One of the primary goals of the global change science community is to quantify the terrestrial sources and sinks of atmospheric carbon dioxide. Among the highest priorities in this area are the need to:

identify and quantify the northern latitude terrestrial carbon sink
quantify the contribution of deforestation in tropical regions to the atmospheric CO₂ budget
determine the rate at which carbon is being sequestered in regrowing forests in tropical and temperate regions

The impetus behind these goals and priorities is developing the ability to predict atmospheric concentrations of CO₂ and greenhouse warming potential over the next century. Achieving this goal will require that we identify not only the current spatial distribution fo terrestrial source and sinks, but also the driving mechanisms. One source of information about the mechanisms regulating interannual variability in atmospheric CO₂ comes from its seasonal amplitude. In the Northern Hemisphere, the seasonal cycle of CO₂ is driven primarily by fluxes from the terrestrial biosphere. Over the last several decades, the seasonal amplitude of atmospheric CO₂ at high northern latitude stations has risen significantly (Keeling et al. 1996, Manning 1993). In addition to the long-term trend, the time sercies of CO₂ records shows condisterable interannual variability, with variations of 3 ppmv existing between high and low amplitude years (Figure 1).

In terms of understanding the terrestrial sources for the observed interannual variations in the atmospheric concentration of CO₂ most of the efforts have focused on improving the understanding of net primary production (NPP), heterotrophic respiration, and net ecosystem production (NEP). However, the atmospheric record reflects net biome production (NBP), which includes CO₂ emissions from biomass burning. Efforts are underway to estimate the spatial and temporal emissions of CO₂ from fire (Hao and Liu 1994; French et al. 1999; Mack et al. 1996) and to relate these emissions to the atmospheric CO₂ record (Witternburg et al. 1998).

Fires are ubiquitous to the boreal forest and can burn large areas when dry, warm conditions are present. Through combining information derived from the records of fire management agencies and analysis of satellite imagery, we know that fire ocurrence is highly variable in the boreal forest, ranging between 2 and 18 million ha yr^-1. It is also estimated that between 10 and 40t carbon ha^-1 burned are released by fires in boreal forests, resulting in a potential range of fire carbon emissions in high northern latitudes between 0.02 and 0.72 Gt C yr^-1.

The effect of biomass burning on the seasonal amplitude can be roughly assessed by examining the growing season net flux (GSNF) in high northern latitudes (Fung 1983). GSNF is defined as the net atmosphere-biosphere flux integrated over the growing season (i.e. when photosynthesis is greater than respiration). For tundra and boreal forest biomes, GSNF has been estimated within a range of 2 to 4 Pg C yr^-1. With most atmospheric transport models, GSNFs of this magnitude generate seasonal amplitudes at high northern latitudes on the order of 15 to 30 ppmv, comparable with flask observations (Law et al. 1996). With the GISS transport model (Fung 1991), a tundra and boreal forest GSNF of 2.5 Pg C yr^-1 translates to a seasonal amplitude of >16 ppmv at Point Barrow (Randerson et al. 1997). To a first approximation, it follows that a year with a 0.5 Pg C yr^-1 anomaly from biomass burning will cause am amplitude change of ~3 ppmv. This anomaly is large relative to observed variability (Figure 1c) and relative to the extremely high precision of the flask measurements (<0.1 ppmv) (Conway et al. 1994). While this example illustrates the possible magnitude of the impact, information abou the timing of fires within the growing season, the spatial distribution of fire, and burn severity is critical for a more thorough analysis.

Back to Boreal Forest Fire Home Page