Near-real time episode identification: Long-range transport of Canadian wildfire aerosol

Air quality has significant impacts on human health, the environment and climate, with atmospheric aerosol playing a key role. As such, the development of methodologies for aerosol typing at a highly resolved time frame is of great value to monitoring networks as they can aid in identifying at a fast pace and at a fine time resolution which aerosol sources are present in peak events of particulate matter pollution. In this context, optical data can be exploited for fast aerosol typing purposes as instruments which measure absorption and scattering coefficients operate at a time resolution of up to minute time frames. In this work, we refined a previously developed aerosol typing methodology, namely a hierarchical clustering of Absorption/Scattering Ångström Exponents, to identify which aerosol sources are responsible for peak absorption and scattering episodes. As a case-study for testing the methodology, a dataset collected in Milan (Italy) was used; it comprises one-month multi-wavelength absorption and scattering coefficients detected at ground-level with 15-min resolution. The methodology proved to be effective in characterising the role of local sources in peak events and led to the identification of a long-range transport of Canadian wildfire aerosol which occurred in June 2025 and affected the whole of Europe including Milan. • Ground-based multi-wavelength optical data provided aerosol typing. • Absorption and scattering data over their 95th percentile used to detect episodes. • Aerosol types identified by AAE-SAE hierarchical clustering. • The cluster-based refined methodology was proved to be effective on a case-study. • The long-range transport of Canadian wildfire aerosol was effectively detected.