Several different inventories of global and regional anthropogenic and biomass burning emissions are assessed for the 1980–2010 period. The species considered in this study are carbon monoxide, nitrogen oxides, sulfur dioxide and black carbon. The inventories considered include the ACCMIP historical emissions developed in support of the simulations for the IPCC AR5 assessment. Emissions for 2005 and 2010 from the Representative Concentration Pathways (RCPs) are also included. Large discrepancies between the global and regional emissions are identified, which shows that there is still no consensus on the best estimates for surface emissions of atmospheric compounds.
At the global scale, anthropogenic emissions of CO, NOx and SO2 show the best agreement for most years, although agreement does not necessarily mean that uncertainty is low. The agreement is low for BC emissions, particularly in the period prior to 2000. The best consensus is for NOx emissions for all periods and all regions, except for China, where emissions in 1980 and 1990 need to be better defined. Emissions of CO need better quantification in the USA and India for all periods; in Central Europe, the evolution of emissions during the past two decades needs to be better determined.
The agreement between the different SO2 emissions datasets is rather good for the USA, but better quantification is needed elsewhere, particularly for Central Europe, India and China. The comparisons performed in this study show that the use of RCP8.5 for the extension of the ACCMIP inventory beyond 2000 is reasonable, until more global or regional estimates become available. Concerning biomass burning emissions, most inventories agree within 50–80%, depending on the year and season. The large differences between biomass burning inventories are due to differences in the estimates of burned areas from the different available products, as well as in the amount of biomass burned.