EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Random Forest-Based Landslide Risk Assessment for Mountain Roads Under Extreme Rainfall: Implications for Infrastructure Resilience

Renfei Li, Jun Li, Yang Zhou, Dingding Han, Dongcang Sun, Yingchen Cui, Modi Wang, Mingliang Li

Sustainability · 01.05.2026 · 18 (9) · 4427

DOI: 10.3390/su18094427 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sustainability, online erschienen am 01.05.2026. Verfasst von Renfei Li und 7 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 66 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Central Public-interest Scientific Institution Basal Research Fund.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Extreme rainfall poses an increasing threat to mountainous transportation systems by frequently triggering landslides along road corridors. Most existing studies focus on long-term landslide susceptibility, whereas event-scale assessments remain limited, particularly in road environments. This study develops an event-scale framework for assessing landslide risk along mountain roads under extreme rainfall conditions, using the July 2023 “23·7” rainfall event in Mentougou District, Beijing, as a case study. A Random Forest model was constructed by integrating multi-source geospatial data with an event-specific inventory of 8930 landslides. The model achieved high predictive performance, with ROC–AUC values of 0.9187 and 0.9166 for the validation and test datasets, respectively. Feature importance analysis further indicates that landslide occurrence is controlled by the combined effects of rainfall, terrain conditions, vegetation cover, and anthropogenic disturbance, with rainfall acting as the primary trigger. High-risk road segments are mainly concentrated in the southeastern part of the study area, showing clear spatial clustering. These results highlight the value of event-scale analysis and demonstrate the effectiveness of the road-oriented framework for identifying hazardous segments under extreme rainfall conditions. The proposed approach provides practical support for landslide monitoring, risk mitigation, and resilient management of mountainous transportation infrastructure.

Förderung

  • Central Public-interest Scientific Institution Basal Research Fund

Zitieren

APA

Li, R., Li, J., Zhou, Y., Han, D., Sun, D., Cui, Y., Wang, M., & Li, M. (2026). Random Forest-Based Landslide Risk Assessment for Mountain Roads Under Extreme Rainfall: Implications for Infrastructure Resilience. Sustainability, 18(9), 4427. https://doi.org/10.3390/su18094427

Vancouver

Li R, Li J, Zhou Y, Han D, Sun D, Cui Y, et al. Random Forest-Based Landslide Risk Assessment for Mountain Roads Under Extreme Rainfall: Implications for Infrastructure Resilience. Sustainability. 2026;18(9):4427. doi:10.3390/su18094427

BibTeX
@article{Li2026,
  author = {Li, Renfei and Li, Jun and Zhou, Yang and Han, Dingding and Sun, Dongcang and Cui, Yingchen and Wang, Modi and Li, Mingliang},
  title = {Random Forest-Based Landslide Risk Assessment for Mountain Roads Under Extreme Rainfall: Implications for Infrastructure Resilience},
  journal = {Sustainability},
  year = {2026},
  volume = {18},
  number = {9},
  pages = {4427},
  doi = {10.3390/su18094427}
}
Kostenlos starten →