Combining Earthquake Ground Motion and Ambient Vibration Recordings to Evaluate a Local High-Resolution Amplification Model—Insight From the Lucerne Area, Switzerland

• 出版年: Frontiers in Earth Science
• 主要な著者: Paulina Janusz, Vincent Perron, Christoph Knellwolf, Donat Fäh
• フォーマット: 論文
• 言語: 英語
• 出版事項: Frontiers Media S.A. 2022-05-01
• 主題: seismic site effects; seismic hazard; urban areas; microzonation; ambient vibration; earthquake ground motion
• オンライン･アクセス: https://www.frontiersin.org/articles/10.3389/feart.2022.885724/full

Amplification factors are often estimated using empirical methods based on earthquake ground motion; however, especially in low-seismicity urban areas, recording a statistically representative number of high-quality signals may take years. Hence, the attempts to use ambient vibration instead have progressed. This includes the development of the hybrid site-to-reference spectral ratio (SSRh) method that combines earthquake and ambient vibration recordings. We applied the method in the Lucerne area in central Switzerland that is characterized by low-to-moderate seismicity but was struck by several strong earthquakes in historical times (i.e., Mw 5.9 in 1,601) and is located in a glacial basin filled with unconsolidated deposits prone to significant amplification. To develop the high-resolution local site amplification model for the city of Lucerne using the SSRh method, we took advantage of a small seismic monitoring network installed in the Lucerne area in total for about a year and the stations of the Swiss Strong Motion Network (SSMNet). In addition, we performed two extensive surveys to record ambient vibrations and used dozens of measurements performed in the area since 2001. The resulting amplification model referring to the Swiss reference bedrock conditions indicates high-amplification factors (up to 10-fold) for a broad range of frequencies. The model is consistent with geological data and site response proxies such as f0 values. The direct comparison of our results with the SSR amplification functions for several sites shows good agreement. However, the model is characterized by high uncertainty and influenced by daily variation of the noise wavefield, as well as the spatial distribution of the stations of the seismic network. We also discussed the extent of the applicability of the method, concluding that the main factor influencing its performance is not the distance but the similarity of the site condition between the stations.