Estimating VS30 at Korean Peninsular seismic observatory stations using HVSR of event records
Introduction
Time-averaged shear-wave velocity down to 30 m depth (VS30) has been used as a major site condition constraint in ground motion model (GMM) development [[1], [2], [3], [4]]. By definition, the VS30 indicates time-averaged shear wave velocity (VS) of surficial soil (i.e., surface to 30 m depth). Being closer to the surface, soil constraint becomes less so that ground motions amplify. Although VS30 only reflects VS information down to 30 m depth, it correlates to deeper soil depth if an abrupt change is not present in soil layers [5]. From the correlation, VS30 can implicitly indicate the long-period ground motion amplification occurred by deep soil column. Hence, for the GMM development, assessing VS30 on seismic stations is a key work to characterize site amplification.
It had been known that the Korean Peninsula was relatively safe from earthquake hazard. However, recently two damaging mid-magnitude earthquakes occurred at the south-east part of the peninsula during 2016–2017 have altered the recognition against earthquake hazard [6]. Hence, to provide rapid and precise information after event occurrences, the Korea Meteorological Administration (KMA) who maintains national seismic networks and announces the event information in Republic of Korea has been expanding the seismic network. As a result, 318 KMA stations are in operation to date, but only 48 stations have geophysical investigation results at this time. The KMA keeps working on additional site investigations, but some site investigations are difficult because of civil complaints and concern of national forest destruction. This difficulty requires an alternative way of site characterization. Among site parameters, the VS30 of seismic stations is important since it is used as the site parameter in the ShakeMap module that is used for intensity service by KMA. Currently, slope-based VS30 developed for world [1], [7] is used to assign VS30 for KMA stations without measured values.
Since Nogoshi [8] and Nakamura [9] introduced correlation between a peak frequency of horizontal-to-vertical spectral ratio (HVSR) from ambient vibrations and the fundamental frequency (f0) of a site, great number of studies have proved the effectiveness of HVSR on estimating f0 of sedimentary deposits [[10], [11], [12]]. Also, it has been revealed that HVSR of both seismic motions and ambient vibrations are equally effective on f0 estimation [[13], [14], [15]]. The HVSR amplitude is lower than the amplitude of empirical site amplification [[16], [17], [18], [19]]; however, peak frequencies between the HVSR and site amplification are found to be similar [20].
The f0 is correlated to the soil stiffness and depth. Thus, the VS30 and f0 have a strong correlation. Also, because f0 is correlated to the HVSR peak frequency (fp), we can use fp to find VS30. Ghofrani and Atkinson [21] examined correlation between VS30 and HVSR parameters (i.e., fp and peak amplitude, Apeak) using NGA West2 and National Research Institute for Earth Science and Disaster Resilience (NIED) data sets in Japan. Hassani and Atkinson [22] used ground motions recorded in Central and Eastern North America (CENA) for VS30 and fp relationship. Kwak and Seyhan [23] used NIED dataset and proposed two kinds of site frequencies: 1) fp correlated to VS30 and 2) lowest peak frequency correlated to the bedrock depth. These models are empirically or semi-empirically developed using local data sets, so it is necessary to develop a newly optimized model or validate existed models if the target region is changed.
Using HVSR parameters is a very attractive technique to find VS30 of seismic stations where performing geophysical investigations is limited. Seismic stations themselves provide ground motion recordings and ambient vibrations in real time so that additional effort for data acquisition is not necessary. The purpose of this study is the estimation of VS30 at the Korean seismic stations using HVSR parameters. In this study, we estimate fp and Apeak of HVSR from KMA seismic records, select a best-fitted model comparing VS30 and fp for sites with geophysical investigation, and apply the model to stations without geophysical investigation to find VS30. The VS30 estimated in this way has lower uncertainty than the current practice, which will improve the intensity service of KMA.
Section snippets
Dataset of ground motions and process
Total of 12,153 three-component seismic records for 389 events occurred during 2009–2018 with local magnitude (ML) 2.5–5.5 observed in 162 stations with accelerometers located on surface in the south part of the Korean Peninsula have been accessed and processed for HVSR estimation. Stations in two regional networks maintained by KMA and Korea Institute of Geoscience and Mineral Resources (KIGAM) are collected. To remove ground motion records with low amplitude and avoid high variance of HVSR,
Site classification group
To investigate shapes of HVSR per site class, we group VS profiles based on two site classification criteria: 1) NEHRP site classification [27] and 2) MOIS (Ministry of Interior and Safety in Republic of Korea) site classification based on the MPSS (Ministry of Public Safety and Security) report [28]. The NEHRP site classification classifies sites as A to E based on the VS30. This classification had been used in the seismic design in Republic of Korea until 2016 [29]. However, since the shallow
Models from literature
Ghofrani and Atkinson [21] (GA14) proposed a global model between VS30 and HVSR parameters (i.e., fp and Apeak) obtained from HVRS of RS. For shallow soil regions (regions including Japan, China, and Taiwan), the VS30 is predicted as:
GA14 also suggest VS30 prediction model solely based on fp for global region:
Eqs. (3), (4) are only effective if fp is greater than 1 Hz. GA14 do not suggest VS30 for fp < 1 Hz because VS30 is not
Conclusion
The objective of this study is to find VS30 using the fp from the HVSR at seismic stations in Republic of Korea. Stations with VS profiles are gathered and the relationship between fp from HVSR and VS30 is analyzed by carefully selecting fp from HVSR and comparing with existed fp-VS30 international models. The main findings of this study are:
- 1)
Sites with shallow soil layer or/and non-uniform layers have low HVSR amplitude (e.g., KAW). This induces a non-distinct peak of HVSR resulting in
Funding
This work was supported by the Hanyang University [grant numbers HY-2019-N]. Authors greatly appreciate their support. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect those of the Hanyang University.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References (35)
- et al.
Summary of the ASK14 ground motion relation for active crustal regions
Earthq Spectra
(2014) - et al.
NGA-West2 equations for predicting PGA, PGV, and 5% damped PSA for shallow crustal earthquakes
Earthq Spectra
(2014) - et al.
NGA-West2 ground motion model for the average horizontal components of PGA, PGV, and 5% damped linear acceleration response spectra
Earthq Spectra
(2014) - et al.
Update of the Chiou and Youngs NGA model for the average horizontal component of peak ground motion and response spectra
Earthq Spectra
(2014) - et al.
An NGA model for the average horizontal component of peak ground motion and response spectra
Earthq Spectra
(2008) Earthquakes in the southeast Korean peninsula: focusing on the 2016 gyeongju and the 2017 pohang earthquakes
(2018)- et al.
Topographic slope as a proxy for seismic site conditions and amplification
Bull Seismol Soc Am
(2007) - et al.
On the amplitude characteristics of microtremor (part 2)
J Seismol Soc Jpn
(1971) A method for dynamic characteristics estimation of subsurface using microtremor on the ground surface
Railway Tech Res Inst Q Rep
(1989)- et al.
Topographic site effects and HVSR. A comparison between observations and theory
Bull Seismol Soc Am
(1966)