Intraplate seismicity in Brazil - Intraplate Earthquakes - page 56

whereas large earthquakes ( 5m b ) in other parts of the country tend to occur as a single

event with a short aftershock sequence.
We calculated the Gutenberg-Richter magnitude-frequency relation using the “uniform”
catalogue for both continental and offshore earthquakes ( Figure 3.2b ) , with the time-
variable detectability thresholds above, using the method of Weichert ( 1980 ) for partially
complete catalogues. This catalogue was declustered (reducing from 800 to 650 events)
and a maximum magnitude of 7.0 was used. We obtained a b -value of 0.97 0.06. The

seismicity rate can best be evaluated for the cumulative frequency of earthquakes with
magnitudes

±

5.0 m b , for which we get a return period of 4 years, making this mid-plate
area of South America one of the SCRs with the lowest seismicity rate in the world, as
shown by Johnston et al . ( 1994 ) .

3.4 Seismotectonic correlations
We now test the statistical significance of some general hypotheses and models regarding
the distribution of seismicity in mid-plate South America, as seen in Figure 3.2 . Onemust
be aware, however, that the Brazilian historical and instrumental catalogue spans a very
short time window and may not be representative of the long-term behaviour of intraplate
seismicity, especially when events of magnitude 6 have return periods of the order of a

century. In addition, it has been suggested that long-termmigration of activity in SCRs may
be a common characteristic, such as in North America (Stein et al ., 2009 ; Li et al ., 2009 )
and Australia (Clark et al ., Chapter 2, this topic). Thus, the statistical tests below should be
regarded as an initial study, and interpretation of the results should consider the relatively
short time span of the catalogue.

+

3.4.1 Lower seismicity in Precambrian cratonic provinces

A first hypothesis to be tested is the possible lower seismicity of Mid-Proterozoic and older
cratonic provinces. Cratons are usually regarded as more “stable” and rigid blocks, com-
pared with the surrounding younger fold belts. We compared the distribution of continental
seismicity only (east of the catalogue limit defined by the thick solid line in Figure 3.2 ) ,
excluding the offshore area. For simplicity, we considered only the two exposed large cra-
tonic areas in the Amazon and Atlantic shields, and have not included small cratonic blocks,
such as the Luis Alves craton in Southern Brazil and the Apa block near the Brazil/Paraguay
border, or cratonic blocks hidden beneath sedimentary basins such as the Rio de La Plata
craton in northeast Argentina or the hypothesized cratonic blocks beneath the Parnaıba and
Parana basins. Figure 3.2b shows that the exposed cratons (Amazon and Sao Francisco
cratons) correspond to 35% of the continental area (i.e., between the “catalogue limit” in
Figure 3.2 and the coastline). If the seismicity distribution had no correlation with these
large cratonic areas, we would expect about the same fraction of epicentres to be located in
those cratonic areas. For the “whole” catalogue ( Figure 3.2a ) we would expect around 390