Machine Learning Method Application to Detect Predisposing Factors to Open-Pit Landslides: The Sijiaying Iron Mine Case Study

Jiang Li, Zhuoying Tan (), Naigen Tan, Aboubakar Siddique, Jianshu Liu, Fenglin Wang and Wantao Li
Additional contact information
Jiang Li: School of Resources and Safety Engineering, University of Science and Technology Beijing, Beijing 100083, China
Zhuoying Tan: School of Resources and Safety Engineering, University of Science and Technology Beijing, Beijing 100083, China
Naigen Tan: School of Resources and Safety Engineering, University of Science and Technology Beijing, Beijing 100083, China
Aboubakar Siddique: School of Resources and Safety Engineering, University of Science and Technology Beijing, Beijing 100083, China
Jianshu Liu: School of Resources and Safety Engineering, University of Science and Technology Beijing, Beijing 100083, China
Fenglin Wang: Hebei Iron & Steel Group, Luanxian Sijiaying Iron Mine Co., Ltd., Tangshan 063700, China
Wantao Li: Hebei Iron & Steel Group, Luanxian Sijiaying Iron Mine Co., Ltd., Tangshan 063700, China

Land, 2025, vol. 14, issue 4, 1-27

Abstract: Slope stability and landslide analysis in open-pit mines present significant engineering challenges due to the complexity of predisposing factors. The Sijiaying Iron Mine has an annual production capacity of 21 million tons, with a mining depth reaching 330 m. Numerous small-scale landslides have occurred in the shallow areas. This study identifies four key factors contributing to landslides: topography, engineering geology, ecological environment, and mining engineering. These factors encompass both microscopic and macroscopic geological aspects and temporal surface displacement rates. Data are extracted using ArcGIS Pro 3.0.2 based on slope units, with categorical data encoded via LabelEncoder. Multivariate polynomial expansion is applied for data coupling, and SMOTENC–TomekLinks is used for resampling landslide samples. A landslide sensitivity model is developed using the LightGBM algorithm, and SHAP is applied to interpret the model and assess the impact of each factor on landslide likelihood. The primary sliding factors at Sijiaying mine include distance from rivers, slope height, profile curvature, rock structure, and distance from faults. Safety thresholds for each factor are determined. This method also provides insights for global and individual slope risk assessment, generating high-risk factor maps to aid in managing and preventing slope instability in open-pit mines.

Keywords: open pit; landslide susceptibility; interpretable machine learning; LightGBM; SHAP (search for similar items in EconPapers)
JEL-codes: Q15 Q2 Q24 Q28 Q5 R14 R52 (search for similar items in EconPapers)
Date: 2025
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2073-445X/14/4/678/pdf (application/pdf)
https://www.mdpi.com/2073-445X/14/4/678/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jlands:v:14:y:2025:i:4:p:678-:d:1618516

Access Statistics for this article

Land is currently edited by Ms. Carol Ma

More articles in Land from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().