Detecting local and regional air pollution from biomass burning at a suburban site

Identifying pollution plumes is challenging in cities where in situ air pollution data are scarce, monitoring stations are placed close to roadways, and the smoke from biomass burning might be of local and regional origin. This work characterizes local and regional smoke events at a suburban site in Londrina, southern Brazil, during the dry season. We measured black carbon at 370 and 800 nm (BC370 and BC880), fine particulate matter (PM2.5), particle number (PN), nitrogen oxides (NOx) and ozone (O3) concentrations with 1-2 min resolutions. Ancillary data, such as meteorological variables, aerosol optical depth (AOD) and air mass backward trajectories, aided identifying the sources. The pollution datasets were analyzed by using descriptive statistics, diurnal cycles, temporal correlations, and conditional probability functions coupled with cluster analysis of backward trajectories. Hourly mean (+/- standard deviation) concentrations were 1.33 +/- 1.28 lig m-3 (BC370), 0.91 +/- 0.84 lig m-3 (BC880), 15.90 +/- 12.95 lig m-3 (PM2.5), 5691 +/- 4872 cm-3 (PN), 7.23 +/- 7.53 lig m-3 (NOx) and 49.77 +/- 17.66 lig m-3 (O3). The prevalence of aerosols from biomass burning was indicated by Angstro center dot m exponent (A370/880) values predominantly larger than 1 (1.51 +/- 0.20). Pollutant concentrations peaked on weekend evenings and, occasionally, scattered along the day due to the short-lived open burning of residential solid waste. These events led to fast changing concentrations, increasing from 282 to 929% between consecutive 2-min measurements, reaching 69.29 lig m-3 (BC370), 456.77 lig m-3 (PM2.5) and 232,500 cm-3 (PN). Emissions from biomass-fueled boilers also contributed to deteriorate the local air quality, under favorable local winds. Three regional events advected pollution from Cerrado (tropical savanna) wildfires lasting two-three days each, and were associated with the South American Low Level Jet. During these event, the AOD reached up to 0.37 and mean PM2.5, BC370, and O3 concentrations were (2.5-3.3), (1.9-2.9) and (1.3-1.5) times higher compared to a clean period. Lowering the population exposure to biomass burning requires immediate action to control local sources, and international cooperation is needed to combat regional wildfires in a warming climate.