How does a brittle-ductile fault nucleate and grow in dolostone? A lesson learnt from a structural, geochemical and K-Ar chronological study of a reactivated Paleozoic thrust fault

Abstract:

Carbonate-hosted faults in the upper crust are mechanically strong, yet, under certain environmental conditions, carbonates may decompose into mechanically weak minerals, with major consequences for faults´ rheological behavior. We combine structural analysis, geochemistry, stable isotopes and K-Ar dating of synkinematic illite/muscovite to investigate the processes that control localization and weakening of initially strong, seismogenic brittle faults. We aim at better understanding how the constantly evolving architecture and composition of brittle-ductile faults affect their seismogenic properties.

The Kvenklubben fault in northern Norway is part of a Caledonian compressional imbricate stack. It juxtaposes greenschist facies metabasalts in the hanging wall against meta-dolostones and has a 2.5 m thick fault core consisting of talc-bearing calc-phyllonites and chlorite phyllonites. Petrographic and geochemical results indicate that the phyllonites formed mainly through fluid-rock interaction and progressive decomposition of the adjacent wall rocks. K-Ar dating and chlorite geothermometry documents that the fault damage zone developed from the base upwards with fault initiation at 530 Ma around 200°C and the main development during reactivation around 440 Ma at c. 285°C.

Early strain increments were accommodated in the dolostone by pressure-solution, formation of optimally oriented tensional fractures and cataclasis along geometrical irregularities of the growing fault plane. Fluids caused sequential decarbonation of the dolostones and carbonation of the metabasalts, resulting in the formation of phyllosilicate-decorated planar fabrics. The newly formed phyllosilicate levels weakened the fault under overall viscous creep conditions. The strongly anisotropic fluid-flow within the phyllonites, together with vein sealing following localized and transient high pore pressure-driven embrittlement, caused strain hardening. Together, the interaction between strain weakening and hardening mechanisms allowed for the fault core to become gradually wider with increasing shear strain, clearly demonstrating the dynamic character of the fault. An important conclusion is that mechanical models need to account for the evolving character of faults and their structural maturity.