| The Spokane PM database currently contains one of the largest continuous PM2.5 daily mass databases available. Coverage includes, but is not limited to: total carbon, elemental and organic fraction carbon, particulate ions, and trace element information, from 1995--2002, taken from an area of Eastern Washington subjected to agricultural, industrial, urban, and rural pollution. Daily PM2.5 mass samples were analyzed with several measurement methods: Kevex and Jordan Valley EDX-771 energy dispersive spectrometers, Instrumental Neutron Activation Analysis, for trace element species; total carbon by thermal manganese oxidation; total carbon, elemental and organic fractions by thermal/optical transmittance; sulfate and nitrate by ion chromatography; and ammonium by colorimetry. The data were modeled to identify pollution sources with the new Positive Matrix Factorization (PMF) model, version 1.1, released by the United States Environmental Protection Agency.; The PMF models revealed seven source features, similar to profiles identified in earlier work using various sets of the PM2.5 data. PMF analyses using the XRF data yielded airborne soil, nitrate, biomass burning, chlorine-rich, metal processing, and vehicle combustion features. INAA data yielded the seventh feature, a Cr-rich feature, which was associated with elemental carbon. Bootstrapping and t test methods confirm the features' mass contributions to the overall ambient PM2.5 concentrations. Airborne soil was associated with late summer dust events, with little to no carbon. Metal processing contains organic carbon, whereas the Cl-rich source contained elemental and light organic carbon. The Cl-rich source determined using Jordan Valley data is confused with the vehicle exhaust and biomass burning sources, due in part to the lack of a calibration standard for the instrument. These source features were tied to their specific industrial sources, which can vary sporadically throughout the study period. Nitrate is associated with domestic burning and increases in winter, with organic carbon components. Biomass burning occurred in late fall or early spring, when fields were burned for future growing seasons, and shares similar feature species with vehicle exhaust, which was present consistently throughout the entire study period. Both biomass burning and vehicle exhaust contain elemental and organic carbon; the exhaust exhibited heavier organic compounds, and biomass burning, mostly lighter compounds. |
