1.2.1.2. A More Detailed Analysis of the Carbon Budget and its Change during the Past 20 Years

Figure 1-1: The global carbon cycle, showing the carbon stocks in reservoirs (in Gt C = 1015 g C) and carbon flows (in Gt C yr-1) relevant to the anthropogenic perturbation as annual averages over the decade from 1989 to 1998 (Schimel et al., 1996, Tables 2.1 and 2.2). Net ocean uptake of the anthropogenic perturbation equals the net air-sea input plus runoff minus sedimentation (discussed by Sarmiento and Sundquist, 1992).

Carbon in the form of inorganic and organic compounds, notably CO2, is cycled between the atmosphere, oceans, and terrestrial biosphere (Figure 1-1). The largest natural exchanges occur between the atmosphere and terrestrial biota (GPP about 120 Gt C yr^-1, NPP about 60 Gt C yr^-1) and between the atmosphere and ocean surface waters (about 90 Gt C yr^-1). The atmosphere contains about 775 Gt C; the residence time for a CO2 molecule in the atmosphere is therefore only about 2.5 years. The characteristic adjustment times between reservoirs in response to perturbations to the system, however, are on the order of decades to centuries (Schimel et al., 1996).

The oceans, vegetation, and soils are significant reservoirs of carbon; they actively exchange CO2 with the atmosphere. Oceans contain about 50 times as much carbon as the atmosphere, predominantly in the form of dissolved inorganic carbon. Ocean uptake of carbon is limited, however, by the solubility of CO2 in seawater (including the effects of carbonate chemistry) and the slow rate of mixing between surface and deep-ocean waters. Terrestrial vegetation and soils contain about three and a half times as much carbon as the atmosphere; the exchange is controlled by photosynthesis and respiration.

The amount of carbon stored globally in soils is much larger than that in vegetation (Table 1-1). Soil is a major carbon pool in all biomes, whereas carbon stocks in vegetation are predominantly in the forest biomes. Boreal forests have a larger proportion of carbon stored in soils than in trees, compared with temperate or tropical forests. There are wide local variations, however, in the amounts and proportions of carbon per unit ground area in vegetation and soil within each biome (see Section 1.3).

The average global carbon budget for the 1980s (1980 to 1989) (Schimel et al., 1996) has been reassessed for the most recent decade from 1989 to 1998. There are some significant differences between the two decades (Table 1-2). Emissions from fossil fuel combustion and cement production have increased by about 0.8 Gt C yr^-1 (based on estimates through 1996 by Marland et al., 1999, and energy statistics for 1997 and 1998 by British Petroleum Company, 1999). There has been a slight decrease, however, in these emissions from Annex I countries in aggregate, with a marked decrease from Annex I countries with "economies in transition." The increase in these emissions from non-Annex I countries in aggregate has been about 0.9 Gt C yr^-1. The rate of increase in the atmospheric stock of carbon, on the other hand, has remained about the same (Keeling and Whorf, 1999). Although the net ocean uptake appears to have increased somewhat (Jain et al., 1995; Harvey et al., 1997), the difference between emissions resulting from the burning of fossil fuels and cement production, on the one hand, and atmospheric and oceanic uptake, on the other, has increased-with the result that the net terrestrial uptake of carbon for the period 1989-1998 was probably 0.7 ± 1.0 Gt C yr^-1.

Precise molecular oxygen (O2) measurements in the atmosphere make it possible to quantify the net global terrestrial carbon flux and the oceanic uptake of carbon in an independent manner. Reconstruction of the mean atmospheric O2 trend from air enclosed in bubbles in glacier ice (Battle et al., 1996) and air archived in tanks yields a net terrestrial uptake of 0.6 ± 0.9 Gt C yr^-1 (±1 standard deviation) for the 1980s. High-precision atmospheric observations (Keeling et al., 1996b) yield a value of 0.9 ± 0.7 Gt C yr^-1 for the period 1990-1997. Thus, there is satisfactory consistency between the estimates from the two approaches.

Factors that influence the net terrestrial uptake of carbon include the direct effects of land use and land-use change (e.g., deforestation and agricultural abandonment and regrowth) (see Section 1.4) and the response of terrestrial ecosystems to CO2 fertilization, nutrient deposition, climatic variation, and disturbance (e.g., fires, wind-throws, and major droughts) (see Section 1.3). These natural phenomena may partially be indirect effects of other human activities: Many ecosystems are in some state of recovery from past disturbances. For the 1980s, the combination of estimates of the strength of these factors (Schimel et al., 1995) yields a value for net terrestrial uptake that is consistent with, but more uncertain than, the residual calculated in line 4 of Table 1-2. For the 1980s, Houghton (1999) estimates the net CO2 source from land-use change to be 2.0 ± 0.8 Gt C yr^-1, which was later revised to 1.7 ± 0.8 Gt C yr^-1 considering newer regional data (Houghton et al., 1999, 2000). Estimates for the most recent decade are 1.6 ± 0.8 Gt C yr^-1 based on regional data up to 1995 (Houghton et al., 1999, 2000). Yet from the revised carbon budget (Table 1-2) we can infer that the net global effect of all other factors has offset the source from land-use change, yielding a significant net terrestrial sink over the past 20 years. The residual terrestrial uptake for both decadal periods in Table 1-2 is comparable in size to the oceanic uptake.

IPCC Homepage