Evaluation of landslide susceptibility at Guilin City， Guangxi Zhuang Autonomous Region based on hyperparameter optimization and shapely additive explanations

Liu Zihan; Su Huiwei; Zeng Mingyue

doi:10.13961/j.cnki.stbctb.2025.06.036

您当前的位置：

首页 >

文章列表页 >

Evaluation of landslide susceptibility at Guilin City， Guangxi Zhuang Autonomous Region based on hyperparameter optimization and shapely additive explanations

Soil and Water Conservation Monitoring and Applied Technology

- Evaluation of landslide susceptibility at Guilin City， Guangxi Zhuang Autonomous Region based on hyperparameter optimization and shapely additive explanations
- Bulletin of Soil and Water Conservation Vol. 45, Issue 6, Pages: 190-201(2025)
- 作者机构：
  
  桂林旅游学院广西文化和旅游智慧技术重点实验室，广西桂林 541006
- 作者简介：
- 基金信息：
- DOI：10.13961/j.cnki.stbctb.2025.06.036
  CLC： P642.22
- CSTR：32312.14.stbctb.2025.06.036
- Received：04 July 2025，
  
  Revised：2025-08-23，
  
  Published：10 December 2025
- 稿件说明：
移动端阅览
刘子晗, 苏会卫, 曾铭玥.基于超参数优化与SHAP的广西桂林市滑坡灾害易发性评价[J].水土保持通报,2025,45(6):190-201.

Liu Zihan, Su Huiwei, Zeng Mingyue. Evaluation of landslide susceptibility at Guilin City， Guangxi Zhuang Autonomous Region based on hyperparameter optimization and shapely additive explanations [J]. Bulletin of Soil and Water Conservation,2025,45(6):190-201.
刘子晗, 苏会卫, 曾铭玥.基于超参数优化与SHAP的广西桂林市滑坡灾害易发性评价[J].水土保持通报,2025,45(6):190-201. DOI： 10.13961/j.cnki.stbctb.2025.06.036. CSTR： 32312.14.stbctb.2025.06.036.

Liu Zihan, Su Huiwei, Zeng Mingyue. Evaluation of landslide susceptibility at Guilin City， Guangxi Zhuang Autonomous Region based on hyperparameter optimization and shapely additive explanations [J]. Bulletin of Soil and Water Conservation,2025,45(6):190-201. DOI： 10.13961/j.cnki.stbctb.2025.06.036. CSTR： 32312.14.stbctb.2025.06.036.

摘要

目的

研究构建整合喀斯特地貌特征的滑坡易感性评价框架，揭示主导因素和相互作用机制，为喀斯特地貌区滑坡风险的精准控制提供理论依据。

方法

以广西桂林市为研究区，创新性融合岩溶特征，构建包含10个关键因子的地理空间数据库。采用逻辑回归（LR）、支持向量机（SVM）、随机森林（RF）、梯度提升机（GBM）和XGBoost5种机器学习模型进行滑坡易发性建模，利用网格搜索优化超参数，并通过准确率和AUC等指标评估模型性能。同时，引入SHAP算法量化因子贡献及交互效应。

结果

集成学习模型（RF，XGBoost）性能最佳，RF和XGBoost的准确率（0.85，0.84）和AUC（0.93，0.92）最高。所有模型呈现“面积递减—灾害密度激增”规律，RF在极高风险区的滑坡密度最高，为0.164例/km²。SHAP分析显示，岩溶因子中的地下水位是多数模型中最具影响力的因子，与NDVI、距河流距离、土壤类型存在一定交互效应，集成模型在特征解释上的一致性较高。

结论

随机森林等集成模型结合SHAP框架，可显著提升喀斯特地貌区滑坡易发性制图的精度与可解释性，证实了地下水位与土壤类型的协同致灾机制。

Abstract

Objective

A landslide susceptibility evaluation framework that integrates the characteristics of Karst landforms was constructed， in order to reveal the dominant factors and their interaction mechanisms， and provide a theoretical basis for the precise control of landslide risks in Karst landform areas.

Methods

The study area was Guilin City， Guangxi Zhuang Autonomous Region. We innovatively integrated Karst characteristics to construct a geospatial database containing nine key factors. Five machine learning models-logistic regression （LR）， support vector machine （SVM）， random forest （RF）， gradient boosting machine （GBM）， and XGBoost-were used for landslide susceptibility modeling. Hyperparameters were optimized via a grid search， and the model performance was evaluated using accuracy and area under the receiver operating characteristic curve （AUC） metrics. The shapley additive explanations （SHAP） algorithm was applied to quantify the contributions of the various factors and their interactive effects.

Results

The ensemble models RF and XGBoost exhibited the best performance， achieving the highest accuracy （0.85 and 0.84， respectively） and AUC values （0.93 and 0.92， respectively）. All models showed a trend of a sharp increase in landslide density with decreasing land area. The RF model yielded the highest landslide density in the extremely high-risk zones （0.164 events/km

）. The SHAP analysis indicated that the groundwater content was the most influential Karst factor in most models. It also revealed the interactive effects of the normalized difference vegetation index， distance to rivers， and soil type. The ensemble models exhibited high consistency in terms of feature interpretation.

Conclusion

Integrating ensemble models such as RF with the SHAP framework significantly improves the accuracy and interpretability of landslide susceptibility mapping in Karst regions. The results of this study confirm the synergistic disaster-forming mechanism between the groundwater content and soil type.

关键词

Keywords

references

Sun Liying ， Ma Bingjuan ， Pei Liang ， et al . The relationship of human activities and rainfall-induced landslide and debris flow hazards in central China ［J］. Natural Hazards ， 2021 ， 107 （ 1 ）： 147 - 169 .

吴淼锐，杨美雪，钟艺倩，等 . 桂林喀斯特世界自然遗产地峰丛区传统村落景观演变与植物景观特征分析［J］. 广西科学， 2024 ， 31 （ 6 ）： 1233 - 1247 .

Wu Miaorui ， Yang Meixue ， Zhong Yiqian ， et al . Analysis of landscape evolution and plant landscape characteristics of traditional villages in peak cluster area， Guilin karst world natural heritage site ［J］. Guangxi Sciences ， 2024 ， 31 （ 6 ）： 1233 - 1247 .

Zhao Zeyang ， Chen Tao ， Dou Jie ， et al . Landslide susceptibility mapping considering landslide local-global features based on CNN and transformer ［J］. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing ， 2024 ， 17 ： 7475 - 7489 .

Jin Bijing ， Zeng Taorui ， Yin Kunlong ， et al . Dynamic landslide susceptibility mapping based on the PS-InSAR deformation intensity ［J］. Environmental Science and Pollution Research ， 2024 ， 31 （ 5 ）： 7872 - 7888 .

吴雅睿，娄春辉，侯龙君，等 . 基于信息量与机器学习耦合模型的滑坡易发性评价对比分析［J］. 西安科技大学学报， 2024 ， 44 （ 6 ）： 1140 - 1153 .

Wu Yarui ， Lou Chunhui ， Hou Longjun ， et al . Comparative analysis of landslide susceptibility evaluation based on the coupling model of information quantity and machine learning ［J］. Journal of Xi’an University of Science and Technology ， 2024 ， 44 （ 6 ）： 1140 - 1153 .

孙才，铁永波，宁志杰，等 . 基于频率比-支持向量机耦合模型的四川省喜德县滑坡易发性评价［J］. 沉积与特提斯地质， 2024 ， 44 （ 3 ）： 547 - 559 .

Sun Cai ， Tie Yongbo ， Ning Zhijie ， et al . Landslide susceptibility mapping in Xide County， Sichuan Province based on frequency ratio-support vector machine coupling model ［J］. Sedimentary Geology and Tethyan Geology ， 2024 ， 44 （ 3 ）： 547 - 559 .

穆成林，周海鹏，阳博，等 . 基于信息量-地理逻辑回归耦合模型的滑坡灾害易发性评价：以四川省德昌县为例［J］. 地理科学， 2025 ， 45 （ 7 ）： 1575 - 1589 .

Mu Chenglin ， Zhou Haipeng ， Yang Bo ， et al . Evaluation of landslide hazard susceptibility based on information value-coupled geographical logistic regression coupling model： A case of Dechang County in Sichuan Province ［J］. Geographical Science ， 2025 ， 45 （ 7 ）： 1575 - 1589 .

Chang Zhilu ， Huang Jinsong ， Huang Faming ， et al . Uncertainty analysis of non-landslide sample selection in landslide susceptibility prediction using slope unit-based machine learning models ［J］. Gondwana Research ， 2023 ， 117 ： 307 - 320 .

Rodrigues S G ， Silva M M ， Alencar M H . A proposal for an approach to mapping susceptibility to landslides using natural language processing and machine learning ［J］. Landslides ， 2021 ， 18 （ 7 ）： 2515 - 2529 .

吴润泽，胡旭东，梅红波，等 . 基于随机森林的滑坡空间易发性评价：以三峡库区湖北段为例［J］. 地球科学， 2021 ， 46 （ 1 ）： 321 - 330 .

Wu Runze ， Hu Xudong ， Mei Hongbo ， et al . Spatial susceptibility assessment of landslides based on random forest：A case study from Hubei section in the Three Gorges reservoir area ［J］. Earth Science ， 2021 ， 46 （ 1 ）： 321 - 330 .

李文杰，巨能攀，王栋，等 . 融合SVM-RFE与层次分析-信息量模型的地质灾害易发性评价［J］. 自然灾害学报， 2025 ， 34 （ 3 ）： 99 - 109 .

Li Wenjie ， Ju Nengpan ， Wang Dong ， et al . Geological disaster susceptibility evaluation based on SVM-RFE and AHP-information model ［J］. Journal of Natural Disasters ， 2025 ， 34 （ 3 ）： 99 - 109 .

Zhang Junyi ， Ma Xianglong ， Zhang Jialan ， et al . Insights into geospatial heterogeneity of landslide susceptibility based on the SHAP-XGBoost model ［J］. Journal of Environmental Management ， 2023 ， 332 ： 117357 .

ŞAHiN E ， Arslan N N ， Özdemir D . Unlocking the black box：An in-depth review on interpretability， explainability， and reliability in deep learning ［J］. Neural Computing and Applications ， 2025 ， 37 （ 2 ）： 859 - 965 .

张虹，辜庆渝，孙诚彬，等 . 基于可解释性机器学习的丘陵缓坡地区滑坡易发性区划研究［J］. 重庆师范大学学报（自然科学版）， 2022 ， 39 （ 3 ）： 78 - 92 .

Zhang Hong ， Gu Qingyu ， Sun Chengbin ， et al . Landslide susceptibility mapping in hilly and gentle slope region based on interpretable machine learning ［J］. Journal of Chongqing Normal University （Natural Science）， 2022 ， 39 （ 3 ）： 78 - 92 .

Sun Deliang ， Wang Jing ， Wen Haijia ， et al . Insights into landslide susceptibility in different karst erosion landforms based on interpretable machine learning ［J］. Earth Surface Processes and Landforms ， 2024 ， 49 （ 3 ）： 1006 - 1027 .

Zhu Yanping ， Taffese W Z ， Chen Genda . Enhancing FRP-concrete interface bearing capacity prediction with explainable machine learning：A feature engineering approach and SHAP analysis ［J］. Engineering Structures ， 2024 ， 319 ： 118831 .

罗路广，裴向军，黄润秋，等 . GIS支持下CF与Logistic回归模型耦合的九寨沟景区滑坡易发性评价［J］. 工程地质学报， 2021 ， 29 （ 2 ）： 526 - 535 .

Luo Luguang ， Pei Xiangjun ， Huang Runqiu ， et al . Landslide susceptibility assessment in Jiuzhaigou scenic area with GIS based on certainty factor and logistic regression model ［J］. Journal of Engineering Geology ， 2021 ， 29 （ 2 ）： 526 - 535 .

Wu Huaying ， Jiang Zhongcheng ， Huang Qibo ， et al . Geochemistry of weathering cover and the main influencing factors in karst area of Guilin， southwest China ［J］. Water ， 2023 ， 15 （ 16 ）： 2944 .

Zhao Fancheng ， Miao Fasheng ， Wu Yiping ， et al . Refined landslide susceptibility mapping in township area using ensemble machine learning method under dataset replenishment strategy ［J］. Gondwana Research ， 2024 ， 131 ： 20 - 37 .

Gong Peng ， Wang Jie ， Huang Huabing . Stable classification with limited samples in global land cover mapping：Theory and experiments ［J］. Science Bulletin ， 2024 ， 69 （ 12 ）： 1862 - 1865 .

Li Xiangqian ， Peng Qiongyan ， Shen Ruoque ， et al . Long-term reconstructed vegetation index dataset in China from fused MODIS and Landsat data ［J］. Scientific Data ， 2025 ， 12 ： 152 .

Chen Yu . Spatial prediction and mapping of landslide susceptibility using machine learning models ［J］. Natural Hazards ， 2025 ， 121 （ 7 ）： 8367 - 8385 .

Ahmad G N ， Fatima H ， Ullah S ， et al . Efficient medical diagnosis of human heart diseases using machine learning techniques with and without GridSearchCV ［J］. IEEE Access ， 2022 ， 10 ： 80151 - 80173 .

Lu Jiangang ， He Yi ， Zhang Lifeng ， et al . Ensemble learning landslide susceptibility assessment with optimized non-landslide samples selection ［J］. Geomatics， Natural Hazards and Risk ， 2024 ， 15 ： 2378176 .

Zhou Xinzhi ， Wen Haijia ， Li Ziwei ， et al . An interpretable model for the susceptibility of rainfall-induced shallow landslides based on SHAP and XGBoost ［J］. Geocarto International ， 2022 ， 37 （ 26 ）： 13419 - 13450 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Landslide susceptibility evaluation and driving force analysis for Jiuzhaigou scenic area based on explainable machine learning

Comparative analysis of landslide susceptibility based on different evaluation units——A case study of Ledu District in Haidong City， Qinghai Province

Estimation of soil moisture and organic matter content in saline-alkaline farmland of Hetao Plain based on interpretable machine learning

Landslide Susceptibility Evaluation Based on Coupling of GBDT-LR Model and Information Model

Remote Sensing Estimation of Forest Volume in Typical Karst Mountainous Areas

Related Author

Zhao Peng

Pei Zengli

Lin Yongming

Wang Daojie

He Songtang

Shen Zhenhong

Li Huamin

Li Yuanyuan

Related Institution

University of the Chinese Academy of Sciences

Key Laboratory of Mountain Hazards and;Engineering Safety， Institute of Mountain Hazards and Environment， Chinese Academy of Sciences and Ministry of;Water Resources

Key Laboratory for Forest Ecosystem Process and;Management of Fujian Province

College of Juncao Science and Ecology， Fujian Agriculture and Forestry;University

School of Energy， Henan Polytechnic University

⁰