Using outdoor radon and radon flux to predict areas with high radon risk

Radon is a radioactive noble gas built in the uranium – radium decay chain. Accumulated indoor radon concentrations can cause lung cancer. The reduction of indoor radon concentrations is a health political topic, noticing indoor radon is a large source of radiation exposure. The delineation of areas with high radon risk is an essential task to effective implement radon protection measures. The methods for delineation range from aggregate statistics of indoor radon concentrations to data driven machine-learning techniques with multiple predictors.

The main factors determining indoor radon concentrations are geogenic parameters (e.g. uranium content, permeability of the soil), building characteristics (e.g. sealing against the underground) and using habits (e.g. air exchange rate). Radon is not only a radiation protection topic. Outdoor radon and radon flux is used in atmospheric sciences as tracer for greenhouse gases and as input variable for atmospheric modelling.

We investigate the possibility using outdoor radon and radon flux to predict areas with high radon risk, by comparing these parameters with other parameters used for radon risk prediction as geological information, uranium content of the soil or weather data. We use the gridded indoor radon concentrations of the European Atlas of Natural Radiation as basis to define if an area shows high indoor radon concentrations. We perform a correlation analysis of the above-mentioned parameters. Further, we predict the gridded indoor radon concentrations with a random forest model and calculate feature importance for the selected model to investigate which parameters have the most impact on the prediction.

This research is part of project 19ENV01 traceRadon.  The project 19ENV01 traceRadon has received funding from the EMPIR programme co-financed by the Participating States and from the European Union's Horizon 2020 research and innovation programme.