Zooming in on distributed brittle deformation across the Rio Grande rift shoulder: implications for strain weakening of the upper crust

Tectonic plate boundaries are shaped by localized and distributed brittle-plastic deformation such as slip on major faults and folding of fault-bounded blocks. However, the microstructural to outcrop-scale mechanisms that enable such deformation and the factors that control the onset of localization remain a matter of debate. Here we combine field-based strain measurements and numerical modeling of a half-graben to investigate patterns of distributed inelastic strain induced by footwall-flexure of the upper crust. We focus on the Sandia Mountains (New Mexico, USA), which have marked the eastern edge of the Rio Grande rift's middle section for the last ~10–25 Myrs. This half-graben is uniquely suited for our study: it consists of a layer of Pennsylvanian limestone which experienced little deformation prior to Cenozoic rifting and lies uncomformably above Proterozoic granite. Furthermore, most of the present-day topography and up-warping of the limestone can be attributed to slip on the Sandia fault system and is well modeled as the deflection of an anomalously weak elastic upper crust. The Sandia limestone thus constitutes a unique record of distributed brittle strain related to inelastic shoulder flexure.

In the field, deformation within the up-warped footwall-block primarily manifests as small faults (<10s of m slip) and sub-mm to cm-scale mode-I calcite-filled fractures. We identified two sets of veins: a N-striking set subparallel to the axis of flexure, and an E-striking set. Fold tests indicate that the veins formed during the onset of flexure and were mostly tilted with bedding. We measured the aperture of thousands of veins sampled by 31 scan lines distributed along an E-W transect through the Sandia footwall. Vein apertures generally follow a power law distribution of slope ~1. Profiles of E-W fracture-borne strain show clear maxima of ~0.1 with 1-mm fracture densities of ~20 cracks/m at outcrops located 12–15 km away from the range bounding fault. This location represents the hinge of the flexure where bending stresses were apparently large enough to exceed the Mohr-Coulomb failure criterion, yet did not result in the localization of a crustal-scale fault.

To test this idea, we designed 2-D numerical simulations of half-graben growth using a visco-elasto-plastic rheology coupled with plastic strain softening to enable spontaneous fault localization. Our models predict spatial patterns of distributed inelastic strain within the footwall block that are consistent with our field-based fracture intensity profiles. We find that the strain softening rate is a key control on (1) the distribution of footwall inelastic strain and (2) whether distributed strain can localize onto a new crustal fault. This enables us to constrain values of weakening rate (~100 MPa/strain) that reproduce the observed pattern of distributed cracking while allowing prolonged slip on a single master fault. Our results demonstrate that numerical geodynamic simulations can be benchmarked against microstructural observations to quantify the strain localization properties of the lithosphere. They also suggest that the low effective rigidity of warped crust stems from the growth and interaction of tensile defects on a range of spatial scales, as is commonly observed in rock deformation experiments.