Black Carbon Emissions Generally Underestimated in the Global South as Revealed by Globally Distributed Measurements

Characterization of black carbon (BC) on a fine scale across the world is essential for understanding its climate and health impacts. However, a paucity of BC mass measurements in different parts of the world and the coarse scale resolution of global models have inhibited evaluation of global BC emission inventories. Here, we apply a novel dataset of globally distributed BC mass measurements from the Surface Particulate Matter Network (SPARTAN) and complementary measurement networks to evaluate contemporary BC emission inventories. We use a global chemical transport model (GEOS-Chem) in its high-performance configuration (GCHP) for high-resolution simulations to relate BC emissions to ambient concentrations for comparison with the measurements. Simulations using the Community Emissions Data System (CEDS) emission inventory exhibit skill (r ² = 0.73) in representing variability in SPARTAN measurements across primarily developed regions with low BC concentrations but exhibit pronounced discrepancy (r ² = 0.00019) across regions with high BC concentrations in the Global South. The model underestimates BC at most Global South sites with a normalized mean bias of -38%. Alternative inventories (EDGAR, HTAP) yield similar results. These findings motivate renewed attention to the challenging task of characterizing BC emissions from developing regions.