Simulation of the influence of residential biomass burning on air quality in an urban area

Abstract

Residential biomass burning is often one of the dominant local sources of air pollution in urban areas during the winter months. However, the corresponding particle emissions are quite uncertain and chemical transport models (CTMs) often have difficulties reproducing the observed winter particulate matter (PM) concentrations. In this work, we combine measurements from a low-cost PM sensor network and the CTM PMCAMx (Particulate Matter Comprehensive Air quality Modeling with extensions) to estimate the spatial and temporal distribution of biomass burning emissions in an urban area.An estimate of 8100 kg d−1 (40 g d−1 per person) PM2.5 emissions from residential biomass burning was calculated for the city of Patras including organic compounds with effective saturation concentrations C* up to 104 μg m−3. The spatial distribution of the emissions was based on the density of fireplaces in the city. The temporal distribution of emissions was based on measurements of the biomass burning organic aerosol (bbOA) and are higher from 18:00 to 22:00 LT peaking at 21:00 LT. The nighttime (18:00-22:00 LT) bbOA emissions were 4 times higher than the corresponding morning (8:00-13:00 LT) ones. Estimated biomass burning emissions vary from day to day based on ambient temperature, with higher emissions during the colder days. The Volatility Basis Set (VBS) was used to simulate bbOA that is treated as semi-volatile and chemically reactive.PMCAMx predicts that bbOA concentration reaches on average 15 μg m−3 in the high-density bbOA emission area during the simulated period, while the corresponding peak hourly concentrations are higher than 40 μg m−3 most of the nights. The average predicted bbOA concentration in the city center is 3–6 μg m−3 and is lower than 1 μg m−3 at the suburbs. The model predicts that bbOA concentrations peak at 9:00 LT in the morning and at 21:00 LT during the nighttime, reproducing the bbOA measurements.