Peak Concentrations and Plume Diffusion in the Atmospheric Surface Layer

Abstract

To improve our understanding of pollutant behavior in the atmosphere, tracer experiments were conducted in 1997 at a field site near Galen, Montana. We developed an empirical method to estimate instantaneous plume spread as a function of standard wind statistics and travel time. Predicted diffusion coefficients are within a factor of 2, or better, of observed values for the 1997 data, and the average predicted-to-observed ratio was 1.03 for an expanded dataset of 97 samples representing a range of terrain types, meteorological conditions, and travel times. The 1997 field data are also used to test a meandering plume model, and modeled time series are similar to observed data including fluctuation statistics such as concentration mean, intensity, intermittency, and peak-to-mean ratio. Finally, we propose a simple approach to predict normalized peak concentrations for ground-level sources. The technique is shown to be realistic, yet conservative, with an average predicted-to-observed ratio of 1.90 for a variety of field studies.