Roadside levels and traffic emission rates of PM2.5 and BTEX in Ho Chi Minh City, Vietnam

A monitoring program was designed and implemented to characterize roadside levels of PM2.5 and BTEX in Ho Chi Minh City, Vietnam, and to generate input data for CALINE (California LINE Source Dispersion Model) inverse modeling. Monitoring was done during Dec 2007-Jan 2008, on both weekdays and weekends, and yielded 284 hourly BTEX samples (adsorption tubes), 24 samples of 8 h-PM2.5 and 42 samples of 24 h-PM2.5 (by MiniVol samplers). The air sampling was done at 8 points on both sides of one street that had an average traffic flow, simultaneously meteorology data and vehicle flows were recorded. Roadside 24 h-PM2.5 levels were 97 +/- 31(53-151) mu g m(-3), higher on weekdays than weekends. Diurnal BTEX variation patterns were consistent with the diurnal flows of 6 vehicle categories moving on the street. BTEX levels were reduced with the increase in downwind distance from traffic lanes (approximately by 15% for each 5 m increment). Principal component analysis also confirmed the association between roadside pollution levels and traffic. A calculation algorithm was developed to remove the urban background, contributed by other sources than traffic in the selected street, from the roadside measured pollution levels. Urban background contributed a majority of PM2.5 (90-98%) and hourly BTEX (67-97%) with higher contributions at upwind side of the street and at late evening hours when less traffic was observed. CALINE inverse modeling produced explainable fleet hourly emission rates (g km(-1) h(-1)) and vehicle emission factors (EF, mg veh(-1) km(-1)). The obtained EF for gasoline and diesel vehicles were comparable with recent measurements made in Asian cities, as well as with calculated EFs for European and US urban fleets about 10-15 years ago. To minimize the collinearity problem encountered in inverse traffic modeling it is essential that the monitoring is done at different times to capture significant variations in the street traffic compositions. (C) 2014 Elsevier Ltd. All rights reserved.