Detailed site effect estimation in the presence of strong velocity reversals within a small-aperture strong-motion array in Iceland
Introduction
Iceland is the largest subaerial part of the Mid-Atlantic Ridge where the North American and Eurasian crustal plates are drifting apart with an average rate of approximately 2 cm/year (Fig. 1) [1], [2], [3]. Passing across Iceland from south to north, the onshore part of the plate boundary is shifted eastward, resulting in two transform zones: the South Iceland Seismic Zone (SISZ) in the south and the Tjörnes Fracture Zone (TFZ) in the north. The largest and most populous agricultural region in Iceland is located in the SISZ for which the seismic potential and characteristics has been well documented on the basis of historical seismicity. It is known as a region in which destructive earthquakes occur, either as strong single earthquakes or in earthquake sequences of magnitude 6–7 events over a period lasting from weeks to years. The causative faults of strong earthquakes in the SISZ occur as parallel and near vertical north-south striking faults, which is perpendicular to the underlying east-west trending plate boundary [4], [5], [6], [7], [8], [9], [10], [11], [12], [13]
Earthquake strong-motion in Iceland has been monitored over the last three decades by the Icelandic strong-motion network (ISMN) which is owned and operated by the Earthquake Engineering Research Centre of the University of Iceland. At present, the network consists of 40 free-field stations that are primarily located in the SISZ and the TFZ, along with several key strong-motion stations in urban centers and key infrastructures such as hydroelectric and geothermal powerplants, dams, hospitals, bridges etc. Additionally, the first Icelandic strong-motion array (ICEARRAY I) was deployed in 2007 in the town of Hveragerði in the SISZ. The ICEARRAY I consists of 13 strong-motion stations with interstation distances of only 50–1900 m [8], in contrast to the more typical ISMN interstation distances of 5–10 km in the SISZ [14]. During the 29 May 2008 Ölfus earthquake and its aftershock sequence, the ICEARRAY I recorded the strong-motions of the mainshock and 1705 of its aftershocks [8], [15]. The main shock recordings were characterized by intense ground accelerations of relatively short durations (5–6 s) and large amplitude near-fault velocity pulses. Despite the relatively small interstation distances of the array considerable variations of earthquake ground motion amplitudes and frequency content were observed. The geometric mean of the horizontal peak ground acceleration (PGA) varied from about 44% to 88% of the acceleration of gravity (g) and peak ground velocity (PGV) from 26 to 62 cm/s [8]. Similar variations of relative amplitudes of the recorded aftershocks have also been reported but not yet fully investigated [15].
The spatial variation in amplitude and frequency content of earthquake ground motions can be attributed to wave propagation effects and localized site effects. During recent decades, it has been recognized that propagation of seismic waves may vary significantly due to local geological and geostructural settings, even over relatively small distances [16], [17]. In general, motions recorded on sites classified as "soil" are larger in amplitude relative to those recorded on "rock" sites [18], [19]. This is due to impedance contrast where soil deposits acting as filters to incoming seismic waves and amplifying motions at certain frequencies. Consequently, site effects is a major aspect of geotechnical earthquake engineering and has a major influence on seismic hazard [e.g., [20], [21], [22], [23], [24], [25]]. It is noteworthy that in earthquake engineering practice in Iceland, site effects are generally not considered to be a key factor, presumably due to the relatively thin topsoil which is in most cases is easily removed from the uppermost competent rock (e.g., lava rock, hyaloclastite, dolerite, etc.). However, for lava rock the presence of pronounced site effects has been reported [17]. Namely, in geologically younger parts of Iceland the interplay of repeated glaciation/deglaciation and fluctuating sea levels with the primary basaltic volcanism has resulted in the geological profiles consisting of recurring layers of basaltic lavas, as well as tuff layers, often with intermediate layers of sediments or alluvium [26]. This is especially true in the SISZ where the topography is approximately flat and of low elevation, and largely covered with postglacial lava flows underlain by Quaternary sediments of mainly fluvial, glacial, and glaciofluvial origin [27]. In such cases the lack of consolidation of the sedimentary deposits between the layers of lava rock causes reversal in the velocity profile with depth, leading to significant differences in site response relative to the older bedrock and should therefore not be ignored [17]. We expect this situation to be the case for Hveragerði because based on geological and borehole information the uppermost lava rock layer (B|C in Fig. 2, ~5000 year-old,) lies on top of a softer sedimentary layer, which in turn lies on top of another lava rock layer (A in Fig. 2, ~10,000 year-old) resulting in a velocity reversal [14]. The extensive ICEARRAY I strong-motion dataset from the aftershocks of the Ölfus earthquake now provide the opportunity not only to quantify the local site effects at the array stations on such soil structure, but also the relative differences over short distances.
There are several techniques that can be used to evaluate and quantify site effects [e.g., [21], [28]] and the choice of the method is usually based on the importance and nature of the project. One of the most popular and widely used techniques to characterize site amplification is the standard spectral ratio (SSR) method [29]. The SSR is defined as the ratio of the Fourier amplitude spectra of an earthquake motion recorded on a soil site to that recorded on a selected reference site. Essentially, the SSR method for estimating site response is based on the comparison of ground motions at soil sites of interest to nearby rock site that is considered a reference motion [29], [30]. The result of the SSR method is a site-specific “amplification curve” which is a function of frequency and reveals both the “predominant frequency” of horizontal vibrations of the site, corresponding to the peak in the ratio, and its amplitude. Pragmatically, the SSR technique can only be used for cases where data are available from dense local arrays, to include a station on a reference site that has negligible site response and is located in close proximity to the soil sites of interest. Careful examination of the reference site is required for acceptable amplification estimation at the sedimentary sites [30]. However, finding a proper reference site can be challenging [30], [31], [32], and large spatial separation of the soil and rock sites may require correcting the recordings for path and finite-source effects [33]. Therefore, the horizontal to vertical spectral ratio (HVSR) method, which was introduced by Nakamura (1989), is also used herein, along with the SSR method. The HVSR method does not require recordings on a reference site, making it a more convenient approach to estimate site response. The HVSR method entails using the spectral ratio of the horizontal to the vertical component of ground motion [34]. The method, also called Nakamura´s technique, was first introduced by Nogoshi and Igarashi based on the initial studies of Kanai and Tanaka [35], [36], [37]. Since then, many investigators worldwide used this relatively easy to implement approach. The fundamental assumption of the HVSR method is that when the soil stratigraphy is comprised of horizontal layers and that the vertical component of the ground motion is free of any kind of influence related to the soil conditions at the recording site. The primary parameter obtained from the HVSR method is the predominant frequency of the soil profile, which corresponds to a peak in the HVSR. The estimate of the predominant frequency is deemed more reliable than the amplitude of the site amplification and is an indicator of a significant velocity contrast at some depth beneath the station that results in the amplification of horizontal ground motion relative to the vertical motion [24], [38], [39]. Although seismologists still debate the HVSR method's physical and theoretical bases, the approach has attracted the attention of many researchers [40]. Noticeably, the SSR and HVSR methods yield similar estimates of the predominant frequencies of soil profiles, but the amplification values determined by HVSR method are generally less than those determined using the SSR method.
In the present study the strong-motions of the aftershocks of the Ölfus earthquake recorded on the ICEARRAY I as well as recordings of microseismic noise at the array stations have been analyzed using the HVSR method. Additionally, the earthquake recordings were analyzed using the SSR method. The results are presented as frequency dependent amplification curves from which the predominant frequency and the amplitude of site amplification are estimated. The results of the different methods are compared and interpreted in terms of local geological conditions, ground motion amplitudes, source-site distances and azimuth, and earthquake parameters such as magnitude and depth. Finally, we interpret the HVSR results by considering that a vertical column of a soil structure consisting of a lava rock layer on top of a softer sedimentary layer, representing a velocity reversal with depth, can be modeled as a simple dynamic system.
Section snippets
Array strong-motion data
The 29 May 2008 Mw6.3 Ölfus earthquake occurred on two parallel, near vertical north-south striking right-lateral strike slip faults that are approximately 4.5 km apart (see Fig. 1a). While the epicenter was located on the eastern fault, ~6.5 km S-E of Hveragerði, the majority of the aftershocks occurred on the western fault that lies only 1–2 km from the town and ruptured ~2 s after the eastern fault [11], [41]. The ICEARRAY I recorded the strong-motions during the main shock and those from 1705
Results
Despite small aperture of ICEARRAY I there are noticeable differences in the HVSR results across the array. The HVSR amplification curves for all earthquake recordings for each of the twelve ICEARRAY I strong-motion stations were calculated over the frequency range of interest to this study (0.5–20 Hz), along with the geometric mean HVSR and the associated (Fig. 4). Some stations exhibit bimodal amplification curves with one mode being more dominant and of relatively larger amplitude than
Spatial distribution of HVSR characteristics
To provide further insight into the HVSR results, the spatial distribution of fundamental frequencies and maximum amplitudes of the amplification curves based on earthquake and microseismic data, respectively, are shown as surface plots in Fig. 7 over the area confined by the array stations. While the slight differences between the HVSR characteristics between stations may become more pronounced when presented spatially in this way, both methods show the same general trend that from north to
Summary and conclusions
The earthquake strong-motion during the Ölfus earthquake on 29 May 2008 and 1705 aftershocks were recorded on the first small-aperture urban strong-motion array in Iceland, the ICEARRAY I. The array consists of 12 stations in the town of Hveragerði, located in the western part of the South Iceland Seismic Zone, and makes possible detailed studies of many aspects of engineering seismology ranging from source to site effects. In this vein, the first comprehensive study of localized site
Acknowledgements
This work was supported by the Icelandic Centre for Research (Grant of Excellence no. 141261-051), the Icelandic Catastrophe Insurance (Grant no. S112-2013) and the University of Iceland Research Fund. Christian I. Olivera was supported by the Leifur Eiriksson Foundation and Virginia Tech, and the Earthquake Engineering Research Centre of the University of Iceland provided him material support. The seismographs used in this study were rented from the LOKI instrument bank at the Icelandic
References (65)
- et al.
Geometry and segmentation mechanisms of the surface traces associated with the 1912 Selsund Earthquake, Southern Iceland
Tectonophysics
(2005) Earthquakes and present-day tectonism in Iceland
Tectonophysics
(1991)- et al.
The Mw6.3 Ölfus earthquake at 15:45 UTC on 29 May 2008 in South Iceland: ICEARRAY strong-motion recordings
Soil Dyn Earthq Eng
(2009) - et al.
Strain release and strain build-up in the South Iceland seismic zone
Tectonophysics
(1988) - et al.
Site amplification in lava rock on soft sediments
Soil Dyn Earthq Eng
(2002) Local site effects on weak and strong ground motion
Tectonophysics
(1993)- et al.
The SIL data acquisition system-at present and beyond year 2000
Phys Earth Planet Inter
(1999) ModelHVSR—A Matlabs tool to model horizontal-to-vertical spectral ratio of ambient noise
Comput Geosci
(2008)Plate boundaries, rifts and transforms in Iceland
Jökull
(2008)Current plate motions
Geophys J Int
(1990)
Effect of recent revisions to the geomagnetic reversal time scale on estimates of current plate motions
Geophys Res Lett
The 1912 Iceland earthquake rupture: growth and development of a nascent transform system
Bull Seism Soc Am
Seismicity pattern in the South Iceland seismic zone
Earthq Predict
A fast and efficient simulation of the far-fault and near-fault earthquake ground motions associated with the June 17 and 21, 2000, earthquakes in South Iceland
J Earthq Eng
Triggered fault slip on June 17, 2000 on the Reykjanes Peninsula, SW-Iceland captured by radar interferometry
Geophys Res Lett
A note on the Mw 6.3 earthquake in Iceland on 29 May 2008 at 15:45 UTC
Bull Earthq Eng
ICEARRAY: the first small-aperture, strong-motion array in Iceland
J Seism
Site effect evaluation in the basin of Santiago de Chile using ambient noise measurements
Geophys J Int
Microzonation of the city of Basel
J Seism
Are microtremors useful in site response evaluation?
Bull Seism Soc Am
The relation between site amplification factor and surficial geology in central California
Bull Seism Soc Am
Effects of surface topography on ground shaking prediction: implications for seismic hazard analysis and recommendations for seismic design
Geophys J Int
Site response as a function of near-surface geology in the south iceland seismic zone
Nat Hazards
Can site responsebe predicted?
J Earthq Eng
Effects of local geology on ground motion near San Francisco Bay
Bull Seism Soc Am
What is a reference site?
Bull Seism Soc Am
Cited by (20)
Vertical seismic ground shaking in the volcanic areas of Italy: prediction equations and PSHA examples
2023, Soil Dynamics and Earthquake EngineeringFrequency-dependent site amplification functions for key geological units in Iceland from a Bayesian hierarchical model for earthquake strong-motions
2023, Soil Dynamics and Earthquake EngineeringHigh spatial-resolution loss estimation using dense array strong-motion near-fault records. Case study for Hveragerði and the M<inf>w</inf>6.3 Ölfus earthquake, South Iceland
2022, International Journal of Disaster Risk ReductionBayesian inference of a physical seismological model for earthquake strong-motion in south Iceland
2020, Soil Dynamics and Earthquake EngineeringIntegrated use of ambient vibrations and geological methods for seismic microzonation
2019, Journal of Applied Geophysics