OR-Corr a New Outlier-Resistant Image Correlation Method: Examples of Surface Deformation and Complex Fracturing from the 2019 Ridgecrest Earthquake Sequence Imaged by High-resolution Optical and Radar Imagery

Chris Milliner, Andrea Donnellan, Rui Chen, Robert Zinke, Xiaohua Xu, Alex E. Morelan, Timothy Dawson, Christopher M. Madugo, James F. Dolan, Alexandra Sarmiento, Yousef Bozorgnia, & Adnan Ansar

Published August 15, 2019, SCEC Contribution #9732, 2019 SCEC Annual Meeting Poster #228

The 2019 Ridgecrest earthquake sequence, CA, involved a Mw 6.4 foreshock on July 4th and the Mw 7.1 mainshock that occurred 34 hours later located ~12 km to the north. This sequence resulted in a complex array of surface fractures that were measured from a suite of geodetic imaging sensors which we use to study the faulting mechanics and kinematics. Here we present coseismic surface deformation measurements from a range of geodetic imaging datasets, including optical and radar, processed using a new MATLAB-based image correlation method that we have developed called OR-Corr (Outlier Resistant Correlator). This image correlator is designed to be more resistant to pixel outliers that may occur between the before and after images (e.g., due to non-coherent changes from building damage or areas of vegetation change), providing a more robust retrieval of pixel and surface motion. Applying OR-Corr to Planet Labs optical imagery (3 m pixel resolution), which acquired images between the Mw 6.4 and Mw 7.1 events, we have been able to successfully separate the surface deformation of the two earthquakes and gain a first-order understanding of which faults did and did not rupture prior to the mainshock. Second, we show an additional capability of OR-Corr, which is to solve for rotations of the coseismic deformation field directly from the images themselves, which allows for measurement of block rotation and provides a new functionality for image correlation techniques. Lastly, combining the optical correlation results with detailed fracture mapping from Sentinel-1 radar azimuthal gradient maps we discuss how these high-resolution datasets, which can measure the decay of inelastic strain away from the primary rupture, can be used to constrain next-generation geodetic-based probabilistic fault displacement hazard models.

Key Words
Ridgecrest earthquake optical radar image correlation deformation

Citation
Milliner, C., Donnellan, A., Chen, R., Zinke, R., Xu, X., Morelan, A. E., Dawson, T., Madugo, C. M., Dolan, J. F., Sarmiento, A., Bozorgnia, Y., & Ansar, A. (2019, 08). OR-Corr a New Outlier-Resistant Image Correlation Method: Examples of Surface Deformation and Complex Fracturing from the 2019 Ridgecrest Earthquake Sequence Imaged by High-resolution Optical and Radar Imagery . Poster Presentation at 2019 SCEC Annual Meeting.

Related Projects & Working Groups
Ridgecrest Earthquakes