| |
Researchers
Abstract
Keywords
Introduction
Data processing
Discussion
Conclusions
References |
|
Heba Karmash
Higher Institute of Earthquakes Studies and Research
heba93karm@gmail.com
Mohamad Daoud
Department of geology-Faculty of Science -Damascus university
mdaoud2360@gmail.com
Raed Ahmad
National Earthquake Center
raedali_2000@yahoo.com
.
|
|
|
Abstract
The characteristics of the local site have a strong influence on the intensity of the ground shaking. There are many methods for estimating site response including numerical and experimental methods (spectral ratio and Fourier spectrum).
In this study, we analyzed 742 real seismic records (magnitude >3) recorded at Syrian Seismological Network Stations (24 stations) using horizontal to vertical components spectral ratio (HVSR) on both earthquakes and noise recordings. Consequently, HVSR curves were obtained and the predominant period of soil (T0) and amplification factor (A0) were determined for each site. The results indicate that this compatibility is better in rocky sites where the variation in the seismic impedance is bigger.
It was noted that most of the sites classified as B or C class, according to the applied methods in this study. The design response curves have been drawn according to the mentioned criteria in the Syrian Anti-Seismic code 2012.
Finally, we developed Prediction Equations of Horizontal Ground Motion components for acceleration and velocity (PGAh and PGVh) using Regression analysis, which gave a good accuracy of the analysis.
|
|
|
Keywords:
site response, spectral ratio h/v, amplification, predominant period, Design response curve.
|
|
| up |
|
| |
|
|
Introduction
Local site effect assessment has become very important in earthquake studies and the spectral ratio method is one of the most common methods which are used to determine the seismic site response. First, H/V Spectral ratio was applied by (Nakamura, 1989) on the ambient noise recordings then it has been extended to include real seismic recordings. (Lermo and Chavez-Garcia, 1993) tested this method on the S-wave portion of recordings at three different sites in Mexico and found very encouraging similarities in results with Standard Spectral Ratios (SSR) method which have developed by (Borcherdt & Gibbs,1970).
Also, the development of a modern and applicable attenuation equation has been become one of the main concerns of seismologists.
Tabel-1- shows an abbreviation for some researches about the spectral ratio method and developing an attenuation relationship of peak ground motion using Earthquakes.
In this study, We apply H/V spectral ratio method using 56 recorded earthquakes by 24 stations of Syrian Seismological Network to determine the predominant period (T0) and amplification factor (A0) for each site as well as site classification.
Also, we calculated Vs30 at each station using topographic slope and geological structure to obtained the design response spectra and developed an attenuation relationship (Prediction Equations) which never been discussed before in Syria using Syrian earthquake data.
Tabel-1- Literature review for the H/V spectral ratio method on earthquakes recordings and attenuation models.
|
|
|
Researcher
|
Data
|
Results
|
|
Mucciarelli & Gallipoli, 2004
|
608 earthquakes recorded at 4 stations in Italy and 1,280 microtremors
|
· Great similarities in the results for two stations.
· Difference appeared in the remaining two stations due to the complex geological location.
|
|
Laouami et al, 2018
|
1391 records of the strong motion in Algeria
|
· Amplitudes and shapes of the spectral H/V ratios for each site class.
· Develop a prediction equation for ground motion using predominant period.
· it was found that the use of Algerian data leads to an exaggerated prediction of ground motions.
|
|
Khalil, 2012
|
HVSR method on the ambient noise in Eastern Ghouta in Damascus – Syria
|
· The effect of topsoil on site response.
· Plotting HVSR curves for each site.
|
|
Chiou and Youngs,2008
|
Depending on VS30
|
· The effect of the fault type is weaker for the aftershocks of the main shock. So his relationship does not include the study of aftershocks and he chose a reference velocity value of 1130 m/s.
|
|
Idriss,2008
|
The effect of VS30 was not taken into account
|
· It was noticed that at magnitude (7.2 < ML > 5.2), the inferred relationship well expresses the damping of the ground motion as it gave a good fit at distances between the surface center and the recording point greater than 15 km.
|
|
Malkawi et al,1995 &
El Ssayed et al,2012
|
|
· The main objective is to study seismic hazard.
· Attenuation relationship based on global relations.
|
|
Alchalabi et al,2003
|
stations of the Syrian Seismological network in syria
|
· Analyzing the recordings of strong motions only.
· obtain Fourier spectra and They didn’t derive a specific attenuation relationship to Syria.
|
|
This study
|
56 earthquakes recorded by 24 stations in syria
(82 accelerograms and 660 velocity recordings)
|
· Determine site response curve and their classification. considering that the inferred response curves represent reference curves.
· In this study, ground motion prediction equations were done based on the value of Vs30 at each site for the first time in syria.
· Our equations are based on earthquake database recorded in Syria.
|
|
|
The geological, tectonic and seismic situation of Syria
The geologic evolution of Syria has created more suitable conditions for the preservation of hydrocarbons. Reservoirs were formed in the extensive clastic and carbonate deposits, most particularly in the Mesozoic, with source rocks throughout the section. The traps are mostly structural in the form of fault blocks or fault-propagation folds (Brew et al,2001).
Most of the earthquakes occur along with the four major tectonis zones which are:
Abdel-Aziz- Sijar uplifts, Eurphrates fault system, Palmyride fold belt and Dead sea fault system.
Figure -1- & Figure -2- show the Distribution of the Stations on the geological map of Syria (After Brew et al,2001) and the earthquakes were recorded since 1995.
|
|
 |
|
|
Figure -1-Distribution of Syrian seismological Network Stations on the geological map of Syria
( scale 1/50000) (after Brew et al,2001).
|
|
 |
|
| Figure -2- The earthquakes were recorded in syria since 1995 till 2012 by the Syrian Seismological Network in Syria (Internal report, National Earthquake Center). |
|
|
Data processing
In this study, 56 earthquakes with a magnitude ranged between 3 and 5.5 were analyzed (are shown in Figure 3 ), recorded in The Syrian Seismological Network during the years: 2009, 2010 and 2011 using weak and strong motion instruments (CMG-40T and CMG-5T).
|
|
 |
|
| Figure -3- Geographical distribution of Earthquakes used in our study. |
|
The regression analysis was also applied to the seismic recordings in order to establish an attenuation relationship describing peak acceleration and velocity ground motion recorded on the horizontal component (PGAh and PGVh) in Syria, taking into account the site effect represented by the value of the shear wave velocity within the depth of the first 30 meters from the surface(VS30).
In addition, Vs30 was calculated based on high-resolution digital images of the topographic elevation and the geological situation at each site, according to the equation developed by (Ahmad, 2014) (who used the topographic slope) and a developed model by (Stewart et al, 2014) which combines the geological conditions of the site and topographic slope.
|
|
Most of the studied stations are classified as B and C class, while ZALF and MNKR are D class, we depended on the predominant period calculated from the HVSR method according to the definitions were set by (Zhao et al, 2006).
Figure _8_ shows the design response curves resulted from this study for 4 major cities in Syria and their comparison with the mentioned values in the Syrian Anti-seismic Code
|
|
 |
 |
|
 |
 |
|
| Figure -8-Design response curves resulted from our study and compare them with the curves from syrian code 2012 for each site |
|
| Based on the data and the ground attenuation model developed by (Boore et al, 1997), we obtained the relationships of the data recorded in the Syrian network, taking into account the value of the reference shear velocity set by (Chiou and youngs, 2008) which was 1130 m/s: |
|
 |
|
| Figure (11) shows the distribution of the residual values of the inferred PGAh and PGVh relationship. The results gave a good fit for the values of the PGVh analysis while the fit was weak in the PGAh analysis and that is well known in this type of studies when the values of accelerations are overesimated. |
|
| |
|
| up |
|
|
Discussion
The analysis window was chosen to contain the complete seismic signal from the arrival of the P-wave until the end of the seismic signal according to the studies mentioned above. the spectral ratio H/V was also calculated with a smoothing constant of 15.
In this study, the seismic noise that precedes the arrival of the real seismic signal was analyzed individually (Nakamura method), and the results of the analysis of velocity, acceleration and seismic noise recordings were compared with each other for each site as shown in the Figure-6.
Subsequently, Nakamura method gave similar results to the real seismic signal analysis in 10 stations, While it gave different results in the remaining 14 stations. This is due to the difference in frequency content from one earthquake to another. The stations are characterized by different amplification, according to the geological situation and topograpghy of the site.
|
|
 |
|
| Figure-4- H/V spectral ratios curves and the mean curve of BIDA station |
|
 |
|
| Figure -5- Fourier amplitude spectrum curves and the mean curve of BIDA station. |
|
 |
 |
 |
|
 |
| Figure -6- Mean H/V spectral ratio curves of 3 stations for Noise, velocity and acceleration records analysis |
 |
 |
|
| Figure -7-Mean Curve of Fourier amplitude spectrum and sort by magnitude for BIDA & ARNB. |
|
|
Most of the studied stations are classified as B and C class, while ZALF and MNKR are D class, we depended on the predominant period calculated from the HVSR method according to the definitions were set by (Zhao et al, 2006).
Figure _8_ shows the design response curves resulted from this study for 4 major cities in Syria and their comparison with the mentioned values in the Syrian Anti-seismic Code
|
|
|
|
|
|
|
|
| |
|
| Figure -8-Design response curves resulted from our study and compare them with the curves from syrian code 2012 for each site. |
|
| Based on the data and the ground attenuation model developed by (Boore et al, 1997), we obtained the relationships of the data recorded in the Syrian network, taking into account the value of the reference shear velocity set by (Chiou and youngs, 2008) which was 1130 m/s: |
|
 |
|
| Figure (11) shows the distribution of the residual values of the inferred PGAh and PGVh relationship. The results gave a good fit for the values of the PGVh analysis while the fit was weak in the PGAh analysis and that is well known in this type of studies when the values of accelerations are overesimated. |
|
 |
 |
|
| Figure -9-Relation between PGA with Magnitude and distance. |
|
 |
 |
|
| Figure -10-Relation between PGV with Magnitude and distance |
|
 |
 |
|
| Figure -11- Residuals values for PGV and PGA. |
|
| up |
|
|
Conclusions
This study completes the characterization of the Syrian Seismological Network. We analyzed 56 events recorded with a magnitude more than 3 in 24 stations during the years: 2009, 2010 and 2011 using the Sesarray-Geopsy analysis program.
Then we compared with the results of noise analysis Pre-arrival seismicity of the real signal (Nakamura method) using HVSR method and Fourier analysis to calculate soil predominant period (T0) and amplification (A0), the results of the two analyses agreed in 10 stations, while they gave different values in the remaining 14 stations.
Also, it’s noted that the great influence of the topography and local geology of the sites on the results, especially in MNKR and QASN stations and most stations were classified in B and C. the code gave higher spectral accelerations because it represents the large earthquakes while our study was for the weak earthquakes, also due to the difference in the amplification factor and the applied methodology.
Finally, the inferred attenuation relationships are suitable to be used in estimating ground motion coefficients for earthquakes with M<5, and it is expected that they will lead to high deviations in estimating ground motion coefficients for large and destructive earthquake, due to the absence of these earthquakes in the database.
|
|
| up |
|
|
References
• Ahmad, r. (2014). Amplification maps for syrian territory with respect to ground motion levels and based on slope_velocity model. Seismol.Earthw.Eng(16), 231_245.
• Alchalabi,A.,Costa,G. and Suhadolc,P.(2003)."Strong motion records from syria: a preliminary analysis. Skopje,Earthquake 40 years of European Earthquake Engineering,1-10.
• Boore, D., Joyner, W., & Fumal, T. (1997). Equations for estimating horizontal response spectra and peak acceleraions from western north america earthquakes: A summary of recent work. Seism.Res.letter(68), 128-153.
• Borcherdt, R., & Gibbs, J. (1970). Effects of local geological conditions in the San Francisco Bay region on ground motions and the intensities of the 1906 earthquake. Bull.Seism.Soc.Am(66), pp. 467-500.
• Brew, G., Sawaf, T., Al_Maleh, K., & Barazangi, M. (2001). Tectonic and geologic evaluation of syria. GeoArabia(6), 573_615.
• Chiou, B., & Youngs, R. (2008). An NGA model for the average horizontal component of peak ground motion and response spectra. Earthq.Spectra(24), 173_215.
• El Ssayed,H., Zaineh,H.,Dojcinovski,D. and Mihailov,V.(2012)."Re_Evaluations of seismic hazard of syria. International journal of Geosciences vol.3 No.4A, 9 pages.
• Husein Malkawi, A.I .,Liang, R., Nusairat,J and Al hamoud,A.(1995).”Probabilistic seismic hazard zonation of syria”. Nat.Hazard 12(2), 139-151.
• Idriss, I. (2008). An NGA empirical model for estimating the horizontal spectral values generated by shallow crustal earthquakes. Earthq.spectra(24), 217_242.
• Internal report- National Earthquakes center.2013. Syrian seismicity analysis using catalog data for the period 1365 BC until the year 2013
• Khalil, Ahlam, 2012." ESTIMATION OF SURFACE GEOLOGY EFFECT ON THE SITE RESPONSE BASED ON NAKAMURA'S TECHNIQUE _A CASE STUDY IN EASTERN GHOOTA, DAMASCUS CITY, SYRIA". Thesis, Higher institute of earthquakes studies and researches.
• Laouami, N., Slimani, A., & Larbes, S. (2018). Ground motion prediction equations for Algeria and surrounding region using site classification based on H/V spectral ratios. Bull Earthquake Eng.
• Lermo, J., & Chavez-Garcia, F. (1993). Site effect evaluation using spectral ratios with only one station. Bull.Seism.Soc.Am(83), 1574_1594.
• Mucciarelli, m., & Gallipoli, M. (2004). The HVSR technique from microtremor to strong motion: Empirical and Statistical consideration. 13th world conference on earthquake engineering, (p. Paper No.45). Vancouver,B.C,Canada.
• Nakamura, Y. (1989). A method for dynamic characterstics estimation of subsurface using microtremor on the ground surface. Quarterly report, Railway Technical research institute .
• Stewart, J., Klimis, N., Savvaidis, A., Theodoulidis, N., Zargli, E., Athanasopoulos, G., et al. (2014). Compilation of a local VS profile database and its application for inference of VS30 from Geologic and Terrain Based proxies. Bulletin of the seismological society of america (104).
• Syrian anti-seismic code 2012.
• Zhao, J., Irikura, K., & Zhang, J. (2006). An empirical site classification method foe strong motion stations in japan using H/V response ratio. Bull Seism Soc Ame(96), 914_925.
|
|
| up |
|
|
