Simulation of large particle transport near the surface under stable conditions: comparison with the Hanford tracer experiments

A plume model is presented describing the downwind transport of large parti cles (1-100 mum) under stable conditions. The model includes both vertical variations in wind speed and turbulence intensity as well as an algorithm f or particle deposition at the surface. Model predictions compare favorably with the Hanford single and dual tracer experiments of crosswind integrated concentration (for particles: relative bias = - 0.02 and 0.16, normalized mean square error = 0.61 and 0.14, for the single and dual tracer experimen ts, respectively), whereas the US EPA's fugitive dust model consistently ov erestimates the observed concentrations at downwind distances beyond severa l hundred meters (for particles: relative bias = 0.31 and 2.26, mean square error = 0.42 and 1.71, respectively). For either plume model, the measured ratio of particle to gas concentration is consistently overestimated when using the deposition velocity algorithm of Sehmel and Hodgson (1978. DOE Re port PNL-SA-6721, Pacific Northwest Laboratories, Richland, WA). In contras t, these same ratios are predicted with relatively little bias when using t he algorithm of Kim et al. (2000. Atmospheric Environment 34 (15), 2387-239 7). (C) 2001 Elsevier Science Ltd. All rights reserved.