PM10 source apportionment in a coastal african megacity (Luanda, Angola): A quantitative basis for air quality management

Abstract

Urban air quality in African megacities remains under-represented in observational studies, and Luanda (Angola) is no exception, despite rapid urbanisation and evidence of high particulate levels. This study presents the first PM10 source apportionment for Luanda, based on daily samples collected from June to November 2023 using gravimetric samplers and chemically characterised for water-soluble ions, carbonaceous fractions, elements and levoglucosan. Sources were resolved with the Positive Matrix Factorisation model, supported by local meteorology, HYSPLIT back-trajectories and NASA FIRMS active-fire detections. Model diagnostics (Q(Robust)/Q(Expected), residuals, Bootstrap, DISP and Fpeak) indicated a stable six-factor solution that reproduced measured PM10 with high agreement (R2 = 0.95; RMSE = 3.95 μg m−3). Traffic-related emissions dominate PM10 (37.6%), represented by a mixed exhaust/non-exhaust factor encompassing tailpipe emissions, brake and tyre wear, and road-dust resuspension. Metal-related industrial emissions are represented by a Zn–Pb–Cd–Sn-rich factor attributed to non-ferrous metallurgy and galvanised/electronic-scrap handling and burning (10.8%), and a Ni-centred metallurgical factor with a potential aviation influence (6.4%). Natural and geogenic sources contribute 29.6%, divided between crustal material + resuspended dust (16.5%) and marine aerosol (13.1%). A mixed secondary aerosol and biomass-burning factor (SA + BB) accounts for the remaining 15.5% and is associated with regionally transported, chemically aged plumes linked to inland fire activity. Overall, traffic and metal-related activities dominate PM10 in Luanda and provide a first quantitative basis for air quality management in this coastal African megacity.