Crustal faulting has long been known as the source of shallow seismicity, and the seismogenic zone is the depth (3-15 km) within the crust that is capable of co-seismic slip, largely under brittle conditions. However, some continental seismicity occurs at depths >> 15 km. I performed thermobarometry of mylonitic pseudotachylyte to determine the P-T of a seismogenic extensional fault in the Caledonian Norwegian margin. Two shear zones (Eidsfjord and Fiskfjord) located in northern Norway exhibit brittle extension propagating into the ductile regime of the lower crust as evidenced by the presence of pseudotachylyte. Averages from Eidsfjord (653 ± 38°C and 570 ± 115 MPa) and Fiskfjord (680 ± 70°C and 1121 ± 219 MPa) correspond to depths of co-seismic slip of 21 ±4 km and 41 ± 9 km, respectively. These depths are 5-25 km below the depth of the standard seismogenic zone in mature fault systems, and require another mechanism (e.g. dynamic downward rupture, unusually high shear stresses) to account for seismogenic rupture at such depths. Assuming Eidsfjord and Fiskfjord were uplifted at the same time, and considering they are currently at the same crustal level, Fiskfjord was uplifted a greater amount and at a faster rate as it was initially located at a greater crustal depth.

The largest earthquakes occur along the subduction thrust interface known as the seismogenic zone. Until recently, erosive margins like Tonga and Honshu have been thought to be unable to support earthquakes with magnitudes higher than 8.5. However, Mw 9, 2011 Tohoku-oki earthquake in Honshu requires a reevaluation of this notion. The seismic potential of Tonga is likely affected by the vertical spatial extent of the up-dip and down-dip limits, which confines the seismogenic zone. The larger the area of the seismogenic zone, the higher the potential for larger earthquakes. Some models suggest that down-dip limit coincides with the fore-arc Moho while others suggest that they are coincident with thermally controlled mineralogical phase changes during slab descent. Tonga is an ideal place to discriminate between these possibilities, as the incoming Pacific plate is cold and thick with rapid convergence, extending cool isotherms deep into the system. In contrast, the fore-arc Moho is only ~16 km deep. This study tests the hypothesis that the down-dip limit of the Tonga seismogenic zone coincides with the fore-arc Moho and thus ceases the seismicity by initiating a stable sliding between the mantle and the subducting crust. We determine the depth of the down-dip limit in Tonga by mapping the distribution of earthquakes recorded for a six-month period from January 1, 2010 to June 30, 2010 by a deployment of ocean bottom seismographs above the Tonga subduction zone. The earthquakes are located by a combination of grid-search method and least-square inversion of the observed arrival times. We identified a down-dip limit at a minimum depth of about 40 km below the sea level suggesting that the hypothesis is failed. Therefore, the commonly held idea that down-dip limit is coincides with the fore-arc Moho is not true in the Tonga case. It is likely controlled by the degree of serpentinization in the mantle wedge controlling the transition from stick-slip to stable sliding.