Susceptibility assessment of landslides triggered by earthquakes in the Western Sichuan Plateau

Abstract Seismic landslides are the most harmful natural events in mountainous areas worldwide. They are secondary disasters triggered by earthquakes and can cause a great number of casualties and significant damage to infrastructure. Three areas, Wenchuan, Lushan and Jiuzhaigou, which are highly prone to seismic landslides, were selected for this study. The aim of this study is to compare the prediction accuracy and spatial generalization ability of the logistic regression (LR) and random forest (RF) models in seismic landslide susceptibility mapping. First, using the LR and RF techniques, the susceptibility models of seismic landslides were developed in Wenchuan and Lushan based on thirteen influencing factors. Then, the accuracy of the susceptibility mapping was evaluated by the area under the curve (AUC) values of the receiver operating characteristic (ROC) curves. The results showed that both RF models have excellent prediction capability and strong robustness over large areas compared with the LR models. The AUC values of the RF and LR models were 0.811 vs. 0.946 and 0.905 vs. 0.969 in Wenchuan and Lushan, respectively. Second, we used the seismic landslides of Wenchuan and Lushan to develop the better model (RF), and then applied the developed RF model to produce the landslide susceptibility mapping of Jiuzhaigou County. The prediction accuracy for Jiuzhaigou dropped to 0.704. The results show that the RF model developed in Wenchuan and Lushan has robustness and spatial generalization ability over large areas. However, many landslides that had not been identified by the model, were located in the scenic area of Jiuzhaigou, a popular tourist destination. This suggests that intensive human activities could increase the landslide susceptibility. Thus, we added a new factor related to tourism to produce an improved RF model. The prediction accuracy of the altered model improved significantly, with the AUC value rising from 0.704 to 0.987. Thus, human engineering activities have a significant impact on landslides. Our study indicates that human engineering activities can increase the likelihood of landslides; hence, human engineering activities related to tourism should not be neglected in landslide susceptibility research. Appropriate disaster prevention and mitigation measures should be taken for the twelve geoparks of western China with peak ground acceleration ≥0.2 g.

[1]  C. Bonnard,et al.  Hazard and risk assessment of earthquake-induced landslides—case study , 2017, Landslides.

[2]  A. Ozdemir,et al.  A comparative study of frequency ratio, weights of evidence and logistic regression methods for landslide susceptibility mapping: Sultan Mountains, SW Turkey , 2013 .

[3]  L. Ayalew,et al.  Landslides in Sado Island of Japan: Part II. GIS-based susceptibility mapping with comparisons of results from two methods and verifications , 2005 .

[4]  Kellie J. Archer,et al.  Empirical characterization of random forest variable importance measures , 2008, Comput. Stat. Data Anal..

[5]  Peizhen Zhang.  A review on active tectonics and deep crustal processes of the Western Sichuan region, eastern margin of the Tibetan Plateau , 2013 .

[6]  K. Seto,et al.  The Vegetation Adjusted NTL Urban Index: A new approach to reduce saturation and increase variation in nighttime luminosity , 2013 .

[7]  Zohre Sadat Pourtaghi,et al.  Landslide susceptibility assessment in Lianhua County (China); a comparison between a random forest data mining technique and bivariate and multivariate statistical models , 2016 .

[8]  A. Trigila,et al.  Comparison of Logistic Regression and Random Forests techniques for shallow landslide susceptibility assessment in Giampilieri (NE Sicily, Italy) , 2015 .

[9]  Lewis A. Owen,et al.  GIS-based landslide susceptibility mapping for the 2005 Kashmir earthquake region , 2008 .

[10]  H. Pourghasemi,et al.  Random forests and evidential belief function-based landslide susceptibility assessment in Western Mazandaran Province, Iran , 2016, Environmental Earth Sciences.

[11]  Comparison of statistical methods and multi-time validation for the determination of the shallow landslide rainfall thresholds , 2018, Landslides.

[12]  Meng Xiangang,et al.  The Ms7.0 Lushan earthquake and the activity of the Longmenshan fault zone , 2013 .

[13]  X. Pei,et al.  On the initiation and movement mechanisms of a catastrophic landslide triggered by the 2008 Wenchuan (Ms 8.0) earthquake in the epicenter area , 2017, Landslides.

[14]  Rui Liu,et al.  A novel genetic algorithm for optimization of conditioning factors in shallow translational landslides and susceptibility mapping , 2017, Arabian Journal of Geosciences.

[15]  S. Bai,et al.  GIS-based logistic regression for landslide susceptibility mapping of the Zhongxian segment in the Three Gorges area, China , 2010 .

[16]  Julian J. Bommer,et al.  Earthquake losses due to ground failure , 2004 .

[17]  Dalia Kirschbaum,et al.  Modeling landslide susceptibility over large regions with fuzzy overlay , 2016, Landslides.

[18]  Runqiu Huang,et al.  The characteristics and failure mechanism of the largest landslide triggered by the Wenchuan earthquake, May 12, 2008, China , 2012, Landslides.

[19]  Ling Peng,et al.  Quantitative risk analysis for landslides: the case of the Three Gorges area, China , 2015, Landslides.

[20]  Chong Xu,et al.  Landslides triggered by slipping-fault-generated earthquake on a plateau: an example of the 14 April 2010, Ms 7.1, Yushu, China earthquake , 2013, Landslides.

[21]  Anne-Laure Boulesteix,et al.  AUC-RF: A New Strategy for Genomic Profiling with Random Forest , 2011, Human Heredity.

[22]  Chong Xu,et al.  Three (nearly) complete inventories of landslides triggered by the May 12, 2008 Wenchuan Mw 7.9 earthquake of China and their spatial distribution statistical analysis , 2014, Landslides.

[23]  Netra R. Regmi,et al.  A comparison of logistic regression-based models of susceptibility to landslides in western Colorado, USA , 2014, Landslides.

[24]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[25]  Ping Sun,et al.  Landslide hazards triggered by the 2008 Wenchuan earthquake, Sichuan, China , 2009 .

[26]  H. Pourghasemi,et al.  Landslide susceptibility mapping by binary logistic regression, analytical hierarchy process, and statistical index models and assessment of their performances , 2013, Natural Hazards.

[27]  Qing Zhou,et al.  Earthquake-triggered landslides in southwest China , 2012 .

[28]  M. M. Wang,et al.  Brief Communication: Landslides triggered by the M s = 7.0 Lushan earthquake, China , 2013 .

[29]  E. Yesilnacar,et al.  Landslide susceptibility mapping : A comparison of logistic regression and neural networks methods in a medium scale study, Hendek Region (Turkey) , 2005 .

[30]  C. F. Lee,et al.  Assessment of landslide susceptibility on the natural terrain of Lantau Island, Hong Kong , 2001 .

[31]  Xi-wei Xu,et al.  The characteristics of the landslides triggered by the Wenchuan Ms 8.0 earthquake from Anxian to Beichuan , 2010 .

[32]  H. A. Nefeslioglu,et al.  Susceptibility assessments of shallow earthflows triggered by heavy rainfall at three catchments by logistic regression analyses , 2005 .

[33]  D. M. Duc,et al.  Landslide susceptibility mapping by combining the analytical hierarchy process and weighted linear combination methods: a case study in the upper Lo River catchment (Vietnam) , 2016, Landslides.

[34]  Yurui Li,et al.  Accelerated Urban Expansion in Lhasa City and the Implications for Sustainable Development in a Plateau City , 2017 .

[35]  Hamid Reza Pourghasemi,et al.  Erratum to: Landslide susceptibility mapping using random forest, boosted regression tree, classification and regression tree, and general linear models and comparison of their performance at Wadi Tayyah Basin, Asir Region, Saudi Arabia , 2016, Landslides.

[36]  P. Cui,et al.  Seismogenic fault and topography control on the spatial patterns of landslides triggered by the 2017 Jiuzhaigou earthquake , 2018, Journal of Mountain Science.

[37]  B. Pradhan,et al.  A comparative study of logistic model tree, random forest, and classification and regression tree models for spatial prediction of landslide susceptibility , 2017 .

[38]  Chong Xu,et al.  Database and spatial distribution of landslides triggered by the Lushan, China Mw 6.6 earthquake of 20 April 2013 , 2015 .

[39]  J. Malet,et al.  Recommendations for the quantitative analysis of landslide risk , 2013, Bulletin of Engineering Geology and the Environment.

[40]  E. Harp,et al.  Interpretation of earthquake-induced landslides triggered by the 12 May 2008, M7.9 Wenchuan earthquake in the Beichuan area, Sichuan Province, China using satellite imagery and Google Earth , 2009 .

[41]  Anita Simic,et al.  Landslide detection and susceptibility mapping using LiDAR and an artificial neural network approach: a case study in the Cuyahoga Valley National Park, Ohio , 2016, Landslides.

[42]  Zohre Sadat Pourtaghi,et al.  Landslide susceptibility mapping using random forest, boosted regression tree, classification and regression tree, and general linear models and comparison of their performance at Wadi Tayyah Basin, Asir Region, Saudi Arabia , 2015, Landslides.

[43]  D. Bui,et al.  Spatial prediction of landslides using a hybrid machine learning approach based on Random Subspace and Classification and Regression Trees , 2018 .

[44]  Steeve D. Côté,et al.  Spring Normalized Difference Vegetation Index (NDVI) predicts annual variation in timing of peak faecal crude protein in mountain ungulates , 2009 .