Liquefaction Phenomenon Identification from post-Pohang Earthquake 2017, South Korea based Integrated SAR and optical imagery data using a machine-learning approach

Abstract

Abstract An interesting liquefaction phenomenon occurred after the 5.6 Mw Pohang earthquake on November 15, 2017. Liquefaction changes soil density due to earthquake vibrations, which in turn causes water to rise and merge with solid soil. Remote sensing data, particularly those obtained with differential interferometric synthetic aperture radar (DInSAR), can be used to determine surface changes and changes in soil moisture. This study used data from the Sentinel-1 C-band from 2016–2020. In addition, optical satellite data were used to determine the changes in water content and soil moisture that occurred after the Pohang earthquake. A combination of spectral bands sensitive to changes in water content were used based on Sentinel-2 and Landsat-8 data. This was correlated with changes in water content, which were analyzed using optical satellite and DInSAR data with a spatial accuracy of 60–80% in comparison with field data. Moreover, the liquefaction susceptibility map has been generated using convolutional neural network with the AUC value is 0.813. However, this research was the initial research to determine the potential of liquefaction in the future based on 2017 Pohang earthquake, and the results can be the reference to improve our understanding of this liquefaction phenomenon.