Imaging the San Andreas Fault between Parkfield and the Salton Sea Using Wavelet Analysis of Airborne Laser Swath Mapping Data
Abstract
The distribution of fault related landforms may be used to divulge the spatial and temporal evolution of fault ruptures within a fault zone. In this study, wavelet analysis was performed on high-resolution Airborne Laser Swath Mapping (ALSM) topographic data to image the morphologic structure of the San Andreas Fault Zone (SAFZ) between Parkfield, CA and the US-Mexico border. ASLM data were collected by the National Center for Airborne Laser Mapping as part of the B4 project and were processed these data to produce a 2-m-resolution Digital Elevation Model (DEM). The DEM tiles were imported to ArcMap, which was used to mosaic, rotate, and crop them. Matlab was used to perform a progressive filling of NODATA values within each of the tiles using an iterative nearest-neighbor averaging scheme on these data. Next, scarp-like features roughly paralleling the average trend of the SAFZ were identified using a previously developed wavelet analysis method. This method convolves the second derivative of an elongated template of a scarp-like topography with the directional curvature of the ALSM DEM that is represented by each of the tiles. In this way, the analysis recovers, in a least-squares best-fitting sense, the amplitude of a particular scarp geometry and orientation. The Signal-to-Noise Ratio (SNR) is then computed at each point in the ALSM DEM for a given template scarp geometry and orientation-- this process is repeated for all scarp geometries and orientations to determine those that have the highest SNR. Such scarp forms are automatically identified as the best-fitting scarp geometry, amplitude, and orientation at each point in the DEM. The geometry gives a quantitative measure of the "roundness" of the profile of the scarp form, and supposing that sharper scarps have been created more recently than those whose forms have been rounded by prolonged erosion, a relative chronology of activity of various fault strands within the fault zone can be reconstructed. With the data from Matlab, the tiles are then imported back to ArcMap for further processing. In ArcMap, the individual tiles from Matlab were mosaicked together to create a model of the swath. To complete this process, the mosaicked dataset was then rotated to mimic the orientation of the fault in the real world. The comparison of the location of the identified scarps to those that have been mapped in the field is currently being made to determine the fidelity of this method. Preliminary results indicate that this analysis reveals the locations of recently active fault ruptures, and as such, these computer-generated maps may provide valuable information about the history of rupture within the fault zone and evaluation of its seismic hazard potential.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFMED41A0491C
- Keywords:
-
- 0815 EDUCATION / Informal education