Transcript calaisgsa - Northwestern University

Triggering of New Madrid
Seismicity by Late Pleistocene
Erosion
Eric Calais & Andy Freed
Purdue University
Roy Van Arsdale, University of Memphis
Seth Stein, Northwestern University
Interplate Earthquakes
Plate motions steadily &
quickly reload faults,
making locations of large
earthquakes and average
time between them
consistent with faults’
geological, paleoseismic,
and seismic histories
Plate B
Plate A
Earthquakes at
different time
Intraplate Earthquakes
Unclear what controls
activation of a particular
mid-continental fault and
the duration of its seismic
activity
Stein, Liu & Wang 2009
M 7 events in 1811-12
Small earthquakes
continue, outlining faults
thought to have ruptured
in 1811-1812
Paleoseismology shows
large events ~ 500
years apart in past
2,000 years
Previously taken as
evidence that strain
accumulates steadily
and is periodically
released during large
infrequent events
New Madrid
However, twenty years of GPS
measurements find no
detectable deformation with
progressively higher precision,
constraining present motions
across the NMSZ to be slower
than 0.2 mm/yr
Because the recent
earthquakes correspond to
strain release at a rate
equivalent to a slip of at least
1-2 mm/yr over the past
~2,000 years, deformation
varies with time
Hence, the NMSZ must have
been recently activated,
consistent with the lack of
significant topography, the
jagged fault, and seismic
reflection and trenching studies
that find an increase in slip rate
on the Reelfoot fault by four
orders of magnitude about
10,000 years ago
This recent reactivation of the
NMSZ argues against Holocene
fault activity being a direct
manifestation of tectonic
stresses, which change on
timescales of millions of years.
Forte et al., 2007
Similar conclusion from GPS data showing at
most slow platewide deformation
Plate interior contains many fossil faults
developed at different times with different
orientations but only a few appear active today
Although New Madrid
earthquakes probably
reactivate favorably
oriented faults
associated with
Palaeozoic rifting,
a stress source localized
in space & time must
have recently triggered
these particular faults
Marshak and
Paulson, 1997
GIA – Glacial Isostatic
Adjustment - is unlikely
stress source for
seismicity
May explain seismicity along old ice
sheet margin in Eastern Canada &
elsewhere (Stein et al., 1979; 1989;
Mazzotti et al., 2005)
GPS shows nothing unusual at New
Madrid
Stresses decay rapidly away from ice
margin, so can’t explain NMSZ (Wu
and Johnson, 2000) unless order of
magnitude weaker than surroundings
(Grollimund and Zoback, 2001)
Sella et
al., 2007
No evidence for such weakening
NMSZ not hot or weak
NMSZ heat flow no higher
than surroundings
NMSZ and surroundings
have essentially the same
temperature & thermallycontrolled strength
No strength reason for
platewide stresses to
concentrate in NMSZ
rather than other faults
McKenna, Stein
& Stein, 2007
Similar difficulty for models in
which earthquakes result from
-sinking of “rift pillow” ancient
high density mafic body
(Grana and Richardson, 1996;
Stuart et al., 1997) due to
weakening of the lower crust in
past 9 kyr (Pollitz et al., 2001)
-sudden recent weakening of
lower crust (Kenner & Segall,
2000)
Braile et al., 1986
Problems: no evidence for weak zone and no obvious
reason for why weakening occurred here at this time
Local stress source for seismicity:
postglacial erosion in Mississippi Embayment
Flexure caused by
unloading from river
incision 16 - 10 ka
reduces normal stresses
sufficiently to unclamp
pre-existing faults
Fits location & timing of
recent seismicity
Doesn’t require
assumption of weak zone
Model predicts NMSZ faults continue
being unclamped by relaxation even
10,000 years after alluvial denudation
stopped, although at a slow and
decaying rate
Maximum stress that can be
transferred into the upper crust from
viscoelastic relaxation following a
large earthquake more than one
order of magnitude less than typical
stress drop value
After a large earthquake releases stresses on an intraplate
fault segment, flexure and viscoelastic relaxation are
inefficient at bringing the rupture back to failure equilibrium
unless faults weaken with time
Fault segments that ruptured are
unlikely to fail again soon,
although stress changes from
erosional unloading or large
earthquakes may eventually
bring to failure nearby segments
that have not yet ruptured
Tuttle
(2009)
This process may be how NMSZ
seismicity migrated in the past
and may eventually activate yet
unruptured segments
Other localized stress sources
may have or will generate
earthquakes elsewhere in
midcontinent
Marshak
and
Paulson,
1997
Stress due to Late Pleistocene erosion could have
triggered New Madrid seismicity
Localized mechanism consistent with recent initiation
and localization in NMSZ
Doesn’t require assuming sudden localized crustal
weakening for which no evidence
Fault segments that ruptured unlikely to fail again soon
Stress changes from erosion or large earthquakes may
eventually cause failure on nearby segments that have
not yet ruptured