Organic Aerosol Concentration, Composition, and Sources Analysis at Pretoria, South Africa Employing Fourier-Transform Infrared Spectroscopy (FT-IR) and Positive Matrix Factorization (PMF)
MUHAMMAD NAVEED ANWAR, Satoshi Takahama, Christopher Oxford, Randall Martin, Adele L. Igel, Rebecca Garland, Ann M. Dillner, University of California, Davis
Abstract Number: 55
Working Group: Source Apportionment
Abstract
The World Health Organization (WHO) estimates that about 91% of the global population live in regions with polluted air – resulting in approximately 7 million premature deaths annually. These deaths occur primarily in Low- and Middle-Income Countries (LMICs). Organic Matter (OM) often dominates PM2.5 mass warranting its quantification particularly in LMICs having limited PM measurements. In this study, we demonstrate the capability of Fourier-transform Infrared Spectroscopy (FT-IR) method for providing OM quantification and characterization in the Pretoria, South Africa (LMIC) at a Surface Particulate Matter Network (SPARTAN) site. The samples affected with Christiansen peak effect (scattering effect) were identified and a revised algorithm was employed to address this phenomenon. Polytetrafluoroethylene (PTFE) filters (MTL Corp., PT25DMCAN-PF03A), used for gravimetric, black carbon, FT-IR, ionic and elemental analysis, were collected with an AirPhoton SS5 Sampler over 24-hour intervals, every third day, from April 2021 to Dec 2022 (120 samples). The mean concentration of PM2.5 was 18.43 ug/m3. We present the daily, seasonal, and overall variation of organic and inorganic composition of collected PM2.5. Source classes and their geographical origins were determined through Positive Matrix Factorization (PMF) model, using the baseline corrected FT-IR Spectra, pollution roses, back trajectory analysis, correlation plots with inorganic species, and MODIS fire data products. Two sources were found: biomass burning & biogenic SOA (BB&SOA) and fossil fuel combustion (FFC) factors. BB&SOA was the dominant source in the winter and early spring while the FFC factor was ubiquitous with the dominant contribution in summer. The BB&SOA factor was further investigated for the individual factors with distinct origins: biogenic SOA, biomass burning, and industrial emissions mix source.