Long-Memory and Fractal Traces in KHz-MHz Electromagnetic Time Series Prior to the ML=6.1, 12/6/2007 Lesvos, Greece Earthquake: Investigation through DFA and Time-Evolving Spectral Fractals

Stochastic Models for Radon Daily Time Series: Seasonality, Stationarity, and Long-Range Dependence Detection

“…The estimated fractional integration parameters of this ARFIMA model is positive and smaller than 0.5 (d 0.278). It corresponds to a Hurst's coefficient of H 0.778 that is consistent with the results obtained with the rescaled range analysis and in general with other literature results (Cuculeanu et al, 1996;Nikolopoulos et al, 2018). The proposed model has been also assessed in terms of its predictive capacity, however the performances are quite poor especially increasing the lag of prediction (moving from 1-day forecast to 5-day forecast).…”

“…For instance, in Cuculeanu et al (1996), the determined values of Hurst's coefficient (0.809) highlight a persistent behavior of the gas. Also, Nikolopoulos et al (2018) compute the rescaled range analysis for several time intervals obtaining a persistent Hurst exponent between 0.7 and 0.9 and in some periods between 0.9 and 1.…”