Abstract. Our understanding of the global black carbon (BC) cycle is essentially qualitative due to uncertainties in our knowledge of its properties. This work investigates two source of uncertainties in modelling black carbon: those due to the use of different schemes for BC ageing and its removal rate in the global Transport-Chemistry model TM5 and those due to the uncertainties in the definition and quantification of the observations, which propagate through to both the emission inventories, and the measurements used for the model evaluation.

The schemes for the atmospheric processing of black carbon that have been tested with the model are (i) a simple approach considering BC as bulk aerosol and a simple treatment of the removal with fixed 70% of in-cloud black carbon concentrations scavenged by clouds and removed when rain is present and (ii) a more complete description of microphysical ageing within an aerosol dynamics model, where removal is coupled to the microphysical properties of the aerosol, which results in a global average of 40% in-cloud black carbon that is scavenged in clouds and subsequently removed by rain, thus resulting in a longer atmospheric lifetime. This difference is reflected in comparisons between both sets of modelled results and the measurements. Close to the sources, both anthropogenic and vegetation fire source regions, the model results do not differ significantly, indicating that the emissions are the prevailing mechanism determining the concentrations and the choice of the aerosol scheme does not influence the levels. In more remote areas such as oceanic and polar regions the differences can be orders of magnitude, due to the differences between the two schemes. The more complete description reproduces the seasonal trend of the black carbon observations in those areas, although not always the magnitude of the signal, while the more simplified approach underestimates black carbon concentrations by orders of magnitude.

The sensitivity to wet scavenging has been tested by varying in-cloud and below-cloud removal. BC lifetime increases by 10% when large scale and convective scale precipitation removal efficiency are reduced by 30%, while the variation is very small when below-cloud scavenging is zero.

Since the emission inventories are representative of elemental carbon-like substance, the model output should be compared to elemental carbon measurements and if known, the ratio of black carbon to elemental carbon mass should be taken into account when the model is compared with black carbon observations.