Natural radioactivity mapping via gamma-ray spectroscopy: integrating different techniques and multivariate information

Among naturally occurring terrestrial radioisotopes, K and some radioisotopes in the U and Th decay chains emit γ-rays having energy of the order of MeV and can be easily detected. Gamma-ray surveys performed in laboratory, in field setting and from aircraft allow for monitoring the environmental radioactivity as well as prove valuable insights on geological mapping, structural geology and soil surveying. Natural radioactivity mapping provides the opportunity to integrate γ-ray spectroscopy data acquired with different measurement methods and field of views of the investigated area. In this framework the elaboration of an unique cartographic product with an appropriate descriptive legend cannot be pursued without critically dealing with some delicate issues. We dealt the problematic aspects regarding the statistical distributions of analyzed datasets, the heterogeneous experimental uncertainties and the spatial resolution of the measurements. Adopting a rigorous study of the data distributions and appropriated geostatistical interpolators, we integrated the results coming from the statistical treatment of the rock analysis and the spatial interpolation of the airborne measurements to which are associated distinct frequency distributions. Taking on the challenge to treat heterogeneous input uncertainties data in the spatial interpolation, we took into account the degree of the confidence associated to different gamma-ray techniques during the estimation process. Multivariate spatial estimations could enhance the estimation of radioelement distribution taking advantage of the correlation existing between the under-sampled γ-ray measurements and the continuous distributions of geological formations. Adopting this approach on the basis of a sophisticated geostatistical interpolator, we obtained radiometric maps tackling the complex aspects related to the performing of spatial interpolation of quantitative data using the qualitative information of the geological map as constraint. The described methods were validated through several surveys that cover approximately 50000 km^2 of the Italian territory: specific cases of studies will be presented and discussed