Cluster Analysis of Urban Ultrafine Particles Size Distributions

Research output: Contribution to journalArticle

Authors

  • Dayana Milena Agudelo-Castaned
  • Elba C. Teixeira
  • Marcel Braga
  • Silvia Beatriz Alves Rolim
  • Luis F.O. Silva
  • David C.S. Beddows
  • Xavier Querol

Colleges, School and Institutes

External organisations

  • Birmingham University

Abstract

Measurements of particle size distribution was made in one location of an urban area in the period January–September/2015 in order to understand the sources and mechanisms influencing ultrafine particle (UFP) number concentrations (PNC2.5-250) using a Scanning Mobility Particle Sizer Spectrometer (SMPS). k-means cluster analysis was applied to interpret the sources, temporal and spatial trends of UFP. Eight clusters were obtained. Main PSD patterns of each cluster, mean concentration of other air pollutants tracing specific sources and processes, and that of meteorological variables, as well as the hourly and seasonal frequencies of occurrence were used to support the interpretation of their origin. Thus, clusters were attributed to traffic rush hours, midday summer new particle formation, diurnal new particle formation and growth, growth of nucleated and other urban particles, urban background, regional and urban background and regional and urban background on cold nights. Many PSDs of the clusters were dominated by nucleation mode particles: midday nucleated fresh particles, photochemically induced (NPF); diurnal nucleation episodes (NPF2); growth of nucleated particles in nocturnal aging (GNPF). Origins of the clusters were related to local/regional sources (mostly traffic and biomass burning), atmospheric processes (photochemical formation and growth) and urban/regional background. Results clearly shows that traffic is a major UFP source in nucleation mode and occurred in higher concentrations in winter (08:00 to 12:00 h) during traffic rush hours, and at night. Photochemical nucleation occurred with a relatively low frequency but yielding very high PNC.

Details

Original languageEnglish
Pages (from-to)45-52
Number of pages8
JournalAtmospheric Pollution Research
Volume10
Issue number1
Early online date12 Jun 2018
Publication statusPublished - 1 Jan 2019

Keywords

  • Clusters analysis, Nanoparticles, Particle number concentration, Particle size distribution, Ultrafine particles