Basin inversion and contractional reactivation of inherited normal faults: A review based on previous and new experimental models.

Compressional inversion of former extensional sedimentary basins is a particularly common phenomenon, given that passive margins share the same destiny of being subsequently incorporated into fold-and-thrust belts. During basin inversion pre-existing faults may concentrate stress and localise future thrust ramps, or may be reactivated recording reversal of movement from extension to reverse. This process has attracted attention largely because of its economic implications related to petroleum and ore deposit prospects, and seismic hazard assessment. Since more than two decades, analogue modelling has been used to indentify and/or test a number of parameters relevant for the inversion process. This paper aims therefore to offer a reasoned review of the analogue modelling work done on the subject, coupling such previous results with those of a new experimental series inspired to the Northern Apennines stratigraphy and evolution. Past models have explored the role of main factors governing fault reactivation susceptibility, such as fluid pressure, fault weakness, fault steepness, angle of shortening, and sediment loading. The new models have addressed the role played by geometry and strength of a basal ductile layer during inversion. Two sets of sand-silicone models were first extended orthogonally at different velocities to produce dissimilar pre-inversion internal geometries, and then shortened coaxially at different velocities to vary the brittle-ductile coupling. The experimental results confirm the selectiveness of fault inversion, and reveal strong similarities with structural styles of tectonic inversion reported from the Northern Apennines and various areas worldwide. The preferred geological model inferred from this modelling involves contribution of (1) syn-inversion rotation of pre-existing faults to shallower dip, and (2) rotation of principal stress axes, which could explain the invariable reactivation of only one of the two oppositely-dipping graben bounding faults. These factors would thus interact with the other well-established parameters controlling the fault reactivation process.