Singular spectrum and principal component analysis of soil radon (Rn-222) emanation for better detection and correlation of seismic induced anomalies

Abstract. In the recent years there are several reporting’s of anomalous seismic induced temporal changes in soil radon emanation. It is however well known that radon anomalies apart from seismic activity are also governed and controlled by meteorological parameters. This is the major complication which arise for isolating the seismic induced precursory signals. Here in the investigation the soil radon emanations temporal variability at MPGO, Tezpur, is scrutinized in the lime light of singular spectrum analysis (SSA). Further prior applying SSA Digital filter (Butterworth low pass) is applied to remove the high frequency quasi periodic component in the time series of soil radon emanation. It was scrutinized that sum of just 9 eigenfunctions were sufficient enough for reproducing the prominent characteristics of the overall variation. This perhaps also evinces that more significantly produced fluctuations are mostly free from natural variations. The variations in soil temperature was observed to be dominated by daily variations similar to radon variation which account to 97.99 % whereas soil pressure accounts for 100 % of the total variance which suggests that daily variations of soil radon (Rn-222) emanation are controlled by soil pressure in MPGO, Tezpur during the investigation period followed by soil temperature. The study concludes that SSA eliminates diurnal and semidiurnal components from time series of soil radon emanation for better correlation of soil radon emanation with earthquakes.