A physics-informed machine learning framework for landslide susceptibility assessment based on probabilistic physically-based model
CUI Hongzhi1, 2, 3, PEI Te4, JI Jian1, 5*
(1. State Laboratory of Intelligent Deep Metal Mining and Equipment, Shaoxing University, Shaoxing, Zhejiang 312000, China;
2. Zhejiang Key Laboratory of Rock Mechanics and Geohazards, Shaoxing University, Shaoxing, Zhejiang 312000, China;
3. School of Civil Engineering and Transportation, Yangzhou University, Yangzhou, Jiangsu 225127; 4. Department of Civil Engineering, The State University of New York at Stony Brook, New York 11794, America; 5. Geotechnical Research
Institute, Hohai University, Nanjing, Jiangsu 210024, China)
Abstract：Landslide susceptibility assessment (LSA) is critical for predicting potential landslides at regional scales. Existing machine learning based data driven models are highly dependent on data quality and generally lack physical interpretability, which often leads to limited generalization ability and physical inconsistency under complex geological and hydrological conditions. In contrast, physically-based models in mechanical principles can characterize the relationship between rainfall infiltration and slope stability, but their application at the regional scale remains constrained by spatial heterogeneity of parameters and limited model applicability. To address these challenges, this study proposes a probabilistic physics-informed machine learning (PIML) framework that explicitly incorporates geotechnical domain knowledge into the model training process. A simplified transient infiltration model and the first-order reliability method (PRL-STIM) are utilized to compute the factor of safety and failure probability, respectively. These physics-based outputs are integrated into a composite loss function designed to ensure both physical and risk consistency, thereby guiding the training of a neural network. The framework is validated using a rainfall-induced shallow landslide event that occurred in Niangniangba, Gansu Province, China, in 2013. A high-quality dataset is compiled, and a spatial cross-validation strategy is employed to assess the model′s generalization ability and predictive uncertainty in previously unseen areas. Results from five-fold spatial cross-validation indicate that the PIML model enhances the average AUC by 13.6% and reduces the scientific inconsistency index (SI) by 88.6% compared to the baseline model. These improvements demonstrate the proposed model′s enhanced robustness, physical consistency, and interpretability for regional LSA.
[1] 宋 健，陆朱汐，谢华威，等. 基于极限平衡原理的地震边坡浅层和深层耦合滑移分析[J]. 岩土工程学报，2023，45(6)：1 141–1 150. (SONG Jian，LU Zhuxi，XIE Huawei，et al. Analysis of coupled shallow and deep sliding of slopes induced by earthquake based on limit equilibrium method[J]. Chinese Journal of Geotechnical Engineering，2023，45(6)：1 141–1 150.(in Chinese))
[2] EMBERSON R，KIRSCHBAUM D，STANLEY T. Global connections between El Nino and landslide impacts[J]. Nature Communications，2021，12(1)：2262.
[3] LÜ Q，WU J，LIU Z，et al. The Fuyang shallow landslides triggered by an extreme rainstorm on 22 July 2023 in Zhejiang，China[J]. Landslides，2024，21(11)：2 725–2 740.
[4] 杨仲康，张世殊，邓建辉，等. 斜坡单元尺寸和滑坡取样方式对区域滑坡易发性评价的影响[J]. 岩石力学与工程学报，2025，44(3)：602–617.(YANG Zhongkang，ZHANG Shishu，DENG Jianhui，et al. The impact of slope unit scale and landslide sampling methods on regional landslide susceptibility assessment[J]. Chinese Journal of Rock Mechanics and Engineering，2025，44(3)：602–617.(in Chinese))
[5] 曾韬睿，邬礼扬，金必晶，等. 基于stacking集成策略和SBAS-InSAR的滑坡动态易发性制图[J]. 岩石力学与工程学报，2023，42(9)：2 266–2 282.(ZENG Taorui，WU Liyang，JIN Bijing，et al. Landslide dynamic susceptibility mapping based on stacking ensemble strategy and SBAS-InSAR[J]. Chinese Journal of Rock Mechanics and Engineering，2023，42(9)：2 266–2 282.(in Chinese))
[6] JI J，CUI H Z，ZHANG T，et al. A GIS-based tool for probabilistic physical modelling and prediction of landslides：GIS-FORM landslide susceptibility analysis in seismic areas[J]. Landslides，2022，19(9)：2 213–2 231.
[7] BAUM R L，SAVAGE W Z，GODT J W. TRIGRS： a Fortran program for transient rainfall infiltration and grid-based regional slope-stability analysis，version 2.0[R]. [S. l.]：[s. n.]，2008.
[8] PICIULLO L，CALVELLO M，CEPEDA J M. Territorial early warning systems for rainfall-induced landslides[J]. Earth-Science Reviews，2018，179：228–247.
[9] WHITELEY J S，CHAMBERS J E，UHLEMANN S，et al. Geophysical monitoring of moisture-induced landslides：A review[J]. Reviews of Geophysics，2019，57(1)：106–145.
[10] ZHU D S，XIA L，GRIFFITHS D V，et al. Reliability analysis of infinite slopes with linearly increasing mean undrained strength[J]. Computers and Geotechnics，2021，140：104442.
[11] 郭万里，蔡正银，简富献. 滑坡堰塞体材料空间变异性研究的试验方法[J]. 岩土工程学报，2025，47(10)：2 076–2 085.(GUO Wanli,，CAI Zhengyin，JIAN Fuxian. Test method for studying spatial variability of landslide dam materials[J]. Chinese Journal of Geotechnical Engineering，2025，47(10)：2 076–2 085.(in Chinese))
[12] 姬 建，崔红志，佟 斌，等. 基于物理过程不确定性的降雨诱发浅层滑坡易发性快速区划：GIS-FORM技术开发与应用[J]. 岩石力学与工程学报，2024，43(4)：838–850.(JI Jian，CUI Hongzhi，TONG Bin，et al. Fast zoning of rainfall-induced shallow landslide susceptibility based on physical process uncertainty：development and application of GIS-FORM[J]. Chinese Journal of Rock Mechanics and Engineering，2024，43(4)：838–850.(in Chinese))
[13] CHANG Z L，JIANG S H，HUANG F M，et al. Prediction of overburden layer thickness based on spatial heterogeneity analysis and machine learning models in hillslope regions[J]. Geoscience Frontiers，2025，16(5)：102109.
[14] MERGHADI A，YUNUS A P，DOU J，et al. Machine learning methods for landslide susceptibility studies：A comparative overview of algorithm performance[J]. Earth-Science Reviews，2020，207：103225.
[15] REICHENBACH P，ROSSI M，MALAMUD B D，et al. A review of statistically-based landslide susceptibility models[J]. Earth-Science Reviews，2018，180：60–91.
[16] WOODARD J B，MIRUS B B. Overcoming the data limitations in landslide susceptibility modeling[J]. Science Advances，2025，11(8)：eadt1541.
[17] MONDINI A C，GUZZETTI F，MELILLO M. Deep learning forecast of rainfall-induced shallow landslides[J]. Nature Communications， 2023，14(1)：2466.
[18] 窦 杰，向子林，许 强，等. 机器学习在滑坡智能防灾减灾中的应用与发展趋势[J]. 地球科学，2023，48(5)：1 657–1 675.(DOU Jie，XIANG Zilin，XU Qiang，et al. Application and development trend of machine learning in landslide intelligent disaster prevention and mitigation[J]. Earth Science，2023，48(5)：1 657–1 674.(in Chinese))
[19] MIAO F，ZHAO F，WU Y，et al. Landslide susceptibility mapping in three gorges reservoir area based on GIS and boosting decision tree model[J]. Stochastic Environmental Research and Risk Assessment， 2023，37(6)：2 283–2 303.
[20] ZHANG J，ZHU C，XIE L，et al. Comparative study on landslide susceptibility assessment of different models：A case study of alpine mountainous region in Xinjiang[J]. Bulletin of Engineering Geology and the Environment，2025，84(2)：95.
[21] 郭 飞，赖 鹏，黄发明，等. 基于知识图谱的滑坡易发性评价文献综述及研究进展[J]. 地球科学，2024，49(5)：1 584–1 606.(GUO Fei，LAI Peng，HUANG Faming，et al. Literature review and research progress of landslide susceptibility mapping based on knowledge graph[J]. Earth Science，2024，49(5)：1 584–1 606.(in Chinese))
[22] 黄发明，吴敦筱，常志璐，等. 滑坡样本缺失下易发性规律和潜在滑坡识别的易发性-InSAR多源信息法[J]. 岩石力学与工程学报，2025，44(3)：584–601.(HUANG Faming，WU Dunxiao，CHANG Zhilu，et al. Landslide susceptibility law under the absence of landslide sample and the susceptibility-InSAR multi-source information method considering the potential landslide identification[J]. Chinese Journal of Rock Mechanics and Engineering，2025，44(3)：584–601.(in Chinese))
[23] 仉文岗，何昱苇，王鲁琦，等. 基于水系分区的滑坡易发性机器学习分析方法：以重庆市奉节县为例[J]. 地球科学，2023，48(5)：2 024–2 038.(ZHANG Wengang，HE Yuwei，WANG Luqi，et al. Machine learning solution for landslide susceptibility based on hydrographic division：Case study of Fengjie County in Chongqing[J]. Earth Science，2023，48(5)：2 024–2 038.(in Chinese))
[24] LIU S L，WANG L Q，ZHANG W G，et al. A physics-informed data-driven model for landslide susceptibility assessment in the Three Gorges Reservoir area[J]. Geoscience Frontiers，2023，14(5)：101621.
[25] PENG M，WANG Y，MA C，et al. Landslide susceptibility assessment integrating deep transfer learning and physical models in the Baihetan reservoir area，China[J]. Journal of Rock Mechanics and Geotechnical Engineering，2025，DOI：10.1016/j.jrmge.2025.07.007.
[26] WANG H，WANG L，ZHANG L. Transfer learning improves landslide susceptibility assessment[J]. Gondwana Research，2023，123：238–254.
[27] MEHRABI N，MORSTATTER F，SAXENA N，et al. A survey on bias and fairness in machine learning[J]. ACM Computing Surveys，2022，54(6)：1–35.
[28] PEI T，QIU T. Landslide susceptibility mapping using physics-guided machine learning：A case study of a debris flow event in Colorado Front Range[J]. Acta Geotechnica，2024，19：6 617–6 641.
[29] PEI T，QIU T. Machine learning with monotonic constraint for geotechnical engineering applications：An example of slope stability prediction[J]. Acta Geotechnica，2024，19(6)：3 863–3 882.
[30] PEI T，QIU T，SHEN C. Applying knowledge-guided machine learning to slope stability prediction[J]. Journal of Geotechnical and Geoenvironmental Engineering，2023，149(10)：4023089.
[31] WEI X，ZHANG L L，GARDONI P，et al. Comparison of hybrid data-driven and physical models for landslide susceptibility mapping at regional scales[J]. Acta Geotechnica，2023，18(8)：4 453–4 476.
[32] CUI H Z，TONG B，WANG T，et al. A hybrid data-driven approach for rainfall-induced landslide susceptibility mapping：Physically-based probabilistic model with convolutional neural network[J]. Journal of Rock Mechanics and Geotechnical Engineering，2025，17(8)：4 933–4 951.
[33] 蒋水华，刘 贤，黄发明，等. 融合观测信息的降雨诱发斜坡失稳机理及可靠度分析[J]. 岩土工程学报，2022，44(8)：1 367–1 375. (JIANG Shuihua，LIU Xian，HUANG Faming，et al. Rainfall-induced slope failure mechanism and reliability analyses based on observation information[J]. Chinese Journal of Geotechnical Engineering，2022，44(8)：1 367–1 375.(in Chinese))
[34] SEGONI S，ROSSI G，CATANI F. Improving basin scale shallow landslide modelling using reliable soil thickness maps[J]. Natural Hazards，2012，61(1)：85–101.
[35] LOW B K，TANG W H. Efficient spreadsheet algorithm for first-order reliability method[J]. Journal of Engineering Mechanics，2007，133(12)：1 378–1 387.
[36] ALEOTTI P，CHOWDHURY R. Landslide hazard assessment：summary review and new perspectives[J]. Bulletin of Engineering Geology and the Environment，1999，58(1)：21–44.
[37] TSAI W P，FENG D，PAN M，et al. From calibration to parameter learning：harnessing the scaling effects of big data in geoscientific modeling[J]. Nature Communications，2021，12(1)：5988.
[38] SHEN C，APPLING A P，GENTINE P，et al. Differentiable modelling to unify machine learning and physical models for geosciences[J]. Nature Reviews Earth and Environment，2023，4(8)：552–567.
[39] ROBERTS D R，BAHN V，CIUTI S，et al. Cross-validation strategies for data with temporal，spatial，hierarchical，or phylogenetic structure[J]. Echography，2017，40(8)：913–929.
[40] BELKINA A C，CICCOLELLA C O，ANNO R，et al. Automated optimized parameters for T-distributed stochastic neighbor embedding improve visualization and analysis of large datasets[J]. Nature Communications，2019，10(1)：5415.
[41] 何简吟. 降雨型浅层滑坡稳定性研究——基于TRIGRS模型和Scoops3D模型[硕士学位论文][D]. 西安：西北大学，2021.(HE Jianyin. Stability analysis of rainfall-induced shallow landslide——Based on TRIGRS model and Scoops3D model[M. S. Thesis][D]. Xi′an：Northwest University，2021.(in Chinese))
[42] HENGL T，MENDES DE JESUS J，HEUVELINK G. B. M，et al. SoilGrids250m：Global gridded soil information based on machine learning[J]. Plos One，2017，12(2)：e0169748.
[43] MEDINA V，HÜRLIMANN M，GUO Z，et al. Fast physically-based model for rainfall-induced landslide susceptibility assessment at regional scale[J]. Catena，2021，201：105213.
[44] BOWLES J E. Physical and geotechnical properties of soils[M]. New York，USA：McGraw-Hill，1981：448–449.
2026, Vol. 45 
