Time Dependence of Aerosol Light Scattering Downwind of Forest Fires

In the first phase of BBOP (Biomass Burn Observation Project), a Department of Energy (DOE) sponsored study, wildland fires in the Pacific Northwest were sampled from the G-1 aircraft via sequences of transects that encountered emission whose age (time since emission) ranged from approximately 15 minutes to four hours. Comparisons between transects allowed us to determine the near-field time evolution of trace gases, aerosol particles, and optical properties. The fractional increase in aerosol concentration with plume age was typically less than a third of the fractional increase in light scattering. In some fires the increase in light scattering exceeded a factor of two. Two possible causes for the discrepancy between scattering and aerosol mass are i) the downwind formation of refractory tar balls that are not detected by the AMS and therefore contribute to scattering but not to aerosol mass and ii) changes to the aerosol size distribution. Both possibilities are considered. Our information on tar balls comes from an analysis of TEM grids. A direct determination of size changes is complicated by extremely high aerosol number concentrations that caused coincidence problems for the PCASP and UHSAS probes. We instead construct a set of plausible log normal size distributions and for each member of the set do Mie calculations to determine mass scattering efficiency (MSE), angstrom exponents, and backscatter ratios. Best fit size distributions are selected by comparison with observed data derived from multi-wavelength scattering measurements, an extrapolated FIMS size distribution, and mass measurements from an SP-AMS. MSE at 550 nm varies from a typical near source value of 2-3 to about 4 in aged air.