Ambient sulphur dioxide (SO2) and particulate matter (PM10) concentrations measured in selected communities of north and south Durban.

Abstract

The industrial basin in the south of Durban is an area of ongoing contention between the residents and major industries, regarding environment health issues especially poor air quality resulting from industrial air pollution. This region is a result of poor urban planning that began in the early 1960's which saw rapid industrialisation alongside expanding communities, which has now resulted in a major environmental dilemma for the city of Durban, eThekwini Municipality. Durban is seen as a key area of growth in South Africa: it has the busiest harbour on the continent; it is a regional hub of the chemical industry and a major motor and metal manufacturer centre; and Durban's population continues to grow at ~4% per annum, all of which are destined to have significant environmental impacts. As part of a large epidemiological study and health risk assessment in this industrial basin, particulate matter smaller than 10 microns (PM10) and sulphur dioxide (S02) were measured in seven community sites across Durban: four in the South, (Wentworth, Bluff, Merebank and Lamontville) and three in the north (Kwamashu, Newlands East and Newlands West). The south sites are located in an industrial basin near two petroleum refineries and a paper mill, while the north comparison sites are ~25 km North West from major industries in the Basin. 24 hour PMIO samples were collected gravimetrically every day during four, three-week intensive phases and thereafter every 6th day using high, medium and low volume samplers. S02 was monitored every 10 minutes with active continuous analysers (European monitor labs and API) according to internationally accepted methods. Rigorous quality assurance methods were followed for both pollutants. S02 followed a distinct spatial distribution where the mean difference in S02 concentrations between the southern and the northern region was 6.7 ppb, while for PMIO similar concentrations were found across all sites with the highest mean concentration at Ngazana in the north (59 Ilg/m3) > Assegai in the south (~58 Ilg/m3) with all other sites ~ I to 10 Ilg/m3 less in mean concentration. S02 diurnal variations display two maxima from 5:00AM to 10:00AM being repeated in the latter part of the day from 20:00PM to 24:00PM. Seasonal pattern of PMIO and SOz to a lesser extent, display very similar mean variations for all sites - the highest levels seen in the colder months of May; June, July of 2004 and 2005. Weak to strong intersite correlations were found for SOz ranging from 0.16 to 0.22 among the south sites; 0.06 to 0.64 among the north sites, while PM10 ranged from 0.73 to 0.88 among the south sites and 0.86 to 0.91 among the north sites. A cross regional correlation of PMIO by sites displayed a moderate to strong correlation ranging from 0.73 to 0.88, while intrasite SOz with PMIO correlations displayed weak to moderate correlations from 0.35 to 0.53. Meteorological conditions wind speed, temperature, pressure and humidity differed across Durban. The difference in temperature and humidity between "summer" and "winter" was on average, approximately 7 °c and 10-15% respectively. These temperature and humidity patterns closely track the increase in SOz and PM10 during "winter" explaining the effects of winter inversions on pollutant levels. Also of interest is the variability of meteorological parameters between south and north Durban with the two regions being 35 km apart. Meteorological conditions impact differently on each pollutant e.g. rain is more likely to decrease PMIO concentrations than it would SOz In general relationships between pollutants and meteorological parameters differ on a site-bysite basis. For instance, wind direction at Assegai increases SOz levels whereas wind direction at Ngazana decreases SOz levels. Another independent variable that proved to be a consistent and important predictor for SOz and PM10 across most sites was the previous day's pollution events; this was a much stronger predictor for PM10 rather than SOz. These findings suggest that pollutants are not fully removed from the atmosphere during a 24 hour period and that the previous day's pollution levels will contribute to current levels, a finding that has important implication when implementing early warning pollution systems as envisaged for the Durban South Basin.