Geogenic radon potential through geostatistical analysis of uranium concentration

Dosimetric measurements are customarily conducted in dwellings to evaluate the radon hazard. The measurement sites are often unevenly distributed. This makes challenging direct data interpolation and their extrapolation to under-sampled areas, as well as the prediction of hazard. Geostatistical techniques, such as logistic regression, help address this issue because they allow for using proxy data to infer the probability of radon hazards where no direct measurements are available. The rock U content can be an appropriate proxy for indoor 222Rn concentration. Considering uranium concentrations in combination with other variables, such as bedrock nature and surface geology, has emerged as an effective method for producing reliable maps of Geogenic Radon Potential (GRP), a hazard indicator of radon generated by the radioactive decay of elements in rocks and soils and released into the air.  In this paper, we investigated the relationship between uranium and radon to map the radiological hazard linked to lithology also in unsampled areas. We used field gamma-ray spectrometry to determine the uranium concentration on the exposed bedrock and radon dosimetric records in indoor environments in direct contact with the ground. In addition to passive radon determinations, we measured the radon in soils by means of an active device. Logistic regression was used to examine the correlation between the concentration of uranium and the indoor radon measured in the same geological formation. This technique was tested in Western Liguria (Northern Italy), an area including a wide range of rocks spanning from sedimentary and metasedimentary to metavolcanic. The approach led to determining the probability of exceeding the threshold of 200 Bq/m3 for each lithology based on U concentration and defining a detailed picture of the investigated area's GRP. A background uranium content of 4 ppm implies a 50% probability of exceeding the safety threshold of indoor radon. Although the dataset of soil radon measurements so far collected is far from being representative, the results indicate that higher concentrations (up to 250 kBq/m3 as an upper bound) roughly correspond to indoor radon > 200 Bq/m3. In summary, our work highlights the relationship between indoor radon concentration and uranium content in rocks and reinforces the use of geological data to identify areas with a higher susceptibility to radon exposure.