The predictive power of airborne gamma ray survey data on the locations of domestic radon hazards in Norway: A strong case for utilizing airborne data in large-scale radon potential mapping

Abstract

It is estimated that exposure to radon in Norwegian dwellings is responsible for as many as 300 deaths a year due to lung cancer. To address this, the authorities in Norway have developed a national action plan that has the aim of reducing exposure to radon in Norway (Norwegian Ministries, 2010). The plan includes further investigation of the relationship between radon hazard and geological conditions, and development of map-based tools for assessing the large spatial variation in radon hazard levels across Norway. The main focus of the present contribution is to describe how we generate map predictions of radon potential (RP), a measure of radon hazard, from available airborne gamma ray spectrometry (AGRS) surveys in Norway, and what impact these map predictions can be expected to have on radon protection work including land-use planning and targeted surveying. We have compiled 11 contiguous AGRS surveys centred on the most populated part of Norway around Oslo to produce an equivalent uranium map measuring 180 km × 102 km that represents the relative concentrations of radon in the near surface of the ground with a spatial resolution in the 100 s of metres. We find that this map of radon in the ground offers a far more detailed and reliable picture of the distribution of radon in the sub-surface than can be deduced from the available digital geology maps. We tested the performances of digital geology and AGRS data as predictors of RP. We find that digital geology explains approximately 40% of the observed variance in ln RP nationally, while the AGRS data in the Oslo area split into 14 bands explains approximately 70% of the variance in the same parameter. We also notice that there are too few indoor data to characterise all geological settings in Norway which leaves areas in the geology-based RP map in the Oslo area, and elsewhere, unclassified. The AGRS RP map is derived from fewer classes, all characterised by more than 30 indoor measurements, and the corresponding RP map of the Oslo area has no unclassified parts. We used statistics of proportions to add 95% confidence limits to estimates of RP on our predictive maps, offering public health strategists an objective measure of uncertainty in the model. The geological and AGRS RP maps were further compared in terms of their performances in correctly classifying local areas known to be radon affected and less affected. Both maps were accurate in their predictions; however the AGRS map out-performed the geology map in its ability to offer confident predictions of RP for all of the local areas tested. We compared the AGRS RP map with the 2015 distribution of population in the Oslo area to determine the likely impact of radon contamination on the population. 11.4% of the population currently reside in the area classified as radon affected. 34% of ground floor living spaces in this affected area are expected to exceed the maximum limit of 200 Bq/m, while 8.4% of similar spaces outside the affected area exceed this same limit, indicating that the map is very efficient...

Detailed Geogenic Radon Potential Mapping Using Geospatial Analysis of Multiple Geo-Variables—A Case Study from a High-Risk Area in SE Ireland

“…Ideally, a GRP estimation is derived from radon soil gas concentrations [ 22 ]. An alternative method to detect radon on the ground is to use airborne gamma spectrometry [ 18 , 23 ]. In Ireland, there are insufficient SGR data to generate useful radon potential maps.…”

Estimation of radon potential through measurement of uranium concentrations in granite geology

“…Radon potential in the study area was estimated by linking the indoor radon concentrations to the measured uranium concentrations in the soil on Baviaansberg. Smethurst et al 11 noted that the relationship between indoor radon and uranium concentrations depends on the characteristic of the geographical locations. The Tusham Ring Complex and the Himachal Pradesh region of India were consequently selected to determine these relationships because of the geographical similarity of these regions to the study area.…”

“…The initial relationship that was extracted from the empirical data was therefore used to relate the indoor radon concentrations and the uranium concentrations of the study area. This relationship was employed to estimate indoor radon potential, similarly to the method used by Smethurst et al 11 The uranium concentrations and indoor radon potentials were interpolated and mapped using ArcGIS and QGIS software. The results for the study area were compared to indoor radon measurements taken in Paarl.…”

Towards a new radon hazard map of Norway using Trøndelag county as a test area

“…The thorium content is of much less importance. Heincke et al (2008) and Smethurst et al (2008Smethurst et al ( , 2017 found simple, linear relationships between the uranium concentration in the ground from airborne surveys and the proportion of dwellings with average annual radon concentrations above the action level of 200 Bq/m 3 and concluded that airborne surveying in the Oslo Fjord region provides important information on radon hazard. The study by Smethurst et al (2008) consider airborne gamma-ray surveys to be better than mapped geology at accounting for and predicting differences in radon potential across the Oslo Fjord area.…”