An Application of Geostatistical and GIS Techniques to Indoor Radon Risk Mapping

Abstract

Radon is a naturally occurring radioactive gas that can cause lung cancer in cases of high-level long-term exposure. An indoor radon data set of long-term measurements in about 5,200 single-family houses has been collected from 1995 to 1999 in southern Belgium where a high radon potential exists. This national survey is to develop cost-effective approaches for mitigation and to select priorities for the government on the radon problem. Radon concentrations were found to be strongly related to geological context in previous radon surveys. The study region (southern Belgium) is divided into three geologically homogeneous zones within which stationarity may be assumed for geostatistical analysis. The radon data are logormally distributed. Logarithmic transforms of radon concentrations were used to perform the geostatistical analysis. A map of radon concentrations was produced using the lognormal kriging technique. A GIS is used to digitise, process and integrate a variety of data, including geological maps, radon maps and measured radon values associated with house locations, an administrative map and road networks etc. It also allows evaluation of the relationships between various spatial data sets with the goal of producing radon risk maps. The results show that high and medium radon risk areas are associated with bedrocks from the Cambrian to Lower Devonian. High radon concentrations are also related to strongly folded and fractured context. Filling materials of the fractures containing radioactive minerals are a main radon source in western and central Ardenne, while opened faults or fractures play a principal role in high radon concentrations in eastern Ardenne.