Mobile noise measurements as a proxy for BC exposure: spatiotemporal and spatial analysis

Abstract

Background Personal exposure to traffic related particulate matter is strongly related to the person’s activity pattern and in-traffic exposure. Measuring the personal exposure is invasive for the subjects and expensive in both equipment and support costs. The spatial and temporal variability of Black carbon exposure is much stronger than for PM10 and PM2.5 and shows similar spatial effects as the mobile noise measurements. Aims The aim is to enhance diurnal exposure prediction by using mobile noise measurements and the derived local traffic dynamics as a proxy for in-traffic Black Carbon (BC) exposure. Mobile noise measurements contain interesting data on traffic characteristics and are cheaper to perform and require less support. Methods We first constructed a spatiotemporal model for bicyclists including instantaneous wind speed and traffic dynamics. Next a spatial model, aggregating mobile measurements over a one year period is build to investigate the effects of local traffic dynamics, local features of bicycle facilities and the long term meteorological corrections. Results A strong correlation has been established between the mobile noise exposure and the BC exposure of bicyclists. The bicyclists’ spatiotemporal prediction model reaches an average trip exposure correlation of 0.86. In the long term spatial model the meteorological effects fade out when sufficient measurements are available. The mobile noise measurement increases the quality of the prediction of the spatial variation of BC exposure significantly compared to models that use only traffic data as primary input. Conclusions Mobile noise measurements can be used as a proxy for BC to map cities at an unprecedented spatial resolution, including the actual local traffic dynamics effects. After correcting the BC exposure for background concentrations, the mobile noise measurements are a valid proxy for personal BC exposure.