Mercury emissions from global biomass burning: Spatialand temporal distribution

This chapter represents a new addition to the UNEP global mercury budget: the mercury emissions from biomass burning, here defined as emissions from wildfires and prescribed burns, and excluding contributions from bio-fuel consumption and charcoal production and use. The results cover the 1997-2006 timeframe. The average annual global mercury emission estimate from biomass burning for 1997-2006 is 675 ± 240 Mg yr^-1. This accounts for 8% of all current anthropogenic and natural emissions. The largest Hg emissions are from tropical and boreal Asia, followed by Africa and South America. They do not coincide with the largest carbon biomass burning emissions, which originate from Africa. Our methodology for budget estimation is based on a satellite-constrained bottom-up global carbon fire emission database (GFED version 2), which divides the globe into regions with similar ecosystems and burn behaviour. To estimate mercury emissions, the carbon model output is paired with regional emission factors for Hg, EF(Hg). There are large uncertainties in the budget estimation associated with burned area, fuel mass, and combustion completeness. The discrepancy between the model and traditional ground based assessments (e.g. FRA, 2000) is unacceptably large at this time. Of great urgency is the development and validation of a model for mercury cycling in forests, accounting for the biogeochemistry for each region. This would provide an understanding of the source/sink relationship and thus mercury accumulation or loss in ecosystems. Limiting the burning of tropical and boreal forests would have two beneficial effects: reducing the source of mercury releases to the atmosphere from burning, and maintaining a sink for atmospheric mercury. Restricting the global release mercury would reduce the vegetation/soil pools, and the potential Hg release in case of fire.