As we approach our next Assembly on 9 April inNice, we can report that six new members will bejoining the EMSC community: SeismologicalInstitute (Albania), CRAAG (Algeria), Universityof Zagreb (Croatia), Institute of Geophysics(Georgia), DEUC (Monaco) and MontenegroSeismological Observatory (Yugoslavia), So, ourmembership continues to expand both in ournumber of institutions, and through ourgeographical coverage. We are now 56organisations in 36 countries spread throughoutour region of focus. As an agreed policy, we havecontinued to maintain a modest subscription feeto encourage even wider participation in order togain from the increased communal benefits whichthis brings to the service we provide, both to ourscience and to the populations of our membercountries. More affluent members continue to beencouraged to contribute more than one unit ofsubscription to compensate for this strategy ofmaintaining the basic fee without indexation forinflation.

In two further initiatives, EMSC is seeking toraise its profile whilst earning revenue to supportthe co-ordination centre in Bruyeres. Aconsortium bid to the European Space Agency,with partners from both the public and privatesectors, has been successful. It is within ESA’sGMES programme (Global Monitoring For theEnvironment and Security), and covers initialphases through 2003/04. A strong delivery againstproject goals in this period will place us in aposition to bid into a very substantial secondphase covering a further 3 years. Members areencouraged to help the core group, whenrequested, to support its mission of engaging withcity authorities concerned with ground movementimpacts from all sources; tectonic, geological andman-made. The approach is to map and monitorsuch movements at millimetric resolutionutilising new processing techniques (PSINSAR)on the ten years’ of satellite radar data which nowexists in the archives.

In its second new initiative, EMSC ispartnering ORFEUS in leading anFP6 bid to build a stronger networkinfrastructure for Europeanseismology, supported by newresearch projects. The proposal willbe completed and submitted in April(for more details, see the Newssection below).Against the above positive moves, I bring to youthe sad news of the loss of one of our strongestsupporters who has provided our rapid momenttensor service through our Key Nodal Member,GFZ. Guenter Bock was tragically killed when theplane he was travelling in crashed-landed inLuxembourg on 6 November. Our condolences andthoughts continue to be with his family, friendsand colleagues, in Germany and across the World.In addition to the considerable impact he has hadin our science and in his unstinting service to theEMSC, its Executive and its members, Guenterwas a true gentleman, liked and respectedeverywhere. We shall miss him in Nice and in thefuture.

Chris BrowittPresident

EDITORIAL

Centre Sismologique Euro-MéditerranéenEuropean-Mediterranean Seismological Centre

www.emsc-csem.org

N° 19 APRIL 2003

Newsletter

No ISSN : 1607-1980

CONTENT

• The ESC-SESAME Unified Hazard Model for the European-Mediterranean region p. 2

• B.U. KANDILLI OBSERVATORY and EARTHQUAKERESEARCH INSTITUTE SEISMOLOGY LABORATORY p. 5

• Observations and monitoring of the seismicity in Bulgaria p. 8

• News of the EMSC p. 11• Seismic Hazard Assessment for Nuclear Power

Plants Safety Laboratory (BERSSIN),Radioprotection and Safety Nuclear Institute (IRSN, France). p. 12

Guenter Bock

1. Projects on seismic hazardassessment in Europe and theMediterranean: objectives,strategies and results

During the last ten years several projects onseismic hazard assessment were active atglobal and regional scales. Within theEuropean-Mediterranean region a number ofmultinational programs were set up toproduce earthquake catalogues, seismic sourcezoning and hazard assessment, through thefollowing three main project frameworks: (1)GSHAP, (2) IGCP-382 project SESAME, and(3) the ESC Working Group on SeismicHazard Assessment.

Within the framework of GSHAP (GlobalSeismic Hazard Assessment Program, 1992-1999), a UN/IDNDR demonstration program,which completed in 1999 the first global mapof seismic hazard in terms of peak groundacceleration (Giardini, 1999); IGCP-382SESAME (Seismotectonics and SeismicHazard Assessment of the MediterraneanBasin, 1996-2000), which provided the firstunified seismic source model andhomogeneous assessment of seismic hazardfor the whole Mediterranean region (e.g.Jiménez et al., 2001); and the EuropeanSeismological Commission Working Group onSeismic Hazard Assessment (ESC/WG-SHA,1996-2002), aiming at the development of ahomogeneous probabilistic seismic hazardassessment procedure for Europe and theMediterranean, the whole European-Mediterranean region has been unified.

GSHAP produced in 1999 the first seismichazard map for the European-Mediterraneanregion in terms of peak ground acceleration, aspart of the GSHAP global hazard map, andwas based on the compilation and assemblageof hazard results as obtained independently indifferent test areas and national andmultinational programs. As was pointed out inGrünthal et al. (1999), although all of theseindependent hazard maps were producedfollowing the same basic seismotectonicapproach, the harmonization of the hazards inthe assemblage of the final GSHAP maprequired several iterations of smoothing andborder matching between the differentregions. The greatest difficulties were met inthe Mediterranean, owing to the large numberof independent areas.

IGCP-382 SESAME developed andcompleted a more detailed, integrated seismicsource model and homogeneous hazardmapping for the Mediterranean region. Mainefforts focused in the development of a unified

source model throughout the region to allowfor a homogeneous hazard assessmentprocedure. The strategy was based on theintegration of regional and national models toavoid ambiguities coming from differentapproaches, and also to avoid gaps in thegeographical coverage through thedevelopment of new source models in areaswhere these were not yet available.Preliminary SESAME results were presentedin September 2000 on occasion of the XXVIIGeneral Assembly of the EuropeanSeismological Commission, in Lisbon,Portugal. Improved results incorporatingupdates to source model and hazardcomputation can be found in Jiménez et al.(2001).

ESC/WG on SHA has completed in 2002 aunified seismic hazard modeling for Europeand the Mediterranean. Our approach toobtain a reference seismic hazard model forEurope and the Mediterranean has beenentirely based on the integration of regionalmodels and the adoption of a homogeneoushazard assessment procedure. The strategy

was based on integrating GSHAP CentralNorthern Europe results with those fromSESAME for the Mediterranean to allow forthe first ever homogeneous seismic hazardcomputational procedure which for the firsttime is based upon a unified source modelthroughout the whole European-Mediterranean region. This comprehensivemodel for seismic hazard assessment allows,for the first time, the generation of hazardmaps, expressing ground motion in differentparameters, for different soil conditions andprobability levels.

2. Development of a unifiedseismic hazard model for theEuropean-Mediterranean region

The European-Mediterranean ESC-SESAMEunified seismic hazard model is based on theSeismotectonic Probabilistic approach andthus based on a regional model of seismicsource zones (established according to tectonic,geophysical, geological and seismological data)with associated parameters (magnitude-frequency parameters, maximum expected

CSEM /EMSC Newsletter

2

The ESC-SESAME Unified Hazard Model for the European-Mediterranean region

M-J. Jimenez1,2, D. Giardini1, G. Grunthal31. Swiss Seismological Service, ETHZ, Zurich, Switzerland,

2. Institute of Earth Sciences-CSIC, Barcelona, Spain,3. GeoForsschungsZentrum, Potsdam, Germany

Figure 1: Unified seismogenic source model for the European-Mediterranean region (463 source zones).

magnitude), through which expected groundmotion is computed based on an appropriateattenuation relationship.

The unified source model consists of atotal of 463 seismic sources (455 shallow and 8intermediate-depth). Figure 1 shows the finalsource model. Each source is characterized bythe corresponding seismicity parameters interms of minimum and maximum magnitude,and earthquake occurrence rates with anassociated sub-catalogue which stems fromthe corresponding regional catalogue. Sourcemodels developed in regional and multi-national programs within GSHAP have beencompiled and then complemented withexisting models in the literature to avoid gapsin the geographical coverage. Originalbackground sources, established in theindividual models to account for seismicity inneighbouring regions, have been eliminated;and new zones at overlapping border areaswere redesigned to harmonize geometrieswhere differences existed. These areas mostlycorrespond to the Pyrenees, the Alps, theCarpathians, Northern Greece and theAegean, among others. In the Mediterranean,a new regional model for the EasternMediterranean region has been developed incooperation with GII (Geophysical Institute ofIsrael), within SESAME and RELEMR(Reducing Earthquake Losses in the EasternMediterranean Region) programmes. Atdifferent stages during the development of thework, regional source models and associatedparameters have gone through improvementsand updates according to any new informationmade available.

Ground motion attenuation modelsdeveloped by Ambraseys et al. (1996) in termsof peak ground acceleration, PGA, andabsolute spectral acceleration, SA, areconsidered to be adequate for the unifiedcomputations for shallow sources, since theserelationships were obtained on the basis of awide European strong motion data set withmagnitudes between 4.0 and 7.9 and fourcategories of soil condition (rock, stiff, soft andvery soft soil). Specific attenuationrelationships are considered for the eightsources of intermediate-depth seismic activitythrough the specific attenuation relationshipsderived in Musson (1999) for Vranceaintermediate source, and in Papaioannou andPapazachos (2000) for intermediate-depthseismic activity sources in the Helenic Arc.

Homogeneous Hazard computation iscarried out inside the area stretching from10ºW-30ºE and 27ºN-72ºN and 30ºE-40ºE and27ºN-47ºN at a grid interval of 0.15 degreesand is performed through SEISRISK III(Bender and Perkins, 1987). Non-isotropicattenuation for intermediate-depthearthquakes originating in Vrancea (Romania)is handled and computed independently byapplying the procedure and code used for theregional hazard mapping of North Balkanregion (Musson, 1999). Ground motion

variability is incorporated in the computationsassuming a lognormal distribution of theground-motion parameter with standarddeviation sa. The number of computationnodes is over 70,000. To ensure that thecomputation through the established unifiedprocedure gave fully compatible results withthe original regional hazards, individual testswere performed for all regions to detectpossible misfits and therefore identify thecausative reasons. The resulting differences inthe hazard results through the unifiedprocedure should arise solely in relation to theharmonization of the basic input data (e.g.,source geometries at border areas, attenuationrelationship) or specific to the computationsfor a large geographical region (e.g. larger gridspacing).

Generation of regional probabilistichazard maps on the basis of the developedunified seismogenic source model, and theadopted regional and specific ground-motionattenuation relationships, is carried outthrough a homogeneous probabilistic seismichazard assessment (PSHA) procedure. Itallows for the first time to obtainhomogeneously computed regional hazardmaps for the European-Mediterranean regionin terms of different ground motionparameters (e.g. PGA, 0.3s SA, 1.0s SA),

different soil conditions (e.g. rock, stiff soil)and different probability levels (e.g. 1%, 10%and 65% of exceedance in 50 years). The mapin Figure 2 depicts the results of homogeneousseismic hazard computation of peak groundacceleration at a 10% probability ofexceedance in 50 years for stiff soil; areas inthe map not covered by the ESC-SESAMEseismic source model (Iceland and Russia) aretaken from the GSHAP Global Seismic Hazardmap.

3. ESC-SESAME main resultsfor PSHA in Europe and theMediterranean

Main results achieved through the European-Mediterranean final unified model for PSHAcan be summarized as follows:

• First ever common model of seismic sourcesfor Europe and the Mediterranean

• Hazard computations are now based on aunified source model of 463 seismic sources(455 shallow and 8 intermediate-depth)

• Homogeneous computational procedure forPSHA

• Generation of hazard maps: ground motionexpressed in different parameters, fordifferent soil conditions and probability levels

3 April 2003

Figure 2: ESC-SESAME European-Mediterranean seismic hazard map for peak ground acceleration [g] with 10% probability of exceedance in 50 years for stiff soil condition.

• Establishment of databases incorporatingfor each seismic source: seismicityparameters (minimum and maximummagnitude), earthquake occurrence rates,and associated subcatalogue (from regionalcatalogue)

Publications, reports, procedures, maps andresults will be loaded on the web and the finalseismic hazard map for Europe and theMediterranean (peak ground acceleration at a10% probability of exceedance in 50 years forstiff soil) is now published under the auspicesof the European Seismological Commission in5000 copies by the Institut Cartogràfic deCatalunya in March 2003.

4. OutlookThe ESC-SESAME is the first ever unifiedmodel for PSHA for Europe and theMediterranean. It was developed within theframework of several recent projects onglobal and regional seismic hazardassessment and allows for homogeneoushazard computation throughout the wholeEuropean-Mediterranean domain. Still someaspects in its realization have remainedunavoidably heterogeneous. Futuredevelopments to harmonize and improvemodels and data can be achieved in theframework of future initiatives at Europeanlevel through regional close-cooperation andefforts in reasonable periods of time, butthese cannot go beyond the limits posed bythe differences in the status on backgroundknowledge and quality of the basic data.These differences, if existing, will remainunsolved and will reflect unavoidably in anyfinal regional hazard map.

Nevertheless, this final unified hazardmodeling for Europe and the Mediterraneanwill contribute to the establishment of aregional seismic hazard framework for theregion in terms of peak ground and spectralacceleration from which seismologists,geologists and earthquake engineers canprofit as a general guideline.

The compiled data bases (e.g. source zoning,attenuation, seismic activity parameters) forthe whole European Mediterranean domainand the homogeneous hazard computationscheme constitute a unique tool which opensnew possibilities for future research ofinterest to the seismological and engineeringcommunities. The ESC-SESAME backgroundhazard model for PSHA can serve for re-evaluation of hazard according to differentcriteria or for improved source modelsincorporating mixed areal/fault sources, forimproved ground motion models (both forsub-regions or for the whole European-Mediterranean region), as the basis forcomparative regional studies dealing withboth methodological and assessment issues,also as an aid to model seismicity inneighbouring regions for national hazardmaps, to establish the basis for a European-

Mediterranean seismic hazard server, and foreducational projects, among many otherapplications.

5. Events where the differentstages in the developmentand results were presented

At different stages on the development of theESC-SESAME unified seismic hazard model,results where presented on occasion of :

• XXVII General Assembly of the EuropeanSeismological Commission, Lisbon,Portugal, 10-15 September 2000.

• Mitigation of Seismic Risk. Support toRecently Affected European Countries,EC-Joint Research Centre, Belgirate,Italy, 27-28 November 2000.

• XXVI General Assembly of the EuropeanGeophysical Society (EGS). Nice, France,26–30 March 2001.

• American Geophysical Union Fall Meeting,San Francisco, 10-14 December 2001.

• 3ª Asamblea Hispano-Portuguesa deGeodesia y Geofísica, Valencia, 4-8 February2002.

• 12th European Conference on EarthquakeEngineering, Londres, UK, 9-13 September2002

• XXVIII General Assembly of the EuropeanSeismological Commission, Genoa , Italy,1-6 September 2002.

• Primer Centenario del Observatorio deCartuja: 100 años de Sismología enGranada, Granada, Spain, 8-11 October2002

a number of invited conferences where givenat :

• European Seismological CommissionWorkshop on “Seismicity Modeling inSeismic Hazard Mapping”, Poljce,Slovenia,22-24 May 2000

• XXV General Assembly of the EuropeanGeophysical Society Nice, France, 26 April2000.

• UNESCO Workshop on Earthquake HazardAssessment Practice and Velocity Modelsand Reference Events in the MediterraneanRegion, Santa Susanna, Barcelona, Spain,20 May 2001.

• PILAR (Program For Increasing TechnicalCapacity on Natural Disaster Reduction inthe Mediterranean Region) planing Meeting,UNESCO, Paris, 24 June 2002.

and a special session on:

• “European Seismology Projects for Hazardand Risk: Sesame, EC8 and the Way Ahead”at the 12ECEE meeting in London,September 2002, was convened by R.Musson as an open discussion to provide aforum to discuss the results achieved,

actual status and future direction ofearthquake hazard research, andsupporting projects, in Europe.

6. What made it possible? The contributions on data and efforts of manyyears of work of many individuals andinstitutions which were active in differentprojects related to hazard in Europe and theMediterranean, specially all those groups andindividuals active within GSHAP, SESAMEand the ESC/SCF WG on SHA, have made itpossible. In particular, Mustafa Erdik,Mariano García-Fernández, Roger Musson,Christos Papaioannou, Avi Shapira, DarioSlejko, for your patience and support - thankyou!

We are also grateful to every contributor at thedifferent stages of development of the differentprograms and projects as referenced in thepublished ESC-SESAME seismic hazard map,but to name all of them here would beimpractical.

Propietary software for hazard computationwas made available for ESC-SESAME by R.Musson (BGS,UK). Figures were preparedusing GMT software (Wessel and Smith ,1998).

7. ReferencesAmbraseys N.N., K.A. Simpson and J.J.Bommer.1996. Prediction of horizontal responsespectra in Europe. Earthq. Eng. Struct. Dyn., 25,371-400.

Giardini, D. Editor.1999. The Global SeismicHazard Assessment Program 1992-1999. SpecialIssue. Annali Geofis., 42 (6).

Grünthal G., C. Bosse, S. Sellami, D. Mayer-Rosaand D. Giardini. 1999. Compilation of theGSHAP regional seismic hazard for Europe,Africa and the Middle East. Annali Geofis., 42,1215-1223.

Jiménez M.J., D. Giardini, G. Grünthal, andSESAME Working Group. 2001. UnifiedSeismic Hazard Modelling throughout TheMediterranean Region. Boll. Geof. Teor. Appl.,42, 3-18.

Musson R. 1999. Probabilistic seismic hazardmaps for the North Balkan region. AnnaliGeofis., 42, 1109-1124.

Papaioannou C. and C. Papazachos. 2000. Time-Independent and Time-Dependent SeismicHazard in Greece based on Seismogenic Sources.Bull. Seism. Soc. Am., 90, 22-33.

Wessel P., W. Smith. 1998. New, improvedversion of Generic Mapping Tools released. EOSTrans. Am. Geophys. U. 1998; 79(47): 579.

CSEM /EMSC Newsletter

4

5 April 2003

Mission of Seismology laboratoryThe seismological laboratory of the KOERI is consisting 67 seismic stations is the center for data analysis and it determines the parameters ofearthquakes (origin time, epicenter coordinates, depth, magnitude, intensity) that occur in Turkey.

B.U. KANDILLI OBSERVATORY and EARTHQUAKE RESEARCH INSTITUTE

SEISMOLOGY LABORATORYKalafat

Kandilli Observatory, Seismology Lab., 34680 Çengelköy- ISTANBUL, TURKEYDogan

Events are collected continuously at thecenter from all of the stations in Turkey byreal time. At the same time, those earthquakeinformation are disseminated as rapidly aspossible to government agencies, togovernment public information centers and tonews media.

The Seismology laboratory provides a twenty-four hour information service continuouslyand it also monitors the seismic activitythroughout Turkey even on holidays. All of the

historical information and parametric dataare collected at the center as a databank thus,that plays a very important role forinstitutions and researchers. After adestructive earthquake, the earthquakeinformation is also disseminated to emergencyand civil centers in order to mitigate theseismic risk and panic.

Our laboratory provides the important seismicinformation for preparing of theseismotectonic and seismic risk maps of thecountry. Today it is unique organization thatgives continuously data to internationalseismological centers such as NEIC, ISC andEMSC through internet.

KOERI continues to provide SeismologicalObservation services with its continuouslyexpanding network throughout Turkey.

OrganizationThe Kandilli Observatory and EarthquakeResearch Institute (KOERI) is Turkey’s uniqueorganization encompassing earthquakesobservation, research, education and

application services within a single, integratedbody. Established 135 years ago in 1868 as Theimperial Meteorological Observatory, KandilliObservatory extended its activities in the 20 thcentury into various observational fields suchas astronomy, astrophysics, geophysics,geomagnetism and seismology. In 1982 theKandilli Observatory was joined to BogaziciUniversity. A new Earthquake ResearchInstitute was founded to merge with theObservatory. The new body, named KOERI,was reorganized with its main emphasisoriented towards earthquake, research,education and relevant service actives.

CSEM /EMSC Newsletter

6

Through the 1980’s KOERI has evolved into a multidisciplinary research organization providing graduate education to M. Sc. and Ph.D. levels in threedepartments, namely Earthquake Engineering, Geophysics and Geodesy, with full-time and part-time faculty members.

Seismic Network andInstrumentsThe Kandilli Observatory and EarthquakeResearch Institute (KOERI), being the oldestinstitution on seismological studies, haveobservation networks having seismological andengineering purposes. Since 1926, The Kandilli

Seismological Division, an integral arm ofKOERI, continues to provide seismologicalobservation services with its continuouslyexpanding network distributed throughoutTurkey. The first station was a mechanicalMainka seismograph installed at Kandilli inIstanbul abbreviated as ISK. The KOERI’s

seismic stations started to be a network in theearly 1970’s with the installation ofconventional seismic stations in WesternTurkey. First real-time data recording has beenstarted in Marmara Region, TURKEY by thecommencement of MARNET sub-networkwhich is consisting 9 radio-link stations. TheKOERI has started to expand its networkthroughout Turkey in 1993 by the installationof on-line and digital seismic stations. Today,there are three kind of stations according tothe signal transmission; radio-link, on-line,broad-band seismic stations.

Radio-Link Seismic Stations: MARNET andIZINET are the earthquake monitoringsystem covering the populous Marmararegion and Lake Iznik in NorthwesternTurkey. Seismometers placed at remote sitestransmit signals continuously by radio to thecenter in Istanbul. Nanometrics telemetricsystem is located in the Marmara Sea regionand it has five digital seismic stations. Realtime data are recorded in digital forms at thecenter. In addition, all of them have verysensitive seismic sensors (BB) andNanometrics network can make automaticdetection of earthquakes with magnitudelarger than 3.0.

KANDILLI OBSERVATORY & ERI

Observatories

Seismology Magnetism Meteorology Astronomy-solar physics

Geophysics EarthquakeEngineering

Geodesy

AdministrationEngineeringdepartment

7 April 2003

On-Line Seismic Stations provide continuousanalog data to the seismological center at theKOERI, Istanbul by leased lines. Real timedata are recorded in analog and digital formsat the center.

Broad-Band Seismometer System is widelybeing utilized in the world. Their extremelyhigh dynamic range and stable transfercharacteristics make them ideal for a wide-range of applications. The KOERI started torun a broad-band seismic station at Isparta(ISP) according to the memorandum ofunderstanding between the KOERI, GeoForschungs Zentrum (GFZ) and the InstitutoNazionale di Geofisica (ING) in the mid ofOctober,1996. The KOERI has a dial-upconnection with the ANTO (Ankara-Turkey,Observatory)- IRIS (Incorporated ResearchInstitutions For Seismology) broad-bandseismic station located in the Middle EastTechnical University, Ankara. Also, the centralstation ISK was being replaced by a broad-bandstation. In addition, a new broad-band stationso called VAN and MALT stations located in the

Eastern part of Turkey. BAL and EDR stationsinstalled in the Western part of Turkey.

The number of broad-band stations will beincreased to 30 in near future.

Data Analysis and datadistributionEvents are recorded at the center both inanalog and digital forms obtained fromtelemetred, broad-band and on-line seismicstations. These are processed using HYPO71for the hypocenter determination. Theseismological division of KOERI determines,as rapidly and accurate as possible, locationand size of all earthquakes of magnitudelarger than 3.0 that occur in the country. Itprovides twenty-four hour information serviceto government agencies, to government publicinformation centers and to news media. Thisinformation is also disseminated immediatelyto the relevant international seismologicalcenters by fax and by internet channels.Observatory has been supplying mainly 3 kindof seismological data: phase readings,

waveform and catalogue to the earth scientistsin Turkey and all over the world throughinternet facilities. The data are alsodisseminated to the relevant internationalseismological centers such as NEIC, ISC, andCSEM through fax and internet.

Future planNowadays, several earthquakes in Turkey haveoccurred more frequently. In particular, theepicentre of these earthquakes are located innorthwestern part of Turkey . Izmit and Düzceearthquakes caused the deaths of more than18.000 people, injured 49.000 people and over108.000 buildings either collapsed or heavilydamaged especially at Yalova-Gölcük-Izmit-Adapazarı-Gölyaka-Düzce- areas.

The earthquake was felt in the Marmararegion, Central Anatolia and the Aegean Seain an approximately 480.000km2 area. Tomitigate the seismic risk and prevent seismicpanic, the Kandilli Observatory andEarthquake Research Institute have draft thefuture plan for improving the seismicmonitoring system by establishment thebroadband digital seismic network in Turkey.Such monitoring system should permit rapidacquisition of seismic wave form data,automatic detection of earthquake event andhypocenter determination. The system shouldalso provide high quality data appropriate tostudy on seismology and tectonic as well asearthquake prediction.

Kaynaslı

CSEM /EMSC Newsletter

8

The observations and monitoring of theseismicity in Bulgaria are carried out by theSeismological Department of theGeophysical Institute, Bulgarian Academyof Sciences.The Geophysical Institute was establishedin 1961 and it incorporated the academicresearch (staff and equipment) ofSeismology, Earth Magnetism andGravimetry, Physics of the Ionosphere andAtmosphere. The main part of theSeismological Department is the NationalSeismological Survey, which is responsiblefor the maintenance and regular operationof the National Seismic Network consistingof 14 permanent seismic stations and twolocal networks with 7 stations.The observed earthquakes in Bulgaria aregenerated as a result of the collisionbetween the African and the Europeanplates. The subduction zone is clearlyexpressed along the Hellenic arc-trenchsystem. Not so far – less than 600-800 km,the strong Bulgarian earthquakes showmore or less typical intercontinentalbehavior with a dominant extensionalregime. The complicated recent tectonicenvironment needs special attention inconnection with the seismicity observationson the territory of Bulgaria. Strong seismicevents and local increased seismicity arethe clear expression of the seismogenesis ofBulgaria.

The strongest seismic events inBulgariaBulgaria is interested by moderate to highseismicity: since 347 it experienced at least30 events of magnitude M ≥ 6 (at least 12 ofthem with magnitude M ≥ 7). The strongestevent in the country is the earthquake ofApril 4, 1904 with a magnitude Ms 7.8,which occurred in the SW, near the borderwith Macedonia. Since 1700 there havebeen at least 3 events with magnitudegreater than 7 in each century, with a burstof activity in the period 1858-1928, whenabout 20 events with magnitude greaterthan 6 (six of which with magnitude M ≥ 7)occurred.13th September 1858Sofia region has been destroyed, estimatedmagnitude over 6.5, observed intensity - IXMSK (Fig.1), coseismic normal fault, boiledsands, a mineral spring appeared, largedestruction of the buildings, somecasualties and many injured reported, longlasted aftershock activity - more than 5months.14th October 1892northern part of northeast Bulgaria,maximum intensity 8 according to MSK is

reported for region of Dulovo, magnitudearound 7.31st March, 1901Shabla-Kaliakra region, epicenter in theaquatory of the Black sea (NE coast ofBulgaria), estimated magnitude - 7.1,observed intensity - up to X MSK, aforeshock (M~4) reported several hoursprior to the main shock, many coseismic andpost-seismic events reported (landslides,stonefalls, liquefaction, tsunami effects - upto 3 meters, about 5 years aftershockactivity). A large destruction of houses,deaths and injured reported.4th April 1904Kresna-Kroupnik region in southwestBulgaria, two very strong shocks in the timedomain of about 20 minutes occurred(M=7.2 and M=7.8). Intensities up to XMSK reported. All coseismic and post-seismic events observed and reported -landslides, stonefalls, surface normalfaulting (a river has been barraged and alake observed), liquefaction, springsappearance, etc. Many deaths and injuredreported. Large destruction of the houses.Long lasted aftershock sequence - morethan 7 years. This is the most active part ofBulgaria up to now.16th June 1913Gorna Oriahovitza region in central part ofnorth Bulgaria. Magnitude 7.0, intensity -up to IX degree MSK. Large destruction,many deaths and injured reported.Landslides, sand boils, liquefaction andaftershocks reported. 7th December 1986 amagnitude 5.7 earthquake occurred in thesame region. Large destruction and a fewdeaths and injured have been reported.

14 and 18th April 1928again two shocks (M=6.8 and 7.0) with IXand X degree MSK reported for central partof south Bulgaria. Large liquefaction area,sand boils, landslides, surface normalfaulting reported and geodeticallymeasured (mentioned in the Richter’s“Elementary seismology”). More than 120deaths and several hundreds injured. Morethan three years aftershock activity.It is remarkable that there are no eventswith M>6.0 within the Bulgarian territory(and only 9 with M > 5.0) after 1928 (withthe exception of 2 events with M≥6 whichoccurred in 1931 in the closed vicinity ofneighbouring Macedonia). After thefrequent moderate events until 1940,seismicity is lower during the recent years,when especially high magnitude events aremissing. Only two earthquakes withmagnitude Ms=5.2 and Ms=5.7 (Io=VIII),with long aftershock sequences, occurred in1986 near by the Strazhitza town (GornaOriahovitza region in central northernBulgaria).

The history of observationsThe first network of correspondents forobservations of felt earthquakes wasorganized and it has collected macroseismicdata since 1891. It was initiated by SpasWatzof, the director of the CentralMeteorological Station, Sofia. In 1903, theBulgarian seismology survey became amember of the International SeismologicalAssociation. The annual reports ofearthquake impacts in Bulgaria (publishedin French and Bulgarian) go on more than75 years. After 1965, many studies of the

Figure 1: 1000 years shakeability map of Bulgaria and the last strongest events

Observations and monitoring of the seismicity in BulgariaR. Glavcheva, E. Botev, B. Rangelov

Geophysical Institute,Bulgarian Academy of Sciences

9 April 2003

moderate size earthquakes and the atlasesof isoseismal maps have been published.The contributions of the Bulgarianspecialists to the long-term seismicityinvestigation of the Balkan region havebeen done during the execution of manyrecently completed European projects(Glavcheva and Radu, 1994; Albini andStucchi, 1997, Ranguelov et al, 2000).The instrumental seismology started in1905, when a Bosch-Omori seismographwas installed in Sofia. This action wasmotivated by the occurrence of severalviolent earthquakes, some of them followedby long-lasting aftershock sequences: the1892 earthquake in northeast Bulgaria(maximum intensity VIII MSK); the BlackSea earthquake in 1901 (observed intensityX MSK in northeast Bulgaria); and twotransborder Bulgarian-Macedonianearthquakes (maximum intensity inBulgaria up to X MSK).Two facts concerning the earliest stage ofBulgarian instrumental seismology arenotable:• the first bulletins, which systematized the

registrations during 1905 and 1906,appeared soon after the instrumentinstallation (Watzof, 1907); and

• selected Bulgarian registrations used byA. Mohorovicic and H.F. Reid in theirworld-recognized works published in 1910(s. References).

The Sofia station remained the only wellinstrumented site up to 1961 (in 1905Agamennone seismoscopes were installedtoo at several localities in Bulgaria),although its equipment was modernized atthe end of 1934 by two horizontal Wiechertseismographs (To = 9-11 sec, magnification

ca. 200). The first Bulgarian catalogue(Kirov et al., 1960) covers all seismic eventsaffected Bulgaria by intensity of minimumV MSK occurring both inside and outsidethe country during the time-period 1892-1958. In this catalogue, beside theearthquake occurrence time, epicentral ormaximum intensity together with thecorresponding area, the most likelyepicenter location for foreign events hasbeen presented; there are not anymagnitude determinations therein.Grigorova and Rizhikova (1966) compile thefirst parametric catalogue in Bulgaria. Aninstrumental magnitude, based on Wiechertrecords, appears as energy characteristics ofthe earthquakes. Later on, an enrichednational catalogue, together with someisoseismal maps of the strongest Bulgarianearthquakes, prepared by Grigorova in theframework of a UNESCO project, has takenplace in the first Balkan earthquakecatalogue (Shebalin et al., 1974) and in theisoseismal atlas (Shebalin, 1974). Thiscatalogue covers most of the seismic eventswith a lower magnitude threshold 4occurring since the antiquity till 1970. Withthe aim to make the seismic zonationupdating, a revised catalogue is compiled inthe end of the 70’s (Grigorova et al., 1978)using the systematization already producedand techniques performed during theBalkan catalogue creation. The newcatalogue presents more than 1400 entrieswithin the territory outlined bygeographical coordinates 40 - 46oN, 20 - 30oE.This catalogue has been used as input datafor the seismic zonation maps of Bulgaria(Boncev et al., 1982). The shakeability mapof 1000 years has been accepted as a basic

map (Fig.1) for the seismic rules and code inBulgaria (Bulgarian seismic rules and code,1987). The corresponding design coefficientshave been attached to the intensities on the1000 years shakeability map. At present theseismic code of Bulgaria is in the process ofmodification according to the EUROCODE 8.In the early 60’s one two componentmechanical seismograph “Krumbach” wasinstalled in South Bulgaria (Dimitrovgradstation). In the 1960's and 1970's theseismic network was expanded to cover theseismogenic areas of Bulgaria by the high-sensitive 3-component seismographs SKD,SKM, VEGIK with galvanometricregistration. The instrumental registrationsof that time have been stored on a smokedpaper (for the old instruments), pen inkrecords and photo paper.After the 1977 Vrancea earthquake amodern seismological network wasestablished in Bulgaria starting itsoperation on 1 August 1980.

Recent observational networkIn 1980 the new National OperativeTelemetric System for SeismologicalInformation (NOTSSI) started operating.Initially it consisted of 6 short periodvertical seismographs (S-13 / TeledyneGeotech) which were situated to monitorthe most active seismic zones of Bulgaria.During the next several years the telemeternetwork has been expanded up to 14registration sites. At present all the 21analog stations in Bulgaria ( Fig.2 ) areequipped with the same one-componentvelocitygraphs (vertical S-13). Most of thestations have a visual recorder unit of thetype “Helikorder” where the record of thesignals is performed by ink pens on a paper.NOTSSI is with a near real time signaltransmission, by telephone connectionsmainly, to the main building of theGeophysical Institute, Sofia. Here therecorded signals are collected and processedin the Operative Center (Fig.3). Two localnetworks have started operation in the mid-90’s in connection with the seismicityobservation around a Nuclear Power Plant“Kozloduy” site (3 local network stations innorthwest Bulgaria) and a salt depositexploitation site near the Provadia town (4local network stations around theProvadia(PRV) permanent station innortheast Bulgaria).All the analog records of the local networkaround NPP “Kozloduy” are digitized by aNanometrics acquisition system inGeophysical institute according to acontract with the corresponding authoritiesof NPP. According to this contractGeophysical institute is responsible for theentire seismic monitoring and analysis ofthe seismicity around the NPP site. One 3-component digital station “Quantera” andbroad-band STS1 seismometer are inoperation at the moment in Vitosha

Figure 2: Seismological network of NOTSSI

CSEM /EMSC Newsletter

10

Geophysical Observatory (near the city ofSofia), according to the contract withMEDNET international project. In theOperative Center of Geophysical institutethe seismological data from all 21 stationsare manually interpreted (P- and S- arrivalsidentification, first onsets, signal duration)and processed by computer. For the focalparameters determination an adaptation ofthe widespread software HYPO'71 and the“four layers” regional velocity model areused (Solakov, 1992). In the earthquake filethe duration magnitude Md is computedaccording to the regional formula byChristoskov and Samardzieva (1983). Twopersons on duty (a seismologist and atechnician) are available 24 hours each day.These two specialists belong to a qualifiedteam of seismologists and technicians,which are in charge to ensure:• permanent high-quality recording and

analysis of the seismic signals,• determination of the earthquake

parameters,• relevant assessment of the potential

dangerous impact on the people orbuildings and to inform the responsiblegovernmental bodies if this impactbecomes dangerous. In a case of a strongearthquake on the Balkans, theseismologist on duty sends Bulgarianseismic data to the neighboring andinternational seismological centers. Atthe same time the seismological team isresponsible for the processing, analyzingand classification the daily seismologicaldata.

The regular international data exchange isbased on the later compiled weeklyseismological bulletin. The information andanalysis of the bulletins of the neighbor andinternational seismological centers areaccepted as a natural and necessarycondition for the preparation of the finalearthquake catalogue in NOTSSI.The location accuracy in the seismic catalogdepends on the instrumental sensitivity atthe different location sites and on thespatial position of the seismic source within

the frame of the recording network. Thehigh sensitivity of the seismographs inSouthwest Bulgaria allows records anddata processing of a great number ofearthquakes with a minimum magnitude Mless than 1.0. The different magnitudelevels of the well-solved local, regional andlong distance earthquakes are establishedas follows: M=2.0 for the territory ofBulgaria, M=3.0 for the Balkans, M=5.0 forthe long distance events. About 12000events recorded during the operation of theNational network are located on theterritory of Bulgaria and its close vicinity(Fig.4). Most of them are micro earthquakes– more than 95% with a magnitude M<3.0.The strongest recorded by the NOTSSIevent is the 1986 magnitude M5.7Strazhitza earthquake in the central NorthBulgaria.The earthquake monitoring in Bulgariaensures the necessary base for permanentinvestigations of space, time and energydistribution of seismicity, the deep Earth’sstructure and the stress field in the earthcrust and leads to better understanding ofthe earthquake genesis nature. Detailedinformation and analysis of the realized

seismic energy (Botev et al., 1991,1992...2001, 2002) are also periodicallyproposed as a generalization andsupplementation of the monthlypublications of the preliminaryseismological bulletin of NOTSSI. Thesestudies help the sought for correlationbetween seismicity and some geophysicalparameters aiming to find out eventualprecursor anomalies.The strong motion network also exists(equipped with SMA1, and other analogueor digital instruments) to support thepractical application in the construction ofbuildings and facilities. The CentralLaboratory of Seismic Mechanics andEarthquake Engineering manages this.The earthquake monitoring promotesmany scientific and practical challenges ofcontemporary Bulgarian seismology suchas:

• monitoring of natural, artificial andinduced seismic events in Bulgaria andits surroundings;

• earthquake statistics (general, aftershocks,etc);

• seismic zonation of the Central Balkans;• long-term seismicity investigations;• earthquake source mechanisms,

kinematic and dynamic specification ofthe large sources,

• seismic energy attenuation;• tsunami modeling (Black Sea);• seismic hazard assessment;• seismic waves propagation (kinematic

and dynamic aspects);• seismic monitoring and site-effects

studies in areas of state important sites(nuclear power plants, dams, lifelines,etc);

• delineating earth structures and dynamicprocesses in tectonically active regions byseismological data;

• searching for earthquake's precursors ofseismic, electro-magnetic, or of othergeophysical nature.

Figure 4: Seismicity of Bulgaria (after 1980, M>=2.0)

Figure 3: Operative center in Geophysical Institute

11 April 2003

Near future activities.To improve the quality of the monitoring ofthe seismicity the establishment of a digitalnetwork (with minimum three broad band 3-component seismographs) is planned to startoperating in the very near future. One morenetwork near the future second NPP site“Belene” (under construction) is intended tobe deployed around the site. Theestablishment of a mobile group for the fieldinvestigations after the strong event is alsoan important task for the post earthquakeactivity research and data collection. Thedetailed study of the well-expressedearthquake sources of the strongearthquakes observed on the territory ofBulgaria is an essential task for the betterunderstanding of the seismic process usingthe monitoring information.It is important to recommend, in theframework of the different Europeanprograms and projects, a wide co-operationamong the scientists of Central andSoutheast Europe regions for theseismological, building’s structural andprevention research. The area is under thestrong influence of earthquakes and othergeodynamic disasters. The improvement ofthe instrumental networks connected by themodern communication networks is thenecessary condition for the successfuldevelopment of the region and the seismicprotection measures. We consider that theaims of seismology, earthquake engineeringand the people’s protection and preventionare a governmental task and each country aswell as the EU community must give theirsupport by funding from different sources.

References:Albini, P. and M. Stucchi (1997): A BasicEuropean Earthquake and a Catalogue for theevaluation of long-term seismicity and seismichazard (BEECD), in Seismic risk in the EuropeanUnion, edited by A. Ghazi A. and M. Yeroyanni,(Brussels-Luxembourg) vol. I, pp. 53-77.

Boncev E., V.I.Bune, L.Christoskov et al.(1982).A method for compilation of seismic zoningprognostic maps for the territory of Bulgaria.Geologica Balcanica, v.12 N2, 3-48.

Botev E., B.Babachkova, B.Dimitrov,S.Velichkova, I.Tzoncheva, K.Donkova,S.Dimitrova .(1991,1991,1992,1992, . . .2002,2002). Preliminary data on the seismicevents recorded by NOTSSI in January - June1991 (July – December 1991), .. ., (July –December 2001). Bulg. Geophys. J., 18,18,. . .28,28 (1,2,3,4).

Bulgarian seismic rules and code (1987). CTSU– BAS, Sofia, 67 pp. (in Bulgarian).

Christoskov L. and E.Samardjieva, (1983).Investigation on the duration of the seismicsignals like an energetic characteristic of theearthquakes. BGJ, vol.IX, N1, , Bulg. Geophys. J.9 (1), 28-37. (in Bulgarian)

Glavcheva, R. and C. Radu (1994): Theearthquake of October 14th, 1892 in CentralBalkans: a transfrontier case. In: Albini P. andMoroni A. (eds.), Materials of the CEC project"Review of Historical Seismicity in Europe", 2,CNR - Istituto di Ricerca sul Rischio Sismico,Milano, 215-223.

Grigorova, E., S.Rizhikova (1966): Tremblementsde terre en Bulgarie au cours de 1961 a 1964. NN62-65, Editions de l’Acad. Bulg. Sci., Sofia, 105.

Kirov, K. (1931, 1945): The earthquakes inBulgaria. Report on the felt earthquakes during1917-1927, 1928-1930 (Sofia), NN18-28, 29-31(in Bulg.).

Mohorovicic, A. (1910): Das Beben vom 8.X.1908.Jahrbuch Meteor. Obs. in Zagreb (Agram) furdas Jahr 1909, Jahrg. IX, IV Teil, Zagreb, 63 S.

Ranguelov B. et al., (2000): The earthquake (M7.8) source zone (SW Bulgaria)., Acad. Publ.House “M.Drinov”, Sofia, 279 pp.

Reid, H.F. (1910): The California earthquake ofApril 18, 1906. Report of S.E.I. Comm., V. II -The mechanics of the Earthquake. CarnegieInst., Washington D.C., 195 p.

Shebalin, N.V., V. Karnik, D. Hadzievski (Eds.)(1974): Catalogue of Balkan Earthquakes,UNDP/UNESCO Surv. Seism. Balk. Reg.,Skopje.

Sokerova, D., S.Rizhikova, R.Glavcheva (1982):Catalogue of Earthquakes in Bulgaria duringthe period 1970-1980. In: National Report ofBulgaria, UNDP/UNESCO and UNDRO ProjectRER/79/014 "Earthquake Risk Reduction in theBalkan Region, W.Gr. A: Seismology,Seismotectonics, Seismic Hazard andEarthquake Prediction, Final Report, Athens,Dec. 1982, A47 - A49.

Solakov, D. (1993): An algorithm for hypocenterdetermination of near earthquakes. Bulg.Geophys. J. 19 (1), 56-69.

Watzof, S. (1902, 1903, ..., 1923): Theearthquakes in Bulgaria. Report on theearthquakes felt in XIXc., and during 1901,1902,..., 1913-1916. (Centr. Meteorol. Inst.,Sofia), NN 1-17 (in Bulgarian, in French).

Watzof S. (1907) Bulletin seismographique del’Institute meteorologique central de Bulgarie de1905, 1906, NN1, 2 (Sofia) 1907.

NEWS OF THE EMSC

Euro.-Med. bulletinThe main objective of the EPSI project was the production of the Euro.-Med.seismological bulletin. This EU-funded project lead by EMSC ended last December andfirst results of the Euro.-Med. bulletin are already available on our web site. OlivierPiedfroid, who has been involved for more than 2 years in this project, has recently leftEMSC. Stéphanie Godey who defended her PhD thesis last year at Utrecht University,has just taken over his position.Stéphanie is now in charge of the production of the bulletin from 01/01/1998 up to now.We expect to have a complete 6-year bulletin by the end of this year and then startdetailed performance analysis of the results.The Euro.-Med. bulletin is based on data contributions from network operators. EMSChas set up procedures to help data operator in the registration of stations to the WorldData Centre. Nearly 200 stations have been registered following these procedures. Sofar, about 70 networks and 1500 stations have contributed to the bulletin. A descriptionof the database content is available by month and by network at www.emsc-csem.org/Html/DATA_main.html If your network does not appear on this page or if youcould help filling some reported gaps, please contact Gilles Mazet (mazet@emsc-csem.org) as soon as possible, otherwise we may not be able to include your data in theEuro.-Med. bulletin!

EMSC-ORFEUS Infrastructure proposalEMSC and ORFEUS are co-ordinating the preparation of an Infrastructure proposal forthe EU Sixth Framework Program. This five-year project is based on our commonExpression of Interest named NERIES and it aims at improving the integration of theseismological community and at improving the services provided by our organisations.

Details of this proposal to be submitted on April 15, are available on both EMSC andORFEUS web site. An information meeting will be held during the EGS meeting on April8, from 19:30 to 21:00 in lecture room R7

New prototype service for active membersFollowing requests from active members, EMSC has developed a tool to automaticallydisseminate by email the result of automatic relocations. Every thirty minutes, all availabledata concerning very recent seismic events are merged, relocations (names MIX) arecomputed for new events and they are updated for events for which new data has beenmade available. The results are automatically displayed on the Real Time Seismicity Page(for more details on this web page, see Newsletter n°17).The prototype service automatically sends a single email by seismic event. In order to avoidthe dissemination of poorly constrained location, the message is sent for events whichoccurred at least 3 hours before, a delay which is large enough to ensure that the majorityof data has already reached EMSC. As an average, these relocations have proved to be morereliable that individual automatic locations.Today, EMSC is looking for active members interested in receiving these messages, helpingus to finalise this tool and make it operational by providing feedback on the reliability ofthe provided locations. If you are interested, please let us know!

Next Newsletter issue The next issue is planned for September 2003. We would like to have a special part onearthquake monitoring in Northern Africa. A number of authors have already beenidentified and new propositions are welcome and should be addressed to Rémy Bossu(bossu@emsc-csem.org).

CSEM /EMSC Newsletter

12

IRSNIRSN is a governmental organization recentlycreated by a law (28th of February 2002).IRSN is the association of two entities, theNuclear Protection and Safety Institute andthe Protection against the Ionizing RadiationsOffice. IPSN was originally part of the AtomicEnergy Commission, which is a nuclearoperator. In order to clearly distinguishbetween a safety organization and a nuclearoperator the French Government decided toseparate IPSN from CEA. Approximately1,500 experts and researchers work at IRSN,on radiation protection and nuclear safety. Themain issue for IRSN is to provide advices forthe French Nuclear Safety Authority (calledDGSNR). IRSN is its main technical support.

Scientific research is carried out in the areasassociated with the use of nuclear energy,ionizing radiation and natural radioactivity.Actual tasks consist in risk assessment,expertise and consulting for public authorities,public opinion information, environmentalmonitoring and radiation-dose follow-up foremployees and training of medical personnel.

THE BERSSINThe Seismic Hazard Assessment for theNuclear Power Plants Safety Laboratory(hereafter BERSSIN), whose activities andmissions are presented in this paper is one ofthe four laboratories that composed the WasteElimination and Geosphere service. The mainassignment of the BERSSIN consists in theexpertise of the safety demonstration of thenuclear operators from the seismic point ofview, for new and old power plants. Accordingto this task, the BERSSIN conducts researchstudies to improve seismic hazard assessmentin France. Finally the division has to maintainthe regulation for the nuclear power plantsafety in coherence with the «state of the art».The team is composed of three geologists, aseismotectonical specialist, five seismologists,a technical engineer and a technical assistant.At this time, three PhD and three postdoctoralstudents work in the team.

SEISMIC HAZARDASSESSMENTIn all the countries, seismic hazardassessment requires three main steps: thefirst one consists in the seismic sourcesdefinition, i.e. the cartography of active faults.The second one is to evaluate the seismicpotential of these sources. The last point is theseismic motion prediction, in term of response

spectrum or acceleration time series or anyother pertinent indicator, at the site ofinterest. This motion is the input data for thestructural engineers for the power plantdesign or for the resistance tests.

The definition of active faults in France is stilla controversial scientific debate. Differentmethodologies allow improving our knowledgeon active faults. Classical geologicalapproaches such as field and geomorphicstudies are completed with Digital ElevationModel (hereafter D.E.M.) and aerialphotography analyses. Several main faults arealready known in France. The study ofneotectonical indices, indicators of seismicruptures in recent (i.e. quaternary)sedimentary formations is also a way to studythe active faults. Whatever the location and thegeometry of a fault, a question still remains:can the fault produce an earthquake ?

Geologists of BERSSIN published this year asynthesis of the quaternary deformationindices for the French territory (Baize et al.,2003). The BERSSIN collaborated withacademic teams and others specialists, andcontributed for the publication of the 1993French sismotectonic map (Grellet et al.,1993). This publication presented the state ofthe art in the 1993, using a very broad sourceof data (from geology to geophysicalinterpretations, seismic profiles, magneticanomalies…).

Historical Seismicity

In addition to these geological investigationsthe seismic history of the country providesimportant information on seismic sources. Theinstrumental seismicity, which covers a 40 yearsperiod considering the installation of thenational network in the 1960’s.This short periodhas been complemented with a project started25 years ago, whose aim is the study of thehistorical seismicity. The BERSSIN with theElectrical French Operator (EDF) and theGeological and Mines Research Division(BRGM) constructed the French historicaldatabase (http://www.sisfrance.net). Thisdatabase is today a reference with all theearthquakes and the associated archivesdescribing the effects of the event. UsingSisFrance, one can find the location of any eventand a map of observation points. An evaluationof all the intensity points is given, with theepicentral intensity (MSK scale). Because theconfidence or the descriptions themselves arevery scattered, quality factors for the locationand the intensity values are given. SisFrancedatabase has more than 80 000 observations

and 9000 archives describing 6000 earthquakes.The time period covered by SisFrance is aroundten centuries. The interpretation of historicaldocument, such as damage summaries afterearthquakes, or description of the event inancient newspapers or church archives is notcommon for seismologists.That is the reason whycollaboration with historians is very important.They are able to interpret these old documents,taken into account the historical social andeconomical context (Quenet et al., 2002).

A sensible point in seismic hazard assessmentis the estimation of magnitude frommacroseismic observations for historicalevents. Levret et al. (1994), using a set of 73earthquakes simultaneously documented bymacroseismic intensities and recorded by theFrench National Seismic Network (LDG) haveproposed an attenuation relation. Levret et al.,computing 238 isoseismals proposed a relationbetween the instrumental magnitude, thefocal distance and the epicentral intensity ofthe event. Using this relation, any historicalevent with sufficient intensity observationscan be associated with a magnitude value. TheBERSSIN still continues to improve thisrelation (Scotti et al., 1998).

Archeosismicity and Paleoseismology

The last sources of data that contribute toimprove our knowledge on the seismicity of thecountry, for a period longer than the historicalperiod described before, are the archeo and thepaleoseismological studies. Archeoseismicityconsists in the study of deformation on oldRoman or Middle Age structures that could berelated to a seismic event. In France, severalstudies have been carried out in this field, andthe BERSSIN conducted several of theseprojects. At present, the BERSSIN is compilinga ten years long project in archeoseismicityperformed on the Pont du Gard (SouthernFrance). Architects, archeologists, geologists andseismologists discovered structuralreinforcements on this bridge that could beassociated to past earthquake damages.Numerical simulations confirmed the possibilityof seismic effects (Volant et al., 2003). But allthese observations and computations do notallow to describe precisely the position and themagnitude of the seismic event.

As in other countries, several paleoseismicindicators have been discovered in France sincethe 1990’s. This recent research field allows todefine another tool to characterize the seismicactivity of faults, in addition with theinstrumental and historical activities. InFrance, several reliable indicators have been

Seismic Hazard Assessment for Nuclear Power Plants Safety Laboratory(BERSSIN), Radioprotection and Safety Nuclear Institute

(IRSN, France).Catherine Berge-Thierry, Stéphane Baize, David Baumont, Fabian Bonilla, Marc-Edward Cushing, Pierre Dervin, Francis Lemeille,

Stéphane Nechtschein, Gérard Peyridieu, Oona Scotti and Philippe Volant.

13 April 2003

published (Lemeille et al., 1999, Figure 1). Thedifficulty of such kind of data remains thedating of the event (with a large timeuncertainty) and the estimation of the size of theevent, in terms of magnitude. The BERSSINwas involved in the European Project PaleoSis(ENV4-CT97-0578 EC).

The BERSSIN uses all these approaches toassess the location of active faults and to definetheir seismic potential. But all these data do notallow answering precisely key questions such aswhat is the maximal size of an earthquake on asegmented fault. This kind of questions stillremains unanswered today. Numericalsimulations based on dynamic rupturepropagation could help to progress on thisquestion.

Ground Motion Prediction and Site Effects

In addition to the geological sourcecharacterization and the assessment ofseismic potential, the BERSSIN works on

ground motion prediction, to propose to civilengineers, the ground motion that can beconsidered as an input for the structuredesign. Empirical approaches withattenuation relations are currently used in theFrench Basic Safety Rule (methodology toassess the seismic hazard for a FrenchNuclear Power Plant). This rule has beenrecently modified, introducing new topics,such as paleoseismicity and site effects. Theattenuation law developed for the rule (Berge-Thierry et al., 2003), with mainly Europeanstrong motion records (Ambraseys et al.,2000), allows the calculation of the groundmotion for rock (Vs>800m/s) or sedimentary(300m/s<Vs<800m/s) sites. Specific site effects,due to very soft soil, or 2D-3D geometricaleffects are not taken into account in theempirical law. For such cases, a specific studyis required to evaluate the linear andnonlinear soil response. Such complex effectsare sometimes cumulative, such as in theGrenoble basin (France), which is a deep valleyfilled with thick soft sediments (Figure 2).

The BERSSIN collaborated during more thanten years with the University of SantaBarbara (California) on the Garner valleyexperiment, to study the 1D linear and non-linear effects, using instrumented deepboreholes. Now, the BERSSIN continues onsite effects studies in the active seismic zone ofthe Corinth Gulf. In the framework ofCORSEIS European Program, IRSNcollaborates with ENS-Paris, IPG-Paris,AUTH (Greece) and NKUA (Greece) to installan array of accelerometers and pore pressureprobes at different depths dedicated to thestudy of liquefaction in the Aigion harbor. Thisexperiment would allow to collect groundmotion data to improve seismic motionmodeling. In addition, BERSSIN is alsoinvolved in the study of site effects, andparticularly, nonlinear soil behavior. Suchphenomena occurred during the Aigionearthquake on June 15, 1995 (M 6.2).Geotechnical measurements and geophysicalexperiments improved the knowledge of thesite providing basic physical parameters. Themain objectives for the BERSSIN in thisexperiment are (1) to increase the strongmotion database with a good knowledge of siteconditions (including static and dynamic soilparameters), and (2) to improve the strongmotion modeling considering linear andnonlinear soil response (Bonilla et al., 2002).We expect that the results could be applied onweak seismicity and low deformation rateareas such as France.

A good strong motion assessment usingempirical relationships requires the collectionof numerous and reliable data. The BERSSINcollaborates with academic teams sinceseveral years to collect and disseminate strongmotion data. For example BERSSINparticipated to the development and to theproduction of the European Strong MotionDatabase (Ambraseys et al., 2000): this

Figure 1: Paleoseismological study in east of France (Lemeille et al., 1999)

Figure 2: Two seismological records in the Grenoble area, one on the rock and one on the sediments (left).3D seismic simulation of the horizontal peak velocity (right).

database can be obtained by simple request atBERSSIN (request addressed toCatherine.berge@irsn.fr). In France, thestrong motion data are centralized by theFrench Accelerometric Network (RAP:http://www-rap.obs.ujf-grenoble.fr/) .TheBERSSIN is associated to the RAP. Theimprovement of strong motion assessment isstrictly related to the increase of reliable data,with accurate site conditions.

Empirical relationships use simple physicalmodels to describe the seismic energyattenuation with few parameters, such as thesource to site distance, the event magnitudeand sometimes the geological site condition.Source complexity, such as the geometricalextension of the fault, which has a strongeffect close to the fault (directivity) is nottaken into account in this empiricaldescription, neither hanging wall effects, orfocal mechanism. These parameters can onlybe estimated in areas with high seismic levelactivity. Nevertheless, synthetic simulationallows to take into account realisticcomplexities of the source. The BERSSINdevelops numerical approaches, especiallyusing a kinematic description of the source.Directivity is then modeled, and peak groundacceleration maps and broadbandaccelerograms can be calculated (Baumontand Berge-Thierry, 2002). This code has beenused in the European Project, dedicated to thePredictability of the Aftershocks after a MainEvent (PRESAP EVG1-CT-1999-0001). In thisproject, the BERSSIN was leader of WP2related to the Coulomb stress change method,as a tool of predictability for the aftershocklocation (Baumont et al., 2002).

The Durance MultidisciplinaryResearch Program

Considering all these research fields, seismichazard assessment in moderate seismiccountries appears to be a real challenge. Inorder to improve the knowledge of the seismic

behavior of an active fault in a lowdeformation rate area, IPSN decided in 1990’sto conduct a multidisciplinary study on theMoyenne Durance Fault, located in SouthEastern France. Historical earthquakescharacterize this fault system (4 events withmagnitude 5 and 5.5 since 1509). This is theonly fault in France with such a periodichistorical seismic activity. A completegeological study of the region has been done,combining field investigations, with aerialphotography interpretations, in collaborationwith academic teams (Paris XI and Cerège).Seismic profiles released by private or publiccompanies were re-interpreted. The resulting3D model constrained by borehole data showsthe complex 3D geometry of the fault(Cushing et al., in prep.). In 1992 IRSNdecided to install a permanent seismologicalnetwork surrounding the fault area (Figure3). It is the first time in France that apermanent seismic network is completelydevoted to one specific fault zone. Althoughmajor historical earthquakes are clearlyassociated with this structure, fewearthquakes have been recorded since 1962with the national seismic network. Ournetwork shows a small seismic activity, withepicenters well aligned along the faultdirection (Volant et al., 2000). Focalmechanisms computed for two events agreewith the regional microstructural studies(Cushing et al, 1997). The Moyenne DuranceFault is characterized by a complex 3Dgeological structure. The fault of the MoyenneDurance is segmented. Standardized locationprocedures with 1D velocity models usinglinear algorithms are not adapted for suchareas and do not constrain the event locationespecially in depth (Lomax et al., 1998). A 3Dnon-linear location program has beendevelopped (Lomax et al.,2000). To completethe seismic potential assessment of this fault,IRSN installed three years ago two permanentGPS stations on each side of the fault, whichwill allow to constrain the deformation rate in

the fault area within the next five or ten years.These sensors will be completed by semi-permanent network in order to identify strainalong a cross section through the fault.

Deterministic and ProbabilisticApproaches

Seismic hazard assessment for nuclear plantsis guided by a specific regulation, which isbased on a deterministic approach (RFS2001-01). Nevertheless, in the context ofProbabilistic Seismic Assessment (PSA)studies, IRSN has always developedprobabilistic seismic hazard codes (Bottardand Gariel, 1995). Since the beginning of2001, the BERSSIN is involved in therevision of the French seismic zonation, toproduce a map describing the differentregions with their associated seismic levels.The Environment Ministry conducts thisrevision, and requested IRSN for technicalsupporting. This seismic zonation has to beevaluated following a probabilistic approach,according to the Eurocode 8, and should bethe bases for the regulation on conventionalstructures. A private company produced theseismic hazard study, and the BERSSIN didthe expertise of their work. The objective wasto give to the French Government and theexpert group, recommendations to evaluatethe results. During this experience theBERSSIN developed a probabilistic seismichazard code using a logic tree approach.Furthermore a Ph.D Thesis (Beauval et al.,2002) is dedicated to the probabilistic seismichazard feasibility in moderate seismicitycountries. In the same time, IRSN itselfproposed to realize a whole ProbabilisticSafety Assessment for Nuclear Power Plants.This type of approach is currently beingdeveloped for the Tricastin NPP site (south ofFrance). Comparison of deterministic andprobabilistic approaches enables theBERSSIN to define and computeuncertainties and margins that should beincluded or explicited in all assessments.

CSEM /EMSC Newsletter

14

Figure 3: DEM of the Durance area with the seismological network (left). Proposed 3D velocity model (stations are the red diamonds), right.

Underground Waste Storage

IRSN is especially involved on seismic hazardassessment related to surface nuclear powerplants. Nevertheless, the possibility to storenuclear wastes in deep geological formationsappeared in the 1990’s. This possibilitydeveloped new research fields. The FrenchGovernment decided to conduct research inFrance in order to define the conditions wherea reversible or irreversible disposal site couldbe achieved and operated in deep geologicalformations. The underground researchlaboratories give the opportunity to answerimportant questions on this topic. Theprogram is conducted by a governmentalagency –ANDRA- (agency responsible for thenuclear wastes management). Since thebeginning, IRSN expertises all the researchresults of ANDRA, for the safety authority. TheBERSSIN is often requested to give answer onspecific seismological and geological questions.The BERSSIN conducts several researchprojects devoted to this deep storage concept.For example, through postdoctoral researchthe BERSSIN works on the growth of faultnetworks, using numerical approaches, or onthe use of the calcite macles to construct thedeformation history of a region (Rocher et al.,2003). Geophysical methodologies are tested inthe IRSN underground natural laboratory, (anancient one-century years old tunnel),especially to characterize the fracturation ofthe argillite due to a rock excavation(Excavation Damaged Zone). The old tunnel isthe reference and new galleries have beenexcavated. A network of 20 high frequencyaccelerometers has been cemented inboreholes before the excavation of a newgallery. The study of the microseismicity willallow to characterize and to image thedamaged zone evolution. The main advantageof this method is that it is a non-destructivemethod which has been used in several placesall over the world. But it is the first time thatthis method is used in argillite. The BERSSINis also involved in the seismic motionprediction in depth, the BERSSIN proposedtwo Ph.D thesis subjects on this topic: usingthe KikNet Japanese accelerometric network,which offer thousands of stations (a surfacereference and a station in a borehole) wheregeological conditions are known (Lussou et al.,2001), we hope to improve the linear and non-linear strong motion prediction from depth tosurface

The wide range of fields covered by theIRSN/BERSSIN contributes to improve theseismic hazard assessment for both nuclearand conventional structure safety.

Références:Ambraseys, N., P. Smit, R. Beradi, D. Rinaldis, F.Cotton et C. Berge-Thierry, Dissemination ofEuropean Strong Motion Data. CDROM-collection, European Commission, Environmentand Climate Research Program, 2000.

Baize S., E. M. Cushing, F. Lemeille, T. Granier,B. Grellet, D. Carbon, P. Combes, C. Hibsch,Sismotectonique de la France métropolitaine -volume 3 : Inventaire des indices de ruptureaffectant le Quaternaire en relation avec lesgrandes structures connues, Mémoire Soc. Géol.Fr, in press, 2003.

Baumont D. et C. Berge-Thierry, ContributionIRSN to the final report for WP4, PRESAPproject, 2002.

Baumont D., F. Courboulex, O. Scotti, N. Melis,G. Stavrakakis, Slip distribution of the Mw 5.9,1999 Athens earthquake inverted from regionalseismological data., Geophys. Res. Lett, 29, No15, 2002.

C. Beauval, D. Bertil, F. Bonilla and O. Scotti,Logic tree approach for PSHA for nuclear powerplants in France, ESC, Gènes, 2002.

Berge-Thierry C., D. Griot-Pommera, F. Cottonand F. Fukushima, New empirical responsespectral attenuation laws for moderateEuropean earthquakes, Jour. of EarthquakeEng., in press, 2003.

Bonilla L. F. Computation of Linear and Non-linear Site Response for Near Field GroundMotion, Ph.D. dissertation, University ofCalifornia, Santa Barbara, 2000.

Bottard S., Gariel J.C, Une méthodologieprobabiliste pour l'évaluation de l'aléa sismiqueen France. Seventh International Conference onApplications of Statistics and Probability inCivil Engineering, Paris, 1995.

Cushing M., Ph. Volant, O. Bellier, M. Sebrier, E.Baroux , B. Grellet, Ph. Combes and Th. Rosique.1997. A multidisciplinary experiment tocharacterize an active fault system in moderateseismic activity area : the example of the Durancefault (south eastern France). EuropeanGeophysical Society ; Annales Geophysicae,supplement I to vol. 15, p. C233. Vienne, April 1997.

Cushing M. , F. Lemeille. B. Grellet and D.Carbon IPSN/BERSSIN, 1998 :Paloseismological studies in the upper rhinegraben area. XXVI General assembly of theEuropean Seismological commission. 23-28August 1998, Tel Aviv, Israel.

Grellet B., Ph. Combes, Th. Granier and H.Philip, Sismotectonique de la Francemétropolitaine dans son cadre géologique etgéophysique, Mémoires de la société géologiquede France, Vol. 1, N 164, Vol 2, 1993.

Quenet G. , A. Levret, O. Scotti and D. Baumont,The Study of Poorly Documented HistoricalEarthquakes in France through amultidisciplinary approach, AmercicanGeophysical Union Meeting, 2002.

Lemeille F., M.Cushing , F. Cotton, B. Grellet, F.Ménillet, J.C. Audru, F. Renardy and C. Fléhoc.Evidences for Middle to Late Pleistocenefaulting within the Northern Upper RhineGraben (Alsace Plain, France). Traces d’activitépléistocène de failles dans le nord du Fossé duRhin supérieur (Plaine d’Alsace, France.Comptes Rendus de l’Académie des Sciences,t.328, série II, n°12, Juin 1999, p. 839-846, 1999.

Levret A., J.C. Backe and M. Cushing, Atlas ofmacroseismic maps for French earthquakes withtheir principal characteristics. Natural Hazard,p. 19-46, 1994.

Lomax, A., P. Volant, C. Berge and J. Virieux,Probabilistic, grid-search earthquake location inthree-dimensional media: Application to theDurance Network in the South of France,Shalheveth Freier 1st Intl. Workshop on AdvancedMethods in Seismic Analysis, Israel, 1998.

Lomax, A., J. Virieux, Ph. Volant and C. Berge-Thierry, probabilistic earthquake location in 3Dand layered models – Introduction of aMetropolis-Gibbs method and comparison withlinear locations, in Advances in Seismic EventLocation, p. 101-134, eds. C.H. Thurber and N.Rabinowitz, 2000.

Lussou P., P.Y. Bard, F. Cotton, and Y.Fukushima, Seismic design regulation codes :contribution of K-NET data to site effectevaluation, Journal of Earthquake Engineering,Vol. 5 (1), p. 13-33, 2001.

Rocher M., S. Baize, S. Jaillet, M. Cushing, Y.Lozac'h. and F. Lemeille. Quaternary stressesrevealed by calcite twinning inversion: Insightsfrom observations in the Savonnières undergroundquarry (eastern France), submitted, 2003.

Scotti O, Levret A.,.and Hernandez B.. Verificationof macroseismic methods on five M > 5instrumental earthquakes in France. Physics andChemistry of the Earth (A), 24, 6, p. 495-499, 1998.

Volant, P., C. Berge, P. Dervin, M. Cushing, G.Mohammadioun and F. Mathieu. Thesoutheastern Durance fault permanent network:preliminary results, Journal of Seismology, 4(2),p. 175-189, 2000.

Volant P., A. Levret, D.Carbon, D. Combescure, T.Verdel, Archeosismicity : a multidisciplinarycase study on the Roman aqueduct of Nîmes(France), in preparation, 2003.

15 April 2003

EMSC informationEMSC/CSEM - c/o LDG, Bât. Sables - BP 12 91680 Bruyères-le-Châtel, France

Chris BrowittPresident president@emsc-csem.orgRémy Bossu +33-1-69267814Secretary General bossu@emsc-csem.orgGilles Mazet-Roux +33-1-69267813Alert system / Data Exchange mazet@emsc-csem.orgStéphanie Godey +33-1-69267813Euro.-Med. Bulletin godey@emsc-csem.org

Véronique GuerdetAdministration +33-1-69267808Fax +33-1-69267000Alert messages alert@emsc-csem.org

urg@ing.esBulletins csem@emsc-csem.orgJSOP data jsop@emsc-csem.orgOther matters csem@emsc-csem.org

16

CSEM /EMSC Newsletter

Syn

op6

Nan

cy -

ww

w.s

ynop

6.fr

501949 1999

CONSEILDE L’EUROPE

COUNCILOF EUROPE

EMSC membersInstitute Country CorrespondantActive MembersNational Seismological Centre (NSC) Armenia Dr. A. Arakelyan Central Institute for Meteorology and Geodynamics (ZAMG) Austria Dr. Edmund FiegweilCentre of Geophysical Monitoring of NAS of Belarus (CGM) Belarus Dr. A. AronovObservatoire Royal de Belgique (ORB) Belgium Dr. Roland VerbeirenBulgarian National Operating Telemetric System for Seismological Information (NOTSSI) Bulgaria Dr. B. RanguelovGeological Survey Department (GSD) Cyprus Dr. George PetridesGeophysical Institute of the Academy of Sciences (GFU) Czech Republic Dr. Jan ZednikInstitute of Physics of the Earth, Brno (IPE) Czech Republic Dr. Jan SvancaraNational Survey and Cadastre, Copenhagen (KMS) Denmark Dr. Soren GregersenNational Research Inst. for Astr. and Geophysics (NRIAG) Egypt Prof. Ali TealebInstitute of Seismology (ISUH) Finland Dr. Pekka HeikkinenSeismic Risk Evaluation for the Safety of Nuclear Facilities (BERSSIN) France Dr. Catherine Berge-ThierryBureau Central de Sismologie Français (BSCF) France Dr. Michel CaraBureau de Recherches Géologiques et Minières (BRGM) France Dr. Pascal DominiqueLaboratoire Central des Ponts et Chaussées (LCPC) France Dr. Pierre-Yves BardBGR Seismologisches Zentralobs. Gräfenberg (BGR) Germany Dr. Klaus KlingeNational Observatory of Athens (NOA) Greece Dr. George StavrakakisUniversity of Thessaloniki (AUTH) Greece Dr. Manolis ScordilisInstitute of Engineering, Seismol., and Earthq. Engineering (ITSAK) Greece Dr. Christos Papaioannou Icelandic Meteorological Office (IMO) Iceland Dr. Ragnar StefanssonDublin Institute for Advanced Studies (DIAS) Ireland Dr. Peter ReadmanGeophysical Institute of Israel (GII) Israel Dr. Yefim GittermanOsservatorio Geofisico Sperimentale (OGS) Italy Dr. Marino RussiStoria Geofisica Ambiente srl (SGA) Italy Dr. Emanuela Guidoboni Geophysics Centre at Bhannes (SGB) Lebanon Dr. Alexandre SursockCentre National de la Recherche (CNR) Morocco Prof. Aomar Iben BrahimNorwegian Seismic Array (NORSAR) Norway Dr. Jan FyenUniversity of Bergen (BER) Norway Dr. Jens Havskov Instituto de Meteorologia (IMP) Portugal Dr. Maria-Luisa SenosInstituto Superior Tecnico (IST) Portugal Dr. Joao FonsecaNational Institute for Earth Physics (NIEP) Romania Dr. GheorgheKing Abdulaziz City for Sciences and Technology (KACST) Saudi Arabia Dr. Tariq Al-KhalifahAgencija Republike Slovenije za Okolje (ARSO) Slovenia Dr. Ina CecicUniversidad Politecnica de Madrid (UPM) Spain Dr. Sonia Alvarez RubioInstitut Cartografic de Catalunya (ICC) Spain Dr. Antoni RocaSchweizerischer Erdbebendienst (SED) Switzerland Dr. Manfred BaerRoyal Netherlands Meteorological Institute (KNMI) The Netherlands Mr. Reynoud SleemanKandilli Observatory and Earthquake Research Institute (KOERI) Turkey Prof. A. Mete IsikaraEarthquake Research Institute (ERD) Turkey Dr. Ferhat TasçiBritish Geological Survey (BGS) United Kingdom Dr. Chris Browitt

Key Nodal MembersLaboratoire de Détection et de Géophysique (LDG) France Dr. Bruno FeignierGeoForschungsZentrum (GFZ) Germany Dr. W. HankaIstituto Nazionale di Geofisica (INGV, Roma) Italy Dr. M. OliveiriIstituto Nazionale di Geofisica (INGV, Milano) Italy Dr. Massimiliano Stucchi Center of Geophysical Computer Data Studies (CGDS) Russia Dr. Alexei Gvishiani Instituto Geografico Nacional (IGN) Spain Dr. E. Carreno

Corporate Members Mediterranean Re Ireland Mr. Tim Hennessy

Members by Right European Seismological Commission (ESC) - Ms. Alice WalkerObservatories and Research Facilities for European Seismology (ORFEUS) - Dr. Bernard DostInternational Seismological Centre (ISC) - Dr. Ray Willemann

Applications for EMSC membershipSeismological Institute, (ASN) Albania Dr. E. DushiCentre de Recherche en Astronomie, Astrophysique et Géophysique (CRAAG) Algeria Dr. A. K. Y. ChaoucheDepartment of Geophysics, University of Zagreb (IRB) Croatia Prof. M. HerakInstitute of Geophysics (TIF) Georgia Prof. T. ChelidzeDirection Environnement Urbanisme et Construction (DEUC) Monaco Dr. P. MondielliMontenegro Seismological Observatory (MSO) Yugoslavia Dr. B. Glavatovic

EMSC,coordinator

of an E.C. funded project

EMSC,specialized European Centre

for the Open Partial Agreement

Marmureanu