In this paper, based on non-negative matrix factorization (NMF), we analyzed the ionosphere magnetic field data of the Swarm Alpha satellite before the 2016 (Mw = 7. 8) Ecuador earthquake (April 16, 0.35°N, 79.93°W), including the whole data collected under quiet and disturbed geomagnetic conditions. The data from each track were decomposed into basis features and their corresponding weights. We found that the energy and entropy of one of the weight components were more concentrated inside the earthquake-sensitive area, which meant that this weight component was more likely to reflect the activity inside the earthquake-sensitive area. We focused on this weight component and used five times the root mean square (RMS) to extract the anomalies. We found that for this weight component, the cumulative number of tracks, which had anomalies inside the earthquake-sensitive area, showed accelerated growth before the Ecuador earthquake and recovered to linear growth after the earthquake. To verify that the accelerated cumulative anomaly was possibly associated with the earthquake, we excluded the influence of the geomagnetic activity and plasma bubble. Through the random earthquake study and low-seismicity period study, we found that the accelerated cumulative anomaly was not obtained by chance. Moreover, we observed that the cumulative Benioff strain S, which reflected the lithosphere activity, had acceleration behavior similar to the accelerated cumulative anomaly of the ionosphere magnetic field, which suggested that the anomaly that we obtained was possibly associated with the Ecuador earthquake and could be described by one of the Lithosphere–Atmosphere–Ionosphere Coupling (LAIC) models.

In this paper, we revisit this earthquake by simultaneously analyzing the magnetic field data of Swarm satellite A and satellite C based on principal component analysis (PCA), and the eigenvalues and principal components are calculated throughout 2016. We find that the first principal component mainly contains the signal originating from solar-terrestrial effects such as geomagnetic activity since the first eigenvalue and the geomagnetic index are highly correlated. Therefore, the second principal component is used to extract the anomalies associated with the Ecuador earthquake in terms of skewness and kurtosis. The anomalous tracks of the S-K (Skewness-Kurtosis) coefficient are accumulated from 90 days before the event to 30 days after. The cumulative number follows an accelerating power-law behavior before the earthquake and decelerating recovery behavior after the earthquake; moreover, the inflection point of the sigmoidal fitting curve is close to the time of the earthquake. The cumulative number of anomalous tracks in the random regions shows a linear increase, further verifying the correlation between anomalies extracted and the Ecuador earthquake. This phenomenon could be related to the preparation phase and the aftershock phase of the Ecuador earthquake.

A YRY-4 borehole strainmeter installed at the Guza Station recorded anomalous changes in borehole strain data preceding the Lushan earthquake on April 20, 2013 (UTC) (M w = 7.0). To identify earthquake-induced abnormal strain changes, we apply principal component analysis (PCA) for the first time to analyse the borehole strain data from the Guza Station. The first principal component eigenvalues and eigenvectors demonstrate that the anomalous days are mainly concentrated within two time periods: 1) October 25-December 30, 2012, and 2) April 15-19, 2013. A combined eigenvalue and eigenvector analysis reveals that the abnormal days exhibit a clustered distribution that is aggregated in the same location for both periods, intuitively indicating that there is a forceful correlation between the two anomalies. We tentatively infer that a similar process contributed to the formation of both anomalies and that these two anomalies are both earthquake precursors associated with the Lushan earthquake. These findings also indicate that the PCA approach exhibits potential for the extraction of earthquake precursor anomalies.

To investigate the nonlinear spatio-temporal behavior of earthquakes, a complex network has been built using borehole strain data from the southwestern endpoint of the Longmenshan fault zone, Sichuan-Yunnan region of China, and the topological structural properties of the network have been investigated based on data from 2011–2014. Herein, six observation sites were defined as nodes and their edges as the connections between them. We introduced Multi-channel Singular Spectrum Analysis (MSSA) to analyze periodic oscillations, earthquake-related strain, and noise in multi-site observations, and then defined the edges of the network by calculating the correlations between sites. The results of the daily degree centrality of the borehole strain network indicated that the strain network anomalies were correlatable with local seismicity associate with the earthquake energy in the strain network. Further investigation showed that strain network anomalies were more likely to appear before major earthquakes rather than after them, particularly within 30 days before an event. Anomaly acceleration rates were also found to be related to earthquake energy. This study has revealed the self-organizing pre-earthquake phenomena and verified the construction of borehole networks is a powerful tool for providing information on earthquake precursors and the dynamics of complex fault systems.