EXPLOITING ADVANCED ATMOSPHERIC AEROSOL CHARACTERISATION TO DEVELOP CUTTING-EDGE APPROACHES FOR SOURCE APPORTIONMENT

Atmospheric aerosols are very complex systems due to the huge variety of emission sources, size distributions, chemical properties, and transformation processes. Although the characterisation of atmospheric aerosols is still challenging, their study is essential since they have different impacts both at global and local scales. Among the various aerosol-related topics under investigation, the scientific community is thus putting much efforts into identifying and quantifying atmospheric aerosol emission sources (source apportionment). In this way, policymakers can develop effective control strategies to mitigate atmospheric aerosol adverse impacts. In recent years, in parallel to more traditional filter-based aerosol characterisation, advanced instrumentation has been developed to gain more insights on atmospheric aerosol properties (e.g., by increasing the sampling time resolution, developing size-segregated samplers, or identifying specific compounds that can be useful to trace specific sources). However, this detailed characterisation is still not properly exploited to the maximum extent in source apportionment models, since it is difficult to set up models which can deal with such sophisticated information together with more traditional data. This PhD work was thus aimed at developing and improving cutting-edge modelling approaches for source apportionment which can put to the maximum use advanced and comprehensive characterisation of atmospheric aerosol. Firstly, in the framework of the PRIN-2017 RHAPS project, the recently developed dispersion-normalised multi-time receptor modelling approach was explored to decouple the influence of atmospheric dilution from emission strength and atmospheric chemistry. The application of this approach is still one of the few in the literature, and it is the first at a Po Valley site (one of the most hot-spot polluted areas in Europe) where aerosol concentrations are so largely affected by atmospheric dilution. This study allowed to better understand the origin of the observed aerosol mixtures impacting the receptor site. Secondly, the power of size-segregated aerosol data was exploited to gain detailed insights into emission sources. A methodology to robustly determine the size distributions of atmospheric aerosol chemical components was developed. Moreover, the completely novel multi-time and multi-size receptor modelling approach was developed to use as input size-segregated compositional datasets with their native time resolution, allowing to obtain more refined information on emission sources. Lastly, the source apportionment method relying on radiocarbon measurements on separated carbon fractions was exploited in this PhD work in two different frameworks. On one hand, inside the INFN-ISPIRA research project, a new sample preparation line for radiocarbon measurements was set up in Milan, validated, and exploited to perform a first source apportionment study on samples collected in the RHAPS project. On the other hand, during my Erasmus+ Traineeship at the LARA (Laboratory for the Analysis of Radiocarbon with Accelerator Mass Spectrometry) at the University of Bern, the analysis of a multi-year dataset was performed by injecting the samples in gaseous form into the LARA facility. Both laboratories are still among the few ones in the world able to separate in detail the sources of carbonaceous aerosol thanks to the possibility of performing radiocarbon measurements on separated elemental and organic carbon fractions.