Poster #067, Earthquake Geology

Effectiveness and reproducibility of remote mapping of the 2019 Ridgecrest earthquake ruptures with airborne lidar and imagery

Elaine K. Young, Michael E. Oskin, & Alba M. Rodriguez Padilla
Poster Image: 

Poster Presentation

2021 SCEC Annual Meeting, Poster #067, SCEC Contribution #11366 VIEW PDF
Remote rupture maps are useful contributions to earthquake response, probabilistic fault displacement hazard analysis, and as training datasets for remote mapping with artificial intelligence. We use post-earthquake lidar data to remotely map surface ruptures produced by the 2019 Ridgecrest Earthquake sequence. The purpose of this study is to 1) develop an objective, uniform map product from which we test the reproducibility of remote surface-rupture mapping and the accuracy of remote compared to field-derived surface-rupture mapping 2) evaluate the utility of lidar compared with other remotely collected datasets in post earthquake response. This study presents remote mapping of the surface ...rupture by three independent mappers using the post-earthquake airborne lidar. Rupture map shapefiles were converted into 1m rasters to remove line length and continuity from the analysis and buffers of various widths were used to account for differences in line placement on a feature with some width (e.g. fault scarp). Map comparisons used the percent overlap between one mapper’s buffered and another mapper’s unbuffered rasterized rupture map. Our maps show ~70-80% overlap for features with steep slopes that are > 50 cm, or one pixel in the lidar data. For features that are less steep, interpretations of the data vary across mappers, with ~40-60% overlap. Compared with published rupture maps using field observations and other data, we have ~30% overlap with 1m buffers and ~40% with 2m buffers, likely due to the limited horizontal resolution of the lidar. Remote mapping using drone imagery captures significantly more details than lidar alone, but is much more time consuming to collect and manually map remotely for large areas. On its own, post-event lidar should not be used without field observations or another method of verifying that mapped features are from the recent rupture, especially for a rupture surrounded by abundant lineaments formed by other mechanisms, (e.g. shorelines, pre existing fault scarps, etc.) as are present at Ridgecrest. Compared with other remote datasets, like image differencing or radar (e.g. InSAR), post-event lidar collection is slow and less sensitive to features produced by horizontal slip. Lidar is most useful for capturing vertical deformation, especially combined with horizontal motion captured in other datasets and when differencing from pre-event lidar.