Rollovers in salt tectonics: The inadequacy of the listric fault model

Rollover anticlines are syn-sedimentary extensional structures, common in salt tectonics, that are classically considered to result from layer bending during hangingwall displacement above a listric normal fault. However, theoretical considerations and seismic images of natural examples put in question the validity of the listric fault model from which rollover growth is steady state, by nature, and controlled by the relative displacement of two blocks. We used laboratory experiments of gravity driven deformation of brittle–ductile models simulating salt tectonics processes to study in which circumstances rollovers are obtained both in time and space. It is argued that rollover growth is a nonsteady process that results from a mechanical instability whose critical lifetime can be divided in three characteristic stages: birth, growth, and decay. In kinematic terms, the development of a synthetic or antithetic rollover requires a three-block system: an upslope block and a downslope block that move downslope and that separate from each other on top of an extending décollement layer lying above a basement block. Synthetic and antithetic rollovers developed with upward concave shapes whatever the sense of shear in the salt layer and can pass laterally to each other without transform fault. This points out that the concave shape of a rollover base is not an argument in favour of listric faulting. The concave shape simply results from the connection between a steeply dipping normal fault and a flat lying or gently dipping décollement.