Intraplate seismic hazard: Evidence for distributed strain and implications for seismic hazard - Intraplate Earthquakes - page 316

Although the association of elevated modern-day seismicity with failed rifts dates back
to the earliest efforts to develop models for CEUS (and intraplate) seismogenesis, some
key questions clearly remain unanswered. For example, the Precambrian Mid-Continent
(Keweenawan) Rift (e.g., Keller et al ., 1983 ) through the upper Midwest has not been
characterized by elevated seismicity during the instrumental era. One possibility is that
GIA and the presence of a failed rift together are a necessary but not a sufficient condition
for elevated activity. Another possibility is that, as I discuss more in a later section, the
historical catalog is simply too short to reveal all active source zones if recurrence intervals
are long. I note, for example, that the largest historical earthquake in the state of Nebraska,
an estimated M5.1 event on November 15, 1877, is inferred to be associated with the
western flank of the Keweenan mafic belt; it is possible, given the uncertainties in the event
location, that this event was within the rift.

12.3.3 Distributed strain release: a simple model

I now explore simple calculations to consider the possibility that CEUS moment release
is driven by strain accrual that is low but more spatially distributed than current hazard
maps suggest. The precise level of strain rate is unknown. Although attempts have been
made to estimate strain rates from historical moment release rates (e.g., Anderson, 1986 ) ,
these are not expected to be reliable given the short catalog and the enormous uncertainties
associated with the magnitude estimates of the largest events. Somewhat better constraint
is available from GPS investigations. Calais et al .( 2011 ) estimate only an upper bound of
1.3

10 − 9 /yr in the NMSZ, while Galgana and Hamburger ( 2011 ) recently estimated a
value of 1-2

×

10 − 9 /yr in the Wabash valley. The model of Grollimund and Zoback ( 2001 ) ,
which includes an imposed zone of weakness centered on the NMSZ, predicts a strain rate
from GIA on the order of 10 − 9 /yr.
As an initial conservative estimate, I assume a strain rate of 0.5-1.0

×

10 − 9 /yrisdis-
tributed over the failed rifts that have elevated seismic potential (Coppersmith et al ., 2012 ) ,
and that these zones cumulatively make up approximately 10% of the CEUS-SCR crust
( Figure 12.4 ) . A moment rate corresponding to distributed strain accumulation can be
estimated from

×

dε/dt = dM o /dt/ (2 . 67 μ Ah) (12.1)

10 11 dyne/cm 2 ,
and A is area (Anderson, 1986 ) . As an illustrative simple calculation, I consider a total CEUS
area of 2800

where h is the thickness of the brittle layer, μ is shear modulus taken to be 3 ×

2400 km 2 in which there is an elevated strain rate of 0.5-1.0 10 − 9 within

× ×

failed rifts that make up 10% of the total area. Assuming h 15 and no aseismic strain

release, Equation ( 12.1 ) predicts an overall moment accrual rate of (4-8)

=

10 24 dyne cm/yr.
If this rate were released solely in M max events, it would be sufficient to produce one
M w 7.0 earthquake somewhere within failed rifts every 45-90 years. Assuming that 10%
of the moment rate is released by smaller earthquakes, the result is 50-100 years. Further

×