A geostatistical autopsy of the Austrian indoor radon survey (1992–2002)

Indoor radon (Rn) concentrations have been measured intensively in many countries to assess the burden of diseases associated with exposure to this radioactive gas. So-called radon-risk maps have consequently been produced to delineate areas with high levels. Geostatistical techniques are commonly used nowadays to map a range of environmental variables, in particular to generate probability maps of exceeding a given threshold. However, very few case studies in which indoor radon measurements have been investigated using geostatistical techniques have been published so far. By analyzing around 12,000 Rn measurements made in Austrian ground floors during a 10-year survey, we aim here to review and discuss the potential of geostatistics for mapping an environmental variable that shows very strong local variability. In particular, we show how kriging of the scale components can shed new light on various factors that affect the very high spatial variability of the variable.