Modeling spatial variability of airborne levoglucosan in Seattle, Washington

In many urban areas residential wood burning is a significant source of wintertime fine particles and has an important influence on spatial variability of particle concentrations. Although woodsmoke fine particles are usually within the size range thought to be most damaging to human health, their chemical composition is different from those derived from fossil fuel combustion, on which most health-effects studies have focused. Development of an exposure assessment tool for identification of the spatial distribution of woodsmoke will improve future epidemiological studies that rely on such intra-urban variability. For land-use regression (LUR) models, uniform buffers (i.e., circular areas or grids) are often applied to model spatial variability of pollutant concentrations. However, when winter woodsmoke levels are expected to be at a maximum, the surface wind is influenced by drainage flow and a given receptor location is systematically downwind of uphill sources. This research extends our previously developed GIS-based catchment air flow modeling approach of wintertime average woodsmoke levels to Seattle, WA, with emphasis on the use of levoglucosan as a marker of wood combustion. We further compare our regression model to a historical data set of mobile light-scattering measurements taken 15–20 years ago. Although fine particle levels have decreased significantly over this period, the spatial models for current levoglucosan (R2=0.57) and historical light scattering (R2=0.49) predict similar spatial patterns. This research demonstrates the usefulness of using both light scattering and levoglucosan to model ambient woodsmoke concentrations and further demonstrates the usefulness of the concept of drainage catchments to identify elevated, persistent nighttime levels of fine particles.