New interpretation methods for rockhead determination using passive seismic surface wave data: Insights from Singapore

Accurate determination of rockhead is crucial for underground construction. Traditionally, borehole data are mainly used for this purpose. However, borehole drilling is costly, time-consuming, and sparsely distributed. Non-invasive geophysical methods, particularly those using passive seismic surface waves, have emerged as viable alternatives for geological profiling and rockhead detection. This study proposes three interpretation methods for rockhead determination using passive seismic surface wave data from Microtremor Array Measurement (MAM) and Horizontal-to-Vertical Spectral Ratio (HVSR) tests. These are: (1) the Wavelength-Normalized phase velocity (WN) method in which a nonlinear relationship between rockhead depth and wavelength is established; (2) the Statistically Determined-shear wave velocity (SD-V_s) method in which the representative V_s value for rockhead is automatically determined using a statistical method; and (3) the empirical HVSR method in which the rockhead is determined by interpreting resonant frequencies using a reliably calibrated empirical equation. These methods were implemented to determine rockhead depths at 28 locations across two distinct geological formations in Singapore, and the results were evaluated using borehole data. The WN method can determine rockhead depths accurately and reliably with minimal absolute errors (average RMSE = 3.11 m), demonstrating robust performance across both geological formations. Its advantage lies in interpreting dispersion curves alone, without the need for the inversion process. The SD-V_s method is practical in engineering practice owing to its simplicity. The empirical HVSR method reasonably determines rockhead depths with moderate accuracy, benefiting from a reliably calibrated empirical equation.