Intraplate seismicity in Brazil - Intraplate Earthquakes - page 66

or in craton edges, in agreement with similar findings on a global scale (Mooney et al .,
2012 ) .
Areas near proposed neotectonic faults (less than 20 km) also seem to be more prone
to seismic activity compared with areas farther away from these features. However, this
correlation only holds on a continental scale: several neotectonic faults have no evidence of
seismic activity ( Figure 3.5 ) and no neotectonic features have been mapped yet in several
seismic zones, such as the northern Parecis and the Pantanal basins. The compilation of
neotectonic faults by Saadi et al . ( 2002 ) is a preliminary result and many more active
features still remain to be mapped. Also, the sampling of neotectonic features compiled
by Saadi et al . may have been influenced by known regions of seismicity, such as the
Codajas fault (BR-05) in the Amazon basin, and the Manga fault (BR-47) in the middle of
the Sao Francisco craton. However, despite the uneven sampling of possible neotectonic
faults, the correlation with epicentral distribution is statistically highly significant. On
the other hand, comparisons of the fault-plane solutions of several events close to these
faults (especially in northeastern Brazil) reveal inconsistencies between the earthquake fault
planes and the geologically mapped structures. We propose that the statistical correlation
between epicentres and neotectonic features is due to some other mechanism, such as stress
concentration near intersecting structures, as proposed by Gangopadhyay and Talwani
(2003, 2007), or unfavourable orientation of the main regional fault with respect to the
current intraplate stresses. At any rate, the relationship of neotectonic features and intraplate
seismicity deserves further studies.
Seismic zones such as the eastern border of the Amazon craton and the northern Parecis
basin (zones 2 and 4 in Figure 3.2c ) could be attributed to stress concentration due both
to lateral density variation in cratonic keels and to flexural stresses from intracrustal loads.
Seismicity in northeastern Brazil (zone 3) and in the Pantanal basin (zone 5) could be
related to stress concentration in the upper crust in areas of thin and weak lithosphere,
such as proposed by Assump¸ ao et al .( 2004 ) for central and southeastern Brazil. The
seismic zone in the middle of the Amazon basin (zone 1) could be influenced by flexural
stresses from intracrustal loads along the Amazon rift ( Figure 3.7 ) , as proposed by Zoback
and Richardson ( 1996 ) , but probably other causes are also necessary as no seismicity is
observed directly above the gravity high.
An interesting aspect of the seismicity distribution in southeast Brazil is the com-
plete lack of epicentres along the Serra do Mar coastal ranges (“SM” in Figure 3.2b and
Figure 3.4 ) . A system of Cenozoic continental rifts, together with the high topography
of the coastal ranges, have been subjected to several neotectonic studies (e.g., Riccomini
and Assump¸ ao, 1999 ; Modenesi-Gauttieri et al ., 2002 ; Cogne et al ., 2012 ) , which have
mapped many faults, some with Holocene reactivation (carbon-14 dating). The compilation
of Saadi et al .( 2002 ) includes several neotectonic faults along the southeast coastal ranges
( Figure 3.5 ) . The extensive evidence of Quaternary activity and lack of epicentres in the
Brazilian catalogue led Assump¸ ao and Riccomini ( 2011 ) to propose the possibility of
long-term seismic migration in southeast Brazil.
Clearly, no simple model seems to explain the location of all seismic zones, and a
combination of several factors may be necessary to explain the concentration of activity in