SESAME: Site Effects Assessment Using Ambient Excitations EVG1-CT K. Atakan 1, A-M. Duval, N. Theodulidis, B.Guillier, J-L.Chatelain, P-Y. Bard, and the SESAME Consortium 1 Department of Earth Science University of Bergen Allégt.41, N-5007 Bergen, Norway Tel: Fax: The H/V Spectral Ratio Technique: Experimental Conditions, Data Processing and Empirical Reliability Assessment

The reliability of the H/V technique is evaluated systematically through the following aspects: Experimental conditions Data processing Empirical evaluation On the reliability of the H/V spectral ratio technique: Motivations SESAME EC-RDG EVG1-CT

SESAME EC-RDG EVG1-CT Experimental conditions H/V technique: Experimental conditions Objectives: This part is dedicated to investigations on the required experimental conditions (instrumental characteristics, data acquisition, external conditions) for warranting the stability and reproducebility of measurements Participants: CETE, Nice UJF/LGIT, Grenoble UiB, Bergen ITSAK, Thessaloniki ICTE/UL, Lisbon INGV, Rome

SESAME EC-RDG EVG1-CT Experimental conditions Description of the work: Instrument calibration and testing, and comparison of the instruments that are used in the project were completed, following a dedicated workshop in Bergen, Norway Assessment of the effects of the experimental conditions is performed and a large number of tests are conducted (see Poster 183 Paper ID 306). Milestones and results: Get together experiment for instrument calibration and testing Field work for studying the influence of experimental conditions Guidelines for measurements

Experimental Conditions: Examples of instrumentation from UiB SESAME EC-RDG EVG1-CT GBV-316´s are used with different casing Data acquisition is a Pocket-PC based SEISLOG system

Experimental Conditions: Instrument tests SESAME EC-RDG EVG1-CT Instrument tests in the lab Instrument tests in the free-field

Sensor tests: seismometers vs accelerometers SESAME EC-RDG EVG1-CT seismometers and 5 accelerometers

Instrument comparisons: Lab vs free-field SESAME EC-RDG EVG1-CT

Experimental Conditions SESAME EC-RDG EVG1-CT Effect of weather conditions Rain and strong wind Snow Reference

Experimental Conditions SESAME EC-RDG EVG1-CT Effect of ground coupling Asphalt Grass Pavement

Experimental Conditions SESAME EC-RDG EVG1-CT The aim is to evaluate the influence of various experimental conditions on the H/V results. In total 58 different tests were performed and the results are classified into three main categories: (1)The experimental conditions that do not influence the results (2)Those that influence the results (3)Those that may influence the results (1)Beyond some limits that can easily be controlled and set up (2)Beyond a limit where there is no control In order to reach a conlusion statistical tests (Student-t test) are applied systematically to all measurements and standard ”Technical cards” are produced

Experimental Conditions SESAME EC-RDG EVG1-CT Technical card example

Experimental Conditions SESAME EC-RDG EVG1-CT Out of 58 different tests that were performed 22 tests in general do not influence (OK) the H/V results 17 tests may influence (May Infl.) the results 13 tests show that the results influence the H/V ratios and are not recommended (N.R.) 6 tests have not enough data (N.E.D) to reach a conclusion

Experimental Conditions SESAME EC-RDG EVG1-CT Following test results influence the H/V ratios and are not recommended (N.R.): Some ground-sensor coupling conditions (gravel, karstic limestone, etc.) Modified sensor coupling (i.e. artificial interface between the sensor and the ground) Weather conditions (rain and especially wind) Buried underground structures

SESAME EC-RDG EVG1-CT Data Processing H/V technique: Data processing Objectives: This part is devoted to investigations on the various data processing alternatives and produce a standard processing software based on the most robust alternative Participants: UiB, Bergen UJF/LGIT, Grenoble CNR-IRRS, Milano ICTE/UL, Lisbon ETH, Zurich INGV, Rome CETE, Nice Résonance, Zurich

SESAME EC-RDG EVG1-CT Data Processing Description of the work: Existing algoritms for each processing step were collected The best (robustness-simplicity-resolution) solution for processing is found based on comparisons with representative data sets Design and implementation is done by a dedicated sub-group Distributing for evaluation and review within consortium Final version and user guidelines Milestones and results: Report with processing options Report with these options on different noise data sets Recommendations and software implementation (J-SESAME) (see Poster 199 Paper ID 2270)

Data processing Software: J-SESAME SESAME EC-RDG EVG1-CT H/V Software (J-SESAME) development is mainly done by: Browsing module (Bladimir Moreno, UiB Bergen) Window selection module (Pierre-Yves Bard, LGIT Grenoble) Processing module (Alberto Tento, CNR Milano; Fortunat Kind, ETHZ) Display module (Pedro Roquette, UL Lisbon) Testing of the processing module was performed by three groups taking in to account the machine dependence ad functionality. A specific parametric survey was conducted and some format conversion programs are developed: CETE, Nice (Etor Querendez, Anne-Marie Duval) INGV, Rome (Fabrizio Cara, Giuseppe Di Giulio, Giovanna Cultrera) IRD-LGIT, Grenoble (Jean-Luc Chatelain, Bertrand Guillier)

Data processing Software SESAME EC-RDG EVG1-CT Browsing Module

Data processing Software SESAME EC-RDG EVG1-CT Window Selection Module

Data processing Software SESAME EC-RDG EVG1-CT Processing Module

Data processing Software SESAME EC-RDG EVG1-CT Display Module

Data processing Software SESAME EC-RDG EVG1-CT Display Module

SESAME EC-RDG EVG1-CT Empirical evaluation H/V technique: Empirical evaluation Objectives: Systematically evaluate and compare (using the J-SESAME software developed) the H/V ratios with other more reliable estimates (reference site). Compare H/V results with observed damage on recent earthquakes Participants: ITSAK, Thessaloniki UJF/LGIT, Grenoble UiB, Bergen ETH, Zurich ICTE/UL, Lisbon INGV, Rome CNR-IRRS, Milano CETE, Nice

SESAME EC-RDG EVG1-CT Empirical evaluation Description of the work: Collecting existing data sets (earthquake and noise) Perform experimental measurements and processing of ambient vibrations at selected sites Systematic comparisons with weak- and strong-motion data as well as damage disribution in urban areas Comparing experimentally and theoretically estimated transfer functions with H/V ratios Milestones and results: Collection of different data sets Report on the experimental measurement results Comparison of the results (noise vs other data sets) (see Poster 200 Paper ID 2323)

SESAME EC-RDG EVG1-CT Empirical evaluation SESAME Standard Information Sheets (SIS) A relational database is created storing the standardized information for each experiment´s data set Data are collected from: EQ + Noise 200 sites EQ + damage + noise 6 locations

SESAME EC-RDG EVG1-CT Empirical evaluation INSTITUTE SITENAMESISEVENTSRECORDS ITSAK S.M. NETWORK EUROSEISTEST INGV BENEVENTO CATANIA CITTA-CASTELLO VERCHIANO COLFIORITO CNR FABRIANNO PREDAPPIO ROVETTA LGIT NICE EBRON GUADELUPE TEHRAN LOURDES GRENOBLE ETHZ S-M NETWORK TOTAL

SESAME EC-RDG EVG1-CT Empirical evaluation

SESAME EC-RDG EVG1-CT Empirical evaluation Comparison between the H/V ratios of noise (black) and earthquake (red) data from two sites in Greece. 110 freq. (Hz) 1 ( h / v ) aml1-N/Z-run Frequency (Hz) H/V 110 freq. (Hz) 1 ( h / v ) arg1-N/Z-run Frequency (Hz) H/V

SESAME EC-RDG EVG1-CT Conclusions Instrumental and experimental conditions, which may influence the H/V ratios are identified. Among the 58 different experimental conditions 13 gave clear influence. The most critical seems to be the effect of wind. Dedicated software (J-SESAME) is developed to allow standard processing routines. A database of earthquake and noise records are compiled for comparison. Standard Information Sheets (SIS) are prepared for easy access.

SESAME EC-RDG EVG1-CT Conclusions Correlations that were performed between the H/V ratios and near-surface geology gave promising results. However, correlation with damage seems difficult to obtain due to the complex nature of damage distribution. Guidelines are being prepared for the use of H/V spectral ratio technique, including experimental factors affecting the results, the standardized processing software with a user manual and recommendations with regard to the empirical evaluation. In parallel to the work presented here, within the SESAME project numerical simulations are also perfomed to better understand the nature of the noise field and to investigate the actual capability of H/V technique to retrieve useful information on site conditions.

SESAME EC-RDG EVG1-CT SESAME Project Main results and reports on the ongoing work can be found at the SESAME Project web-site: See also the Posters 183,199 and 200

SESAME EC-RDG EVG1-CT

Experimental Conditions SESAME EC-RDG EVG1-CT

Experimental Conditions SESAME EC-RDG EVG1-CT TEST Number P1-1-1P1-1-2 P1-2P1-3P1-4P1-5P1-6P1-7P1-8 CONCLUSIONOK May Infl. OKMay Infl. OK May Infl. TEST NumberP2-1P2-2P2-3P2-4P2-5 P2-6-1P2-6-2 P2-7P2-8P2-9P2-10 CONCLUSIONMay Infl. N.R.OKMay Infl. OKMay Infl. N.R. TEST Number P3-1-1P3-1-2P3-1-3P3-1-4P3-1-5P3-1-6P3-1-7 CONCLUSIONN.R. May Infl. OK

Experimental Conditions SESAME EC-RDG EVG1-CT TEST Number P3-1-8P3-1-9P P P P CONCLUSIONOKN.R.May Infl. OK TEST NumberP3-2P3-3P3-4P3-5P3-6P3-7 CONCLUSIONOKN.R.OKMay Infl. OKMay Infl. TEST NumberP4-1P4-2 CONCLUSIONN.R.

Experimental Conditions SESAME EC-RDG EVG1-CT TEST NumberP6-1P6-2P6-3P6-4P7-1 CONCLUSIONN.R. N.E.DOKN.E.D TEST NumberP8-1P8-2P8-3P8-4 CONCLUSIONN.E.DOK TEST Number P9-1P9-2-1P9-2-2P9-3P9-4-1P9-4-2P9-5P9-6P9-7 CONCLUSIONMay Infl. N.E.DMay Infl. N.E.DOKN.E.DOK