HR: 15:10h
AN: G22D-07    [PDF]
TI: Characteristic and Uncharacteristic Earthquakes as Possible Artifacts: Applied to the New Madrid and Wabash Seismic Zones
AU: Stein, S
EM: seth@earth.northwestern.edu
AF: Department of Geological Sciences, Northwestern University, 1850 Campus Drive, Evanston, IL 60208
AU: * Newman, A V
EM: anewman@lanl.gov
AF: Earth and Environmental Sciences, EES-9, Los Alamos National Laboratory, MS-D462, Los Alamos, NM 87545
AB: Generally, the largest events, characteristic earthquakes (CE), appear more often than expected from the Gutenberg-Richter relation. Whether this effect is real or apparent is an interesting question since apparent differences can arise from several possible situations. If the known history is shorter than or comparable to the recurrence time of largest events, apparent CEs can occur because sampling bias allows anomalous short recurrence to be observed more frequently (fractions of earthquakes cannot be observed). Alternatively, apparent CEs can occur if paleoseismic data overstimate magnitudes or underestimate recurrence. In the New Madrid Zone (NMSZ), simulations suggest that because the 2000 year paleoseismic record is 4$\times$ the accepted recurrence for CEs, there is a small probability of observing apparent CEs. A more significant effect is that paleoliquifaction data regionally appear to overestimate magnitudes. It is commonly assumed that the distribution of liquifaction, used for paleomagnitudes, near the NMSZ is smaller that seen globally. Liquefaction features extending 250 km from an earthquake in the NMSZ are interpreted as evidence for an M 8.3 event, rather than an M 7.6 as would be inferred from the global curve. This practice arose because the curve was calibrated assuming an M 8.3 for the 1811-12 events whereas more recent analysis finds M$\sim$7.4. Hence either the largest earthquakes are CEs or the paleoevents were smaller than those in 1811-12. The opposite effect occurs in the Wabash Valley Seismic zone, where the paleoearthquakes appear ``uncharacteristic'', less frequent than would be inferred from instrumental seismicity. The paleoseismic record is long enough that the discrepancy is unlikely to be a sampling artifact. Hence either the uncharacteristic behavior is real, or the paleoseismic record captures only a small fraction of the large preinstrumental earthquakes.
DE: 1208 Crustal movements--intraplate (8110)
DE: 7209 Earthquake dynamics and mechanics
DE: 7212 Earthquake ground motions and engineering
DE: 7221 Paleoseismology
SC: Geodesy [G]
MN: 2003 Fall Meeting