A physics-informed machine learning solution for landslide susceptibility mapping based on three-dimensional slope stability evaluation

Abstract

Landslide susceptibility mapping is a crucial tool for disaster prevention and management. The performance of conventional data-driven model is greatly influenced by the quality of the samples data. The random selection of negative samples results in the lack of interpretability throughout the assessment process. To address this limitation and construct a high-quality negative samples database, this study introduces a physics-informed machine learning approach, combining the random forest model with Scoops 3D, to optimize the negative samples selection strategy and assess the landslide susceptibility of the study area. The Scoops 3D is employed to determine the factor of safety value leveraging Bishop’s simplified method. Instead of conventional random selection, negative samples are extracted from the areas with a high factor of safety value. Subsequently, the results of conventional random forest model and physics-informed data-driven model are analyzed and discussed, focusing on model performance and prediction uncertainty. In comparison to conventional methods, the physics-informed model, set with a safety area threshold of 3, demonstrates a noteworthy improvement in the mean AUC value by 36.7%, coupled with a reduced prediction uncertainty. It is evident that the determination of the safety area threshold exerts an impact on both prediction uncertainty and model performance.

Keywords

machine learning / physics-informed model / negative samples selection / interpretability / landslide susceptibility mapping

References

[1] HeJ-y, QiuH-j, QuF-h, et al. . Prediction of spatiotemporal stability and rainfall threshold of shallow landslides using the TRIGRS and Scoops3D models. Catena, 2021, 197: 104999 J]

[2] WangL, WuC-z, YangZ-y, et al. . Deep learning methods for time-dependent reliability analysis of reservoir slopes in spatially variable soils. Computers and Geotechnics, 2023, 159105413 J]

[3] WANG Yun-hao, WANG Lu-qi, LIU Song-lin, et al. A comparative study of regional landslide susceptibility mapping with multiple machine learning models [J]. Geological Journal. 2023. DOI: https://doi.org/10.1002/gj.4902.

[4] TangH-m, WasowskiJ, JuangC. Geohazards in the three Gorges Reservoir Area, China Lessons learned from decades of research. Engineering Geology, 2019, 261105267 J]

[5] DengM-l, HuangX-h, YiQ-l, et al. . Fifteen-year professional monitoring and deformation mechanism analysis of a large ancient landslide in the Three Gorges Reservoir Area, China. Bulletin of Engineering Geology and the Environment, 2023, 82(7): 243 J]

[6] ZhangW-g, TangL-b, LiH-r, et al. . Probabilistic stability analysis of Bazimen landslide with monitored rainfall data and water level fluctuations in Three Gorges Reservoir, China. Frontiers of Structural and Civil Engineering, 2020, 1451247-1261 J]

[7] GuoZ-z, FerrerJV, HuerlimannM, et al. . Shallow landslide susceptibility assessment under future climate and land cover changes: A case study from southwest China. Geoscience Frontiers, 2023, 14(4): 101542 J]

[8] AchuAL, AjuCD, Di NapoliM, et al. . Machine-learning based landslide susceptibility modelling with emphasis on uncertainty analysis. Geoscience Frontiers, 2023, 146101657 J]

[9] HuQ, ZhouY, WangS-x, et al. . Machine learning and fractal theory models for landslide susceptibility mapping: Case study from the Jinsha River Basin. Geomorphology, 2020, 351106975 J]

[10] WangH-j, ZhangL-m, YinK-s, et al. . Landslide identification using machine learning. Geoscience Frontiers, 2021, 121351-364 J]

[11] ZhangW-g, LiH-r, HanL, et al. . Slope stability prediction using ensemble learning techniques: A case study in Yunyang County, Chongqing, China. Journal of Rock Mechanics and Geotechnical Engineering, 2022, 14(4): 1089-1099 J]

[12] WangH-j, ZhangL-m, WangL, et al. . Machine learning powered high-resolution co-seismic landslide detection. Gondwana Research, 2022, 123217-237 J]

[13] SunD-l, WenH-j, WangD-z, et al. . A random forest model of landslide susceptibility mapping based on hyperparameter optimization using Bayes algorithm. Geomorphology, 2020, 362: 107201 J]

[14] ZhouX-z, WenH-j, ZhangY-l, et al. . Landslide susceptibility mapping using hybrid random forest with GeoDetector and RFE for factor optimization. Geoscience Frontiers, 2021, 12(5): 101211 J]

[15] ZhangW-g, HeY-w, WangL-q, et al. . Landslide susceptibility mapping using random forest and extreme gradient boosting: A case study of Fengjie, Chongqing. Geological Journal, 2023, 5862372-2387 J]

[16] SunD-l, XuJ-h, WenH-j, et al. . Assessment of landslide susceptibility mapping based on Bayesian hyperparameter optimization: A comparison between logistic regression and random forest. Engineering Geology, 2021, 281105972 J]

[17] Reid M, Christian S, Brien D, et al. Scoops3D: software to analyze three-dimensional slope stability throughout a digital landscape [J]. U. S. Geological Survey. 2015. DOI: https://doi.org/10.3133/TM14A1.

[18] LiuS-l, WangL-q, ZhangW-g, et al. . A physics-informed data-driven model for landslide susceptibility assessment in the Three Gorges Reservoir area. Geoscience Frontiers, 2023, 14(5): 101621 J]

[19] WeiX, ZhangL-l, LuoJ-y, et al. . A hybrid framework integrating physical model and convolutional neural network for regional landslide susceptibility mapping. Natural Hazards, 2021, 1091471-497 J]

[20] WeiX, ZhangL-l, GardoniP, et al. . Comparison of hybrid data-driven and physical models for landslide susceptibility mapping at regional scales. Acta Geotechnica, 2023, 18(8): 4453-4476 J]

[21] RashidB, IqbalJ, SuL-jun. Landslide susceptibility analysis of Karakoram highway using analytical hierarchy process and scoops 3D. Journal of Mountain Science, 2020, 17(7): 1596-1612 J]

[22] PeiT, QiuT, ShenC-peng. Applying knowledge-guided machine learning to slope stability prediction. Journal of Geotechnical and Geoenvironmental Engineering, 2023, 149(10): 04023089 J]

[23] ZhangZ-l, PanQ-j, YangZ-h, et al. . Physics-informed deep learning method for predicting tunnelling-induced ground deformations. Acta Geotechnica, 2023, 1894957-4972 J]

[24] LanP, SuJ-j, MaX-y, et al. . Application of improved physics-informed neural networks for nonlinear consolidation problems with continuous drainage boundary conditions. Acta Geotechnica, 2023, 19(1): 495-508 J]

[25] SheilB. Hybrid framework for forecasting circular excavation collapse: combining physics-based and data-driven modeling. Journal of Geotechnical and Geoenvironmental Engineering, 2021, 147(12): 04021140 J]

[26] ZhangP, YinZ-y, SheilB. Interpretable data-driven constitutive modelling of soils with sparse data. Computers and Geotechnics, 2023, 160105511 J]

[27] ZHANG Pin, YIN Zhen-yu, SHEIL B. A physics-informed data-driven approach for consolidation analysis [J]. Géotechnique. 2023, DOI: https://doi.org/10.1680/jgeot.22.00046.

[28] ZhangS, LanP, LiH-c, et al. . Physics-informed neural networks for consolidation of soils. Engineering Computations, 2022, 3972845-2865 J]

[29] LiuS-l, WangL-q, ZhangW-g, et al. . A comprehensive review of machine learning-based methods in landslide susceptibility mapping. Geological Journal, 2023, 58(6): 2283-2301 J]

[30] HuangB-l, YinY-p, LiuG-n, et al. . Analysis of waves generated by Gongjiafang landslide in Wu Gorge, three Gorges reservoir, on November 23, 2008. Landslides, 2012, 9(3): 395-405 J]

[31] GuD-m, HuangDa. A complex rock topple-rock slide failure of an anaclinal rock slope in the Wu Gorge, Yangtze River, China. Engineering Geology, 2016, 208165-180 J]

[32] ZhouC, HuY-j, XiaoT, et al. . Analytical model for reinforcement effect and load transfer of pre-stressed anchor cable with bore deviation. Construction and Building Materials, 2023, 379: 131219 J]

[33] XiaoL-l, WardSN, WangJ-jia. Tsunami squares approach to landslide-generated waves: application to Gongjiafang landslide, Three Gorges Reservoir, China. Pure and Applied Geophysics, 2015, 172(12): 3639-3654 J]

[34] HuangB-l, YinY-p, WangS-c, et al. . A physical similarity model of an impulsive wave generated by Gongjiafang landslide in Three Gorges Reservoir, China. Landslides, 2014, 11(3): 513-525 J]

[35] ReichenbachP, RossiM, MalamudBD, et al. . A review of statistically-based landslide susceptibility models. Earth-Science Reviews, 2018, 18060-91 J]

[36] ZhangH-j, SongY-x, XuS-l, et al. . Combining a class-weighted algorithm and machine learning models in landslide susceptibility mapping: A case study of Wanzhou section of the Three Gorges Reservoir, China. Computers & Geosciences, 2022, 158: 104966 J]

[37] JanizadehS, BateniSM, JunC, et al. . Potential impacts of future climate on the spatio-temporal variability of landslide susceptibility in Iran using machine learning algorithms and CMIP6 climate-change scenarios. Gondwana Research, 2023, 1241-17 J]

[38] TongZ-l, GuanQ-t, ArabameriA, et al. . Application of novel ensemble models to improve landslide susceptibility mapping reliability. Bulletin of Engineering Geology and the Environment, 2023, 82(8): 309 J]

[39] BreimanL. Random forests. Machine Learning, 2001, 4515-32 J]

[40] ParkD-s, ShinM-b, ParkW-j, et al. . Slope stability analysis model for the frost-susceptible soil based on thermal-hydro-mechanical coupling. Computers and Geotechnics, 2023, 163: 105715 J]

[41] PanQ-j, QuX-r, WangXiang. Probabilistic seismic stability of three-dimensional slopes by pseudodynamic approach. Journal of Central South University, 2019, 26(7): 1687-1695 J]

[42] BishopAW. The use of the Slip Circle in the Stability Analysis of Slopes. Géotechnique, 1955, 5(1): 7-17 J]

[43] HungrO. An extension of Bishop’s simplified method of slope stability analysis to three dimensions. Géotechnique, 1987, 37(1): 113-117 J]

[44] HungrO, SalgadoFM, ByrnePM. Evaluation of a three-dimensional method of slope stability analysis. Canadian Geotechnical Journal, 1989, 26(4): 679-686 J]

[45] YanF-p, WeiS-g, ZhangJ, et al. . Depth-to-bedrock map of China at a spatial resolution of 100 meters. Scientific Data, 2020, 7(1): 2 J]

[46] TangM-g, XuQ, YangH, et al. . Activity law and hydraulics mechanism of landslides with different sliding surface and permeability in the Three Gorges Reservoir Area, China. Engineering Geology, 2019, 260105212 J]

[47] MaoY-m, ZhangM-s, WangG-l, et al. . Landslide hazards mapping using uncertain Naïve Bayesian classification method. Journal of Central South University, 2015, 2293512-3520 J]

[48] YangC, LiuL-l, HuangF-m, et al. . Machine learning-based landslide susceptibility assessment with optimized ratio of landslide to non-landslide samples. Gondwana Research, 2023, 123198-216 J]

[49] KavzogluT, TekeA. Advanced hyperparameter optimization for improved spatial prediction of shallow landslides using extreme gradient boosting (XGBoost). Bulletin of Engineering Geology and the Environment, 2022, 81(5): 201 J]

[50] HuangW-b, DingM-t, LiZ-h, et al. . Landslide susceptibility mapping and dynamic response along the Sichuan-Tibet transportation corridor using deep learning algorithms. Catena, 2023, 222106866 J]

[51] WangY, FangZ-c, WangM, et al. . Comparative study of landslide susceptibility mapping with different recurrent neural networks. Computers & Geosciences, 2020, 138104445 J]

[52] TehranyMS, JonesS, ShabaniF. Identifying the essential flood conditioning factors for flood prone area mapping using machine learning techniques. Catena, 2019, 175174-192 J]

[53] LIU Song-lin, WANG Lu-qi, ZHANG Wen-gang, et al. Physics-informed optimization for a data-driven approach in landslide susceptibility evaluation [J]. Journal of Rock Mechanics and Geotechnical Engineering. 2024, DOI: https://doi.org/10.1016/j.jrmge.2023.11.039.