Observação e modelagem estatística das concentrações de material particulado em paradas de ônibus urbanos

Abstract

Every day, people are exposed to pollutants from vehicle emissions, especially in transport microenvironments, such as bus stops and terminals. However, most studies on public transport traffic do not assess the concentrations of pollutants in these microenvironments. The behavior of fine particulate matter (with aerodynamic diameter less than 2.5 µm, PM2,5), black carbon (BC) and particle number (PN) was evaluated at bus stops in of Londrina (PR) city center during monitoring campaigns, conducted in 2015 and 2019. The 2015 experiment was performed with equipment mounted on bicycles that traveled the city’s core quantifying the particulate concentrations at 31 bus stops. 2019 experiment collected BC and PN data at five points, three of them selected from 2015 experiment plus two new ones. In the mobile monitoring, the mean PM2.5, BC and PN concentrations (+/- standard deviation) were 11,00 ±19,30 µg m-3,7,60 ±14,90 µg m-3 and 27.552 ±22.570 # cm-3 , respectively. In the fixed campaign, the mean BC and PN concentrations were 9,17 ±27,50 µg m-3 and 34.980 ±33.508 # cm-3 , respectively. The PM2,5, BC and PN concentrations at the bus stops were on average 0,30-2,90, 1,58-19,62 and 1,50-6,69 times higher than at an urban background site. At four stops from the 2019 campaign, three zones were evaluated: deceleration zone (located 10 m before the stop), boarding/alighting area and acceleration zone (located 10 m after the stop). The data allowed to pinpoint increases between 2,39- 12,05 µg m-3 and 11.380-23.220 # cm-3 in BC and PN, respectively, at the bus stops during two situations: when passengers were boarding and when buses accelerated at departure. The mean BC concentrations in the acceleration and stop zones were 17% higher than at the deceleration zone. However, the mean PN concentrations in the acceleration zone were 11% higher compared to the stop zone. Thus, it is not advisable to place several consecutive bus stops since the stop located immediately after the first one will be mainly affected by the previous accelerating bus emissions. Two forecast models for PM2.5 and BC concentrations were developed using multiple linear regression, including traffic, weather, background concentration, street, and bus stop characteristics as predictive variables. The PM2,5 and BC models had coefficient of determination (R2 ) of 0,36 ±0,09 and 0,28 ±0,08, respectively, and identified the following variables as the best predictors: atmospheric pressure, the concentration of BC with a wavelength of 880 nm at the background site and the height of the buildings: width of the street ratio. Likewise, the BC model identified the truck volume as a predictive variable. Finally, the PM2,5 model also identified that the directions of the southern sector (S, SE and SW) could favor an increase in PM2,5 concentrations at bus stops, possibly affected by the large traffic rate on avenues in those sectors.