The Geographical Pattern of Local Statistical Dispersion of Environmental Radon in Europe

Abstract

Acknowledged as a significant health hazard, increasing attention has been given to indoor and geogenic radon for some 20 years. One part of the efforts is surveying in order to assess the geographical extent of the hazard. Results acquired in surveys serve to support decisions in radon abatement policy, aimed to reduce exposure and consisting in prevention, mitigation and remediation measures. One particular element is delineation of so-called radon priority areas, or areas in which the hazard is such that abatement measures should be implemented with priority. These areas are estimated from radon measurements supported by different modelling methods. Methods which are based on estimating local probabilities that radon concentration exceeds a reference level often rely on measures of local variability, expressed for example by the geometric standard deviation or the coefficient of variation, because these describe the shape of the distribution whose tail areas are the sought-after probabilities. Evidently, delineation of radon priority areas thus depends, apart from mean concentration, on dispersion within the area whose priority status shall be assessed. A second use of spatial variability measures of radon is survey planning, because the sample size necessary to estimate a mean with given precision depends on the dispersion of the quantity to be assessed. It is estimated through pilot surveys or derived from general knowledge. The large radon databases accumulated for years allow more detailed insight into spatial properties of dispersion, some of which are discussed in this paper, in the first place the relation between local dispersion and mean and sampling density. Not least, they also grant insight—so far mostly speculative—about the process, understood as a stochastic process, which generates the spatial dynamic. Main conclusions are that a proportional effect in logarithmic scale exists and has an effect on estimation of local exceedance probability and that one should keep in mind that results, for example the local status as a radon priority area, depend on the estimation method.