Uncertainties in observational data on organic aerosol: An annual perspective of sampling artifacts in Beijing, China

- Positive artifact can cause an overestimation of OA concentrations by up to 40%.
- It remains a challenge to measure semivolatile OA based on filter sampling.
- The positive artifact can be effectively removed by an ACM denuder.
- The ACM denuder is small in size, easy to use and multi-functional.
- The ACM denuder is recommended for routine measurement of OA.

Current understanding of organic aerosol (OA) is challenged by the large gap between simulation results and observational data. Based on six campaigns conducted in a representative mega city in China, this study provided an annual perspective of the uncertainties in observational OA data caused by sampling artifacts. Our results suggest that for the commonly-used sampling approach that involves collection of particles on a bare quartz filter, the positive artifact could result in a 20-40 % overestimation of OA concentrations. Based on an evaluation framework that includes four criteria, an activated carbon denuder was demonstrated to be able to effectively eliminate the positive artifact with a long useful time of at least one month, and hence it was recommended to be a good choice for routine measurement of carbonaceous aerosol.