[CEUS-earthquake-hazards] Article on Central U.S. Earthquake hazard by expert panel

Tue May 3 22:01:46 UTC 2011

Dear Buddy:

Having spent many years buried inside the fatality data, what is pretty well
known now is:

1.       You cannot subset the world into nice neat areas and say the hazard
here is X in terms of death, that would be nice, but it is not possible at
all,

a.       you can do this for design purposes accepting that in using this
hazard design method  and using Chicago as a guide you have about a 1 in
10,000 risk per annum of  1 million deaths in the place 

b.      in reality you have to have 3,000,000 in one place for the 1,000,000
toll.  There are only a limited number of places where this occurs. 

c.       The earthquake that causes the event can have a 25,000 year repeat
period, but the fatal event is still likely as a 100 year event, for the
want of simple number. 

d.      The earthquake return period has nothing to do with the fatality
return period, one is a localized event the other can only be measured
against the world data. Haiti was expected, where was unknown, the likely
size range was known and the approximate period, i.e. a mean of about 30
years after 1976. It is to freaking regular. Expect another 250,000 dead in
the next 30 years, in one of 500 places. Until Kafka's method to resolve the
stats is used it cannot be quantified better.

2.       Engineering is important, but the most often cited excuse for the
deaths is the size of the event was not allowed for in design terms, such
as,  we built a 20 foot wall and the wave was 30 feet. I have read this in
every large fatal event in one form or the other.  The recent NZ event
includes this type of issue on first look.

3.       You are never ever going to predict the hazards completely and the
big earthquakes are such long return periods and we have not been around
long enough as history writing humans to use this solution, physics is the
solution not length of records. Forensics  geology will help, but too much
is buried.

4.       Greater urbanization is increasing the risk, this is part of the
Generalized Poissonian Process.

5.       25 of the 125 most urbanized areas are in the CEUS, a large
earthquake risk extends across this area. The early data for the last decade
and before suggests a 7.2 in the wrong place is enough for 1,000,000 deaths.

6.       Your building design hazard mapping is interesting, but it is
merely a simple building standard to hopefully limit the total deaths this
century to 4 million.  We are going to have at least 2 million deaths and it
would appear closer to 4 million.  You are making an ethical decision in
your mapping as to risk level, you are not protecting the population other
than at a general minimum level. Any simple thought experiment shows you
will always be at the minimum of the probability estimate for most places. 

7.       You are relying on the fact that it will only happen once a century
in this very high range according to the stats, in effect the ethical
question is do we design all 125 urban areas for this standard or do we
allow one million to potentially die. In making your maps the decision is
made and the large tolls will continue.  It is a random act of chance using
GPD as to whether it is Mumbai or Chicago.

8.       My guess is you have a 25% chance of being wrong and a 75% chance
of being right about it occurring in the CEUS. Kafka's method would refine
the estimate. I also think that your luck might just run out at some time in
the CEUS.  You cannot roll the dice that many times and miss forever. ( My
thoughts however are that Kafka's method will refine the estimates better
than assuming an equal probability in each of the 3M+ people cities. 

9.       The large event will most likely  have an internal 1 Hz pulse that
will knock the crap out of some part of our engineered world. Newmark and
Hall were looking at pulses in the 1960 to 1980 period, this was stopped
with the cessation of Nuclear Plant design.  I talked to Hall about this
matter and his comment in Aussie terms was no money - no research interest.

10.   This risk will not change till we have a decent method to determine
the location of the next event, probably in the year 2525 we may be there, a
bright physics type will solve this problem probably using techniques as yet
uninvented.  You will save lives by moving everyone out of the buildings in
time.  Nothing else will work, because we cannot engineer safe buildings,
because the population does not want to, a large F350 is more important to
some.

11.   You will not get 1,000,000 deaths in an interplate event, it is just
not statistically possible unless something way outside the norm occurs like
an M10 under SF. I am happy to send you all the papers where this point has
been looked at, 1,000,000 deaths will probably occur and if it does, most
likely it will be in an major intraplate urban area, such as Boston, NY,
London etc. It is irrelevant to me where it occurs,  but if you want to
protect the 125 areas, then the need is for a general M8 design standard to
mitigate for all unknowns. 

12.   Unsellable idea, hence your ethical quandary - this rather interesting
series of emails, and my counting of the dead every two weeks or so.

I agree in part with Seth, the actual risk of 1,000,000 dead out of NMZS is
low at the moment, but I disagree with his method for reaching this
conclusion.  I worry not about the people in Memphis, I merely remember the
words of John Donne, 

"Any man's death diminishes me because I am involved in Mankind; 

And therefore, never send to know for whom the bell tolls; 

It tolls for thee."

Regards

JMN

Dr. John M. Nichols, MIE(Aust), Chartered Professional Engineer

Construction Science Department

College of Architecture

A134 MS 3137 TAMU

Langford AC

College Station, TX, 77843-3137

Phone:   +1 979 845 6541

FAX: +1 979 862 1572

Email : jm-nichols at tamu.edu

From: ceus-earthquake-hazards-bounces at geohazards.usgs.gov
[mailto:ceus-earthquake-hazards-bounces at geohazards.usgs.gov] On Behalf Of
Eugene S Schweig
Sent: Friday, 29 April 2011 12:46 PM
To: Central and Eastern U.S. Earthquake Hazards Listserve
Subject: Re: [CEUS-earthquake-hazards] Article on Central U.S. Earthquake
hazard by expert panel

Dear Colleagues, 

I think that saying the hazard methodology has had a series of spectacular
failures overstates the case.  The maximum earthquake used in the models may
have been wrong, but we work with the understanding we have at the time.
Perhaps these "failures" should caution us to act conservatively in
considering lowering the hazard when faced with conflicting evidence, as in
the case of the New Madrid seismic zone. 

Buddy 
__________________________________________________________________
Eugene Schweig
Director
Geology and Environmental Change Science Center
U.S. Geological Survey

MS980 Federal Center                                      (303) 236-5344
Denver, CO 80225-0046                         Fax: (303) 236-5690
___________________________________________________________________

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://geohazards.usgs.gov/pipermail/ceus-earthquake-hazards/attachments/20110503/6f2f7f6d/attachment-0001.html>