Ambient organic carbon to elemental carbon ratios: influence of the thermal-optical temperature protocol and...

Ambient organic carbon (OC) to elemental carbon (EC) ratios are strongly associated with not only the radiative forcing due to aerosols but also the extent of secondary organic aerosol (SOA) formation. An inter-comparison study was conducted based on fine particulate matter samples collected during summer in Beijing to investigate the influence of the thermal-optical temperature protocol on the OC to EC ratio. Five temperature protocols were used such that the NIOSH (National Institute for Occupational Safety and Health) and EUSAAR (European Supersites for Atmospheric Aerosol Research) protocols were run by the Sunset carbon analyzer while the IMPROVE (the Interagency Monitoring of Protected Visual Environments network)-A protocol and two alternative protocols designed based on NIOSH and EUSAAR were run by the DRI analyzer. The optical attenuation measured by the Sunset carbon analyzer was more easily biased by the shadowing effect, whereas total carbon agreed well between the Sunset and DRI analyzers. The EC(IMPROVE-A) (EC measured by the IMPROVE-A protocol; similar hereinafter) to EC(NIOSH) ratio and the EC(IMPROVE-A) to EC(EUSAAR) ratio averaged 1.36 ± 0.21 and 0.91 ± 0.10, respectively, both of which exhibited little dependence on the biomass burning contribution. Though the temperature protocol had substantial influence on the OC to EC ratio, the contributions of secondary organic carbon (SOC) to OC, which were predicted by the EC-tracer method, did not differ significantly among the five protocols. Moreover, the SOC contributions obtained in this study were comparable with previous results based on field observation (typically between 45 and 65%), but were substantially higher than the estimation provided by an air quality model (only 18%). The comparison of SOC and WSOC suggests that when using the transmittance charring correction, all of the three common protocols (i.e., IMPROVE-A, NIOSH and EUSAAR) could be reliable for the estimation of SOC by the EC-tracer method.