Wang Youxiang, Kang Liangqiang, Wang Jianping
Shaanxi institute of Geological Exploration, Sinochem general administration of geology and mines, Xi'an, 710000, China.
Heliyon. 2024 Sep 3;10(17):e37352. doi: 10.1016/j.heliyon.2024.e37352. eCollection 2024 Sep 15.
Landslide risk mapping can be an effective reference for disaster mitigation and land use planning, but the modelling process involves multidisciplinary knowledge which leads to its complexity. In this study, Jiaxian County in Shaanxi Province on the Loess Plateau of China, served as the study area, primarily characterized by Quaternary loess-covered geomorphology, with an average rainfall of about 400 mm annually. Soil erosion and human engineering activities have contributed to significant slope failures, posing threats to local residents and infrastructure. A reasonable inventory of landslides in the region was established by field survey combined with aerial imagery, allowing for characterization of their development and spatial distribution. Nine thematic maps related to landslide occurring and three vulnerability maps were prepared as influencing factors for landslide risk assessment. Subsequently, landslide susceptibility and hazard were evaluated using a kernel extreme learning machine (KELM) and information value (IV) model, followed by map validation. A decision table was then employed to generate the landslide risk map. The results of landslide hazard mapping showed that the historical landslide events were mainly developed in the central part of the study area, particularly concentrated near the developed river network. Integration of overall risk elements suggested that landslide risks in the study area were generally at a low level. Besides, a total of 0.25 % and 2.05 % of the areas were classified as having very high and high landslide risk levels, respectively, where 65.11 % of inventory landslides occurred. Therefore, the proposed procedure is a valuable tool for assessing landslide risk in Jiaxian Country.
滑坡风险制图可为减灾和土地利用规划提供有效的参考，但建模过程涉及多学科知识，导致其具有复杂性。本研究以中国黄土高原陕西省佳县为研究区域，其主要特征为第四纪黄土覆盖地貌，年平均降雨量约400毫米。土壤侵蚀和人类工程活动导致了大量的边坡失稳，对当地居民和基础设施构成威胁。通过实地调查结合航空影像，建立了该地区合理的滑坡清单，从而能够对滑坡的发育和空间分布进行特征描述。编制了与滑坡发生相关的9张专题地图和3张脆弱性地图，作为滑坡风险评估的影响因素。随后，利用核极限学习机（KELM）和信息值（IV）模型对滑坡易发性和危险性进行了评估，随后进行了地图验证。然后采用决策表生成滑坡风险图。滑坡危险性制图结果表明，历史滑坡事件主要发生在研究区域的中部，尤其集中在发育的河网附近。综合总体风险要素表明，研究区域的滑坡风险总体处于较低水平。此外，分别有0.25%和2.05%的区域被归类为具有非常高和高的滑坡风险等级，其中65.11%的已编目滑坡发生在这些区域。因此，所提出的程序是评估佳县滑坡风险的一个有价值工具。
