Spectral Characteristics of Strong Ground Motion Time Series for Low to Medium Seismicity Regions with Deep Soil Atop Deep Geological Sediments ^†

Abstract

In this study, we examine the features of horizontal and vertical strong motion time series for sites with deep soil above deep geological deposits, in low-to-medium seismicity zones. New empirical regional equations were derived for spectral and peak ground acceleration attenuation based only on the strong-motion time series that were recorded in the north-western Balkans region. Results show that short-period horizontal spectral amplitudes at rock locations can be larger than those resulting from the combination of the deep soil and the deep geological deposits. However, the results also show that there is a significant increase in spectral amplitudes for larger vibration periods.

1. Introduction

Time series analysis plays a crucial role in understanding earthquake ground motion temporal patterns, trends, and dynamics, ultimately aiding in more reliable strong ground motion prediction, risk assessment, and disaster management efforts.

This study was motivated by the observation that, despite the fact that the 2004 Eurocode 8 [1] acknowledges the possible significance of deeper (geological) site surroundings, indicating in Clause 3.1.2(1) that the National Annex may create a classification scheme that takes deep geology into account, most countries, including Croatia, have not yet incorporated deep geology into their ground classification schemes.

We examined features of both horizontal and vertical strong motion for deep soil sites above deep geological deposits, in low-to-medium seismicity zones. We selected Osijek, Croatia—a city in the southern portion of the Pannonian Basin and on the right bank of the river Drava (Figure 1)—for our case study, since previous research revealed that the city’s building stock is very vulnerable [2,3,4]. Osijek’s geological sediments are up to 2.7 km deep, and the strata above geotechnical rock (the s-wave velocity larger than 800 m/s) have a thickness of 150–180 m (Figure 2).

2. Methodology

Using only the strong-motion time series (i.e., surface ground acceleration time histories) that were recorded in the region of the north-western Balkans, new empirical regional equations were derived for spectral and peak ground acceleration attenuation. A total of 436 horizontal and 218 vertical strong strong-motion acceleration time series from 112 events with magnitudes between 3 and 6.8 that were recorded in the northwest Balkans are included in the acceleration time history database that was utilized to develop the predictive equations. The proposed equations simultaneously account for the influence of deep geological conditions and local soil characteristics.

Empirical scaling coefficients were obtained via multiple linear regression analyses using the MATLAB2015a^® function "regress." As the results for vibration periods longer than 2 s would be questionable, since most of the accelerograms were high-pass filtered with a very large corner frequency (because the majority of the acceleration time series had a low signal-to-noise ratio for longer periods), only the coefficients for spectral amplitudes up to a period of 2 s were computed. Further details on the utilized data are given by Bulajić et al. [6,7]. Figure 3 shows how the empirical predictions based on the new scaling equations for horizontal ground motion [5] correlate with the pseudo-acceleration spectra recorded in the analyzed region at deep soil sites. Figure 4 shows how the empirical predictions based on the new scaling equations for vertical ground motion [8] correlate with the pseudo-acceleration spectra recorded in the analyzed region at deep soil sites.

3. Analysis of Empirical Predictions

3.1. PGA Values

Regarding the peak ground acceleration values, the results indicate that horizontal PGA values at sites with deep soil atop deep geological sediments are only 6% bigger than those seen at rock soil sites over geological rocks. The vertical PGA for the area under investigation can be roughly estimated as 0.61 of the horizontal PGA values. This is a 37% greater ratio than the one provided by Eurocode 8 (2004) for its Type 2 spectra.

3.2. Spectral Amplitudes

Empirical estimates show that short-period horizontal spectral amplitudes at rock locations can be, however, larger than those resulting from the combination of deep soil and deep geological deposits. It was also demonstrated that, at deep soil sites, there might be a 26% de-amplification of the vertical pseudo spectral acceleration (PSA) amplitudes for (vibration) periods between 0.10 s and 0.15 s and greater than 1.50 s. Consistent with other recent investigations of soft sediment non-linear behavior, this implies that dissipation of the seismic energy in deep soil sites overcomes the amplification. Furthermore, it is demonstrated that there is also a significant increase in spectral amplitudes for some vibration periods. For the horizontal motion, the maximum amplification was obtained for T = 0.5 s and is equal to 2.37. When it comes to vertical ground motion, deep soil on top of deep geological deposits showed the greatest amplification of 1.48 times for a 0.3 s vibration period when compared to the rock sites.

4. Uniform Hazard Spectra

Finally, the microzonation maps were made using the derived scaling equations [5,6,7,8]. Figure 5 shows microzonation maps for six different horizontal PSA amplitudes and four different probabilities of p in t years (analog to return periods of 95, 475, 975, and 2475 years), for the combination of the deep soil sites and the deep geological sediments [7].

The resultant Uniform Hazard Spectra (UHS) at deep soil sites and for three distinct deep geology types were compared to the ground type C Eurocode 8 spectra, scaled by the PGA values (Figure 6 and Figure 7). The highest ratios of UHS to PGA values deviate from the 2.5 factor that Eurocode 8 suggests for horizontal spectra. For UHS amplitudes at deep soil on top of deep geological strata, this ratio is 3.25 for the return period of 95 years, 3.4 for that of 475 years, 3.5 for that of 975 years, and 3.6 for that of 2475 years. This represents a 31–46% increase over the 2.5 factor that Eurocode 8 recommends. Furthermore, a notable underestimating of vertical spectra occurs when Eurocode 8 ratios of vertical to horizontal spectra are used. The horizontal to vertical PSA ratios recommended by Eurocode 8 for Ground Type C and Type 2 spectra are 2.0 to 2.3 times smaller than the empirical ratios at deep soil on top of deep geological deposits and with a (vibration) period of 0.3 s. It was demonstrated that deep geological conditions, the probability of the empirical estimates, and the source-to-site distance all had an impact on the vertical to horizontal PSA ratios for a particular soil type. Therefore, contrary to what Eurocode 8 suggests, they cannot be described by a constant that solely depends on the magnitude of an earthquake.

5. Discussion and Conclusions

For the case study area, i.e., the city of Osijek in Croatia, which is situated in a low-to-medium seismicity region and founded on deep soil sites over deep geological deposits, we analyze aspects of horizontal and vertical strong motion time series. Solely based on the strong-motion time series acquired in the northwest Balkan region, new empirical regional equations were established for spectral and peak ground acceleration attenuation. The findings indicate that, at rock locations, short-period horizontal spectral amplitudes can be higher than those arising from the combination of deep geological deposits and deep soil. Nevertheless, the outcomes also demonstrate that spectral amplitudes significantly rise with longer vibration periods.

The provided microzonation maps and UHS can be considered as a first step towards more accurate hazard estimates for the studied region, despite the limited data used for the defining of the ground motion prediction equations used for hazard calculations. The created scaling equations can easily be updated and the hazard recalculated as further data become available, increasing the reliability of the risk estimations for the city of Osijek’s building stock.