Geoscience Reference
In-Depth Information
12.2.4 Prehistoric earthquakes in other regions
Paleoseismic investigations in the CEUS have generally focused in the regions where large
earthquakes occurred during historical times. However, paleoliquefaction features have
been identified in other places in the CEUS, away from the NMSZ and the CHSZ. Significant
paleoliquefaction features have been documented in the Wabash Valley (e.g., Munson
recently developed field and analytical techniques, estimating an approximate magnitude of
M
w
7.5. The results of these investigations have been incorporated in the USGS hazard
map. The Wabash Valley remains seismically active in modern times: since 1900 the region
has produced a higher rate of M
5 events than the NMSZ, which has not produced an
earthquake larger than M5 since 1895.
oliquefaction feature near Marianna, Arkansas, more than 100 km south of the currently
seismically active segments of the NMSZ. These sand blows, dated to approximately 5,500
years BP, have very large dimensions, comparable to those found in the NMSZ. As noted by
in space and time within the Reelfoot fault system (i.e., beyond the NMSZ proper), although
more field investigation is needed to assess the history of activity at the Marianna site. Addi-
tionally, a number of geological investigations, including seismic reflection profiling and
trenching, within or adjacent to the greater NMSZ have pointed towards Holocene activity
on faults other than those apparently involved with the 1811-1812 sequence (e.g., Cox
faulting along the Commerce Geophysical lineament, which trends northeastward through
central Arkansas, southeastern Missouri, southern Illinois and south/central Indiana (Bald-
of Holocene faults adjacent to the central NMSZ could provide additional evidence for
localized hazard, but without magnitude estimates or recurrence rates it is impossible to
predict the implications for hazard.
Recent NSHMP maps assume a characteristic earthquake distribution for the NMSZ
and the CHSZ; i.e., the rate of large events estimated from paleoseismic evidence is signifi-
cantly higher than an extrapolation of background seismicity rates assuming a Gutenberg-
acteristic events is higher than the extrapolated Gutenberg-Richter rate by a factor of
3 to a factor of 10 or more, depending on the assumed magnitude of the character-
and the Cheraw and Meers faults in Colorado and Oklahoma, respectively, earthquake
rates for the United States national seismic hazard maps are estimated from background
seismicity rates assuming a Gutenberg-Richter distribution, with an imposed M
max
of
7.5 in rifted crust and 7.0 in SCR regions apart from rift zones (e.g., Frankel
et al
.,