Quantifying fault activity over different time scales in the Lower Rhine Graben, towards an improved fault database for seismic hazard assessment.

The Lower Rhine Graben (LRG) is an area of slow intra-plate extension in north-western Europe. Located in a densely populated area, this rift, with moderate but rather continuous seismic activity, poses significant seismic hazard. The LRG NW-trending fault system is 200-km long and accommodates a total extension of ~0.1 +/- 0.03mm/yr. While the major active faults are well known, the activity of this complex system as a whole remains poorly understood. This is partly due to the fact that the tectonic signal issued from such low strain rates deformation is often overprinted by other natural or anthropogenic processes. Thus, previous fault models do not integrate minor structures associated with limited deformation and remain elusive about precise fault geometry and branching. A high-resolution DEM, created from Lidar-based DEMs recently available in the surrounding countries, allows us to retrieve detailed tectonic information and refine the fault traces and scarp geometry. We thus present, for the entire region, a revised and homogeneous fault map, based on morphological observations of fault scarps and offset alluvial terraces, complemented by external information from paleoseismological surveys and geophysical profiles. The high-resolution topography shows a clear difference in fault morphological expression between the eastern and western sides of the graben, with clear scarps and sharp boundaries along the eastern side and smoother cumulative scarps in the west, suggesting contrasting fault behavior across the graben. Based on this detailed mapping, we propose a new active faults model for the whole LRG, reflecting the uncertainties in fault geometry. This is compiled in a database, including several levels of fault mapping (traces, fault sections, faults, main faults), where the fault traces are ranked according to the certainty of their identification and location.

Another limitation for seismic hazard assessment in the area is the relative scarcity of fault-displacement data compared to the large number of structures. In the southern part of the graben, a well-developed terrace allows us to estimate the activity of most faults over the Quaternary, but such an extended marker is missing in the northern part of the LRG, resulting in only few localized data. To complement these offset observations, we use several 3D-geological models. After a selection of the most representative geological layers, we automatically retrieve their offsets at several locations along each fault, to obtain the spatial slip distribution at different timescales.  We observe that along individual faults, the slip profile evolves laterally and in time, showing some fault linkage, while at the scale of the graben borders the total slip does not show significant lateral variations. Moreover, although the surface-expression differs between the two sides of the graben, the total slip rates are fairly equivalent on both sides, suggesting a symmetrical extension, at least for the northern area.

All offset measurements available for different marker horizons are also included in the new LRG fault database, thus providing an integrated tool which allows the user to choose the most relevant timescale and degree of geometrical complexity for advanced seismic hazard assessment.