Quantitative analysis of distributed normal faulting patterns in 3D thermal-mechanical simulations of continental rifting

Thermal-mechanical simulations of lithospheric deformation provide a key tool for testing hypotheses on the processes controlling continental rifting and breakup. While such simulations have qualitatively reproduced many first-order observations of continental rift structure and evolution, most comparisons between modeled and natural data are qualitative in nature. As the processes controlling lithospheric deformation are highly non-linear, quantitative comparisons are required to carefully assess competing hypotheses and derive unique combinations of parameters that successfully reproduce geologic and geophysical observations. Here, we present work that uses a range of open-source and open-access software to run 3D thermal-mechanical simulations of continental rifting and extract quantitative statistics of the modeled fault network evolution. The simulations of continental rifting are run with the open-source mantle convection and lithospheric deformation code ASPECT, which uses advanced solvers and parallel computing to efficiently solve nonlinear and computationally massive problems. Using a combination of velocity boundary conditions, initial random strength perturbations and rapid strain-weakening produces a distributed normal fault network in the upper crust, which evolves as individual faults grow, deactivate and interact with surrounding faults. To derive quantitative statistics characterizing the evolution of the fault network, we utilize open-source and python-based image processing routines to identify individual faults and extract associated properties. Our presentation will focus on the key assumptions in performing this analysis and how they affect the time-integrated statistics characterizing the fault network evolution.