Search

Real-time measurements of particles in the 15–736 nm range have been obtained by a Scanning Mobility Particle Sizer to characterize the evolution of particle size distribution and new particle formation (NPF) events in an urban background area. The annual, weekly and diurnal variations of the modal (nucleation (N nuc), Aitken (N Ait) and accumulation (N acc)) particle concentrations were characterised. The N Ait and N acc registered their maximums in cold months during rush hours, in the morning (0600–0900 UTC) and in the afternoon (1700–2000 UTC), while the maximums for N nuc were reached in warm months during midday hours. N Ait , N acc and N total showed a significant negative correlation with wind speed and a different relationship with the planetary boundary layer (PBL) height by periods. In the warm period, a positive significant correlation between PBL and N nuc was registered, indicating that the higher dispersion promoted by a high PBL causes favourable conditions for the occurrence of NPF events (a low polluted atmosphere). NPF processes are one of the main sources of ultrafine particles (<100 nm) in the warm period. After a visual-based classification, 45 NPF events of type Ia (strong and with a good confidence level) were identified and analysed, occurring primarily between 1100 and 1500 UTC, mainly in spring and summer. In addition, a two-step method was developed for identifying NPF events: cluster analysis followed by discriminant analysis. The application of discriminant analysis to one of the clusters, grouping 93 days, enabled us to identify 55 of the 56 NPF events days included in the cluster. This method is a valuable tool for identifying NPF events quickly and effectively. [Display omitted] • Nucleation mode peaks were recorded in warm period due to NPF events at midday hours. • In the warm period, the higher PBL, the higher nucleation particle concentration. • O 3 and condensation sink were limiting factors for strong and clear NPF events. • A 2-step method was built for finding NPF events: a cluster plus a discriminant analysis. [ABSTRACT FROM AUTHOR]