Fault Damage Zones Q Value Determination From Spectra Decay Analysis and Its Application to the 2019 Ridgecrest Aftershock Sequence

Abstract

An earthquake-hosting fault may present complicated geometry, which has relatively straight and simple segments to host major slips, and bifurcates into “horse tail” structures near ends. The “horse tails” play an important role in controlling the nucleation and termination behaviors of great earthquakes. Despite the obvious geometrical differences between the two structures, understanding the damage zones formed by multiple ruptures is also important in understanding their physical properties. The seismic attenuation parameter (Q) is important in understanding fault damage zones, which are commonly considered to be related to the damage intensity and rock fragmentation, and further implies the fault maturity and porosity. However, Q is not well resolved in damage zones by traditional methods because damage zones only span a few hundred meters in width. The emergence of dense portable seismic arrays and high-precision microseismic catalogs now provides a new opportunity to resolve fine-scale attenuation structures. In this study, we propose a new method to invert for the attenuation properties of seismic waves trapped in fault damage zones (fault zone trapped waves) through spectral decay analysis of amplitude ratios from events with varying propagation distances along the fault. First, we develop the trapped wave attenuation theory by analogy with Love waves. Second, we validate the feasibility of inverting attenuation curves by theoretical analysis and extracting the parameters from the attenuation curves. Furthermore, we adopt the inversion methods for two dense portable arrays (B2 and B4) deployed across the ruptured fault segments of the 2019 Mw 7.1 Ridgecrest earthquake. Our inverted results reveal markedly different attenuation structures for damage zones at different locations: the damage zone at the “horse tail” (along array B4) shows a Q value of ∼30–40 and a velocity reduction of ∼20–30%, whereas at the central fault (along B2 array) exhibits a Q of ∼50–70 and a velocity reduction of ∼10–15%. Overall, the “horse tail” faults may exhibit a more fragmented, fluid-rich, and compliant zone in comparison with the central segment. Our method provides a generalized framework to quantify near-fault attenuation from dense seismic observations. Our method is effective and can be generalized to invert attenuation parameters for other fault zone observations.