Short-term monitoring of cut slope along highway using the ambient noise method

Abstract

Landslides pose significant risks to transportation infrastructure, with cut slopes along highways being particularly vulnerable to stiffness degradation and potential failure during rainfall events. Advanced monitoring for cut slopes can support landslide risk assessment. Traditional monitoring techniques often struggle to capture precursory internal mechanical changes before surface displacement. In contrast, seismic velocity changes (dv/v) derived from ambient noise interferometry provide a direct method to trace the evolution of subsurface material stiffness. This study investigates an artificial cut slope along a highway in Fuyang District, Hangzhou, Zhejiang Province. Ambient noise data were collected using seismometers over two stages from July to September 2025. Temporal variations in subsurface seismic velocity were derived through cross-correlation of station pairs, followed by the application of the stretching technique. During the first stage, rainfall events resulted in a maximum seismic velocity reduction of 4%. In the second stage, dv/v values recovered under dry conditions but dropped by more than 1% on 13 September, indicating progressive weakening of slope materials due to prolonged wet conditions. The influence of stacking duration on the stability of these measurements was also evaluated. Narrowband filtering analysis suggests that the 4–6 Hz band was stable daily cyclic variations dominated by thermoelastic effects, while the 6–8 Hz band captured a velocity drop of approximately 3% on 27 August, which fully recovered within 24 h. This study validates the feasibility of using ambient noise interferometry for monitoring cut slopes and provides a proof-of-concept demonstration for complementary monitoring of cut slopes. Quantitative analysis of the rainfall–velocity relationship and longer-term observations are required to establish reliable early warning thresholds.