Nonlinear dynamic analysis of the fault activities induced by groundwater level variations

Abstract

This study explored the possible repercussions of water-table reduction on the activity of faults in California, USA. The statistical period was regarded between 1972 and 2020. The cross-correlation and dynamic analysis were performed on the constructed time-series of the number and magnitude of the earthquakes. The cross-correlation results indicated a correlation between faults activity and water-table fluctuations. The dynamic analysis results of the monthly earthquake numbers indicated that declining groundwater levels produced irregularities and the conversion of periodicity in the time-series. Moreover, chaos and sensitivity to the initial conditions were diminished, while the randomness in time-series developed (m increased from 6 to 7). Additionally, the correlation and long-range persistency in time-series were reduced (H(2) decreased from 0.6493 to 0.6304). Furthermore, the number of earthquakes in a month and multifractality were limited (for ΔH from 1.8557 to 1.1701 and Δα from 2.0726 to 1.3743). The dynamic analysis results of the earthquake magnitude time-series suggested that the periodicity and predictability were reduced, while irregularities and randomness in time-series were extended. Moreover, long-memory, correlation, and multifractal characteristics in time-series were diminished (for H(2) from 0.9506 to 0.6310, ΔH from 0.1277 to 0.0536, and Δα from 0.27 to 0.1665). Besides, the possibilities of large and small earthquakes were balanced, while the probability of an earthquake of a greater magnitude shifted. The cross-correlation results were consistent with the dynamic analysis findings and confirmed water-table reduction correlation with faults’ seismic activity. In summary, seismicity pattern (rates and magnitude) was altered due to a reduction in water-table.