INTRODUCTION--Mobile monitoring can effectively capture spatial patterns in on-road exposure. Here, we report measured on-road concentrations of black carbon and ultrafine particles in urban and peri-urban Bangalore, India.METHODS--Our mobile platform, a CNG car, was equipped with an aethalometer for black carbon (BC) and a condensation particle counter for ultrafine particles (UFP). We sampled on roadways in four parts of Bangalore: a central business district, a residential urban neighborhood, a peri-urban neighborhood, and along an urban-rural transect; in total, we collected ~400 h (~5000 km) of on-road data. Our study design involved oversampling each road segment (n=25 repeated samples per road). We then quantify, via subsequent subsampling, the data requirements for robustly estimating time-integrated concentrations. We computed average concentration for each 30m road segment and computed mean and median concentrations for each region.RESULTS--Mean on-road concentrations were highest in the central business district (BC: 50 µg/m3; UFP: 145,000 cm-3), intermediate in the urban residential neighborhood (32 µg/m3; 87,000 cm-3), and lowest in the peri-urban neighborhood (29 µg/m3; 30,000 cm-3). Median values are 2% - 38% less than mean. The urban-rural transect reveals analogous spatial patterns: average concentrations decline when going from downtown to the peri-urban area; spatial gradients are larger for UFP than BC. Subsampling analyses reveal how the robustness of estimated mean concentrations varies with number of observations and the timing of those observations, and how that relationship varies in space.CONCLUSION--Average on-road concentrations decline from downtown to the peri-urban area; spatial gradients are larger for UFP than BC. Our subsampling analyses inform how many visits are needed to obtain robust estimates of long-term average concentrations, as well as how that result varies in space and time.
Read full abstract