From eugene@amelia.nas.nasa.gov Thu May 25 13:05:14 PDT 1995
Article: 15002 of ca.earthquakes
Newsgroups: sci.geo.earthquakes,ca.earthquakes
Path: agate!ames!cnn.nas.nasa.gov!amelia.nas.nasa.gov!eugene
From: eugene@amelia.nas.nasa.gov (Eugene N. Miya)
Subject: [lm 5/1/95] Frequently asked earthquake references, part III
Message-ID: 
Followup-To: poster
Summary: This file exists to combat scientific illiteracy
Keywords: This file is temporarily being cross-posted to sci.geo.earthquakes
Lines: 2595
Sender: news@cnn.nas.nasa.gov (News Administrator)
Nntp-Posting-Host: amelia.nas.nasa.gov
Reply-To: eugene@amelia.nas.nasa.gov
Organization: NAS/NASA Ames Research Center
Date: Mon, 22 May 1995 13:56:10 GMT
Xref: agate sci.geo.earthquakes:2482 ca.earthquakes:15002

Archive-name: ca-earthquakes, Part III


Cathy Smither (PhD, Caltech) asked that this post be dedicated to
Alfred Wegener.


The purpose of this FAQ is NOT to circumvent conventional media.
If it did, the FAQ would easily be as long as library text books.
If you are unwilling to look up information in a book,
this FAQ could easily be just as long as any book.


A source for most general questions about earthquakes is
THE classic reference work on the subject:

%A Charles F. Richter
%Z Caltech
%T Elementary Seismology
%I W. H. Freeman Publishers (try a library)
%C New York
%C San Francisco (it appears this office is now closed)
%D 1958

Why do we suggest THIS book?
Because most readers are familiar with Richter's name.  Other people will
question other authors (either authorities [whom the cranks question]
or cranks [who the authorities question], you are welcome to read anyone).
Although this text was published before the theory of plate tectonics,
it is quite readable by the general public (the extensive math being
in the Appendix).  It covers earthquake dynamics, questions like animals,
weather, etc.  And he is THE authority.  Also it's harder to question the dead.

For instance page 133:
"Earthquake weather," as commonly described, is merely a popular fable.
 . . .can be traced to classical writing.  The Greeks . . .

Visit a library to find this book to answer your questions before thinking
about posting to the net.  Do you really think an 100+ line net posting can
answer everything?  Richter's 1.5 inches thick and packed with information.

There are numerous other good books.  Bruce Bolt (UC Berkeley),
Norris (UCSB), and others.  Just check a library out.  Community
college classes are fun and useful (field trips!).  Additionally,
the Bay Area is especially fortunate to have the Menlo Park offices
of the US Geological Survey.  The USGS does work on cartography
(map making), seismology, volcanology, land slides, hydrology, etc.
They have tours, open house, and have speakers available on a wide
variety of subjects.  Just visit them.  They have maps, reports, etc.


IF YOU receive a prediction: earthquakes happen practically everyday
around the world.  Any GOOD prediction will have, epicenter:
	latitude
	longitude
	depth
	date (with time)
	magnitude
Insist on knowing these things and you will get ride of 80% of the cranks.
Simplying saying "A quake will happen" is like dowsing for water.


Richter specifically wrote:
  Amateur predictors are legion, and will continue to be, so long as claiming
  to predict earthquakes is an easy way to get one's name into the newspaper.
  Many of them are honestly self-deceived; they usually have
  (1) no concept of the frequency of small earthquakes (100,000 a year is
  a good figure to think of, although it needs definition in terms of the
  lower limit of the magnitude included),
  (2) no means of knowing what earthquakes have occured or how large they are,
  beyond mention in the press and the space allotted there, and
  (3) no effective training in scientific thinking.
  Some "predictors" select a large number of quakes through the year and then
  claim as predicted anything which happens within a few days of any such
  date -- so that the earthquakes of half the year or even more are called
  in as "verification."
   ...Predictions based on positions of the sun and moon have to be regarded
  a trifle more seriously, since there is evidence that tidal forces may
  occasionally act as triggers for earthquakes otherwise on the point of
  taking place; in this way the date and hours of occurence may show a
  slight statistical correlation with the tides.
Page 386
Chapter 24, Earthquake Risk and Protective Measures
Part I, Earthquake Nature and Observation
Charles Richter


The front of your California phone book, the Red Cross, Sunset Magazine,
your state and county offices of emergency services are
all good sources of information concerning earthquake preparedness.
Topics include:

+Preparing your home to prevent property damage and injury
 from falling objects (e.g. bolting houses to foundations,
 securing water heaters and bookcases, NOT having your bed under a window).

+Stocking a few days of food and water (albeit it could do you little good
 if you are not near your cache).

+Knowing what to do during (get under a doorway or desk, DO NOT run outside)
 and after an earthquake (Do you know how to turn off your utilities?).

+Try to stay off of the telephone except in EXTEREME emergencies.
 Emergency services personnel need the lines for disaster assistance.
 Temporarily disable UUCP connections as well.  If you need to make a
 call, wait patiently for a dial tone.  Clicking the receiver will
 prevent you from ever getting one.  To contact family, set up a contact
 person out of state whom everyone can call.  Calls out are given
 preference over calls into the damaged region.  Pay telephones (at
 least the ones run by Pacific Bell) are on priority service and should
 provide reliable service.

+LEARN FIRST AID.  SF 1989 showed that volunteer rescuers are critical.
 If you want to learn something useful, get a Red-Cross certification
 card, and even then, that's not a guarantee, it's First-Aid.  Any smart
 class will prepare you with pre-cautions.

 Remember STOP!
	  Think
	  Observe
	  Plan

+One previous net.thread involves fire-arms.  Guns have no use in this
 natural disaster.  That silly discussion belongs in other newsgroups.


Lastly, on a frequently mentioned topic peripheral to earthquakes:
The following message comes from a former Vice-President of Caltech:
As for "Caltech" vs. "Cal Tech," it is DEFINITELY supposed to be CALTECH,
one word, not Cal Tech.  Linda is a stickler on this, and woe unto him
who makes this mistake! She frequently finds the error in written stuff,
often stuff coming from Caltech itself.  She blasts the miscreant.
Best wishes  -----Barclay
As far as I know, it was always one word.  Supposedly, the powers that
be (or were) -- Millikan, Hale, and Noyes -- decided at the beginning that
Caltech would be different from Georgia Tech etc.  -----Barclay

This is useful first order check of outgoing Tech material.

END BOILER PLATE.

From: greg@perry.berkeley.edu (Greg Anderson)
Subject: Quake magnitudes
Summary: arbitrary, arbitrary, arbitrary

In article <24eo56INNjlf@darkstar.UCSC.EDU> andy@cse.ucsc.edu
(Andy John) writes:
>Is the magnitude the magnitude of the total energy released, or the amount of
>shaking at it's worst point?

In short, Richter magnitude is none of the above.

In long, I offer the following:

	Richter's definition of the local magnitude scale came about because
he was looking for a quantitative way to compare earthquakes, based on 
instrumental recordings rather than statements of 'the quake felt like XX to
me here'.  Basically, he wanted to be able to say, based on instrumental
recordings, `This earthquake here on this date was larger (or smaller)
than this earthquake over there on this other date.'  
	Richter borrowed the idea of a magnitude scale from astronomers, who
use such to classify the brightness (either apparent or absolute) of stars.
What he did was to define a magnitude scale based on the distance to the quake
and maximum amplitude of that quake as recorded on the photographic Wood-
Anderson instrument.

	Here's how it works:

	You have an earthquake and you have a Wood-Anderson instrument at some
distance from that earthquake.  You locate the quake and calculate the distance
in kilometers from your instrument to the quake's location.  Next, you get out
your Wood-Anderson record and your ruler and measure the maximum amplitude of
the earthquake's recording on that instrument, like so:


                                    ^
	                          /   \
				 /     \
         ----------          ----       \        /--------- <-
					 \      /            |
    record w/o quake      record	  \    /        max amplitude
		           with		   \  /              |
		          quake		    \/		    <-


The formula for determining the Richter magnitude of an earthquake is:

		ML = log10(max. amplitude) - log10(A0)

The term -log10(A0) is a correction for the distance from the quake to the
instrument, and is designed so that an earthquake 100 kilometers from your
station will have a maximum amplitude of 1 millimeter if the quake's magnitude
is 3.0.  The important thing to note here is that that number, ML = 3.0, is
completely arbitrary.  Richter could have defined it to be anything he wanted
it to be -- 0.0, 1.0, 10.0, whatever.

	You have now read the maximum amplitude (in mm) from your record and 
have determined the distance from the quake to your location, so now you look up
the distance correction factor, take the log (base 10) of the max amplitude,
add those two numbers together, and you have your Richter magnitude.


Now, several important things about Richter magnitude:

	1)  Richter magnitude is physically meaningless.  
		The Richter magnitude is based solely on how a specific
		type of instrument responds to the motion of a quake, and
		tells you nothing about the physical quake itself.

		But this is OK, because, and this is critical to keep in
		mind, the ONLY reason for having the Richter Scale is to 
		have an objective means of comparing two quakes.  It is
		NOT an inherent thing of quakes -- it is an arbitrarily
		designed scale created by Richter for his convenience.

		There have been several papers written which attempted to
		find an equation linking energy release and Richter magnitude
		or any of several other physical parameters to Richter
		magnitude, but these are all EMPIRICAL relationships, not
		something inherent to the scale itself.

	2)  Richter magnitude is inaccurate, particularly for large quakes.
		This is due to the fact that, as quakes get larger, they
		release more and more energy into the ground, and this will
		eventually overwhelm the recording device's capabilities.
		For the Wood-Andersons, this happens at about ML=6.5 or so.

	3)  Richter magnitude does NOT depend on how people felt the quake.
		Posts on ca.earthquakes often contain the statement, 'I am
		in city XX, and it felt like a Y.Y.  What was it in city
		ZZ?'  These people are NOT describing magnitude.  They are
		describing intensity -- in other words, how the quake 
		affected people or objects near them.  Intensity varies
		widely with distance, what the person was doing at the time,
		and what kind of soil were they on.  Other factors enter into
		it, such as what kind of building they are in and how sensitive
		they are.

		Magnitude, on the other hand, uses only distance and amplitude.
		From station A at a distance of, say, 20 km from the quake,
		we might get an ML=4.2, while at station B at a distance of,
		say, 100 kilometers in the other direction from the quake, we
		might get an ML=4.4, and at a third station at 50 km and 
		along a third azimuth, we might get an ML=4.1.  The magnitude
		for the quake would be given as a 4.2, and that would be it.
		The intensity might vary drastically from place to place and
		distance to distance, but the magnitude would not.

	4)  Little differences in magnitude are meaningless.
		These little differences, such as those between UC Berkeley
		and the USGS in Menlo Park and Caltech, are caused by two
		major factors.  First, slight differences in the methods 
		and instruments used to measure Richter magnitude, and second,
		variations in the earthquake's energy release (i.e. more energy
		might go north than would go south, which would cause the 
		northern stations to give higher magnitudes than average, while
		southern stations might give lower than average magnitudes) as
		well as variations in the earth's crust.  They are also due to
		slight differences in reading techniques from one person to the
		next (for example, Suzanna Loper at UCB and I have found that 
		the average difference in magnitude between her mags and mine
		is about 0.1 magnitude units) and other, less important factors.

		But all of this is really meaningless, especially considering
		that Richter himself recommended only using magnitudes to about
		0.5 magnitude units.   Basically, magnitude differences from
		one group to another of +/- 0.2 magnitude units are very 
		common, and essentially meaningless.

	5)  The magnitude scale is logarithmic in amplitude.
		A very pompous way of saying that, for each whole unit 
		increase (or decrease) in magnitude, the recorded maximum
		amplitude goes up (or down) by a factor of 10.  So a quake
		with ML=4.0 has 10 times higher recorded amplitude than a 
		quake with ML=3.0 and 10 times lower than a quake with ML=5.0.

		An EMPIRICAL relationship between ML and energy release has
		been found and basically states that, for each increase (or
		decrease) in magnitude of a 1.0 unit, the energy released
		by the quake goes up (or down) by a factor of about 32.  But
		keep in mind that this is something found by calculating the
		energy and magnitude individually, and then trying to connect
		them, not something inherent in the magnitude scale itself.

Whew!

Now, to complicate the issue further, there are many different types of
magnitudes in use by seismologists, each with its own purpose.  The Richter
magnitude is the most familiar to the public because it is the one that gets
reported to the press for local quakes.  And all of the scales, except one,
give you no physical information about the quake, and are only designed to
help compare one quake to the next.

That one exception is moment magnitude, which is based not on the instrumental
recordings of a quake, but on the area of the fault that ruptured in the quake.
This means that the moment magnitude does tell you something physical about 
a single quake.  But that's another post, and there are others better suited 
than me to do that one.

Hope that helps and doesn't confuse anybody.

Greg Anderson

_____

From: anderson@mahi.ucsd.edu
Subject: Rates of Earthquakes

Recently, someone asked about the rate of occurrence for earthquakes
of a given size in California.  Dr. Robert Uhrhammer at UC Berkeley has
examined the UC Berkeley catalogs for the past forty years, and has
come up with some statistics on rate and probabilities for earthquakes
of a given size or larger in certain areas of California and for California
as a whole. Below, I have summarized the results in a few
tables.  Thanks go to Bob for the data, and any and all mistakes (either
typos or, more importantly, errors in interpretation) are mine.

Below, I present the tables for the Central Coast Ranges of California,
a region that generally extends from Santa Rosa in the north to San Luis
Obispo in the south, Northern and Central California (north of the transverse
ranges), and California overall.  These tables give the rate of earthquakes
per year and the probability of earthquakes occurring in one hour, day, week,
month, and year, for a range of magnitudes.

If you want all the details on how these tables are generated, they are at
the bottom of this section.

CENTRAL COAST RANGES, CALIFORNIA

Having looked at all the earthquakes with M >= 2.5 that were recorded by the
UC Berkeley network between 1949 and 1988, and declustered them into about 
3000 groups, the following equation was found:

	log N = 3.97 - 0.832*M

This can be translated into the following table:

 				Probability in one
 ML	 eqs/yr		 hour	  day	 week     month	    year
_____________________________________________________________________
2.5	 77.625		0.8816	19.1465	77.5254	 99.8449  100.0000
3.0	 29.785		0.3392	 7.8313	43.6051	 91.6432  100.0000
3.5	 11.429		0.1303	 3.0807	19.7308	 61.4186   99.9989
4.0	  4.385		0.0500	 1.1935	 8.0875	 30.6110   98.7541
4.5	  1.683		0.0192	 0.4596	 3.1841	 13.0835   81.4124
5.0	  0.646		0.0074	 0.1766	 1.2340	  5.2383   47.5681
5.5	  0.248		0.0028	 0.0678	 0.4753	  2.0434   21.9439
6.0	  0.095		0.0011	 0.0260	 0.1826	  0.7890    9.0682
6.5	  0.036		0.0004	 0.0100	 0.0701	  0.3035    3.5818
7.0	  0.014		0.0002	 0.0038	 0.0269	  0.1166    1.3898


NORTHERN AND CENTRAL CALIFORNIA

Having looked at all the earthquakes with M >= 3.0 that were recorded by
the UC Berkeley network between 1949 and 1988, and declustered them into
about 2900 sequences, the following equation was found:

	log N = 4.44 - 0.857*M

which translates to the following table:

 				Probability in one
 ML	 eqs/yr		 hour	  day	 week     month	    year
___________________________________________________________________
3.0	 73.961		0.8402	18.3312	75.8846	 99.7895  100.0000
3.5	 27.574		0.3141	 7.2716	41.1554	 89.9524  100.0000
4.0	 10.280		0.1172	 2.7754	17.9380	 57.5431   99.9966
4.5	  3.833		0.0437	 1.0439	 7.1054	 27.3407   97.8348
5.0	  1.429		0.0163	 0.3905	 2.7105	 11.2258   76.0426
5.5	  0.533		0.0061	 0.1457	 1.0192	  4.3422   41.2995
6.0	  0.199		0.0023	 0.0544	 0.3812	  1.6415   18.0130
6.5	  0.074		0.0008	 0.0203	 0.1423	  0.6151    7.1371
7.0	  0.028		0.0003	 0.0076	 0.0531	  0.2298    2.7228

CALIFORNIA

This table is likely to be somewhat of an underestimation, because of
the fact that some magnitude 3.5 or smaller earthquakes may have been missed.
But, assuming the data are complete, the equation

	log N = 4.816-0.82*M

is found, which translates to the following table:

 				Probability in one
 ML	 eqs/yr		 hour	  day	 week     month	    year
___________________________________________________________________
3.0	226.986		2.5562	46.2847	98.7287	100.0000  100.0000
3.5	 88.308		1.0024	21.4772	81.6993	 99.9363  100.0000
4.0	 34.356		0.3912	 8.9775	48.3504	 94.2902  100.0000
4.5	 13.366		0.1524	 3.5933	22.6661	 67.1701   99.9998
5.0	  5.200		0.0593	 1.4136	 9.5162	 35.1653   99.4483
5.5	  2.023		0.0231	 0.5524	 3.8157	 15.5140   86.7744
6.0	  0.787		0.0090	 0.2153	 1.5022	  6.3483   54.4812
6.5	  0.306		0.0035	 0.0838	 0.5871	  2.5194   26.3758
7.0	  0.119		0.0014	 0.0326	 0.2288	  0.9878   11.2302
7.5	  0.046		0.0005	 0.0127	 0.0891	  0.3855    4.5287


Now, armed with these tables, you can ask yourself a very important
question regarding earthquake 'predictions'.  How good is the prediction?
How probable is it that the predicted quake will happen just by random 
chance in the given interval of time?  

A good earthquake prediction will have the following components:

	(1) a specific geographic location
	(2) a specific time window
	(3) a specific magnitude
	(4) a probability estimate, and how it was determined

A specific geographic location means, for example, 'in the area within 
20 km of Cape Mendocino' or 'within 15 km of San Jose'.  Saying something
like, 'somewhere between San Francisco and Mexico' or 'on the eastern
side of the Sierra Nevada' won't cut it, because it's by far too large
an area.

A specific time window means, for example, 'between 20 July and 22 July, 
1993', not 'sometime within the next three months'.  

A specific magnitude range means, for example, 'between 4.0 and 4.5 on the
Richter scale' or 'larger than 6.5 on the Richter scale'.

And the probability estimate means, 'There is a 65% chance of this happening.'
And it's important that the person doing the 'predicting' explain how he or 
she arrived at the probability estimate.  Simply guessing that there is, oh, 
about a 65% chance is not good enough for people to judge the prediction.

So a good earthquake prediction will read something like this:

	There is a 65% probability that there will be an earthquake of
	magnitude approximately 5 on the Richter scale within 20 kilometers
	of Cape Mendocino between 20 July and 22 July, 1993.  The probability
	estimate is based on my historic success rate, as shown by the
	published predictions I have made.

This kind of a prediction gives you enough specific information to (a) decide
if you believe the person *before* the quake, (b) decide if it's worth worrying
about or making some sort of change in your life, and (c) test it when it 
either does or does not happen.

A bad earthquake prediction reads something like this:

	On 24 April 1992, there was a magnitude 6 earthquake near Cape 
	Mendocino.  I predicted this quake, saying that there would be
	a magnitude 6 earthquake somewhere between Japan and Washington
	State.

The reasons this one is bad should be obvious, but in case they aren't, here 
are some of them:

	(1) the 'prediction' is announced after the fact
	(2) the geographic area is much too large to be of any use,
		and in fact is so large as to almost guarantee the success
		of the prediction by chance.
	(3) there is no time estimate.  Magnitude 6 earthquakes happen 
		with regularity in this zone, and not giving a time window
		practically guarantees success (unless the quakes simply
		stop happening).
	(4) there is no probability estimate.

Anyway, sermon over.  With the above tables, when someone 'predicts' a quake in
California, you can take the time and magnitude window given and estimate the
chances that such an earthquake would happen simply by chance.  So, for 
argument's sake, let's say that I make the prediction I did above for Cape
Mendocino between the 20th and 22nd of July, 1993.  Look in the table for
Northern California, with magnitude equal to 5.  The table says that on an
average day, there is a 0.39% chance of such an earthquake occurring somewhere
in Northern and Central California.  Given that we are talking about a much
smaller area, a 400 square kilometer area, there is a much smaller chance that
such an earthquake will happen at random.  So, if I make this prediction, and
the earthquake does come, it is a strong indication that my 'prediction' method
works.  If I do this a number of times, and succeed a large percentage of the
time, it looks good for my method.

Whew!  This turned out to be considerably longer than I had expected.  Once 
again, I should thank Dr. Robert Uhrhammer for his data and help.

I should also state (although I would hope it is obvious) that the opinions 
in here are mine, and not UC Berkeley's, and not those of the Scripps 
Institution of Oceanography, or of my mother, of of your mother, or of Mother 
Teresa.  At least, not that I know of. ;-)

****
Now, here are the details on how the tables are generated:

Simply munging a catalog together and counting earthquakes of a given size in
a given length of time will give you a false idea of the average rate of 
seismicity.  The reason is that simply counting quakes doesn't take into 
account the fact that many of the earthquakes will be aftershocks of some other
principal earthquake.  What you really need to count are the principal 
earthquakes, because doing otherwise will bias your counting.  This process of
deciding whether a given earthquake is a principal earthquake or part of an
aftershock sequence is called 'declustering'.

Once the data are declustered, you make a graph with the logarithm (base 10) 
of the cumulative number of earthquakes of a given size or larger in a given 
period of time on the y-axis, and magnitude on the x-axis.  Since the number 
of quakes goes up rapidly with decreasing magnitude, your chart will look 
something like this:

	|
	|
	| *
	|   *
 log N	|      *
	| 	*
	|         *
	|           *
	|_____________________
		  M

Then you attempt to fit a straight line through the data, with the form:

	log N  =  a - b*M

This leads to 

	rate = 10^(a - b*M),

where 'rate' is defined as the number of earthquake sequences of a given 
magnitude or larger in a given period of time, usually one year.

(Of course, all this is really done using a computer, using a number of 
different line-fitting methods.)

OK.  Now, we have a way of calculating the rate at which earthquakes of a given
size or larger occur, on average, in the given area.  For a concrete example,
take a look at the "Northern and Central California Seismicity" table.  From
this table, you can see that there are, on average, 1.44 principal earthquakes
with magnitude 5.0 or larger in Northern and Central California in a given year.
Or that there are, on average, about 10.3 principal earthquakes with magnitude
of 4.0 or larger in Northern and Central California in a given year.

Now, you are ready to answer the following question:

	What is the probability that an earthquake of a given size or larger
	will occur in my area of interest in a given amount of time?

To answer that, you have to make one final assumption, and that this that the
earthquakes are somehow distributed in time.  Usually, it is assumed that 
earthquake occurrences in time follow the Poisson distribution (for an 
explanation of the Poisson distribution, I refer you to the nearest probability 
text, since I'm highly likely to get the explanation wrong, anyway.)  Suffice 
to say that, because of the structure of the Poisson distribution used in the 
tables above, with a high rate of seismicity for a given magnitude, the 
probability rapidly approaches 100%, while for a lower rate, it takes much 
longer.  An equivalent way of saying that is that, the lower the magnitude 
you look at, the faster the probability of an earthquake of that magnitude 
happening in a given time rises to nearly 100%.

Greg Anderson

_____

From: tcsmith@netcom.com (Ted Smith)
Subject: CDMG BBS
Robert Stroh wrote:

Not only do I have it, but here are all the other GeoInfo Net BBS
numbers, too.  We'll be posting an updated network list very shortly.

 Hubs:
  GeoNet BBS              316-265-6457  2400b      Wichita, Kansas
                     and  316-265-1994  14.4K bps
  GeoFuel GeoScience BBS  416-829-4097  16.8K bps  Oakville, Ontario
  CDMG ONLINE             916-327-1208  14.4K bps  Sacramento, California
  Computer Solutions BBS  504-542-9600  14.4K bps  Hammond, Louisiana
  GISnet BBS              303-447-0927  14.4K bps  Boulder, Colorado
  Dark Matter             604-534-7667  14.4K bps  Langley, British Columbia

 Nodes:
  PreCambrian BBS         602-881-5836  14.4K bps  Tucson, Arizona
  Computer Plumber BBS    319-337-6723  14.4K bps  Iowa City, Iowa
  PSN San Jose            408-226-0675  2400b      San Jose, California
  PSN Pasadena            818-797-0536  2400b      Pasadena, California
  PSN Memphis             901-360-0302  2400b      Memphis, Tennessee
  COGSNET BBS             303-526-1617  14.4K bps  Golden, Colorado
  SurveyNet BBS           207-549-3213  14.4K bps  Whitefield, Maine
  Snowshoe BBS            304-572-2531  14.4K bps  Snowshoe, West Virginia

_____


From: Richard Stead

> >: The traditional Richter measurement is based on the amplitude of
> >: surface movement.  
> >
> >Officially, it's based on the movement _on_paper_ of a particular
> >class of seismograph.
> >
> 
> This is semantics...
> All scientific measurements are based on instruments which were designed
> to report "natural" events.

No.
The Richter magnitude has a very specific definition.  There is a particular
instrument called a Wood-Anderson torsion seismometer that the measurement
must be made on.  This instrument only functions as a horizontal seismometer.
It has a wire stretched vertically to which a small cylindrical weight is
welded (against the wire, not with the wire through the middle).  There
is also a small mirror mounted.  A beam of light is directed at the mirror
and reflects onto a rotating drum over which a sheet of paper film is mounted.
When the earth moves horizontally, the inertia of the weight creates torsion
in the wire, deflecting the beam of light from the mirror.  The weight and
wire are tuned such that the instrument has a natural period of 0.8 seconds.
(The 0.8 version must be used for Richter magnitude, but other periods
have been used, particularly 6.0 seconds).  The computation of Richter magnitude
begins by measuring the peak amplitude of the deflection on the paper film.

Today, of course, real Wood-Anderson seismometers are rarely used.  Instead,
we use instruments that are much more capable.  To compute a Richter magnitude,
however, we first convert the records of these instruments to exactly what
an 0.8 second Wood-Anderson torsion would have recorded on its film.  We then
measure the peak amplitude and procede from there.  We can do this only because
modern instruments include all the information the old Wood-Anderson would
have recorded, but more in addition.  The best are called broadband wide
dynamic range seismometers.  Kinda sounds like the promotions you see on
certain records or CDs.  A modern instrument will respond in a band from
microHertz to 100 Hertz and have a dynamic range of over 150 dB.  The
Wood-Anderson isn't even close to that.  It has a useful range of about
0.01 to 10 Hz and a useful dynamic range of about 30 dB.

The limits of the Wood-Anderson cause the Richter magnitude to "saturate"
at about M8.5-8.7.  It can never get bigger no matter how big the quake
really is.  The reason for this is something called the "corner frequency".
As quakes get bigger, a given site will no longer receive more and more
of the higher frequencies.  The lowest frequency that is constant in this
respect is the "corner".  Frequencies below that continue to increase as
quakes get bigger.  Really large quakes have to be distinguished at frequencies
in the range of 0.01 to 0.001 Hz (and even lower - the energy at DC is
what is most useful).  These frequencies still cause ground motion,
but it is not perceptable and not damaging except at very close range to
the fault.  The Kanamori-Richter or open-ended Richter scale addresses
this problem, and basically eliminates the physical Wood-Anderson
instrument from the computation.

Modern seismology prefers Mw as a magnitude measure for the large and
great earthquakes.  It is computed from the estimate of energy released
by the quake.  This is the measure that makes Chile, 1960 the largest
quake ever at Mw=9.6.  All magnitude scales are adjusted to be in good
agreement to Richter over the range that Richter is most applicable,
so Mw=9.6 is like saying the quake would be a Richter 9.6, if Richter
could measure that high.

_____

From: stead@seismo.CSS.GOV (Richard Stead)
Subject: Re: Pacific Northwest questions

In article <27t0p9INNpk5@darkstar.UCSC.EDU>, andy@cse.ucsc.edu
(Andy John) writes:
> My parents live in Seattle, and have recently heard about the subduction
> zone etc... They said that the entire area is bowed upwards about a meter
> or so, due to the plate tectonics.  I think it was the plate coming in 
> from the ocean, hitting a mountain range, and being

That's not the best way to picture it.  The plate is going down because it
is unstable on top.  Unlike ice, which is lighter than water, solidified
mantle material (the ocean crust) should not float above the viscous mantle.
This is complicated by the fact that the mantle is also solid on short time
scales, and that pressure from the miles of rock above it make the mantle
more dense than the ocean crust.  When you can move ocean crust into the
mantle, the pressure also makes the crust more dense, and then it is denser
than the mantle at the same pressure.  Thus, it will sink if it starts going
down.  (But sink slowly, the timescale over which the mantle is fluid is
millenia).  Thus, no mountain is necessary to push the plate down.  In fact,
they can just start going down on their own - there are several subduction
zones that are out in the middle of the ocean.  However, subduction can
create mountains two ways.  First, the force of collision, if a plate wants
to move over a subduction zone, can cause the over-riding plate to buckle
and thus produce mountains.  Second, as the plate "de-waters" (the ocean
crust under the ocean undergoes chemical reactions in which water is
incorporated in the chemical composition of the minerals, for example,
in the formation of hydrates and hydroxides - heat and pressure release
the water), the water lowers the melting point of the surrounding rock, and
the now molten rock moves to the surface.  It breaks through in the form
of strato-volcanoes, characterized by explosive eruptions.  This always
happens as the down-going slab reaches a particular depth.  The slab rarely
goes down straight, so the volcanoes form in back of the subduction zone on
the over-riding plate.  The Cascades are an example of this kind of volcanic
activity.

> pushed under. They said there could be a magnitude 9 earthquake, and the 
> area could settle about a meter!!!!!!!! Is this true!!!!!! Is it even 
> possible to survive a 9?

Actually, I would expect more than 1 meter maximum subsidence in a magnitude 9.
Certainly, areas of Chile in 1960 and Alaska in 1964 had more than 1 meter.
It is certainly possible to survive a 9.  As far as what you can feel or most
damage it would cause, a 9 is no different than an 8.  It just covers more area
and lasts a lot longer.  Of course, subsidence is larger in a 9, and thus more
land can get flooded.  Also, there is more damage, because some structures
can outlast the shaking of an 8, but when asked to survive the same shaking
for a few more minutes, just aren't up to it.  Finally, and more significantly,
tsunamis are likely to be much worse in a 9.  That's the big trouble.
If you're up on a nice high, flat plateau in a 9, just lie back and enjoy
the experience.  You will be in no danger, and you will get to experience
one of the rarest natural wonders.  I won't vouch for you safety in a building,
under a cliff or even among trees.  Trees snap off, cliffs fall and buildings
may collapse under the shaking.  Trees and cliffs can't be helped, but there
is no reason buildings should not withstand a 9, it is only bad design that
causes them to collapse.

_____

Article 2687 of ca.earthquakes:
From: ho@helen.CS.Berkeley.EDU (Kinson Ho)
Subject: Re: Sliver Creek Fault in San Jose

[I submitted the following info to the former/current maintainer of
 the earthquake FAQ ages ago, but apparently it was decided that
 the info is not relevant.  I was never told why.]


If you are planning to buy a house in the Bay Area, 
I would strongly recommend a visit to the ABAG (Assoc of Bay Area
Governments) building in Oakland, near Chinatown.  There are
two places of interest:

a. BAREPP (Bay Area Regional Earthquake Preparedness Project)
   101 8th (510-540-2713)

   They have lots of handouts, a huge technical library, and perhaps
   lots of experts right there.  They will also send you a "home pack"
   if you call them up.

b. ABAG library (in the same building)
   They have maps of fault lines, maps of regions with different 
   magnitude of ground shaking in case of a quake, and earthquake
   planning scenario documents (i.e. what the experts think will
   probably fail in the BIG ONE) etc.  You may be able to buy some
   of these in one of the offices there.  Ask the librarian for help.


   Plan to spend a LOT of time there.  BTW, they get a lot of visits 
   from people who are planning to buy a house. 


Welcome to earthquake country...


Kinson Ho

_____

Article 2947 in ca.earthquakes:
From: Ted Smith
Subject: Lone Pine Scarp and Special Studies Zones

K>Many years ago, on a geology field trip, I went to see the scarp
 >of the 1872 quake near Lone Pine. I'd like to go there again, but can't
 >remember exactly where we went. I looked in my copy of Earthquake Country,
 >and it says "Take the Whiney Portal highway west out of town. After crossing
 >the Los Angeles aqueduct, walk (or drive on jeep roads) about a quarter-
 >mile north to where a very obvious 23-foot east-facing scarp cuts
 >across an alluvial deposit of a former channel of Lone Pine Creek".
 >Unfortunately, this book is 30 years old.  Are these directions useful, or
 >does someone have more up-to-date or accurate information? Any help
 >would be appreciated. Thanks.

The fault has been zoned as a fault-rupture hazard by the California
Department of Conservation's Division of Mines and Geology.
1:24,000-scale maps are available showing the zones and the faults.  An
index to the maps appears in DMG Special Publication 42 ("Fault-Rupture
Hazard Zones in California").  Copies of the maps are available at
county planning offices and from Blue Print Service Company, 1157
Mission Street, San Francisco, CA 94105; phone 415-495-8700, ext. 550.

_____

Article 2933 of ca.earthquakes:
From: stead@seismo.CSS.GOV (Richard Stead)
Subject: Re: Circular quake diagrams???

In article <2c8jr2$8ni@darkstar.UCSC.EDU> andy@cse.ucsc.edu (Andy John) writes:
>Is there a easy to explain relationship between those circular
>diagrams that always come with quakes? Is the plane the figure is
>in supposed to be the surface of the ground? From the figure can
>you tell where shaking does and doesn't happen?

I haven't been reading for a few days, but it looks like no one answered this.

The diagram is informally named a "beachball".  The plane is the ground (though
I have seen them used in scientific papers with the plane oriented differently).
The diagram, given that the plane is the ground, represents the lower
hemisphere of the focal sphere.  The focal sphere is an imaginary sphere
large enough to completely contain the fault rupture.  The diagram represents
the permanent distortion of that sphere that the quake created.  Normally,
colored areas represent zones that now protrude outside the original sphere,
and the other area has intruded toward the center of the sphere.  The
boundaries between these are called the nodal lines.  They still intersect
the original sphere surface.

So you can see, since it is just a hemisphere, the plane could be other
than the ground - for example, it might be useful to see it from the side,
or oriented at another angle, like the surface of a subducting slab.

The sphere is not determined by actually measuring displacements of a
spherical surface.  Instead, the relative motion of this surface is
projected from the observation of the motion of waves as recorded at
sesimometers.  Many simplifying assumptions are made:
1) at the very least, the "moment tensor" is assumed to be symmetric.
   This allows a hemisphere to represent the entire sphere.
   The asymetric information must be carefully extracted and evaluated,
   but it is valuable for identifying the fault plane, directivity, etc.;
   but it is difficult to do and rarely used.
2) normally, the earthquake is assumed to have a point source.  This means
   that all the rupture happened at a single point.  This never happens,
   but is reasonable to first order.  Some seismologists study how the
   mechanism changes as a function of time and position on the fault -
   in that case they really do break the fault and the quake process
   into individual points in space and time.
3) Often, the mechanism is examined in terms of the best-fitting "double
   couple".  The double couple is the theoretical exact motion of a
   point source fault.  This makes the nodal lines run precisely along
   lines that divide the sphere into hemispheres ("great circles") and
   that represent planes that intersect each other at exactly 90 degrees.
   On an ordinary fault, one plane would then be the fault plane and the
   other is called the "auxillary plane".  In most faulting environments,
   the stress producing the faulting is equally satisfied by slip on
   either of the two planes.  This is why so many faults are at right
   angles to each other in CA.  The most noticeable is the San Andreas
   and the Garlock, but Landers/Big Bear counts and the Superstition
   Hills/Elmore Ranch quakes also did this.
   You will notice that Doug's mechanism for the recent Parkfield quake
   did not show the double couple, but showed the full moment tensor.
   However, you can easily see where the double couple lines would go,
   the mechanism is mostly double couple.  The other component is
   "CLVD" - compensated linear vector dipole.  This is a pure compression
   or extension motion.  Explosion/implosion could be part of a moment
   tensor, but it is normally factored out, particularly when drawing
   focal spheres.

Up to this point, I have described only one kind of focal sphere plot.
This is that for radial distortion of the sphere, and is appropriate
for P-waves.  The shear distortion may also be represented in the same
way, except the different areas become twists in opposite directions.
Since shear is two-dimensional, while radial distrotion is only
one-dimensional, the shear motion must be separated into to focal sphere
plots.  This is normally done for shear parallel to the ground (SH for
shear horizontal), and normal to the ground (SV).  These are much more
difficult to describe in netnews.  The moment tensor plot for radial
distortion contains sufficient data to extract the SH and SV spheres,
assuming an elastic solid, so the mechanisms are generally used by
seismologists only to visualize the shear wave patterns, rather
than represent any information the P-wave one can't.  (The focal sphere
is often also called a "radiation pattern" - by analogy to radiation
patterns of antennae.  In fact, a map of the world can be projected
onto the focal sphere, given a propagation model, to determine first
motion anywhere in the world).


You also ask if you can tell where shaking happens.  No.  First, shaking
happens everywhere.  However, even if we were imbedded in a perfectly
uniform, infinite elastic solid, and the quake was a point source, there
would only be a single infinite line along which there was no motion -
that line is the intersection of the two planes of the double couple.
The nodal line for the P-wave is the peak for the S-wave and vice-versa.
(The case for a full moment tensor is tougher).
Beyond purely theoretical considerations, the earth is not infinite or
uniform.  This causes energy to refract, reflect and diffract around such
that, while a point may miss the first motion, it will certainly get
plenty of shaking after that.  However, since the S waves are stronger
than the P, it is better to be nodal to S - shaking will be a small bit
less.
What the figure really tells you is what the first motion is expected to
be everywhere, but things get a lot less certain after that first swing.

_____

Article 3049 of ca.earthquakes:
From: jtchew@csa3.lbl.gov (Ad absurdum per aspera)
Subject: PHYSICS NEWS UPDATE #156, 24 Dec 93

[Written by the American Institute of Physics and posted by us.
Respond to  or other references below.  Always
posted here on sci.research; sometimes crossposted to other
interested groups with followups directed here. Back issues,
along with FYI and the American Physical Society news/opinion
column WHAT'S NEW, are archived on NIC.HEP.NET for your anonymous
FTP'ing pleasure.  

I won't be logging back on until early January, so this might be
your last PNUp for a while.  Have a merry Christmas, happy Hannukah,
rootsy Kwanzaa, enchanted solstice, or poignant seasonal feast day
appropriate to your religion, ethnicity, or other identity group
(this *is* coming to you from Berkeley, you know! :)  -jc]

GEOELECTRIC SIGNALS: DO THEY PRECEDE
EARTHQUAKES?  Speaking at the recent meeting of the American
Geophysical Union, Anthony Fraser Smith of Stanford (415-723-
3687) reviewed his data from four years ago which showed that local
measurements of Earth's magnetic field fluctuated much more
vigorously than usual in the days and hours before the 7.1-magnitude
Loma Prieta earthquake.  Many scientists hesitate to infer any
correction between the signals and the quake, particularly on the basis
of only one such data set.  To study the matter further, Fraser Smith
has set up several detectors around California near faults. Simon
Klemperer, also of Stanford (415-723-7344), attempts to model Fraser
Smith's signals by suggesting that in the buildup to a quake, a flexing
fault system might squeeze pockets of water together (which are
sparse at these depths--18 km), altering the electrical conductivity of
the fault, which in turn can act like an antenna to modulate the
measured magnetic field at the surface.  Other types of geoelectric
signals possibly related to quakes were reported at the AGU meeting.
Seiya Uyeda of Tokai University in Japan and Texas A&M cited data
linking four recent earthquakes in Japan with anomalies in the static
voltage differences between various measurement stations.  Jean Chu
of MIT presented a small portion of an extensive Chinese study (over
20 years) of earthquakes and possible precursors in the form of
changes in the conductivity of the Earth.

_____

From: shirriff@cs.berkeley.edu (Ken Shirriff)
Subject: magnitude & power

This posting is a long explanation of scales for earthquake magnitudes.
In short, the magnitude is the base 10 log of the ground movement
amplitude, with a bunch of fudging to make results come out the
same at different measuring sites and to make the results comparable
between different earthquakes.

In general, the magnitude is determined by a formula of the form
mag = log(a/T) + f(delta,h) + Cs + Cr
where mag is the magnitude, a is the ground amplitude in microns,
T is wave period in seconds, f is a function to correct for the
effects of distance and focal depth, delta is the epicentral distance
in degrees, h is the earthquake focal depth in km,
Cs is a correction for the local structures at the station and
Cr is a regional correction.

The original Richter scale was designed in 1935 for comparing local
earthquakes in Southern California and cannot be used directly for
comparing earthquakes in other areas.  It is defined as M sub L =
the logarithm of the maximum recorded trace amplitude in microns
of a Wood-Anderson torsion seismograph with specified constants
(free period=0.8s, magnification=2800, damping=0.8) at an epicentral
distance of 100 km.  (Note: all logarithms are base 10).

The magnitude M was designed in 1945 by Gutenberg based on surface
waves.  Considering Rayleigh surface waves in a period range of
20+-2 sec for earthquakes of normal depth, the equation becomes M
= log a + c1 log delta + c2, where a is the horizontal ground
amplitude in microns, delta is the epicentral distance in degrees,
and c1 and c2 are constants.  (As best as I can tell, this is the
magnitude normally reported.)

There is a third magnitude m, similar to M, based on body waves.

These magnitudes are related by:
m = 1.7 + 0.8 ML - 0.01 ML^2 (where ML is M sub L)
m = 0.56 M + 2.9

The International Geophysical Assembly in Zurich in 1967 adopted the
following recommendations for magnitude determinations of distant events:

1.  Magnitudes should be determined from (a/T)max for all waves for
    which calibrating functions f(delta,h) are avaliable: PZ, PH, PPZ, PPH,
    SH, LH, (LZ).  (Z=vertical component, H=horizontal component,
    L=surface wave).
2.  Amplitudes and periods used ought to be published.  Two magnitudes
    (m=body-wave magnitude, M=surface-wave magnitude) should be used.
    For statistical studies M is favoured.  The conversion formula m =
    0.56M+2.9 is recommended.
3.  For body waves the f(delta,h) values of Gutenberg and Richter are
    used.  For surface waves, the Moscow-Prague formula:
    M = log(a/T) + 1.66 log(delta) + 3.3 is used.
    (a is the horizontal componente of Rayleigh surface waves;  T should
    be in the period range of 10-30 sec.)
4.  If short period records are used exclusively, too low magnitudes
    result.  In order to eliminate this error, it is strongly recommended
    that for short-period readings either a/T or f(delta,h) be
    adjusted such that the agreement with long-period instruments is
    achieved.

The energy in ergs released by the earthquake is given by:
log E = 12.24 + 1.44 M, where M is the magnitude > 5.
This is an empirical equation, derived by integrating over the whole wave
train in time and space.  From this equation, each increase of 1 in
magnitude increases the energy released by a factor of 27.5.

The maximum acceleration a0 (in cm/sec^2) is related to magnitude by
M = 2 + 2 log(a0).

The information in this posting came from "Introduction to Seismology",
Markus Bath, 1973, John Wiley&Sons, New York.  This book explains
most of what I wanted to know, with lots of formulas.  Check it out.


_____

Article 4101 of ca.earthquakes:
From: gnelson@megatest.com (Glenn Nelson)
Subject: Speed of Seismic Waves

Here's a brief tutorial about estimating the distance to an earthquake.

You can figure the distance to a lightning strike by timing the difference
between the flash and the thunder, then calculating based on the speed
of sound, approximately 1 mile in 5 seconds. Similar for earthquakes
except both P and S waves travel slow, whereas light is virtually instantaneous.

Here's how you figure how far away the quake is. The total travel time
for S is Ts = dist / Vs and for P is Tp = dist / Vp. The S wave arrives
after P, P wave feels like a hammer blow, mostly vertical, S is stronger
and longer and often has sideways components. Time the difference of the
two arrivals, then
(Ts - Tp) = dist/Vs - dist/Vp ==> dist = (Ts - Tp) / (1/Vs - 1/Vp).

If you are within 80-100 km, this is a reasonable formula:
	dist = 8 * (Ts-Tp) kilometers.

The P wave travels fastest, the S wave slower. P stands for primary and
S for secondary, or maybe P is for pressure (the type of wave) and S for
shear. In normal matter shear waves always travel slower. In the earth's
crust, 25 to 50 km thick on the continents, P velocity is close to sqrt(3)
faster than S (that's the year George Washington was born). In California
we often use 1.75 or even 1.78. In California, for distances up to
about 40 km, the P waves travel about 6 km/sec, after that, maybe 7 km/sec.
In the formula above you substitute Vs=Vp/1.75, then choose some number
for Vp, I usually use 6. Substitute above to get the simplified formula.


>From chucko@rahul.net Thu Mar  3 22:16:39 1994

I tried plowing through all three sections of the FAQ on my little 24x80
home terminal.  It seemed to me there was a great deal of repetition
about the basics (e.g. intensity vs. magnitude, why the Richter scale is
not particularly accurate for measuring large quakes, etc.).

I think some of this material would be best made available via FTP, and
the FAQ should contain a pointer to that site and a very brief summary
of all that, but then I have high-bandwidth Internet access at work and
am spoiled.

In any event, thanks for organizing this.
 -- Chuck

>From hannah@ai.sri.com Thu Mar  3 10:16:43 1994

   I am posting this to ask for opinions on the structure of the
   FAQ for ca.earthquakes, which is cross-posted to sci.geo.geology.
   The FAQ has grown to a fairly large size, now consisting of three
   individual parts, the last of which contains the new sections added
   this month.  I am soliciting opinions as to the idea of splitting
   the FAQ a little differently into N (currently three) sections, 
   which will contain the current FAQ, and posting the new sections 
   in their own individual post.  This would allow readers of the FAQ
   who are solely interested in the newest chunks to read those 
   sections fairly quickly, and skip the sections they may have already
   read.  I was thinking I would label this "new" section something
   like 

	   [lm mm/dd/yy] Earthquake FAQ New Sections

   to allow easy identification.

   Greg Anderson

I agree that the current version of the ca.earthquake FAQ is a little
cumbersome.  There's good info in there, but the FAQ has "just
growed", and I suspect that the info in it could be presented in a
more usable fashion.  Currently, topics are all mixed together, so a
newbie might have to wade thru a lot of incomprehensible (to him)
technical stuff to find the answer to a common "civilian" question.
Topics are in chronological order, so long-time readers have to wade
thru a lot of old info to find what (if anything) has been added.
I've seen this handled in a couple of different ways on other
newsgroups.

The FAQs for rec.pets.dogs (and .cats) are maintained by a single
individual, who saves interesting information from net discussions and
"weaves" it into the text of the FAQ (which reads more like a pamphlet
than a compendium of messages).  The first section of the dog-FAQ
(which is up to something like 10 or 15 files, now) gives an outline
of what is in the FAQs, as well as their names and how to access them
via FTP or the e-mail request version thereof.  Each FAQ subsection is
pretty much a self-contained document on a single topic (such as "Your
New Dog"), with some references to the other FAQs.  Each begins with a
statement of when it was updated last, with changes marked by "|"s in
the margin.  Thus, when the dog-FAQs are posted, I can glance at the
beginning of each one and decide whether it has anything new, then use
my text editor to search for "|", and read only the new sections.  I
believe that the FAQs are posted monthly, with a weekly posting that
summarizes what they are, where they are, and how to get them.  The
keeper of the FAQ has an abbreviated version of the instructions in
her .sig lines, so every time she posts, folks get a clue about the
FAQ.

The FAQs for rec.equestrian are rarely posted.  What is posted
semi-regularly is a FAQ-FAQ, which tells what topics have FAQs, and
how to access them.  Each of these topics is a file, more like the
current ca.earthquakes FAQ, that mainly quotes someone's posting(s),
with only minimal cleanup.

I'm mainly a consumer of information on ca.earthquakes, from the
perspective of a scientist in a different field, with the usual
"civilian" experiences with Bay Area quakes over the last 24 (yikes,
how time flies!) years.  Consequently, I'm not in a good position to
tell you how best to subdivide the technical information.  However, I
think that there should be a shorter, less technical document written
just to answer the obvious questions (and calm the emotions) about how
quakes happen, aftershock sequences, Richter vs Mercalli (apologies
for botched spellings), suggested reading (my personal favorite is
Yanev's "Peace of Mind..."), etc.  Perhaps the very capable ladies at
Caltech have some "canned" verbage they could donate?

Certainly, a FAQ on preparedness is in order---things like what to
have on hand, things to do in advance to minimize structural damage
and injury, etc.  Is there a possibility of getting an on-line version
of the widely-distributed pamphlet that USGS prepared after Loma
Prieta?  (I realize it'll lose something in the translation to ASCII.)

The technical discussions are interesting, but probably should go in a
separate file(s), which will probably be skipped by the scientifically
challenged, and those in a panic over an all-too-recent temblor.

Well those are my thoughts.  Just what you needed---lots of "advice",
and no volunteers.... ;-)

>From ho@helen.CS.Berkeley.EDU Wed Mar  2 14:55:50 1994

I would suggest splitting the FAQ into different posts, by subject.
e.g. Magnitudes, why earthquakes happen, preparedness etc.
There does not seem to be too much of the "car kit" or "home kit"
lists included.  (I've seem them posted by individual readers
from time to time --- perhaps we should integrate them.)  

I would also suggest that each FAQ posting includes some form
of Index/table of contents, right at the beginning.

Thanks for maintaining the FAQ.

>From iva@monty.rand.org Thu Mar  3 09:56:38 1994

I like the idea of "FAQ new sections", perhaps with periodic (every 
three or six months?) reworking of the basic FAQ.

>From simutis@ingres.com Thu Mar  3 08:50:00 1994

Yes, breaking it up seems the right thing to do.

Possibly a 'major changes' post, with the info getting grouped
with other similar topics; I'm afraid this will mean developing
an overall and per-post table of contents...

FAQ could also use the Mercalli Scale someone posted, and Richard
Stead's explanation of the 'beachball' diagram; I have both of those
saved - except I don't have an actual 'beachball' diagram!

It's amazing how much work a 'labor of love' turns out to be - thank
you for doing the FAQ.

From: e_gs18@va.nmh.ac.uk
Subject: Re: quake-l mailing list?
Sender: news@c1.nkw.ac.uk (Ed Marchewicz)
Date: Fri, 1 Apr 1994 18:56:09 GMT

In article <1994Mar31.075406.503@amoco.com>, awrobinson@amoco.com (Andrew W. Robinson) writes:
> Elsewhere I saw a reference to an email mailing list, quake-l. Can
> anyone pass along any information about this mailing list? Is it a
> general discussion of earthquakes? How does one subscribe?

QUAKE-L was started by folks interested in the social consequences of
earthquakes, so is not technically oriented.  It carries a mix of 
first-hand accounts, occasional reposts from SEISM-L, basic enquiries 
and general chit-chat.  In recent months, a few other seismologists 
(all of them, I think, also s.g.g participants) have been active there,
so I feel less alone :-) .  The tone has been popular and the volume 
moderate (maybe 3/4 messages a day).  Subscribe by sending the message :
       SUBSCRIBE QUAKE-L 
to LISTSERV@VM1.NODAK.EDU   In the traditional LISTSERV fashion, 
subscriber lists, back issues &c are available from that address.
The LISTSERV program will send you details when you subscribe.

There are quite a few other list- and mail-servers covering the 
seismological community.  Most are rather specialised, but the following 
might be of interest to a wider community.  Note: *all* are much more 
technically focussed than s.g.g or QUAKE-L!

SEISM-L redistributes data messages from the US National Earthquake
Information Service.  It's not a discussion list, but new subscribers
all too regularly post "What's this stuff mean?" messages instead of looking
through the archives :-(.  About 3/4 *data* messages a day, so requires
serious interest!  Subscribe by sending the message :
        SUBSCRIBE SEISM-L 
to LISTSERV@BINGVMB.BITNET  (or possibly LISTSERV@bingvmb.cc.binghamton.edu)

SEISM-L has a parallel discussion list, SEISMD-L, which appears to be
moribund.

There is also a VOLCANO list from the LISTSERV@ASUACAD.BITNET. This is 
again fairly technical, and I am under the impression that list membership 
as well as posts are screened by the moderator.   The HAZARDS list can be 
obtained from the LISTSERV@LISTS.COLORADO.EDU -- this takes the form of an 
edited newsletter; also, the server appears to have been a bit shaky of late.

Russ Evans
British Geological Survey, Edinburgh             e_gs18@va.nmh.ac.uk


From: ted.smith@mtnswest.com (TED SMITH)
Subject: Re: quake-l mailing list?
Date: Fri,  1 Apr 1994 22:23:00 GMT


In <1994Mar31.075406.503@amoco.com>, awrobinson@amoco.com (Andrew W.
Robinson) wrote:

A>Elsewhere I saw a reference to an email mailing list, quake-l. Can
 >anyone pass along any information about this mailing list? Is it a
 >general discussion of earthquakes? How does one subscribe?

The following excerpts are from ftp.nisc.sri.com
netinfo/interest-groups [dated 14 June 1993]:

QUAKE-L@VM1.NODAK.EDU
     QUAKE-L@NDSUVM1.BITNET
     QUAKE-L%NDSUVM1.BITNET@VM1.NODAK.EDU

     Mailing list for discussion of the ways various national and
     international computer networks can help in the event of an
     earthquake, or the help can be enhanced.  One of the basic problems
     discussed might be network reconfigurations which would be temporarily
     required; others might be in actually putting various groups in
     electronic contact with each other.

     Public notebooks for the list will be available from LISTSERV, can be
     searched with the LISTSERV database facility (send LISTSERV the
     command info database for details), and are available via anonymous
     FTP from VM1.NODAK.EDU (134.129.111.1) after entering CD LISTARCH (use
     DIR QUAKE-L.* to see any notebooks/archives).

     BitNet or Internet users may subscribe to the list by sending a
     message or e-mail to LISTSERV@NDSUVM1 or LISTSERV@VM1.NODAK.EDU,
     respectively.  On the first line of the text or body of the message
     enter the command; SUB QUAKE-L your_full_name where your_full_name is
     your real name, not your login Id.

     Coordinator: Marty Hoag 
     


SEISM-L%BINGVMA.BITNET@MITVMA.MIT.EDU

     Seismological topics of general interest.

     To subscribe send the following command to LISTSERV@BINGVMB
     (non-BitNet users send mail to LISTSERV%BINGVMB.BITNET@MITVMA.MIT.EDU)
     SUBSCRIBE SEISM-L your_full_name To unsubscribe, send UNSUBSCRIBE
     SEISM-L

     Coordinator: Jim Blake 


 * QMPro 1.52 * Why experiment on animals with so many lawyers out there?


From: Joe Dellinger 
Subject: FAQ

	I've collected some stuff I found interesting; it's in
sepftp.stanford.edu
under
pub/geology/information

	I just collected stuff I happened to read that I thought was
interesting, or stuff that I thought would be useful and happened to get
a chance to get.

	I've been doing my small part to help out NASA the last couple of
weeks; I installed an e-mail daemon to produce SEDS-2 visual pass predictions
for those who needed individually tailored info but couldn't calculate it
themselves. (Try it out, if you want; send e-mail to
seds@montebello.soest.hawaii.edu
)
	The people who are researching the tether thought this was a good
idea but evidently couldn't do it themselves...

From: cochrane@netcom.com (Larry Cochrane)
Subject: Re: Personal Seismic Network? FAQ?
Date: Thu, 21 Apr 1994 01:40:06 GMT

First it's the Public Seismic Network. PSN has 4 bulletin boards around the
country, the numbers are:
	San Jose, Ca 		1(408)226-0675
	Menlo Park USGS, Ca 	1(415)327-1517
	Pasadena, Ca		1(818)797-0536
	Memphis, Tenn 		1(901)360-0302  

There is also a gopher site at gopher.ceri.memst.edu Port 70
I also have some of the PSN files in my ftp directory at ftp.netcom.com
in /pub/cochrane.

>If you are part of the PS network, would a 1 axis seismometer be helpful?

Most PSN stations only have one sensor, it would be nice to have three, two
horizontal, one pointed north-south and the other west-east, and one vertical 
sensor. My Lehman (a horizontal seismometer) is pointed north-south since most
of the local quakes I receive here in Redwood City (between San Jose and
San Francisco) Ca. originate north or south of me. This sensor can also
receive quakes from all over the world. I was able to get a 7.3 in 
Indonesia a few months ago. The plans to build a Lehamn are in file 
lehmansei.zip on any of the PSN systems and my ftp directory.

>How about a Endevco accelerometer (also 1 axis). 

I think that accelerometers are only good for strong motion detection.
To receive magnitude 3's 100km-200km away, or large distant quakes,
you need a sensor that is very sensitive to very small ground motions. 
I don't know if an accelerometer would be sensitive enough. One of the 
problems is that large quakes can saturate the system very easily. 
The 6.7 LA quake in January saturated my system and LA is 500km (300m) away.
The ideal station would have a set of high sensitivity sensors and a set 
to strong motion sensors. Maybe someday...

>If I can monitor, I would want to digitize the signal(s) from the sensor 
>to 8 or 10 bits depending on micro controller, is this good enuf? 

You really need 12 or more bits, 8 or 10 bits will also cause saturation
problems. There are plenty of 12 bit A to D cards out there that work
with the software on the PSN BBS's that most stations run to collect the 
data.
 
>What is the typical sampling rate?

Anywhere from 30 to 100 samples per second. I run at 36 sps and have
a hi-gain low noise 10 Hz low pass amp/filter card between the sensor 
and my A to D card. I sample at the low end because I save all info to disk.
This creates a 4.7Meg file each day. I save everything to disk because
I live in a very noisy location, a freeway about 500 yards from me and
train tracks about 2/3 of a mile from me. The software on the PSN BBS's 
only saves data when an event happens and I could not come up with
a good trigger point. I ended up writing my own data collecting 
software to get around this problem. I can then go back and see what
I got after I hear about a quake from either Andy Michael (USGS) finger
service or his weekly quake report. I then create a PSN compatible
quake file from this logged data. My system does have an alarm on it so
I do know about larger quakes right away.

BTW: I have just released a new version of my Windows PSN quake file viewer
called WinQuake, in file winqk15.zip. This new release is in my ftp 
directory or any of the PSN systems. It is also available on 
Ted Smith's CDMG ONLINE BBS system at 1(916)327-1208 (thanks Ted). I plan
to make a more formal announcement about WinQuake in a few days to 
ca.earthquakes and sci.geology.
 
Hope this helps.

Larry Cochrane
San Jose, Ca PSN
cochrane@netcom.com



From: andy@pangea.Stanford.EDU (Andy Michael USGS Guest)
Subject: Re: Earthquake Predications
Date: 27 Apr 1994 19:49:50 GMT

In article <2pkhcd$47m@search01.news.aol.com> gigi55@aol.com (GIGI 55) writes:
>A couple of years ago I heard
>something about the ability of seismologists to predict (scientifically of
>course) the possibility of an upcoming earthquake. The story, if I'm not
>mistaken, said that they (the Official Seismological Institutes here in CA)
>would start issuing some kind of warning. Something like the weathermen do. X%
>of probability of an earthquake in the next 3 days. I hope this doesn't sound
>too off the wall but although I don't remember the exact details, the story
>really stuck in my mind. Any comments?

The short term probabilities are based on the clustering of earthquakes
into foreshock-mainshock-aftershock sequences.  One of the primary features
of earthquake catalogs is that earthquakes cluster like this.  Of course,
aftershocks are much more common than foreshocks, and it is hard to tell
foreshocks from the more common background (or unclustered) earthquakes.  But
we have enough historical experience to be able to estimate the odds that
a given earthquake is a foreshock which would mean that it would be followed
by something larger.  I will resist explaining any of the math because this
is Lucy Jones' specialty.  Such forecasts are also made for the odds of
large aftershocks occurring, and these have been discussed here recently.



>Also, what does Jack Cole do for a living? Is he/was he a Seismologist? What
Jack worked in a electronics store (actually I'm pretty sure it was "The
Good Guys" which is a stereo/TV store chain that has branched out into general
consumer electronics now.


>methods do you use as an investigator to check these predictions out. And how
>do you decide what predictions are even worth checking out?

His predictions were checked out because they were being reported on in the
press and were starting to create public concern.  It was the effect of his
predictions on the public that got us interested.

To investigate the methods I got together a team that consisted of one
of our electronics people that has followed electrical methods and
earthquake prediction for many years, Tony Fraser-Smith from Stanford
who has been involved with ultra-low frequency electromagnetic waves
and earthquake prediction, and at Tony's suggestion another radio
frequency expert from Stanford.  We then visited Jack to look at his
equipment, get him to describe his techniques, and attempted (and
completely failed) to get a documented list (or actually any list) of
past predictions.  The makeup of the team was designed to be able to
verify if there was any link between what Jack was doing and what Tony
Fraser-Smith was doing (such a link had been claimed and did not
exist), to be able to analyze the probable source of any signals he
could show us (those that he could show us on his signal analyzer were
clearly manmade noise, however his predictions were made by listening
to noise bursts on AM radios.  He, at least then, was unable to record
these noise bursts so we couldn't do much with them), and to analyze
his record of success and failures (my part of the team, but of course
there wasn't a record).

I should add that Jack fully cooperated with our visit, although he
did grandstand a bit by inviting a TV news crew to show up near the
end of it.  So, I got stuck doing an interview at his home before we
could finish our report.  They also filmed footage of us getting into
our car complete with close-ups of the US Government plates (oh, the
drama of it all :-) ).

We then wrote a short report that was then used by the state and the USGS
to issue joint statements refuting his prediction.  That report had the
effect of largely getting people to ignore his predictions (e.g. no
schools or large businesses closed and there were no runs on supermarkets
for supplies.  These things have happenned in other cases and some schools
were considering closing before our findings were released).  I consider
this one of the more direct contributions I have made to society, although
I'd rather have had the time to do my own research.

Andy


From: cochrane@netcom.com (Larry Cochrane)
Subject: Re: Is there a description of the format ...
Date: Wed, 27 Apr 1994 03:23:57 GMT

In article Pete Carah wrote:
>Is there a description of the format or at least file-read source around so
>those of us with unix (or clone) workstations can view them too?
>(e.g. parse the file into a simple list and use gnuplot, or better...)

>I have looked at the PSN (Pasadena) file list and can't see anything
>like this at first glance.

I got the format for the Public Seismic Network quake data files from the 
documentation for SDAS the DOS program that most PSN stations use to record
quake data. Here is the part of the doc file that explains the format:

	 User's Guide for Seismic Data Acquisition System 

APPENDIX A - SEISMIC DATA FILE INTERNAL FORMAT

The seismic data file is  saved in BASIC  BSAVE format.  The
BSAVE statement creates an unencoded,  binary file, which is
an exact image of  memory contents.  Seven  bytes of control
information are  written  at the  beginning of the file, and
these  bytes  are followed by  the  data  bytes  copied from
memory.  The file length shown  in  the  DOS directory entry
will be the BSAVE-specified length plus 7,  rounded up to  a
multiple of the BASIC buffer size.  The  format  of the file
is as follows:

Offset Length Contents
  0      1    X'FD' (Constant BSAVE file format identifier)
  1      2    BASIC DS segment value at BSAVE execution time
  3      2    Offset in the DS segment specified in the 
	      BSAVE statement
  5      2    Data length specifed in the BSAVE statement
  7      *    Memory image data

The  segment value, offset, and length values are all stored
in Intel low-high-order format.

If a BLOAD  statement is executed with an offset term, then
that offset  value and the BASIC DS  segment value in effect
at  BLOAD  execution  time will be used to place the  loaded
data; if a  BLOAD statement is  executed  without  an offset
term,  then  the  stored segment and  offset values  will be
used.  In both cases, the stored length value determines the
amount of data loaded.  The memory image data portion of the
file  is described below.  These  2-byte  integer fields are
also  stored  in Intel  low-high-order  format.  Offsets  as
listed should be used as shown to index into the array after
it  has been BLOAD'ed into memory, i.e.  ARRAY%(1)  contains
the year  of the start of data collection.  This assumes you
have not changed the BASIC default array origin of zero.

The file format is shown below:

    OFFSET       NAME     CONTENTS

       0         Format   Flag describing format of data.
                          Earlier versions of SDAS saved data
                          in a different format.  If this field
                          contains the value 2, then the following
                          format description applies to the file.
       1         SYEAR    Year of start of data collection
       2         SMON     Start month of data collection
       3         SDAY     Day of start of data collection
       4         SHOUR    Hour of start of data collection
       5         SMIN     Minute of start of data collection
       6         SSEC     Second of start of data collection
       7         STENTH   Tenths of second at start
       8         FHOUR    Hour of finish of data collection
       9         FMIN     Minute of finish of data collection
      10         FSEC     Second of finish of data collection
      11         FTENTH   Tenths of seconds at finish
      12         COUNT    Count of valid elements in file
                          including the 100-byte header
      13         BASE     BASELINE value used for this file
      14         MIN      Min. sampled value in this file
      15         MAX      Max. sampled value in this file
      16         ORIENTATION CHARACTER N = North-South, E = East-West,
				       Z = Vertical

The next 4 values are dependent on the location of the station, and
are copied from values supplied in the SDAS.PRO file.

      17         LATINT   Station latitude N of equator (neg for S)
                          Integer portion only
      18         LATDEC   Decimal portion of latitude multiplied by
                          100 and rounded to nearest integer
                          Also negative if South Latitude
      19         LONGINT  Longitude E of Greenwich (neg. for W)
                          Integer portion only
      20         LONGDEC  Decimal portion of longitude multiplied by
                          100 and rounded to nearest integer
                          Also negative if West Longitude
      21                  NEIC hour of quake origin
                          Set to -1 by this program, modified by the
                          user via the SDASCOM program later after
                          checking with the National Earthquake
                          Information Center or other authority.
      22                  NEIC minute of quake origin
      23                  NEIC second of quake origin
      24                  NEIC tenths of second of origin
      25-39      RECLOC   Name of recording location (15 chars max)
      40-99      COMMENT  Description of quake.  Added to quake file
                          with the SDASCOM program at a later time.
      100-25099           Two-byte-integer digitized seismometer
                          data values.

WinQuake was written in C++ so I don't use the first 7 bytes. The program
skips the 7 bytes in the file, then reads in the next 200 bytes in to a
structure, I get the samples count in COUNT (offset 12), subtract 100, 
then read in that many samples.

Hope this help. If anyone has any questions about this format or WinQuake
feel free to e-mail me. I'll also put a copy of the format in file 
format.psn in my ftp directory.

BTW: We could also use a good Mac program. Any Mac programmers out there... 



>From michael@garlock.wr.usgs.gov  Thu Apr 28 17:56:16 1994
Date: Thu, 28 Apr 94 18:04:16 PDT
From: michael@garlock.wr.usgs.gov (Andy Michael)
Message-Id: <9404290104.AA14139@garlock. wr.usgs.gov>
To: eugene
Subject: Re: Earthquake Predications


Jack Coles, who previously worked in a consumer electronics store, has
made earthquake predictions for several years.  His methods were
investigated by the U.S.G.S. in 1991 because one of his predictions
had made the press and was starting to create public concern.

To investigate his methods, I put together a team that consisted of
another geophysicist from the USGS that has a lot of electronics
expertise and who has followed electrical methods and earthquake
prediction for many years, Tony Fraser-Smith from Stanford who has been
involved with ultra-low frequency electromagnetic waves and earthquake
prediction, and at Tony's suggestion another radio frequency expert
from Stanford.  We then visited Jack to look at his equipment, get him
to describe his techniques, and attempted (and completely failed) to
get a documented list (or actually any list) of past predictions.  The
makeup of the team was designed to do the following:

    To verify if there was any link between what Jack was doing
    and what Tony Fraser-Smith was doing (such a link had been
    claimed and did not exist).

    To analyze the probable source of any signals he could show
    us.  Those that he could show us on his signal analyzer were
    clearly manmade noise, however his predictions were made by
    listening to noise bursts on AM radios.  He, at least then,
    was unable to record these noise bursts so we couldn't do
    anything with them.

    To analyze his record of predictions to see if he was doing
    better than random chance, but there was no record.

After this visit I wrote a short report and our joint efforts led to a
press release by the State of California's Office of Emergency
Services.  The key part of it is, "In the opinion of Dr. Davis [State
Geologist], of the State Department of Conservation/Division of Mines
and Geology, as well as USGS and Stanford University scientists who
have discussed his efforts with him, Mr. Coles' work does not provide
systematic evidence that radio signals were precursory to the
[aforementioned] earthquakes.  In comments to OES, Dr. Davis stated
that 'the present status of this work does not warrant its use in
public policies by local or state government and would not justify
special efforts in preparedness on the dates specified in his
forecasts.'"

The report and press release had the effect of largely getting people
to ignore his predictions.  E.g. no schools or large businesses closed
and there were no runs on supermarkets for supplies.  These things have
happenned in other cases and some schools were considering closing
before our findings were released.  I consider this one of the more
direct contributions I have made to society.


Article: 8007 of ca.earthquakes
Newsgroups: ca.earthquakes
From: flowers@lanai.cs.ucla.edu (Margot Flowers)
Subject: Re: WHY MORNING EARTHQUAKES?

After pointing out differences in characteristics between the moon and
its quakes, and the earth and earthquake, such as mass differences,
differing rock characteristics, Bruce Bolt observes that the deep
quakes on the moon:

     ... commonly occur within an interval of a few days during
     perigee, ... About equal numbers of deep moonquakes occur at
     these centers at opposite phases of this tidal pull, so that the
     most active periods are 14 days apart.  These periodic properties
     at least suggest that the tidal pull of the Earth on the moon
     triggers the occurrence of the deep seismic-energy releases.
                   [Bolt, _Earthquakes_ 1993, p. 178]

>From what I understand, these kinds of patterns have been sought on
the earth but not confirmed.  Is anyone aware of any successes?


Article: 8067 of ca.earthquakes
From: kjn@netcom.com (Ken Navarre)
Subject: Re: Plate Tectonic FAQ?

One of the best books that I have found is EARTHQUAKES AND GEOLOGICAL 
DISCOVERY by Bruce Bolt, 1993, Scientific American Library, distributed by 
W.H. Freeman & Co., available thru most bookstores. Bruce bolt is a 
Professor of Seismology and former Director of the UC Berkely Seismographic 
Stations. In his book he discusses seismology, the various waves and 
their propagation, plate tectonics, how and why scientist study 
earthquakes, and how earthquakes affect buildings. Quite an undertaking 
in a mere 225 pages! It has some excellent photos and illistrations and 
is an easy read.


Article: 8070 of ca.earthquakes
From: hough@seismo.gps.caltech.edu (Susan Hough)
Newsgroups: ca.earthquakes
Subject: Re: Plate Tectonic FAQ?
Date: 13 May 1994 18:32:19 GMT

In article  isis@netcom.com (Mike Cohen) writes:
>What are the causes of earthquakes inland away from the plate boundaries &
>how big/common are they?

Inland ('intraplate') quakes can be caused by a number of different
kinds of stress--large scale 'glacial rebound' (the slow flexure of
the crust back up after a large sheet of ice is removed), for example,
or the broad compressional stress caused within eastern North America
by the 'push' forces from the mid-Atlantic spreading center.  The New
Madrid region is classified as a 'failed rift'--a place where the
continental crust started to rift apart, but then stopped.  Mary Lou
Zoback (USGS, Menlo Park) has shown theoretically how the presence of 
this rift will translate into the occurence of (infrequent) large 
earthquakes.  The rift extends northward from the New Madrid region,
and may imply a seismic hazard for the Wabash Valley region (Illinois/
Indiana)--there is paleoseismic evidence that large prehistoric
earthquakes have occurred in this region as well.

Overall, the rate of earthquakes in the stable part of the continent
is something like 1/10-1/100 of the rate in California; much more
infrequent, but still present.  Arch Johnston (Memphis State) has
tried to classify world wide intraplate earthquakes and argues that
all of the very large intraplate events may occur at failed rifts,
with some large events at continental margins (the 1888 Charleston
earthquake being an example of the latter).  Moderate-sized events,
magnitude 6 or so, are considered more 'fair game' for other areas,
as was illustrated by last year's deadly earthquake in India.

I confess to very little direct knowledge of Wyoming's seismotectonics.
It is not linked to the New Madrid zone. A map of historic seismicity
reveals a spattering of fairly small-moderate size earthquake across
the state, more than, say, North Dakota, but still a very low
overall level.  There has been a concentration of events at the
_very_ NW corner of the state, part of a zone that continues
southward into Idaho (Snake River areas) and Utah.  Probably somebody
out there knows more about this zone than I do(?)

Sue 


From: geomagic@seismo.do.usbr.gov (Dan O-Connell)
Subject: Re: Richter scale
Date: Mon, 23 May 1994 23:18:46 GMT

In article <1994May23.161537.10493@neutron.nacamar.de> jui@neutron.nacamar.de (Uwe Harmening) writes:

>   maybe this is a bit of a silly question, but we had a discussion today about
>   the Richter scale. How is an earthquake measured, is the Richte scale still
>   used and if yes, will it be replaced?
>   Thanx in advance!

The Richter scale is no longer used to report most earthquake magnitudes,
although the much of the press continues to incorrectly call some
magnitudes "Richter magnitudes." The Richter scales was developed in
southern California by studying the decay of of maximum seismic amplitudes
observed on Wood-Anderson short-period torsional seismographcs with
distance. He developed distance (attenuation) corrections so seismic
amplitudes could be corrected to zero distance and thereby characterize
the "size" of the earthquake. Since the attenuation relations were
developed from southern California earthquakes and seismographic
stations, the application of the Richter scale elsewhere 
is suspect unless area specific attenuation curves are developed. 
Even then, it does say much about the physical size of the earthquake.

The moment-magnitude, Mw, is now the most popular magnitude scale. Seismic
moment is defined as 

Mo = udA, where u is average rigidity, d is average fault slip, and
A is fault area. Mo is most commonly derived from moment tensor
inversions of long-period seismograms from worldwide seismic networks.
The Hanks and Kanamori (1979) relation,

Mw = 2/3 *log(Mo) - 10.7

is used to calculate the moment magnitude. Through Mo, Mw is directly
proportional to the actual "size" of the earthquake. The methods
used to derive Mo essentially correct for the radiation pattern and 
attenuation effects. The Richter magnitude scale begins to "saturate"
between 6.5 and 7.0. (it does not increase with increasing earthquake
size above these magnitudes due to the frequency band it samples).
Hanks and Kanamori (1979) developed Mw so it corresponded as well
as possible to various existing magnitude scale (within the
appropriate frequency bandwidths of the various magnitude scales).

Let's use the 1992 Landers earthquake for an example. The fault
length was ~80 km, fault width about ~10 km, average slip about 4.5m,
and average rigidity about 3E11 dyne/cm^2. You could infer these 
parameters from field observations, aftershock seismicity, geodetic 
deformation measurements, long-period moment tensor inversions, etc.
The moment, Mo, is 1.08e27 dyne-cm and Mw = 7.3. 

This scratches the surface, but hopefully, it helps.

Dan O'Connell
geomagic@seismo.do.usbr.gov
Seismotectonics Group, U.S. Bureau of Reclamation
Denver Federal Center, P.O. Box 25007 D-3611, Denver, CO 80225
"We do custom earthquakes (for food)" 
                  or 
"Just more roadkill on the information superhighway"

                   /\
                  /  \
                 /    \        /\            /\
    /\          /      \      /  \          /  \    /\  /\/\  /\/\
___/  \  /\/\/\/        \    /    \    /\  /    \  /  \/    \/    \  /\_______
       \/                \  /      \  /  \/      \/                \/
                          \/        \/



-- 
Dan O'Connell
geomagic@seismo.do.usbr.gov
Seismotectonics Group, U.S. Bureau of Reclamation


From: geomagic@seismo.do.usbr.gov (Dan O-Connell)
Subject: Re: Richter scale

In article <2rrd4j$gcp@quartz.ucs.ualberta.ca> buri@probe2.phys.ualberta.ca (Michael Burianyk) writes:

   In article   
   geomagic@seismo.do.usbr.gov (Dan O-Connell) writes:
   > The moment-magnitude, Mw, is now the most popular magnitude scale. Seismic
   > moment is defined as 
   > 
   > Mo = udA, ......
   > 
   > Mw = 2/3 *log(Mo) - 10.7
   > 
   > is used to calculate the moment magnitude. 

> Next question ...  what is the relationship between moment-magnitude, 
> or seismic moment, and the energy released during an earthquake?

The radiated seismic energy could be calculated by integrating
the square of the absolute value of the slip velocity radiated by
all portions of the fault or

 /+x /max(z)
 |   |      |v(x,z)|^2 dz dx                                 (1)
/-x  /min(z)

where v(x,z) is the slip velocity on the fault. Using the Parseval-Rayleigh
theorem, (1) can be calculated from the integral of the velocity
spectra of broadband seismic recording (if propagation, attenuation,
and radiation effects can be estimated and removed). To first order,
slip velocity is proportional to stress drop. So you could have two
earthquakes on the same size fault that release different amounts
of energy if the stress drops are significantly different (assuming
equivalent slip durations). The problem with (1) is that slip velocity
on the fault is one of the "mysteries of the universe" parameters
that has not been directly measured. Fault slips are inferred from
inversions of geodetic and/or seismic recordings, but there are lots
of different slip velocity-slip duration-stress drop combinations 
that can produce the same slip distribution on a fault.

So to provide an "easy" answer to the question, you can using
the empirically derived Gutenberg-Richter energy-surface wave
magnitude relation

log  E = 11.8 + 1.5*M                                        (2)
   10 S              S

where M is surface wave magnitude and each unit increase in M corresponds
       S                                                     S
to a 32-fold increase in energy. Substituting Mw for M gives a quick
                                                      S
and dirty estimate in ergs.


or (From: shirriff@cs.berkeley.edu (Ken Shirriff)) 
Markus Bath's log E = 12.24 + 1.44 M                         (3)
where M is the magnitude > 5. This is an empirical equation, derived by 
integrating over the whole wave train in time and space.  From this 
equation, each increase of 1 in magnitude increases the energy released 
by a factor of 27.5.

Both (2) and (3) represent attempts to approximate the integral (1)
by accounting for the loss of amplitude with distance of propagation
of various seismic phases (surface waves for (2) and ensembles of
all wavetypes for (3)).

Anyway, seismic energy increases proportionately with moment, as each
represents integrals over the fault. For seismic moment, its the integral
of slip over the fault. For energy, its the integral of squared slip
velocity over the fault. You can infer slip distribution from various
data sets. Slip velocity is a transient that is difficult to
measure, but it is an extremely important factor in determining
the behavior (time and space distributions) of strong motions 
(accelerations).

Dan O'Connell
geomagic@seismo.do.usbr.gov
Seismotectonics Group, U.S. Bureau of Reclamation
Denver Federal Center, P.O. Box 25007 D-3611, Denver, CO 80225
"We do custom earthquakes (for food)" 
                  or 
"Just more roadkill on the information superhighway"

                   /\
                  /  \
                 /    \        /\            /\
    /\          /      \      /  \          /  \    /\  /\/\  /\/\
___/  \  /\/\/\/        \    /    \    /\  /    \  /  \/    \/    \  /\_______
       \/                \  /      \  /  \/      \/                \/
                          \/        \/


-- 
Dan O'Connell
geomagic@seismo.do.usbr.gov
Seismotectonics Group, U.S. Bureau of Reclamation


From: stgprao@st.unocal.COM (Richard Ottolini)
Subject: Re: World-wide earthquake frequencies?

>From Bolt "Earthquakes" Appendix A:

M(s)		#>
8		2
7		20
6		100
5		3000
4		15000
3		100000

Seismicity size-rates tend to follow power laws, with each
increase in magnitude N times more common.
N varies from region to region, and in southern CA, from
decade to decade..
(The numbers above 7 look too high and the jump from 5 to 6
high too.  Are Bolt's numbers correct?)



From: John Gunn 
Subject: Re: World-wide earthquake frequencies?
Date: Sat, 11 Jun 94 17:13:21 -0500

Telnet neis.cr.usgs.gov and login as qed.  This BBS is run by the USGS
and has many interesting facts about earthquakes worldwide in the last
century.


From: emeth@beretta.ramp.com (David F Jones)
Subject: Re: Northern California quake info?

World Wide = finger quake@gldfs.cr.usgs.gov

Hawaii Volcano Obs = finger quake@tako.wr.usgs.gov

IRIS Teleseisms = spyder@dmc.iris.washingtion.edu ( non-finger )

finger quake@gldfs.cr.usgs.gov         (Worldwide Earthquakes)
finger quake@andreas.wr.usgs.gov       (Northern California)
finger quake@scec.gps.caltech.edu      (Southern California)
finger quake@fm.gi.alaska.edu          (Alaska)
finger quake@slueas.slu.edu            (Central U.S.)
finger quake@seismo.unr.edu            (Nevada)
finger quake@eqinfo.seis.utah.edu      (Utah and Yellowstone)
finger quake@geophys.washington.edu    (Washington & Oregon)

file://garlock.wr.usgs.gov/pub/earthquake.html is one place to start


From: kjn@netcom.com (Ken Navarre)
Subject: Re: Scientific America home seismograph

The article was published in Scientific American, 1979, v. 241, No.1, 
pg 152-161. It described the construction of a horizontal seismic sensor
and preamp circuit designed by James D. Lehman. At the time, Mr. Lehman 
was with the Physics Dept., James Madision University, Harrisonburg, Va., 22807.

Another source of information that describes modification to the Lehman 
sensor was published in the:
Journal of Geological Education, 1987, v.35, pg. 124 by
Richard Lawrence Koll
Dept. of Geology and Meteorology
Kean College of New Jersey
Union, NJ, 07083

The modifications include the use of common pipe fittings to construct 
the frame and support for the sensor. Construction is simple and 
straightforward. The amplifier, however, was prone to significant noise 
and oscillation. Newer circuits and components are available that eliminate 
this problem.

=====

Web (WWW) Pages:
http://vulcan.wr.usgs.gov/home.html

Article 11338 of ca.earthquakes:
Path: cnn.nas.nasa.gov!ames!lll-winken.llnl.gov!unixhub!news.Stanford.EDU!morrow.stanford.edu!pangea.Stanford.EDU!andy
From: andy@pangea.Stanford.EDU (Andy Michael USGS Guest)
Newsgroups: ca.earthquakes
Subject: Re: usgs bbs number?
Date: 4 Jan 1995 01:55:36 GMT
Organization: Stanford Univ. Earth Sciences
Lines: 11
Distribution: ca
Message-ID: <3ecv6o$6g5@morrow.stanford.edu>
References: 
NNTP-Posting-Host: pangea.stanford.edu

In article  tbear@calon.com (tbear) writes:
>I went to the usgs open house in Meno Park, CA and I was told that there 
>would be a number to access the usgs bbs. Does anyone know the number?

The bbs in Menlo Park is (415)327-1517 and from within the Bay Area
1-800-328-1517.

A list of available services is can be gotten by sending email to
help@quake.wr.usgs.gov as long as your return path is set correctly.

Andy


Article 246 of sci.geo.earthquakes:
Path: cnn.nas.nasa.gov!ames!hookup!news.mathworks.com!uhog.mit.edu!bloom-beacon.mit.edu!senator-bedfellow.mit.edu!polycarp!gibson
From: gibson@polycarp.NoSubdomain.NoDomain (Rick Gibson)
Newsgroups: sci.geo.earthquakes
Subject: Re: Who was the first seismologist?
Date: 9 Jan 1995 22:17:43 GMT
Organization: Massachvsetts Institvte of Technology
Lines: 35
Sender: gibson@polycarp (Rick Gibson)
Distribution: world
Message-ID: <3escm7$sc0@senator-bedfellow.MIT.EDU>
References: 
NNTP-Posting-Host: polycarp.mit.edu

In article , rmwm@va.nmh.ac.uk (Roger Musson) writes:
|> Who was the first seismologist?
|> 
|> Of course, it rather depends on your definition of seismologist. Perhaps the 
|> question should be who was the first person to be referred to as a 
|> seismologist? First use of the word "seismometer" I can date precisely, but 
|> "seismologist" I can't.
|> 
|> Any views, opinions?
|> 
|> Roger Musson
|> British Geological Survey
|> e_rmwm@va.nmh.ac.uk 
 
For one perspective, the online  Oxford English DIctionary gives the following  as its 
early references to the word:

   1858 Mallet in Rep. Brit. Assoc. i. 1 The few physicists who are engaged in
   Seismology. 

   1879 Rutley Stud. Rocks iii. 9 The branches of physical geology known as
   Vulcanicity and Seismology.

For the word seismometer, they cite the following as the earliest reference:

   1841 J. D. Forbes in Edin. Phil. Trans. XV. i. 220 The self-registering part of
   the apparatus, which Mr. David Milne has termed a Seismometer. 

How do these correspond with your results?


Rick Gibson
MIT
gibson@erl.mit.edu



Article 248 of sci.geo.earthquakes:
Path: cnn.nas.nasa.gov!ames!hookup!uwm.edu!lll-winken.llnl.gov!unixhub!news.Stanford.EDU!morrow.stanford.edu!pangea.Stanford.EDU!andy
From: andy@pangea.Stanford.EDU (Andy Michael USGS Guest)
Newsgroups: sci.geo.earthquakes
Subject: Re: Who was the first seismologist?
Date: 10 Jan 1995 01:55:44 GMT
Organization: Stanford Univ. Earth Sciences
Lines: 68
Distribution: world
Message-ID: <3espf0$orj@morrow.stanford.edu>
References:  <3escm7$sc0@senator-bedfellow.MIT.EDU>
NNTP-Posting-Host: pangea.stanford.edu

In article <3escm7$sc0@senator-bedfellow.MIT.EDU> gibson@polycarp.NoSubdomain.NoDomain (Rick Gibson) writes:
>In article , rmwm@va.nmh.ac.uk (Roger Musson) writes:
>|> Who was the first seismologist?
>|> 
>|> Of course, it rather depends on your definition of seismologist. Perhaps the 
>|> question should be who was the first person to be referred to as a 
>|> seismologist? First use of the word "seismometer" I can date precisely, but 
>|> "seismologist" I can't.
> 
>For one perspective, the online  Oxford English DIctionary gives the following  as its 
>early references to the word:
>
>   1858 Mallet in Rep. Brit. Assoc. i. 1 The few physicists who are engaged in
>   Seismology. 
....

To proceed down this route for a bit...

That's for seismology, for seismologist the OED gives the very similar:

1859: R. Mallet in Rep. Brit. Assoc. 1858 133 "The subject appears to me worthy
of more examination at the hands of Vulcanologists and Seismologists."

(Apparently this issue came out late.)

Richter, in his textbook, cites a few studies earlier than Mallet's work on the
1857 Naples event although that one gets a lot of credit because of its very
systematic nature and the attempt to use physics to learn something about the
event.  It is also the first one in Richter's list with a clear author.

In any case it seems clear that the words seismology and seismologist
are both credited to Mallet.  He is also a good candidate to have been
called a seismologist first.

The problem with an OED search is that it looks at the usage of the word
not how it could have been applied in retrospect.

Probably all civilizations made studies of earthquakes and their
effects, but the Chinese seismoscopes were in use at least as early as
132 A.D.  Richter credits these to Chang Heng, so perhaps he is the
first recorded seismologist.  At least he lived long before Mallet and
invented a device we now call a seismoscope and therefore is arguably
a seismologist.

As noted above, it also depends on how broadly you want to apply the
term seismologist.  The original usage was for the study of earthquakes
and their causes and effects.  Now, we tend to apply it more to those
who use waveforms to study either the earth structure or earthquakes.
The people who study effects without waveforms are often earthquake
geologists or earthquake engineers.  For the more limited modern
definition it might go to Gray, Milne, and Ewing for developing the
first good seismographs (motion versus time).  For the broader
definition you can probably go back as far as the written record
allows, although I would restrict the search to those that display some
of what we would currently call scientific method to their studies.
That may allow you to skip Chang Heng depending on how he interpreted
the results of his device.  It also may allow you to skip whoever wrote
the Old Testament.  However, Amos Nur has argued that some of the
descriptions from the Old Testament fit the actual motion on faults in
that area.  So, do you depend on good observations or good
interpretation.  Yes, I am having fun clouding the issue.

That said, Mallet does deserve a lot of credit.  If you can ever get
your hands on a copy of his 1862 book "Great Neapolitan Earthquake of
1857" its worth a look.  I sometimes wonder just how much I would be
willing to pay for a copy.  Anyone got one they want to sell?

Andy


Article 217 of sci.geo.earthquakes:
Newsgroups: sci.geo.earthquakes
Path: cnn.nas.nasa.gov!ames!agate!spool.mu.edu!bloom-beacon.mit.edu!gatech!swiss.ans.net!paperboy.amoco.com!apctrc!zjad49
From: zjad49@trc.amoco.com (Joe Dellinger)
Subject: Re: Increasing Frequency and Magnitude
Message-ID: 
Sender: usenet@trc.amoco.com
Organization: Amoco Production Company, Tulsa Research
X-Checksum-Snefru: 800b5cd2 f63bf959 5bd4cb28 283884e6
References: <3ehg8a$fpd@canopus.cc.umanitoba.ca> <3ekdc1$4jg@tekadm1.cse.tek.com>
Date: Sat, 7 Jan 1995 01:26:51 GMT
Lines: 113

Here is an article on the subject of increasing quake death rates that's
worth dusting off again:

ftp://sepftp.stanford.edu/pub/geology/information/from_ca.earthquakes/quake_death_rates

-------------------------------------------------------------------------
From: alanf@tekig6.PEN.TEK.COM (Alan M Feuerbacher)
Subject: Re: Natural Disasters in the 20th Century
Date: 6 Nov 93 01:42:07 GMT
Organization: Tektronix, Inc., Beaverton,  OR.

In article <2b8fn6$qgg@seismo.CSS.GOV> stead@seismo.CSS.GOV (Richard Stead) writes:
>In article <12269@tekig7.PEN.TEK.COM> alanf@tekig6.PEN.TEK.COM (Alan M Feuerbacher) writes:
>>A religious article entitled "Natural Disasters -- A Sign
>>of the Times?" said the following:
>>   _New Scientist_ warns that "the world can expect more
>>   disasters in the 1990s than in past decades."....
>
>Without question, there is no evidence that the rate or intensity of natural
>processes has changed at all, not this century, not this millenium, not
>in millions of years.  All the evidence available points to a very constant
>rate for these things.....
>Quakes are more likely to have been
>constant with time, though it is difficult to measure.  There is a
>possibility that there have been times of increased tectonic activity
>in the past - probably causing more quakes.  There is no evidence however
>for any time having less seismicity than we do now.
>
>I sure hope you are asking these questions because you really want to know
>the answers.  Too often, we get fruitcakes who are trying to push some
>bizarre theory or religious doctrine and trying to get things out of the
>data that aren't there.  They come here with their minds made up and think they
>are going to educate the dumb, stubborn scientists.  They refuse to listen
>to reason or accept the evidence.  A lot of people here will react
>negatively if they get the impression that someone is doing this, and I
>wouldn't really blame them.

You hope right!  I do not agree with the writer of the article I quoted.
I was hoping to get a number of answers like yours, which will help me
show certain people that their confidence in sources like this is quite
misplaced.

To all who responded, many thanks!

I've done a bit of my own research on earthquake frequency and casualty
rates.  Below is presented some data I found through a variety of sources.
What do you professionals think?
___________________________________________________________________________

             A Comparison of Earthquake Victims

          1715-1783:                        1915-1983:

Year Location           Deaths   Year Location            Deaths

1715 Algeria            20,000   1915 Italy               29,970
1717 Algeria            20,000   1920 China              180,000
1718 China              43,000   1923 Japan              143,000
1719 Asia Minor          1,000   1927 China              200,000
1721 Iran              100,000   1932 China               70,000
1724 Peru (tsunami)     18,000   1933 USA                    115
1725 Peru                1,500   1935 India (Pakistan)    60,000
1725 China                 556   1939 Chile               30,000
1726 Italy               6,000   1939 Turkey              23,000
1727 Iran               77,000   1946 Turkey               1,300
1730 Italy                 200   1946 Japan                2,000
1730 China             100,000   1948 Japan                5,131
1730 Japan             137,000   1949 Ecuador              6,000
1731 China             100,000   1950 India                1,500
1732 Italy               1,940   1953 Turkey               1,200
1736 China                 260   1953 Greece                 424
1737 India             300,000   1954 Algeria              1,657
1739 China              50,000   1956 Afghanistan          2,000
1746 Peru                4,800   1957 Iran (Northern)      2,500
1749 Spain               5,000   1957 Iran (Western)       2,000
1750 Greece              2,000   1960 Chile                5,700
1751 Japan               2,000   1960 Morocco             12,000
1751 China                 900   1962 Iran                10,000
1752 Syria              20,000   1963 Yugoslavia           1,100
1754 Egypt              40,000   1964 Alaska                 131
1755 China                 270   1966 Turkey               2,529
1755 Iran                1,200   1968 Iran                11,588
1755 Portugal           60,000   1970 Turkey               1,086
1755 Morocco            12,000   1970 Peru                66,794
1757 Italy              10,000   1971 USA                     65
1759 Syria              30,000   1972 Iran                 5,057
1763 China               1,000   1972 Nicaragua            6,000
1765 China               1,189   1973 Mexico (Western)        52
1766 Japan               1,335   1973 Mexico (Central)       700
1771 Japan (tsunami)    11,700   1974 Pakistan             5,200
1773 Guatemala          20,000   1975 China                  200
1774 Newfoundland          300   1975 Turkey               2,312
1778 Iran (Kashan)       8,000   1976 Guatemala           23,000
1780 Iran (Tabriz)     100,000   1976 Italy                  900
1780 Iran (Khurasan)     3,000   1976 Bali                   600
1783 Italy (Calabria)   60,000   1976 China              242,000
1783 Italy (Palmi)       1,504   1976 Philipines           3,373
1783 Italy (Monteleone)  1,191   1976 Turkey               3,790
                                 1977-1983 addition:      44,623
                      _________                        _________
Total 1715-1783:      1,373,845  Total 1915-1983:      1,210,597
Annual average:          19,911  Annual average:          17,545
___________________________________________________________________________

Alan Feuerbacher
alanf@atlas.pen.tek.com

-------------------------------------------------------------------------
-- 
     /\    /\    /\/\/\/\/\/\/\.-.-.-.-.......___________
    /  \  /  \  /Amoco Production Research Tulsa Oklahoma\/\/\.-.-....__
___/    \/    \/Joe A. Dellinger      Internet: joe@sep.stanford.edu    \/\.-.__
-------------- Uh oh, Toto, I think we're back in Kansas! ----------------------


Article 509 of sci.geo.earthquakes:
Newsgroups: sci.geo.earthquakes
Path: cnn.nas.nasa.gov!ames!usenet.hana.nm.kr!jagalchi.cc.pusan.ac.kr!news.kreonet.re.kr!insosf1.infonet.net!newshost.marcam.com!news.mathworks.com!udel!gatech!howland.reston.ans.net!ix.netcom.com!netcom.com!jgk
From: jgk@netcom.com (Joe Keane)
Subject: Human Involvement
Message-ID: 
Summary: We need reasonable precautions.
Keywords: acceleration
Organization: none
Date: Fri, 20 Jan 1995 09:38:46 GMT
Lines: 85

This is from _Geology of California_ by Norris and Webb:

In an unstable region with large population and continued growth, maximum
safety precautions to offset earthquake hazards are extremely important.
In California, about 75 percent of the population resides in the areas of
greatest instability.  It is difficult to implement earthquake precautions,
however.  They are expensive, and because earthquakes may never affect
the majority, appropriate precautions are often ignored.  Building for
earthquake resistance may increase costs by 10 percent or more.  In the past,
incorporating earthquake safety into structures was hampered because geologic
science could not provide adequate data for structural engineers, architects,
and construction firms.  Today better information is available.

An interesting example of this problem relates to the development and
deployment of strong-motion seismographs.  For many years, seismologists had
been most interested in developing extremely sensitive instruments that would
allow scientists working at places like Berkeley or Pasadena to record and
analyze in detail the nature of distant earthquakes.  Nearby strong shocks
were expected either to disable or deactivate these sensitive instruments.
Structural engineers, on the other hand, were much less concerned with the
nature of a quake in Kamchatka, for example, and instead wanted to know what
forces local earthquakes imposed on buildings and other structures.  Beginning
about 1965, strong-motion instruments were perfected and placed in buildings,
on dams, bridges, and other structures.  These began to yield data that
surprised nearly everyone.  During the moderately strong 1971 San Fernando
earthquake (magnitude 6.4), accelerations equal to or slightly greater than
gravity were measured.  This meant that objects not tied down tended to
float when the acceleration exceeded gravity, just as they do in spacecraft;
rocks tossed in the air often landed upside down, for example.  These
high accelerations forced considerable review of the adequacy of existing
construction standards and design.  Such large accelerations were thought
to have been extremely rare, but it is now clear that they are not.

Establishing safety standards assumes willingness to recognize earthquake
hazards, expenditures of large sums of money required, and counteracting the
apathy that arises from the ``it can't happen to me'' philosophy.  This last
obstacle might be overcome quickly if many continuing, sufficiently strong but
not disastrous earthquakes were to occur.  Instead, infrequent, localized, but
unfortunately sometimes severe tremors affect comparatively few people at a
time.  For example, the 1987 Whittier Narrows earthquake (magnitude 5.9, $400
million damage) was termed a ``wake-up call'' to the Los Angeles metropolitan
area which, at present, is bracing for ``the big one.''

*Engineering*

It is now feasible to construct modern earthquake-resistant (not
earthquake-proof) buildings.  Since the locally disastrous 1933 Long Beach
earthquake (magnitude 6.3), the Field Act has required that all California
public school buildings be earthquake resistant.  Moreover, the uniform
building codes adopted in the 1950s have applied the Field Act standards to
virtually all buildings.  Yet damage sustained in the 1971 San Fernando quake
showed that even these standards were not completely adequate.  Nonetheless,
dams, waterworks, highways, and utility structures may all be built with
reasonable safety provided certain precautions are followed and provided
that they are not built directly on faults or on unstable ground subject
to liquefaction or sliding.

Public lack of awareness and failure to demand reasonable precautions are
the main impediments to earthquake safety, although some hazards will always
remain.  In the 1971 San Fernando earthquake, damage to newly constructed
(built with reasonable safeguards) was quite extensive.  The quake's
surprisingly large forces led some geologists to assert, however, that the
San Fernando case can reasonably be expected to occur every 5,000 years.  Most
seismologists are now much less sure that these high accelerations are rare,
and on the basis of what is known about the setting and history of the San
Fernando fault system it is now estimated that the recurrence interval is
between 100 and 300 years.  Should huge expenditures be allocated to safeguard
structures, roads, and bridges that have an estimated life of 50 years or so?
Reasonable precautions are required, but some degree of risk must be accepted
when people choose to live in an unstable, earthquake-prone region.

*Psychological Factors*

The psychological discord experienced when an earthquake strikes can be
severe for a few people.  There is often a tendency to accept rumors about
earthquakes without regard for the facts.  The psychological consequences of
other destructive natural phenomena such as tornados, hurricanes, and floods
are also extremely traumatic, but they do not seem to be accompanied by
false rumors to the same degree as earthquakes.  Perhaps other natural
disaster-producing phenomena are less damaging to the psyche because they can
be seen--people see them arrive and see them leave--something rarely possible
with earthquakes.  Storms, floods, landslides, windstorms, and fires take far
larger tolls of life and property in the United States than do earthquakes.
Furthermore, safeguards against earthquakes are no more costly than those
required to prevent damage from other natural phenomena.


>From cjones@mantle.colorado.edu  Fri Jan 27 09:33:53 1995
Received: from data.nas.nasa.gov (data.nas.nasa.gov [129.99.23.20])
	by wilbur.nas.nasa.gov (8.6.8.1/NAS.5.b) with ESMTP id JAA27109
	for ; Fri, 27 Jan 1995 09:33:52 -0800
Received: from mantle.colorado.edu (mantle.Colorado.EDU [128.138.136.28])
	by data.nas.nasa.gov (8.6.8.1/NAS.5.b) with ESMTP id JAA08839
	for ; Fri, 27 Jan 1995 09:33:38 -0800
Received: (from cjones@localhost) by mantle.colorado.edu (8.6.9/8.6.9/CNS-3.5) id KAA06996 for eugene@nas.nasa.gov; Fri, 27 Jan 1995 10:33:30 -0700
Date: Fri, 27 Jan 1995 10:33:30 -0700
From: Craig Jones 
Message-Id: <199501271733.KAA06996@mantle.colorado.edu>
To: eugene
Subject: EQ magnitudes
Status: RO

This got bounced back using the address first given to me; hope
this one works....

Craig Jones

>From daemon Fri Jan 27 10:29:39 1995
Received: from lace.Colorado.EDU (lace.Colorado.EDU [128.138.240.25]) by mantle.colorado.edu (8.6.9/8.6.9/CNS-3.5) with ESMTP id KAA06759 for ; Fri, 27 Jan 1995 10:29:38 -0700
Received: from localhost (localhost) by lace.Colorado.EDU (8.6.9/8.6.9/CNS-3.5) with internal id KAA00289; Fri, 27 Jan 1995 10:29:49 -0700
Date: Fri, 27 Jan 1995 10:29:49 -0700
From: Mail Delivery Subsystem 
Subject: Returned mail:  Host unknown (Name server: wilber.nas.nasa.gov: host not found)
Message-Id: <199501271729.KAA00289@lace.Colorado.EDU>
To: 
Status: R

The original message was received at Fri, 27 Jan 1995 10:29:48 -0700
from tectquad.Colorado.EDU [128.138.136.29]

   ----- The following addresses had delivery problems -----
  (unrecoverable error)

   ----- Transcript of session follows -----
501 ...  550 Host unknown (Name server: wilber.nas.nasa.gov: host not found)

   ----- Original message follows -----
Return-Path: cjones@mantle.colorado.edu
Received: from tectquad.colorado.edu. (tectquad.Colorado.EDU [128.138.136.29]) by lace.Colorado.EDU (8.6.9/8.6.9/CNS-3.5) with SMTP id KAA00287 for ; Fri, 27 Jan 1995 10:29:48 -0700
Message-ID: 
Date: Fri, 27 Jan 95 10:30:49 MST
From: "Craig Jones" 
Subject: magnitudes
To: eugene@wilber.nas.nasa.gov
X-Mailer: VersaTerm Link v1.1.1

Hi!--the following note sent to me from Greg Anderson will explain the rest
of this message...

  Craig --

  As one of the others who finally wore out to the point of no longer
  posting *anything* to either sci.geo.geology or sci.geo.earthquakes
  (not to mention ca.earthquakes), I want to say that your post did a
  very great job of explaining things.  I seriously think you should
  consider posting it (or auto-posting it) every couple weeks, or
  perhaps months.

  You might well also want to send a copy to Eugene Miya at NASA Ames,
  as he took over from me (after I took over from him) the duties of
  posting the ca.earthquakes FAQ...

  In short, your post was great, and I can now rest with a clear
  conscience ( ;-) )

  Greg Anderson
  anderson@mahi.ucsd.edu

  PS Eugene Miya's e-mail address is eugene@wilber.nas.nasa.gov

----
Anyway, though I doubt there is anything really all that special about the
post he referred to, I am including a copy here for your perusal.

Craig Jones, cjones@mantle.colorado.edu
Research Associate, CIRES, Univ. of Colorado

---------
In Article <3g1ggfINNpt0@titan.oit.umass.edu>, cyndal@titan.oit.umass.edu
(Cyndal) wrote:
>  Hi all,
>
>  I happened to see a post last week where the sender was a bit
>out of sorts (a polite term) because the media was taking the
>magnitude of the Kobe quake and confusing it with a Richter scale
>measurement.  I've seen 3(and maybe there are 1-2 more) different ways 
>recently that the Kobe quake was measured: Magnitude, x.y Richter, and one
>other one.
>  For those of us who aren't seismologists/geologists/etc, could someone
>explain(in layman's terms) what the differences are?  I've puttered around
>a few places that list daily/weekly quakes and have seen things like 
>"Mag 4.5" and always assumed it was 4.5 on the Richter scale.  Now I see
>that this is probably a very wrong assumption.  So, please, what ARE
>the differences, and why are quakes given different measurements?
>                                   Thanks!
>                                   Michele
>
This qualifies as the most often asked question on this newsgroup and
sci.geo.geology; there should be a permanent post on what magnitudes mean,
how they are defined, etc.  The reason this question usually goes unanswered
in these newsgroups is that all of us who have answered it a few times are
just getting worn out.

So let's go ahead and see what's there....

First, there are two fundamental ways of describing the "size" of an
earthquake that often make it to newspapers and the public: magnitude and
intensity.  We'll cover magnitude in a minute, but it is equivalent to the
wattage on a light bulb--it to some degree should be proportional to the
total energy output of the earthquake regardless of where the earthquake is.
Intensity is how strongly the earthquake is felt at a particular spot, which
depends on distance from the earthquake, local geology, the presence of
observers (often), etc.  It is equivalent to how bright the light from a
bulb is at different places in a room.  Intensity in the U.S. is generally
reported as Modified Mercalli Intensity and is usually assigned a roman
numeral (in part to distinguish it from magnitude, in part tradition, and in
part because these observations are no more precise than an integral value);
when reported in the media, usually the peak value is reported.  This
occasionally is confused with magnitude.  In some countries other intensity
scales are used (and I seem to recall that Japan is one of those countries).

OK, so now magnitude.  The fundamental to all magnitude measures is that for
a 1 point increase in any magnitude for fixed location of earthquake and
seismometer, the amount of ground motion measured for that magnitude scale
will increase by a factor of 10.  It happens that this works out to mean
that an increase of 1 in magnitude will also represent an increase in the
total energy of an earthquake of a factor of about 28.  To calculate the
magnitude, corrections are made for the distance from the epicenter, depth
of the earthquake, model of seismometer, and Earth structure; thus the
magnitude of an earthquake should be identical regardless of where it is
measured if using a common scale (discussed below).  In reality, the
corrections for Earth structure are complex and are affected as well by the
radiation pattern of seismic waves from an earthquake.  As a result,
individual measurements on an earthquake will vary quite a bit.  Most of the
regional seismic networks in the U.S. will average readings from several
stations to make their magnitude estimate.  Again, because of some
difficulties in correction and the way seismic energy might radiate from an
earthquake somewhat differently to two different networks, the estimates of
magnitude might vary by a few fractions of a magnitude.  Although this
occasionally seems like incompetence, it just reflects the uncertainty of
the whole proceedure.  If you return to the light bulb analogy, if two
observers tried to estimate the wattage of the light bulb at two spots, they
would in fact measure the intensity and then try and correct for the
distance to the bulb.  But if observer A had, say, a moth between her and
the bulb, her wattage (magnitude) might be a little low, and observer B
lacked the moth but caught a reflection from a window as well, his estimate
might be a little too high.

Why different magnitudes?  It boils down to the way the Earth transmits
seismic energy.  First, there are body waves and surface waves.  Surface
waves only travel (create motion) within the upper few kilometers to a few
hundred km of the Earth; body waves can travel within the entire Earth.  An
analogy in water are ocean waves and sound.  Scuba divers are probably
familiar with the rapid decrease of motion from ocean swells (waves) with
depth--a ship can bob about a lot at the surface while a diver 100 feet down
might not move much at all.  That is a surface wave.  Sound, however,
travels equally well throughout the ocean, as SONAR demonstrates (we'll omit
all the peculiarities of SOFAR channels and the like, thank you very much
ex-submariners)--it is a body wave.  Second, the Earth has different flavors
of seismic waves.  For instance, body waves come in P and S waves.  P waves
are compressional waves and S waves are shear waves; P waves travel faster
than S waves.  If you want to see P and S waves, get a Slinky, lay it on a
nice, slick table, stretch it out some, and then push and pull one end
towards the other end quickly--you will see a wave travel down the Slinky of
coils close together and further apart than the Slinky at rest--this is a P
wave.  Now rapidly move one end of the Slinky at right angles to the length
of the Slinky--you'll see a wave move down where the Slinky slides over to
one side and then back again.  This is an S wave (more or less).  Surface
waves also have different flavors, but for the most part this isn't
exploited in magnitude determinations.

OK, so what magnitude scales exist?  There are more than I will go over, but
these are the most common: duration magnitude (Md), local (Richter)
magnitude (ML), body wave magnitude (Mb), surfave wave magnitude (Ms),
moment magnitude (Mw). The reason for using the different scales is that
each type of seismic wave is easily observed only over certain magnitude or
distance ranges.

Duration magnitude is used for small earthquakes or in areas with generally
poorly calibrated seismometers.  It relies on the observation that the time
it takes for the motion from an earthquake to fall into the background noise
is proportional to other, more physically defensible measures of magnitude.  

Local (Richter) magnitudes are measured for most earthquakes up to somewhere
around M 7; beyond that the measurement fails to increase with earthquake
size.  This is basically Richter's technique--you measure the maximum
amplitude of ground motion (often restricted to a certain part of the
seismogram but not always), take that and the distance from the earthquake
and you can get the Richter magnitude.  While the original definition was
based on a specific instrument at a specific distance, corrections for
distance and other seismometers have been worked out over the years. Large
earthquakes generally swamp the instruments used to make these measurements
and the Earth does an increasingly inefficient job of transmitting these
particular waves for very large events.

Body wave magnitude is a very similar measure to local magnitude except it
can be applied to the body waves from distant earthquakes and it has quite a
number of corrections for which body wave (P or S, for instance), the
Earth's structure, and the period of the wave being examined.  It is
proportional to the log of the amplitude of ground motion divided by the
period of the wave being measured (A/T).

Surface wave magnitudes also use the log of the amplitude divided by period,
but surface waves are only generated by relatively shallow earthquakes. 
Again corrections exist for distance, depth, and seismometer type.  Surface
waves tend to saturate in the M 8+ range.

Moment magnitude has confused many people because it isn't really a direct
magnitude measure (log of an amplitude of a seismic wave) as are the others.
Instead it is derived from seismic moment, which is simply the product of
the shear modulus (a property of the rock), the fault area, and the slip on
the fault.  Seismic moment is usually derived from fits made to entire
seismograms or large parts of seismograms using certain physical models;
this has been automated for several years by a group at Harvard and has been
applied in near-real time by several other groups including the USGS. 
Unlike other magnitude estimates, you can also derive an estimate of seismic
moment from geodetic observations (how far points have moved on the surface
of the Earth due to an earthquake) or from geologic observations because you
only need fault area and displacement and some estimate of shear modulus. 
In practice you will not see such estimates in the public media because they
take time to measure and are not necessarily measuring the same exact
phenomena (for instance, if sizable slip occurs without generating seismic
energy, a geodetic or geologic moment will exceed a seismic moment).  The
measure "moment magnitude" is somewhat like the duration magnitude in that
it has been made to correlate with the other scales, though there are good
physical arguments for this relation.  In practice, moment magnitudes are
the best measure because they can include all the observations that go into
the other magnitudes plus the duration of the arrivals plus additional
very-long period oscillations of the Earth.

In general these magnitudes have all adjusted been to agree where they
overlap (the worst case is usually Mb and Ms, which often disagree a fair
bit--mostly due to the way surface waves are generated).  Because of this,
you could say that the ground motion would be 100,000 times greater in a
magnitude 7 than a magnitude 2 earthquake at a given spot despite the
different measures probably used (though the reality is that the ground
motion would have probably saturated).  A statement less affected by the way
the Earth transmits energy would be that a magnitude 7 earthquake generates
about 16,000,000 times more energy that a magnitude 2.

Hopefully this goes some distance in explaining magnitudes.

Craig Jones
CIRES, University of Colorado, Boulder
cjones@mantle.colorado.edu

[Return to Aaron Brown's Large List of Links]
[Return to Aaron Brown's Web Page]