Slowly deforming metropolitan France: what can GNSS tell about physical processes ?

Analysis of lithospheric deformation is key to understanding current tectonics and other active deformation processes. The Alceste project, conducted in the framework of the Résif Seismicity Transverse Action, aims at proposing an updated seismic hazard model in metropolitan France built from the most recent data and academic consensus. One of the contributions will come from geodetic observations through strain rate integration in seismotectonic zoning and seismic hazard models.

Most of Western Europe in general, and metropolitan France in particular, is located within the Eurasian Plate, which has very low deformation and seismicity rates. These GNSS-derived secular velocity field could be related to the combination of different deformation processes, with a minor contribution from plate tectonics (relative plate motions, mantle convection, etc), while most of the measured velocities could be explained by non-tectonic long-term or transient processes (gravitational motions, Glacial Isostatic Adjustment, erosion, anthropogenic deformation, etc). Some of these physical processes causing surface deformation also reflect stress changes at depth that may be associated with loading on active faults and seismicity. Properly mapping this deformation is therefore a key to better assess seismic hazard in slow straining areas.

In order (i) to assess the variability due to the diversity of the strain-rate calculation methods used in the scientific community and (ii) to test their capacity to resolve low-amplitude consistent surface deformation, we conduct a benchmark exercise. We build sets of synthetic velocity fields sampled at the existing GNSS permanent stations from the RENAG (REseau NAtional GNSS Permanent), RGP (Réseau GPS Permanent) and other permanent and non-permanent benchmarks. Our synthetic velocity fields have the same characteristics (noise, uncertainties) as the observed velocities in metropolitan France but they contain surface deformation signals from known physical processes (block rotations, fault elastic loading, large scale flexure, etc). We compare the strain rate invariants derived independently by nine different RENAG research teams (using different software) to the expected strain rate patterns and discuss drawbacks and advantages of each approach. In a second step, we analyzed the strain rate tensors derived from the synthetic velocity fields to discuss potential regional style of the deformation in metropolitan France. Previous studies have shown that the computation of strain rate tensors is impacted by the user-defined parameters and the algorithm specificity used. Exploring these different biases in the strain rate solutions represent the opportunity to improve the understanding of the conventional problem of the standard interpolation.