SciELO - Scientific Electronic Library Online

vol.70 suppl.1La corriente sub-superficial del Perú-Chile y la variabilidad de la zona de mínimo oxígeno frente a Chile centralDeficiencia de oxígeno en el Oceano Indico norte índice de autoresíndice de materiabúsqueda de artículos
Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados


Gayana (Concepción)

versión impresa ISSN 0717-652Xversión On-line ISSN 0717-6538

Gayana (Concepc.) v.70  supl.1 Concepción oct. 2006 

Suplemento Gayana 70: 46-52, 2006


Global warming projection of the change in dissolved oxygen concentrations in low oxygen regions of the oceans

El calentamiento global y la proyección del cambio en la concentración de oxígeno disuelto en las regiones de bajo oxígeno de los oceanos

Richard Matear

CSIRO Marine and Atmospheric Research, Marine Laboratories, GPO Box 1538
Hobart, Tasmania, Australia 7001,


Global warming projections using a range of climate models included in the IPCC 4th assessment report (AR4) suggest the oceans will warm, the stratification of the upper ocean will increase and the ventilation of the ocean interior will change. These physical changes will impact dissolved oxygen levels in the ocean. Using a global warming projection from the CSIRO (Australian Commonwealth Scientific and Research Administration) climate model linked to a simple ocean biogeochemical model I investigated how dissolved oxygen levels in the ocean interior change under global warming. The climate simulations project the low oxygen regions like the eastern equatorial Pacific will expand. By the end of the century it is projected that the volume of hypoxic water (<10 mmol/kg) in the thermocline of the eastern Equatorial Pacific Ocean will double.

Keywords: Global warming, oxygen changes, anoxia, biogeochemical cycles, climate change.


Las proyecciones sobre el calentamiento global utilizando un rango de modelos climáticos incluidos en el 4° informe de evaluación del IPCC (AR4), sugiere que los océanos se calentarán, incrementará la estratificación del océano superficial y cambiará la ventilación del océano interior. Estos cambios físicos impactarán sobre los niveles de oxígeno disuelto en el océano. Utilizando una proyección del calentamiento global originado del modelo climático del CSIRO (Australian Commonwealth Scientific and Research Administration) acoplado a un modelo biogeoquímico simple investigo cómo cambiarán los niveles de oxígeno disuelto en el océano interior bajo el calentamiento global. Las simulaciones del clima proyectan que se expandirán las regiones con bajo oxígeno como el Pacífico ecuatorial oriental. Hacia el final del siglo se proyecta que se duplicará el volumen de agua hipóxica (<10 mmol/kg) en la termoclina del Océano Pacifico Ecuatorial.

Palabras Claves: Calentamiento global, cambios en el oxígeno disuelto, anoxia, ciclo biogeoquímicos, cambio climático.


The atmospheric concentration of CO2 and other greenhouse gases are increasing and changing the radiative properties of the atmosphere. To study the response of the ocean-climate system to prolonged climate perturbation, investigators have integrated ocean-climate models with an atmospheric CO2 concentration that rises over time to double, triple or quadruple the control CO2 level (Bi et al. 2001, Hirst 1999, Manabe & Stouffer 1993, Goosse & Renssen 2001). The oceanic response in these experiments typically includes widespread surface warming, retreat of sea ice and a general increase in upper ocean stratification, all of which may impact upon the dissolved oxygen concentrations of the ocean.

Several climate change simulations have predicted outgassing of oxygen from the ocean into the atmosphere and large declines in the dissolved oxygen concentrations in the ocean interior by the end of this century (Bopp et al. 2002, Matear et al. 2000, Plattner et al. 2001, Sarmiento et al. 1998). This study uses a biogeochemical model that includes carbon, nutrient and oxygen cycling to examine the effect of anthropogenic global warming on the oceanic oxygen distribution in a climate-change simulation that displays major changes in water mass formation rates and densities. The study investigates whether global warming will expand the present anoxic regions in the ocean.


We utilize the CSIRO climate model (Gordon & O'Farrell 1997, Hirst et al. 2000) together with a simple oceanic biogeochemical model (Matear & Hirst 1999) to investigate the impact of climate change on anoxic ocean regions. The CSIRO climate model is a comprehensive coupled general circulation model with atmospheric, oceanic, sea ice and land surface components. Further details of the ocean component are found in Hirst et al. (2000). The biogeochemical model is described by Matear & Hirst (2003) and is run using archived climate model output following a technique described by Aumont et al. (1998).

The biogeochemical model includes a simple representation of the surface export production of biological matter as a function of the temperature, mixed-layer depth and nutrient concentration (phosphate) in the euphotic zone, with the sinking particulate organic matter (POM) remineralizing according to prescribed functions of depth. The euphotic zone production of organic matter consumes dissolved phosphorous, nitrogen and carbon and releases oxygen according to the Redfield et al. (1963) ratio P:N:C:O2 of 1:16:106:-138, while the subsurface remineralization conversely releases/consumes these elements in like ratio.

We perform two simulations with the CSIRO climate model: 1) a control experiment of duration 850 years with a constant atmospheric level of equivalent CO2 (330 µatm), and 2) a greenhouse gas forcing experiment where the level of equivalent CO2 follows the IS92a radiative forcing scenario (Houghton et al. 1995) from the year 1880 until the year 2083, and thereafter is held constant for another 650 years (Fig. 1a). The elevated stabilized level of equivalent CO2 corresponds to three times 1880 value (990 µatm).

3. Results

The three primary factors affecting oceanic oxygen concentrations are changes in (1) surface temperature, (2) export production, and (3) ventilation. An increase in surface ocean temperature decreases oxygen solubility hence decreases surface oxygen concentration. The signature of the reduced surface concentration may then be advected or mixed into the ocean interior. An increase in export production increases the remineralization of sinking particulate organic matter causing more rapid depletion of subsurface oxygen, hence reduces interior oxygen concentration. Finally, a decrease in the ventilation rate increases the residence time of water in the subsurface ocean allowing for more POM remineralization to occur within a given water mass and so reduces oxygen concentrations in the ocean interior. A fourth affect on the oceanic oxygen concentrations is the increase in the rate of air-sea oxygen exchange caused by the reduction in sea-ice extent with climate change. However, since the subducted Antarctic Bottom Water (AABW), Antarctic Intermediate Water (AAIW) and North Atlantic Deep Water (NADW) are over 95% oxygen saturated when they are subducted in the control experiment, the increase in gas exchange can only increase the oxygen content of these water masses by a maximum of 10 µmol/kg.

3.1 Physical Changes

Here we summarize the physical changes most relevant to future dissolved oxygen content of the ocean. The time series of globally-averaged sea surface temperature (SST) for the two experiments are shown in Fig. 1b. As in other similar climate model experiments (e.g., Sarmiento et al. 2004), the SST shows a rapid increase during the period of increasing equivalent CO2 concentration in the atmosphere, followed by a gradual increase during the subsequent period of stable elevated equivalent CO2 concentration. The change in surface water density follows a similar pattern, with rapid reduction during the former period and a gradual further reduction during the later period. The density reduction in the surface is mainly caused by surface warming in the tropics and mid-latitudes, and by surface freshening in the polar regions. The simulated freshening of the surface polar ocean with global warming is caused by an increase in pole-ward atmospheric moisture transport and high latitude precipitation and, in some regions, by reduced rate of brine rejection associated with reduced sea-ice formation. Sea ice extent declines by 2730 to about 15-20% of that at the outset.

While the upper ocean responds rapidly to changes in buoyancy forcing, the deep ocean density adjusts to the altered surface conditions on much longer, millennial, time scale. The consequence of these differences in time-scale is that upper ocean stratification increases cause a reduction and/or shoaling of the oceanic overturning circulation. Time series of our simulated North Atlantic and Antarctic deep water overturning are shown in Figs. 1c and 1d. The strength of the North Atlantic overturning is reduced by nearly 50% under global warming, and it shoals so that by about the 23rd century, new North Atlantic Deep Water penetrates to only about 1.5 km depth. Antarctic overturning also declines to minimal levels by about the 23rd century. These changes, combined with general reductions in high-latitude oceanic convection and high-latitude outcropping of density surfaces result in markedly reduced ventilation of deep water during the period of elevated stable equivalent CO2.

Figure 1. Time series of model indices for the control and climate change ( transient,) integrations: a) Prescribed atmospheric equivalent CO2 concentration (relative to 1880 level); b) Global average sea surface temperature; e) Strength of overturning of the Antarctic Bottom Water formation circulation which penetrates downward through the 1250-m depth level (106 m3 s-1); d) Strength of overturning in the North Atlantic Deep Water formation circulation, and strength of this overturning which penetrates downward through the 1600-m depth level (106 m3 s-1). Dotted vertical lines indicate the present time (calendar year 2000), and the times of equivalent CO2 doubling (2033) and tripling (2082) from the 1880 level.

3.2 Oxygen Changes

To assess the model simulations of dissolved oxygen we compare them to observations (Levitus et al. 1993). Figure 2a-b shows the observed and simulated zonal averaged dissolved oxygen concentrations in the Pacific Ocean. In the Pacific, the model displays the major features evident in the observations. The model reproduces the penetration of oxygen rich Antarctic Intermediate Water (AAIW) and Antarctic Bottom Water (AABW) into the subtropical gyre of the South Pacific. The tongue of AAIW has similar form and magnitude to the observations. However, the modeled AABW is about 10 µmol/kg too high in dissolved oxygen at 50°S. Generally dissolved oxygen concentrations are 10 µmol/kg too high in the Southern Ocean, below 1000 m. In the North Pacific, the model reproduces the oxygen minimum water at the correct depth but the magnitude is 10 µmol/kg greater than the observations. In the equatorial Pacific, the oxygen minimum water is at the right depth but again the value is 10 µmol/kg greater than the observations.


Figure 2. Zonal mean field of oceanic dissolved oxygen concentration for the Pacific ocean from (a) the observations (Levitus et al. 1993), (b) the climate change integration for the year 1980 in (mmol kg-1).

Figure 3. Dissolved oxygen concentrations (µmol/kg) at 300m from (a) the observations (Levitus et al. 1993) and b) the climate change model for the year 1990.

Excess oxygen simulated by model the in the South Pacific and South Atlantic is attributed to excess ventilation of the water in this region which is supported by simulated natural 14C concentrations that are too young compared to observations. If the high oxygen concentrations were caused by too little export production we would expect to simulate phosphate concentrations in the Southern Ocean that are less than the observed values and this does not occur. The zonal averaged sections show that the deep and intermediate water of the model contain too much dissolved oxygen. Below 1000 m the global averaged dissolved oxygen concentration in the model exceeds the observations by 5 µmol/kg. Hence, the model is starting from a state that is less prone to deep anoxia than the present ocean.

In both the model and the observations the major regions of hypoxia (less 10 mmol kg-1 dissolved oxygen concentrations) are found in the thermocline of the northern Indian and in the eastern equatorial Pacific Oceans. To assess the model's ability to simulate regions of low oxygen concentrations we compare the observed oxygen concentrations at 300 m with the simulated values (Fig. 3). In the tropical Atlantic and Indian Oceans one finds the model has larger regions of hypoxic water than the observations. This disagreement is greatest at 300 m. In the tropical Pacific the hypoxic water is concentrated closer to the coast than in the observations. In all three oceans, the model appears to have excess export production in the equatorial upwelling regions, which tends to concentrates the oxygen depletion closer to the site of upwelling at the expense of the rest of the basin. In the model, the hypoxic regions are larger than the observations (by a factor of four in overall volume), and true anoxic conditions occur in some locations, which are not apparent in the observed climatology (Levitus et al. 1993). Although the model over-estimates the volume of hypoxic water most of the discrepancy occurs in the Indian and Atlantic Oceans, two regions that do not have expanding hypoxic water in our climate change projections. In the equatorial Pacific, the volume of hypoxic water is comparable to the observations.

The extent of the hypoxic regions may change with greenhouse warming. Overall, the oxygen concentration of tropical thermocline waters increases in the climate change integration because the reduction in export production in the tropics more than offsets the decline in concentration resulting from solubility decrease of the warmer surface source waters. However, in regions where export production remains unchanged or increases, such as the North Pacific and eastern Equatorial Pacific Oceans, oxygen levels decline. Figure 4 shows the change in the global volume of hypoxic water in the model during the control and climate change integrations. The total ocean volume hypoxic water increases slightly by the end of the climate change integration. The volume of hypoxic water in the North Indian Ocean declines during the climate change integration, while that in the eastern equatorial Pacific increases and is responsible for the overall growth in the volume seen in Fig. 4.

Figure 4. Time series of the total volume of hypoxic water (O2 concentration is less than 10 mmol kg-) present in the global ocean.

Most of the increase in volume of hypoxic water in the eastern tropical Pacific occurs between 10°N and 10°S, where this volume approximately doubles during the course of the climate change integration. In contrast, oxygen concentrations in the thermocline exhibit only small changes over most of the remaining tropical Pacific during the same period. This difference in dissolved oxygen trends is related to a tendency for export production to become more concentrated near the tropical Pacific upwelling zones. Over the tropical Pacific as a whole, total export production falls during the climate change integration, but over the eastern equatorial Pacific (115°W-78°W, 10°N-10°S), export production increases by 20 percent. The elevated production region is associated with the more rapid utilization of nutrient as a result of warmer surface waters. Thus, over most of the tropical Pacific, the decline in thermocline oxygen concentration resulting from the change in solubility of the source water tends to be offset by declines in the consumption resulting from reduced remineralization of POM, leading to a minor net change in concentration. In the eastern equatorial Pacific, the effect of the changed solubility is augmented by the effect of the increased oxygen consumption by POM remineralization, yielding lower thermocline oxygen concentrations and more extensive hypoxia. The effect of the change in solubility on the volume of hypoxic water can be quantified; the volume of hypoxic water in the eastern equatorial Pacific increases by only 60 per cent in the climate change integration performed with fixed solubility, compared to a doubling of volume in the full integration.


We have performed a suite of integrations with a coupled ocean-atmosphere climate model and an ocean biogeochemical model to examine the long-term impact of global warming on oceanic dissolved oxygen levels. A state of collapsed Antarctic Bottom Water formation and shoaled North Atlantic overturning develops by about year 2150, leaving the ocean below about the 2-km depth unventilated. This situation persists through to the end of the integration. The physical background is therefore one of major disruption to the deep ocean overturning, which is known to be important in the cycling of nutrients and the replenishment of oxygen to the deep ocean (Bopp et al. 2002, Matear 2000).

Overall, oxygen concentrations in the upper 1.5 km of the ocean fall during the first several centuries of the principal climate change integration, but approach equilibration thereafter. The decline in the oxygen concentration in the thermocline waters is mostly caused by the reduction in solubility of the surface source waters due to warming. Oxygen concentrations actually rise in the thermocline across parts of the tropical oceans, as a result of reduced oxygen consumption by remineralization of POM associated with reductions in export production. Export production in the tropics is reduced generally as part of an overall decline associated with the increased stratification and reduced supply of nutrients to the upper ocean. Export production also becomes more concentrated towards regions of nutrient supply (i.e., towards regions of upwelling or up-mixing of nutrient rich water), as the warmer conditions facilitate more rapid utilization of available nutrients. Export production actually increases in the vicinity of the upwelling zones in the eastern equatorial Pacific, and this leads to enhanced subsurface oxygen consumption and to a doubling of the volume of hypoxic thermocline water in that region. The response of export production to climate change determines the response of the low ocean regions to climate change and this response needs to be better understood.


The author would like to acknowledge the financial support he receives from the Australian Greenhouse Office Climate Change Science Program.


Aumont, O., J. C. Orr, D. Jamous, P. Monfray, O. Marti, & G. Madec. 1998. A degradation approach to accelerate simulations to steady state in a 3-D tracer transport model of the global ocean. Climate Dynamics, 14, 101-116.         [ Links ]

Bi, D. H., W. F. Budd, A. C. Hirst & X. R. Wu. 2001. Collapse and reorganisation of the Southern Ocean overturning under global warming in a coupled model. Geophys. Res. Let., 28:3927-3930.         [ Links ]

Bopp, L., C. Le Quere, M. Heimann, A. C. Manning & P. Monfray. 2002. Climate-induced oceanic oxygen fluxes: Implications for the contemporary carbon budget. Global Biogeochem. Cycles, 16, art. no.1022.         [ Links ]

Goosse, H. & H. Renssen. 2001. A two-phase response of the Southern Ocean to an increase in greenhouse gas concentrations. Geophys. Res. Let., 28: 3469-3472.         [ Links ]

Gordon, H.B. & S.P. O'Farrell. 1997. Transient climate change in the CSIRO coupled model with dynamical sea ice., Mon. Wea.Rev. 125:875-907.         [ Links ]

Hirst, A.C. 1999. The Southern Ocean response to global warming in the CSIRO coupled ocean-atmosphere model, Environmental Modeling and Software: Special issue on Modelling Global Climate Change, 14:227-242.         [ Links ]

Hirst, A. C., S. P. O'Farrell & H. B. Gordon. 2000. Comparison of a coupled ocean-atmosphere model with and without oceanic eddy-induced advection. Part I: Ocean spinup and control integrations Journal of Climate, 13:139-163.         [ Links ]

Houghton, J. T., L. G. M. Filho, J. Bruce, H. Lee, B. A. Callander, E. Haites, N. Harris & K. Maskell (Eds.). 1995. Climate Change 1994, 339 pp., Intergovernmental Panel on Climate Change, Cambridge University Press.         [ Links ]

Levitus, S., M. E. Conkright, J. L. Reid & R. G. Najjar. 1993. Distribution of nitrate, phosphate and silicate in the world oceans, Prog. Oceanogr., 31:245-273.         [ Links ]

Manabe, S. & R. J. Stouffer. 1993. Century-scale effects of increased atmospheric CO2 on the oceanic-atmosphere system. Nature, 364:215-218.         [ Links ]

Matear, R.J. 2000. Climate change impacts on marine systems, Australia Microbiology, 21:17-20.         [ Links ]

Matear, R.J. & A.C. Hirst. 1999. Climate Change Feedback on the Future Oceanic CO2 uptake. Tellus, 51B:722-733.         [ Links ]

Matear, R. J. & A.C. Hirst. 2003. Long term changes in dissolved oxygen concentrations in the ocean caused by protracted global warming, Global Biogeochem. Cycles, 17:1125.         [ Links ]

Matear, R. J., A. C. Hirst & B. I. McNeil. 2000. Changes in dissolved oxygen in the Southern Ocean with climate change, Geochemistry Geophysics Geosystems. (         [ Links ]

Plattner, G. K., F. Joos, T. F. Stocker & O. Marchal. 2001. Feedback mechanisms and sensitivities of ocean carbon uptake under global warming. Tellus, B, 53:564-592.         [ Links ]

Redfield, A., B. Ketchum & F. Richards. 1963. The influence of organisms on the composition of sea water, in The Sea, edited by M. Hill, pp. 26-77, Interscience, New York.         [ Links ]

Sarmiento, J. L., T. M. C. Hughes, R. J. Stouffer & S. Manabe. 1998. Simulated response of the ocean carbon cycle to anthropogenic climate warming. Nature, 393:245-249.         [ Links ]

Sarmiento, J.L. et al. 2004. Response of ocean ecosystems to climate warming. Global Biogeochem. Cycles 18:GB3003, 1-23         [ Links ]

Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons