Multi-source InSAR Earth Observation for mapping clay shrink–swell hazard in France: from urban-scale time series to risk-relevant indicators

Clay shrink–swell is a major geohazard affecting buildings and infrastructure in France, driven by seasonal soil moisture variations and exacerbated by recurrent droughts. Interferometric Synthetic Aperture Radar (InSAR) time series provide spatially dense measurements of ground deformation, but the robustness of shrink–swell indicators and their relevance for hazard assessment depend on data sources and processing strategies.
This study investigates the contribution of multi-source InSAR Earth Observation data to the characterization of shrink–swell dynamics at scales relevant for hazard mapping. It builds on a detailed urban-scale analysis over the Toulouse metropolitan area, where two independent Sentinel-1 InSAR pipelines (an academic, fully transparent workflow (Flatsim) and an industrial operational product (SatSense)) were harmonised within a common time-series framework. Identical analyses were applied to both datasets, including trend estimation, harmonic modelling of seasonal deformation, sliding-window analysis, clustering, and the derivation of a composite shrink–swell indicator (RGA index).
Although the two pipelines differ substantially in processing strategy and noise characteristics, they retrieve consistent deformation patterns: weak long-term subsidence combined with spatially coherent seasonal signals controlled by clay-rich formations. Differences mainly affect spatial smoothness and noise levels and do not alter hazard-relevant metrics. These results indicate that shrink–swell signals derived from Sentinel-1 time series are robust to processing choices.
Finally, we discuss how this approach can be extended by integrating European Ground Motion Service (EGMS) products and complementary InSAR datasets to support national-scale screening of shrink–swell hazard in France.