Spatio-temporal distribution of particulate matter and wood-smoke tracers in Temuco, Chile: a city heavily impacted by residential wood-burning

Temuco, like most Southern Chile cities, has a long-standing air pollution problem due to residential wood heating. Here, we study the temporal and spatial distribution of PM_2.5 and the biomass-burning tracers levoglucosan and water soluble potassium (K_sol). PM_2.5 samples were collected at 40 sites throughout the city and sampled four times in four different sessions between 2017 and 2018. Large concentrations were observed for all species, particularly in winter. Signs of spatial and temporal correlation were observed from multilevel models for the different concentrations and ratios. Ratios of levoglucosan and K_sol to PM_2.5 were lower in summer. Spatial clustering was observed both from spatial autocorrelation tests and kriging, with clusters of higher ratios for of all three species (comparing rotating to central site) observed in some neighborhoods. Clusters correlated with lower socioeconomic status (SES), which may reflect population density and lower quality stoves. Clusters of higher K_sol to PM_2.5 ratios in some neighborhoods were moderately correlated with higher SES, which may reflect the use of better-quality stoves (e.g., pellets). The current work provides a case study and example of the investigation of wood burning as an air pollution source in a city where many inhabitants use wood as a fuel for heating and cooking. It reveals key temporal and spatial patterns which can assist the design of mitigation strategies.