obvious in the forests. There is about as much living carbon on land as there is atmospheric carbon, perhaps 500 Gton C. The amount of carbon in Earth’s soil depends on how deep the boundary is between soil and the underlying rocks, where soils might be affected by changes in climate but the carbon below that is out of reach. The usual soil depth is assumed to be one meter, resulting in a soil carbon pool of about 1,500 Gton C as soil carbon, twice as much as is in the atmosphere.

The oceans contain about fifty times more carbon than the atmosphere does, about 38,000 Gton C. Most of the carbon in the ocean is in the inorganic forms: dissolved \(\text{CO}_2\), carbonic acid (\(\text{H}_2\text{CO}_3\)), bicarbonate ion (\(\text{HCO}_3^{–}\)), and carbonate ion (\(\text{CO}_3^{=}\)).

Sedimentary rocks contain most of the Earth’s carbon, in chemical forms of limestone (\(\text{CaCO}_3\)) and organic carbon, mostly in the form of a random indescribable goo called kerogen. These total to approximately 1,200,000 Gton C. The sediments that are now rocks were originally deposited in water but are now found over most of the surface of the Earth, including the highest mountaintops. Fossil fuels make up only a small fraction of the buried organic carbon in the Earth, but even so, there is enough carbon to knock the carbon cycle significantly out of whack. Fossil fuels alone total about 5000 Gton C.

Fossil fuel combustion takes carbon that was sleeping in sedimentary rocks and injects it into the atmosphere. Most of the fossil carbon is coal, and of that there is enough to increase the \(\text{CO}_2\) concentration of the atmosphere to about ten times its natural concentration, if it were all released at the same time.

The Atmospheric CO₂ Thermostat

Earth’s climate has navigated a path that stayed within a relatively narrow temperature range, the freezing and boiling points of water, since the very first sedimentary rocks appeared, shortly after the birth of the Earth. Meanwhile, the heat source for the surface of the Earth, the sun, has gotten 25% brighter over geologic time. This ‘faint young sun paradox’ was first noted by the astronomer Carl Sagan. Sagan and Mullen, Earth and Mars: Evolution of atmospheres and surface temperatures, (1972).

Part of the explanation of this stability has to do with the carbon cycle and a mechanism called the weathering-\(\text{CO}_2\) thermostat, which stabilizes Earth’s climate by regulating the \(\text{CO}_2\) concentration in the atmosphere. This mechanism is based on a chemical reaction called the Urey reaction Urey, The planets, their origin and development, (1952).: \[\text{CaSiO}_3 + \text{CO}_2 \longleftrightarrow \text{CaCO}_3 + \text{SiO}_2 \; . \tag{1}\] The first compound, \(\text{CaSiO}_3\) , is the simplest possible chemical formula for igneous rocks. Most real rocks have other elements in them, such as aluminum and sodium, and much more complicated chemical formulas than \(\text{CaSiO}_3\). This equation operates in the world by igneous rocks reacting with atmospheric \(\text{CO}_2\), which ultimately winds up in the chemical form of \(\text{CaCO}_3\) in sediments of the ocean. Chemical weathering, the chemical breaking down of rocks, acts as a carbon sink (the opposite of a source) and provides a pathway for carbon to exit the fast surface carbon cycle into the solid Earth. This process represents the Urey reaction working from left to right: \(\text{CaSiO}_3 + \text{CO}_2 \rightarrow \text{CaCO}_3 + \text{SiO}_2\).

The main idea behind the atmospheric \(\text{CO}_2\) thermostat is that when the temperature of the Earth rises, it accelerates the hydrologic cycle (the amount of freshwater running off the continents) and therefore the rate of chemical weathering.Walker, Hays, and Kasting, A negative feedback mechanism for the long-term stabilization of Earth’s surface temperature, (1981). This effect is due to the fact that warmer air can hold more water vapor. More water vapor generates more rain when it gets squeezed out of the air as it rises and cools. Runoff rates on Earth today are much higher in the warm tropics than in the cooler high latitudes. Holland, The chemistry of the atmosphere and oceans, (1978). Everything else being equal, the weathering carbon sink would be stronger on a warmer planet than a cooler one.

The carbon sink from chemical weathering is balanced in a steady state by a carbon source, the Urey reaction operating from right to left: \(\text{CaSiO}_3 + \text{CO}_2 \leftarrow \text{CaCO}_3 + \text{SiO}_2\). This process comes from the melting down of carbonate rocks in the solid Earth and \(\text{CO}_2\) coming to the surface in volcanic gases and hot spring fluids.

This figure summarizes the source and sink components of the weathering-\(\text{CO}_2\) thermostat:

The combination of independent source and concentration-dependent sink for \(\text{CO}_2\) makes for a stable negative feedback system that ultimately controls the \(\text{CO}_2\) concentration and thus the Earth’s temperature. For example, if the atmospheric \(\text{CO}_2\) concentration were too high, weathering would consume more \(\text{CO}_2\) than could be replenished by the volcanic carbon source, and the atmospheric \(\text{CO}_2\) concentration would drop. The planet would cool, slowing the hydrologic cycle, and decreasing chemical weathering. Eventually, after a few hundred thousand years, the weathering sink would return to balance with the \(\text{CO}_2\) source.

The processes of the weathering-\(\text{CO}_2\) thermostat are sketched out below. For \(\text{CO}_2\) and temperature perturbations (yellow), a corresponding hydrological-weathering response (red or blue) will try to bring Earth’s climate back into balance (asterisk).

The weathering \(\text{CO}_2\) thermostat is at least partly responsible for the stability of Earth’s temperature through geologic time, even as the heat output of the sun gradually increases (Carl Sagan’s faint young sun paradox). Geologists find sedimentary rocks from the very earliest days of the history of the Earth, indicating that the hydrological cycle has continuously functioned and thus that Earth’s temperature has managed to stay between the boiling and freezing temperatures of water for this entire time (although there may be a temporary exception to this rule in the ‘snowball Earth’ scenario).

It’s not clear that the weathering \(\text{CO}_2\) thermostat can explain the entire span of Earth’s history—the sun was so cool during the earliest days of Earth’s history that some other greenhouse gas, perhaps methane or ammonia, might have been helping keep the Earth warm. But the \(\text{CO}_2\) thermostat serves as a sufficient explanation of the stability, and the vagaries, of global climate for the Phanerozoic era, the time of formation of the fossil-bearing deposits from the last 600 million years into the present, about which we know a lot more than about earlier times.

The strongest evidence in support of a weathering \(\text{CO}_2\) thermostat mechanism is the stability of the \(\text{CO}_2\) concentration in the atmosphere through the last 800,000 years, for which we have ice core \(\text{CO}_2\) data. If the \(\text{CO}_2\) sources (volcanoes) and sinks (weathering) were out of balance by 10%, they could double the \(\text{CO}_2\) concentration in the air in about 50,000 years. There are glacial-interglacial cycles in \(\text{CO}_2\) concentration (below), but there is no long-term trend upward or downward in the concentrations. On long time averages, the fluxes balance to within 1-2%.Zeebe and Caldeira, Close mass balance of long-term carbon fluxes from ice-core \(\text{CO}_2\) and ocean chemistry records, (2008).

The Slowest Climate Changes

The weathering \(\text{CO}_2\) thermostat tries to maintain the \(\text{CO}_2\) concentration of the atmosphere, and the climate, at some set-point, but there are ways in which factors such as the placement of the continents on the Earth’s surface and the creation of mountains by tectonic plate collisions can alter the set-point of the thermostat.Berner, The Phanerozoic carbon cycle: \(\text{CO}_2\) and \(\text{O}_2\), (2004). On time scales of tens and hundreds of millions of years, the temperature of the Earth drifts up and down between ‘icehouse’ and ‘hothouse’ climates: icehouse climates have large ice sheets at or near both poles and ‘hothouse’ climates do not have permanent ice and are when Earth remains, on an annual average, above freezing even at the poles.

Mountains tend to accelerate the global rate of chemical weathering by exposing fresh igneous rock to chemical scouring by freshwater. When igneous rocks weather, not all of the atoms are soluble in water. Some recrystallize back out of solution, forming another class of mineral called clay minerals. Clay minerals tend to have more water in their chemical structure than igneous rock minerals do, and they are softer. The clays left behind form the basis for the soils that cover most of the land’s surface. Soils impede weathering by limiting contact between the rocks and the water.

Tectonic uplift, driven by plate collisions, therefore accelerates the rate of chemical weathering. But in steady state, after the weathering-\(\text{CO}_2\) thermostat has had time to equilibrate, the total rate of \(\text{CO}_2\) consumption by weathering on Earth needs to balance the rate of \(\text{CO}_2\) emerging from the Earth in volcanoes and hot springs. The end result of uplift, then, is to pull the atmospheric \(\text{CO}_2\) concentration to a lower value, cooling the planet. The increased ‘weatherability’ of a planet with mountains on it will be compensated for by a cooler climate and lower atmospheric \(\text{CO}_2\) concentration. One reason for a generally cooling climate of the Earth over the last tens of millions of years (orange arrow over temperature reconstruction below), since the last ‘hothouse’ era, about 50 million years ago, is the uplift of the Himalayas, driven by the collision of the Indian and Asian tectonic plates.

There is more runoff in lower latitudes, so if by chance the continents of the Earth find themselves mostly in the low latitudes near the equator, it would tend to stimulate weathering and draw down atmospheric \(\text{CO}_2\). Flooding and exposure of the continents by fluctuating sea levels also affects weathering, because weathering happens only on land, not under the oceans. On million-year time scales, global sea levels seem to rise and fall by hundreds of meters, as reconstructed by the area of the continental crust that is covered in ocean water, the way the continental shelves ringing most land areas are covered today. It is not clear why or how sea level varies this much, but it could be driven by changes in the average density of ocean crust, driven perhaps by the rate of seafloor spreading, or it could be that there is enough water dissolved in the minerals of the mantle to affect sea level as the water flows in and out of the solid Earth.

Plants and trees have the capacity to change the weatherability of the land surface, and therefore the set-point of the \(\text{CO}_2\) thermostat and the climate of the Earth. Plants take up \(\text{CO}_2\) through their leaves and release most of it below the soil surface, raising the soil gas \(\text{CO}_2\) concentration to many times higher than the atmospheric concentration.

Since \(\text{CO}_2\) is on the left-hand, ‘reactant’ side of the Urey reaction, higher \(\text{CO}_2\) concentrations should push the equilibrium chemical reaction to the right, dissolving more igneous rock. Plants, by pumping \(\text{CO}_2\) into the soils, might thereby be accelerating the rate of chemical weathering. By stimulating the chemical weathering reaction, the conquest of the land surface by plants in the Silurian (ending about 416 million years ago) may have acted to cool the planet.

The rate of \(\text{CO}_2\) degassing from the Earth is also likely to vary through time. Most of the carbon is assumed to be recycled, from sedimentary rocks that subducted and ultimately released their \(\text{CO}_2\) in the reverse, right-to-left direction of the Urey reaction, \(\text{CaSiO}_3 + \text{CO}_2 \leftarrow \text{CaCO}_3 + \text{SiO}_2\), this transformation is called metamorphism because it takes place at temperatures at which the rocks begin to deform and change chemically into what are called metamorphic rocks. Marble, for example, is metamorphosed \(\text{CaCO}_3\).

It appears, however, to take exceptional circumstances to drive \(\text{CaCO}_3\) into a subduction zone. Subduction zones tend to be found in the deepest parts of the ocean, where the crust has cooled, become denser, and sunk further down into the fluid-like mantle of the Earth. Yet \(\text{CaCO}_3\) tends to deposit in shallower and mid-depth waters, because it usually dissolves under the higher pressure in the deep ocean. So in order for \(\text{CaCO}_3\) to be subducted efficiently, the subduction must be of a relatively shallow ocean floor. One such time this happened was when the Indian plate collided with Asia (red arrow below), squeezing the \(\text{CaCO}_3\)-rich sediments of the shallow Tethys Sea off the face of the Earth:

The resulting \(\text{CO}_2\) flux into the atmosphere may have been the reason for the warm early Eocene climate Edmond and Huh, Non-steady-state carbonate recycling and implications for the evolution of atmospheric pCO2, (2003). (circled in orange below).

There is a computer model of the geologic carbon cycle, developed over decades by Bob Berner, called Geocarb. Berner and Kothavala, Geocarb III: A Revised Model of Atmospheric \(\text{CO}_2\) over Phanerozoic Time, (2001). The drivers that Geocarb is most sensitive to are the plant invasion of land in the Silurian; the rate of seafloor spreading, which controls sea level and \(\text{CO}_2\) degassing; and tectonic effects such as mountain building and the geographic distribution of the continents. Over the past 600 million years, the Geocarb model predicts high \(\text{CO}_2\) concentrations in eras of Earth’s history that geologists know to be warm, and low \(\text{CO}_2\) concentrations in glaciated times:

There are no ice core \(\text{CO}_2\) concentration measurements from millions of years ago. Estimates of past \(\text{CO}_2\) concentrations from these times are based on more indirect, ‘proxy’ data rather than on direct measurement. Some proxies used to estimate past \(\text{CO}_2\) concentrations include: the number of vents in the waxy seal, called stomata, on the bottoms of fossilized leaves; the carbon isotopic composition of \(\text{CaCO}_3\) that forms in arid soils; the carbon isotopic compositions of particular biomolecules; and the isotopic composition of boron atoms trapped in \(\text{CaCO}_3\). None of the available proxies for pre-ice core atmospheric \(\text{CO}_2\) concentration are completely bullet-proof, but it is reassuring that most of the available credible proxy techniques seem to give the same general answers about when \(\text{CO}_2\) levels in the past were higher than today. Over at least the past 66 million years, temperatures have generally followed the changes in \(\text{CO}_2\):

The weathering-\(\text{CO}_2\) thermostat operates on the order of hundreds of thousands of years. If the rate of \(\text{CO}_2\) degassing from Earth were to increase suddenly, it would therefore take hundreds of thousands of years for \(\text{CO}_2\) and the climate to fully respond to that change. The grand climate shifts between ‘hothouse’ and ‘icehouse’ in the geologic past generally took place on time scales much longer than the thermostat response time, drifting up and down over tens of millions of years. The pacing of these climate shifts is therefore set by the time scales of plate tectonics, the mountain building and soil formation, the expansion of land plants, and changes in sea level and continental area, rather than by how fast the weathering-\(\text{CO}_2\) thermostat can respond.

\(\Uparrow\) To the top

\(\Rightarrow\) Next chapter

\(\Leftarrow\) Table of contents

Earth’s Geologic Carbon Cycle

Earth’s Carbon Reservoirs

The Atmospheric CO2 Thermostat

The Slowest Climate Changes

The Atmospheric CO₂ Thermostat