SIGNAL DECOMPOSITION AND ANALYSIS FOR THE IDENTIFICATION OF PERIODIC AND ANOMALOUS PHENOMENA IN RADON TIME-SERIES.

This work concerns continuous monitoring of radon and thoron specific activity in soil gas within the framework of identifying possible anomalies. It is based on the analysis of a medium-term data record obtained from soil gas in an area of geophysical interest. The RaMonA spectrometric system is also used to measure the climatic parameters and a specific analysis of the alpha spectra is performed to better determine the alpha lines intensity. Since radon emission is also influenced by meteorological parameters, it is mandatory to differentiate the changes due to the deep phenomena. Different procedures are utilized to reach the above objective: statistical analysis using the Empirical Mode Decomposition technique, the Multiple Linear Regression method and the Remote Radon Estimation by using of the thoron trend to eliminate the locally produced radon fraction. The results of such methods are compared to recognize and to highlight radon anomalies.