ANALYZING SEISMIC SIGNATURES: SDFA FOR DISCRIMINATING EXPLOSIONS AND EARTHQUAKES

Abstract

The Smoothed Detrended Fluctuation Analysis (SDFA) method, introduced by Linhares in 2016, offers a valuable tool for characterizing long-range correlations in time series data. Building upon the foundations of Detrended Fluctuation Analysis (DFA) and the wavelet shrinkage procedure, SDFA computes statistical fluctuations measures, denoted as F(l), using varying window lengths (l). By analyzing these measures across different window lengths, SDFA derives a scaling exponent, specifically the slope coefficient obtained through regression analysis of F(l) against ln(l), where l varies within a specific range determined by g(n). Linhares, in a subsequent work in 2016, established an optimal choice for g(n), where g(n) is defined as ⌊(ln(n))⌋ with ⌊⋅⌋ representing the integer part function and n denoting the length of the time series. This abstract explores the fundamental concepts and procedures underlying the SDFA method, highlighting its significance in uncovering long-range correlations in time series data.