Intraplate earthquakes in North China - Intraplate Earthquakes - page 118

Figure 5.11 (a) Seismicity in the source region of the 1976 Tangshan earthquake. Dots are epicenters
of aftershocks; circles are background seismicity (M

3.0, 1970-2011). The focal mechanism
solutions are for the two mainshocks of 1976 and the May 5, 2012 event. The inset map shows the
location of the Tangshan region. (b) Earthquake sequence in the source region of the great Tangshan
earthquake since 1970. (c) Seismicity rates (number of events per year) of the Tangshan earthquakes
sequence. Solid lines are least-square fitting.

rest of the North China Plain. Again, because these regions have no evidence of unusually
high tectonic activity relative to other fault zones in the North China Plain, the higher strain
rates indicate that postseismic deformation continues to the present (the Haicheng region
does not have sufficient GPS stations to allow a meaningful strain calculation).
The long-distance roaming of large earthquakes and their long sequences of aftershocks
hence make seismic hazard assessment, a challenging task for any region (Stein et al ., 2011 ,
2012), more difficult in North China and other mid-continents. If the earthquakes roam
between widespread faults, the current practice, which uses previous large earthquakes as
hints to locate future earthquakes, tends to overestimate the hazard in places where previous
large earthquake occurred, and underestimate the hazard elsewhere. And if the aftershock
sequences last a long time, small aftershocks may be misread as precursors.
This situation cannot be fundamentally improved by simply extending the earthquake
records alone. This is because earthquakes in North China and other mid-continents do
not fit the model of well-identified faults being steadily loaded. Instead, these earthquakes
are the products of widespread faults interacting with each other in a complex dynamic
system (Stein et al ., 2009 ) . Hence, we cannot treat each individual fault or fault segment
as an isolated system and expect some regular recurrence time. In a complex system
where changes of any part have nonlinear impacts on all other parts (Stein et al ., 2009 ) ,
the concepts of stress cycles, characteristic earthquakes, and recurrence time may not
hold - their validity has been questioned even for interplate earthquakes (Jackson and
Kagan, 2006 ) . Recognizing the fault systems in North China and other mid-continents
as complex dynamic systems does not make hazard assessment easier, but it would be a
necessary paradigm shift. Geodynamic modeling will need to explore how faults interact