2007 NSTA: St. Louis, Missouri Earthquake Prediction and Forecasting: A Case Study of the San Andreas and New Madrid Faults Sponsored by: IRIS (Incorporated.

2007 NSTA: St. Louis, MissouriEarthquake Prediction and Forecasting: A Case Study of the San Andreas and

New Madrid Faults

Sponsored by: IRIS (Incorporated Research Institutions for Seismology)NSF (National Science Foundation)

Presented by:John Taber, IRISMichael Wysession, Department of Earth and Planetary Sciences, Washington University, St. Louis, Missouri

For all IRIS NSTA activities

www.iris.edu/joomla

Go to: NSTA 2007

Earthquake Prediction: Precursory Events

Ex/ 1989 Loma Prieta Earthquake

Also: Resistivity, water pressure and well levels, geyser activity, changes in seismicity

Earthquake Prediction: Animal Behavior??

The exception: Feb 4, 1974: Haicheng, China

* Almost no details are known of this.

Seismic “Gap” Hypothesis

Another Approach: Forecasting

OLD Map of “Seismic Gap” Predictions

Parkfield was in correct place, but very late.

Northridge, Landers, Joshua Tree and Big Bear Earthquakes were not even on this map!!!

“New York City Bear Gap” Hypothesis

Parkfield, California, showed evidence of the recurrence of similar-sized (M 6.0) earthquakes

In 1985 a 6.0 Parkfield earthquake was predicted with 95% confidence to occur by 1993.

Mean = (1966-1857)/5 = 22 yearsExpected date = 1988

Didn’t occur until 2004

(16 years late!)

Was it a success?

Right size, right location, wrong date.

Paleoseismology

M >7 mean = 132 yr = 105 yr Estimated probability in 30 yrs 7-51%

Sieh et al., 1989Extend earthquake history with geologic record

The NMFZ (New Madrid Fault Zone) is a region of elevated seismicity at the intersection of Missouri, Arkansas, Tennessee, Kentucky, and Illinois

The NMFZ is associated with the Reelfoot Rift – a failed rift zone that was active 750 million years ago, as the supercontinent Rodinia began to break up.

The Reelfoot Rift is a region of lower elevation (which is why the Mississippi River flows down through it!).

In 1811-1812, swarm of earthquakes occurred. The largest were:

12/16/1811: M7.2 1/23/1812: M7.0 2/07/1812: M7.4

[Hough et al., 2000]

“the big one”

FA

UL

T L

EN

GT

H

FAULT WIDTH

A Magnitude 8 (M8) earthquake is much bigger than a M7 earthquake:

Roughly 10x greater displacements

Roughly 30x more energy released

NMSZ FREQUENCY-MAGNITUDE RELATIONSHIP

For New Madrid, combine instrumental seismology with earlier data to explore large earthquake recurrence

Large paleoearthquakes occurred at~ 1450 and 900 AD(Magnitudes unknown)

Gives a M7 every few thousand years.

Might NEVER get a M8 earthquake

Stein & Newman, 2004

?

“Seismologists have predicted a 40-60% chance of a devastating earthquake in the New Madrid seismic zone in the next ten years. Those odds jump to 90% over the next 50 years. The potential magnitude of a catastrophic New Madrid quake dictates that we approach the preparedness on a regional basis"

Unjustified given geologic evidence

Are the seismic hazards as great in NMSZ as California? Of course not.

2% chance of shaking within 50 years: = seismic hazard within next 2500 years

The Global Seismic Hazard Assessment Program (GSHAP), with the support of the International Council of Scientific Unions (ICSU), and endorsed as a demonstration program in the framework of the United Nations International Decade for Natural Disaster Reduction (UN/IDNDR).

More reasonable assessment, based upon actual seismicity

Earthquakes occur where earthquakes occur

Earthquakes occur where seismometers exist

Earthquakes occur where earthquakes occur

Earthquakes occur where seismometers exist

Almost no seismometers!!

o Historical o Instrumental

Notice all the earthquakes EAST of the NMSZ!

The NMSZ might be finished (for now?). It might be time for a different failed rift to undergo some continental creaking.

GPS SITE MOTIONS: Vertical show glacial rebound

Horizontal show no pattern

DON’T SEE MOTION AT NEW MADRID OR ELSEWHERE IN EASTERN US:

If you don’t bend it, you can’t break it

GPS site motions within their ellipse of uncertainty - no

motion!

No significant or coherent intraplate deformation visible

GPS data show little or no (< 2 mm/yr)

motion

Little - or none -accumulating for

future earthquakes

GPS CONSTRAINTS ON LARGE EARTHQUAKE MAGNITUDE & RECURRENCE

GPS

Newman et al, 1999

Intersection of Paleoseismology and GPS evidence suggests low M7 earthquakes, but larger events would have VERY long recurrence times.

Will we ever have a real-time warning system for St. Louis?

Not likely. Not much you can do with 25 seconds!

25 s !!