Land Use, Land-Use Change and Forestry

Other reports in this collection

1.2.1.3. Inter-Annual and Decadal Variability of Atmospheric CO2 Concentrations

The uncertainty ranges in Table 1-2 result partly from our limited ability to determine accurately the gradual changes in the carbon balance resulting from human-induced emissions. In addition, however, variations in the atmospheric CO2 growth rate that have been recorded since 1960 imply that global terrestrial and oceanic carbon sources and sinks may vary significantly in time (Conway et al., 1994; Francey et al., 1995; Keeling et al., 1996a). Fossil fuel emissions, on the other hand, do not fluctuate much from one year to the next, whereas the exchange of atmospheric CO2 with the oceans and the terrestrial biosphere responds to inter-annual climate variations. High atmospheric CO2 growth rates have been recorded during three recent El Ni�o events-in 1983, 1987, and 1998-indicating a lower than normal uptake of atmospheric CO2 by the terrestrial biosphere and the oceans (Gaudry et al., 1987; Keeling et al., 1989; Keeling and Whorf, 1999). Conversely, low atmospheric CO2 growth rates were observed between 1991 and 1993, indicating enhanced uptake-particularly over the northern hemisphere (Ciais et al., 1995a,b; Keeling et al., 1996b).

Table 1-2: Average annual budget of CO2 perturbations for 1980 to 1989 (consistent with values given in Schimel et al., 1996) and 1989 to 1998 (note the 1-year overlap in the two decadal periods). Flows and reservoir changes of carbon are expressed in Gt C yr-1; error limits correspond to an estimated 90-percent confidence interval.

  1980 to 1989 1989 to 1998

1) Emissions from fossil fuel combustion and cement production
5.5 � 0.5
6.3 � 0.6(a)

  a) from Annex I countriesd

3.9 � 0.4a
3.8 � 0.4a

    i) from countries excluding those with economies in transition

2.6 � 0.3
2.8 � 0.3

    ii) from countries with economies in transition(d)

1.3 � 0.3a
1.0 � 0.3a

  b) from rest of worldd

1.6 � 0.3a
2.5 � 0.4a
 
2) Storage in the atmosphere
3.3 � 0.2
3.3 � 0.2b
 
3) Ocean uptake
2.0 � 0.8
2.3 � 0.8c
 
4) Net terrestrial uptake = (1) - [(2)+(3)]
0.2 � 1.0
0.7 � 1.0
 
5) Emissions from land-use change
1.7 � 0.8e
1.6 � 0.8f
 
6) Residual terrestrial uptake = (4)+(5)
1.9 � 1.3
2.3 � 1.3

a Based on emission estimates through 1996 by Marland et al. (1999) and estimates derived from energy statistics for 1997 and 1998 (British Petroleum Company, 1999).
b Based on atmospheric CO2 concentrations measured at Mauna Loa, Barrow, and South Pole (Keeling and Whorf, 1999).
c Based on ocean carbon cycle model (Jain et al., 1995) used in the IPCC Second Assessment Report (IPCC, 1996; Harvey et al., 1997) consistent with an uptake of 2.0 Gt C yr-1 in the 1980s.
d Annex 1 countries and countries with economies in transition (a subset of Annex 1 countries) defined in the FCCC. Emissions include emission estimates from geographic regions preceding this designation and include emissions from bunker fuels from each region.
e Based on land-use change emissions estimated by Houghton (1999) and modified by Houghton et la.(1999, 2000), which include the net emissions from wood harvesting and agricultural soils.
f Based on estimated annual average emissions for 1989-1995 (Houghton et al., 1999, 2000).

Ocean carbon models and available data suggest that the oceans contribute less to observed year-to-year changes in atmospheric CO2 concentration than does the terrestrial biosphere (Winguth et al., 1994; Le Qu�r� et al., 1998; Lee et al., 1998; Feely et al., 1999; Rayner et al., 2000). The terrestrial biosphere therefore appears to drive most of the inter-annual variation in CO2 flows. The way ecosystems respond to climate variability is not well understood, although the correlation and lag-correlation of inter-annual variability between CO2 growth rates, climate, and the remotely sensed "greenness" normalized difference vegetation index (NDVI), which is related to photosynthesis, is illustrative (Braswell et al., 1997; Myneni et al., 1997).

When terrestrial biogeochemical models are forced with realistic year-to-year changes in temperature and precipitation, they can simulate changes in the global and regional biosphere and associated changes in CO2 exchange with the atmosphere (Kindermann et al., 1996; Tian et al., 1998). These models can reproduce the magnitude and to some extent the phase of observed inter-annual variability of atmospheric CO2 concentrations, though different processes have been implicated in attempts to explain the observed fluctuations (e.g., Heimann et al., 1997). There are still differences in detail that have not been resolved.

Shifts in magnitude and phase of atmospheric CO2 fluctuations on a decadal time scale suggest that seasonality of terrestrial biotic fluxes has been changing slowly at mid to high northern latitudes (Keeling et al., 1996b; Randerson et al., 1997). Remotely sensed data (Myneni et al., 1997), as well as phenological observations (Menzel and Fabian, 1999), independently indicate a longer growing season in the boreal zone and in temperate Europe during recent decades.



Other reports in this collection