On Sat, 10 Jul 1999, Nick & Sophie Caporossi wrote:
> listing quakes and other seismic activity on the East Coast of the USA. It
> had been covering an area from Maine to Virginia and parts of Ontario
> Canada. It is no longer doing so. Does any one know of another site that
> has the same information?

   There is a good global catalog hosted by Berkeley's seismo lab web site.
All you do is input the min/max latitude and longitude and the beginning and
end time and a few other parameters and the computer will build a list of all
quakes in that area.  I am not sure how often it is updated, but it seems
fairly thorough.  I don't remember the address, but you can find it by
clicking through www.berkeley.edu to their site.

>  On the subject of seismic activity, it would be informative if there was
> some discussion on what we are seeing when the sensor detects an event.
> Since I am novice in the field of seismology, I have a few of questions.
> 
> 1: In some Earthquakes, why is it that the "P" and "S" waves do not arrive
> at the time where I would expect them to? 

   The way that "expected" travel times are found is only by reference to a
large compilation of seismograms for various epicentral distances.  The
arrivals are averaged and formed into one of the standard earth models (like
IASPEI, PREM, J-B, etc.).  However, since this averaging is done over the
whole globe regardless of location, the result is that differences in velocity
caused by anomalies in various locations are not resolved.  When the ray path
for the waves between a quake and source encounter one of these so-called
"lateral heterogeneities" the expected travel time differs from the real
travel time.  Newer techniques have been used for quite some time now that use
knowledge of the wave ray path and the difference in real travel time and
expected travel time from a model to make a map of these irregularities based
on changes in their seismic velocities.  This is usually called "tomography,"
a word which is borrowed from medical techniques that use X-rays and other
sources.  Tomography is a fairly simple technique that uses a matrix inversion
to solve for the velocity map of the medium, but it produces pretty pictures
of the interior that everyone loves to look at and attempt an interpretation.

   Lateral heterogeneities are probably always caused by changes in chemistry
or temperature in the Earth's interior that are in turn caused by large scale
churning in the mantle.  The reason this churning or convection occurs is
because the core releases heat into the mantle, which is a relatively poor
conductor of heat.  Hence heat pools up in spots, making the mantle material
expand and become more buoyant.  This causes the mantle to deform and move to
accomodate the transfer of heat.  Another important lateral heterogeneity
occurs because some portions of the Earth's crust are much thicker than
others. For instance, the ocean crust is usually half as thick as continental
crust...

> 2: In some quakes, the Event starts with a large Spike and then gradually
> decreases. Why?
> 3: Why do some quakes gradually swell. They look like a rubber garden hose
> that is on the verge of blowing up. The 6.3 in the Solomon Islands on
> 7/9/99 made that kind of an unusual foot print.

Both these questions are related.  There are many possibilities:

1) Source: The simplest answer to your question is that faults move at
different rates and usually in a very jumpy manner.  The first P-arrival is
usually very good to use to identify the behavior of the fault motion that
created the quake.  Also, faults are not point sources, but lie along a plane. 
Deep earthquakes will also look different.  Also, depending on how your
station is oriented with respect to the fault plane will change the manner of
waves coming toward you. 

2) Medium: The second way the shape of the first arrival can be changed is by
the material the waves pass through.  Small differences can be smoothed out in
the Earth when the waves pass through regions with variations in seismic
velocities.  Imagine a bunch of pin balls being dropped into a large pin ball
machine with many bumpers and obstacles.  No matter how you arrange their
pattern when starting the drop, they will bounce around and come out more
randomly at the bottom than at the top.  This randomness is a smoothed version
of the original pattern. 

3) Instrument: Your recording instrument will also impose a change in shape of
the wave arrivals because of its characteristics as a pendulum.  The
seismogram is the "convolution" of the real wave form and your instrument's
"impulse response function."  The impulse response is the way your seismometer
responds to a sharp input spike.  A convolution can be found analytically by
taking the product of the LaPlace or Fourier Transforms for each function and
then inverting back to the time domain.  It can also be found numerically (in
the case of digital data) by summing the two records a(i) and b(j) as
SUM[a(i)*b(i-j)].  If a(i) is the real wave form and b(j) is the impulse
response you can see that a response of b(j) = 1,0,0,0,0,0,... is that of a
seismometer that imposes no change in the wave form and most accurately
represents the ground motion.  The above sum returns the exact copy of a(i). 
However such an impulse response is not physically possible for any real
seismometer.  Because of this fact, the seismogram will always be more
broadened than the real wave form.  The amount of broadening depends on your
specific instrument.  A broadband instrument will partially correct for this
effect for some frequency range by using feedback electronics. 


John Hernlund
E-mail: hernlund@.......
WWW: http://www.public.asu.edu/~hernlund/

******************************************************************************


_____________________________________________________________________

Public Seismic Network Mailing List (PSN-L)