Recommendations for reporting equivalent Black Carbon (eBC) concentration based on long-term pan-European in-situ observations

To incorporate Equivalent Black Carbon (eBC) as a new variable in air quality (AQ) guidelines and to develop effective mitigation strategies, it is crucial to estimate its mass concentration consistently throughout the AQ monitoring networks (AQMNs) with minimal uncertainties. A reliable determination of eBC mass concentrations derived from filter absorption photometers (FAPs) measurements depends on the appropriate quantification of the mass absorption cross-section (MAC) for converting the absorption coefficient (b_abs) determined from FAPs measurements to eBC. Several studies have shown substantial variability in MAC due to local and regional variability in e.g., eBC sources, burning conditions, and eBC internal mixing among others. This MAC variability may lead to considerable uncertainty in eBC estimation. This study investigates the spatial-temporal variability of the MAC obtained from simultaneous elemental carbon (EC) measurements and b_abs determination performed following the ACTRIS procedures at 22 sites. We compared different methodologies for retrieving eBC integrating different options for calculating MAC, including locally derived MAC, median MAC value calculated from 22 sites, and site-specific rolling regression MAC. The eBC concentrations that underwent corrections using these methods were identified as MeBC (median MAC), LeBC (local MAC), and ReBC (Rolling MAC) respectively. These corrected eBC concentrations were compared with eBC as directly provided by FAPs (NeBC; nominal instrumental MAC). The median MAC values were 7.8 ± 3.4 m² g⁻¹ from 12 aethalometers at 880 nm, and 10.6 ± 4.7 m² g⁻¹ from 10 MAAPs at 637 nm. Combining datasets obtained from these two types of FAPs resulted in a median value of about 10.7 ± 4.8 m² g⁻¹ at 637 nm. However, the experimental MAC values showed significant site and seasonal dependencies, with heterogeneous patterns between summer and winter in different regions. Pronounced differences (up to more than 50%) were observed between NeBC from FAPs and ReBC due to the differences observed between the experimental and nominal MAC values. Moreover, long-term trend analysis revealed a statistically significant (s.s.) decreasing trend in EC mass concentrations. Interestingly, we show that the corresponding corrected eBC trends are not independent of the way eBC is calculated, due to the variability of MAC. NeBC and EC decreasing trends were consistent at sites with no significant trend in experimental MAC. Conversely, where MAC showed a s.s. trend, the NeBC and EC trends were not consistent while ReBC concentration followed the same pattern as EC. These results underscore the importance of accounting for MAC variations when deriving eBC measurements from FAPs and emphasize the necessity of incorporating EC observations to constrain the uncertainty associated with eBC. Thus, this study recommends the use of co-located measurements of b_abs and EC mass concentrations by expanding monitoring networks to include regular EC sampling. However, in situations where EC observations are unavailable, we recommend applying the default MAC value of around 10 m² g⁻¹ recommended by ACTRIS when b_abs is provided by MAAP at 637 nm and the MAC value of 7.8 m² g⁻¹ when b_abs is provided by aethalometers at 880 nm.