Unified model for intraplate earthquakes - Intraplate Earthquakes - page 280

is concentrated by a heterogeneity with a different elastic modulus than the surrounding
space. The same is true for faults, which are equivalent to planar heterogeneities. Three
types of LSCs were identified and incorporated into a testable local stress concentrator
model for intraplate earthquakes by Talwani and Gangopadhyay ( 2000 ) . These are shallow
plutons, rift pillows, and fault intersections and kinks. These are described below.

11.4.1 Stress amplification around plutons

A spatial association between seismicity and stress accumulation on the periphery of mafic
and ultramafic plutons embedded in felsic country rocks was identified by Long ( 1976 ) ,
Kane ( 1977 ) , and McKeown ( 1978 ) , and ascribed as being due to rigidity contrast between
the two. Analytical modeling by Campbell ( 1978 ) showed that the accumulated differential
stress was largest when an elliptical inclusion was parallel to S T , and it scaled with the size
of the pluton. McKeown ( 1978 ) also noted that ancient rift zones may be a primary control
on the locations of mafic intrusions.
With increasingly better seismic, geophysical, and geological data, a spatial association
between seismicity and stress accumulation on the periphery of plutons was observed in the
NMSZ (Hildenbrand et al ., 2001 ) , and in South Carolina (Stevenson et al ., 2006 ; Talwani
and Howard, 2012 ) . Agreement of calculated stresses on the periphery of plutons, and the
predictions of the analytical model of Campbell ( 1978 ) , with the observed locations of
seismicity has further strengthened the suggestion of a causal association between the two
(Stevenson et al ., 2006 ) .

11.4.2 Seismicity associated with rift pillows
A rift pillow forms as a result of mafic magmatic intrusion into the lower crust during rift
formation. In failed rifts this high-density rift pillow forms in the lower crust and the excess
mass of the pillow must be supported by the strength of the cooled lithosphere, inducing
deviatoric stresses in the plate (Zoback and Richardson, 1996 ) . A causal association of
seismicity with buried rift pillows has been inferred in Brazil, in the NMSZ, and at two
locations in India.
The east-west trending Amazonas rift in central Brazil is one of the largest continental
rifts in the world (Zoback and Richardson, 1996 ) . Two deep (23 km and 45 km) moderate (M
5.1 and 5.5) earthquakes occurred on the northern margin of the rift in 1963 and 1983. Based
on the theoretical models by Sonder ( 1990 ) and Zoback ( 1992a ) , Zoback and Richardson
( 1996 ) suggested that these earthquakes were associated with local stress concentration
around the rift pillow at those depths. The depth of the rift pillow was inferred from gravity
data. The observed direction of the maximum horizontal stress, S Hmax , in the vicinity of
the rift inferred from bore-hole data and from focal mechanisms was found to be rotated

counter-clockwise relative to the east-west direction of S T . Based on modeling, they
suggested that the induced rift normal stresses, S L , associated with the rift pillow can be as
large as the tectonic stress, and that its influence on S T depends both on the orientation of
the rift pillow relative to S T and on the ratio of the magnitude of the two stresses.

75 °