<i>Rolling</i> vs. <i>seasonal</i> PMF: real-world multi-site and synthetic dataset comparison

Abstract. Particulate matter (PM) has become a major concern in terms of human health and climate impact. In particular, the source apportionment (SA) organic aerosols (OA) present submicron particles (PM1) gained relevance as an atmospheric research field due to diversity complexity its primary sources secondary formation processes. Moreover, relatively simple but robust instruments such Aerosol Chemical Speciation Monitor (ACSM) are now widely available for near-real-time online determination composition non-refractory PM1. One most used tools SA purposes is source-receptor positive matrix factorisation (PMF) model. Even though recently developed rolling PMF technique already been OA on ACSM datasets, no study assessed added value compared more common seasonal method using practical approach yet. this paper, both techniques were applied synthetic dataset nine European datasets order spot main output discrepancies between methods. The advantage was that methods' outputs could be expected “true” values, i.e. original values. This revealed similar results amongst methods, although profile's adaptability feature proved advantageous, it generated profiles moved nearer truth points. Nevertheless, these highlighted impact profile anchor solution, use different with respect led significantly multi-site study, while differences generally not significant when considering year-long periods, their importance grew towards shorter time spans, intra-month or intra-day cycles. As far correlation external measurements concerned, performed better than globally ambient investigated here, especially periods seasons. comparison coincide rolling–seasonal similarity reporting moderate improvements. Altogether, provide solid evidence robustness methods overall efficiency proposed approach.