Short-term associations between traffic-related noise, particle number and traffic flow in three European cities

- 20-min measurements of air pollution, noise and road traffic were taken at 141 sites.
- Traffic noise levels and traffic counts were far more constant over time than ultrafine particles number concentrations.
- Simultaneous measurements of traffic count and noise were moderately to well correlated.
- Simultaneous measurements of ultrafine particles and noise were poorly correlated.
- This should allow future studies to disentangle the short-term effects of ultrafine particles and noise.

Outdoor noise and particulate matter concentration share common sources, including road traffic in urban areas, raising the potential for mutual confounding in epidemiological studies of their health effects. While some studies evaluated their long-term correlation, little is known about their short-term correlation. Our aim was to study the correlation of short-term noise, ultrafine (<0.1 μm) particulate matter number concentration (UFP), and traffic flow in urban areas. A secondary aim was to document the temporal variability of these short-term measurements. We simultaneously measured traffic noise levels, UFP concentrations as well as motor vehicles' flows for 20 min in 141 locations, on one to three occasions, in three middle size European cities (Basel, Girona, Grenoble). The reproducibility of the short-term noise measurements and traffic counts over time was high, as reported by the intraclass correlation coefficient (ICC), which quantified the agreement between repeated measurements (ICC = 0.86-0.97, according to city, for noise and ICC = 0.93-0.94 for traffic counts); this was not the case for UFP number concentrations (ICC = −0.11 to 0.14). The Pearson correlations of simultaneous 20-min measurements of UFP number concentrations and noise levels were in the 0.43-0.55 range, depending on the city; correlations between noise levels and vehicle counts varied from 0.54 to 0.72; and correlations between UFP concentrations and vehicle counts were lower (r = 0.15-0.37 depending on the city). Measurements during as little time as 20 min of outdoor noise and traffic, but not of UFP, were strongly reproducible over durations of a couple of days or months in middle-size European cities. In these areas, on the short-term, noise levels and UFP concentrations exhibited relatively moderate correlations, which may allow adjustment for mutual confounding in epidemiological studies, thus allowing to disentangle their possible short-term health effects.