Distributions of throws, widths and scarp slopes on normal faults and their relations to fault growth: Insights from Auto_Throw code

Cumulative displacements on faults can tell us how they have accumulated over time, and thus how faults have grown. We approach this question from two angles, focusing on normal faults with topographic escarpment. First, we develop a code, Auto_Throw, to automate the measurements of throws on normal fault scarps. Using a piecewise linear fitting approach and basic slope and length constraints, Auto_Throw mathematically explores all possible configurations of scarps and offset surfaces on a topographic profile, so that a throw and its uncertainty are defined from the statistical analysis of tens to hundreds of measurements. Auto_Throw also measures the scarp slopes and the across-fault widths. We test Auto_Throw performance on 35 normal faults of different lengths and contexts, performing a million measurements. Second, we address the tectonic implications of the measurements. We conduct a global analysis of all fault data in order to smooth out site-dependent complexities and identify common features that may result from the fault growth process. We find that throws, scarp slopes and fault widths significantly vary along faults, yet in a similar manner for each fault quantity. Throw distributions show a common envelope shape, asymmetric with a long slip taper. Steepest scarp slopes are sub-constant along faults but systematically drop down towards fault tips. Fault width scales with throw, but a rough zone of constant width is identified off the faults. Tapering slip on a fault, shallowing of its scarp slope, and distributed deformation off its trace, could all be indicative of the fault lateral propagation.