SCEC Award Number 07181 View PDF
Proposal Category Individual Proposal (Data Gathering and Products)
Proposal Title Fault Surface Bumps: Prevalence, Geological Constraints and Consequences
Investigator(s)
Name Organization
Emily Brodsky University of California, Santa Cruz
Other Participants Amir Sagy
Mai-Linh Doan
Undergraduate Student
SCEC Priorities A10, A7, A9 SCEC Groups Geology, FARM
Report Due Date N/A Date Report Submitted N/A
Project Abstract
 The goal of the project was to measure the irregularities of fault surfaces together with the fault zone internal structure, and thus generate the 3D picture of faults that is necessary for the physical modeling of earthquakes. Our measurements demonstrated that slip surfaces of small-slip faults are relatively rough at all measured scales, whereas those of large-slip faults are polished at small scales, but surprisingly contain elongated bumps and depressions at scales of a few to several meters. This project focussed on clarifying the last observation by connecting the surface topography to the internal geometry at a particularly well-exposed normal fault in a quarry near Klamath Falls, Oregon.

We found that the fault zone has a layered damage architecture. Slip primarily occurs inside a 1-20 mm wide band that contains principal slip surfaces with individual widths of ~100 m. The slip band sits atop a cohesive layer which deforms by granular flow. Several fault strands with total slips of 0.5-150 m also have cohesive layers with thicknesses increasing monotonically with slip. The thickness added to the cohesive layer per unit slip decreases with increasing displacement indicating that slip progressively localizes. The main fault is a continuous surface with 10-40 m long quasi-elliptical geometrical asperities, i.e., bumps. The bumps reflect variations of the thickness of the granular cohesive layer and can be generated by a pinch-and-swell instability. As the granular layer is rheological distinct from its surroundings, the asperities are both geometrical and rheological inhomogenities.
Intellectual Merit Our earlier data yielded two major conclusions (Sagy et al., 2007). First, small-slip faults (slip <1 m) are rougher than large-slip faults (slip 10 to 100 m or more) on profiles parallel to the slip direction at the scale of slip during moderate earthquakes (a few meters). Secondly, large-slip fault surfaces show elongate bumps that are meters long and up to several decimeters high. The origin of these irregularities was unknown. The bumps are a clear departure from the self-similar or self-affine models previously used to model fault geometry. Such geometrical asperities in the fault surfaces can be connected to seismological asperities which are an important part of rupture nucleation and arrest. This project identified these geometric bumps with a thickening of an internal layer and thus demonstrated that they functioned as both geometrical and rheological asperities.
Broader Impacts This project required analysis of an unwieldy large-point cloud of LiDAR data. We needed a customized point cloud viewer and editor to make the problem tractable. Using SCEC ACCESS funds, a UCSC computer science major, Michael Steffeck, built a suitable viewer, SlugView, and made the software freely available to the community through the UCSC website. This position moved Michael from a bagging groceries to support his education to a fully capable and employable engineer who has moved on to a career in technical programming for national security applications. SlugView has become a core tool of fault roughness analysis from ground-based LiDAR.
Exemplary Figure Figure 4. Variations of the width of Layer II under protrusions and depressions of the fault surface. a) Variations of Layer II within the exposure of the middle part of Flowers Pit Fault surface. The biggest bump (marked by B) overlies the most thickened Layer II with a width >1 m at the tip of the arrow. The largest depression (marked by D) overlies the most thinned Layer II with a minimum width of 5 centimeters. Smaller perturbations of the width of Layer II are also observed under smaller protrusions and depressions. b) Two examples of local thickening of Layer II. Under protrusions (left) while thin appearance of the layer is observed under depressions (right). c) Maximum observed width of Layer II under large bumps measured from six protrusions, and minimum width of Layer II measured under ten large depressions. The background values measured in areas with relatively small amplitude variations of the surface (32 measurements). Error bars indicate 1 standard deviation.