No abstract

Predicting the potential depth of wave-induced liquefaction is of vital importance of hazard assessment. Prevention and mitigation of liquefaction hazard can be carried out after predicting the liquefaction potential depth. Sediment geotechnical parameters were summarized from historical engineering projects and calculated liquefaction depth with data of actually measured water depth, sediment types and different wave recurrence periods. Then the study area was divided into different liquefaction zones. Results showed that both liquefaction range and depth increased with wave return period, which remained steady when increased to a certain extent. The calculated liquefaction depth was compared with the thickness of liquefied layers, which obtained by geophysical methods. Result shows that the average difference was about 1m, which means the method of liquefaction depth calculation was applicable in this study area. At last, in order to reduce harm of liquefaction, potential maximum liquefaction depth (PMLD) is put forward. PMLD is the smaller one between the thickness of seabed liquefiable sediment and the maximum calculated liquefaction depth. The calculation and partition of liquefaction depth can be more accurate after considering PMLD.

A submarine landslide is a well-known geohazard that can cause significant damage to offshore engineering facilities. Most standard predicting and mapping methods require expert knowledge, supervision, and fieldwork. In this research, the main objective was to analyze the potential of unsupervised machine learning methods and compare the performance of three different unsupervised machine learning models (k-means, spectral clustering, and hierarchical clustering) in modeling the susceptibility of the submarine landslide. Nine groups of geological factors were selected as the input parameters, which were obtained through field surveys. To estimate submarine landslide susceptibility, all input factors were separated into three or four groups based on data features and environmental variables. Finally, the goodness-of-fit and accuracy of models were validated with both internal metrics (Calinski–Harabasz index, silhouette index, and Davies–Bouldin index) and external metrics (existing landslide distribution, hydrodynamic distribution, and liquefication distribution). The findings of k-means, spectral clustering, and hierarchical clustering performed commendably and accurately in forecasting the submarine landslide susceptibility. Spectral clustering has the greatest congruence with environmental geology parameters. Therefore, the unsupervised machine learning model can be used in submarine-landslide-predicting studies, and the spectral clustering method performed best. Furthermore, machine learning can improve submarine landslide mapping in the future with the development of models and the extension of geological data related to submarine landslides.