Seismology [S]

S33C  ACC:03   Wednesday

Operating High-Performance Seismic Networks: Challenges and Tools II


Presiding: L Gee, US Geological Survey; P Hellweg, Univ. of California, Berkeley

S33C-01 INVITED  

Seismic Data Monitoring at U.S. Tsunami Warning Centers

Whitmore, P M (paul.whitmore@noaa.gov), NOAA/NWS/West Coast and Alaska Tsunami Warning Center, 910 South Felton Street, Palmer, AK 99645, United States
* Ferris, J C (justin.ferris@noaa.gov), NOAA/NWS/West Coast and Alaska Tsunami Warning Center, 910 South Felton Street, Palmer, AK 99645, United States
Weinstein, S A (stuart.weinstein@noaa.gov), NOAA/NWS/Pacific Tsunami Warning Center, 91-270 Fort Weaver Road, Ewa Beach, HI 96706, United States

Seismic data are critical to Tsunami Warning Centers (TWCs) as most tsunamis are caused by earthquakes: 1) directly through static sea floor displacement, or 2) indirectly by sub-aqueous landslides triggered by strong shaking. While tsunami size does not have a direct correlation with earthquake magnitude, seismic data typically provides the first indication that a tsunami may have been generated. TWCs require a certain signal reliability and seismometer density throughout their region which are related to response time and mission requirements. For example, in order to consistently issue messages within five minutes of an earthquake's origin time, the following seismometer network criteria must be met: 1) 12 properly-distributed stations, 2) within 900 km of the source region, 3) 80 percent seismometer uptime, 4) maximum 30 seconds data latency, and 5) high-quality digital, broadband seismic data. This seismic network configuration will provide 9 to 10 P-wave arrivals within 150 seconds of the earthquake origin time. Approximately sixty seconds of the P-wave signal is required to measure moment magnitude for large earthquakes. This provides an analyst 30 seconds to review the event and an additional 60 seconds to compose and transmit the appropriate message (a total of five minutes). Increasing seismometer density will lead to decreased response times, though the decrease in response time is limited by source process time, analyst review, and message creation time. Conversely, decreasing seismometer density and uptime, or increasing latency will have a negative effect on response time. There are currently sixteen regional, national, and global seismic networks supplying data to the TWC's virtual seismic network. In total, approximately 250 stations are monitored in real-time and processed to determine the likelihood of tsunamigenesis. TWCs can monitor data latency and seismometer uptime with earthworm modules, and latency times and outages are summarized yearly to monitor network performance.


S33C-02  

New Perpectives at MedNet

* Mazza, S (mazza@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
Olivieri, M (olivieri@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
Mandiello, A (mandiello@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
Scognamiglio, L (scognamiglio@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
Pondrelli, S (pondrelli@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
Amato, A (amato@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy

MedNet (Mediterranean Network) is a network of very broad band seismic stations installed in countries surrounding the Mediterranean Sea. The project started in 1987, with the aim of providing high quality real-time data to the comprehension of one of the most active seismic regions of the World. Its main objectives were mapping the structure of the Mediterranean region, studying the seismic source properties of intermediate and large events, and applying this knowledge to measures for hazard mitigation and civil protection. Although general goals still hold, the network has now a different valence in the Euro-Mediterranean seismological landscape. National networks are now equipped with similar high quality instrumentation, real time transmission is adopted everywhere for seismic monitoring and infrastructures can support high rate data exchange among networks. MedNet objectives have then shifted toward contributing to real time monitoring of the Euro-Med region. The network presently comprises 22 operating seismic stations installed and maintained in cooperation with 13 geophysical institutions in Italy and in most of the Euro-Mediterranean. All the stations are equipped with 24-bit digitizers and very broad band sensors (Streckeisen STS2, with a few STS1). The MedNet Data Center (MNDC) exchanges data in real time with many seismological observatories, as well as ORFEUS and IRIS DMS. Over 15 years of archived very broad band data are distributed at users request by standard NetDC and AutoDRM protocols (in SEED and GSE formats respectively). Within the EC Project NERIES, MNDC hosts the European Integrated Data Archive, providing support and/or backup to the partner institutions. Presently, fully automatic network functions include: i)daily monitoring of state of health; ii) data recover after link failures; iii) triggered retrieval of event waveforms (with magnitude- and region- specific selection criteria) from continuous; iv) web pages update (http:mednet.ingv.it) for events and station information. Two different techniques for rapid semiautomatic moment tensor solutions are running at MedNet. The first one makes use of the Regional CMT routines (Arvidsson & Ekström, 1998). The procedure is very stable, but the human intervention is essential. The second procedure, as proposed by Dreger & Romanowicz (1994), yields a first unmanned solution, then a successive solution revised by a seismologist. This second approach uses higher frequencies, in order to lower the Mw threshold for which moments are estimated in areas with proper station coverage. In the framework of IOC initiative on "North East Atlantic, Mediterranean and connected seas Tsunami Warning System" (NEAMTWS), INGV is developing a Regional Center for a tsunami watch in the Mediterranean. In this regard, different systems are under test to locate earthquakes in the Euro-Med region, taking advantage of the stations available from the Virtual European Broadband Seismograph Network (VEBSN) and from national networks.


S33C-03  

The Berkeley Digital Seismic Network

Romanowicz, B (barbara@seismo.berkeley.edu), Berkeley Seismological Laboratory, UC Berkeley 215 McCone Hall #4760, Berkeley, CA 94720, United States
Dreger, D (dreger@seismo.berkeley.edu), Berkeley Seismological Laboratory, UC Berkeley 215 McCone Hall #4760, Berkeley, CA 94720, United States
Neuhauser, D (doug@seismo.berkeley.edu), Berkeley Seismological Laboratory, UC Berkeley 215 McCone Hall #4760, Berkeley, CA 94720, United States
Karavas, W (karavas@seismo.berkeley.edu), Berkeley Seismological Laboratory, UC Berkeley 215 McCone Hall #4760, Berkeley, CA 94720, United States
* Hellweg, M (peggy@seismo.berkeley.edu), Berkeley Seismological Laboratory, UC Berkeley 215 McCone Hall #4760, Berkeley, CA 94720, United States
Uhrhammer, R (bob), Berkeley Seismological Laboratory, UC Berkeley 215 McCone Hall #4760, Berkeley, CA 94720, United States
Lombard, P (lombard), Berkeley Seismological Laboratory, UC Berkeley 215 McCone Hall #4760, Berkeley, CA 94720, United States
Friday, J (john), Berkeley Seismological Laboratory, UC Berkeley 215 McCone Hall #4760, Berkeley, CA 94720, United States
Lellinger, R (rickl), Berkeley Seismological Laboratory, UC Berkeley 215 McCone Hall #4760, Berkeley, CA 94720, United States
Gardner, J (jarrett), Berkeley Seismological Laboratory, UC Berkeley 215 McCone Hall #4760, Berkeley, CA 94720, United States
McKenzie, M R (rick), Berkeley Seismological Laboratory, UC Berkeley 215 McCone Hall #4760, Berkeley, CA 94720, United States
Bresloff, C (cyndyb), Berkeley Seismological Laboratory, UC Berkeley 215 McCone Hall #4760, Berkeley, CA 94720, United States

Since it began monitoring earthquakes in northern California 120 years ago, the Berkeley Seismological Laboratory (BSL) has been striving to produce the highest quality and most complete seismic data possible in the most modern way. This goal has influenced choices in instrumentation, installation and telemetry, as well as the investment in expertise and manpower. Since the transition to broadband (BB) instrumentation in the mid- 1980s and to a fully digitally telemetered network in the early 1990s, we have continued these efforts. Each of our 25 BB installations includes three component BB seismometers (STS-1s or STS-2) and digital accelerometers to capture the full range of ground motion from distant teleseisms to large, nearby earthquakes (almost 250 dB). The ground motion is recorded on-site by 24 bit dataloggers. Additional environmental parameters, such as temperature and pressure, are also monitored continuously. Many stations record also C-GPS data that is transmitted continuously to the BSL via shared real-time telemetry. The BDSN's first stations were installed in abandoned mines. In the last 15 years, we developed installations using buried shipping containers to reduce environmental noise and provide security and easy access to the equipment. Data are transmitted in real-time at several sampling rates to one or more processing centers, using frame relay, radio, microwave, and/or satellite. Each site has 7-30 days of onsite data storage to guard against data loss during telemetry outages. Each station is supplied with backup batteries to provide power for 3 days. The BDSN real-time data acquisition, earthquake analysis and archiving computers are housed in a building built to "emergency grade" seismic standards, with air conditioning and power backed up by a UPS and a large generator. Data latency and power are monitored by automated processes that alert staff via pager and email. Data completeness and timing quality are automatically assessed on a daily basis for all stations, and the results are reviewed using the program calqc. Missing data are semi-automatically retrieved from the stations and merged into the data streams, while timing corrections are made manually. Calqc also controls the final transfer of the data to the archive. We assess waveforms from all BDSN components using several tools, such as daily noise spectra. BDSN waveforms are available immediately after they arrive at the data center through the DART (Data Available in Real Time), while archived data can be retrieved from the Northern California Earthquake Data Center. The BDSN contributes data to the joint USGS/UC Berkeley earthquake notification program for northern California, which produces, automatically and in quasi-real- time, earthquake locations and size as well as other source parameters. In particular, three component broadband data from the BDSN are essential for the automatic and rapid determination of Ml, and since 1994, Mw and moment tensors for events with Mw>4. Since 2003, finite source models are calculated for events with Mw>6. The BDSN is part of the California Integrated Seismic Network and contributes data for the determination of ShakeMaps in that framework.
http:seismo.berkeley.edu


S33C-04  

A New Design of Seismic Stations Deployed in South Tyrol

* Melichar, P (peter.melechar@zamg.ac.at), ZAMG, Hohe Warte 38, Vienna, CA A-1190, Austria
Horn, N (nikolaus.horn@zamg.ac.at), ZAMG, Hohe Warte 38, Vienna, CA A-1190, Austria

When designing the seismic network in South Tyrol, the seismic service of Austria and the Civil defense in South Tyrol combined more that 10 years experience in running seismic networks and private communication systems. In recent years the high data return rate of > 99% and network uptime of > 99.% is achieved by the combination of high quality station design and equipment, and the use of the Antelope data acquisition and processing software which comes with suite of network monitoring & alerting tools including Nagios, etc. The new Data Center is located in city of Bolzano and is connected to the other Data Centers in Austria, Switzerland, and Italy for data back up purposes. Each Data Center uses also redundant communication system if the primary system fails. When designing the South Tyrol network, new improvements were made in seismometer installations, grounding, lighting protection and data communications in order to improve quality of data recorded as well as network up-time, and data return. The new 12 stations are equipped with 6 Channels Q330+PB14f connected to STS2 + EpiSensor sensor. One of the key achievements was made in the grounding concept for the whole seismic station - and aluminum boxes were introduced which delivered Faraday cage isolation. Lightning protection devices are used for the equipment inside the aluminum housing where seismometer and data logger are housed. For the seismometer cables a special shielding was introduced. The broadband seismometer and strong-motion sensor are placed on a thick glass plate and therefore isolated from the ground. The precise seismometer orientation was done by a special groove on the glass plate and in case of a strong earthquake; the seismometer is tide up to the base plate. Temperature stability was achieved by styrofoam sheets inside the seismometer aluminum protection box.


S33C-05  

ANZA Seismic Network- From Monitoring to Science

* Vernon, F (flvernon@ucsd.edu), IGPP, UCSD, IGPP UCSD, La Jolla, CA 92o93-0225, United States
Eakin, J (jeakins@ucsd.edu), IGPP, UCSD, IGPP UCSD, La Jolla, CA 92o93-0225, United States
Martynov, V (vladik@epicenter.ucsd.edu), IGPP, UCSD, IGPP UCSD, La Jolla, CA 92o93-0225, United States
Newman, R (rlnewman@ucsd.edu), IGPP, UCSD, IGPP UCSD, La Jolla, CA 92o93-0225, United States
Offield, G (goffield@ucsd.edu), IGPP, UCSD, IGPP UCSD, La Jolla, CA 92o93-0225, United States
Hindley, A (ahindley@ucsd.edu), IGPP, UCSD, IGPP UCSD, La Jolla, CA 92o93-0225, United States
Astiz, L (lastiz@ucsd.edu), IGPP, UCSD, IGPP UCSD, La Jolla, CA 92o93-0225, United States

The ANZA Seismic Network (http:eqinfo.ucsd.edu) utilizes broadband and strong motion sensors with 24-bit dataloggers combined with real-time telemetry to monitor local and regional seismicity in southernmost California. The ANZA network provides real-time data to the IRIS DMC, California Integrated Seismic Network (CISN), other regional networks, and the Advanced National Seismic System (ANSS), in addition to providing near real-time information and monitoring to the greater San Diego community. Twelve high dynamic range broadband and strong motion sensors adjacent to the San Jacinto Fault zone contribute data for earthquake source studies and continue the monitoring of the seismic activity of the San Jacinto fault initiated 24 years ago. Five additional stations are located in the San Diego region with one more station on San Clemente Island. The ANZA network uses the advance wireless networking capabilities of the NSF High Performance Wireless Research and Education Network (http:hpwren.ucsd.edu) to provide the communication infrastructure for the real-time telemetry of Anza seismic stations. The ANZA network uses the Antelope data acquisition software. The combination of high quality hardware, communications, and software allow for an annual network uptime in excess of 99.5% with a median annual station real-time data return rate of 99.3%. Approximately 90,000 events, dominantly local sources but including regional and teleseismic events, comprise the ANZA network waveform database. All waveform data and event data are managed using the Datascope relational database. The ANZA network data has been used in a variety of scientific research including detailed structure of the San Jacinto Fault Zone, earthquake source physics, spatial and temporal studies of aftershocks, array studies of teleseismic body waves, and array studies on the source of microseisms. To augment the location, detection, and high frequency observations of the seismic source spectrum from local earthquakes, the ANZA network is receiving real-time data from borehole arrays located at the UCSD Thornton Hospital, and from UCSB's Borrego Valley and Garner Valley Downhole Arrays. Finally the ANZA network is acquiring data from seven PBO sites each with 300 meter deep MEMs accelerometers, passive seismometers, and a borehole strainmeter.


S33C-06  

The Regional Data Center at the Seismological Observatory Graefenberg (SZGRF)

* Stammler, K (klaus@szgrf.bgr.de), Seismological Observatory Graefenberg, Federal Institute for Geosciences and Natural Resources, Mozartstr. 57, Erlangen, 91052, Germany

The SZGRF operates a regional data center comprising broadband waveform data of the Graefenberg array (GRF) and the German Regional Seismological Network (GRSN). The continuous archive reaches back to 1976 (for GRF) and 1991 (for GRSN) and is completely automatically accessible via WWW and AutoDRM (details see http:www.szgrf.bgr.de) . The total number of permanently archived stations is currently 42. The majority of the stations is transmitted in near-realtime, others are copied with a delay between 20 min and 1 day. The data quality is assured by automatic check procedures and by a daily manual event analysis. Currently, the archiving system is modernized from a semi-automatic procedure to a fully automated system based on a mysql database and self developed archiving tools (Python scripts). The station sites are equipped with STS-2 seismometers and 24bit digitzers. The instruments are installed in vaults keeping stable temperature. The GRF and some of the GRSN sites are additionally equipped with pressure-tight housings to reduce the influence of air pressure variations on the recordings.


S33C-07  

The California Integrated Seismic Network

* Hellweg, M (peggy@seismo.berkeley.edu), UC Berkeley, Seismological Laboratory 215 McCone Hall # 4760, Berkeley, CA 94720, United States
Given, D (doug@usgs.gov), USGS Pasadena, 525 S Wilson Ave., Pasadena, CA 91106, United States
Hauksson, E (hauksson@gps.caltech.edu), Caltech, Seismological Laboratory MS 252-21, Pasadena, CA 91125, United States
Neuhauser, D (doug@seismo.berkeley.edu), UC Berkeley, Seismological Laboratory 215 McCone Hall # 4760, Berkeley, CA 94720, United States
Oppenheimer, D (oppen@usgs.gov), USGS Menlo Park, 345 Middlefield Rd. MS 977, Menlo Park, CA 94025, United States
Shakal, A (tshakal@consrv.ca.gov), California Geological Survey, 801 K St. MS 12-35, Sacramento, CA 95814, United States

The mission of the California Integrated Seismic Network (CISN) is to operate a reliable, modern system to monitor earthquakes throughout the state; to generate and distribute information in real-time for emergency response, for the benefit of public safety, and for loss mitigation; and to collect and archive data for seismological and earthquake engineering research. To meet these needs, the CISN operates data processing and archiving centers, as well as more than 3000 seismic stations. Furthermore, the CISN is actively developing and enhancing its infrastructure, including its automated processing and archival systems. The CISN integrates seismic and strong motion networks operated by the University of California Berkeley (UCB), the California Institute of Technology (Caltech), and the United States Geological Survey (USGS) offices in Menlo Park and Pasadena, as well as the USGS National Strong Motion Program (NSMP), and the California Geological Survey (CGS). The CISN operates two earthquake management centers (the NCEMC and SCEMC) where statewide, real-time earthquake monitoring takes place, and an engineering data center (EDC) for processing strong motion data and making it available in near real-time to the engineering community. These centers employ redundant hardware to minimize disruptions to the earthquake detection and processing systems. At the same time, dual feeds of data from a subset of broadband and strong motion stations are telemetered in real- time directly to both the NCEMC and the SCEMC to ensure the availability of statewide data in the event of a catastrophic failure at one of these two centers. The CISN uses a backbone T1 ring (with automatic backup over the internet) to interconnect the centers and the California Office of Emergency Services. The T1 ring enables real-time exchange of selected waveforms, derived ground motion data, phase arrivals, earthquake parameters, and ShakeMaps. With the goal of operating similar and redundant statewide earthquake processing systems at both real-time EMCs, the CISN is currently adopting and enhancing the database-centric, earthquake processing and analysis software originally developed for the Caltech/USGS Pasadena TriNet project. Earthquake data and waveforms are made available to researchers and to the public in near real-time through the CISN's Northern and Southern California Eathquake Data Centers (NCEDC and SCEDC) and through the USGS Earthquake Notification System (ENS). The CISN partners have developed procedures to automatically exchange strong motion data, both waveforms and peak parameters, for use in ShakeMap and in the rapid engineering reports which are available near real-time through the strong motion EDC.
http:www.cisn.org


S33C-08  

The Integrated Seismic Monitoring System in Italy

Cattaneo, M (cattaneo@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
* Amato, A (amato@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
Anzidei, M (anzidei@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
Badiali, L (badiali@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
D'Anna, G (danna@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
Delladio, A (delladio@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
Mazza, S (mazza@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
Francesco, M (f.mele@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
Michelini, A (michelini@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
Salvaterra, C (salvaterra@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy
Giulio, S (selvaggi@ingv.it), Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata, 605, Rome, 00143, Italy

In the past 5 years the Italian monitoring system, managed by the Istituto Nazionale di Geofisica e Vulcanologia (INGV) has been strongly improved, thanks to support from the National Dept. of Civil Protection (DPC). Basic goals of the network are the rapid evaluation of earthquake parameters and the acquisition of high quality data for seismological research on earthquake source and deep structure. Particular emphasis has been given to real- time data transmission, network robustness and redundancy, multisensor remote stations. Presently, data from more than 250 stations are received in real-time at the INGV acquisition system in Rome, mostly through satellite links (about 100) and terrestrial digital lines (about 100), while other data are received through radio links and Internet. Our stations are equipped with either commercial (mostly Nanometrics) or INGV- produced digitizers (GAIA2). MedNet data in Italy and EuroMed regions are also used for real time monitoring. MedNet data are shared in real-time with ORFEUS and IRIS, while data from the National Seismic network are presently distributed for triggered events a few minutes after an earthquake. Data latencies are mostly in the range 0-10 s, with the upper limit (constrained by satellite channel sharing) presently being reduced down to 3 s. Most of the stations are equipped with either Trillium 40s, 120s or 240s, or STS2 seismometers, coupled with Episensor accelerometers. In about 100 sites, continuous GPS's send stream data in real time to the INGV centres, at 1s or 30s sampling rate. Data from GPS receivers are available at http:ring.gm.ingv.it for a subset of the network. A few dense dial-up local networks in central Italy contribute to the monitoring system. Beside a centralized acquisition in Rome, data from wide regions covered by the network are transmitted in real time to other INGV centres (Grottaminarda, Catania) to guarantee redundancy and disaster recovery. In volcanic areas, such as Etna and Vesuvio, dense multiparametric networks are operated in real time from INGV centres in Napoli and Catania. Present targets of the national network, which is still under development, include off-shore monitoring through OBS development and connection, integration with strong motion national network, Shake Maps rapid computation (http:earthquake.rm.ingv.it/shakemap/shake/), early warning for targeted areas.