An examination of the Pasquill-Gifford-Turner dispersion scheme

The Pasquill-Gifford-Turner (PGT) dispersion curves are used in most air quality models recommended for regulatory analysis of small industrial sources. In this paper, we first examine the origin of these dispersion curves. We then examine the theoretical foundation of these curves in the light of current understanding of dispersion, which we refer to as surface-layer similarity theory (SST). Numerical estimates of vertical spread from PGT are then compared with those from SST for conditions corresponding to the Prairie Grass experiment (Barad, 1958). We find that while the PGT dispersion estimates are broadly consistent with those from SST, they only represent a limited description of dispersion in the surface boundary layer. The use of the PGT dispersion scheme requires careful consideration of the reference height used for the wind speed that appears in the Gaussian dispersion equation. In fact, choosing a reference height (as PGT does) is not meaningful, because vertical spread is not independent of the transporting wind speed. The shortcomings of the PGT scheme can be readily corrected by adapting a version of a scheme based on surface-layer similarity theory. We show that the recommended method is just as easy to use as the PGT dispersion scheme, even within the Gaussian framework.