Simultaneous measurements of black carbon, PM10, ozone and NOx variability at a locally polluted island in the southern tropics

This paper shows a comparative study of particle and surface ozone concentration measurements undertaken simultaneously at two distinct semi-urban locations distant by 4 km at Saint-Denis, the main city of La Reunion island (21.5 degrees S, 55.5 degrees E) during austral autumn (May 2000). Black carbon (BC) particles measured at La Reunion University, the first site situated in the suburbs of Saint-Denis, show straight-forward anti-correlation with ozone, especially during pollution peaks (similar to 650 ng/m(3) and 15 ppbv, for BC and ozone respectively) and at night-time (similar to 90 ng/m(3) and 18.5 ppbv, for BC and ozone respectively). NOx (NO and NO2) and PM10 particles were also measured in parallel with ozone at Lislet Geoffroy college, a second site situated closer to the city centre. NOx and PM10 particles are anti-correlated with ozone, with noticeable ozone destruction during peak hours (mean similar to6 and 9 ppbv at 7 a.m. and 8 p.m. respectively) when NOx and PM10 concentrations exhibit maximum values. We observe a net daytime ozone creation (similar to 19 ppbv, DeltaO(3) similar to +4.5 ppbv), following both photochemical and dynamical processes. At night-time however, ozone recovers (mean similar to 11 ppbv) when anthropogenic activities are lower ([BC] similar to 100 ng/m(3)). BC and PM10 concentration variation obtained during an experiment at the second site shows that the main origin of particles is anthropogenic emission (vehicles), which in turn influences directly ozone variability. Saint-Denis BC and ozone concentrations are also compared to measurements obtained during early autumn (March 2000) at Sainte-Rose (third site), a quite remote oceanic location. Contrarily to Saint-Denis observations, a net daytime ozone loss (similar to 14.5 ppbv at 4 p.m.) is noticed at Sainte-Rose while ozone recovers (similar to 17 ppbv) at night-time, with however a lower amplitude than at Saint-Denis. Preliminary results presented here are handful data sets for modelling and which may contribute to a better comprehension of ozone variability in relatively polluted areas.