Modelling oxidative potential of atmospheric particle: A 2-year study over France

The oxidative potential (OP) of particulate matter (PM) has emerged as a promising indicator of the adverse effects of PM on human health. In particular, OP is an indicator for oxidative stress in biological media through formation of reactive oxygen species. To provide a mapping of the spatial and temporal OP variability over France, we have developed a strategy to simulate the volume-normalized oxidative potential (OP_v) in the state-of-the-art CHIMERE air quality model over the metropolitan French territory for the years 2013 and 2014. To do so, we combined a measurement-derived and source specific intrinsic OP (OP_i) receptor modelling approach with Particle Source Apportionment Technology (PSAT) in CHIMERE. First, the model's ability to reproduce PM₁₀ concentrations and speciation was verified using in situ observations in mainland France. Furthermore, a mostly satisfying correspondence between receptor model and PSAT outputs was obtained considering their source specific chemical profiles. Simulated versus observed OP_v values showed median correlations ranging from 0.35 to 0.60 and mean fractional biases from -30 % to zero, depending on the OP assay considered (ascorbic acid AA, or dithiothreitol DTT) and the PM sources taken into account (i.e. two methods with different PM sources have been used, the reduced and the extended set methods). The modelled two-year average OP_v fields show greater spatial hot spots over large urban areas (especially along roadsides) compared to those for PM₁₀ distributions, due to elevated intrinsic OP_i values for the primary anthropogenic sources such as traffic and biomass burning. These effects are stronger for the AA compared to the DTT assays, and for a method with a reduced set compared to an extended set of sources. Overall, through the OP apportionment, these results advocate for reinforcing action plans to reduce emissions from road traffic as well as biomass burning emissions.