driven by the stochastic nature of turbulence. For point sources,

the root-mean-square (rms) concentration fluctuation

observed at some downwind location is typically as large as the mean

concentration. This statistical concentration variability is

important in a number of problems such as estimating the peak

concentration of toxic or hazardous materials and evaluating

dispersion models.

This paper focuses on the computation of the statistical variability

and dispersion model evaluation using concentration

fields generated by a Lagrangian particle model driven by velocity fields

from large-eddy simulations of the PBL. This coupled approach

was used previously to simulate the mean concentration fields from

sources in the convective boundary layer (CBL) (Weil et al., 2004) and

found to agree well with laboratory and field data. We

extended the approach to compute the statistical variability in the

concentration due to a point source in a highly convective boundary

layer. For a surface source, the computed variability was in good

agreement with results from two field experiments. In this paper, we

extend the calculations to sources in weaker convection and show how the

results can be applied to evaluate simpler or more routine dispersion

models for applications.