Seismic site response analysis of soil sites during the Christchurch earthquakes

Abstract

The seismic response of soils near the earth’s surface can have a significant effect on the nature of ground shaking over a region. This is because of the fact that near-surface soil conditions respond significantly different to soils at depth (which have higher overburden pressures) and their geotechnical properties can also vary appreciably over very short distances as a result of various dominant geological processes. This project focused on better understanding the seismic response of soils based on investigation of the shallow and deep properties of soils in the Canterbury region, and empirical analysis of the ground motions observed in the 2010-2011 Canterbury earthquakes with a particular emphasis on the possible role of site response effects.

Active and passive surface wave analyses were conducted to determine the deep properties of soils at 22 location in the Canterbury region. The active surface wave analyses utilized “T-Rex”, the world’s largest tri-axial vibration truck, which was transported to New Zealand from Austin, Texas via the support of the NZ Earthquake Commission and the US National Science Foundation. These analyses were able, for the first time, to quantify the stiffness of the significant depth of soft soils below Christchurch. Such a quantitative estimate can be utilized in various numerical simulation methods to better understand their significance in producing the observed ground motions.

Empirical analyses of the ground motions recorded at the same location in multiple earthquakes were used to understand the proportion of ground motion variability which is due to “systematic site effects”, which was found to be approximately 40%. This result has significant implications for placing impetus on advanced characterization of site conditions to better understand the future ground motion hazard at specific locations.

Numerical “site response analyses” are on-going to attempt to provide a direct physical link between the ground conditions which have been experimentally observed, and the systematic nature of the observed ground motions at various locations, which were the two aforementioned aspects of the current research.

---

Technical Abstract

This project focused on better understanding the seismic response of soils based on investigation of the shallow and deep properties of soils in the Canterbury region, and empirical analysis of the ground motions observed in the 2010-2011 Canterbury earthquakes with a particular emphasis on the possible role of site response effects.

Our research team conducted deep Vs profiling at 22 strategic locations throughout Christchurch with the goal of developing a seismic velocity model for the Canterbury basin. The deep Vs profiling was performed using a combination of large active-source and ambient-wavefield surface wave testing. At each location, the following data was gathered: (a) active-source (sledge hammer) surface wave data collected with a linear array composed of 48 4.5-Hz vertical geophones equally spaced at 2 m (94-m long array) with source offset distances of 5, 10, 20 and 40 m; (b) active-source (NEES@UTexas T-Rex vibroseis) surface wave data collected with a linear array composed of up to 24 1-Hz vertical geophones equally spaced at 10 m (up to 230-m long array) with source offset distances of 20, 40 and 80 m; and (c) ambient-wavefield (microtremor) surface wave data collected with circular arrays (60-, 200- and 400-m diameter) composed of 10 broadband seismometers.

The active-source surface wave data from both the sledge hammer and T-Rex sources was processed using the frequency domain beamformer method. The ambient-wavefield surface wave data was processed using both the high-resolution f-k method and the spatial autocorrelation (SPAC) method. The dispersion estimates from the active-source and ambient-wavefield data at each site were combined to produce an inter-method composite dispersion curve with associated data uncertainty bounds. After which, the open-source software package Geopsy was used to perform a multi-mode, joint inversion of the dispersion data, the SPAC curves and the peak frequency of the mean horizontal-to-vertical (H/V) spectral ratio curve for the site. Rather than providing a single, deterministic Vs profile for each site, these inversions have been used to provide a suite of profiles that fit the experimental data equally well, given the estimated data uncertainty. While results are still preliminary, this work has resulted in Vs profiles that extend a minimum of several hundred meters and a maximum of greater than 1km below the ground surface. This information, coupled with available seismic reflection, refraction, and travel time tomography datasets will be used to develop a numerical model for the region that will be used in kinematic broadband simulations with seismic effective stress analyses of surficial soils. Such physically-based simulations will enable further fundamental insight into the mechanics of the observed ground motion during the 2010-2011earthquakes, and enhance our ability to predict such effects in future earthquakes worldwide.

This research has more than tripled the available comparisons between large active-source and ambient-wavefield surface wave methods utilized for deep Vs profiling. These comparisons are needed before confidence in utilizing passive-wavefield methods independently can be achieved. Generally, the dispersion estimates from the large active-source (T-Rex vibroseis) and ambient-wavefield data agreed very well at most sites. However, at approximately 20-25% of the sites the dispersion estimates obtained from the large active-source testing yielded a different interpretation of the dispersive trend than those obtained from ambient-wavefield testing alone. These variations in dispersive trend occurred over band-limited frequency/wavelength ranges, and are believed to be caused by mode jumps in the ambient-wavefield dispersion estimates. The data is still being examined to determine why.

Finally, as a result of a high-density of strong motion instruments in the Canterbury region a significant number of high amplitude near-source ground motion have been recorded. Of particular note are the large number of strong motions which have been recorded at the same location over these multiple events. Such a relatively unique ground motion dataset allows for the opportunity to directly examine systematic and repeatable ground motion phenomena which can be attributed to site effects. Systematic effects in the Canterbury earthquakes are examined within the non-ergodic empirical ground motion prediction framework. It was observed that the location-to-location residual (i.e. systematic feature of the between-event residuals) had values close to zero for short-to-moderate vibration periods, but became increasingly positive for T>1s, likely indicative of important Canterbury-specific sedimentary basin and near-surface soil effects which are not adequately accounted for in the Bradley (2010) GMPE. On the basis of the similar site-to-site residuals, 15 of the 20 stations were adequately grouped into four sub-regions, while the remaining 5 stations did not fit any of these general sub-regions. The grouping of sites into sub-regions allows the possibility of non-ergodic ground motion prediction over sub-regions of Canterbury, rather than site-specific predictions only at strong motion stations. Examination of the standard deviations in the residuals illustrates that, on average, the non-ergodic standard deviation is 65-85% of the ergodic standard deviation of the Bradley (2010) model. However, on a site-by-site basis this percentage can easily vary by 20%.

 

References for final report:

Bradley, BA, 2012 - Ground motions observed in the Darfield and Christchurch earthquakes and the importance of local site response effects, New Zealand Journal of Geology and Geophysics, 55, 279-286

Bradley, BA, 2013 - Systematic ground motion observations in the Canterbury earthquakes and region-specific non-ergodic empirical ground motion modelling, Earthquake Spectra (submitted)

 

Conference Papers:

Cox, B, Bradley, BA, Wotherspoon, L Stokoe II, K, Cubrinovski, M, Teague, D, Wood, C, 2013 - Deep vs profiling for dynamic characterisation of Christchurch, New Zealand: Towards reliably merging large active source and ambient-wavefield surface wave methods, International Conference on Geotechnical Earthquake Engineering: From case history to practice. In honour of Prof Kenji Ishihara: Istanbul, Turkey