home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!usc!isi.edu!venera.isi.edu!tar
- From: tar@ISI.EDU (Thomas A. Russ)
- Newsgroups: ca.earthquakes
- Subject: Earthquake Warnings (was Re: earthquakes during surgery)
- Message-ID: <23282@venera.isi.edu>
- Date: 25 Jan 93 21:57:49 GMT
- References: <1993Jan21.232601.15723@kestrel.edu> <51918@seismo.CSS.GOV>
- Sender: news@isi.edu
- Reply-To: tar@isi.edu
- Organization: USC-ISI
- Lines: 50
- In-reply-to: stead@skadi.CSS.GOV's message of 25 Jan 93 21:17:15 GMT
-
- In article <...> stead@skadi.CSS.GOV (Richard Stead) writes:
-
- [justification for warnings deleted]
-
- > Anyway, I think early warning is justified, given the many possible existing
- > uses. And there are undoubtedly uses that cannot be anticipated. (the only
- > disadvantages are 1) possibility of panic and
-
- It would seem that given the fairly short warning times possible (on
- the order of 5-30 seconds?), that any panic caused by the alarm going
- off would be short-lived (or justified :)
-
- > 2) possibility of false alarms).
-
- This is much trickier to get right. If the false alarm rate were too
- high, then the alarms would end up being ignored. I'm not quite sure
- what rate of false alarms would be acceptable.
-
- Something along the lines of one or two per year might be tolerable,
- although that would still be a large number of false warnings for each
- true one. Damaging earthquakes are really not all that common. Even
- a very accurate (specific) detector for extremely uncommon events will
- generate a lot of false alarms.
-
- Whether this is problematic depends a lot on what actions are taken in
- response to the alarm. Some actions which cause only short term
- disruptions (stepping back from the surgery table, moving fire
- equipment outdoors) won't be a problem. Unfortunately, not all of the
- possible actions that one would want to take have only short term
- ramifications. If you do an emergency shutdown of a nuclear power
- reactor, then you can't just switch it back on a minute later. The
- ramifications can spread much further than just that power plant.
- If a lot of computers crash because of power grid problems in response
- to false alarms, the economic costs begin to rise. If this were to
- happen once a month, I imagine people would not want to put up with it.
-
- I would also add the problem of adequately testing the system.
- Although you can use simulations for a lot of the testing, you won't
- really know whether it works or not until a major earthquake strikes.
- You might end up having to run the system for a decade or more without
- even knowing if it works.
-
- In spite of these reservations, I'd still like to have one for myself
- (if the receiver weren't too expensive).
-
-
- --
-
- Thomas A. Russ tar@isi.edu
- USC/ISI, 4676 Admiralty Way, Marina del Rey, CA 90292 (310) 822-1511
-