Active–passive multichannel analysis of surface waves integration for reconnaissance-scale characterization of shallow limestone bedrock in northern Vietnam
Abstract
Reliable bedrock-depth information is essential for foundation design and seismic site characterization, yet borehole-only investigations provide sparse coverage and active-source surface-wave surveys may lack the low-frequency energy needed to resolve deeper interfaces. This study evaluates co-located active and roadside-passive multichannel analysis of surface waves at three sites in Quang Hanh Ward, Quang Ninh Province, northern Vietnam. Twenty-four 4.5 Hz geophones were deployed at 3 m spacing; active records were generated with a 10 kg sledgehammer, and 32 s passive records were recorded using traffic noise. Normalized frequency-phase-velocity images were combined and inverted to preferred one-dimensional shear-wave velocity (VS) profiles. Depth-averaged velocities and successive 5 m velocity ratios were compared with two boreholes and Standard Penetration Test data. The combined dispersion information extended from approximately 3 Hz to site-dependent upper limits of about 15–20 Hz and was interpreted to a depth of 50 m. The preferred models show a low-velocity surficial unit (VS = 164–364 m/s) over a high-velocity unit assigned to fractured limestone (approximately 1,500–2,300 m/s). The principal transition occurs at 10–15 m at Site 1 and 15–20 m at Sites 2 and 3, overlapping borehole evidence that places limestone at approximately 14–16 m. VS30 values of 513, 591, and 451 m/s place all three sites in Site Class C under the American Society of Civil Engineers 7-22/International Building Code 2024 and Ground Type B under Eurocode 8. The field example indicates that active–passive integration can extend usable bandwidth while retaining shallow resolution. Because only three lines and two non-blind boreholes were available and inversion uncertainty was not quantified, the proposed velocity-ratio minimum should be treated as a site-specific screening indicator rather than a generally validated bedrock estimator.