--
Edward Cranswick                Tel: 303-273-8609
US Geological Survey, MS 966    Fax: 303-273-8600
PO Box 25046, Federal Center    cranswick@........
Denver, CO 80225-0046  USA      E.M. Forster said, "Only connect".

I thought I would chime in on this issue, as a CA network operator.  
As Bob pointed out, we (in N CA) are swamped trying to maintain our existing
network of stations on what staff we have left after the Reduction in
Force a few years ago.  In addition, adding in exotic instruments
to a real-time network can be a very time consuming task, particularly
when the data are intermittent, noisy, etc.  In other words, if the value-added
is small in relation to the effort, then I probably cannot be actively involved.

In particular, I don't think we have a problem locating earthquakes in CA,
except for the hinterlands.  In well instrumented portions of the state, 
a few more traveltimes is unlikely to improve our locations 
significantly.  Of course, the recent work on doublets using
waveform correlations could change this picture.

Where I see the value of these PSN stations is in strong motion data.
There are several grey-literature papers documenting a need for 1000's
of observations in a metropolitan area, and the PSN sites are potentially
quite useful.  I suspect most PSN operators would be more interested
in running weak-motion sensors, as there is more to see, but with the
improvements in sensor quality and ADC dynamic range, perhaps they can 
have the best of both worlds in an accelerometer.  

A few questions to the group (forgive my complete ignorance of
the PSN):

Does anyone know how many PSN operators exist, say, in the SF Bay
region?  My level of interest goes with the power of 10.

Would it be reasonable to ask that the PSN work with, say, the Earthworm
developers to transmit data via the Internet in a common format, so regional
seismic network operators could limit the amount of software integration?

Likewise, could we ask for some set of standards, so that we don't start
importing data that corrupts instead of enhances our dataset?

-David