Geoscience Reference
In-Depth Information
epicentres (35% of 1,118 epicentres) but we observe only 200. Getting half the expected
rate is statistically highly significant. If we use the “uniform” catalogue, we get a similar
result: about 40% of the expected rate (33 observed out of 82 expected).
We can then conclude that cratonic areas in mid-plate South America are about 50% less
seismic than the average mid-plate seismicity in Brazil and neighbouring areas. Including
smaller cratonic blocks (with areas less than 10% of the major Amazon and Sao Francisco
cratons) would not change this result significantly. In fact, it would probably make the
result even more clear, as most of these other small blocks have almost no earthquakes. We
conclude that large cratonic areas are about four times less seismic than the other younger
provinces: Brasiliano fold belts (
750-540 Ma) and intraplate basins. This may not seem
surprising at first. Similar results can be found in other continents such as Australia, where
Clark et al . (this volume) showed that cratonic crust is less seismic than non-cratonic areas.
The statistically significant result for Brazil could only be obtained on a continent-wide
scale. Some cratonic areas can have high earthquake activity (such as in themiddle of the Sao
Francisco craton; e.g., Chimpliganond et al . [2010]) and regional comparisons of cratonic
versus non-cratonic areas (such as done by Assump¸ ao et al . [2004] for southeastern Brazil)
have not produced statistically significant differences.
3.4.2 Intraplate seismicity and cratonic roots
Recently, Mooney et al .( 2012 ) analyzed the global distribution of intraplate earthquakes and
concluded that intraplate seismicity tends to concentrate around cratonic edges, especially
for large magnitudes. The craton edges were defined by the S-wave anomalies from global-
scale tomography. The global scale of that study included both mid-continent as well as
passive margin earthquakes, and earthquakes more properly characterized by continental-
shelf dynamics were included as “craton-edge” events. Here we compare the distribution of
continental earthquakes with S-wave anomalies. We used the S-wave anomaly at 100 km,
obtained from joint inversion of surface waves and receiver function point constraints
(Assump¸ ao et al ., 2013 ) , as a proxy for lithospheric depths ( Figure 3.3 ) .
The map of Figure 3.3a clearly shows a trend of higher seismicity above areas of low
S-wave velocities, which is confirmed by the high frequency of epicentres above areas with
S-wave anomalies between
4% ( Figure 3.3c , d). This result confirms the findings
of Assump¸ ao et al .( 2004 ) , who showed that in Central Brazil higher seismicity is observed
above areas of low P-wave velocities at lithospheric depths. This can be interpreted as stress
concentrations in the upper crust due to lithospheric thinning.
Figure 3.3c , d also shows that the number of events above areas of very high S-wave
velocities (anomalies
1 and
+
5% corresponding to cratonic cores or roots) are lower than
expected. This is consistent with the finding above (Section 3.4.1) that cratonic areas tend
to have lower seismicity. However, an interesting feature in these histograms is the peak
of high seismicity above areas with anomalies in the range
5%. This seems to
correspond to the cratonic edge effect found by Mooney et al . ( 2012 ) on a global scale.
+
3to
+
Search WWH ::




Custom Search