Temporal Analysis of Ground Movement at a Metal Mine in China

• Published in: Remote sensing (Basel, Switzerland) Switzerland), 2022-10, Vol.14 (19), p.4993
• Main Authors: Li, Guang, Hui, Xin, Ma, Fengshan, Guo, Jie
• Format: Article
• Language: English
• Subjects: Autocorrelation functions; Behavior; Earth sciences; Earthquakes; Fractals; Geological hazards; Global positioning systems; GPS; Ground motion; ground movement; Land subsidence; Mining; Monitoring; Nickel; Periodicity; Power law; Power spectra; Resonant frequencies; Scale invariance; self-affine time series; Time series; Underground construction
• ISSN: 2072-4292; 2072-4292
• DOI: 10.3390/rs14194993

Mining-induced ground movement is a complicated nonlinear process and a regional geological hazard. Time series in Earth sciences are often characterized as self-affine, long-range persistent, where the power spectra exhibit a power-law dependence on frequency. Whether there exists a periodic signal and a fundamental frequency in the time series is significant in analyzing ground-movement patterns. To evaluate whether a power law describes the power spectra of a ground-movement time series and whether a fundamental frequency exists, GPS monitoring records taken over 14.5 years describing ground movement in the Jinchuan Nickel Mine, China, were analyzed. The data sets consisted of 500 randomly selected GPS monitoring points, spanning the April 2001–October 2015 time period. Whether a periodic signal in the ground movements existed was determined through the autocorrelation function. The power spectra of the ground-movement time series were found to display power-law behavior over vastly different timescales. The spectral exponents of the horizontal and vertical displacements ranged from 0.47 to 3.58 and from 0.43 to 3.37, with mean values of 2.05 and 1.79, respectively. The ground movements of minefields No.1 and No.2 had 1.1-month and 1.4-month fundamental periods, respectively. Together with a discussion of the underlying mechanisms of power-law behavior and relevant influencing factors, these results indicate that ground-movement time series are a type of self-affine time series that exhibit long-range persistence and scale invariance and show a complex periodicity. These conclusions provide a basis for predicting land subsidence in the study area over a timescale.