Enhancing EGMS Products for Local-Scale Ground Deformation Analysis: Ascending–Descending Data Integration and Geostatistical Modelling

The European Ground Motion Service (EGMS) provides freely available Sentinel-1 InSAR products for large-scale ground-deformation monitoring, although the spatial resolution of the standard combined product limits its use for detailed investigations of localized phenomena. This study presents a workflow for enhancing the spatial characterization achievable from EGMS Calibrated products by integrating ascending and descending observations at the local scale, combining temporal harmonization with ground-based monitoring, site-specific geometric co-registration, spatial aggregation on regular grids, ascending–descending decomposition, and geostatistical modelling through variogram analysis and Ordinary Kriging. The methodology was applied to a slow-moving landslide affecting the urban area of Chieuti, southern Italy, and validated against high-precision geometric levelling acquired over the same May 2021–November 2023 period. A sensitivity analysis of grid resolutions between 10 and 25 m identified the 15 × 15 m grid as the most appropriate compromise between spatial detail, availability of ascending–descending observations, and validation performance. Two datasets were generated and investigated: a rigorous dataset based exclusively on cells containing observations from both acquisition geometries and an extended dataset in which missing geometries were locally reconstructed from neighbouring observations to increase spatial coverage. The reconstructed vertical velocity field showed satisfactory agreement with levelling measurements, particularly within the actively deforming sector, whereas cumulative displacement exhibited larger discrepancies. The extended dataset substantially increases spatial coverage while maintaining validation statistics comparable to those of the rigorous dataset, although the latter shows lower residual dispersion within the actively deforming sector. Larger residuals were concentrated near the main kinematic discontinuity, highlighting the limitations of continuous kriging across sharp deformation gradients. The results demonstrate the potential of EGMS products for retrospective local-scale deformation analysis, while emphasizing that successful application depends on favourable acquisition geometry, adequate persistent-scatterer density, identifiable co-registration features, and compatibility between the actual displacement direction and the assumptions adopted for LoS decomposition.