Paper No. 17-5
Presentation Time: 3:10 PM
EXPLORING THE USE OF UNOCCUPIED AERIAL SYSTEMS FOR STUDYING EARTHQUAKE-INDUCED LIQUEFACTION DEPOSITS
ROGERS, Stephanie, Auburn University, Department of Geosciences, 2050 Beard Eaves Coliseum, Auburn, AL 36849, WOLF, Lorraine, Department of Geosciences, Auburn University, 2050 Beard Eaves Coliseum, Auburn, AL 36849, GUVEN, Can, Auburn University
The Department of Geosciences, 2050 Beard Eaves Coliseum, Auburn, AL 36849-0001 and MATTHEWS, Steffen, Auburn University, 2047 Beard Eaves Coliseum, Department of Geosciences, Auburn, AL 36849
Paleoliquefaction deposits (prehistoric sand blow deposits and large sand-filled fissures formed as the result of strong ground shaking from prehistoric earthquakes) serve as the principal source of information about the occurrence of large (magnitude (M) > 6) earthquakes. Establishing earthquake history and identifying seismic sources is critical for estimating the region’s seismic hazard and the recurrence times of large, damaging earthquakes that could affect the cities and towns within the Mississippi River valley, including the metropolises of Memphis and St. Louis. This proof-of-concept study aims to obtain new knowledge of earthquake-induced paleoliquefaction deposits in the New Madrid Seismic Zone (NMSZ) using multispectral sensors on Unoccupied Aerial Systems (UASs). We hypothesize that a correlation exists between sand blow age and the amount of organic content in the sediment, with older deposits having higher organic content than younger deposits, which we expect to be sandier, and that these characteristics are distinguishable from multispectral imagery from UASs.
We collected UAS data with a Phantom 4 Pro with Parrott Sequoia sensor and a Matrice 200 with an Altum sensor in October 2019 and November 2020, respectively. We flew at three paleoliquefaction sites in Yarbro, AR, one of which has historic sand blow deposits and the other two having prehistoric sand blows deposits. Pix4D Mapper was used to photogrammetrically process the data to create high-resolution (~2 cm) Digital Surface Models (DSMs), orthoimagery, and multispectral raster bands (red, red-edge, near-infrared, green, blue, and LWIR (thermal)). In the next step of the project, these bands will be used to calculate indices for estimating soil development in sand blows based on differing spectral characteristics. Additionally, spectral information will be compared to soil samples to determine if relative sand blow ages can be established. Results from the study, if the technique proves effective, will contribute to reconnaissance efforts to identify prehistoric deposits from which to establish the history of large earthquakes in the NMSZ.