Abstract
Eulerian observations of wind direction fluctuations in slightly unstable conditions are analysed in terms of Taylor's statistical theory of dispersion which, with the HayβPasquill assumption of similar shapes for Lagrangian and Eulerian spectra, relates the standard deviation of the crosswind material distribution, Ο~y~, to the standard deviation of fluctuations in wind direction, Ο~ΞΈ~. The dependence of the predicted crossβwind dispersion parameter, S=Ο~y~/(Ο~ΞΈ~x), where Γ is the distance downwind from the source, on travel time, T, is examined, and it is concluded that the data are well represented by an empirical form originally proposed by Draxler to fit observed crosswind dispersion.
In contrast to direct dispersion observations, the present data show little scatter and unambiguously demonstrate that the common assumption of an exponentially decaying correlogram results in a less accurate representation. In particular, the exponential form approaches the largeβtime limit, S^2^ Ξ± 1/T, too rapidly, and thus may be expected to perform poorly in extrapolating from a limited range of data.
The consistency of the Draxler form in representing both the Eulerian measurements presented here, and direct Lagrangian dispersion observations, serves to support not only the form itself, but also, indirectly, the HayβPasquill hypothesis.
Evidence of a correlation between the intensity and scale of turbulence is also presented, which, if confirmed over a larger range of atmospheric stability conditions than considered here, will be very useful for estimating dispersion from wind statistics.