Paul,

I am getting the impression that you did not see the NEIS info
about the 20 April quake? VIS:

The following is from the United States Geological Survey,  National
Earthquake   Information   Center:    Preliminary   hypocenter   for
earthquake of 2000 Apr 20, NEW YORK:  latitude 44.0  degrees  north,
longitude  74.3  degrees  west,  origin  time  08 46 54.0 utc, depth
shallow, magnitude 3.7 mbLg.  The earthquake was felt in eastern New
York  and  at  Montpelier,  Vermont.   There have been no reports of
damage.  This is located in the same general area as a magnitude 5.1
earthquake on October 7, 1983, that caused minor damage and was felt
in 12 U.S.  states and 2 Canadian provinces.
Stations used:  NCB P 084656.2 HNH P 084720.5 LBNH P 084724.7
BINY P 084731.3 LSCT P 084734.9 WES P 084737.1 PAL P 084742.0
SSPA P 084800.5 MCWV P 084823.8

This is their calculation of the origin time based on the P arrival time
at these 9 stations. Any additional arrival time data should only refine 
the origin time estimate. A single seismogram can only provide the unique
arrival time at the station site; the reading is used with data from other
stations, and a minimum of three are needed to calculate a hypocenter
location, which results in the origin time estimate.

A VERY approximate method we use with the wall maps is to use the P and 
S times to guess the distance of a station to an event, letting each 1 
second difference to represent 8 kilometers of travel time. This lets us 
draw three circles on the wall map around the reporting stations, so their 
intersection is the epicenter guess.

An origin time can also be guessed at with a single station P and S,
again using 8 seconds/S-P time to get a distance, then figuring the
P travel time. SO if S-P is 10 seconds, the event is about 80 km from
the station (this is only good out to about 200 km). At a velocity
of 5 km/second, the P wave took 16 seconds to get to the station, so
the origin time is the P time minus 16 seconds. THis VERY approximate
estimating can be improved by averaging the available data. 

MUCH better results are obtained using hypocenter programs that use
recursive algorithms to minimize the error ellipses.

Regards,
Sean-Thomas
__________________________________________________________

Public Seismic Network Mailing List (PSN-L)