Refining Seismic Risk: Site Characterisation of 12 Seismic Stations in Victoria

Title, Authors, and Publication

The paper, titled "Site Characterisation of 12 Seismic Monitoring Stations in SE Victoria, Australia," was authored by Reza Ebrahimi, Trevor Allen, Abraham Jones, Hadi Ghasemi, and Januka Attanayake and presented at the Australian Earthquake Engineering Society (AEES) 2023 National Conference, held in Brisbane, Queensland, Australia, from November 23–25, 2023.

Objective and Background

Understanding ground-motion characteristics is essential for accurate seismic hazard assessments, as near-surface conditions significantly influence earthquake shaking intensity. This study aims to characterize 12 seismic monitoring stations in southeastern Victoria, Australia, using various microtremor survey methods to determine site-specific shear wave velocity (Vs) profiles.

The research is particularly relevant following the 2021 MW 5.9 Woods Point earthquake, the largest instrumentally recorded event in Victoria. By assessing site response characteristics, the study contributes to refining Ground Motion Models (GMMs) used in seismic hazard assessments and infrastructure resilience planning in Australia.

Introduction

Earthquake ground-motion modeling is influenced by local site effects, which depend on the geological and geotechnical properties of the near-surface layers. To quantify these effects, the study evaluates seismic site response at 12 University of Melbourne Seismic Network (UMSN) stations that recorded the 2021 Woods Point earthquake.

This research is necessary for:

1. Improving region-specific GMMs by accounting for local site effects.
2. Benchmarking recorded earthquake data to better interpret future seismic events.
3. Developing accurate Vs30 (average shear wave velocity in the top 30m) estimates for site classification in building codes and hazard models.

Methodology

1. Microtremor Survey Techniques Modified Spatially Averaged Coherency (MSPAC) Method: Measures the coherence of surface waves across different station pairs. High-Resolution Frequency-Wavenumber (HRFK) Method: Analyzes wave propagation parameters such as phase velocity and slowness. Rayleigh-Wave Three-Component Beamforming (RTBF) Method: Estimates Rayleigh wave ellipticity, an indicator of subsurface stiffness.
2. Shear Wave Velocity Inversion Dispersion curves and ellipticity data were jointly inverted to develop 1-D shear wave velocity profiles for each site. The study used Geopsy software to extract velocity models, incorporating multimodal dispersion information.
3. Site-Specific Vs30 Estimation The study calculated Vs30 values for each station using measured shear wave velocities. Results were compared to geological-proxy Vs30 values from national databases.
4. Validation and Uncertainty Assessment Inversion results were validated by comparing: Theoretical SH transfer functions. Rayleigh wave ellipticity models. HVSR (Horizontal-to-Vertical Spectral Ratio) curves from ambient seismic noise.

Key Findings

1. Site Conditions Vary Significantly Across the 12 Stations Vs30 values ranged from 234 m/s (soft rock) to 1040 m/s (hard rock). Stations located near the coastline exhibited lower shear wave velocities, indicating softer subsurface materials.
2. Measured Vs30 Values Differ from Geological Estimates At several stations (e.g., CLIF, NARR, MARD), measured Vs30 values were up to three times lower than geological estimates, highlighting the limitations of proxy-based site classification.
3. Higher-Mode Surface Waves Improve Inversion Accuracy Incorporating Love wave dispersion and Rayleigh wave ellipticity improved depth resolution in shear wave velocity models. The approach helped refine site classifications beyond single-mode Rayleigh wave analysis.
4. Validation Confirms Reliability of Site Characterisation Comparison of observed HVSR peaks, theoretical ellipticity curves, and SH transfer functions confirmed that the estimated shear wave velocities are robust and reliable.

Conclusion

This study provides site-specific characterisation of 12 seismic monitoring stations, enabling more accurate seismic hazard assessments in southeastern Victoria. Key conclusions include:

• Site effects significantly impact recorded earthquake motions and should be accounted for in Ground Motion Models (GMMs).
• Measured Vs30 values differ from geological proxies, reinforcing the need for direct site measurements in hazard modeling.
• Microtremor survey techniques provide reliable site characterisation, essential for updating Australian seismic risk assessments.

The findings will contribute to the Australian Ground-Motion Database, supporting earthquake engineering applications and infrastructure resilience planning.

Future Work and Applications

1. Expanding Site Characterisation to More Seismic Stations Conducting additional microtremor surveys across Australia to refine seismic hazard models.
2. Developing Advanced Ground Motion Models Integrating site-specific velocity profiles into next-generation GMMs for Australia.
3. Assessing Nonlinear Site Effects in Earthquake Shaking Evaluating how high-intensity ground motions affect near-surface response.
4. Refining Building Code Site Classifications Using direct Vs30 measurements to improve site classification in Australian seismic design standards.
5. Utilizing Machine Learning for Site Response Prediction Applying AI-driven inversion techniques to streamline Vs30 estimation for future seismic studies.

?