Analysis of geophysical and meteorological parameters influencing 222Rn activity concentration in Mladeč caves (Czech Republic) and in soils of Phlegrean Fields caldera (Italy)

The present work concerns a detailed analysis of Radon time series to differentiate endogenous from exogenous phenomena which provide anomalous signals. Two-year data from two sites in Czech Republic and in Italy are analyzed in order to contribute to the prevention of natural hazards. A new hybrid forecasting method is implemented and tuned for the identification of Radon anomalies in the time series. It is based on the combination of Multiple Linear Regressions, Empirical Mode Decomposition and Support Vector Regression methods that decompose the signal and analyze the components to distinguish the variations due to Radon originated in depths from those due to environmental parameters. The possible correlations with fumarolic tremors in the Italian site and the faults microdisplacements in the Czech site have been studied, as well as with the earthquakes that have influence on two studied areas. Results show that: (i) the used method is very effective considering the calculated statistical uncertainties; (ii) the outer temperature is the main influencing Radon driving force; (iii) the extracted Radon anomalies due to endogenous phenomena are well correlated with fault displacements, fumarolic tremors, and with earthquakes under a characteristic delay time for each area; (iv) significant correlations among earthquake magnitude and depth with fault displacement and fumarolic tremor are found.