Topographic Amplification of Ground Motions in Mt. Pleasant, Christchurch, New Zealand

Session: Near-Surface Effects: Advances in Site Response Estimation and Its Applications
Type: Oral
Date: 4/28/2020
Time: 03:15 PM
Room: 110 + 140
Description: 

Many historic earthquakes have shown that ground motions are amplified near the top of hills and ridges. Studies have also shown that Rayleigh waves generated at slope surfaces contribute to large vertical motion and ground rocking near the crest. This phenomenon is known as topographic amplification, which is often associated with the concentrated damage pattern observed in hilly areas during large earthquakes.

It is well understood that the interference of seismic waves caused by surface topography and geologic layering can substantially alter the ground motion characteristics in a certain wavelength range related to the size of the topographic feature. However, many observations and experimental data have reported amplification factors far exceeding the predictions, often attributed to the inadequacy of the simplified models to fully reflect the complexity of the phenomena.

This paper presents a case study of topographic amplification in the suburb of Mt. Pleasant in Christchurch, New Zealand, where we deployed an array of nine seismometers. This array recorded several regional earthquakes in March 2017, which were examined using the standard spectral ratio (SSR) technique and the horizontal-to-vertical spectral ratio (HVSR) method. Unlike many previous studies, the amplification of ground motions demonstrated by the recorded data was generally consistent with previous theoretical studies.

We also simulated the dynamic response of a selected 2D cross-section with plane waves of various incidence angles using a finite difference program FLAC. The results of simulations were encouragingly consistent with the experimental results, which demonstrate the possibility of predicting the topographic effects, given that high-resolution elevation and velocity models are available. This study also highlights the need for further studies to examine the performance of predictive models in comparison with well-characterized and well-instrumented case studies.

Presenting Author: Seokho Jeong

Authors