Empirical and modeling studies on the physical and radiative properties of windblown dust over the Columbia Plateau.

摘要

The research presented here is part of a multidisciplinary wind erosion program, Columbia Plateau PM10 Project (CP3), aimed at evaluating the effects of windblown dust on air quality through intensive field campaigns and application of numerical modeling of emission, transport and dispersion of windblown dust. In the first section, wind tunnel experiments were conducted to evaluate the collection efficiencies of several dust samplers commonly deployed in field campaigns. In dust storm conditions (wind speeds of 7 to 15 m s-1, dust mass loading of 0.42 to 1.11 g s-1), collection efficiencies ranged from 40 to 100%. Next, the emission, transport and dispersion of PM10 (particulate matter with an aerodynamic diameter less than 10 mum) were numerically simulated for six regional dust storm events. The modeling system included the prognostic meteorological model, Mesoscale Metorological Model Version 5 (MM5), the CALMET/CALGRID Eulerian modeling pair and a dust emission module (EMIT-PM). Improved parameterizations in EMIT-PM resulted in good agreement between observed and predicted 24-hour average PM10 concentrations (ratios between 0.5 to 6.0). The model performance improved with increasing magnitude of the dust storm. Sensitivity tests of the modeling system were also conducted on PM10 emissions and concentrations by varying emission variables for the different land use categories and soil classes in eastern Washington. Predicted emissions were significantly influenced by soil crusting, random soil roughness and vegetative surface. A 40% reduction in the coverage of dry fallow lands resulted in a factor of 2 to 3 reduction in downwind PM10 concentrations compared to a 7 to 9% reduction when surface cover was increased by 50%. Finally, aerosol optical depths (AOD) were predicted for two dust storm events and values ranged from 0.3 to 0.7, comparable to values in published literature. Sensitivity analyses indicated the magnitude of error for the modeled AOD values ranged from +/-10 to +/-30%. Aerosol optical depths may prove to be much more useful for model validation than point concentrations in receptor areas, and they also provide information on the role of mineral dusts in regional and global radiation balances.