Geological Hazards Theme Services

The service, developed by Geoapp, integrates multiple datasets, including landslide inventories, geological and lithological maps, terrain information and vegetation indicators, within a machine‑learning framework to estimate the intrinsic likelihood of slope failure. Vegetation conditions are characterised through NDVI derived from Copernicus Sentinel‑2, while terrain metrics are extracted from digital elevation models.

Users define an Area of Interest (AOI) and provide a catalogue of past landslide events, together with geological and lithological layers. These datasets are used to train a Random Forest model, which identifies spatial patterns associated with previous landslide occurrences and extrapolates them to similar terrain conditions across the AOI. Model training and validation are performed using independent subsets of the input data to ensure robustness of the susceptibility results.

When InSAR‑derived ground‑motion information is available, the service can refine susceptibility outputs into a landslide potential hazard product. In this case, deformation rates are analysed on slope units to characterise activity levels and to highlight areas where intrinsic predisposition coincides with observed terrain movement.

The outputs support disaster preparedness, mitigation and land-use planning and include:

• A pixel-based landslide susceptibility raster map, typically at 20-metre spatial resolution,
• A slope-unit-based vector dataset indicating susceptibility levels and, when applicable, landslide potential hazard.

The service, developed by the Aristotle University of Thessaloniki (AUTh), is based on interferometric stacking techniques, which aggregate information from multiple Sentinel‑1 acquisitions to estimate mean line‑of‑sight (LoS) ground‑motion velocities and associated uncertainty measures. By combining a sufficient number of interferograms over a user‑defined time period, kinesIS provides a stable representation of long‑term deformation trends.

Users define an Area of Interest (AOI) and a time range for analysis. The service automatically processes the corresponding Sentinel‑1 data archive to compute LoS displacement rates expressed in mm/year. When suitable acquisition geometries are available, the results can be further processed using the 3D geometric decomposition service to derive vertical (up–down) and horizontal (east–west) components of ground motion.

The service is designed for wide‑area coverage, with spatial resolution typically ranging from tens to a few hundred metres, depending on processing configuration and data availability. It is particularly suited to areas where spatial continuity is required and where deformation signals are gradual rather than abrupt.

The outputs support regional geohazard assessment, infrastructure monitoring and screening and prioritisation of areas, and include:

• Line-of-sight ground-motion rate maps (mm/year)
• Uncertainty maps

It provides average ground displacement rates and uncertainty estimates by aggregating multiple differential interferograms.

The service, developed by the Aristotle University of Thessaloniki (AUTh) ingests LoS ground‑motion products generated by Interferometric Stacking (kinesIS) or Persistent Scatterer Interferometry (PSI) services, using observations from different satellite viewing geometries. Based on a geometric decomposition approach, it resolves the LoS measurements into vertical (up–down) and horizontal (east–west) displacement rates, expressed in mm/year.

Users provide a minimal set of inputs, including the deformation datasets and a reference area used to stabilise the solution. The service then computes decomposed motion components and associated uncertainty information, ensuring consistency between input datasets and the resulting outputs.

The products depend on the type of input deformation data and support improved interpretation of ground-motion patterns, discrimination between vertical and horizontal deformation processes and geohazard assessment related to subsidence, tectonic motion, volcanic deformation and slope instability. The outputs include:

• For interferometric stacking inputs, the service delivers raster maps of vertical and east–west ground‑motion rates, together with uncertainty layers.
• For PSI inputs, the service produces decomposed motion rates (and, where applicable, displacement time series) provided as tabular outputs suitable for further analysis.