Geographic Resolution Issues in Ram Transport Risk Analysis

Abstract

Transport risk analyses based on the RADTRAN code have been met with continual demands for increased spatial resolution of variations in population densities and other parameters employed in the calculation of risk estimates for transport of radioactive material (RAM). With the advent of geographic information systems (GISs) large quantities of data required to describe transport routes, which may extend to hundreds of kilometres, with high resolution (e.g. 1 km segments) can be handled without inordinate expense. This capability has raised a question concerning the maximum resolution of available input data and compatibility with RADTRAN computational models. Quantitative examinations are presented of spatial resolution issues in the calculation of incident-free doses and accident dose risks. For incident-free calculations, the effect of decreasing route-segment length on accuracy, in view of the model employed, is examined, and means of reducing total data input to the RADTRAN calculations, without loss of meaningful resolution of population concentrations, are presented. In the case of accident-risk calculations, the ability to detail population density under very large dispersal plumes permits comparison of plume modelling to actual data. In both types of calculations, meaningful limits to geographic extent are suggested.