Detecting anomalies in electromagnetic time series in Ileia, Greece prior to near-area earthquakes

The investigation of anomalies in electromagnetic (EM) signals as a potential precursor to seismic activity is an active research area in seismo-electromagnetic studies. This study aims to examine the anomalies in EM signal and their relation with seismic events as well as to investigate the noise characteristics, fractal properties, and the presence of non-linear behavior in EM signals. EM signals at carrier frequencies of 3 kHz, 10 kHz, and 46 MHz were recorded at one second time interval, over an eighteen day window spanning nine days before and after the seismic event occurred at the study area. Welch power spectral density (PSD) analysis was employed to check the anomalous behavior in EM time series. The PSD examines the presence of underlying features by verifying the presence of broadband fractal backgrounds with superimposed narrow spectral lines indicating the presence of stochastic variations and deterministic periodic components that lies due to the external interference, modulation, or resonant processes. The PSD portrays distinct anomalous peaks before the seismic events of magnitude ML=4.1, ML=5.8, ML=3.5, ML=3.6 and ML=3.5 occurred on 29^th of March 2024, indicating the higher energy localization at various frequency bands. Some anomalous peaks were also observed before the seismic events of magnitude ML=4.7 and ML=4.4 occurred on 7^th of July and 28^th of September, 2024 respectively indicating the possible perturbations in EM signals before the seismic events. Hurst and Lyapunov exponents were employed to investigate the underlying features and to assess the transient variations in EM signals. Results portrays the transition from random to weakly correlated behavior during the pre-seismic phase. Overall, this study supports the assumption that EM signals may portray precursory signatures that are related to seismic events. Still, the observed anomalies are not consistent across all data set indicating that they may be influenced by environmental noise or meteorological parameters. This study provides a strong and important framework by combining spectral, fractal and non-linear methods for representing EM signals in terms of noise, fractal scaling, and non-linear dynamics to identify the robust indicators of the seismic events.