A scalable GIS–database workflow for processing EGMS InSAR time series for landslide susceptibility studies

The European Ground Motion Service (EGMS) provides continental-scale InSAR ground deformation time series, offering new opportunities for investigating slow-moving geohazards such as landslides. However, the direct use of EGMS products in landslide hazard and susceptibility modelling remains challenging due to the large data volumes, the temporal structure of the time series, and the need to integrate deformation data with environmental and climatic variables.

In this contribution, we present a scalable workflow for transforming EGMS data into analysis-ready inputs for dynamic landslide susceptibility studies. The methodology was developed using GRASS GIS and PostgreSQL/PostGIS, exploiting a high-performance multicore computing infrastructure (tens of CPU cores) to efficiently manage very large datasets while preserving the temporal information required for robust interpretation. To optimise computational performance and validate the robustness of the pipeline, the workflow was first tested on a reduced pilot area and subsequently extended to the entire Province of Salerno (southern Italy), a region characterised by complex geomorphology and widespread slope instability.

EGMS Level-2a ascending Persistent Scatterer displacement time series were imported into GRASS GIS and reorganised into complete time series, resulting in a database exceeding 600 million displacement observations. To reduce data dimensionality while retaining physically meaningful information, Persistent Scatterers were spatially associated with slope units and filtered based on extreme displacement values. The deformation observations were therefore integrated with geomorphological, geological and climatic variables, including hourly precipitation data and surface temperature, aggregated at the slope-unit scale.

The resulting spatio-temporal database provides a consistent and comprehensive foundation for training machine learning models aimed at dynamic landslide susceptibility assessment and future early warning applications. The proposed workflow demonstrates how EGMS products can be systematically transformed into scalable and integrated inputs for regional-scale geohazard analysis.