Southern California Seismic Network
(California Institute of Technlogy (Caltech) and
United States Geological Survey (U.S.G.S) Pasadena Office)
Date: March 9, 1998
Location, Contact, and Address:
Southern California Seismic Network recorded at Pasadena, California on
the Caltech Campus.
Philip Maechling (phil@gps.caltech.edu) 626 395-3801
1200 E. California Blvd. MS 252-21
Pasadena CA 91125
Network Characteristics:
Analog Network:
~250 Analog stations with 350 seperate channels and approximately 20
low gain channels.
Analog data are digitized and processed by the Caltech/USGS Seismic
Processing System (CUSP) .
This system utilizes VAX hardware and the VMS operating system.
Digital Network:
~55 24-bit digital stations with both 3 component broad-band sensors,
and 3 component strong motion sensors.
~25 19-bit digital instruments with 3 component strong motion sensors.
The number of digital data channels recorded continuously exceeds 500
channels due to the
multiple sensors per site, and the various sampling rates at which
our dataloggers generate data.
This network is recorded using a Sun Enterprise 4000 running Solaris
2.5.1
Description of Current Earthworm Usage:
We use the Solaris version of Earthworm for association, location, and
earthquake information distribution. We've introduced our own broadband
picker, to supply picks to Binder, and our own magnitude program for use
with Earthworm. This software runs on our operational real-time seismic
monitoring system. It supplies information to the Caltech/USGS staff during
Southern California earthquakes.
We also have demonstrated waveform and parametric data distribution
from the Caltech/USGS CUSP system over to our digital network processing
system using Earthworm messaging techniques. We implemented Earthworm style
messages and messaging code on the VAX/VMS system and have been able to
distribute data from CUSP to our SUN, running our code on CUSP and regular
Earthworm code on the Sun.
Standard EarthWorm Modules in current Use:
Earthworm code (v2.0) running on a SUN Enterprise 4000 under Solaris-2.5.1
-
binder -Carl's associator
-
eqprelim -rapid reporting of binder associations
-
hypomgr, hypoinv -manage and produce standard USGS hypocenter locations
-
ring2coax, coax2ring, copystatus, statmgr - (control and monitor)
-
ad_demux - This routine converts multiplexed earthworm messages
to de-multiplexed Earthworm trace messages.
Associated programs/modules already developed or adapted here:
-
startstop-modified startstop so that it does not require a console.
-
rtp-Real-time picker. This phase picker was designed to pick our
broadband waveforms. We input these picks to binder.
-
eqmag - This modules inputs the hypoinverse results, and calculates
a magnitude for the the event. We calculate Ml's from our broadband and
strong motion data. By combining eqprelim and eqmag, we generate location
and magnitude estimates quite rapidly, frequently in less than one minute
after the origin time of a southern California earthquake.
-
tpagerfeeder -Our digital processing system used already used some
Unix paging software (tpage) and pagerfeeder and tpage were contending
for the same phone line. So we modified Earthworms pagerfeeder to use tpage.
-
rt2udp - This routine runs on a VAX/VMS CUSP system and distributes
earthworm style waveform data. Standard Earthworm coax2ring running on
a Sun Solaris computer can then read in the Earthworm packets from CUSP
and record our analog data on a Sun computer.
Development Plans:
-
Carl_trig - We are interested in integrating a subnet triggering
scheme into our digital processing system. We plan to evaluate the Carl_Trig
routine as a basis for the triggering declarations.
-
ring2mcast - We are interested in earthworm style data distribution
using IP Multicast rather than UDP broadcast. We believe that multicast-based
Earthworm data distribution will have efficiency advantages.
-
ORB server integration and evaluation - We want evaluate the publicly
available version of the JSP/Datascope software. Both the data analysis
tools, and the ORB server/client software, are of interest to us.
-
WaveServer XX - We have a need for a WaveServer type program. We
want to evaluate the latest and greatest WaveServer.
-
Upgrade to Recent Earthworm - Our Earthworm system has been stable
and hasn't required changes for some time. However, Earthworm development
has marched on and it now appears that we are behind the state of the Earthworm
art. We expect to upgrade to the most recent Earthworm software sometime
during 1998.
-
Year2000 Evaluation - We are going to turn our system clocks ahead
to the year 2000 and see how Earthworm likes it.
Our Development Directions:
-
WavePool/WaveServer - We believe that a WaveServer type application
would provide a valuable way to distribute data after a significant earthquake.
We would like to be able to send a data request to a Northern California
WaveServer and received waveform data immediately over the internet. And
we believe we could offer a service such as that if we can find a robust
WaveServer type program.
-
Database integration - Both our real-time systems and our data center
are moving towards use of commercial databases. We believe that the amount
of data that will be generated by the TriNet/SCSN project will require
high capacity, robust, data storage only available from commercial databases.
-
Integration of Digital and Analog Network Data - We would like to
process both our analog network data and our digital network data on the
same real-time processing system. We are particularly interested in understanding
and learning from the seismic systems which current have this capability
such as IceWorm and SunWorm systems.
Return to main Earthworm or
Sites pages.