Generally, two different sampling strategies were used, depending on whether smoke plumes or regional haze were being sampled. For smoke plumes the primary objective was the determination of emission factors by the carbon balance method for a variety of species. This requires measurements of the five major carbon species in the smoke: CO2, CO and hydrocarbons (in the gas phase), and organic and elemental carbon in aerosols. For a "grab" bag sample collected in smoke, a standard set of three filters was exposed: a quartz filter for elemental and organic carbon; a Teflon filter for mass, elemental carbon and ionic species; and a Nuclepore filter for mass, elemental carbon and PIXE analysis. CO, CO2, NOx and SO2 were also measured in the "grab" bag by switching the continuous monitors to the bag for 60 seconds after each bag fill. This fixes the ratio of all species measured from the bag to the excess (i.e., above ambient) concentration of CO2, which is the main tracer of combustion emissions. Hydrocarbon species were measured by filling the canisters from the manifold concurrent with opening the "grab" bag for a smoke sample. To fill a canister completely usually required two passes through the smoke, in which case the air sampled by the bag is not exactly the same as the air captured in the canister. To make comparison possible, CO2 was measured in the canisters so that the ratio of hydrocarbons to excess CO2 can be determined. We then compute the apparent hydrocarbon concentration for each "grab" bag sample by scaling the hydrocarbon concentration to the excess CO2 measured in the bag. This method was chosen in order to take advantage of the cleanliness of the manifold and to avoid any contamination from the "velostat" material of the bag. In SCAR-B, 68 canister samples were collected for hydrocarbon measurements. Twenty-three Teflon filters were exposed simultaneously with canister samples for the specific purpose of measuring emission factors from a variety of individual fires.
A total of 62 Teflon filters were also exposed in regional haze; since aerosol concentrations were generally much lower than in smoke plumes multiple bag samples were generally required to get enough loading on the filters for mass and chemical analysis. Also, since the regional smoke was generally quite uniform, fewer simultaneous hydrocarbon canisters (20 in all) were collected with the Teflon filters.
In addition to the filters collected from bag samples, a set of three samples were collected on each flight from the continuous inlet manifold and integrated for the entire flight. These consisted of a 47 mm Teflon filter, a MOUDI impactor, and a gold tube for mercury. These samples (which will be analyzed for elemental composition) are considered representative of the general mix of old and new smoke.
Forty Teflon field "blanks" were analyzed for aerosol mass and ionic species to determine the uncertainties associated with filter handling (prior to sample exposure) in the field.