Complementary air mass analysis and aerosol classification for improved source apportionment

Photometric measurement parameters such as Single Scattering Albedo (SSA), Aerosol Optical Depth (AOD), and Angstrom Exponent (AE) can be used for aerosol classification and source identification using artificial intelligence mechanisms or simplified classification trees. Sometimes these methods are difficult to apply in incomplete datasets or when using vaguely defined classification algorithms that provide ambiguous responses about their origin. In such cases, the origin of atmospheric aerosols may be identified using air-mass trajectory models of atmospheric pollution. This study uses a NOAA HYSPLIT model, which, in addition to atmospheric photometric properties, significantly improves the analysis of results in discutable situations and may, by its nature, even replace the measurement of aerosol chemical composition when profiling sources is performed. The experimental part was developed using the AERONET dataset for Lampedusa (Italy) for the period 2020-2024.