Submarine Landslide Susceptibility and Spatial Distribution Using Different Unsupervised Machine Learning Models

被引:10
作者
Du, Xing [1 ,2 ]
Sun, Yongfu [3 ]
Song, Yupeng [1 ]
Xiu, Zongxiang [1 ]
Su, Zhiming [1 ]
机构
[1] MNR, Inst Oceanog 1, Qingdao 266061, Peoples R China
[2] Ocean Univ China, Coll Environm Sci & Engn, Qingdao 266100, Peoples R China
[3] Natl Deep Sea Ctr, Qingdao 266237, Peoples R China
来源
APPLIED SCIENCES-BASEL | 2022年 / 12卷 / 20期
基金
中国国家自然科学基金;
关键词
submarine landslide; machine learning; hazard susceptibility; spatial distribution; SLOPE STABILITY ANALYSIS; PREDICTION;
D O I
10.3390/app122010544
中图分类号
O6 [化学];
学科分类号
0703 ;
摘要
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.
引用
收藏
页数:17
相关论文
共 33 条
[1]  
Aranganayagi S., 2007, ICCIMA 2007 INT C CO, VIII, P13, DOI DOI 10.1109/ICCIMA.2007.328
[2]   Spatial prediction of landslides using a hybrid machine learning approach based on Random Subspace and Classification and Regression Trees [J].
Binh Thai Pham ;
Prakash, Indra ;
Dieu Tien Bui .
GEOMORPHOLOGY, 2018, 303 :256-270
[3]   The Kinematics of a Debris Avalanche on the Sumatra Margin [J].
Bradshaw, A. S. ;
Tappin, D. R. ;
Rugg, D. .
SUBMARINE MASS MOVEMENTS AND THEIR CONSEQUENCES, 2010, 28 :117-+
[4]   Landslide spatial modeling: Introducing new ensembles of ANN, MaxEnt, and SVM machine learning techniques [J].
Chen, Wei ;
Pourghasemi, Hamid Reza ;
Kornejady, Aiding ;
Zhang, Ning .
GEODERMA, 2017, 305 :314-327
[5]   CLUSTER SEPARATION MEASURE [J].
DAVIES, DL ;
BOULDIN, DW .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1979, 1 (02) :224-227
[6]   Probabilistic slope stability analysis for practice [J].
El-Ramly, H ;
Morgenstern, NR ;
Cruden, DM .
CANADIAN GEOTECHNICAL JOURNAL, 2002, 39 (03) :665-683
[7]   Geology and tsunamigenic potential of submarine landslides in Santa Barbara Channel, Southern California [J].
Fisher, MA ;
Normark, WR ;
Greene, HG ;
Lee, HJ ;
Sliter, RW .
MARINE GEOLOGY, 2005, 224 (1-4) :1-22
[8]  
[甘华阳 Gan Huayang], 2004, [防灾减灾工程学报, Journal of disaster prevention and mitigation engineering], V24, P177
[9]   Evaluation of Different Machine Learning Methods and Deep-Learning Convolutional Neural Networks for Landslide Detection [J].
Ghorbanzadeh, Omid ;
Blaschke, Thomas ;
Gholamnia, Khalil ;
Meena, Sansar Raj ;
Tiede, Dirk ;
Aryal, Jagannath .
REMOTE SENSING, 2019, 11 (02)
[10]   Slope stability analysis by finite elements [J].
Griffiths, DV ;
Lane, PA .
GEOTECHNIQUE, 1999, 49 (03) :387-403