Published January 1, 2026 | Version v1

Sentinel-1 SAR-based globally distributed co-seismic landslide detection by deep neural networks

  • 1. Univ Cambridge, Dept Earth Sci, Cambridge, England; Univ Cambridge, Dept Geog, Cambridge, England; Univ Padua, Dept Geosci, Machine Intelligence & Slope Stabil Lab, Padua, Italy
  • 2. CNR, Res Inst Appl Math & Informat Technol, Genoa, Italy
  • 3. Univ Padua, Dept Geosci, Machine Intelligence & Slope Stabil Lab, Padua, Italy; Univ Technol, State Key Lab Geohazard Prevent & Geoenvironm, Chengdu, Peoples R China
  • 4. Ctr Tecnol Telecomunicac Catalunya, Dept Remote Sensing, Geomatics Res Unit, Barcelona, Spain
  • 5. British Geol Survey, Nottingham, England
  • 6. Univ Padua, Dept Geosci, Machine Intelligence & Slope Stabil Lab, Padua, Italy

Description

Rapid response to multiple landslide events demands accurate, all-weather, day-and-night detection capabilities. Optical remote sensing has advanced landslide detection but remains limited under adverse weather and lighting conditions. Synthetic Aperture Radar, resilient to these constraints, remains underexplored for automated landslide detection due to challenges such as complex pre-processing and geometric distortions. This study integrates Deep Neural Networks with Sentinel-1 backscatter data for co-seismic landslide detection, utilizing a data-centric approach. We train and test the models using 11 earthquake-induced widespread landslide events, covering ˜ 73 000 landslides across diverse geologic and climatic settings. Inference on unseen events in Haiti (2021) and Sumatra (2022) demonstrates robust transferability, achieving F1-scores up to 82 %. Using explainable artificial intelligence, we highlight the discriminative capability of change detection bands over backscatter alone. Our findings emphasize the potential of SAR-based DNN models for worldwide, generalized, and rapid landslide detection, addressing critical gaps in current methods that solely use optical data. This research lays a foundation for broader applications in automated SAR-based earth surface change detection, particularly in complex, hilly and mountainous terrains. © 2026 Lorenzo Nava et al.

Notes

This research was supported by the \u201CThe Geosciences for Sustainable Development\u201D project (Budget Ministero dell'Universit\u00E0 e della Ricerca\u2013Dipartimenti di Eccellenza 2023\u20132027C93C23002690001) and by Sichuan Science and Technology Program (grant no. 2024JDHJ0038). Alessandro Novellino was funded through the BGS International NC programme \u201CGeoscience to tackle Global Environmental Challenges\u201D (NERC reference NE/X006255/1). Oriol Monserrat was funded by the Spanish Grant PID2020-116540RB-C21 funded by MCIN/AEI/10-.13039/501100011033.

Files

Sentinel-1 SAR-based globally distributed co-seismic landslide detection by deep neural networks.pdf

Additional details

Funding

Ministerio de Ciencia, Innovación y Universidades
Towards a smart exploitation of land displacement data for the prevention and mitigation of geological-geotechnical risks PID2020-116540RB-C21