UC Berkeley

Liquefaction mapping currently relies on visual analysis of satellite and aerial images that covers the area of interest. Sand boils, lateral spreading and ground settlement are common evidences that researchers look for when identifying liquefaction damaged areas after an earthquake. However, this method can be unreliable. For example, irrelevant objects on images like burnt rice stalks in agricultural field can be misclassified as sand boils. Interferometric Synthetic Aperture Radar (InSAR), as a remote sensing technique, is widely used for characterizing ground displacement with high resolution and accuracy. With the increasing availability of free InSAR data sets and the short return period of modern satellites, co-seismic InSAR analysis becomes possible shortly after the occurrence of an earthquake. Therefore, InSAR is well suited for analyzing lateral and vertical displacement caused by liquefaction and works as a reference for field reconnaissance teams. In this thesis three earthquake case histories with significant liquefaction damage, specifically the 2010-2011 Canterbury Earthquake Sequence, including 2010 Darfield Earthquake and 2011 Christchurch Earthquake, the 2016 Kumamoto Earthquake, and the 2018 Hokkaido Eastern Iburi Earthquake are analyzed with four different InSAR methods: differential InSAR analysis, persistent scatter time series analysis, coherence-based method and pixel (dense) offset analysis. The co-seismic displacement results are converted into vertical and horizontal displacements to compare with field observations from reconnaissances right after these earthquakes. After comparing the effectiveness of the different methods and different satellites at each site, an InSAR workflow for analysing earthquake-induced liquefaction observations is presented.