Disentangling Fault Scarp Geometry and Slip-Distribution in 3D

Abstract:

We present a new and inherently 3D approach to the analysis of fault scarp geometry using high resolution topography. Recent advance in topographic measurement techniques (LiDAR and Structure from Motion) has allowed the extensive measurement of single earthquake scarps and multiple event cumulative scarps to draw conclusions about along-strike slip variation and characteristic slip. Present analysis of the resulting point clouds and digital elevation models is generally achieved by taking vertical or map view profiles of geomorphic markers across the scarp. Profiles are done at numerous locations along strike carefully chosen to avoid regions degraded by erosion/deposition. The resulting slip distributions are almost always extremely variable and “noisy”, both for strike-slip and dip-slip faulting scarps and it is often unclear whether this reflects slip variation, noise/erosion, site effects or geometric variation.

When observing palaeo-earthquake and even modern event scarps, the full geometry, such as the degree of oblique slip or the fault dip, is often poorly constrained. We first present the results of synthetic tests to demonstrate the introduction of significant apparent noise by simply varying terrain, fault and measurement geometry (slope angle, slope azimuth, fault dip and slip obliquity). Considering fully 3-dimensional marker surfaces (e.g. Planar or conical) we use the variation in apparent offset with terrain and measurement geometry, to constrain the slip geometry in 3D. Combining measurements windowed along strike, we show that determining the slip vector is reduced to a simple linear problem. We conclude that for scarps in regions of significant topography or with oblique slip, our method will give enhanced slip resolution while standard methods will give poor slip resolution. We test our method using a Structure from Motion pointcloud and digital elevation model covering a ~25 km stretch of a thrust fault scarp in the Kazakh Tien Shan.