The New Madrid seismic zone of the Central United States - Intraplate Earthquakes - page 186

Schweig and Ellis ( 1994 ) speculate that the NMSZ may have initiated between 11 and
3 Ma due to a change in North American plate motion and consequent rotation of the stress
field that resulted in higher resolved shear stress on the Reelfoot Rift faults. However,
this model raises the question of why other Eastern United States rifts have not also
been reactivated. The model also does not account for the Holocene initiation of seismicity
within the NMSZ and surrounding Reelfoot Rift. More recently Forte et al .( 2007 ) argue that
contemporary descent of the Farallon slab in the lower mantle beneath the Central United
States is responsible for New Madrid seismicity. In the Forte et al . model, downwelling
mantle flow, viscously coupled to the ancient Farallon slab, is the driving mechanism.
However, their calculated S Hmax in the NMSZ appears to be of the same magnitude and
direction as much of the Midcontinent Rift System where there is no seismicity or evidence
of Quaternary faulting.
Liu and Zoback ( 1997 ) relate New Madrid faulting to local high heat flow and local
differences in lithospheric strength. McKenna et al .( 2007 ) refute the existence of the
proposed heat flow anomaly and conclude that there is no fundamental difference in litho-
spheric strength between the NMSZ and the surrounding area. Even if subsequent research
should support local high heat flow in the NMSZ, there is no apparent reason why heat flow
should have increased during the Quaternary.
Excess fluid pressure in the NMSZ due to regional groundwater flow has been proposed
to explain its seismicity (McKeown and Diehl, 1994 ) . High pore pressure would reduce the
normal stress across the faults and promote faulting. However, the Holocene groundwater
flow pattern in the NMSZ appears to have been in existence since the formation of the
Mississippi Embayment in the Late Cretaceous and thus this model does not provide an
explanation as to why faulting turned on during the Holocene.
Van Arsdale et al .( 2007 ) have discussed the late Pliocene through Pleistocene erosional
history of the central Mississippi River valley, which they suggest may be tied to the late
Holocene onset of NMSZ activity. The ancestral Mississippi/Ohio River system used to
flow south across this region at an elevation that was 100 m higher than today's Mississippi
River ( Figure 7.6 ) . Entrenchment of the ancestral Mississippi/Ohio River system started at

4 Ma with the most recent 6 m of entrenchment occurring between 12,000 and 10,000
years ago (Calais et al ., 2010 ) . Calais et al .( 2010 ) argue that this most recent entrench-
ment reduced the vertical stress, which reduced the horizontal stress that kept the NMSZ
faults locked. In this model, the erosion between 12-10 ka reactivated the stressed NMSZ
faults.
The Calais et al .( 2010 ) model provides both an explanation for the onset of the NMSZ
and for the onset of the formation of the Lake County Uplift. Denudation of the Missis-
sippi River valley at the end of the Pleistocene caused right-lateral slip on the Axial and
New Madrid North faults thereby causing uplift of the Lake County Uplift compressional
stepover. As discussed above, Sikeston Ridge, Joiner Ridge, and the southern portion of
Crowley's Ridge may also be compressional stepovers. If indeed these are all compressional
stepovers then they could all be a consequence of right-lateral shear across the Reelfoot
Rift that occurred as the valley was denuded over the past

4 Ma. The denudation-driven