• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.42 no.1
• /
• pp.23-33
• /
• 2022

The horizontal-to-vertical spectral ratio (HVSR) of ambient noise is widely used to identify the resonant frequency of a site. The frequency at the largest HVSR is regarded as the resonant frequency. The source of ambient noise is impossible to identify and control. Therefore, obtaining reliable HVSR of ambient noise requires sufficient measurement time and absence of near-field vibration. In this study, we investigated the minimum stabilization time required for a portable seismometer and the effect of the distance between the seismometer and artificial vibration on HVSR estimation. In the case of a soil site, the HVSR was stabilized after 5 minutes after sensor installation. In the case of a rock site, stabilization required more than an hour. Human-footsteps within 10 m of the seismometer strongly influenced the HVSR for the soil site. These results provide a field guideline when measuring ambient noise for HVSR.

• Journal of the Korean earth science society
• /
• v.41 no.6
• /
• pp.658-669
• /
• 2020

The horizontal-to-vertical spectral ratio (HVSR) analysis is conducted to estimate the site amplification effect and the thickness of the sedimentary layer beneath the measurement site. We investigated the effects of meteorological variations (wind and precipitation rate) and sensor burial depths on HVSR analysis. The HVSR results were unstable when seismographs were exposed on the ground. The HVSR results of ambient noise data measured under strong winds were also unstable. It is recommended to measure the ambient noise at wind speeds of <3 m s-1. Stable HVSR results were obtained when seismographs were buried, regardless of the precipitation rates. The results of this study provide the best observations and optimal weather conditions required to acquire accurate and reliable HVSR results.

• Journal of the Korean earth science society
• /
• v.43 no.1
• /
• pp.188-198
• /
• 2022

Some deep faults do not reach the ground surface and are seldom recognized. Gokgang Fault area in the east of the Heunghae area of the Pohang basin has been selected to confirm the feasibility of the Horizontal-to-Vertical Spectral Ratio (HVSR) approach to identify blind faults. Densely spaced microtremor data have been acquired along two lines in the study area and processed to obtain resonance frequencies. An empirical relationship between the resonance frequency and the bedrock depth was proposed using borehole data available in the study area. Resonance frequencies along two lines were then converted to bedrock depths. The resulting depth profiles show significant lateral variations in the bedrock depth. As expected, considerable variation in the resonance frequency is observed near the Gokgang fault. The depth profiles also present additional significant variations in the resonance frequencies and the bedrock depths. The feature is presumably related to a blind fault that is previously unknown. Therefore, this case study confirms the feasibility of the HVSR technique to identify faults otherwise not recognized on the surface.

• Journal of the Korean earth science society
• /
• v.44 no.6
• /
• pp.594-610
• /
• 2023

The horizontal-to-vertical spectral ratio (HVSR) has been widely applied to evaluate ground characteristics such as site response and thickness of the soft sedimentary layer on top of the bedrock via dominant frequencies and amplification factors of microtremors. Eight seismic stations were selected to investigate the HVSR results at the surface and at varying depths, and their variations due to wind speeds. These stations are equipped with seismic sensors on the surface and downhole(s) at depths. The borehole data analysis reveals that the geological condition at burial depth influences the HVSR results. Their dominant frequencies indicate the entire thickness of the soft layer, not the thickness to the bottom or top of the soft sedimentary layer from the seismometer burial depth. Analysis of the background noise observed at the surface showed that the resonance frequency estimation varied with wind speed changes. In the studied cases, the background noise observed in the sedimentary layer at depths of 20 to 66 meters yielded stable and consistent resonance frequency estimation regardless of wind speed fluctuations. The results of the seismic sensors buried deeper than 100 meters are unstable. The result indicates that the background noise from the buried seismometer at shallow depths (~0.3 m) under light wind conditions (wind speeds less than 3 m/s) is sufficient to achieve the purpose of the HVSR analysis.

• Journal of the Korean earth science society
• /
• v.31 no.6
• /
• pp.541-554
• /
• 2010

Fifteen min-microtremor data sets were collected at 136 sites from a coastal area of Kunsan and 117 sites from an inland area of Jeonju located in SW Korea, and were analyzed for the HVSR (Horizontal to Vertical Spectral Ratio) of the sites. The microtremor spectra of the coastal area have stronger energy in the lower frequency range from 1-6 Hz than those of the coastal area. This result can be attributed to the effect of the waves and tides in the Keum river and the Yellow sea. Twenty four hours of measurement of the microtremor indicated that the microtremor spectrum correlates with the human activities, but the microtremor HVSR peak was observed consistently at the characteristic frequency for the site. The HVSR peaks were grouped into 4 types -"single peak", "double peak", "broad peak" or "no peak"- based on their shapes. More than 90% of the data sets exhibit peak frequencies ($F_0$) which can be easily identified. The distribution of $F_0$ reveals a close relationship with the topography and local geology of the areas, exhibiting high F0s in the hillside areas and low $F_0s$ in the reclaimed land area. While the amplitudes of microtremor HVSR peak frequencies are less than 4 in the downstream of the inland area, those of the recently reclaimed land in the coastal area are extremely high (more than 10). The results of this study indicate that detailed HVSR studies are essential for the earthquake hazard reduction of reclaimed lands.

• The Journal of Engineering Geology
• /
• v.31 no.2
• /
• pp.211-222
• /
• 2021

The magnitude 5.4 earthquake occurred in 2017 near 7.5 km north of Buk-gu, Pohang. In order to find out the characteristics of microtremor in Buk-gu, Pohang, Gyeongsangbuk-do, where earthquake damage occurred significantly, a total of 39 points were acquired to analyze the horizontal to vertical spectral ratio (HVSR). Microtremor vary from region to region, so the resonant frequency of the region is obtained by examining the microtremor. For Thickness analysis, we analyze the structure and properties of shear wave velocity (Vs) up to the underlying rock to compare resonance frequencies to match the horizontal to vertical spectral ratio (HVSR) analysis technique against nearby boring data. Using F₀ = V_s/4H with a resonance frequency of alluvium is 1.3 ± 0.07 Hz and a resonance frequency of Yeonil group is 0.69 ± 0.22 Hz, the alluvium thickness was found to be 26~30 m and the Yeonil group thickness was 170~250 m.

• 한국지구물리탐사학회:학술대회논문집
• /
• 2010.10a
• /
• pp.117-120
• /
• 2010

• Journal of the Korean earth science society
• /
• v.23 no.4
• /
• pp.380-392
• /
• 2002

The purpose of the present study is to evaluate the usage of microtremor in estimation of subsurface structure and ground response to strong ground motion. To accomplish the purpose, the current status of microtremor study are reviewed and microtremors recorded at several stations are analysed. First of all, the stability of microtremor is examined through the analysis of microtremors recorded for 80 seconds per hour during the time from 10 p.m. to 6 a.m. for eight hours at night time. It is found that the shape of microtremor spectra of low frequency below 10Hz is approximately invariable with time and the spectra contain informations about subsurface structure. The subsurface structures estimated from the predominant frequency determined from the recorded microtremors are compared with the known ones from geophysical surveys at several stations in Kyungju. The comparison of structures shows rough agreements at most stations. Horizontal to vertical spectral ratio(HVSR) technique for microtremor has been proposed as an indirect method to determine ground response to strong ground motion. The HVSR for microtremors recorded in Kyungju is calculated and compared with theoretical transfer function calculated from the known structures. The comparison shows rough coincidence of the peak frequency of spectra between them.

• Geophysics and Geophysical Exploration
• /
• v.12 no.1
• /
• pp.132-142
• /
• 2009

The Tamar rift valley runs through the City of Launceston, Tasmania. Damage has occurred to city buildings due to earthquake activity in Bass Strait. The presence of the ancient valley, the Tamar valley, in-filled with soft sediments that vary rapidly in thickness from 0 to 250mover a few hundreds metres, is thought to induce a 2D resonance pattern, amplifying the surface motions over the valley and in Launceston. Spatially averaged coherency (SPAC), frequency-wavenumber (FK) and horizontal to vertical spectrum ratio (HVSR) microtremor survey methods are combined to identify and characterise site effects over the Tamar valley. Passive seismic array measurements acquired at seven selected sites were analysed with SPAC to estimate shear wave velocity (slowness) depth profiles. SPAC was then combined with HVSR to improve the resolution of these profiles in the sediments to an approximate depth of 125 m. Results show that sediments thicknesses vary significantly throughout Launceston. The top layer is composed of as much as 20m of very soft Quaternary alluvial sediments with a velocity from 50 m/s to 125 m/s. Shear-wave velocities in the deeper Tertiary sediment fill of the Tamar valley, with thicknesses from 0 to 250m vary from 400 m/s to 750 m/s. Results obtained using SPAC are presented at two selected sites (GUN and KPK) that agree well with dispersion curves interpreted with FK analysis. FK interpretation is, however, limited to a narrower range of frequencies than SPAC and seems to overestimate the shear wave velocity at lower frequencies. Observed HVSR are also compared with the results obtained by SPAC, assuming a layered earth model, and provide additional constraints on the shear wave slowness profiles at these sites. The combined SPAC and HVSR analysis confirms the hypothesis of a layered geology at the GUN site and indicates the presence of a 2D resonance pattern across the Tamar valley at the KPK site.

• Geomechanics and Engineering
• /
• v.37 no.6
• /
• pp.539-554
• /
• 2024

This study explores development of prediction model for seismic site classification through the integration of machine learning techniques with horizontal-to-vertical spectral ratio (HVSR) methodologies. To improve model accuracy, the research employs outlier detection methods and, synthetic minority over-sampling technique (SMOTE) for data balance, and evaluates using seven machine learning models using seismic data from KiK-net. Notably, light gradient boosting method (LGBM), gradient boosting, and decision tree models exhibit improved performance when coupled with SMOTE, while Multiple linear regression (MLR) and Support vector machine (SVM) models show reduced efficacy. Outlier detection techniques significantly enhance accuracy, particularly for LGBM, gradient boosting, and voting boosting. The ensemble of LGBM with the isolation forest and SMOTE achieves the highest accuracy of 0.91, with LGBM and local outlier factor yielding the highest F1-score of 0.79. Consistently outperforming other models, LGBM proves most efficient for seismic site classification when supported by appropriate preprocessing procedures. These findings show the significance of outlier detection and data balancing for precise seismic soil classification prediction, offering insights and highlighting the potential of machine learning in optimizing site classification accuracy.