Past and present natural and anthropogenic contributions

The results of flux inversion can be considered by latitude band. Since the mid-1980s, calculations have shown that global interannual variability is essentially explained by fluctuations in the tropical zone. Variations were marked by two major excursions in the early and late 1990s. The latter is linked to the extreme El Niño event of 1998, during which a drought intensified forest fires in Asia, generating a significant flow of pyrogenic _CH4.

The 1991-1992 anomaly is linked to the major eruption of the Pinatubo volcano in 1991 and the collapse of the Soviet economy. The Pinatubo eruption provides an opportunity to study the complexity of the phenomena involved in _CH4 production and destruction. Volcanic aerosols led to a reduction in UV radiation and therefore in the production of -OH radicals, which limited the destruction of _CH4 by oxidation. In addition, summer cooling linked to the climatic forcing of the eruption led to a decrease in humid tropical sources of _CH4.

Source inversion calculations enable us to draw up a temporal balance sheet by process. Flooded areas are responsible for the greatest inter-annual variations in fluxes, while biomass fires contribute significantly less. Anthropogenic emissions have been declining since the early 1990s, leading to a drop in atmospheric growth rates over the decade. In addition to the collapse of the Soviet economy with its drop in fossil fuel extraction, several developed countries have taken steps to reduce emissions from waste storage. In the early 2000s, anthropogenic emissions began to rise again, which is thought to be linked to Chinese growth associated with intense coal production.

A complementary upstream approach comes from the simulation of methane by explicitly modeling flows with a numerical model representing the processes of the carbon cycle. These modeled fluxes are then used as inputs for an atmospheric chemistry and transport model.

Dynamic Global Vegetation Models(DGVM) represent the terrestrial biosphere with around ten plant functional types(PFT). The biosphere is discretized over the continents on a global scale. Empirical, macroscopic laws parameterize biological and physiological processes: for example, plant respiration as a function of temperature or humidity, the effects of atmospheric _pCO2 on plant productivity..

Specific numerical modules deal with _CH4 formation and degradation processes. Local mechanisms are multiple and complex, from the degradation of soil organic matter by methanogenic bacteria to the diffusion of gaseous _CH4 into the atmosphere. Nevertheless, these processes can be approximated by simple laws at the ecosystem level : for example, the temperature sensitivity of methane fluxes broadly follows an Arrhenius law. The intercomparison of _CH4 flux simulations by these models still reveals numerous disagreements, particularly at wetland level. Another advantage of these models is that they can be used to study their response to major perturbations in temperature, humidity and _pCO2.

The synthesis ofbottom-up andtop-down approaches shows the compatibility of estimates, at regional level and for the different types of processes leading to _CH4 emissions: wetlands, biomass combustion, fossil fuels, agriculture, landfills, etc

The overall picture for the decade 2000 shows natural sources of around 347 Tg/year and anthropogenic sources of around 315 Tg/year. The estimated " sink " of atmospheric _CH4 degradation by -OH and other mechanisms is around 632 Tg/year, which does not fully offset annual emissions (total 662 Tg/year). This imbalance is at the root of the increase in atmospheric content described above.