Abstract. We present a comprehensive overview of particulate air quality across the five major metropolitan areas of South Africa (Cape Town, Bloemfontein, Johannesburg and Tshwane (Gauteng Province), the Industrial Highveld Air Quality Priority Area (HVAPA), and Durban), based on a decadal (1 January 2000 to 31 December 2009) aerosol climatology from multiple satellite platforms and a detailed analysis of ground-based data from 19 sites throughout Gauteng. Data include Aerosol Optical Depth (AOD550, 555) from Aqua (550 nm), Terra (550 nm), and MISR (555 nm) platforms, Ängström Exponent (α550/865, 470/660) from Aqua (550/865 nm) and Terra (470/660 nm), Ultraviolet Aerosol Index (UVAI) from TOMS, and model results from the Goddard Ozone Chemistry Aerosol Radiation and Transport (GOCART) model. Results in Cape Town are distinct, owing to a typically clean, marine airmass origin and infrequent continental influence. At continentally-influenced sites, AOD550, AOD555, α550/865, α470/660 and UVAI reach maxima (0.12–0.20, 1.0–1.8, and 1.0–1.2, respectively) during late winter and early spring (August–October), coinciding with a period of enhanced dust generation and the maximum frequency of close-proximity and subtropical fires identified by MODIS Fire Information for Resource Management System (FIRMS). The adjacent metropolitan and industrial Gauteng and HVAPA areas have been identified as a megacity based on NO2 concentrations, but AOD is a factor of 3–6 lower than other megacities worldwide. GOCART results suggest that the contributions of organics and black carbon to AOD are significantly enhanced during biomass burning season (ASO), but that sulfate is the most significant contributor to AOD (~70–80%) through the rest of the year. Dust appears to be underestimated by GOCART emissions inventories at continentally-influenced metropolitan areas of South Africa. Ground monitoring sites were classified according to site type: (1) township and informal settlement sites with domestic burning influence, (2) urban and suburban residential sites with no domestic burning in the immediate vicinity, (3) industrial sites, and (4) one traffic site situated at a major freeway interchange. PM10 concentrations in township areas are 56% higher than in developed residential areas and 78% higher than in industrial areas as an annual average, with PM10 in townships 63 and 136% higher than developed residential and industrial areas, respectively, in winter (June, July, August). Monthly PM10 and PM2.5 concentrations reach annual maxima during winter at all sites except in industrial areas. At industrial sites, maxima in PM10 and PM2.5 tend to occur during summer (December–February), when photochemical generation of secondary aerosol is expected and when deep and unstable boundary layers allow high stack emissions (emitted above the boundary layer during winter) to reach the ground in close proximity to point sources. Diurnal profiles of PM10 and PM2.5 display maxima during morning (06:00–09:00 LT) and evening (17:00–22:00 LT) at nearly every site – especially during winter – and underscore the importance of domestic burning as a major source of primary particles. Multi-year averages indicate that evening maxima at some township sites average in excess of 400 μg m−3. These results from the urban/industrial Gauteng area quantitatively confirm previous studies suggesting that the lowest-income populations of South Africa experience the poorest air quality, and demonstrate that domestic burning results in frequent exposure to high concentrations of particulate pollution in the region comprising the cities of Johannesburg and Tshwane. While remotely-sensed data are frequently used as a proxy for ground air quality, we report poor correlations between PM concentrations and satellite parameters and suggest that this practice is not appropriate in metropolitan South Africa. Disagreement between satellite and ground data may be attributed to a number of factors: (1) vertical inhomogeneity and stratified pollution layers aloft during much of the year, (2) extremely shallow winter boundary layers, (3) discrepancy between satellite passover times and elevated diurnal PM concentrations, and (4) poor spatial resolution of satellites compared with highly localized PM sources. While remotely-sensed data provide a good picture of regional, seasonal properties of column aerosol, a complete understanding of South Africa's air quality at the ground will necessitate more extensive monitoring at the ground and intensive, multi-platform campaigns to understand the relationship between ground and satellite data.
The Concentrations and Removal Effects of PM10 and PM2.5 on a Wetland in Beijing
