Normal faulting in the upper continental crust: observations from regions of active extension

Abstract

Observations of present-day normal faulting in regions of active continental extension may be helpful when interpreting the geological record in older extensional basins. The most obvious manifestation of active extension is normal faulting in earthquakes. Earthquake foci are mostly confined to the upper (seismogenic) continental crust, whose thickness imposes a scale on the observed deformation. Large earthquakes move faults with lengths similar or large compared with the thickness of the seismogenic layer. Large seismogenic normal faults on the continents appear to be restricted to a dip range of around 30–60°: dips significantly gentler than 20° have not been observed in fault plane solutions of large earthquakes. Large seismographic normal faults are approximately planar in cross-section and cut through the base of the upper seismogenic layer, rotating about a horizontal axis as they move. A reasonable estimate of the regional extension can often be made from the dip of the large normal faults and the tilt of the blocks they bound (the ‘domino model’). Such faults are rarely continuous for more than 15–20 km, but commonly change strike or step in an en échelon fashion. This segmentation may occur on a scale controlled by the thickness of the seismogenic layer, and is an important influence on sedimentation and drainage. These observations are common to large seismogenic normal faults from a variety of tectonic settings, and suggest that the kinematics of normal faulting may not be strongly influenced by the forces responsible for the extension, which can vary widely in nature and magnitude. Small earthquakes, which move faults that are small compared with the seismogenic layer thickness, show no simple pattern and often can be interpreted as internal deformation of blocks bounded by large faults. In some places the seismogenic basement faults do not reach the surface but are decoupled from the sedimentary cover by layers of weak lithology. Faults in the sedimentary cover may be strongly curved in cross-section, requiring their hangingwalls to deform internally. Estimating extension from such faults is not straightforward and requires a knowledge of the hangingwall deformation. Estimates made in this way are often very non-unique and prone to large errors. We postulate that the thickness of the seismogenic upper crust controls both the maximum length of large normal fault segments along strike and the maximum size of blocks that can rotate coherently about a horizontal axis.