Improving pixel-based regional landslide susceptibility mapping

Abstract

Regional landslide susceptibility mapping (LSM) is essential for risk mitigation. While deep learning algorithms are increasingly used in LSM, their extensive parameters and scarce labels (limited landslide records) pose training challenges. In contrast, classical statistical algorithms, with typically fewer parameters, are less likely to overfit, easier to train, and offer greater interpretability. Additionally, integrating physics-based and data-driven approaches can potentially improve LSM. This paper makes several contributions to enhance the practicality, interpretability, and cross-regional generalization ability of regional LSM models: (1) Two new hybrid models, composed of data-driven and physics-based modules, are proposed and compared. Hybrid Model I combines the infinite slope stability analysis (ISSA) with logistic regression, a classical statistical algorithm. Hybrid Model II integrates ISSA with a convolutional neural network, a representative of deep learning techniques. The physics-based module constructs a new explanatory factor with higher nonlinearity and reduces prediction uncertainty caused by incomplete landslide inventory by pre-selecting non-landslide samples. The data-driven module captures the relation between explanatory factors and landslide inventory. (2) A step-wise deletion process is proposed to assess the importance of explanatory factors and identify the minimum necessary factors required to maintain satisfactory model performance. (3) Single-pixel and local-area samples are compared to understand the effect of pixel spatial neighborhood. (4) The impact of nonlinearity in data-driven algorithms on hybrid model performance is explored. Typical landslide-prone regions in the Three Gorges Reservoir, China, are used as the study area. The results show that, in the testing region, by using local-area samples to account for pixel spatial neighborhoods, Hybrid Model I achieves roughly a 4.2% increase in the AUC. Furthermore, models with 30 m resolution land-cover data surpass those using 1000 m resolution data, showing a 5.5% improvement in AUC. The optimal set of explanatory factors includes elevation, land-cover type, and safety factor. These findings reveal the key elements to enhance regional LSM, offering valuable insights for LSM practices.

Keywords

Landslide susceptibility mapping / Logistic regression / Convolutional neural network / Hybrid model / Interpretability / Cross-regional generalization

References

A.L. Achu, C.D. Aju, M. Di Napoli, P. Prakash, G. Gopinath, E. Shaji, V. Chandra. Machine-learning based landslide susceptibility modelling with emphasis on uncertainty analysis. Geosci. Front., 14 (2023), Article 101657

H.A.H. Al-Najjar, B. Pradhan. Spatial landslide susceptibility assessment using machine learning techniques assisted by additional data created with generative adversarial networks. Geosci. Front., 12 (2021), pp. 625-637

K.-T. Chang, S.-H. Chiang. An integrated model for predicting rainfall-induced landslides. Geomorphology, 105 (2009), pp. 366-373

L.X. Chen, Z.Z. Guo, K.L. Yin, D.P. Shrestha, S.K. Jin. The influence of land use and land cover change on landslide susceptibility: a case study in Zhushan Town, Xuan'en County (Hubei, China). Nat. Hazards Earth Syst. Sci., 19 (2019), pp. 2207-2228

X.Y. Chen, L.L. Zhang, L.H. Chen, X. Li, D.S. Liu. Slope stability analysis based on the Coupled Eulerian-Lagrangian finite element method. Bull. Eng. Geol., 78 (2018), pp. 4451-4463

Y.M. Chen, L.L. Zhang, X. Wei, J.B. Xu, S.X. Fu, C.C. Liao. Debris-flow-induced damage assessment for a submarine pipeline network in regional-scale natural terrain. Eng. Geol., 311 (2022), Article 106917

A. Contento, H. Xu, P. Gardoni. Probabilistic formulation for storm surge predictions. Struct. Infrastruct. Eng., 16 (2020), pp. 547-566

J. Corominas, C. van Westen, P. Frattini, L. Cascini, J.P. Malet, S. Fotopoulou, F. Catani, M. Van Den Eeckhaut, O. Mavrouli, F. Agliardi, K. Pitilakis, M.G. Winter, M. Pastor, S. Ferlisi, V. Tofani, J. Hervás, J.T. Smith. Recommendations for the quantitative analysis of landslide risk. Bull. Eng. Geol., 73 (2014), pp. 209-263

Q. Cui, L.L. Zhang, X.Y. Chen, Z.J. Cao, X. Wei, J. Zhang, J.B. Xu, D.S. Liu, C.L. Du. Quantitative risk assessment of landslides with direct simulation of pre-failure to post-failure behaviors. Acta Geotech., 17 (2022), pp. 4497-4514

M. Di Napoli, F. Carotenuto, A. Cevasco, P. Confuorto, D. Di Martire, M. Firpo, G. Pepe, E. Raso, D. Calcaterra. Machine learning ensemble modelling as a tool to improve landslide susceptibility mapping reliability. Landslides, 17 (2020), pp. 1897-1914

J. Dou, A.P. Yunus, D.T. Bui, A. Merghadi, M. Sahana, Z.F. Zhu, C.-W. Chen, Z. Han, B.T. Pham. Improved landslide assessment using support vector machine with bagging, boosting, and stacking ensemble machine learning framework in a mountainous watershed, Japan. Landslides, 17 (2019), pp. 641-658

J. Du, T. Glade, T. Woldai, B. Chai, B. Zeng. Landslide susceptibility assessment based on an incomplete landslide inventory in the Jilong Valley, Tibet, Chinese Himalayas. Eng. Geol., 270 (2020), Article 105572

Z.C. Fang, Y. Wang, L. Peng, H.Y. Hong. Integration of convolutional neural network and conventional machine learning classifiers for landslide susceptibility mapping. Comput. Geosci., 139 (2020), Article 104470

K. Gaidzik, M.T. Ramirez-Herrera. The importance of input data on landslide susceptibility mapping. Sci. Rep., 11 (2021), p. 19334

P. Gardoni, A. Der Kiureghian, K.M. Mosalam. Probabilistic capacity models and fragility estimates for reinforced concrete columns based on experimental observations. J. Eng. Mech., 128 (2002), pp. 1024-1038

J.N. Goetz, R.H. Guthrie, A. Brenning. Integrating physical and empirical landslide susceptibility models using generalized additive models. Geomorphology, 129 (2011), pp. 376-386

J.N. Goetz, A. Brenning, H. Petschko, P. Leopold. Evaluating machine learning and statistical prediction techniques for landslide susceptibility modeling. Comput. Geosci., 81 (2015), pp. 1-11

A. Grant, J. Wartman, G. Abou-Jaoude. Multimodal method for coseismic landslide hazard assessment. Eng. Geol., 212 (2016), pp. 146-160

Z. Guo, Y. Shi, F. Huang, X. Fan, J. Huang. Landslide susceptibility zonation method based on C5.0 decision tree and K-means cluster algorithms to improve the efficiency of risk management. Geosci. Front., 12 (2021), Article 101249

F. Guzzetti, A.C. Mondini, M. Cardinali, F. Fiorucci, M. Santangelo, K.-T. Chang. Landslide inventory maps: New tools for an old problem. Earth-Sci. Rev., 112 (2012), pp. 42-66

F.M. Huang, J. Yan, X.M. Fan, C. Yao, J.S. Huang, W. Chen, H.Y. Hong. Uncertainty pattern in landslide susceptibility prediction modelling: Effects of different landslide boundaries and spatial shape expressions. Geosci. Front., 13 (2022), Article 101317

Y. Huang, L. Zhao. Review on landslide susceptibility mapping using support vector machines. Catena, 165 (2018), pp. 520-529

G.E. Karniadakis, I.G. Kevrekidis, L. Lu, P. Perdikaris, S. Wang, L. Yang. Physics-informed machine learning. Nat. Rev. Phys., 3 (2021), pp. 422-440

S. Lee, B. Pradhan. Landslide hazard mapping at Selangor, Malaysia using frequency ratio and logistic regression models. Landslides, 4 (2006), pp. 33-41

Li, Y.Y., Yin, K.L., Chai, B., Zhang, G.R., 2008. Study on statistical rule of shear strength parameters of soil in landslide zone in three gorges reservoir area. Rock Soil Mech. 29, 1419-1418 (in Chinese with English abstract).

C. Li, A.L. Handwerger, J. Wang, W. Yu, X. Li, N.J. Finnegan, Y. Xie, G. Buscarnera, D.E. Horton. Augmentation of WRF-Hydro to simulate overland-flow- and streamflow-generated debris flow susceptibility in burn scars. Nat. Hazards Earth Syst. Sci., 22 (2022), pp. 2317-2345

M. Liao, H. Wen, L. Yang. Identifying the essential conditioning factors of landslide susceptibility models under different grid resolutions using hybrid machine learning: A case of Wushan and Wuxi counties, China. Catena, 217 (2022), Article 106428

Q.G. Lin, P. Lima, S. Steger, T. Glade, T. Jiang, J.H. Zhang, T.X. Liu, Y. Wang. National-scale data-driven rainfall induced landslide susceptibility mapping for China by accounting for incomplete landslide data. Geosci. Front., 12 (2021), Article 101248

Z.Q. Liu, G. Gilbert, J.M. Cepeda, A.O.K. Lysdahl, L. Piciullo, H. Hefre, S. Lacasse. Modelling of shallow landslides with machine learning algorithms. Geosci. Front., 12 (2021), pp. 385-393

L.N. Liu, S.D. Li, X. Li, Y. Jiang, W.H. Wei, Z.H. Wang, Y.H. Bai. An integrated approach for landslide susceptibility mapping by considering spatial correlation and fractal distribution of clustered landslide data. Landslides, 16 (2019), pp. 715-728

S. Liu, L. Wang, W. Zhang, W. Sun, J. Fu, T. Xiao, Z. Dai. A physics-informed data-driven model for landslide susceptibility assessment in the Three Gorges Reservoir area. Geosci. Front., 14 (2023), Article 101621

Y. Liu, W. Zhang, L. Zhang, Z. Zhu, J. Hu, H. Wei. Probabilistic stability analyses of undrained slopes by 3D random fields and finite element methods. Geosci. Front., 9 (2018), pp. 1657-1664

J.Y. Luo, L.L. Zhang, H.Q. Yang, X. Wei, D.S. Liu, J.B. Xu. Probabilistic model calibration of spatial variability for a physically-based landslide susceptibility model. Georisk, 16 (2021), pp. 728-745

K. Mandal, S. Saha, S. Mandal. Applying deep learning and benchmark machine learning algorithms for landslide susceptibility modelling in Rorachu river basin of Sikkim Himalaya, India. Geosci. Front., 12 (2021), Article 101203

R. Melo, J.L. Zêzere, J. Rocha, S.C. Oliveira. Combining data-driven models to assess susceptibility of shallow slides failure and run-out. Landslides, 16 (2019), pp. 2259-2276

B.M. Meneses, S. Pereira, E. Reis. Effects of different land use and land cover data on the landslide susceptibility zonation of road networks. Nat. Hazards Earth Syst. Sci., 19 (2019), pp. 471-487

B.B. Mirus, E.S. Jones, R.L. Baum, J.W. Godt, S. Slaughter, M.M. Crawford, J. Lancaster, T. Stanley, D.B. Kirschbaum, W.J. Burns, R.G. Schmitt, K.O. Lindsey, K.M. McCoy. Landslides across the USA: occurrence, susceptibility, and data limitations. Landslides, 17 (2020), pp. 2271-2285

V. Moosavi, Y. Niazi. Development of hybrid wavelet packet-statistical models (WP-SM) for landslide susceptibility mapping. Landslides, 13 (2015), pp. 97-114

S.C. Oliveira, J.L. Zêzere, S. Lajas, R. Melo. Combination of statistical and physically based methods to assess shallow slide susceptibility at the basin scale. Nat. Hazards Earth Syst. Sci., 17 (2017), pp. 1091-1109

T. Pei, T. Qiu, C. Shen. Applying knowledge-guided machine learning to slope stability prediction. J. Geotech. Geo-Environ. Eng., 149 (10) (2023), Article 04023089,

H.R. Pourghasemi, Z. Teimoori Yansari, P. Panagos, B. Pradhan. Analysis and evaluation of landslide susceptibility: a review on articles published during 2005–2016 (periods of 2005–2012 and 2013–2016). Arab. J. Geosci., 11 (2018), Article 193

B. Pradhan. A comparative study on the predictive ability of the decision tree, support vector machine and neuro-fuzzy models in landslide susceptibility mapping using GIS. Comput. Geosci., 51 (2013), pp. 350-365

P.E. Quinn, D.J. Hutchinson, M.S. Diederichs, R.K. Rowe. Regional-scale landslide susceptibility mapping using the weights of evidence method: an example applied to linear infrastructure. Can. Geotech. J., 47 (2010), pp. 905-927

P. Reichenbach, M. Rossi, B.D. Malamud, M. Mihir, F. Guzzetti. A review of statistically-based landslide susceptibility models. Earth-Sci. Rev., 180 (2018), pp. 60-91

S. Saha, J. Roy, B. Pradhan, T.K. Hembram. Hybrid ensemble machine learning approaches for landslide susceptibility mapping using different sampling ratios at East Sikkim Himalayan, India. Adv. Space Res., 68 (2021), pp. 2819-2840

R. Schlögel, I. Marchesini, M. Alvioli, P. Reichenbach, M. Rossi, J.P. Malet. Optimizing landslide susceptibility zonation: Effects of DEM spatial resolution and slope unit delineation on logistic regression models. Geomorphology, 301 (2018), pp. 10-20

H. Shahabi, S. Khezri, B.B. Ahmad, M. Hashim. Landslide susceptibility mapping at central Zab basin, Iran: A comparison between analytical hierarchy process, frequency ratio and logistic regression models. Catena, 115 (2014), pp. 55-70

R. Strauch, E. Istanbulluoglu, J. Riedel. A new approach to mapping landslide hazards: a probabilistic integration of empirical and physically based models in the North Cascades of Washington, USA. Nat. Hazards Earth Syst. Sci., 19 (2019), pp. 2477-2495

C.X. Su, B.J. Wang, Y.H. Lv, M.P. Zhang, D.L. Peng, B. Bate, S. Zhang. Improved landslide susceptibility mapping using unsupervised and supervised collaborative machine learning models. Georisk, 17 (2022), pp. 387-405

D. Sun, Q. Gu, H. Wen, J. Xu, Y. Zhang, S. Shi, M. Xue, X. Zhou. Assessment of landslide susceptibility along mountain highways based on different machine learning algorithms and mapping units by hybrid factors screening and sample optimization. Gondwana Res., 123 (2023), pp. 89-106

A.Y. Sun, B.R. Scanlon, Z. Zhang, D. Walling, S.N. Bhanja, A. Mukherjee, Z. Zhong. Combining physically based modeling and eeeep learning for fusing GRACE ssatellite data: Can we learn from mismatch?. Water Resour. Res., 55 (2019), pp. 1179-1195

D.L. Sun, J.H. Xu, H.J. Wen, D.Z. Wang. Assessment of landslide susceptibility mapping based on Bayesian hyperparameter optimization: A comparison between logistic regression and random forest. Eng. Geol., 281 (2021), Article 105972

A. Tabandeh, P. Asem, P. Gardoni. Physics-based probabilistic models: Integrating differential equations and observational data. Struct. Saf., 87 (2020), Article 101981

H.M. Tang, J. Wasowski, C.H. Juang. Geohazards in the three Gorges Reservoir Area, China – Lessons learned from decades of research. Eng. Geol., 261 (2019), Article 105267

P.T. Thi Ngo, M. Panahi, K. Khosravi, O. Ghorbanzadeh, N. Kariminejad, A. Cerda, S. Lee. Evaluation of deep learning algorithms for national scale landslide susceptibility mapping of Iran. Geosci. Front., 12 (2021), pp. 505-519

D. Tien Bui, T.A. Tuan, H. Klempe, B. Pradhan, I. Revhaug. Spatial prediction models for shallow landslide hazards: a comparative assessment of the efficacy of support vector machines, artificial neural networks, kernel logistic regression, and logistic model tree. Landslides, 13 (2015), pp. 361-378

K. Ullah, Y. Wang, Z. Fang, L. Wang, M. Rahman. Multi-hazard susceptibility mapping based on Convolutional Neural Networks. Geosci. Front., 13 (2022), Article 101425

Y. Wang, Z.C. Fang, H.Y. Hong. Comparison of convolutional neural networks for landslide susceptibility mapping in Yanshan County, China. Sci. Total Environ., 666 (2019), pp. 975-993

C.H. Wang, Q.G. Lin, L.B. Wang, T. Jiang, B.D. Su, Y.J. Wang, S.K. Mondal, J.L. Huang, Y. Wang. The influences of the spatial extent selection for non-landslide samples on statistical-based landslide susceptibility modelling: a case study of Anhui Province in China. Nat. Hazards, 112 (2022), pp. 1967-1988

H.J. Wang, L.M. Zhang, K.S. Yin, H.Y. Luo, J.H. Li. Landslide identification using machine learning. Geosci. Front., 12 (2020), pp. 351-364

H.J. Wang, L.M. Zhang, H.Y. Luo, J. He, R.W.M. Cheung. AI-powered landslide susceptibility assessment in Hong Kong. Eng. Geol., 288 (2021), Article 106103

X. Wei, L.L. Zhang, J.Y. Luo, D.S. Liu. A hybrid framework integrating physical model and convolutional neural network for regional landslide susceptibility mapping. Nat. Hazards, 109 (2021), pp. 471-497

X. Wei, L.L. Zhang, P. Gardoni, Y.M. Chen, L. Tan, D.S. Liu, C.L. Du, H. Li. Comparison of hybrid data-driven and physical models for landslide susceptibility mapping at regional scales. Acta Geotech., 18 (2023), pp. 4453-4476

L. Weidner, T. Oommen, R. Escobar-Wolf, K.S. Sajinkumar, R.A. Samuel. Regional-scale back-analysis using TRIGRS: an approach to advance landslide hazard modeling and prediction in sparse data regions. Landslides, 15 (2018), pp. 2343-2356

Wen, H., Hu, J., Zhang, J., Xiang, X., Liao, M., 2023. Explainable machine learning model for rockfall ssusceptibility evaluation. Geo-Risk 2023: Developments in Reliability, Risk, and Resilience, pp. 102-110.

T. Xiao, L.M. Zhang, R.W.M. Cheung, S. Lacasse. Predicting spatio-temporal man-made slope failures induced by rainfall in Hong Kong using machine learning techniques. Géotechnique, 73 (2023), pp. 749-765

Y.W. Xu, S. Kohtz, J. Boakye, P. Gardoni, P.F. Wang. Physics-informed machine learning for reliability and systems safety applications: State of the art and challenges. Reliab. Eng. Syst. Saf., 230 (2023), Article 108900

C. Xu, X.W. Xu, F.C. Dai, Z.D. Wu, H.L. He, F. Shi, X.Y. Wu, S.N. Xu. Application of an incomplete landslide inventory, logistic regression model and its validation for landslide susceptibility mapping related to the May 12, 2008 Wenchuan earthquake of China. Nat. Hazards, 68 (2013), pp. 883-900

H.Q. Yang, L.L. Zhang, Q.J. Pan, K.-K. Phoon, Z.C. Shen. Bayesian estimation of spatially varying soil parameters with spatiotemporal monitoring data. Acta Geotech., 16 (2021), pp. 263-278

H.Q. Yang, L.L. Zhang, L. Gao, K.-K. Phoon, X. Wei. On the importance of landslide management: Insights from a 32-year database of landslide consequences and rainfall in Hong Kong. Eng. Geol., 299 (2022), Article 106578

W. Zhang, L. Tang, H. Li, L. Wang, L. Cheng, T. Zhou, X. Chen. Probabilistic stability analysis of Bazimen landslide with monitored rainfall data and water level fluctuations in Three Gorges Reservoir, China. Front. Struct. Civ. Eng., 14 (2020), pp. 1247-1261

L.L. Zhang, J. Zhang, L.M. Zhang, W.H. Tang. Stability analysis of rainfall-induced slope failure: a review. Geotech. Eng., 164 (2011), pp. 299-316

L.L. Zhang, Y.F. Zheng, L.M. Zhang, X. Li, J.H. Wang. Probabilistic model calibration for soil slope under rainfall: effects of measurement duration and frequency in field monitoring. Géotechnique, 64 (2014), pp. 365-378

X.Z. Zhou, H.J. Wen, Y.L. Zhang, J.H. Xu, W.G. Zhang. Landslide susceptibility mapping using hybrid random forest with GeoDetector and RFE for factor optimization. Geosci. Front., 12 (2021), Article 101211

X. Zhou, H. Wen, Z. Li, H. Zhang, W. Zhang. An interpretable model for the susceptibility of rainfall-induced shallow landslides based on SHAP and XGBoost. Geocarto Int., 37 (2022), pp. 13419-13450

T. Zieher, M. Rutzinger, B. Schneider-Muntau, F. Perzl, D. Leidinger, H. Formayer, C. Geitner. Sensitivity analysis and calibration of a dynamic physically based slope stability model. Nat. Hazards Earth Syst. Sci., 17 (2017), pp. 971-992