Towards a better characterization of submicron aerosol in the Mediterranean basin

Abstract

[eng] Atmospheric aerosol is an ensemble of atmospheric pollutants with a severe impact on human health and Earth climate. Particulate Matter (PM) effects vary depending on the composition and size. However, current air quality guidelines (AQG) from WHO and from the EU Commission only define threshold standards for bulk PM10 and PM2.5 concentrations. In terms of health, the smaller the particles, the deeper they penetrate into the respiratory system, and they can even diffuse into the bloodstream and circulate to other parts of the organism. The adverse effects on the tissues in which they are deposited depend on the PM composition. Regarding climate, PM size and composition relevantly affect many aerosol-radiation direct and indirect interactions, which regulate troposphere temperature. Consequently, the consideration of these properties is relevant for present and future climate descriptions.

Barcelona, the city where this study focuses, is located in the Mediterranean basin, a region with high complexity in terms of air pollution, hence, the nature of this enclave requires thorough monitoring of PM in order to protect the population from exposition accurately. This dissertation focuses on submicronic aerosol evolution over the last decade, with a more detailed study for May 2014-May 2015 and September 2017-October 2018.

The objective of this thesis is to describe the PM1 sources in Barcelona by means of source apportionment (SA) techniques. The Positive Matrix Factorisation (PMF) algorithm is one of the most widely used approaches for SA and is the tool used in this dissertation. A secondary aim of this study is the improvement of the SA methodology itself. This is accomplished by testing the outcomes of new methodologies involving the more automatic analysis and dataset junction with several approaches.

A field-deployed aerosol mass spectrometer was used at the Barcelona site for continuous PM1 measurements, and SA was performed on Organic Aerosol (OA). First, SA was tackled from a conventional methodology, the seasonal PMF, then, the novel rolling PMF methodology was tested and compared to the fore one. Finally, a comprehensive PM1 SA was performed based on an ensemble of different datasets coming from a variety of measurement techniques. These steps enabled a progressive aerosol composition understanding and acknowledgement of subsequent aerosol trends.

A PM1 concentrations decrease was found in the 2014-2018 period, a trend confirmed by other studies at the site. Its relative composition changed significantly; a decrease was found for SO42-, BC, and NH4+, while NO3- increased and OA levels were found stable. The OA SA revealed that its sources were: secondary OA (SOA, >55-70%), road traffic OA (12-19%), cooking-like OA (14-18%), and biomass burning OA (4-6%). These sources are similar to those reported in other sites across the Mediterranean region. In this study, all the primary OA sources were found in a clear decrease from 2014 to 2018. An increasing SOA proportion and SOA oxidation state were also observed. These increments could be explained by a likely increase in the oxidation capacity of the atmosphere, related to the accumulation of oxidative radicals reported in many cities. In

order to bridge the possible inter-annual variability in that period, the time period was elongated (2014-2021) detecting the same underlying trends.

With the aim of further climate and health impact aerosol impact assessment, this thesis provides mid-term PM1 sources diagnosis. SOA is especially concerning in terms of health effects, hence this pollutant is to be continuously monitored to deeply understand its precursors and formation mechanisms to design effective abatement policies.

[cat] L’aerosol atmosfèric és un conjunt de contaminants atmosfèrics amb un impacte sever sobre la salut humana i el clima. Els efectes del material particulat (PM) varien depenent de la seva composició i mida.

Barcelona, la ciutat en què es centra aquest estudi, està localitzada a la conca mediterrània, una regió d’alta complexitat en termes de contaminació de l’aire. Per tant, requereix un monitoratge conscienciós del PM per tal de protegir la població acuradament de la seva exposició. Aquesta tesi es focalitza en l’evolució del PM submicrònic (PM1) al llarg de l’última dècada, amb un estudi més detallat per als períodes maig 2014 - maig 2015 i setembre 2017 - octubre 2018.

L'objectiu d'aquesta tesi és descriure les fonts de PM1 a Barcelona per mitjà de tècniques de contribució de fonts (SA). L'algoritme de Factorització de Matriu Positiva (PMF) és un dels enfocaments més utilitzats per al SA o és l'eina emprada per a aquesta dissertació. Un objectiu secundari d'aquesta tesi és la millora de la pròpia metodologia del SA. Això és executat per mitjà del testatge a través de diferents metodologies.

Les mesures de PM1 es van dur a terme a través d’un espectròmetre de masses instal·lat a l'estació de Barcelona. El SA va ser executat per a l'Aerosol Orgànic (OA) submicrònic mesurat a partir d’aquest instrument, aplicant diferents metodologies per a avaluar-ne l’exactitud. Finalment, un estudi detallat del SA de PM1 va ser executat basat en un conjunt de dades provinents de diferents tècniques de mesura. Aquests passos van permetre la comprensió progressiva de la composició i el reconeixement de les tendències subjacents de l'aerosol.

Un dels resultats més rellevants consisteix en la detecció d'una tendència creixent del SOA independentment de la disminució del PM1, relacionat amb el decreixement de l'OA primari. A més, pel que fa al SA, aquesta tesi proposa diverses modificacions del protocol. El SOA és especialment preocupant pels seus efectes en la salut, per tant, aquest contaminant ha de ser contínuament monitorejat amb tècniques de SA per tal d'entendre els seus precursors i mecanismes de formació amb l'objectiu de dissenyar mesures de mitigació efectives.