A High-Resolution Site Amplification Map for Wellington, New Zealand

Abstract

Although earthquake site effects play a crucial role in the evaluation of local seismic hazard and associated risk, their quantification over the frequency range of interest for engineering applications still remains challenging. Mapping the local amplification at high resolution is difficult even in seismically active cities such as Wellington, New Zealand. Employing traditional methods to map amplification, such as the standard spectral ratio (SSR), is realistic only with sufficient density of strong-motion stations (SMS) across the city and the presence of a suitable rock reference station. Recently, hybrid standard spectral ratio methodologies (SSRh) have been proposed to fill in the gaps and provide estimates at much finer spatial resolution. SSRh combines traditional SSR, calculated on earthquake data between a soil reference and a rock station, with SSR computed from simultaneous ambient vibration recordings (SSRn) at a temporary location and the soil reference site within the sedimentary basin. In the last decade, over 450 single-station ambient noise measurements were undertaken across Wellington, and no collocated soil reference station is available, making the SSRh method as it stands impossible to apply. To overcome this limitation, we propose an adaptation of SSRh to capture the same basin response between a soil site and soil reference station as in the case of the synchronous ambient vibration data. We employ an additional interim step that uses the traditional SSRn between each of the soil sites and a rock reference broadband station recording synchronous long-term ambient vibration. The resulting empirical amplification model using the SSRh adaptation is in good agreement with the available SSR at SMS. Amplification factors up to 10 are present along the Centreport area, where significant damage was observed during the Mw 7.8 Kaikōura earthquake. By employing the adjusted SSRh methodology, we were able to develop a first-level high-resolution empirical site amplification model for Wellington. The approach provides an attractive solution for the evaluation of site effects across regions where a significant number of unsynchronized ambient vibration measurements are available.