Python version of ANTI-FASc
End‑to‑end ML pipeline that produces the first global map of land‑subsidence rates, built from 46k+ labeled points with 23 predictors, estimating >6.3 M km² affected, including 231k km² of urban/dense settlements and ~2 billion people.
Estimate incremental volume changes (erosion and deposition) along the path of a debris flow.
SEPARATE is an open-source, GUI-based software package that provides users with a fast, reliable, and automated method to post-process tipping bucket rain gauge (TBRG) data.
A custom component for Home Assistant that displays geohazard alerts from NVE (Norwegian Water Resources and Energy Directorate) and met.no. Stay informed about avalanches, floods, landslides, and other natural hazards.
GeoPrior-v3: Physics-guided AI for geohazards
Data and code for pixel-based landslide susceptibility modelling in Huancabamba (Peru) using Random Forest and spatial block cross-validation.
Geological Hazard Detection Project based on YOLOv8
DEM generation at Sedongpu Glacier (SE Tibet) for debris flow volume estimation
ProboSed: Bridging legacy core logs and modern machine learning. An open-source suite for probabilistic sediment characterization and nonlinear failure analysis in subduction margins.
Analysis code and derived segment database for morphostructural control on subduction seismicity in the Pacific Ocean: 221 trench segments parameterised morphometrically, flexurally and seismically.
Screening whether the thermal conditions before the August 2026 Langtang Lirung ice rock collapse were climatically unusual, using two independent methods: classical trend statistics and a VAE trained on ERA5 Land spatial fields.
Reproducible Python figure-generation code for a study of crustal deformation and earthquake-triggered cascading geohazards (rockfall/landslide, liquefaction, Coulomb stress transfer, sublacustrine mass wasting) across the Çaldıran–Erçiş–Van fault–lake system, Eastern Anatolia.
Python workflow for multi-criteria landslide susceptibility zonation: AHP weighting with consistency ratio, raster reclassification, weighted linear combination and pooled quantile classification. Multi-software pipeline with NumPy, rasterio, GRASS GIS, GDAL and GMT; fully scripted and reproducible from public geospatial data.
LSTM vs. conceptual model flood forecasting, SHAP explained XGBoost risk classification, and a geospatial risk overlay for Thames and Eden catchments, UK.
Dual-pol Sentinel-1 analysis of a repeatable radar precursor signal before two 2025-2026 slope failures at Langtang Lirung, Nepal
Evidence-driven geoparsing for geohazard text, with constrained agents, OSM/Amap validation, GIS enrichment, and resumable human review.
Independent physics analysis of the 26 August 2026 Langtang Lirung collapse and Trishuli flood, Nepal
Multi-Hazard Risk
A framework for assessing rockfall risk for road networks
To associate your repository with the geohazards topic, visit your repo's landing page and select "manage topics."