*Article* **A Real-Time Comparison of Four Particulate Matter Size Fractions in the Personal Breathing Zone of Paris Subway Workers: A Six-Week Prospective Study**

**Rémy Pétremand <sup>1</sup> , Guillaume Suárez <sup>1</sup> , Sophie Besançon <sup>2</sup> , J. Hugo Dil <sup>3</sup> and Irina Guseva Canu 1,\***


**Abstract:** We developed a Bayesian spline model for real-time mass concentrations of particulate matter (PM10, PM2.5, PM1, and PM0.3) measured simultaneously in the personal breathing zone of Parisian subway workers. The measurements were performed by GRIMM, a gravimetric method, and DiSCmini during the workers' work shifts over two consecutive weeks. The measured PM concentrations were analyzed with respect to the working environment, the underground station, and any specific events that occurred during the work shift. Overall, PM0.3 concentrations were more than an order of magnitude lower compared to the other PM concentrations and showed the highest temporal variation. The PM2.5 levels raised the highest exposure concern: 15 stations out of 37 had higher mass concentrations compared to the reference. Station PM levels were not correlated with the annual number of passengers entering the station, the year of station opening or renovation, or the number of platforms and tracks. The correlation with the number of station entrances was consistently negative for all PM sizes, whereas the number of correspondence concourses was negatively correlated with PM0.3 and PM10 levels and positively correlated with PM1 and PM2.5 levels. The highest PM10 exposure was observed for the station platform, followed by the subway cabin and train, while ticket counters had the highest PM0.3, PM1, and PM2.5 mass concentrations. We further found that compared to gravimetric and DiSCmini measurements, GRIMM results showed some discrepancies, with an underestimation of exposure levels. Therefore, we suggest using GRIMM, calibrated by gravimetric methods, for PM sizes above 1µm, and DiSCmini for sizes below 700 nm.

**Keywords:** occupational exposure; Bayesian spline model; time-series; public transport; particulate matter; inhalation
