6.2.4 Changes in the carbon cycle and climate-carbon feedbacks

In the previous two sections, we have briefly described the influence of anthropogenic forcing on climate. In turns, climate changes have impacts on the biogeochemical cycles, leading to modifications of the radiative forcing with potential feedback effects on climate. Among all the complex mechanisms involved, we will focus here on CO₂ as it is the dominant anthropogenic gas (see Section 4.1.2.1).

As mentioned in Section 2.3.1, about half of the anthropogenic CO₂ emitted by fossil-fuel burning and changes in land use has stayed in the atmosphere. The remaining half is stored approximately equally in the ocean and terrestrial biosphere. However, this division of anthropogenic emissions between atmospheric, oceanic and land reservoirs will change in the future.

First, the changes in atmospheric CO₂ concentration itself modify the atmosphere-ocean and atmosphere/land CO₂ fluxes. The balance between H₂CO₃, HCO₃^- and carbonate ions CO₃^2- explains why the ocean is able to store large amounts of CO₂ (see section 2.3.2.1). In particular, the CO₂ which is transferred from the atmosphere reacts with the water to form H₂CO₃ and with carbonate ions (CO₃^2-) to obtain bicarbonate ions (HCO₃^-), the dominant form of inorganic carbon in the ocean:

The CO₂ flux from the atmosphere to the ocean during the 20^th and 21^st centuries will tend to decrease the availability of carbonates ions (at least on time scales from decades to centuries, see Section 6.3.1). This will reduce the efficiency of reaction (6.1) to form bicarbonates from CO₂. A larger fraction of the dissolved inorganic carbon (DIC) will remain as H₂CO₃, increasing the partial pressure of carbon dioxide in the ocean and thus reduce the oceanic uptake (see Eq. 2.38). As a consequence, the ocean will continue to store some anthropogenic CO₂, but its relative contribution will decrease.

Over land, the increase of CO₂ concentration in the atmosphere generally implies more assimilation and sequestration of carbon by the terrestrial biosphere through photosynthesis (see Equation 2.46). This CO₂ fertilisation effect is not related to any limitation of plant productivity by CO₂ availability in present-day conditions, but rather to the predicted better regulation of the plant/atmosphere gas exchanges through stomata in future. With high levels of CO₂, smaller exchanges are required for the same CO₂ uptake, implying less transpiration and thus an increase in plants’ efficiency of water use. However, many factors limit plant growth, including the availability of nutrients. The long term and large-scale effect of the CO₂ fertilisation effect have not yet been precisely assessed.

These biogeochemical effects (also referred to as concentration effects) will occur even in the absence of any climate change induced by changes in the atmospheric composition. Global warming will also reduce the oceanic solubility of CO₂ (see Section 2.3.2.1). This is one example of a positive climate-carbon cycle feedback. In addition, increased stratification and slower oceanic circulation (see Section 6.2.3) are expected to reduce the exchanges between the surface layers rich in anthropogenic carbon and the deeper layer. The deeper water does not yet contain a significant amount of anthropogenic carbon because of the relatively slow oceanic overturning and diffusion rates (see Section 1.3.2), so the slower renewal of surface waters will tend to induce higher levels of DIC at the surface and thus reduce the oceanic uptake of carbon, providing another positive climate/carbon feedback. Changes in marine biota could also lead to some feedback loops, but they are not currently well understood. Present-day models suggest that their role is relatively unimportant, but the modelling of marine ecosystems is still very simple, and more precise estimates of those effects are required.

Temperature and precipitation changes also affect the carbon cycle on land. Warming tends to accelerate decomposition in soils, which releases CO₂ to the atmosphere. The primary production is enhanced by warming in cold areas and by an increase in precipitation in dry areas. In addition, in warm, dry areas where water availability is a limiting factor, a decrease in precipitation produces a reduction in productivity and thus in the uptake of CO₂ by vegetation. In addition, climate changes influence the distribution of biomes (see, for instance, Section 4.3.3) as well as the frequency and extent of wildfires (savannah and forest fires) which emit substantial quantities of CO₂. This illustrates that both positive and negative carbon/climate feedbacks are expected over land in different regions.

In order to estimate the influence of the feedback between climate changes and the carbon cycle, simulations have been performed with climate models including a representation of the carbon cycle (Friedlingstein et al., 2006). In the first group of numerical experiments, both the carbon cycle and the climate were allowed to change in response to anthropogenic CO₂ emissions (SRES A2 scenario). In the second group, the models were again driven by CO₂ emissions, but the climate was kept constant. In other words, the increase in CO₂ was not associated with any change in the radiative forcing. Because of this constant climate, the climate/carbon feedback loops were inactive, and it was thus possible in this idealised set-up to measure the contribution of biological processes (the concentration effects) to the changes in the carbon cycle. By studying the difference between the two groups of models, it was then possible to make a first-order estimate of the influence of the carbon/climate feedback loops.

In the fully coupled experiments, where climate and the carbon cycle interact, the concentration of atmospheric CO₂ is predicted to be between 20 and 220 ppm higher than in the constant-climate system by 2010 (Fig. 6.14). The net climate/carbon feedback effect is positive for all the models. This clearly indicates that the modifications in climate mean that a larger fraction of anthropogenic CO₂ will remain in the atmosphere in the future. The models suggest that this is mainly due to the terrestrial biosphere, which will display changes in primary productivity and increased soil respiration in future.

The projections made by models including a carbon cycle driven by emissions of CO₂ can be compared to those directly driven by CO₂ concentrations (Fig. 6.15). It must be recalled here that the concentrations in the SRES scenarios (Section 6.1.2) were obtained using a carbon-cycle model that includes its own representation of climate/carbon feedbacks. Consequently, Figure 6.15 does not display the results of simulations with and without climate/carbon feedbacks, but rather illustrates the impact of different representations of those feedbacks. A first important feature is the large increase in the range of projections in the simulations including carbon-cycle models. Changes in the carbon cycle are thus a key source of uncertainty in climate projections. Second, in the majority of the cases, the projected temperature changes in 2100 are larger in the coupled climate/carbon cycle models, leading to a range of temperature increases of 2.3-5.6°C for Scenario A2.

Another consequence of the flux of anthropogenic carbon from the atmosphere to the ocean is oceanic acidification (see Eqs. 2.39 to 2.41). Over the period 1750-1994, the surface pH of the global ocean decreased by about 0.1. The expected decrease by 2100 is about 0.3-0.4 for a standard scenario, the precise value depending on the level of CO₂ emissions. By the year 3000, the decrease may be as high as 0.7. This would lead to pH values lower than those estimated for the last few hundred million years.

This ocean acidification increases the solubility of CaCO₃, (see Section 4.3.1); this could also be related to the reduced CO₃^2- concentration due to oceanic uptake of CO₂. This will have a clear impact on CaCO₃ production by corals as well as by calcifying phytoplankton and zooplankton, and thus on their life cycles (see Section 2.3.2.2). The aragonite produced by, for instance, corals, will be particularly influenced by this change as it is less stable than calcite.