Richard,

Regarding Mb and Ms discrimination of nuclear testing: a very 
simple explanation can be offered:

Mb is the "body wave" magnitude, and is generally derived from the 
amplitude/distance measurements of the P or S wave after they have
traveled through the "body" of the earth., which is a 
compressional wave radiating out from the seismic source. Note
that the polarity of this wave can be negative: ie dilatational,
if your station is located in that quadrant of the focal mechanism.

Ms is the surface wave magnitude, generally derived from the Lg
type waves, which are usually the largest peak to peak amplitude
of the seismic record. These are generated when the body waves reach 
the free surface, and most of their energy comes from the S-waves.

The formulas for determining Mb or Ms are similar, only differing
in the path attenuation parameters. A generic form for Ms is
	Ms = log(A/T) + 1.66*log(delta) + 0.18; A in nanometers,
	T in seconds, and delta in degrees.
Many a PhD thesis has been argued over the values of the constants,
some of which end up "my quakes are bigger than yours, so WE should
get the funding".

For discriminating nuclear tests, the main difference between these
and earthquakes is the focal mechanism, or what happens at the 
source of the event. With a quake, the mechanism is one side of a 
fault segment moving laterally or vertically with respect to the other 
side, over a rupture length that is proportional to the quake magnitude.
The source is a "double couple" that releases about equal P and S
wave energy in a pattern that is a symmetric but mirrored image 
about the fault.

But an explosion is a "point source" that radiates mostly compressive
energy equally in all directions. Nuclear tests do not generate shear
waves, so they produce minimal surface waves. (Strip mine blasts DO
produce large surface waves because of the ripple in the firing).
The Nevada Test Site underground tests were always remarkable in our
records because of the energetic P phase, no S phases, and a wimpy
surface wave. (A further give-away of an NTS event was that they always 
occurred exactly on the UTC hour.)

SO one of the promising methods for policing the CTBT (test ban treaty)
was hoped to be the Mb to Ms ratio. However, this is greatly affected
by the ray paths, and setting off a small nuke off center in a large
oblong cavity can imitate a double couple with shear and surface waves.
Often the blast will release local strain, showing a shear wave.

By the way, I  don't think the improving the S/N of your SG in the range 
from 1 to 20 mhz (1000 to 50 seconds range) would help in this regard, 
since the period of body waves (P and S) are relatively short, in the 
range from 1 to 10 seconds at teleseismic distances.

Regards,
Sean-THomas
__________________________________________________________

Public Seismic Network Mailing List (PSN-L)