The limitations of atmospheric dispersion data and their contribution to uncertainties in dose assessment

The calculation of atmospheric dispersion patterns is often an important component of radiation dose estimates. These dispersion calculations are a possible source of error and such errors or uncertainties need to be quantified. An important source of uncertainty is the meteorological data used in the calculations. Such data may be less than ideal because of constraints imposed by both availability and by the variances associated with population from which the data are obtained. We have studied a simple and much used model of atmospheric dispersion--the Gaussian plume. We discuss the uncertainties on the meteorological data which are input to the model and how these uncertainties could be used to estimate uncertainties for the modeling results. In doing this we have addressed both the uncertainty associated with a recorded climatology and the added uncertainty arising from the year-to-year variability at any given location.