Identification of pre-seismic anomalies of soil radon-222 signal using Hilbert–Huang transform

Concentration of Rn-222 in soil has been monitored continuously at Ravangla in the Sikkim Himalayan Region of eastern India for about 7 months from October 2015 to May 2016 to detect earthquake-induced anomalies. The recorded data clearly show that various physical and meteorological parameters influence the soil radon concentration, leading to very complex soil Rn-222 time series. The components due to such external influences have been removed from the present time series, and Hilbert–Huang transform (HHT) applied for analysis of the data. Two radon anomalies caused due to earthquakes of magnitude M = 5.0 that occurred on 19 November 2015 and 5 April 2016 within an epicentral distance of 500 km from the monitoring station have been identified on the soil Rn-222 time series. These two precursory anomalies occurred 9 and 10 days, respectively, before the occurrence of the earthquakes. The absence of spurious signals or missing anomalies demonstrates that HHT is advantageous for analysis of nonlinear non-stationary data, and hence, it is a promising technique to analyse soil radon behaviour for predicting the possibility of occurrence of earthquakes.