Unified model for intraplate earthquakes - Intraplate Earthquakes - page 281

Seismic refraction data indicate the presence of a buried high-density rift pillow in
the lower crust beneath the NMSZ (Mooney et al ., 1983 ) . Grana and Richardson ( 1996 )
modeled the local stress field associated with a rift pillow at a depth of 30 km beneath
the NMSZ. They found that the direction of S Hmax near the top of the rift axis was rotated
clockwise

relative to S T . They suggested that the perturbed stress field over
this and other stress pillows was adequate to trigger earthquakes.
In their study of the M 5.8 1997 Jabalpur earthquake in central India, Rajendran and
Rajendran ( 1998 ) concluded that this deep earthquake (36

10 ° to 30 °

4 km) located within the
Narmada rift was associated with a rift pillow at that depth. The presence of the rift
pillow has been inferred from gravity, deep seismic sounding, and seismic reflection data
(Rajendran and Rajendran, 1998 ) .
Mandal ( 2013 ) analyzed 10 years of aftershocks of the 2001 Bhuj earthquake occurring
in the Kutch rift zone He inverted more than 450 well-determined focal mechanisms in
10 km depth slices to obtain the directions of the local stress orientations. While the
orientation of S Hmax for the top 30 km was along the regional direction of S T ,

±

N-S, he

found that for the depth range of 30-40 km it was rotated clockwise by . The results of

seismic tomography suggested that the earthquakes at that depth were possibly associated
with fluid-filled mafic intrusions (rift pillow?).
In an alternative explanation for the role of a rift pillow in contributing to seismicity
in NMSZ, Stuart et al .( 1997 ) suggested that the cause of the observed seismicity was the
stress concentrated in a weak sub-horizontal detachment fault in the lower crust directly
above the rift pillow.

50 °

11.4.3 Stress concentration associated with fault geometry: the (fault)
intersection model

It has been known for a long time that fault bends play an important role in the generation
and termination of seismicity (see, e.g., King, 1986 ) . With improving seismicity data the
role of fault intersections in generating earthquakes was recognized. In the intersection
model (Talwani, 1988 ) intersecting faults form a locked volume where stress builds up
in response to S T. This stress build-up can be large enough to cause major earthquakes.
Fault intersections provide locations for the initiation and cessation of rupture and for the
local generation and accumulation of stress and the resulting earthquakes (Talwani, 1999 ) .
The observed pattern of seismicity is consistent with the results of numerical analyses of
the stress fields in the immediate vicinity of fault intersections (Andrews, 1989 ; Jing and
Stephansson, 1990 ) .
Simple 2D numerical models have been used to investigate the response to intersecting
faults corresponding to known geological features subjected to far-field loading (Gangopad-
hyay et al ., 2004 ; Gangopadhyay and Talwani, 2005 ) . These studies focused on the NMSZ
and the Middleton Place Summerville seismic zone near Charleston, South Carolina. The
resultant stress patterns, sign and amplitude of the maximum shear stress were consis-
tent with the observed locations and structural style of faulting. These numerical models
thus support the hypothesis that in a localized volume of previously weak crust, fault