Analysis of Solid Earth Tides Using Singular Spectrum Analysis: Implications for Earthquake Precursors

Abstract

The traditional field of research has been enriched by discovering and applying several unique approaches for extracting relevant information from a time series of Solid Earth tides for a specific location. This study focused on the impactful Mw 9.1 megathrust earthquake of December 26, 2004, in Sumatra-Andaman, which spans 13 earthquakes within a 75-kilometer radius from 1991 to 2021. Employing the univariate singular spectrum analysis (SSA) on Solid Earth tides (SET), a significant anomalous variation in the sixth component of empirical orthogonal functions (EOF) is identified. It is noted that the eigenvalue corresponding to this component exhibits a correlation with earthquake magnitude. Further, this study is extended to a broader spatial region to perform in-depth spatio-temporal analysis exclusively for the December 26, 2004 earthquake, by expanding spatial analysis to latitudes 2° to 5° N and longitudes 94° to 98° E through a sliding window method. A spatio-temporal analysis from 1995 to 2004 reveals distinct patterns preceding higher magnitude earthquakes, offering valuable insights for long-term forecasting. It has been correlated with foreshocks that occur to the same spatiotemporal extent. In a nutshell, singular spectrum analysis SSA of Solid Earth Tides emerges as a promising precursor for long-term earthquake magnitude prediction. Also, EOF alone is insufficient for accurately forecasting the magnitude, as the results show an inverse relationship with magnitude. Finally, it might be combined with other parameters like foreshocks for reliable earthquake forecasting.