Geoscience Reference
In-Depth Information
(1992) suggested that mid-plate deformation resulted from stresses that originated at plate
boundaries and were transmitted through the lithosphere. Numerical modeling (Chapter
10) has confirmed this premise.
Some of the earlier models in the 1970s and 1980s hypothesized that intraplate earth-
quakes occurred as a result of stress build-up on discrete structures in the upper and
lower crust. These models were combined into a testable, unified model that postulates
that these discrete structures act as local stress concentrators, a concept that was con-
firmed at two locations with the results of seismic tomography analysis. A comparison of
Figures 6.3 and 9.5, which show the modeled structures at the Sea of Japan and Kutch Rift
Basin, with Figure 11.3c, which shows the predictions of the unified model, illustrates this
point. Such detailed information about the seismic structures is lacking in other regions.
As newer and better data become available, this model will be tested and other models will
evolve.
13.2 New observations
The temporal and spatial pattern of large historic and prehistoric earthquakes in China
(Chapter 5) and the Kutch Rift Basin (Chapters 6 and 11) show an absence of repeat
earthquakes on the same fault. In each case the earthquakes jumped from one location to
another and were described as “roaming earthquakes” (Chapter 5). Although the spatial
extent was different in these cases - widespread, continental scale in China and within a
single Kutch Rift Basin in India - there was no evidence of repeat earthquakes. A similar
pattern was observed for the morphogenic earthquakes (prehistoric events detected from
paleo-seismological studies) in Australia (Chapter 2). These observations are consistent
with a model that hypothesizes sequential reactivation on different structures. This lack
of stationarity of the seismic source on a particular fault in an intraplate setting is at
odds with the basic assumption of stationarity in probabilistic seismic hazard estimation
methodology, which was developed for plate boundary earthquakes (Chapter 12). This
dichotomy requires a reassessment of methods of seismic hazard estimation in intraplate
settings and the development of new ones.
In mid-plate regions there should perhaps be a lower magnitude threshold used in
seismic hazard estimation, as advocated by the authors of Chapters 4 and 8. They point out
that because of low seismic-wave attenuation in mid-plate regions, the presence of poorly
constructed buildings, and vulnerable local soil conditions, there may be significant damage
potential due to earthquakes with magnitudes as low as 4.5.
The seismicity and geodetic observations in the decade following the 2001 M 7.7 Bhuj
earthquake provide a unique dataset to study the seismogenesis of an intraplate earthquake.
There was evidence of a migration of a seismic pulse away from the source of the 2001
event, reaching 250 km in about 10 years with a sequential activation of faults along the way.
Monitoring of ground motion led to the detection of local, isolated pockets (hundreds of
kilometers) of elevated strain rates 6 to 8 years after the 2001 earthquake. These anomalous
