Earthquake Hazard Assessment Practice and Velocity Models and Reference Events in the Mediterranean Region

Santa Susanna, Spain

20-25 May 2001

With a Short Course on:

Understanding and Applying Probabilistic

Seismic Hazard Analysis

United Nations Educational, Scientific and Cultural Organization

Paris, France

Contact: Dr. Soren T. Malling

Telephone: 33 1 45 68 41 22

s.malling@unesco.org

Executive Summary

The Mediterranean region, because of its geological structure, seismicity, topography and climate, has been frequently subjected to natural disasters resulting in great losses of life and property. Field studies and investigations of disasters indicate that large portions of the land surface, population, infrastructure, and industry of the region have been subjected to earthquakes in the past or will be subjected to earthquakes in the future.

In recent years, there were several major, damaging earthquakes in the Mediterranean region. Therefore, in the interest of reducing earthquake risk in the Mediterranean region, UNESCO and the Instituto de Ciencias de la Tierra 'Jaume Almera' – CSIC convened a workshop in Santa Susanna, Spain, 20-25 May 2001, on Earthquake Hazard Assessment Practice and Velocity Models and Reference Events in the Mediterranean Region. Contributed papers on seismology and engineering were invited. More than 50 participants from the following countries attended the workshop: Algeria, Belgium, Cyprus, Denmark, Ecuador, Egypt, France, Greece, Israel, Italy, Iran, Jordan, Lebanon, Libya, Morocco, Netherlands, Saudi Arabia, Solvenia, Syria, the Palestinian Authority, Spain, Turkey, United States, and Yemen.

In addition to the workshop, Dr. Robert T. Sewell (USA) presented a one-day training course on: Understanding and Applying Probabilistic Seismic Hazard Analysis and panel discussions were held on Earthquake Hazard Assessment in the Mediterranean Region and on Velocity Models and Reference Events in the Mediterranean Region.

Maria-Jose Jimenez presented the Keynote Address on The Unified Seismic Hazard Map of Europe and the Mediterranean.

Thirty contributed papers were presented that discussed various aspects of earthquake hazard assessment and velocity models and reference events in the Mediterranean region.

Participants were taken on a one-day field trip to the ‘La Garrotxa’ volcanic field, the Volcano Museum of Olot, and the area affected by the XVth century damaging earthquake series.

Introduction

In the EMR, these earthquakes are associated with the northward movement of the Arabian plate. The 1,000 km long western boundary of the Arabian plate is a complex plate boundary, extending from zones of sea‑floor spreading in the Red Sea to zones of plate convergence in Turkey, and lies along the line of the Gulf of Aqaba, the Dead Sea rift, and the Ghab depression. The sense of motion along the Dead Sea transform fault system is left lateral, with the eastern side moving northward relative to the western side. Total displacement is estimated at about 107 km since Oligocene time, with an annual rate of about 0.5 cm. over the last 7 to 10 million years.

On 22 November 1995, a Mw 7.2 earthquake occurred in the central Gulf of Aqaba region causing damage in nearby communities in Jordan, Egypt, Israel, and Saudi Arabia and was felt for more than 700 km. An aftershock sequence lasted for more than one year with numerous shocks exceeding Ms 5.0. The size of the main shock and some of the aftershocks demonstrates the threat that earthquakes pose to the EMR. These events occurred during RELEMR’s (Reduction of Earthquake Losses in the Eastern Mediterranean Region) second Joint Seismic Observing Period (JSOP‑II) and therefore we have the ability to accurately locate the earthquake sequence by integrating data from all the national networks in the region. This permits greatly improved accuracy in epicenter and magnitude determinations. In October 1997, a workshop was hosted by the Cyprus Geological Survey Department to locate the main shock and approximately ten aftershocks.

In the western Mediterranean region, which includes portions of Greece, Italy, Spain and northern Africa, seismicity is widely distributed and seismic hazards are high. Modern interpretations of this seismicity suggest the existence of seven micro-plates, with seismic activity concentrated at the micro-plate boundaries, which coincide with the Alps, Appenines, and Hellenic arc. Among the most notable recent (1996) seismic events was the M=6.8 earthquake in the historic city of Assisi in the Italian Appenines which destroyed numerous cultural artifacts, including important frescoes.

Since 1993, the U.S. Geological Survey (USGS) and UNESCO have been cooperating with EMR earth science organizations under the RELEMR program. Countries from the western Mediterranean region have also participated. The European-Mediterranean Seismological Center (EMSC) has coordinated the exchange of data among EMR countries and, since 1996, LLNL has also been cooperating in the program.

The Spain-2001 Workshop

Background

This workshop continued efforts addressed in similar RELEMR workshops in Amman, Jordan, 4-7 May 1998, in Istanbul, Turkey, 14‑17 October 1998, Nicosia, Cyprus, 3-7 May 1999, and in Istanbul, Turkey, 29 May – 2 June 2000 and 25-27 October 2000. The goals of RELEMR workshops are to foster data exchange among countries in the region, to conduct joint activities and experiments that would improve the quality of seismic data, to improve hazard assessments in the Mediterranean region, to improve the dissemination of earthquake engineering data, and ultimately to improve the seismic provisions of building codes in the region. In Amman, working groups were formed on 1) seismic calibration and 2) the development of a RELEMR seismic hazard map. The 1998 workshop in Istanbul was organized along these themes. The Cyprus-1999 workshop added sessions on the role of auxiliary station operators and their responsibilities and the interaction with their earthquake-reporting activities. The first Istanbul-2000 workshop was dedicated to seismic calibration using the Dead Sea explosions and selected natural events, the second Istanbul-2000 workshop was dedicated to large earthquakes in the region. The program and abstracts for this workshop are presented in Annexes B and C, respectively.

Short Course

UNDERSTANDING AND APPLYING PROBABILISTIC SEISMIC HAZARD ANALYSIS (PSHA)—Presented by Robert T. Sewell

During the workshop held in Istanbul in 2000, S. T. Algermissen presented a two-day seminar on Probabilistic Earthquake Hazard and Risk Assessment. This seminar presented an overall view of earthquake hazard and risk assessment so that the participants would have an understanding of the many scientific and engineering elements required for these kinds of studies. Emphasis was on the types of data required, preparation of the data for hazard and risk assessment, the types of assessment possible and the development of suitable hazard and risk models.

The training course presented at this workshop by R. T. Sewell was designed to allow the participants to acquire an enhanced understanding of how to use PSHA software (FRISKSP). Using a laptop computer and projector, Sewell showed participants how to: hand calculate probabilities of exceedance of ground motions for simple cases, hand calculate a uniform hazard spectrum given hazard curves for spectral accelerations at multiple vibration periods, and incorporate near-source and site-response effects in obtaining PSHA results. He also provided participants with an enhanced understanding of how site-specific seismic design criteria are developed and applied.

Workshop Proceedings

Sunday 20 May 2001

Opening Session

M. Garcia-Fernandez, Instituto de Ciencias de la Tierra 'Jaume Almera' – CSIC, welcomed participants to Spain and to the workshop on behalf of the organizers and sponsors.

S. Malling, UNESCO, indicated that previous RELEMR meetings dealt with the eastern Mediterranean with some participation of western Mediterranean countries and that this meeting is a little different as it formally includes the entire Mediterranean region. RELEMR started in 1992 with the help and cooperation of scientists from the region and we are now trying to produce a regional ground-shaking map(s) at scale(s) appropriate for engineers and planners and move to probabilistic hazard assessments.

M. Foose, USGS, also emphasized the need for regional ground-shaking map(s) and the exchange of data and ideas. He also expressed concern for the lack of progress between meetings and how to keep the effort going between meetings.

F. Riviere, EMSC, indicated that the EMSC gathers data for the region on behalf of the EU and is the data agency for RELEMR.

Invited Dinner Speaker/Workshop Dinner

Maria-Jose Jimenez: Unified Seismic Hazard Model for the European-Mediterranean Region

· Unified seismic hazard map of Europe and Mediterranean programs: SESAME/GSHAP/ESC WG-SHA

· GSHAP produced a heterogeneous set of regional hazard maps from the integration of independent hazard results

· SESAME also tried to produce a unified model of Mediterranean region

§ unified model for sources and attenuation

§ source model: 351 sources + attenuation laws + probabilistic seismic hazard assessment = seismic hazard maps

§ now have a new unified seismogenic source model

§ now have a homogeneous map of Mediterranean region and integration with Europe

§ SESAME/ESC WG-SHA – work wrapping up

§ need continuation and cooperation

· A comprehensive model of hazard assessment for the whole European-Mediterranean region has been developed

· Generation of hazard maps, expressing ground motion in different parameters

§ peak ground acceleration (0.3s and 1.0s spectral acceleration)

§ different soil condition (rock, stiff soil)

§ different probability levels (1%, 10% and 65% of exceedance in 50 years)

· A suite of final hazard maps for the whole European-Mediterranean region are being published in 2001 as a closing activity of the ESC/SC-F Working Group on Seismic Hazard Assessment.

Monday 21 May 2001

Keith McLaughlin: Use of Reference Events to Validate 3D Models in the Mediterranean Region for Location Calibration

· Calibrating the Group 2 stations in the Mediterranean/North Africa region

· Goal is to improve locations with sparse global network

· Developing reference events, models, correction surfaces

· Reference events are used to validate 3-D models in the Mediterranean region to improve location calibration

· Improved locations are based on the sparse IMS network: reduce location bias and uncertainties

· Reference events are needed for 3-D models, travel time curves, modeling errors, and validation

· IASPEI (Bonder and Engdahl) are also collecting reference events

· 3-D ray tracing through model(s), correction surfaces referenced to IASPEI91 model

· Reference events:

§ GT0: known explosion

§ GT2: mine explosion

§ GT5: local network, aftershock

§ Reference events validated by cluster analysis

· Collected 837 GT0-5 events (200 more on the way) and 13,012 regional phases as of 5 April 2001

§ Validation is proceeding

§ CMR is looking for additional reference events in the Middle East/North Africa

· CU Boulder model, Harvard Model (smoother)

· SAIC Model of Crust (Vp & Vs)

· Need phase arrivals for 94 IMS and surrogate stations

· Silesia Poland Mine, Gulf of Aqaba cluster, other clusters available

· Feedback is welcome; presently CMR is preparing preliminary surfaces

· Everything will be posted on web site

Torild Van Eck: ORFEUS and MEREDIAN

· Observatories and Research Facilities for European Seismology

· Mediterranean-European Rapid Earthquake Data Information and Archiving Network

· Funded by the European Community

· Data available within a few hours for Mag. >5.5 globally and >5.0 within Europe

· MEREDIAN goal is the real-time waveform data exchange, open rapid access

· Real-time data gathering via internet, Antelope, AutoDRM, SeisComp (GFZ-Potsdam)

· QuakeExplorer (Anthony Lomax, Univ. Nice): Java based package to access data and do basic analysis

· Real time data requires lots of bandwidth, presently only event data available

· Website: http://orfeus.knmi.nl/meridian

· AutoDRM help is available once the database is set up – contact Florence Riviere or Torild van Eck for help in installing software at receiving institution

· Depends on cooperation, access to internet; does not depend on IMS/IDC

· ORFEUS - earthquake waveform data

· ORFEUS also reports to FDSN (IRIS, Pacific21, Geoscope, Geofon)

· ORFEUS access is free through the web

Avi Shapira: Travel-time modelling in the Middle East

· Based on the Dead Sea Calibration Experiment

· Verified regional travel time model

· Locations based on grid search of first arriving P and S waves with manual weighting

· Assess the minimum picking error, assess the location accuracy, test several models

· Dead Sea shots observed out to teleseismic distances

§ DSE recorded in Cyprus and Greece but none in Turkey (7-11 degrees)

§ DSE known as GT within 50 meters and time within 20 milliseconds

· Location and origin time determined from close in observations

· Observations fit IASPEI91 model

· Scatter in picks of first arrivals is 0.2 – 0.3 seconds (1-sigma)

§ This determines a bound on the accuracy of any model or correction scheme

· One-dimensional model(s) fit the data within the uncertainties

· Large differences in crustal structure between Mediterranean and continental regions

§ EMR does not have a homogeneous crustal structure, however, there is a homogeneous TT model

· Searched for differences in Eastern and Western regions, relative to the Dead Sea Transform

§ GII found 4-layer model (e and w of fault) rms is about .247 sec. from picks and scatter, so this is probably the best in the region

§ currently used model by GII fits the TT measurements of the DSE very well

· A number of models produce essentially the same locations, due to the good distribution of stations. Challenge to IMS and global locations is the paucity of stations.

· Model is probably accurate out to 700-800 km from Dead Sea

· 3 model - unified, west model, east model - what is the effect of location/relocation of DSE?

· All models gave same difference in times based on all (different) models

· What is secret of model independence? Get good distribution of stations

A. Rodgers and A. Al-Amri: Seismic velocity models and ground truth development in the Arabian Peninsula: A U.S.-Saudi Arabian cooperative project

· Results from analysis of Saudi, Kuwaiti, Jordanian and Israeli data

· Used KSU, KACST combined with UCSD network in Saudi Arabia

· Seismicity/crustal structure affect propagation

· Showed interesting event in southern Kuwait (in REB) that might originally been considered induced

§ BB waveform analysis argued that depth was 10km

· Future work may involve: 1) further BB deployments, 2) help with hazard analysis, 3) site response studies, 4) tectonics

· Arabian shield: High Q crust, low Q (slower) mantle

· Sediments thicken toward the Gulf—up to 10 km

· RAYN could not see P-waves from DSE

· Pn velocity 7.75 km/s (slow!) to NW Saudi Arabia along Dead Sea Rift.

· BB waveform modeling

§ Looked at Love and Rayleigh wave group velocities across Arabian shield

§ Crustal thickness in shield and platform is 36 Km

§ Red Sea events can only be modeled after using anisotropy

· Determine structure using waveform modeling

§ Red Sea anisotropy has significant S wave anisotropy

§ Gulf of Aqaba GT events from JSOP (Sweeney)

§ Method works also in Kuwait for waveform modeling

§ Used mining event clustering to find mining events for GT

§ Saudi attenuation so low that Lg can be seen as far as 600 km from mining shots

· Zagros events recorded by ~10 KACST BB stations plus Kuwait BB

§ Will model waveform

· Gulf Aqaba GT events

§ Multiple stations within 30 Km

§ Will relocate KACST data based on manual picks

§ Wants to merge Egyptian data

K. Nakanishi: Data sharing in the EMR

· Joint Seismic Observing Period (JSOP) was very successful, but participation has flagged. Why?

§ Proposed the assembly discuss: 1) distribution of parametric data, 2) velocity models, 3) regional magnitudes, reference events

§ Possible next steps 1) common standards, 2) waveform exchange, 3) autodrm, 4) training for UNIX/LINUX, 5) virtual regional data center

· Some observatories are not contributing to the JSOP

§ Need to develop data sharing and possibly a joint bulletin for consistency

§ Need common software/platform/methodology

§ Broadband seismic data is becoming more common

§ Virtual regional seismic data center? Could serve as a data exchange point.

Hadi Gashut: Earthquake activity in Libya and the importance of the Libyan National Seismograph Network

· Libyan Center for Remote Sensing and Space Science (LCRSSS)

§ Established in 1989

§ Has responsibility for seismic monitoring

· Libya affected by regional tectonics

· Most earthquakes are in the Hun Graben (Al-Qadahia) & Al Jabal Al Akhdar regions

· There is some activity in the south and off-shore in the Gulf of Sirt

· Historical earthquakes

§ As large as M 7.1 (Hun Graben)

§ Date to the year 262

· The LNSN is currently being deployed and consists of both broadband and short-period stations

§ 5 stations (3BB, 2SP) in phase I

§ 5 later (2BB, 3SP) in phase II

§ Will start with 3 stations (NE, NW and S parts of country)

· Collaboration exists with ETH (Zurich) MEDSEA experiment (2 BB stations, 2 year project)

Walter Hays: World Congress on Disaster Reduction

· World Congress on Disaster Reduction

§ Pre-meeting August 2001

Ø Reston, Virginia, USA

Ø 19-22 August 2001

§ Meeting in 2002

· Congress themes

§ Living with disasters

§ Building to withstand

§ Learning from disasters

· Congress goals

§ Sustainability

§ Disaster technical assistance

§ Education

§ Create regional centers of excellence for sustainable development

· Strategic goals

§ Solicit “Global Blueprints for Change” for presentation at the First World Congress

Organize regional forums for change

D. Slejko: Hazard Assessment Practice in Italy

· GNDT seismic hazard project - just completed in Italy

· National project for seismic hazard

§ source data

§ seismotectonics

§ attenuation

§ probabilistic methods

§ deterministic method

· Data catalog with 40000 events

§ only about 4000 events studied

· Delineate Italy and the Adriatic in a zonation map into 80 zones

· Attenuation relations

§ Abraseys

§ Sabetta-Pugliese

· Peak ground acceleration common but not necessary best assessment of hazard - better to use spectral attenuation

· Site effects need to be taken into account

· General hazard projects have ended and now regional projects have started

· Differences between the 1996 and 1999 maps

§ Updated earthquake catalog

§ New completeness approach

§ Weighted seismicity rates over entire catalog

§ No maximum magnitude for the seismic zones

§ Hard boundaries

§ No background zones

§ 2 averaged attenuation relations

§ New attenuation relation for the volcanic seismic zone 73

§ Updated differentiated intensity attenuation relations

· Slejko doesn’t like hard boundries

· Slejko doesn’t know which map is better

· Soil types

§ Very soft

§ Soft

§ Rigid

§ Rocky

· Computed PGA with 475 year return period

· General Hazard project ended—regional projects are now being started

Avi Shapira: Hazard Assessment Practice in the Middle East and Application to Building Codes

· More than 95% of the populations of Israel, Jordan, Lebanon, and the Palestinian Authority are susceptible to intensity VIII

· 27 July 1927 earthquake in Jericho (M 6.2) is the first earthquake with magnitude information

· 1995 M7.1 Gulf of Aqaba earthquake was the wake-up call

· Flow chart - seismogenic zones, b value and return period, attenuation and site effects

· Seismogenic zones made by geologists from seismicity information

§ High segmentation of Dead Sea transform means there will not be a M 8.5 on the DST

§ Not much known on Mediterranean seismicity - no faults recognized

· Many different attenuation functions (at least 10)

· Question about magnitudes - unify?

· PGA not a good measure of hazards

§ New building codes will not include PGA

§ Spectral amplitude at 1 and .3 sec

§ Site classification same as FEMA

· Need to start site specific stochastic simulations

§ Start with Brune model of S waves (omega squared)

· Planning (codes) site specific unified hazard acceleration response spectrum risk: site specific acceleration response spectrum

· Question - is there a resonant frequency for a building?

§ Each building has many different resonant frequencies

· Next generation building code Application

§ Spectral amplitudes @ 1.0 and 0.3 sec

§ Site effects

§ PGA without frequency associated with it is meaningless

Y. Alpay-Biro: Current developments for seismic hazard assessment for Istanbul and Marmara region

· Detailed fault analysis and history for the last 300 years

§ complicated with possible gaps or periods of quiescence

§ history of damage in Istanbul

§ geology of area (soil types—NEHRP soil classification)

§ inventory of buildings

§ GIS data

§ vulnerability curves

· Integrated fault map for region around Marmara

§ Data sources

Ø Infrastructure layouts

Ø Digital maps

Ø 1:1000 scale maps

Ø building census

Ø specific studies

§ Analysis of response to recent major earthquakes

§ Segmented the North Anatolian Fault

Ø Seismicity shows gaps

Ø Suggestion of seismic gap between 1912 and 1999 events along the northern fault

Ø Fault maps of Marara Sea from various sources

Ø Considered models of fault segmentation

Ø Trend exists of seismicity through northern Sea of Marmara

· Istanbul rapid response and early warning project - instrumentation along south crescent of Istanbul (European and Asiatic side)

· Conclusions

§ Results to be used for assessment of risk and motivation for additional resources and mitigation

§ Very high risk in old town Istanbul!

Polat Gülkan: Source modelling and seismological properties of the recent Turkey Earthquakes

· Recent seismic zonation in Turkey

§ Catalog: 23BC – current

§ Standard procedures

§ Big uncertainty remained in attenuation relationships

· Review of Izmit & Duzce events

§ Over turned parked vehicles

§ Near-field ground velocities from strong motion

§ ~20,000 killed

§ ~40,000 buildings lost

§ High velocity fling was well documented at a few sites (normal & transverse)

§ Spectra exceeded “California” design criteria but not greatly so

§ Drift demands were excessive however (over 2% several times) in the near-field

· Seismic zonation map for Turkey

§ Tool for the structural engineer

§ Include 100 years of instrumented seismicity

§ 17 seismogenic zones + background zones

§ Parameters estimated for each zone: b-value, Mmax

§ Attenuation relations only recently obtained for Turkey (this is the largest source of uncertainty)

§ Near field strong motion observations of 17 August 1999 and 12 November 1999 earthquakes not enough, given the recurrence periods of large earthquakes

§ Near field (D<10 km) observations are more complicated and accelerations are extremely high – exceeding the Universal Building Code (UBC)

Maria Yeroyanni: The Mediterranean Disaster Information Network (MEDIN)

· Programmatic presentation of European Commission funding opportunities

§ 38 million EU / year

§ 85 projects total

§ 30 projects in seismic risk

§ Call for proposals from EC covering seismic hazard mitigation and engineering

§ Also funding remote sensing

· Goals

§ Exchange of data

§ Improvement of buildings

§ Technologies to monitor EQ

§ Risk reduction/mitigation

· Going toward larger projects but will entertain small projects

· EU-MEDIN – Workshop to draft declaration paper

§ Sharing of disaster related information, data & expertise

§ Promote existing and ongoing efforts

§ Define research activities, explore relationship w/ International Disaster Information Network (GDIN)

· EU contribution to MEDIN

§ Meeting in Brussels 15-17 Nov 2000

§ Foster coordination of research

§ Declaration paper

§ EU representative added to GDIN steering committee

· Declaration

§ Promote sharing of research results

§ Build on existing work

§ EU-MEDIN steering group

§ Define research activities

§ Explore relationship with GDIN

· Sttering group

§ Review MEDIN objectives/deliverables

§ Redress EU component of MEDIN

§ Discuss possibility of newsletter

I. Cecic: The European Seismological Commission Field Investigation Team (FITESC)

· Survey of damage from several recent earthquakes (photos: Croatia, Central Appinine, NW Slovenia, Turkey)

· Macroseismic Survey Team for rapid deployment after major earthquakes to collect non-instrumental data

· Intensity/effects data will help to assess the risk and vulnerability to earthquake damage

· Need to deploy quickly, before recovery begins (bulldozers destroy the evidence)

· ESC team(s) could train local teams

· Data to be shared and posted on the web

· Cost ~15000 euro/trip for 10 people for 10 days

· Estimate ~ 1 mission/year on average

· www.gsrg.nmh.ac.uk/~phoh/fitesc_disc.htm

Tuesday 22 May 2001

Djillali Benouar: A review of the earthquake data bank in western Mediterranean countries (Algeria, Morocco, Tunisia)

· Seismic hazard assessment (deterministic and probabilistic) for North African countries: Morocco to Egypt

· Need complete and homogeneous earthquake catalog

§ Algeria started in 1980 with the El Asnam earthquake

§ El Asnam earthquake made the government aware of the earthquake risk

§ EQ M=5.5 still kills people—4.5 causing damage

Ø Excess damage due to poor buildings—modern buildings are better

§ Use instrumented seismicity ~100 years

§ Use historical information

· Algerian code as of 1988

· Reconstruct macroseismic field from written reports

· Reduce the risk by reducing uncertainties

§ Need reliable database to start

§ Problem - change in definitions over history – example: Magnitude mb to Ml; same for intensity

§ Take historic catalogs, look at macroseismic and/or instrumental information

· Combine everything about country to reconstruct the intensity assessment

· Algeria now has homogeneous catalog complete with all input data; input in probabilistic method

Mehdi Zare: Recent Developments in Iranian Seismic Hazard Studies

· Population density highest in the north and northwest, northern Zagros, sparse population in eastern part

· Tehran built upon two or more significant faults

§ North Tehran Fault borders Tehran-foothills of Alborz, another fault postulated for the southern edge of Tehran

· Kerman region along western border of the Lut Desert has had great earthquakes

· Very complex fault-fault interactions

· Determine seismic source zones, active faults in urban areas, attenuation laws for strong motions, zoning map

§ Determine seismic source zones from satellite photos and field visits

· 1500 strong motion stations in Iran, including ~1000 accelerographs

· H/V spectral ratio to classify the site response

· Greater attenuation in Zagros Mountains compared to Alborz and Central Iran

· 468 three-component stations (sites) have attenuation laws

· Attenuation Law similar to Joyner-Boore

· There are very few near-field observations

· Stress-drop 20-50 bars, implying intra-plate stresses (?)

· Future work

§ Develop Coulomb stress change maps for fault zones

§ Determine active fault zone in urban areas

§ Develop attenuation laws using recent acceleration data in Iran

§ Generate spectral acceleration maps for Iran (for different periods)

Mehdi Zare: Urban Development and Construction Along North Tabriz Fault Zone in NW Iran: A Field Visit Report (May 2001)

· Tabriz is a historical city—several thousands of years of history

§ 16 major earthquakes—12 demolished the city—last during French revolution

§ Construction on hills gives both landslide and earthquake risk

§ Much of the construction is government sponsored

· Right-lateral strike-slip fault (north Tabriz Fault) is primary with left-lateral strike-slip as a secondary fault

· Construction is approaching the fault

§ 10 years ago, only construction was overnight construction (poor people—migrants)

§ Today: 2 million people—200k in region near fault

§ Presented photographs of recent construction near the fault

Florence Riviere and Gilles Mazet-Roux: Contribution of the EMSC-CSEM to Earthquake Hazard Assessment Practice in the European-Mediterranean region

· Contributions to Joint Seismic Observing Program (JSOP) 1997-2001

§ IIEES in Iran and KISR in Kuwait started sending data 2001

Ø Well distributed events throughout the country

Ø No phase picks from other countries for these events

§ Need to re-establish data exchange

§ Involve new participants

§ Use JSOP to make merged bulletin

§ Consistency of magnitudes

· JSOP-3 started April 1997

§ Open ended

§ Data exchange in JSOP or GSE format

§ 3 to 6 contributing institutes

§ Data available on EMSC web site (phase picks, no waveforms): http://www.emsc-csem.org

§ Mostly magnitude 2-4

§ To submit JSOP data (bulletins): jsop@emsc-csem.org

· Create a European-Mediterranean bulletin by merging all available bulletins

§ Preliminary in 1 week

§ Final in 1 month

§ EPSI project funded by EC

§ EMSC can relocate events by merging bulletins

§ Need events outside of contributing countries so that they can be relocated by merging catalogs

· EMSC alerts for M > 5

§ Automatic mesages from 12 national and local networks and forms

§ Some arrivals and some origins

§ Aim is to send out alert within 1 hour

§ Real time data processing

Ø Magnitude threshold of 5.0 in European-Mediterranean region

Ø Analyst runs relocation with available data

Ø Backup by IGN (Spain) and ING (Italy)

§ 40 alerts in 2000

Gideon Leonard: Model inaccuracy and bias in hypocenter estimation

· Simulations of location performance

§ IASPEI91 travel times with maximum errors of 2.5%

§ Used IMS stations

§ Constructed different scenarios

§ Three event regions: Sea of Galilee, Dead Sea, Gulf of Aqaba

Ø 1000 events with no model errors

Ø 1000 events, modeling error of 2.5%

Ø IDC compensates for model errors

· Bias in hypocenter estimation

§ Travel time and model errors separated

§ Model assumed isotropic

· Conclusions

§ Negative bias by IDC if IASPEI91 is correct

§ IASPEI91 error is 2.5%

§ Next step - need calibrations

Tariq Al-Khalifah: Saudi National Seismic Network

· KASCT network: Saudi National Seismic Network (SNSN)

§ Starting to produce a bulletin

§ Telecommunications has been a problem

§ KSU network started in 1985

§ Royal decree in 1991 for KACST to study earthquakes – but no funding

§ 1995 Gulf of Aqaba earthquake led to funding KACST network

§ Kinemetrics deployed system

§ 37 stations on leased lines

Ø BB 0.001-100 Hz

Ø Concentrated along Red Sea

§ King Abdulaziz University runs 7 station subnetwork

§ All waveform data are archived on DLT tapes

§ A large number of teleseismic events now recorded

§ Converting communications to VSAT

§ Perhaps real time internet connections

§ KSU and KACST networks may be connected to produce joint locations

§ Both systems run ANTELOPE – working on EARTHWORM in parallel

· Provided CD-ROM containing the Saudi broadband data for the Dead Sea Experiment (Data are included on the CD-ROM included with this report.)

· Small network run in Medina area

§ Operated by the Saudi Geological Survey

§ Primarily to monitor volcanics

Djillali Benouar: Hazard Mapping in Algeria

· Hazard mapping in the Maghreb countries

§ Used earthquake catalog described earlier

§ Used different attenuation laws to estimate the uncertainty in probabilistic seismic hazard calculations

§ Probability of exceedance for 10/50 years as a function of PGA for Algiers

§ Period dependent spectral amplitude

· Possibly a seismic gap on Morroco-Algeria border

· No earthquakes within the interior south of the coast

Mohammed Al-Haddad: Seismic Microzonation in Saudi Arabia

· Phase I: First study was of the western coast (Al-Wajah, Yanbu, Jeddah, Jizan)

· Phase II: Gulf of Aqaba region (City of Haql and Ad-Durra Customs facility)

§ Seismic design criteria

Ø Proposed using UBC 94

Ø Some people want to use UBC 97

Ø Aramco wants UBC2000

§ Soil (amplification) can lead to increase in ground motion

§ Microzonation used to account for local site effects

§ Field tests: bore holes to 30-50 m

§ Shear wave velocities used to define the UBC soil type

§ Could not contour the average shear wave velocity

· Measurement of Rayleigh wave dispersion

§ Collect samples for laboratory work

§ Depth to bedrock, water table

§ Field work is done at a number of sites within each city

· City of Haql and Ad-Durra Customs facility

§ 50 sites for shear wave velocities

§ Cross-hole shear velocities for 15 sites

§ Vertical electrical sounding

Ø Depth to bedrock

Ø Water level

· Seismic zonation of western Saudi Arabia

§ Refined seismotectonic map

§ Refined seismic source modeling

§ Sensitivity analysis

§ New seismic hazard analysis (NEHRP-97)

Jalal Al-Dabbeek and Abdel-Hakeem Jawhari: Palestinian Common Buildings and Variation in Seismic Building Codes

· Seismic education for the public

§ Need to work with citizens and decision makers

§ Radio and television to raise awareness of seismic hazards

Ø What do we need to do before, during, and after an earthquake

Ø Al Dabbeek was on 70 TV shows

Ø Need to change mentality of the people

§ Elective courses in seismic hazard/risk mitigation for all students

· Building codes: guidelines and formulas for minimum legal requirements for design and construction in an area.

· Main objective: protect life, health and property

· MERC project on building codes

§ NRA, GII are partners

· Many codes considered: Base shear values for four different building styles (10 story, zone 2) solicited from many engineers are quite variable

§ Regional

§ International

· Factors affecting the seismic base sheer force

§ Seismicity zone

§ Soil effect

§ Occupancy

§ Building

§ Structural system

§ Configuration

Ramy El-Khoury: Examples of earthquake resisting system in Lebanon

· Lebanon is a highly vulnerable county

§ Historically recorded earthquakes back to 1^st millennium

§ Example: 16 March 1956 Chhime-Chouf earthquake

Ø Ms 5.4

Ø 136 killed

Ø 6000 houses collapsed

Ø 17000 houses damaged

§ Many felt earthquakes in Lebanon although most are small

· Building design, modeling and retrofit

§ Examples of earthquake resisting systems in Lebanon

Ø Reviewed retrofitting tank structure

Ø Reviewed earthquake properties/construction techniques of new UN building

Ø Reviewed school: part on rock, part on clay

v Fault in between

v Detailed construction

Polat Gülkan: Near-field ground motion effects and their correlation with structural damage

· Disaster management research center

§ Polat Gulkan – Director

§ Established 1997 at METU in Ankara

§ Fields of interest

Ø Science

Ø Engineering

Ø Law

Ø Culture

Ø Politics

Ø Economic consequences

Ø Mandatory disaster insurance

Ø Land use management

Ø Community participation

Ø Remote sensing and GIS

Ø MEDIN

· Strong motion instruments in Turkey

§ 120 instruments installed (not enough)

§ Most along Anatolian Faults

§ Most located in government buildings for ease of maintenance and communications

Ø Building interaction may be important

§ No borehole data

§ Little or no site composition data

· Fit acceleration spectra to Joyner and Boore model

§ 47 records

§ Limited Turkish data set diverges from JB by factors of 1.5 or more

§ All records in buildings – no free field

§ Generally, attenuation relationships are lower than typical values

· Histogram of S-wave velocity

§ 20 150 m/s

§ 12 300 m/s

§ 9 700 m/s

§ 2 900 m/s

Jeannette Fernandez: The Global Earthquake Safety Initiative (GESI)

· Described methodologies for estimating earthquake deaths based on city-specific studies with an evaluation of landslide potential, fire fighting potential, medical preparedness, …

· Depends on feedback on quality of methodology

§ Principles

Ø Meaningful

Ø Easy to understand

Ø Reasonably priced

Ø Motivational

§ Monitor improvements cities are doing to improve seismic risk

· Determine an earthquake fatality potential

§ Buildings

§ Landslides

§ Search and rescue

§ Fire

§ Medical care

· Steps in making the evaluation

1. Represent shaking on firm ground; average PGA with 10% exceedance in 50 years

2. Represent shaking on soft soil; % of city area with soft soil

3. Define building inventory (9 categories; rate quality: design, construction, materials; select vulnerability curve)

4. Assign buildings to damage states for each building type and % of buildings in each damage state

5. Determine % population in each damage state

¨ Assign each building type to light/heavy category for lethality

¨ Determine % of each building type in each height category

¨ Assume building occupancy

¨ Not all at home—reduce by 25%

¨ Assume % occupants with life threatening injuries =% killed

7. Represent deaths, injuries by landslide

¨ % city susceptible to landslides (steep slopes)

8. Represent lives saved by trained SAR teams

9. Represent deaths, injuries from fires

10.

11. Represent effects of medical care

12. Combine results

13. Repeat analysis for schools

· San Salvador is at risk from landslides and Quito is at risk from earthquakes (biggest problem is collapse of structures)

Gideon Leonard and D. Steinberg: Seismic Hazard Assessment: Simultaneous effect of earthquakes at close and distant sites

· Motivated by planning for emergency response

· Six M>6 in last 1000 years, one of these was a M 7 in Israel/S. Lebanon

· Simultaneous effect of close/distant sites

§ Estimate PGA as a function of epicentral distances

§ Based on J and B 1993

§ Bridges the gap between probabilistic and deterministic earthquake hazard assessments

§ Assume distribution of towns near fault and estimate recurrence intervals

Wednesday 23 May 2001

UNDERSTANDING AND APPLYING PROBABILISTIC SEISMIC HAZARD ANALYSIS (PSHA)—A short course presented by Robert T. Sewell

Expected learning outcome

· Be able to hand calculate probabilities of exceedance of ground motions for simple cases

· Be able to hand calculate a uniform hazard spectrum given hazard curves for spectral accelerations at multiple vibration periods

· Be able to incorporate near-source and site-response effects in obtaining PSHA results

· Acquire enhanced understanding of how to use PSHA software (FRISKSP) to obtain PSHS results

· Acquire enhanced understanding of how site-specific seismic design criteria are developed and applied

· Acquire a broader knowledge of PSHA issues and applications

Course format

· Over 60 key topics

¨ Background concepts (13)

¨ Basic PSHA concepts (18)

¨ Intermediate PSHA concepts (16)

¨ Advanced PSHA concepts (11)

¨ Concluding summary (4)

· Later key topics build upon earlier ones

Handouts

· 166-page workbook

· CD-ROM containing spreadsheets and all workbook information and data (in the pocket on the back cover of this report)

Thursday 24 May 2001

FIELD TRIP to Les Volca de Garrotxa

Participants visited ‘La Garrotxa’ volcanic field, the Volcano Museum of Olot, and the area affected by the XVth century damaging earthquake series.

The Garrotxa Volcanic Zone, near the old town of Olot, contains the finest volcanic landscape in Iberia. Set just south of the Pyrenees, it includes a high plateau, thirty Strombolian volcanic cones, diverse explosion craters, numerous lava flows, basalt cliffs and verdant river valleys. The last eruption probably occurred about 11,500 years ago. Altitudes vary between 238 m and 1,026 m. Well preserved medieval villages affected by the damaging 1427-1428 earthquake series are found in this region and were visited.

Friday 25 May 2001

Rami Hofstetter and Josef Leonov: New developments in seismic source zoning in Israel and adjacent regions

· Zones defined by GII for Israel and adjacent regions from geologic/tectonic data

§ Most events are in 3 narrow rectangles along the Gulf of Aqaba and Dead Sea Rift

§ NW-SE rectangular zone in Northern Israel

· Current ISR network

§ Completeness in Israel for ML > 2.1

§ 1980’s to present

§ Calculated b value for israel and adjacent regions = 0.97

§ ~20 seismic zones

· Homogenous magnitude scale for zonation

§ > 5 1900-1955

§ >4 1955-1971 WWSSN

§ >3.5 1971-1981

§ >2.5 1982-1984

§ >1.8 1984-present

· No magnitudes >5.5

· Used Wiechert magnitude-frequency relation

· Lots of activity along dead sea

· Activity in Gulf of Aqaba is misleading because of 1993-1995 swarms

· Site effects

§ Failure of borehole section transfer function and micro-tremor analysis to explain PGA anomalies at EIL led to modified Joyner-Boore (1981/2) attenuation relationship

· Attenuation law for Israel and the adjacent regions

§ Use to revise hazard map and revision of code

§ Data from 16 earthquakes that triggered the strong motion network

§ New data and current attenuation relation are in disagreement and need reanalysis

§ Need to remove site effect from record with deconvolution

§ To get frequency response function, get all available geological information

Ø Borehole logs

Ø Reflection and refraction data

Ø Geophysical prospective data

Ø Free field ambient noise

Ø Long-term data acquisition

· Data analysis and empirical function determination

§ Nakamura method

§ Reference station technique

§ Receiver functions

· Even with processing, attenuation laws indicate that ground acceleration is too low

· Data scarcity does not allow making new attenuation laws

· Alternative is to look at other laws developed recently

§ Considered 11 published attenuation laws from literature

§ Chose B-J-F (1993/4)

X. Goula: Seismic hazard assessment in Spain: New developments in Catalonia

· Spanish code based on seismic hazard studies of Martin 1984

§ Catalog and seismotectonic zones poorly defined

§ PSHA approach (McGuire, 1976)

§ Accelerations deduced from intensity maps

§ Attenuation relationship deduced from isoseismal lines of major earthquakes

§ Elastic response inferred for three soil types

§ Granada, Andalucia (Southern Spain) have highest hazard

· Building Code is currently being updated

§ Minor revision of eq cat

§ Same seismotectonic zonation

§ Same attenuation relationship

§ New spectral forms for soft soils

· New catalog of Catalonian seismicity available

§ Combined from 4-5 catalogs

§ Includes historical earthquakes

§ Published by his institute and available on their website

· Seismotectonic zonation based on available geophysical/geological data

Antonio Roca: Evaluation of seismic vulnerability of dwelling buildings in Catalonia, Spain

· A method for preliminary evaluation of seismic vulnerability of dwelling buildings was developed and applied to the area of Catalonia in order to evaluate earthquake risk for emergency plans.

§ 6.1 million people, 1 million buildings, 841 towns in Catalonia

§ Objective: Preparation and activation of emergency plans

§ Based on the European Macroseismic Scale (EMS-98) plus parameters easily available in the building census (location, age, and number of floors)

§ Methodology: Classification of risk, estimation of damage, assessment of human casualties, evaluation of economic losses

§ webpage at www.icc.es

§ Can also be used to generate damage scenarios

Ø Recurrence of 1428 earthquake in Ripolles would result in 1600 killed, 135,000 homeless, 26,000 injuries

· New broadband VSAT real time seismological network

§ Provides Civil Defence centres with fast estimation of effects just after an event is detected and located by the network

· Future plans: Use the vulnerability database with a real-time strong-motion network to make real-time estimates of damage patterns

· Project RISK-EU

§ Years 2001-3

§ Earthquake risk scenarios for European Union towns

Sultan Al-Shaabi: Changes in b-value prior to large earthquakes in the Gulf of Aqaba

· During the last fifteen years, the Gulf of Aqaba is one of the most seismically active zones in the Middle East region

§ Seismicity is intense in the Gulf 1985-1996

§ Sequences in 1983, 1990-1, 1993 and 1995

§ July 1993foreshock and many aftershocks, concentrated near BMSH (KSU station)

§ 1990, May 1991 swarms M>4.0 in Southern Gulf of Aqaba

§ 1415 events compiled using data from the KSU network and the adjacent neighboring networks

Ø Time span: 1985 – 1995

Ø Area: Latitude 28°-30° N, longitude 34°-36° E

Ø Duration magnitudes > 2.8

· b-value in the Gutenberg-Richter relation investigated

§ Maximum likelihood method was used

§ b-value has apparently increased sharply (1.46±0.21) before the mainshock-aftershock sequence type

§ Moderate and steady increase in the b-value (0.98±0.14) occurs before the foreshock-mainshock-aftershock sequence

· Conclusion: b-value temporal variation may be an effective method for predicting earthquakes in the area of interest—supports of observation of Smith (1986) that the b-value is a medium term precursor of earthquakes

Salah El-Hadidy Aly Youssef: The Role of the Egyptian National Seismological

Network in the better understanding of the seismicity in Egypt and surrounding countries

· Seismicity of Egypt is mostly in the northeast near Cairo, Sinai, and Aqaba

§ 20 million people in greater Cairo area

§ Historical seismicity (Ambrases 1994) follows instrumental trends

· Seismicity before 1980 only from historical or WWSSN (Helwan) instrumentation

· Activity near the Aswan Dam after dam was filled

§ Ali Gharib studied this activity

§ 1981-1990: Aswan Dam (Nasser Lake) activity concluded to be tectonic

· Additional activity in Eastern Desert probably due to new stations

· KEG started 1990, now replaced with a new NRIAG instrument (STS-2)

· Project to study seismic hazards along the Gulf of Aqaba and around Aswan

§ UNESCO funded

§ Need data from Saudi Arabia to improve the locations in the Gulf

§ Abou Elenean (1997) produced focal mechanisms

· Seismic refraction studies

§ Started in 1972 in cooperation with Makris

§ Many profiles from the Eastern Desert and Sinai

§ Aswan also covered

§ Western Desert poorly covered

§ Crustal thickness about 28 km in Eastern Desert, thickens to west

· Surface wave dispersion at KEG for Mediterranean Sea, Continental paths

§ Seismic P-wave tomography around KEG

§ 100 well-located events

§ Aftershock activity lies along the edges of low velocity body in the middle crust

§ This body corresponds with a gravity low, possibly related to the basaltic flow

§ The Cairo earthquake aftershocks are broadly distributed and it’s hard to associate them with specific faults

Abdul El-Ala Amin Mohammed Sayed Ahmed: Seismotectonic and earthquake hazard studies in Egypt

· Egypt concerned about Nile Valley because of population and development

§ Seismic hazard and vulnerability increasing due to urbanization and development in earthquake prone areas

§ Microzonation map for Egypt

Ø Seismic environment

Ø Seismotectonic model and seismic parameters of the faults

Ø Geology

Ø Soils

Ø Seismic response

Ø Amplification, resulting intensity

§ Maximum intensity zonation map

Ø Maximum intensity in Aqaba

§ Isoseismal intensity maps for a number of events

Earthquake Hazard Assessment Practice and Velocity Models and Reference Events in the Mediterranean Region

Santa Susanna, Spain, 20-25 May 2001