Landslide Susceptibility Mapping & Exposure Assessment

Interactive WebGIS showcasing a landslide susceptibility model (Random Forest) and population exposure analysis. Built with QGIS preprocessing, Dzetsaka classification, and OpenLayers for the web map.

What’s inside

• Layers: Terrain derivatives (DTM, slope, aspect, plan/profile curvature) and thematic/proximity factors (land use/Geology, distance to rivers, roads, faults).
• Model: Random Forest (Dzetsaka) trained on landslide vs. NLZ points; validated via confusion matrix.
• Exposure: Population counts summarized by susceptibility class.
• WebGIS: Base map switcher + local TIFF overlays, scale line, mouse position, fullscreen.

Quick start

npm install
npm run start   # dev server at http://localhost:5173/

Build for GitHub Pages

npm run build   # outputs static site to /docs

GitHub Pages can serve directly from the docs/ folder (Settings → Pages → deploy from branch, folder /docs).

Notes

• Local TIFF overlays are bundled; the Polimi geoserver is unavailable outside course hours, so remote layers are not expected to load.
• vite.config.js uses base: './' so the build works from any path (including GitHub Pages).

Credits

• Base HTML/CSS: Twenty by HTML5 UP (CCA 3.0)
• OpenLayers + Vite scaffold adapted from the Polimi WebGIS class template

Landslide Susceptibility Mapping & Exposure Assessment

Multi-layer risk mapping, remote sensing insights, and on-demand web mapping for decision makers in Ticino.