Evolution of Tidal Influence During the ETS Seismic Cycle Reveals Competition Between Tectonic Loading and Fault Healing

Abstract:

Following the discovery of the evolution of tremor response to tidal stress over the duration of ETS slip at a spot (Houston 2015; Royer et al 2015;Yabe et al 2015), we investigate whether and how it may evolve between major large ETSs, which occur quasi-periodically in several subduction zones. Preliminary results show that tidal response does evolve over the average interETS period in northern Cascadia - decaying over the first quarter of the cycle to lowest values then climbing back up in the second half of the cycle part of the way toward the strong response seen late during major ETSs. Thus far, we have ignored the possible role of transient stresses during interETS tremor because tremor bursts are mostly small.

We explore a strength-threshold model where tidal influence is stronger when stress is close to strength and weaker when they are farther apart. Shortly after a major ETS, both stress and strength are presumed to have fallen over the large region where slow slip occurred. Then, however, stress rebuilds quasi-linearly by plate tectonic loading, whereas strength rebuilds as the logarithm of time (e.g., Vidale et al 1994). Thus, model stress and strength diverge the most midway through the interETS cycle, the period of weakest tidal sensitivity. Tidal stresses become more effective in triggering tremor later in the cycle as the linearly-growing stress approaches the logarithmically-growing strength. This model broadly fits our observed evolution of tidal response. However, the tendency of ETSs to initiate downdip may require an additional process that varies along dip. This approach illuminates the competition between healing on the plate interface and reloading with tectonic stress, and can help constrain and perhaps even monitor physical conditions on the deep subduction interface.

The figure shows the evolution of two measures of tidal influence on tremor, consistency and sensitivity (right), and the data on which they are based - probability distributions of stress at the times of tremor and the strength of the exponential relation between tremor rate and tidal stress (left). Tremor and stress data for 6 normalized interETS time periods are averaged together.