Multi-spectral detection of statistically significant components in pre-seismic electromagnetic emissions related with Athens 1999, M=5.9 earthquake

Detection Test for Periodic Signals Revisited Against Various Stochastic Models

“…At these CLs the MTM spectrum indicates spectral features at extra long-scales ( > 30 days), diurnal (24h) and semidiurnal (12h) terms are statistically significant, whereas spectral features at long-scales (2-10 days) are not statistical significant at the 95% CL based on a AR(1)+WH null. Our result is partially in agreement with the findings in [20]. Additionally, we find that most spectral features (except those at diurnal and semidiurnal hours) are not statistical significant in the MTM spectrum after applying the Bonferroni correction [22] by dividing the significance level (e.g., α = 0.05) by the number of tested data as in [7].…”

“…1(c) shows that extra long-scales variability modes (> 30 days) and oscillatory modes at 3-6 days vanish in the high activity epoch (98.5 ≤ DN ≤ 100.5) in the areawise significance test (α = 0.005), which are based on the area and shape of the significant regions proposed by Maraun et al [15] but modified here by just using the chi-square test (for details, see Equation (18) in TC98) but not the MC simulations to reduce the computation cost. In particular, oscillatory terms between 8 and 10 days become significant at about day number DN = 97, and not at DN = 76 as charmed by Kalimeris et al [20]. We also find that oscillatory terms between 8 and 10 days remain in the high activity scale for approximately 3 days after the main seismic shock occurred at DN = 100.498.…”

“…Furthermore, we find that the periodicities shorter than 45 min and the intermittent diurnal/semidiurnal oscillatory modes are still stable and significant. Based on these results, it seems plausible that the significant spectral features claimed by Kalimeris et al [20] are artifacts of a poor choice of null ''noise'' hypothesis and a lack of multiple test correction, and we led to the conclusion that the transition period (82.6 ≤ DN ≤ 90.2) before the Athens' earthquake (M = 5.9, 7 September 1999) defined by Kalimeris et al [20] may be unreliable. Anyway, our new results with those found in [20] can enhance the preseismic character of the recorded kHz EME time series.…”

“…As the first test case we revisit the kHz electromagnetic emission (EME) time series in North-South orientation associated with the Athens' earthquake (M = 5.9, 7 September 1999), which were acquired using a 10 kHz tuned loop antenna installed on a carefully selected mountainous site at Zakynthos Island (37.76 o N, 20.76 o E), Greece. Here we only focus on the filtered version (starts on 00:00:00 (UT) of May 30, 1999, and finishes on 23:59:59 (UT) of September 29, 1999, covering a period of 123 days with a time resolution of 10 min) of the original EME time series, which has been detailed discussed by Kalimeris et al [20] with ML96 and TC98 method against the hypothesis an AR(1) red noise. In this section, we reanalysis the EME time series (see Fig.…”

Recent seismic activity at Cephalonia (Greece): a study through candidate electromagnetic precursors in terms of non-linear dynamics

“…The possible relation of the field-observed fracture-induced electromagnetic emissions (EME) in the frequency bands of MHz and kHz, associated with shallow EQs with magnitude 6 or larger that oc-curred inland or near the coast, has been examined in a series of publications in order to contribute to a better understanding of the underlying processes (e.g. Eftaxias et al, 2001Eftaxias et al, , 2004Eftaxias et al, , 2008Kapiris et al, 2004;Karamanos et al, 2006;Papadimitriou et al, 2008;Contoyiannis et al, 2005Contoyiannis et al, , 2013Potirakis et al, 2011Potirakis et al, , 2012aPotirakis et al, , b, c, 2013Potirakis et al, , 2015Minadakis et al, 2012a, b;Balasis et al, 2011Balasis et al, , 2013Donner et al, 2015;Kalimeris et al, 2016). Additionally, a four-stage model for the preparation of an EQ by means of its observable EM activity has been recently put forward Contoyiannis et al, 2015, and references therein).…”