- Request #147 (P. Bodin): Adding a geographic type attribute
--------------------------

Basically I think etype in the AQMS schema should be a unique characteristic
of an origin that specifies the physical process involved in the energy
release (and which might also reflect its location technique due to the nature
of the physical process). However, the current way we use 'etype' in PNSN
doesn't accomplish this, because we also have etypes that are geographical in
nature (i.e., 'regional', and 'teleseism'). We do have a surprising number of,
for example, regional (i.e. Canadian) probable explosions. We can use either
'px' or 're', but not both. I suppose we could come up with another etype
('regional probable explosion'), but this is distasteful.

So, change #1 would be to add a geographic type attribute, for which at the
start we would use local (i.e. inside network), regional, and teleseism. I
could also see this being useful in our historical catalog, where what is a
'local earthquake' today (say in central Oregon) was for earthquakes in 1975
(but still in the same catalog) a 'regional' event.


- Request #148 (P. Bodin): Loading non-volcanic tremor locations
--------------------------

Changes #2 would prepare us for loading non-volcanic tremor locations, and
about this I'm even less confident, and I'm really interested in provoking a
wider discussion. Currently our 'tremor' locations are done entirely outside
of AQMS and cataloged independently, too. (see http://www.pnsn.org/tremor). We
are supposed to explore how to consolidate reporting of these phenomena.
CISN/AQMS etype already permits tremor, of course, but it's a really different
sort of physical process and there are different strategies and schemes for
locating and characterizing it. The method we use involves making envelopes of
5-minute time windows and cross-correlating them. What results is a binary
decision about whether there was tremor during that period and if so, where it
was (an epicenter and perhaps a depth). Currently there are no 'magnitudes'
(but there could be one day, I suppose). But, like I say, this is just one
possible method. etype of tremor is probably just fine in this case...but
would schema changes be required to describe what is actually being reported
and how it was calculated?

------------------------------------------------------------------------
==> Does it belong in AQMS? Is it a Standard or NIC question?
------------------------------------------------------------------------