A V_S30 Map for New Zealand Based on Geologic and Terrain Proxy Variables and Field Measurements

Abstract

A time-averaged 30-m-depth shear wave velocity ( V S30) map is developed for New Zealand as a weighted combination of a geology-based and a terrain-based model. A Bayesian updating process allows local V S30 measurements to control model estimates where data exist and uses model estimates developed for other parts of the world where local data are sparse or nonexistent. Geostatistical interpolation is performed on the geology-based and terrain-based models using local V S30 measurements to constrain the model in the vicinity of data. Conventional regression kriging is compared with a flexible multivariate normal (MVN) approach that allows for arbitrary assumptions regarding measurement uncertainty at each data location. A modification to the covariance structure in the MVN application allows for more realistic estimates by reducing undesirable extrapolation across geologic boundaries. The results of kriging and MVN approaches are compared. The geology-based and terrain-based MVN models are combined to produce a final model suitable for engineering applications. The 100-m resolution map outputs are publicly available.

Ground motion simulation of hypothetical earthquakes in the upper North Island of New Zealand

“…Basin velocity properties were defined as linearly increasing functions of depth, approximately 2 to 3 times smaller than basement velocities at the interface. A spatial V S30 correction was applied using the map of Foster et al (2019).…”

Ground motion simulations of Hope fault earthquakes

“…The high frequency (HF, f > 1Hz) portion of the methodology uses a simplified physics-based approach which constructs a time-series from the summation of multiple stochastic source sub-faults and a simple 1D velocity model, to which a V s30 -based site amplification is then applied [29]. For this study, the New Zealand-specific V s30 model of Foster et al [30] was adopted. The LF and HF simulations are then merged [5] to form a site-specific broadband ( f = 0 − 50Hz) time-series, from which spectral ordinates over the range T = 0 − 10s can be reliably derived [31].…”

“…(PGA for the reference simulation can be viewed in the appendix). Figure 7 presents empirical estimates for PGV, D s5−95 , pSA(0.5s) and pSA(3.0s) using the Afshari and Stewart [39] (for D s5−95 ) and Bradley [37] (for PGV, pSA(0.5s) and pSA(3.0s)) models (in combination with the Foster et al [30] New Zealand V s30 model and correlations to derive other necessary inputs, e.g. Z 1.0 ).…”

“…As the empirical models vary primarily with source-to-site distance (R rup ) for a given rupture (in a similar way to the HF simulation methodology), the predicted intensity measures from empirical models exhibit a near-elliptical variation about the rupture plane. The principal reason for the localised deviations from a simple source-to-site distance decay is the local site response associated with the spatial variation in V s30 (via the New Zealand-specific model of Foster et al [30] as noted in Methodology). This is in contrast to simulation-derived intensity measures (see Figure 6), that exhibit a more complex spatial variation.…”

Multi-hazard analysis and mapping of coastal Tauranga in support of resilience planning

“…This probabilistic seismic hazard analysis (PSHA) adopted a model of soil conditions represented by Vs30 seismic shear wave velocity [8] which reflected the continuous geotechnical variation across the city. An earthquake rupture forecast was developed, and the seismic hazards then calculated in terms of peak ground acceleration (PGA) which was spatially mapped for return periods of 25, 100, 500, 1,000 and 3,030 years.…”