Site Amplification Factors Research Articles

Machine learning based ground motion site amplification prediction

Site condition impact on seismic ground motion has been a complex but important subject in earthquake hazard analysis. Traditional studies on site amplification effect are either based on site response via wave propagation simulation or regression analysis using parameters such as Vs30, bedrock ground motion and site response period. Ground Motion Prediction Equations (GMPEs) are used for regions where there is limited data of seismic records. The main issues with these approaches are that they cannot demonstrate the complex relationship between site amplification and its various affecting parameters, thus there exists large uncertainty in the results. Recent studies based on machine learning have shown significant improvement in predicting the site amplification, but the result is not well explained. This study assembled the information on 6 parameters including Vs30, magnitude, epicentral distance, earthquake source depth, bedrock ground motion, and altitude of 353,327 records observed during 1997 and 2019 from 698 KiK-net stations. Three machine learning algorithms of Random Forest (RF), XGBoost, and Deep Neural Networks (DNN) were implemented to predict the site amplification factor using these 6 selected parameters. Shapley Additive explanation (SHAP) was used to explain the importance of the 6 parameters. The results show that all three machine learning algorithms performed much better than the traditional GMPE approach with XGBoost’s performance the best. The explanation provided by the SHAP analysis further enhanced the reasonability of this study. It is anticipated that the combination of machine learning and SHAP analysis can provide better assessment for site amplification of ground motion with better potential of future application in seismic hazard analysis.

Subsurface structure identification at the blind prediction site of ESG6 based on the earthquake-to-microtremor ratio method and diffuse field concept for earthquakes

We participated in the blind prediction exercise organized by the committee of the blind prediction experiment during the 6th International Symposium on Effects of Surface Geology on Seismic Motion (CBP-ESG6). In response to the committee's request, we identified the ground velocity structure from microtremors observed at a target site as the first step of the exercise. First, we calculated the horizontal-to-vertical spectral ratio of microtremors (MHVR) at the target site from the distributed microtremor data collected in the vicinity of the target site in Kumamoto Prefecture. Then, we converted the MHVR into a pseudo horizontal-to-vertical spectral ratio of earthquake (pEHVR) using the previously proposed and validated earthquake-to-microtremor ratio (EMR) method, where an empirically obtained EMR is used to convert MHVR into pEHVR. Next, we inverted the S-wave and P-wave velocity structures based on the pEHVR and the diffuse field concept for earthquakes. The theoretical EHVR calculated from the identified velocity structure reproduced the pEHVR quite well in the frequency range of 0.1–22 Hz. After the collection of the blind prediction results by all the participants, the CBP-ESG6 released the observed earthquake records, a preferred model based on the P–S logging data from the in-situ borehole measurement combined with the generic deeper structure, and the average of all the predicted structures by the participants. Notably, our inverted structure was found to be close to the preferred model and the averaged one of all the blind prediction participants, despite some minor differences in the horizontal site amplification factor around the maximum peak frequency at 0.8–1 Hz.Graphical

Scenario-based earthquake damage assessment of highway bridge networks

In earthquake-prone regions, the evaluation of seismic impacts on bridges is crucial to mitigation, emergency, and recovery planning for highway networks. The degree of bridge damage determines the cost and time required for repairs and the level of post-earthquake functionality including disruption of transportation network, increased costs due to reduction of traffic flow and restricted access to emergency routes. The article presents the methodological development and implementation of an interactive web application for rapid geospatial assessment and visualisation of earthquake damage scenarios of municipal highway bridge networks based on open access datasets. The proposed framework consists of the following successive models: hazard, inventory, damage, and impact. The seismic hazard model generates spatial distribution of the shaking intensity for earthquake scenarios in terms of ground motion intensity measure using ground motion prediction equations based on seismic hazard model for Eastern Canada. The shaking intensities are then modified with local site amplification factors based on the Canadian highway bridge design code values. The inventory model provides a database of existing bridges based on open-access data which are then classified according to their seismic vulnerability. The damage model assesses seismic performance of classes of bridges by applying respective fragility functions represented as probabilistic relationships between the intensity measure and the degree of expected damage. The impact model evaluates the post-earthquake traffic-carrying capacity of the highway network based on the predicted damage including repair cost as a percentage of replacement cost of bridges and inspection priority. The web-application is demonstrated with a bridge network in Quebec City including 117 bridges subjected to 180 earthquake scenarios. The proposed methodology is particularly useful to facilitate direct communication of potential impacts to emergency managers and city transport officials.

High-frequency S and S-coda waves at ocean-bottom seismometers

To clarify the characteristics of high-frequency (> 1 Hz) S and S-coda waves at ocean-bottom seismometers (OBSs), we analyzed seismograms observed at permanent OBSs and inland broadband seismometers around the Kii Peninsula in southwest Japan along the Nankai Trough. The coda amplitudes (both horizontal and vertical) at the OBSs were much larger than those at the inland rock-site stations. Because coda amplitudes relative to those at inland rock-site stations have been used as site-amplification factors, large site amplifications for both components can be expected due to the presence of thick oceanic sediments just below the OBSs; however, the observed maximum S-wave amplitudes in the vertical component exhibited similar attenuation trends against epicentral distances at both OBS and inland stations. To clarify the causes of this discrepancy, we conducted numerical simulations of seismic wave propagation using various three-dimensional seismic velocity structure models. The results demonstrated that coda waves at OBSs mostly comprise multiple scattered waves within a thick (> 2 km) sedimentary layer; consequently, coda amplitudes at OBSs become much larger than those at inland rock-site stations. Our numerical simulations also confirmed the generation of large coda amplitudes at regions with seawater depths ≥ 4 km, where no OBS was deployed. However, the thick sedimentary layer and seawater have limited effects on maximum S-wave amplitudes at the OBSs. Given that the effects of a thick sedimentary layer and seawater on S and S-coda waves differ, we concluded that the coda-normalization technique for site-amplification correction against a rock-site station could not be applied if stations are located within regions above the thick sedimentary layer or deeper sea depths. The site amplifications at the OBSs were corrected according to the horizontal-to-vertical ratios at each OBS; we adjusted the simulated horizontal envelopes at the OBSs using these ratios of the observed S-coda waves. As well as inland seismometers, the site-corrected simulation results practically reproduced the observed high-frequency envelopes at OBSs.Graphical

Influence of local site effects on seismic risk maps and ranking of Italian municipalities

The latest studies concerning seismic risk assessment in Italy were developed without considering the site geo-lithological effects, thus limiting the assessment to rock soil type. In this study, for the evaluation of site amplification effects, we used the results of recent works based on the Italian seismic microzonation data. As this first study is performed for the entire Italian territory at the municipality scale, the site amplification factors (AFs) have been assigned to the chief town of each of the 7715 Italian municipalities, assuming that the building stock is concentrated in that small area. The AFs have been compared with those foreseen by the Italian building code (NTC2018). The PGA amplification with the new AFs ranges from 1 to 2.2 bringing to an average increase of 75% in the hazard, compared to 27% if using the NTC2018 amplification. The seismic risk has been evaluated using the probabilistic hazard assessment adopted in the NTC2018 and recent vulnerability/exposure models developed for Italy. The residential building stock was subdivided into 5 vulnerability classes (3 for masonry and 2 for concrete) derived from the national census data and further refined using the construction age and building height. The results show that, for a rock site condition, the loss estimates are comparable with those of previous works in terms of casualties, homeless, and economic damage. The introduction of the site effects brings to a significant increase in the expected losses resulting higher than the historical data retrieved from a careful analysis of the major Italian earthquakes. This suggests the importance of developing new fragility curves based on updated AFs. To support the implementation of risk reduction programs, a seismic risk index is introduced, allowing a municipality ranking by risk, a cost estimation of the seismic retrofitting in the highest risk municipalities, and an assessment of the consequent risk reduction.

U.S. National VS30 Models and Maps Informed by Remote Sensing and Machine Learning

Abstract The shear-wave velocity time averaged over the upper 30 m (VS30) is widely used as a proxy for site effects, forms the basis of seismic site class, and underpins site-amplification factors in empirical ground-motion models. Many earthquake simulations, therefore, require VS30. This presents a challenge at regional scale, given the infeasibility of subsurface testing over vast areas. Although various models for predicting VS30 have thus been proposed, the most popular U.S. national, or “background,” model is a regression equation based on just one variable. Given the growth of community data sets, satellite remote sensing, and algorithmic learning, more advanced and accurate solutions may be possible. Toward that end, we develop national VS30 models and maps using field data from over 7000 sites and machine learning (ML), wherein up to 17 geospatial parameters are used to predict subsurface conditions (i.e., VS30). Of the two models developed, that using geologic data performs marginally better, yet such data are not always available. Both models significantly outperform existing solutions in unbiased testing and are used to create new VS30 maps at ∼220 m resolution. These maps are updated in the vicinity of field measurements using regression kriging and cover the 50 U.S. states and Puerto Rico. Ultimately, and like any model, performance cannot be known where data is sparse. In this regard, alternative maps that use other models are proposed for steep slopes. More broadly, this study demonstrates the utility of ML for inferring below-ground conditions from geospatial data, a technique that could be applied to other data and objectives.

Microtremor array observation in Southern Part of the Bandung basin: deep 1d s-wave velocity structure

Site amplification factor become crucial in seismic hazard microzonation, which depends on local geological condition, therefore identification of bedrock depth is very important related to sediment thickness. One of the Geophysical method that can be applied for subsurface investigation is microtremor array survey method (MAM). This method doesn’t require seismic source, so it is very possible to be applied in the urban area. Bandung basin is a research project for seismic hazard and risk study, where most of this area covered by high population and settlements that becomes the obstacle in this field survey. Geologically, Bandung basin consists of volcanic rock and floodplain, and we can find outcrop of Tertiary rock in some parts the basin. Seismicity in Bandung basin is very active because there are many seismic sources around it, such as shallow crustal and megathrust segment in the southern part. The microtremor array survey was conducted in Southern basin at 3 sites observation represent shallow and deep bedrock depth by using triangular array in shape. The result shows the bedrock depth in BB12 is 139 m, BB13 is 211 m and BB14 is 192 m. This condition represents the feature of bedrock position in Southern Bandung basin.

Regional Spectral Characteristics Derived Using the Generalized Inversion Technique and Applications to Stochastic Simulation of the 2021 Mw 6.1 Yangbi Earthquake

ABSTRACTIn this study, 334 accelerograms of 42 small-to-moderate earthquakes recorded at 36 strong-motion stations were used to investigate the ground-motion characteristics of the southwestern margin of the Sichuan–Yunnan rhombic block (SWM-SYRB), one of the most active seismic regions in China. The high-frequency attenuation parameter was estimated using the spectral decay method, and the site component (κ0) was fitted. The κ0 estimates decrease with the increasing time-averaged S-wave velocity of the uppermost 30 m (VS30). Using the generalized inversion technique, the source spectra, quality factor (Q), and site amplification were derived from the Fourier amplitude spectra (FAS). The obtained average stress drop for earthquakes occurred in SWM-SYRB was the second largest among various boundary areas of SYRB. The inverted Q model was Q(f)=115.1f0.616. The low Q structure that extends southwestward from the Songpan–Garze block to SWM-SYRB could be responsible for the strong regional attenuation of ground motion with distance. At frequencies above 10 Hz, the average site amplifications were influenced by the high-frequency attenuation effect. The site amplification of site class D reached a factor of 6 at 0.7 Hz. Moreover, it was observed that site amplification factors can be even higher when peak ground acceleration is larger than 0.8 m/s2. Finally, the obtained parameters were used in the stochastic finite-fault simulation method to reproduce the FAS, 5%-damped pseudospectral acceleration, and time series of the 2021 Mw 6.1 Yangbi earthquake. The simulated spectra properly matched the observations in a broad frequency band of 0.1–25 Hz. Furthermore, the simulated time series could generally represent the amplitude of the S-wave portion of observed ones.

A hybrid approach for deriving horizontal site amplification factors considering both the similarity of HVSRE and the vertical amplification correction function

AbstractUnder hypothetical and idealized conditions, the horizontal site amplification factor (HSAF) is defined as the ratio of the horizontal Fourier amplitude spectrum (FAS) at a point on the Earth's surface with respect to the horizontal FAS at the identical location but on the outcropping seismological bedrock. The HSAF reflects the profile of local sedimentary soils and weathered rock formations, indicating site effects. In most cases, such an idealized HSAF is difficult to measure directly. Thus, quantitatively estimating a HSAF value is a crucial task in strong‐motion prediction over the last century. Fortunately, many strong‐motion stations have been constructed throughout Japan, facilitating the characterization of HSAFs down to seismological bedrock at more than 1600 observational sites by the generalized inversion technique (GIT). First, this study reported the similarity distribution of the HSAF and the horizontal‐to‐vertical spectral ratio of earthquakes (HVSRE). Subsequently, we proposed a hybrid method from a novel aspect for estimating HSAF in terms of the observed similarity distribution and the vertical amplification correction function (VACF) proposed previously. Compared with the direct use of HVSRE for substituting HSAF, the hybrid method proposed herein demonstrated an improvement of greater than 30% in terms of the residuals between estimated HSAFs and those separated from observations.

An Empirical Site Amplification Model for the North–South Seismic Belt of China

Abstract Considering 869 ground-motion recordings from moderate-to-strong earthquakes that occurred in 2007–2019, this work investigates the variation of site amplification factors and develops an empirical site amplification model of the north–south seismic belt of China. The work first regresses a simple empirical ground-motion model (GMM) for the averaged reference site (the average shear-wave velocity for the top 30 m of the Earth (VS30) is about 760 m/s) by 322 recordings of the dataset. The regressed reference site model indicates that the slow distance attenuation still exists in earthquakes of the north–south seismic belt. Then, using the derived reference site model and 869 recordings, empirical site factors are computed and an empirical site amplification model is regressed. For the earthquakes in the north–south seismic belt, the proposed site amplification model predicts the site amplification for the 50th percentile rotated pseudospectral acceleration (RotD50 SA) with a damping ratio of 5% and periods 0.01–10 s. Moreover, the difference in site amplification factors between the north–south seismic belt of China and those of the Next Generation Attenuation (NGA) database is investigated. In general, variations of site amplification factors of the north–south seismic belt of China with VS30 are less significant than those from the NGA database. Therefore, the regional differences of ground motions caused by site effects are significant for the north–south seismic belt of China. For short periods (<1.0 s) and large values of VS30 (>300 m/s), the median site amplification of the north–south seismic belt is fitted well with most model predictions and the NGA data. The residual regional differences in the north–south seismic belt of China after site corrections can be applied to develop the more efficient global GMMs.

An updated stochastic finite fault modeling: Application to the Mw 6.0 earthquake in Jiashi, China

On January 19, 2020 at 21:27 local time, the Jiashi County in China witnessed an Mw 6.0 earthquake. The earthquake event was recorded by 42 strong ground-motion observation stations and recorded a total of 126 sets of acceleration time history records. In this study, we simulated the Jiashi earthquake based on the updated finite-fault modeling. The zero-distance high-frequency attenuation parameter κ0 and local site amplification factor were calculated based on the recordings, and the other input parameters were obtained from the previous studies. Then the Fourier amplitude spectrum, 5%-damped pseudo-acceleration response spectrum, peak ground acceleration, and accelerograms at 12 selected stations were synthetized using an updated finite-fault modeling. The results indicated that the synthetic results had no evident differences at most of the selected stations, which showed that the updated approach can be adopted to predict the ground motion. In addition, the simulated accelerograms are similar to the observed records at most stations, which indicated that the updated modeling can be used to generate the seismic input records required for structural analysis. In the future research, we should more analyze the interconnection between each input parameter and establish the relationship between parameters. If so, even though we cannot obtain the detailed information of several parameters, we can still accurately estimate the seismic field of imminent earthquakes in the future, which will help to establish the disaster maps of the target region.

Spectral decomposition of S-waves in investigating regional dependent attenuation and improving site amplification factors: A case study in western Greece

Spectral decomposition of S-waves in investigating regional dependent attenuation and improving site amplification factors: A case study in western Greece

Assessment of Seismic Vulnerability using the Ambient Noise Recordings in Cachar Fold Belt, Assam

Abstract Northeast India is a seismically active region, with frequent major earthquakes. The impact of seismicity can be assessed by using the values fundamental resonant frequency (F0) and the site amplification factor (A0) to derive the seismic vulnerability index (Kg). In 2011, a network of 65 stations was installed in the Cachar fold belt,Assam to continuously record data for a continuous period of eleven months. In this study, the horizontal-to-vertical spectral ratio (H/V) method is employed to identify the resonant frequencies and site amplification factors using ambient noise data to assess the seismic hazard potential in the Cachar fold belt. The amplification factors ranged between 1.15 to 6, while the resonance frequencies varied between 0.25 to 5 Hz. The ambient noise interferometry is utilized to estimate surface waves between station pairs and invert the group velocity dispersion curves for the near-surface S-wave velocities. The S-wave velocity for the basement rocks varied between 600 to 1600 m/s. The S-wave velocities are in turn used to derive an empirical relationship between resonant frequency and alluvium thickness valid for the study region. It is found that stations located at regions with subsurface anticlines had a relatively high resonance frequency and low amplification factor, whereas receivers in synclinal locations and with a thick layer of alluvium sedimentary deposits have a substantially higher site amplification factor and lower resonant frequencies. Further, it is observed that regions with a substantial alluvium sediment cover and near the course of the Barak river are more seismically vulnerable and susceptible to natural hazards.

Study of bilinear elastic horizontal displacement spectrum models

To obtain elastic horizontal displacement spectra for different damping ratios for use in displacement-based seismic design, the influences of the moment magnitude, fault distance, and site class on the 5%-damped elastic horizontal displacement spectrum are first investigated using the ASK14 ground motion prediction equation (GMPE). Then, bilinear displacement spectrum models for different damping ratios are proposed based on the statistical characteristics of the displacement spectra of strong motion records with the principle of retaining the major influencing factors and neglecting the minor ones, and the models are used to describe uniform hazard displacement spectra for a point source site with a return period of 475 years. Finally, the method for implementing the proposed models in practical applications is discussed. The results of the study show the following. The moment magnitude is the main factor affecting the shape of the displacement spectrum. The proposed bilinear displacement spectrum models can reflect the influence of the moment magnitude, reflect the statistical characteristics of the displacement spectra of the actual strong motion records, and well describe the uniform hazard displacement spectra of the point source site with a return period of 475 years. The zoning map of the two spectral displacements and the corresponding site amplification factors and damping scaling factors of the bedrock site at each seismic hazard level are provided to enable the implementation of the proposed bilinear displacement spectrum models in practical applications.

Deep-Neural-Network-Based Estimation of Site Amplification Factor from Microtremor H/V Spectral Ratio

ABSTRACT Site amplification factors (SAFs) of S wave at ground surface are crucial for evaluating and predicting seismic ground motions. This study proposed a novel methodology for directly estimating S-wave SAF from microtremor horizontal-to-vertical spectral ratio (MHVR) based on deep neural network (DNN) model. We analyzed site amplifications obtained from generalized spectral inversion technique and microtremor data observed at Kyoshin net and Kiban–Kyoshin network sites in Chugoku district, western Japan. The DNN model was developed using peak frequency and the frequency-dependent relationship between MHVRs and SAFs. The sites were divided into training set, validation set, and test set. The training set and validation set were used in k-fold cross-validation technique to evaluate and select optimal model. Once the optimal model had been determined, the model was employed on the test set that was completely independent of the training and validation set for evaluating the generalization performance. Residuals and root mean square errors between the estimated and observed SAFs were evaluated to discuss the applicability of the proposed model. We also confirmed that the DNN model shows better performance in estimating SAFs compared with the existing double empirical correction method.

Parameter estimation for predicting near-fault strong ground motion and its application to Lushan earthquake in China

Stochastic finite-fault modelling based on the dynamic corner frequency is a widespread method for synthesizing near-fault ground motions at high frequencies and has been proven to be an effective tool for simulating high-frequency ground motions by researchers globally. Many input parameters, which have more uncertain effects on the synthetic result, are related to sources, sites, and paths in stochastic modelling. Thus, it is difficult to select reasonable parameters. When using the stochastic finite-fault method, the key link to be considered is how to determine the input parameters required for simulations. Based on this fact, this study analyzed and discussed the sensitivity of several important input parameters in the simulation parameters to the simulation results. The results showed that for the Class C site stations, the crustal and local site amplification factors must be considered simultaneously. Overall, the influence of the high-frequency attenuation parameter κ0 calculated using different methods on the simulation results is not evident. For near-fault stations, various upper-edge buried depth synthetic response spectra can be set, and finally, the average value is taken as the final synthetic recorded response spectra. When using the random source model, the rupture starting point must be set on the subfault in the middle along the fault length and width direction. The conclusions provide a basis for a highly reasonable and rapid determination of input parameters when using the stochastic finite-fault simulations for practical engineering applications and can improve the simulation accuracy to apply the finite source modelling to rebuild scenarios and evaluate and alleviate the seismic hazard in the target region.

Site Amplifications from Earthquake Data and VS30 in the Fort Worth Basin, Texas

Abstract Following the development of unconventional oil and gas production in the Fort Worth Basin, Texas, a rapid increase in basin-wide seismicity began in 2008 that grew to include earthquakes affecting a substantial portion of the Dallas–Fort Worth metropolitan area. To assess and mitigate the seismic hazard, which in this region is impacted by the thickness of the sedimentary basin and accompanying soft soil layer, we estimate site effects at 22 seismic stations deployed to record these earthquakes. Site responses are derived using two different datasets and approaches: (1) a modified generalized inversion technique (GIT) based on the S-wave Fourier amplitude spectra from earthquakes; and (2) application of the quarter-wavelength approximation (QWA) using estimates of average shear-wave velocities in the upper 30 m, known as VS30. We find that site amplification estimates based on the two techniques are roughly consistent with one another (median of the amplification ratio = 0.92) over frequencies where a quarter-wavelength corresponds to ∼30 m depth. The site amplification factors from the two approaches are found on average to be about 3 at the quarter-wavelength frequency (QWF). These site amplification estimates are not well correlated with geologic characteristics including rock type and geologic age. Finally, QWA values at six sites do not match GIT site amplification at the QWF (outside of ± median absolute deviations boundary), which we attribute to a combination of the underlying assumptions of the QWA, uncertainty in VS30 estimates, and unmodeled site response complexity.

Re-evaluation of New Zealand seismic hazard for geotechnical assessment and design

This paper scrutinises the seismic hazard of New Zealand (NZ) from a geotechnical engineering perspective. The two codified versions of the seismic hazard in NZS1170.5 (structural loading standard) and NZTA Bridge Manual (NZTA-BM) are shown to yield consistently different peak ground acceleration (PGA) hazards throughout NZ. Results from site-specific PSHA for 24 locations in NZ are used to examine key hazard characteristics, including earthquake magnitude and the effects of site conditions (classes) on the PGA hazard. The comparative evaluations show that for most of the locations considered, NZS1170.5 and NZTA-BM overestimate the PGA hazard. However, NZS1170.5, and NZTA-BM in particular, significantly underestimate the PGA hazard for locations that are at a short source-to-site distance from the Hikurangi Subduction Zone (HSZ), and for which HSZ significantly contributes to their hazard. Using the results from this study, an interim PGA hazard is recommended for geotechnical assessment and design in support of the NZ guidelines for geotechnical earthquake engineering practice. The recommended interim PGA hazard is applicable to all site classes without any modification or use of site amplification factors.

Seismic hazard prediction using multispectral amplification maps in a complex topographic area: A case study of Qiaozhuang town, Sichuan Province, Southwest China

Earthquakes can cause widely distributed slope failures and damage in mountainous areas. The accurate prediction of ground motions in mountainous areas is essential for managing the seismic risk of urban cities near mountains but is restricted primarily by complex seismic site amplification effects in areas of uneven terrain. This study selected Qiaozhuang town located in the Qingchuan—Pingwu fault zone, Southwest China, as a case study. A simulator for mapped seismic responses using a hybrid model (SiSeRHMap) was applied to compute the multispectral seismic topographic amplification maps at the three slope units surrounding Qiaozhuang town (Weigan hill, Mt. Dong, and Mt. Shizi). Post-earthquake damage survey maps, 1D seismic site response spectral ratios, and H/V spectral ratios of earthquake data were used to validate the computed seismic site amplification factors and resonance frequencies. The results suggest that strong topographic amplification effects usually occur at distinct slope locations, such as hilltops, convex slope positions, upslope, and narrow ridges. The computed topographic amplification factors in the study area reached up to 2.4 at upslope or hilltops, and the resonance frequencies were between 3 and 10 Hz. Topographic effects can be as important as stratigraphic effects when assessing seismic amplification effects in the study area. We conclude that both topographic and stratigraphic effects should be considered in the comprehensive seismic hazard assessment of the study area or other similar mountain towns.

Can site effects be estimated with respect to a distant reference station? Performance of the spectral factorization of coda waves

SUMMARY The reliability of the ground motion simulation is highly dependent on the quality of the site response (or site effects) evaluation. There are several methods to estimate the Site Amplification Factor (SAF) corresponding to Fourier Amplitude Spectra (FAS), either by using numerical simulation tools or empirical approaches. A widely used empirical method is the ‘Standard Spectral Ratio’ (SSR) technique based on the FAS ratio of the seismic record at a target site to the corresponding one at a nearby amplification-free ‘rock’ site (reference site). The main limitation of this method lies in the availability of a reference site relatively close to the target one. In this study a SAF estimation technique at a target site in relation to a distant reference site is presented and evaluated. This technique is based on the retrieval of the minimum phase Source Time Function (mpSTF) at a pair of examined sites (target-reference), with the Spectral Factorization analysis of Coda waves (SFC) proposed by Sèbe et al. The so derived mpSTF are considered as a convolution of the actual source function, and of the SAF, so that the FAS ratio of the mpSTF, derived at one site (target) and at a distant reference site, should be an estimate of the target SAF. The latter is confirmed in this study. Under the conditions of a common STF at the examined sites and of similar coda waves excitation factor, the ratio of the FAS of the mpSTFs (target over reference site) can safely approach the actual SAF, at least when target-reference distance is up to ∼60 km and provides encouraging results at longer distances. This technique was applied at 24 sites in western Greece in relation to 4 reference sites located at varying distances from the target ones (from 0.4 to 110 km). More than 700 STFs were calculated for 89 moderate magnitude earthquakes (3.9 ≤ M ≤ 5.1) located in this high seismicity area and SAFs were determined from each pair of target-reference stations using common seismic sources. Finally, the average SAFs were computed and compared to the ones computed by other methods (e.g. SSR, GIT and HVSR) demonstrating the reliability and robustness of the proposed technique in site effect estimation.