Source apportionment in PM2.5 in central Japan using positive matrix factorization focusing on small-scale local biomass burning

Abstract

Here, we discuss the influences of small-scale biomass burning (BB) originating from local/regional emissions, such as agricultural waste burning, which are difficult to detect via remote sensing. To evaluate the BB contribution, we used a positive matrix factorization (PMF) model with chemical component data including levoglucosan, a useful BB indicator, at multiple urban, suburban, and background sites in central Japan during two weeks in October and November of 2014 and January and February of 2015. High levoglucosan concentrations were observed not only at the suburban sites but also at the urban sites. At the background site located downwind of Northeast China, the levoglucosan concentrations were elevated when air masses were transported from Northeast China, where fire spots were detected. Meanwhile, at the other sites, small-scale local or regional BB emissions prevailed, and fire spots were rarely detected. At these sites, the levoglucosan levels were high although the relevant air masses did not originate in Northeast China. The source apportionment results demonstrated that in fall, BB was a significant source of particulate matter with diameters of less than 2.5 μm (PM2.5), organic carbon, and elemental carbon at all sites. Notably, the BB contributions to PM2.5 were higher at an urban site located near farmland than at other urban sites in fall and winter. Our results indicate that specific BB emission sources, such as local, small, and irregular open field burning, impact the air quality not only at suburban areas but also at urban areas in central Japan.