Performance evaluation for four GIS-based models purposed to predict and map landslide susceptibility: A case study at a World Heritage site in Southwest China
Article
Article Title | Performance evaluation for four GIS-based models purposed to predict and map landslide susceptibility: A case study at a World Heritage site in Southwest China |
---|---|
ERA Journal ID | 1715 |
Article Category | Article |
Authors | Jiao, Yuanmei, Zhao, Dongmei, Ding, Yinping, Liu, Yan, Xu, Qiue, Qiu, Yingmei, Liu, Chengjing, Liu, Zhilin, Zha, Zhiqin and Li, Rong |
Journal Title | Catena |
Journal Citation | 183, pp. 1-15 |
Article Number | 104221 |
Number of Pages | 15 |
Year | Dec 2019 |
Publisher | Elsevier |
Place of Publication | Netherlands |
ISSN | 0341-8162 |
1872-6887 | |
Digital Object Identifier (DOI) | https://doi.org/10.1016/j.catena.2019.104221 |
Web Address (URL) | https://www.sciencedirect.com/science/article/pii/S0341816219303637 |
Abstract | Landslide susceptibility mapping is a prerequisite for preventing and mitigating hazardous risks, especially in landslide-prone mountainous areas. This study applies two well-known models – the information value model (IVM) and maximum entropy (MaxEnt) model, and two additional models – a biological climatic (Bioclim) model and a mean distance (Domain) model, to evaluate the landslide susceptibility of the Honghe Hani Rice Terraces, a World Heritage site in Yuanyang County of Southwest China. A spatial dataset comprising 235 historical landslide locations were used in the models at a training-to-testing data ratio of 75:25. Fifteen commonly used environmental, geological, and meteorological factors, as well as three new factors including proximity to settlements, human activity intensity, and the multi-year average temperature were selected as the impacting factors to landslide occurance. Model validation and comparison were performed using the area under the receiver operating characteristic curve, and several statistical indices (sensibility, specificity, accuracy, precision, recall, true skill statistic or TSS, F-measure). The results show that the MaxEnt model yields the best overall performance for landslide susceptibility analysis, followed by the Domain model, IVM, and the Bioclim model. The jackknife test result shows that the three most important factors contributing to landslide hazard include proximity to roads, annual rainfall, and proximity to settlements. Furthermore, only a small proportion of areas positioned along roads were identified to be at high or very high risk for landslides given the fact the study site is located in a remote and mountainous region with under-developed economy in southwest China. Findings from this study can be used to facilitate landslide risk mitigation and the sustainable conservation of the Honghe Hani Rice Terrace World Heritage site or other similar areas. |
Keywords | Landslide susceptibility mapping; Information value model (IVM); Maximum entropy model (MaxEnt); Biological climatic model (Bioclim); Mean distance model (Domain); Honghe Hani Rice Terrace World Heritage |
Public Notes | Files associated with this item cannot be displayed due to copyright restrictions. |
Byline Affiliations | Yunnan University, China |
University of Queensland | |
Library Services |
https://research.usq.edu.au/item/w8z84/performance-evaluation-for-four-gis-based-models-purposed-to-predict-and-map-landslide-susceptibility-a-case-study-at-a-world-heritage-site-in-southwest-china
36
total views0
total downloads3
views this month0
downloads this month