[CEUS-earthquake-hazards] Comment for the CEUS web discussion

[Li, Qingsong]

Hi,

 

I guess people may argue one way or the other. But I do see some
indication from the following paragraph of the open file of 2008 hazard
map. I am happy to see these changes although I expect more. So I am
fine with that these changes are called "improvement" instead of "back
off".

 

Best,

 

Qingsong Li

 

 

Earthquake Recurrence

 

Estimated recurrence intervals of large New Madrid earthquakes have been
refined with additional paleoseismic data. In 1996, sparse data loosely
constrained recurrence intervals; Frankel and others (1996) estimated an
average value of 1,000 years. By 2002, paleoliquefaction and
historical-earthquake data provided evidence for a 500-year recurrence
(Cramer, 2001; Frankel and others, 2002; Tuttle and others, 2002). At
least one large earthquake at about A.D. 300 and three sequences of
large earthquakes at about A.D. 900, A.D. 1450, and A.D. 1811-1812 were
recognized from crosscutting relationships and radiometric dating of
liquefaction features (Tuttle and others, 2002; Tuttle and others,
2005). The occurrence of three sequences between A.D. 300 and A.D.
1811-1812 (about 1,500 years) implies an average recurrence of about 500
years. However, it is unclear whether or not the northern arm ruptured
in the A.D. 1450 sequence. If it did not, then this arm has experienced
only two events in the past 1,500 years, suggesting an average
recurrence of about 750 years. We have weighted the 500-year and
750-year earthquake recurrence alternatives for the northern arm equally
(fig. 6). In addition to these alternative recurrence models, some
recent research suggests that the New Madrid seismic zone may rupture on
average every 1,000 years as suggested by abrupt changes in meander
patterns of the Mississippi River that were used to estimate earthquake
activity on the Reelfoot fault (Holbrook and others, 2006). They infer
that large, reverse-faulting ruptures are episodic. Within relatively
active periods, average recurrence intervals are several centuries long,
but during the less common quiescent periods, average recurrence
intervals are interpreted to be about 1,000 years. Accordingly, we added
a 1,000-year branch to the recurrence model. At present the
meander-pattern methodology cannot detect small reverse-slip ruptures of
the Reelfoot fault or large strike-slip ruptures of other faults in the
seismic zone, so we assigned the 1,000-year branch a weight of 0.1.

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://geohazards.usgs.gov/pipermail/ceus-earthquake-hazards/attachments/20090123/68926ba7/attachment.html