So far we have viewed the concentrations of species in the atmosphere as controlled by emissions, transport, chemistry, and deposition. From an Earth system perspective, however, the composition of the atmosphere is ultimately controlled by the exchange of elements between the different reservoirs of the Earth. In the present chapter we examine atmospheric composition from this broader perspective, and focus more specifically on the biogeochemical factors that regulate the atmospheric abundances of N2, O2, and CO2.
The Earth system (including the Earth and its atmosphere) is an assemblage of atoms of the 92 natural elements. Almost all of these atoms have been present in the Earth system since the formation of the Earth 4.5 billion years ago by gravitational accretion of a cloud of gases and dust. Subsequent inputs of material from extraterrestrial sources such as meteorites have been relatively unimportant. Escape of atoms to outer space is prevented by gravity except for the lightest atoms (H, He), and even for those it is extremely slow ( problem 6. 3 ). Thus the assemblage of atoms composing the Earth system has been roughly conserved since the origin of the Earth. The atoms, in the form of various molecules, migrate continually between the different reservoirs of the Earth system ( Figure 6-1 ). Geochemical cycling refers to the flow of elements through the Earth's reservoirs; the term underlines the cyclical nature of the flow in a closed system.
The standard approach to describing the geochemical cycling of elements between the Earth's reservoirs is with the box models that we introduced previously in chapter 3. In these models we view individual reservoirs as "boxes", each containing a certain mass (or "inventory") of the chemical element of interest. The exchange of the element between two reservoirs is represented as a flow between the corresponding boxes. The same concepts developed in chapter 3 for atmospheric box models can be applied to geochemical box models. The reservoirs may be those shown in Figure 6-1 or some other ensemble. Depending on the problem at hand, one may want a more detailed categorization (for example, separating the hydrosphere into oceans and freshwater), or a less detailed categorization (for example, combining the biosphere and soil reservoirs).
Most of the mass of the Earth system is present in the deep Earth, but this material is largely isolated from the surface reservoirs: atmosphere, hydrosphere, biosphere, soil, and lithosphere ( Figure 6-1 ). Communication between the deep Earth and the surface reservoirs takes place by volcanism and by subduction of tectonic plates, processes that are extremely slow compared to those cycling elements between the surface reservoirs. The abundance of an element in the atmosphere can therefore be viewed as determined by two separable factors: (a) the total abundance of the element in the ensemble of surface reservoirs, and (b) the partitioning of the element between the atmosphere and the other surface reservoirs.
Supply of elements to the surface reservoirs was set in the earliest stages of Earth's history when the crust formed and volcanic outgassing transferred material from the deep Earth to the atmosphere. The early Earth was a highly volcanic place due to energy released in its interior by radioactive decay and gravitational accretion. Present-day observations of volcanic plumes offer an indication of the composition of the outgassed material. It was mostly H2O (~95%), CO2, N2, and sulfur gases. There was no O2; volcanic plumes contain only trace amounts of O2, and examination of the oldest rocks on Earth show that they formed in a reducing atmosphere devoid of O2. The outgassed water precipitated to form the oceans. Carbon dioxide and the sulfur gases then dissolved in the oceans, leaving N2 as the dominant gas in the atmosphere. The presence of liquid water allowed the development of living organisms (self-replicating organic molecules). About 3.5 billion years ago, some organisms developed the capacity to convert CO2 to organic carbon by photosynthesis. This process released O2 which gradually accumulated in the atmosphere, reaching its current atmospheric concentration about 400 million years ago.
It is instructive to compare the evolution of the Earth's atmosphere to that of its neighbor planets Venus and Mars. All three planets presumably formed with similar assemblages of elements but their present-day atmospheric compositions are vastly different ( See Atmospheres of Venus, Earth, and Mars ). Venus has an atmosphere ~100 times thicker than that of Earth and consisting mostly of CO2. Because of the greater proximity of Venus to the Sun, the temperature of the early Venus was too high for the outgassed water to condense and form oceans (see section 7.5 for further discussion). As a result CO2 remained in the atmosphere. Water vapor in Venus's upper atmosphere photolyzed to produce H atoms that escaped the planet's gravitational field, and the O atoms left behind were removed by oxidation of rocks on the surface of the planet. This mechanism is thought to explain the low H2O concentrations in the Venusian atmosphere. On Earth, by contrast, the atmosphere contains only 10-5 of all water in the surface reservoirs (the bulk is in the oceans) so that loss of water to outer space is extremely slow and is compensated by evaporation from the oceans.
The atmosphere of Mars is much thinner than that of the Earth and consists principally of CO2. The smaller size and hence weaker gravitational field of Mars allows easier escape of H atoms to outer space, and also allows escape of N atoms produced in the upper atmosphere by photolysis of N2 (escape of N atoms is negligible on Earth). Although the mixing ratio of CO2 in the Martian atmosphere is high, its total abundance is in fact extremely low compared to that of Venus. It is thought that CO2 has been removed from the Martian atmosphere by surface reactions producing carbonate rocks.
Nitrogen (as N2) accounts for 78% of air on a molar basis. Figure 6-2 presents a summary of major processes involved in the cycling of nitrogen between surface reservoirs. Nitrogen is an essential component of the biosphere (think of the amino acids) and the atmosphere is an obvious source for this nitrogen. Conversion of the highly stable N2 molecule to biologically available nitrogen, a process called fixation, is difficult. It is achieved in ecosystems by specialized symbiotic bacteria which can reduce atmospheric N2 to ammonia (NH3). The NH3 is assimilated as organic nitrogen by the bacteria or by their host plants, which may in turn be consumed by animals. Eventually these organisms excrete the nitrogen or die; the organic nitrogen is eaten by bacteria and mineralized to ammonium (NH4+), which may then be assimilated by other organisms.
Bacteria may also use NH4+ as a source of energy by oxidizing it to nitrite (NO2-) and on to nitrate (NO3-). This process is called nitrification and requires the presence of oxygen (aerobic conditions). Nitrate is highly mobile in soil and is readily assimilated by plant and bacteria, providing another route for formation of organic nitrogen. Under conditions when O2 is depleted in water or soil (anaerobic conditions), bacteria may use NO3- as an alternate oxidant to convert organic carbon to CO2. This process, called denitrification, converts NO3- to N2 and thus returns nitrogen from the biosphere to the atmosphere.
An additional pathway for fixing atmospheric N2 is by high-temperature oxidation of N2 to NO in the atmosphere during combustion or lightning, followed by atmospheric oxidation of NO to HNO3 which is water-soluble and scavenged by rain. In industrial regions of the world, the fixation of N2 in combustion engines provides a source of nitrogen to the biosphere that is much larger than natural N2 fixation, resulting in an unintentional fertilization effect.
Transfer of nitrogen to the lithosphere takes place by burial of dead organisms (including their nitrogen) in the bottom of the ocean. These dead organisms are then incorporated into sedimentary rock. Eventually the sedimentary rock is brought up to the surface of the continents and eroded, liberating the nitrogen and allowing its return to the biosphere. This process closes the nitrogen cycle in the surface reservoirs.
A box model of the nitrogen cycle is presented in Figure 6-3 . This model gives estimates of total nitrogen inventories in the major surface reservoirs and of the flows of nitrogen between different reservoirs. The most accurate by far of all these numbers is the mass mN2 of nitrogen in air, which follows from the uniform mixing ratio CN2 = 0.78 mol/mol:
where 1 teragram (Tg) = 1x1012 g. Other inventories and flows are far more uncertain. They are generally derived by global extrapolation of measurements made in a limited number of environments. The flows given in Figure 6-3 are such that all reservoirs are at steady state, although the steady-state assumption is not necessarily valid for the longer-lived reservoirs.
An important observation from Figure 6-3 is that human activity has greatly increased the rate of transfer of N2 to the biosphere (industrial manufacture of fertilizer, fossil fuel combustion, nitrogen-fixing crops), resulting possibly in a global fertilization of the biosphere. This issue is explored in See Global fertilization of the biosphere .
We focus here on the factors controlling the abundance of atmospheric N2. A first point to be made from Figure 6-3 is that the atmosphere contains most of the nitrogen present in the surface reservoirs of the Earth. This preferential partitioning of nitrogen in the atmosphere is unique among elements except for the noble gases, and reflects the stability of the N2 molecule. Based on the inventories in Figure 6-3 , the N2 content of the atmosphere could not possibly be more than 50% higher than today even if all the nitrogen in the lithosphere were transferred to the atmosphere. It appears then that the N2 concentration in the atmosphere is constrained to a large degree by the total amount of nitrogen present in the surface reservoirs of the Earth, that is, by the amount of nitrogen outgassed from the Earth's interior over 4 billion years ago.
What about the possibility of depleting atmospheric N2 by transferring the nitrogen to the other surface reservoirs? Imagine a scenario where denitrification were to shut off while N2 fixation still operated. Under this scenario, atmospheric N2 would be eventually depleted by uptake of nitrogen by the biosphere, transfer of this nitrogen to soils, and eventual runoff to the ocean where the nitrogen could accumulate as NO3-. The time scale over which such a depletion could conceivably take place is defined by the lifetime of N2 against fixation. With the numbers in Figure 6-3 , we find a lifetime of 3.9x109/(80+160+30+20) = 13 million years. In view of this long lifetime, we can safely conclude that human activity will never affect atmospheric N2 levels significantly. On a geological time scale, however, we see that denitrification is critical for maintaining atmospheric N2 even though it might seem a costly loss of an essential nutrient by the biosphere. If N2 were depleted from the atmosphere, thus shutting off nitrogen fixation, life on Earth outside of the oceans would be greatly restricted.
Answer. The probability that a nitrogen atom in the ocean will be transferred to the lithosphere (burial; 10 Tg N yr-1) rather than to the atmosphere (denitrification in the oceans; 100 Tg N yr-1) is 10/(10+100) = 0.09. A nitrogen atom cycles on average 10 times between atmospheric N2 and the oceans before it is transferred to the lithosphere.
where (CH2O)n is a stoichiometric approximation for the composition of biomass material. This source of O2 is balanced by loss of O2 from the oxidation of biomass, including respiration by living biomass and microbial decay of dead biomass:
To assess the potential for this cycle to regulate atmospheric O2 levels, we need to determine inventories for the different reservoirs of Figure 6-4 : O2, CO2, and orgC. Applying equation (6.1) to O2 and CO2 with CO2 = 0.21 v/v and CCO2 = 365x10-6 v/v ( See Mixing ratios of gases in dry air ), we obtain mO2 = 1.2x106 Pg O and mCO2 = 2000 Pg O = 790 Pg C (1 petagram (Pg) = 1x1015 g). The total amount of organic carbon in the biosphere/soil/ocean system is estimated to be about 4000 Pg C (700 Pg C in the terrestrial biosphere, 2000 Pg C in soil, and 1000 Pg C in the oceans).
Simple comparison of these inventories tells us that cycling with the biosphere cannot control the abundance of O2 in the atmosphere, because the inventory of O2 is considerably larger than that of either CO2 or organic carbon. If photosynthesis were for some reason to stop, oxidation of the entire organic carbon reservoir would consume less than 1% of O2 presently in the atmosphere and there would be no further O2 loss (since there would be no organic carbon left to be oxidized). Conversely, if respiration and decay were to stop, conversion of all atmospheric CO2 to O2 by photosynthesis would increase O2 levels by only 0.2%.
What, then, controls atmospheric oxygen? The next place to look is the lithosphere. Rock material brought to the surface in a reduced state is weathered (oxidized) by atmospheric O2. Of most importance is sedimentary organic carbon, which gets oxidized to CO2, and FeS2 (pyrite), which gets oxidized to Fe2O3 and H2SO4. The total amounts of organic carbon and pyrite in sedimentary rocks are estimated to be 1.2x107 Pg C and 5x106 Pg S, respectively. These amounts are sufficiently large that weathering of rocks would eventually deplete atmospheric O2 if not compensated by an oxygen source. The turnover time of sedimentary rock, that is the time required for sedimentary rock formed at the bottom of the ocean to be brought up to the surface, is of the order of 100 million years. The corresponding weathering rates are 0.12 Pg C yr-1 for rock organic carbon and 0.05 Pg S yr-1 for pyrite. Each atom of carbon consumes one O2 molecule, while each atom of sulfur as FeS2 consumes 19/8 O2 molecules. The resulting loss of O2 is 0.4 Pg O yr-1, which yields a lifetime for O2 of 3 million years. On a time scale of several million years, changes in the rate of sediment uplift could conceivably alter the levels of O2 in the atmosphere.
This cycling of O2 with the lithosphere is illustrated in Figure 6-5 . Atmospheric O2 is produced during the formation of reduced sedimentary material, and the consumption of O2 by weathering when this sediment is eventually brought up to the surface and oxidized balances the O2 source from sediment formation. Fossil records show that atmospheric O2 levels have in fact not changed significantly over the past 400 million years, a time scale much longer than the lifetime of O2 against loss by weathering. The constancy of atmospheric O2 suggests that there must be stabilizing factors in the O2-lithosphere cycle but these are still poorly understood. One stabilizing factor is the relative rate of oxidation vs. burial of organic carbon in the ocean. If sediment weathering were to increase for some reason, drawing down atmospheric O2, then more of the marine organic carbon would be buried (because of slower oxidation), which would increase the source of O2 and act as a negative feedback.
Figure 6-5 Cycling of O2 with the lithosphere. The microbial reduction of Fe2O3 and H2SO4 to FeS2 on the ocean floor is discussed in See Short questions on the oxygen cycle .
Answer. One molecule of O2 is produced for each organic C atom incorporated in the sediments. Therefore, formation of the present-day sediments was associated with the production of (32/12)x1.2x107= 3.2x107 Pg O2. That is 30 times more than the 1.2x106 Pg O2 presently in the atmosphere! Where did the rest of the oxygen go? Examination of Figure 6-5 indicates as possible reservoirs SO42- and Fe2O3. Indeed, global inventories show that these reservoirs can account for the missing oxygen.
Ice core measurements show that atmospheric concentrations of CO2 have increased from 280 ppmv in pre-industrial times to 365 ppmv today. Continuous atmospheric measurements made since 1958 at Mauna Loa Observatory in Hawaii demonstrate the secular increase of CO2 ( Figure 6-6 ). Superimposed on the secular trend is a s easonal oscillation (winter maximum, summer minimum) that reflects the uptake of CO2 by vegetation during the growing season, balanced by the net release of CO2 from the biosphere in fall due to microbial decay.
The current global rate of increase of atmospheric CO2 is 1.8 ppmv yr-1, corresponding to 4.0 Pg C yr-1. This increase is due mostly to fossil fuel combustion. When fuel is burned, almost all of the carbon in the fuel is oxidized to CO2 and emitted to the atmosphere. We can use worldwide fuel use statistics to estimate the corresponding CO2 emission, presently 6.0 0.5 Pg C yr-1. Another significant source of CO2 is deforestation in the tropics; based on rates of agricultural encroachment documented by satellite observations, it is estimated that this source amounts to 1.6 1.0 Pg C yr-1.
we find S sinks = 6.0 + 1.6 - 4.0 = 3.6 Pg C yr-1. Only half of the CO2 emitted by fossil fuel combustion and deforestation actually accumulates in the atmosphere. The other half is transferred to other geochemical reservoirs (oceans, biosphere, and soils). We need to understand the factors controlling these sinks in order to predict future trends in atmospheric CO2 and assess their implications for climate change. A sink to the biosphere would mean that fossil fuel CO2 has a fertilizing effect, with possibly important ecological consequences.
Carbon dioxide dissolves in the ocean to form CO2H2O (carbonic acid), a weak diacid which dissociates to HCO3- (bicarbonate) and CO32- (carbonate). This process is described by the chemical equilibria:
with equilibrium constants KH = [CO2H2O]/PCO2 = 3x10-2 M atm-1, K1 = [HCO3-][H+]/[CO2H2O] = 9x10-7 M (pK1 = 6.1), and K2 = [CO32-][H+]/[HCO3-] = 7x10-10 M (pK2 = 9.2). Here KH is the Henry's law constant describing the equilibrium of CO2 between the gas phase and water; K1 and K2 are the first and second acid dissociation constants of CO2H2O. The values given here for the constants are typical of seawater and take into account ionic strength corrections, complex formation, and the effects of temperature and pressure.
The average pH of the ocean is 8.2 ( See Ocean pH ). The alkalinity of the ocean is mainained by weathering of basic rocks (Al2O3, SiO2, CaCO3) at the surface of the continents, followed by river runoff of the dissolved ions to the ocean. Since pK1 < pH < pK2, most of the CO2 dissolved in the ocean is in the form of HCO3- ( Figure 6-7 ).
and NCO2(g) is related to PCO2 at sea level by (1.11) (Dalton's law):
where P = 1 atm is the atmospheric pressure at sea level and Na = 1.8x1020 moles is the total number of moles of air. Assuming the whole ocean to be in equilibrium with the atmosphere, we relate NCO2(aq) to [CO2(aq)] in (6.5) by the total volume Voc = 1.4x1018 m3 of the ocean:
Substituting into equation (6.3) , we obtain for F:
For an ocean pH of 8.2 and other numerical values given above we calculate F = 0.03. At equilibrium, almost all of CO2 is dissolved in the ocean; only 3% is in the atmosphere. The value of F is extremely sensitive to pH, as illustrated by Figure 6-8 . In the absence of oceanic alkalinity, most of the CO2 would partition into the atmosphere.
Figure 6-8 pH dependence of the atmospheric fraction F of CO2 at equilibrium in the atmosphere-ocean system (equation (6.8) ).
One might infer from the above calculation that CO2 injected to the atmosphere will eventually be incorporated almost entirely into the oceans, with only 3% remaining in the atmosphere. However, such an inference is flawed because it ignores the acidification of the ocean resulting from added CO2. As atmospheric CO2 increases, [H+] in (6.8) increases and hence F increases; this effect is a positive feedback to increases in atmospheric CO2.
The acidification of the ocean due to added CO2 is buffered by the HCO3-/CO32- equilibrium; H+ released to the ocean when CO2(g) dissolves and dissociates to HCO3- (equilibrium (R4) ) is consumed by conversion of CO32- to HCO3- (equilibrium (R5) ). This buffer effect is represented by the overall equilibrium
which is obtained by combining equilibria (R3) - (R5) with (R5) taken in the reverse direction. The equilibrium constant for (R6) is K' = [HCO3-]2/PCO2[CO32-] = KHK1/K2. Uptake of CO2(g) is thus facilitated by the available pool of CO32- ions. Note that the buffer effect does not mean that the pH of the ocean remains constant when CO2 increases; it means only that changes in pH are dampened.
To better pose our problem, let us consider a situation where we add dN moles of CO2 to the atmosphere. We wish to know, when equilibrium is finally reached with the ocean, what fraction f of the added CO2 remains in the atmosphere:
We relate dNCO2(g) to the corresponding dPCO2 using equation (6.6) :
Replacing into (6.9) yields:
We now need a relation for d[CO32-]/dPCO2. Again we use equilibrium (R6) :
Replacing d[HCO3-] ª -2d[CO32-] and K' = KHK1/K2 into (6.15) we obtain
and finally replace into (6.13) :
Substituting numerical values into (6.18) , including a pH of 8.2, we obtain f = 0.28. At equilibrium, 28% of CO2 emitted in the atmosphere remains in the atmosphere, and the rest is incorporated into the ocean. The large difference from the 3% value derived previously reflects the large positive feedback from acidification of the ocean by added CO2.
The above calculation still exaggerates the uptake of CO2 by the ocean because it assumes the whole ocean to be in equilibrium with the atmosphere. This equilibrium is in fact not achieved because of the slow mixing of the ocean. Figure 6-9 shows a simple box model for the oceanic circulation.
Figure 6-9 Box model for the circulation of water in the ocean. Inventories are in 1015 m3 and flows are in 1015 m3 yr-1. Adapted from McElroy, M.B., The Atmosphere: an Essential Component of the Global Life Support System, Princeton University Press (in press).
As in the atmosphere, vertical mixing in the ocean is driven by buoyancy. The two factors determining buoyancy in the ocean are temperature and salinity. Sinking of water from the surface to the deep ocean ( deep water formation) takes place in polar regions where the surface water is cold and salty, and hence heavy. In other regions the surface ocean is warmer than the water underneath, so that vertical mixing is suppressed. Some vertical mixing still takes place near the surface due to wind stress, resulting in an oceanic mixed layer extending to ~100 m depth and exchanging slowly with the deeper ocean. Residence times of water in the individual reservoirs of Figure 6-9 are 18 years for the oceanic mixed layer, 40 years for the intermediate ocean, and 120 years for the deep ocean. Equilibration of the whole ocean in response to a change in atmospheric CO2 therefore takes place on a time scale of the order of 200 years.
This relatively long time scale for oceanic mixing implies that CO2 released in the atmosphere by fossil fuel combustion over the past century has not had time to equilibrate with the whole ocean. Considering only uptake by the oceanic mixed layer (V = 3.6x1016 m3) which is in rapid equilibrium with the atmosphere, we find f = 0.94 from (6.18) ; the oceanic mixed layer can take up only 6% of fossil fuel CO2 injected into the atmosphere. Since the residence time of water in the oceanic mixed layer is only 18 years, the actual uptake of fossil fuel CO2 over the past century has been more efficient but is strongly determined by the rate of deep water formation.
An additional pathway for CO2 uptake involves photosynthesis by phytoplankton. The organic carbon produced by phytoplankton moves up the food chain and about 90% is converted eventually to CO2(aq) by respiration and decay within the oceanic mixed layer. The 10% fraction that precipitates (fecal pellets, dead organisms) represents a biological pump transferring carbon to the deep ocean. The biological productivity of the surface ocean is limited in part by upwelling of nutrients such as nitrogen from the deep ( Figure 6-3 ), so that the efficiency of the biological pump is again highly dependent on the vertical circulation of the ocean water. It is estimated that the biological pump transfers 7 Pg C yr-1 to the deep ocean, as compared to 40 Pg C yr-1 for CO2(aq) transported by deep water formation.
Taking into account all of the above processes, the best current estimate from oceanic transport and chemistry models is that 30% of fossil fuel CO2 emitted to the atmosphere is incorporated in the oceans. We saw in section 6.5.1 that only 50% of the emitted CO2 actually remains in the atmosphere. That leaves a 20% missing sink. We now turn to uptake by the biosphere as a possible explanation for this missing sink.
Cycling of atmospheric CO2 with the biosphere involves processes of photosynthesis, respiration, and microbial decay, as discussed in section 6.4 and illustrated in Figure 6-4 . It is difficult to distinguish experimentally between photosynthesis and respiration by plants, nor is this distinction very useful for our purpose. Ecologists define the net primary productivity (NPP) as the yearly average rate of photosynthesis minus the rate of respiration by all plants in an ecosystem. The NPP can be determined experimentally either by long-term measurement of the CO2 flux to the ecosystem from a tower ( section 4.4.2 ) or more crudely by monitoring the growth of vegetation in a selected plot. From these data, quantitative models can be developed that express the dependence of the NPP on environmental variables including ecosystem type, solar radiation, temperature, and water availability. Using such models one estimates a global terrestrial NPP of about 60 Pg C yr-1.
which implies that atmospheric CO2 responds quickly, on a time scale of a decade, to changes in NPP or in decay rates. It is now thought that increased NPP at middle and high latitudes of the northern hemisphere over the past century may be responsible for the 20% missing sink of CO2 emitted by fossil fuel combustion ( See Sinks of atmospheric CO2 deduced from changes in atmospheric O2 ). Part of this increase in NPP could be due to conversion of agricultural land to forest at northern midlatitudes, and part could be due to greater photosynthetic activity of boreal forests as a result of climate warming. The organic carbon added to the biosphere by the increased NPP would then accumulate in the soil. An unresolved issue is the degree to which fossil fuel CO2 fertilizes the biosphere. Experiments done in chambers and outdoors under controlled conditions show that increasing CO2 does stimulate plant growth. There are however other factors limiting NPP, including solar radiation and the supply of water and nutrients, which prevent a first-order dependence of NPP on CO2.
A summary of the processes described above is presented in Figure 6-10 in the form of a box model of the carbon cycle at equilibrium for preindustrial conditions. Uptake by the biosphere and by the oceans represent sinks of comparable magnitude for atmospheric CO2, resulting in an atmospheric lifetime for CO2 of 5 years. On the basis of this short lifetime and the large sizes of the oceanic and terrestrial reservoirs, one might think that CO2 added to the atmosphere by fossil fuel combustion would be rapidly and efficiently incorporated in these reservoirs. However, as we have seen, uptake of added CO2 by the biosphere is subject to other factors limiting plant growth, and uptake of added CO2 by the ocean is limited by acidification and slow mixing of the ocean.
Even after full mixing of the ocean, which requires a few hundred years, a fraction f = 28% of the added CO2 still remains in the atmosphere according to our simple calculation in section 6.5.3 (current research models give values in the range 20-30% for f). On time scales of several thousand years, slow dissolution of CaCO3 from the ocean floor provides an additional source of alkalinity to the ocean, reducing f to about 7% ( problem 6. 9 ). Ultimate removal of the CO2 added to the atmosphere by fossil fuel combustion requires transfer of oceanic carbon to the lithosphere by formation of sediments, thereby closing the carbon cycle. The time scale for this transfer is defined by the lifetime of carbon in the ensemble of atmospheric, biospheric, soil, and oceanic reservoirs; from Figure 6-10 we obtain a value of 170,000 years. We conclude that fossil fuel combustion induces a very long-term perturbation to the global carbon cycle.