27.05.03 · earth-science / oceanography

Ocean chemistry: salinity, carbonate system, ocean acidification

stub3 tiersLean: nonepending prereqs

Anchor (Master): Sarmiento, J. L. & Gruber, N. — Ocean Biogeochemical Dynamics (2006)

Intuition Beginner

Seawater is about 3.5% dissolved salts by weight. Most of that is sodium chloride — ordinary table salt — but seawater also contains magnesium, calcium, and potassium salts. These dissolved ions make ocean water conductive and give it a slightly bitter taste.

Salinity varies across the ocean. The Atlantic is saltier than the Pacific because it loses more water to evaporation relative to its size. Subtropical regions, where evaporation is high and rainfall is low, are saltier than rainy equatorial zones where freshwater dilutes the salt.

The ocean absorbs roughly a quarter of all CO2 humans release into the atmosphere. When CO2 dissolves in seawater it forms carbonic acid, which releases hydrogen ions and lowers pH — a process called ocean acidification. Since preindustrial times, average surface ocean pH has fallen from about 8.2 to 8.1, corresponding to a 26% increase in hydrogen ion concentration.

That shift threatens marine organisms that build shells and skeletons from calcium carbonate, including corals, clams, oysters, and some plankton. Lower pH reduces the availability of carbonate ions, making it harder for these organisms to grow and maintain their structures.

Visual Beginner

Property Typical value Notes
Salinity (average) 35 PSU (~3.5% by weight) Ranges from ~33 (Baltic) to ~40 (Red Sea)
Dominant ions Na+, Cl- Together ~85% of dissolved salts by weight
Other major ions Mg2+, Ca2+, K+, SO4 2- Conservative: ratios nearly constant
Surface pH (preindustrial) ~8.2 Slightly basic
Surface pH (present) ~8.1 26% increase in H+ concentration
CO2 absorption ~25% of anthropogenic emissions Drives acidification
Threatened organisms Corals, shellfish, pteropods, coccolithophores Rely on calcium carbonate

Worked example Beginner

Coral reefs are built by tiny animals called coral polyps that secrete calcium carbonate skeletons. A reef that took thousands of years to build can be weakened in decades by ocean acidification.

Consider the chemistry. Coral polyps combine calcium ions and carbonate ions to form calcium carbonate: Ca2+ plus CO3 2- yields CaCO3. This reaction requires a sufficient concentration of carbonate ions in the surrounding water.

When excess CO2 from fossil fuel burning dissolves in seawater, it reacts with carbonate ions: CO2 plus H2O plus CO3 2- produces 2 HCO3-. Each molecule of added CO2 consumes one carbonate ion, converting it to bicarbonate. Fewer carbonate ions mean the water is less hospitable to calcification.

Field studies on the Great Barrier Reef show that calcification rates have declined by roughly 15% since 1990, consistent with the drop in carbonate saturation over the same period. Laboratory experiments confirm the mechanism: corals grown in water with lower pH and lower carbonate concentration build thinner, more fragile skeletons.

Check your understanding Beginner

Formal definition Intermediate+

Salinity is the total mass of dissolved inorganic solids per unit mass of seawater, conventionally expressed in practical salinity units (PSU) on the Practical Salinity Scale of 1978 (PSS-78). Since 1978, salinity has been defined operationally through the electrical conductivity of seawater relative to a standard potassium chloride solution, rather than by direct chemical analysis. The major ions in seawater — Na+, Cl-, SO4 2-, Mg2+, Ca2+, K+, and HCO3- — occur in nearly constant ratios (Marcet's principle) and are termed conservative ions because their concentrations change almost exclusively through mixing and freshwater fluxes, not through biological or chemical reactions.

Non-conservative constituents include dissolved gases (O2, CO2, N2), nutrients (nitrate, phosphate, silicate), and trace elements whose concentrations are strongly modified by biological uptake, redox reactions, or scavenging. Dissolved oxygen varies from near zero in oxygen-minimum zones to supersaturation in productive surface waters.

The carbonate system describes the suite of chemical equilibria governing the speciation of dissolved inorganic carbon in seawater. Atmospheric CO2 dissolves in seawater and hydrates to form carbonic acid, which dissociates in two steps:

The first dissociation constant governs the equilibrium between CO2(aq) and HCO3-, and the second dissociation constant governs the equilibrium between HCO3- and CO3 2-. Both constants depend on temperature, salinity, and pressure.

Dissolved inorganic carbon (DIC) is the total concentration of all dissolved inorganic carbon species:

In modern surface seawater at typical pH ~8.1, bicarbonate (HCO3-) constitutes roughly 90% of DIC, carbonate (CO3 2-) about 9%, and aqueous CO2 plus carbonic acid about 1%.

Total alkalinity (TA) is a measure of the ocean's capacity to neutralize acid, defined as the excess of proton acceptors over proton donors relative to a reference pH. Approximately:

where the borate term reflects the contribution of boric acid dissociation. TA is conservative with respect to changes in temperature and pressure (it changes only through addition or removal of constituents), making it a useful tracer alongside DIC.

Calcite and aragonite saturation states

The saturation state of a calcium carbonate mineral is defined as:

where is the solubility product of the mineral. For , the water is supersaturated and precipitation (shell formation) is thermodynamically favored. For , the water is undersaturated and the mineral tends to dissolve. Calcite and aragonite are two crystal polymorphs of CaCO3 with different solubilities; aragonite is more soluble (lower at a given carbonate concentration) and is therefore more vulnerable to acidification. Modern surface waters typically have , but values below 1 occur at depth and are shoaling toward the surface as acidification progresses.

The biological pump and the solubility pump

The biological pump is the suite of processes by which organic carbon fixed by photosynthesis in the surface ocean is transported to depth through sinking particles (detritus, fecal pellets, marine snow), active transport by vertically migrating zooplankton, and dissolution of calcium carbonate shells. The biological pump maintains a vertical gradient in DIC, with deep waters enriched by roughly 10% relative to surface waters.

The solubility pump is the physical mechanism by which CO2 dissolves more readily in cold, dense water at high latitudes, is carried to depth by thermohaline circulation, and is sequestered in the deep ocean. Together, the biological and solubility pumps maintain the ocean's capacity to absorb atmospheric CO2.

Nutrient cycling and the Redfield ratio

Marine phytoplankton require dissolved inorganic nutrients — primarily nitrate (NO3-), phosphate (PO4 3-), and silicate (SiO4 4-) — in addition to carbon and light. The stoichiometry of planktonic organic matter was characterized by Alfred Redfield, who observed that the atomic ratio C

in marine plankton and deep ocean nutrients is approximately 106:16:1, now called the Redfield ratio. Deviations from this ratio indicate nutrient limitation, preferential remineralization, or ecological shifts.

Anoxic and euxinic conditions

Where oxygen is depleted by respiration faster than it is resupplied by circulation, waters become hypoxic (O2 < ~2 mg/L) or anoxic (O2 = 0). In anoxic basins where sulfate-reducing bacteria are active, hydrogen sulfide (H2S) accumulates, producing euxinic conditions. The Black Sea is the largest modern euxinic basin, with oxygenated surface water overlying sulfidic deep water separated by a sharp chemocline at roughly 100–150 m depth.

Key result: the Revelle factor and ocean buffer capacity Intermediate+

The ocean does not absorb CO2 in simple proportion to atmospheric concentration. The relationship is modulated by the ocean's carbonate chemistry through a quantity called the Revelle factor (or buffer factor), defined as:

The Revelle factor measures the fractional change in the partial pressure of CO2 at the ocean surface relative to the fractional change in dissolved inorganic carbon. A high Revelle factor means that a large increase in atmospheric pCO2 produces only a small increase in oceanic DIC — the ocean is a poor buffer. A low Revelle factor means the ocean absorbs CO2 more readily.

For modern surface seawater, the Revelle factor is approximately 10. This means that a 10% increase in atmospheric pCO2 produces only about a 1% increase in DIC. The relatively high value arises because added CO2 is converted primarily to bicarbonate (the dominant species), consuming carbonate ions and limiting the capacity for further CO2 uptake. This buffering is a double-edged sword: it moderates atmospheric CO2 growth but means the ocean cannot absorb all anthropogenic emissions, and the CO2 that is absorbed drives acidification.

The Revelle factor varies geographically. Cold, high-latitude waters have lower Revelle factors (better CO2 absorption) because CO2 is more soluble at low temperatures, while warm subtropical waters have higher Revelle factors. This pattern means the high-latitude oceans — particularly the North Atlantic and the Southern Ocean — are the primary regions of oceanic CO2 uptake.

As acidification progresses and carbonate ion concentrations decline, the Revelle factor increases, reducing the ocean's capacity to absorb additional CO2. This positive feedback between acidification and buffer capacity is a significant concern for climate projections: the ocean will become less effective at mitigating atmospheric CO2 growth even as emissions continue.

Exercises Intermediate+

Advanced results Master

Carbonate system equilibrium constants

The thermodynamic equilibrium constants and for the carbonate system in seawater are functions of temperature (), salinity (), and pressure (). The most widely used formulations are those of Mehrbach et al. (1973), refit by Dickson and Millero (1987), and the alternative formulation of Roy et al. (1993). The pressure dependence is significant below the thermocline: increasing pressure stabilizes the dissociated species, effectively increasing and at depth and shifting the carbonate equilibrium toward bicarbonate and carbonate at the expense of aqueous CO2.

The borate dissociation constant governs the equilibrium:

Borate contributes significantly to alkalinity (roughly 10% of total alkalinity in typical seawater) and must be included in any precise carbonate system calculation. The total boron concentration is proportional to salinity, with mol/kg.

Given any two of the four carbonate system parameters (DIC, TA, pCO2, pH), the other two can be calculated by solving the coupled equilibrium equations. This is the foundation of ocean CO2 system software such as CO2SYS (Lewis and Wallace 1998, with subsequent updates by van Heuven et al. 2011 and Orr et al. 2015).

The Revelle factor in detail

The Revelle factor can be expressed analytically in terms of the equilibrium constants and the carbonate system parameters. Following Sarmiento and Gruber (2006), the Revelle factor for a simple carbonate system (ignoring borate and other minor contributions) is:

This expression shows that depends on the ratio DIC/TA and on the speciation of the carbonate system. The global mean surface Revelle factor is approximately 10, but it ranges from about 8 in cold subpolar waters to about 15 in warm tropical waters. As DIC increases relative to TA (i.e., as the ocean absorbs more CO2), the Revelle factor rises, reducing the ocean's buffer capacity in a positive feedback.

Isotopic tracers in ocean chemistry

Stable and radiogenic isotopes provide powerful tools for tracing ocean chemistry and circulation. Carbon-13 (C) in dissolved inorganic carbon reflects the balance between biological production (which preferentially removes 12C) and remineralization, as well as the Suess effect — the progressive depletion of C in surface waters from fossil fuel CO2 (fossil fuels are depleted in 13C). Radiocarbon (C) measures the ventilation age of water masses: deep water that has been isolated from the atmosphere for centuries has lower C due to radioactive decay (half-life 5,730 years).

Oxygen isotopes (O) in the calcium carbonate shells of foraminifera record the temperature and isotopic composition of the water in which the shells formed. This is the foundation of paleoceanographic temperature reconstruction extending back millions of years.

Paleo-pH reconstruction from boron isotopes

The boron isotope proxy is the primary method for reconstructing past ocean pH. Boron in seawater exists as B(OH)3 and B(OH)4-, with the relative abundance controlled by pH. The two species have different B signatures because of an equilibrium isotope fractionation of about 27 per mil. As pH increases, the fraction of B(OH)4- increases, and its B approaches the seawater value. Boron incorporated into marine carbonates (primarily as B(OH)4- substituted for CO3 2-) preserves a record of the B of B(OH)4- at the time of formation, allowing reconstruction of paleo-pH with uncertainties of roughly 0.02–0.05 pH units. Application to foraminifera from deep-sea sediment cores has revealed pH changes of 0.3–0.4 units across glacial-interglacial cycles and larger excursions during mass extinction events.

IPCC projections of ocean acidification

Under the IPCC SSP5-8.5 scenario (high emissions), surface ocean pH is projected to decline from the present ~8.1 to ~7.75 by 2100, with aragonite saturation states dropping below 1 in large portions of the Southern Ocean, subarctic Pacific, and portions of the North Atlantic. Under SSP1-2.6 (strong mitigation), pH stabilizes near 8.05 by 2100. The rate of acidification under high-emission scenarios is estimated at 0.02–0.03 pH units per decade — an order of magnitude faster than any change recorded in the geological record over the past 300 million years, with the possible exception of the Chicxulub impact.

Impacts on calcification and marine ecosystems

Laboratory and field studies demonstrate that reduced pH and carbonate saturation states impair calcification across a wide range of organisms. Coral calcification rates decline roughly 15–30% for a doubling of pCO2. Pteropods (marine snails) show visible shell dissolution within days when exposed to water undersaturated with respect to aragonite. Coccolithophores, which produce calcium carbonate plates (liths), show malformed and reduced lith production under elevated CO2.

Coral bleaching — the loss of symbiotic dinoflagellates under thermal stress — is compounded by acidification because bleached corals have reduced energy for calcification and repair. The combination of warming and acidification is projected to reduce coral reef accretion to below erosion rates by mid-century under high-emission scenarios, causing net reef loss.

Biogeochemical models: CEMENT, OCMIP, and coupled climate-carbon models

The Carbon Cycle Model Intercomparison Project (OCMIP) established standardized protocols for comparing ocean carbon cycle models, specifying boundary conditions, circulation fields, and diagnostic outputs. Key outputs include air-sea CO2 flux, DIC and alkalinity distributions, and biological pump efficiency. The follow-on OCMIP-2 protocol included intercomparison of naturally occurring and anthropogenic CO2 uptake.

CEMENT (Carbon Ecosystem Model for ENvironmental Tracing) and similar box models represent the ocean carbon cycle using a small number of well-mixed reservoirs (surface, intermediate, and deep ocean) connected by prescribed fluxes. While simplified, these models capture the first-order behavior of the carbonate system and allow rapid exploration of parameter space.

Coupled climate-carbon cycle models (C4MIP) embed the ocean carbon cycle within Earth system models that include atmospheric CO2, terrestrial carbon dynamics, and climate feedbacks. A key result from C4MIP is the positive feedback between climate change and the carbon cycle: warming reduces ocean CO2 solubility, strengthens stratification (reducing the supply of deep carbon to the surface), and alters biological productivity, all of which weaken the ocean carbon sink. This feedback amplifies atmospheric CO2 growth by 20–200 ppm by 2100 across different models, representing a major source of uncertainty in climate projections.

The dolomite problem

Dolomite (CaMg(CO3)2) is abundant in the geological record — particularly in Paleozoic and Precambrian carbonates — yet virtually no dolomite forms in the modern ocean at Earth surface conditions. This "dolomite problem" has persisted for over two centuries. Modern seawater is supersaturated with respect to dolomite, but the kinetic barrier to ordered Ca-Mg substitution prevents direct precipitation. Most modern dolomite forms through slow, diagenetic replacement of calcite or aragonite by Mg-rich fluids, often mediated by microbial sulfate reduction in anoxic sediments. The abundance of dolomite in ancient rocks may reflect different ocean chemistry (higher Mg/Ca ratios, different sulfate concentrations) or longer timescales of diagenesis.

Ocean anoxic events and mass extinctions

The mid-Cretaceous ocean anoxic events (OAEs, ~120–90 Ma) were episodes of widespread deep-ocean anoxia recorded as black shale deposits rich in organic carbon. OAE 1a (Selli event, ~120 Ma) and OAE 2 (Bonarelli event, ~93 Ma) are associated with large carbon cycle perturbations, inferred from sharp negative excursions in C (indicating massive carbon injection) followed by positive excursions (indicating organic carbon burial). These events likely involved submarine volcanism (particularly the Ontong Java Plateau) releasing CO2, warming the climate, intensifying the hydrological cycle, increasing weathering and nutrient fluxes, and driving eutrophication and oxygen depletion in restricted ocean basins.

The end-Permian extinction (~252 Ma) provides the most dramatic example of ocean chemistry crisis as an extinction mechanism. Evidence from C, O, boron isotopes, and biomarker proxies supports a scenario in which Siberian Traps volcanism released enormous volumes of CO2, causing rapid warming, ocean acidification, and expansion of anoxic and euxinic conditions. The resultant collapse of marine ecosystems eliminated approximately 90% of marine species. Ocean acidification is inferred from the selective extinction of heavily calcified organisms and the temporary disappearance of carbonate platforms.

CLIVAR and GO-SHIP repeat hydrography

The CLIVAR (Climate Variability and Predictability) program and its successor GO-SHIP (Global Ocean Ship-based Hydrographic Investigations Program) maintain a network of repeat hydrographic sections spanning the global ocean. These sections provide high-precision measurements of dissolved inorganic carbon, total alkalinity, dissolved oxygen, nutrients, and transient tracers (CFCs, SF6, C) along full-depth transects. The repeat measurements, conducted approximately every 5–10 years, allow direct detection of ocean chemistry changes — including the penetration of anthropogenic CO2 to depth, the shoaling of aragonite saturation horizons, and the expansion of oxygen-minimum zones. GO-SHIP data are the gold standard for validating ocean biogeochemical models and calibrating autonomous sensor measurements from Argo floats.

Connections Master

Connections to ocean circulation (Unit 27.05.02)

Ocean chemistry is inseparable from ocean circulation. The thermohaline circulation transports dissolved gases, nutrients, and carbon between the surface and deep ocean on millennial timescales. The solubility pump depends on cold, dense water absorbing CO2 at high latitudes and carrying it to depth. Upwelling zones bring deep, nutrient-rich (and often CO2-rich) water to the surface. Changes in overturning rate alter the distribution of DIC, alkalinity, and oxygen throughout the water column.

Connections to climate change (Unit 27.07.01)

Ocean acidification is a direct consequence of rising atmospheric CO2 — what Roger Revelle called "the other CO2 problem." The ocean absorbs roughly a quarter of anthropogenic CO2 emissions, moderating atmospheric warming but at the cost of changing ocean chemistry. The feedback between climate change and the ocean carbon sink (weakening absorption under warming) is a positive feedback that amplifies atmospheric CO2 growth and therefore warming. Ocean deoxygenation — the expansion of oxygen-minimum zones driven by warming-induced stratification — compounds acidification stress on marine organisms.

Connections to marine ecosystems (Unit 27.05.01)

Ocean chemistry directly controls the habitability of marine environments. pH and carbonate saturation states determine whether organisms can build and maintain calcium carbonate structures. Nutrient availability (nitrate, phosphate, iron) controls primary production and the biological pump. Dissolved oxygen concentrations define the boundary between aerobic and anaerobic ecosystems. The combined stresses of warming, acidification, and deoxygenation threaten marine biodiversity and fisheries.

Connections to Earth history (Unit 27.08.01)

The chemistry of the ocean has changed dramatically over geological time, and these changes are recorded in marine sediments. Carbon isotope excursions mark episodes of massive carbon injection and burial. Boron isotopes record past pH. Oxygen isotopes record temperature and ice volume. The carbonate compensation depth fluctuates with ocean chemistry. Past ocean chemistry crises — the end-Permian, the Paleocene-Eocene Thermal Maximum, the mid-Cretaceous OAEs — provide analogues for understanding the current anthropogenic perturbation and its potential consequences.

Connections to atmospheric chemistry

The ocean and atmosphere exchange CO2, O2, N2, dimethyl sulfide (DMS), and other gases. The direction and magnitude of air-sea CO2 flux depend on the difference between atmospheric pCO2 and surface ocean pCO2, which is controlled by surface temperature, DIC, and alkalinity. The ocean is currently a net sink for atmospheric CO2, absorbing roughly 2.5 GtC per year. The ocean also releases DMS produced by marine phytoplankton, which oxidizes in the atmosphere to form sulfate aerosols that influence cloud formation and climate.

Connections to the carbon cycle and geochemistry

Ocean chemistry is a major component of the global carbon cycle. The ocean holds roughly 38,000 GtC as dissolved inorganic carbon — about 50 times the atmospheric reservoir. The weathering of silicate rocks on land consumes atmospheric CO2 and delivers bicarbonate to the ocean, where it is eventually precipitated as calcium carbonate and buried in marine sediments. This silicate weathering feedback operates on timescales of hundreds of thousands of years and is Earth's long-term thermostat. On shorter timescales, the ocean's capacity to absorb CO2 is governed by the carbonate chemistry described in this unit.

Historical and philosophical context Master

The Challenger expedition and the birth of chemical oceanography

The HMS Challenger expedition (1872–1876) produced the first systematic measurements of ocean chemistry, including salinity, dissolved gases, and dissolved constituents at stations spanning the global ocean. The chemical analysis of the water samples, led by William Dittmar, established that the relative proportions of the major ions in seawater are nearly constant — the foundation of Marcet's principle (sometimes called the Forchhammer principle or Dittmar's law). This result is the basis for the practical salinity measurement through conductivity that is still used today.

Roger Revelle and the discovery of the ocean carbon sink

Roger Revelle, director of the Scripps Institution of Oceanography, and his colleague Hans Suess demonstrated in a landmark 1957 paper that the ocean does not absorb anthropogenic CO2 as readily as had been assumed. Revelle showed that the carbonate chemistry of seawater imposes a buffer — now quantified by the Revelle factor — that limits the fraction of emitted CO2 that dissolves in the ocean. This result was critical for establishing that a substantial fraction of fossil fuel CO2 would remain in the atmosphere and drive warming. Revelle and Suess's work directly motivated Charles David Keeling's program of continuous atmospheric CO2 measurement beginning in 1958, producing the iconic Keeling Curve.

Keeling, Takahashi, and the measurement revolution

Charles David Keeling's meticulous measurements of atmospheric CO2 at Mauna Loa Observatory beginning in 1958 revealed both the steady increase in CO2 concentration and the seasonal cycle driven by Northern Hemisphere photosynthesis. Taro Takahashi of Lamont-Doherty Earth Observatory extended this approach to the ocean, producing the first global maps of air-sea CO2 flux from surface pCO2 measurements. The Takahashi database, maintained and updated over four decades, remains the benchmark for estimating the ocean's CO2 uptake.

The emergence of ocean acidification as a scientific concern

While the chemistry of CO2 in seawater had been understood since the early 20th century, the ecological implications of ocean acidification were not widely recognized until the early 2000s. Ken Caldeira and Michael Wickett's 2003 paper in Nature, which coined the term "ocean acidification" and projected pH declines under various emission scenarios, brought the issue to wide scientific and public attention. Scott Doney's 2009 Annual Review paper synthesized the growing body of evidence for acidification impacts. The first symposium on "The Ocean in a High-CO2 World" was held in 2004, and the issue has since become a central topic in climate science and policy.

The philosophical dimension: planetary boundaries and irreversible change

Ocean acidification represents a perturbation of the Earth system operating on a timescale — centuries to millennia for recovery — that is geologically rapid but politically glacial. The ocean's carbonate chemistry changes faster than marine organisms can adapt, and faster than geological processes (silicate weathering) can restore the original state. The concept of ocean acidification as a "planetary boundary" — a threshold beyond which the Earth system may shift to a qualitatively different state — raises fundamental questions about intergenerational responsibility and the limits of the precautionary principle. Once a coral reef is dissolved, it cannot be rebuilt on human timescales, regardless of future emissions reductions.

Bibliography Master

  1. Tarbuck, F. K. & Lutgens, E. J. (2018). Earth Science (15th ed.). Pearson. Ch. 10: Ocean chemistry.

  2. Pilson, M. E. Q. (2013). An Introduction to the Chemistry of the Sea (2nd ed.). Cambridge University Press. Ch. 1-5: Seawater composition.

  3. Sarmiento, J. L. & Gruber, N. (2006). Ocean Biogeochemical Dynamics. Princeton University Press. Ch. 1-4: Carbonate system.

  4. Doney, S. C., Fabry, V. J., Feely, R. A. & Kleypas, J. A. (2009). "Ocean acidification: the other CO2 problem." Annual Review of Marine Science, 1, 169-192.

  5. Revelle, R. & Suess, H. (1957). "Carbon dioxide exchange between atmosphere and ocean and the question of an increase of atmospheric CO2 during the past decades." Tellus, 9, 18-27.

  6. Dickson, A. G. & Millero, F. J. (1987). "A comparison of the equilibrium constants for the dissociation of carbonic acid in seawater media." Deep Sea Research, 34, 1733-1743.

  7. Caldeira, K. & Wickett, M. E. (2003). "Oceanography: anthropogenic carbon and ocean pH." Nature, 425, 365.

  8. Mehrbach, C., Culberson, C. H., Hawley, J. E. & Pytkowicz, R. M. (1973). "Measurement of the apparent dissociation constants of carbonic acid in seawater at atmospheric pressure." Limnology and Oceanography, 18, 897-907.

  9. Takahashi, T. et al. (2009). "Climatological mean and decadal change in surface ocean pCO2, and net sea-air CO2 flux over the global oceans." Deep Sea Research Part II, 56, 554-577.

  10. Redfield, A. C. (1958). "The biological control of chemical factors in the environment." American Scientist, 46, 205-221.

  11. Orr, J. C. et al. (2005). "Anthropogenic ocean acidification over the twenty-first century and its impact on calcifying organisms." Nature, 437, 681-686.