Group A, Poster #165, Fault and Rupture Mechanics (FARM)

Semi-automated extraction of fault displacement profiles and displacement-length relationships from high-resolution lidar data and standard fault maps

Mercedes Quintana, Alba M. Rodriguez Padilla, Duncan M. Chadly, & Michael E. Oskin
Poster Image: 

Poster Presentation

2022 SCEC Annual Meeting, Poster #165, SCEC Contribution #12491 VIEW PDF
Faults grow over time through incremental slip events. The finite distribution of displacements on faults, and the ratio of the maximum displacement at the surface to the fault length are frequently sought after in fault mechanics, as these metrics reflect the physics of fault growth and the host material. Traditional methods are time-consuming and often spatially sparse which limits the potential of high resolution topographic datasets. We developed a semi-automated MATLAB algorithm to measure throw from high-resolution lidar topography. Our inputs are the lidar DEM for the region and a simple fault map. We first collect fault-perpendicular topographic profiles evenly spaced along each faul...t. We then train a Support Vector Machine to detect scarps based on spatial slope characteristics in a manually curated subset. Finally, we use the second derivative of elevation to fit each scarp in the profile and calculate throw. The algorithm outputs the displacement profile for every fault mapped, and the maximum displacement vs length relationship for the network of faults. This approach enables rapid and standardized collection of fault throw and length metrics for large datasets. We tested our algorithm on normal faults in the Volcanic Tableland in Bishop, California. Going forward, we will validate our method on other landscapes dominated by normal faulting where high-resolution topography is available.