Time-frequency analysis of radon and thoron data using continuous wavelet transform

Continuous exposure to environmental radiation, whether it derives from natural or artificial sources, is thought to pose a substantial risk to public health. In addition to the health effects associated with prolonged exposure to environmental radiations, long-term measurements of these radiations can be used for a variety of beneficial purposes, such as the forecasting of impending earthquakes. Signal processing is an important application used for the purpose of forecasting. Wavelets, being signal-processing tools, are helpful in many applications such as anomaly detection in time series data. However, selection of the best wavelet for a particular application is still a problem that hasn't found a satisfactory solution. In this study, we used continuous wavelet transform (CWT) on environmental radiations, specifically radon time series (RTS) and thoron time series (TTS) data, for the investigation of time-frequency information (TFI). The distribution of energy in the output wavelet decomposition have been investigated by several wavelet families such COIF4, DB4, SYM4 to detect frequency composition of signal and its relation with anomalies hidden in the observed data. Using discrete wavelet transform (DWT), specifically SYM4, DB4, and COIF4, we transformed the radon and thoron time series into a time-dependent sum of frequency components. Using CWT scalograms, the anomalies in the both of time series datasets (TSD) have been identified, and these anomalies have been associated with the seismic events that occurred during the period of the study. The results show that DB4 and SYM4 wavelets are good at identifying anomalies in original radon and thoron TSD, but SYM4 performs better for DWT-decomposed radon and thoron TSD.