![]() Throughout the crust and upper mantle, variable fault zone damage inferred from geophysically constrained mechanical properties (e.g., Froment et al., 2014 Roland et al., 2012), coupled to enhanced fluid circulation and growth of frictionally weak, hydrous phyllosilicates (e.g., Roland et al., 2010), provides one hypothesis to explain persistent along-strike segmentation in seismic behavior. Several recent studies have linked seismic style to alteration and damage of rocks. This earthquake deficit has been explained in two ways: (a) oceanic transforms experience both earthquakes and aseismic creep along the same fault segments, but at different times (e.g., Abercrombie & Ekström, 2001 Hilley et al., 2020 McGuire et al., 1996) and/or (b) oceanic transforms are segmented into “locked patches” hosting quasi-periodic earthquakes of Mw > 6.0, and microseismically active “rupture barriers” dominated by creep (e.g., McGuire, 2008 McGuire et al., 2005, 2012 Sykes & Ekström, 2012 Wolfson-Schwehr et al., 2014). Instead, up to 95% of displacement occurs aseismically, despite the faults cross-cutting the brittle mafic crust (Boettcher & Jordan, 2004). Oceanic transform faults have low seismic coupling, and display far fewer and smaller earthquakes than expected from fault length-magnitude scaling relations, based on the Harvard centroid moment tensor catalog (e.g., Bird et al., 2002). Our findings suggest that variations in the size and number of earthquakes on oceanic transform faults is controlled most of all by how damaged the existing rock is and how much alteration to weak, water-bearing secondary minerals such as chlorite has occurred along fault planes. The same damage and alteration responsible for the weakness also prevents earthquakes. In contrast, we find that already damaged and altered rocks, found within natural faults (containing an increased proportion of the mineral chlorite), are weak. We find that dolerite, one of the primary rock-types of the oceanic crust, is strong and capable of starting earthquakes. We sheared cylindrical samples, holding one half in place and sliding the other over it, creating laboratory equivalents of geological faults. To understand these characteristics, we conducted laboratory deformation experiments using rocks collected from the ocean floor (near Hess Deep in the Pacific) and from an ancient transform fault (in Cyprus). ![]() The reasons for their weakness and lack of large earthquakes are puzzling. Fewer and smaller earthquakes than expected occur along these faults, which are also considered weak structures. Oceanic transform faults are plate boundary faults where motion of oceanic lithosphere is dominantly horizontal and parallel to tectonic motion. This transition implies that seismic behavior is controlled by degree of damage and alteration, such that earthquakes can nucleate within relatively intact oceanic crust, whereas fault segments of increased damage and chlorite content tend to slip aseismically. In contrast, matrix-rich chlorite-bearing fault breccias and gouges are frictionally weak ( μ = 0.25–0.48) and velocity-strengthening (characteristic of stable creep). Dolerites and cemented breccias are frictionally strong ( μ = 0.52–0.85) and velocity-weakening (strength decreases with increasing slip velocity, characteristic of earthquakes). We test the effects of fault-rock evolution on oceanic transform fault frictional strength and stability using direct-shear experiments (at room temperature, 10 MPa normal stress, and fluid-saturated conditions) on dolerite from the East Pacific Rise and natural fault rocks from the exhumed Southern Troodos Transform, Cyprus. Neither their weakness nor tendency to creep are well-explained. Oceanic transform faults are inferred to be weak relative to surrounding oceanic crust and primarily slip aseismically.
0 Comments
Leave a Reply. |