Shifting Dominant Air Pollution Sources in Sub-Saharan Africa’s Changing Urban Environment: Analysis of Elemental Composition and Source Contributions of Ambient Fine Particulate Matter (PM2.5) Pollution in Accra, Ghana

Abstract Urban and economic expansion across Sub-Saharan Africa (SSA) may be shifting the relative contributions of dominant ambient air pollution sources, influencing both pollutant concentrations and composition, along with their associated health risks. However, city-scale data on the composition and source contributions are scarce, hindering source- and sector-specific policy interventions. Using samples from a large-scale, citywide monitoring campaign, we characterized the elemental composition and source contributions of PM2.5 pollution across the Accra metropolis, a major SSA city. From April 2019 to June 2020, we collected over 600 weekly gravimetric (filter-based) PM2.5 samples from 146 unique sites spanning diverse land-use across the city. Filters were analyzed for several elements using energy dispersive x-ray fluorescence (ED-XRF), and positive matrix factorization (PMF) was applied to identify major sources, and changes in their relative contributions, at four legacy sites from a decade prior (2007-2008). Concentrations of elements associated with biomass burning (K, Br, Pb) and vehicle emissions (Ti, Fe, Cu, Zn) were highest in high-density residential and heavily trafficked areas, respectively. There was strong temporal variation in several elements (Al, Si, Fe, Ti, K, Ca, Mn) across seasons, with elevated concentrations during the dusty Harmattan season, when crustal dust accounted for 52-68% of total elemental PM2.5 mass. In the non-Harmattan period, we identified 4-6 major sources including biomass burning (comprising 11-32% of total PM2.5 elemental mass), traffic emissions and road dust (9-36%), solid and electronic waste combustion (0-35%), construction dust (0-30%), crustal dust (11-47%), and sea salt (8-67%). Compared to a decade prior, the contribution from biomass burning to overall PM2.5 has decreased by ∼16% (95% CI: 5.2-25.8; p-value = 0.017). Contrarily, the composite traffic/road dust factor remained stable, though increasing traffic emissions may have been offset by declining road-dust contributions. Most notably, contributions from solid and electronic waste burning have risen significantly (mean difference = 23%; 95% CI: 15.9-30.0; p-value = 0.002) to levels approaching those of biomass burning and traffic-related sources at legacy sites. Our findings indicate that household biomass use remains a dominant source of PM2.5 in Accra, although its relative contribution appears to have declined. This reduction may be partially offset by increasing contributions from solid and electronic waste burning in many locations. These shifts in PM2.5 composition and relative source contributions underscore the need for source- or sector-specific policy actions in the region.