Dispersion measurement from ultra-short ambient noise using phase-weighted stacking in urban areas

Abstract

Passive surface wave methods that utilize ambient noise to investigate underground space have been widely used in densely populated urban areas. As an important step in imaging from ambient noise, dispersion measurement seriously affects the accuracy of the final inversion results. Linear stacking, the most commonly used stacking method in dispersion measurement, usually relies on the long observation time to obtain Green's functions with high signal-to-noise. However, it is sometimes difficult to ensure sufficient observation time for passive surface wave methods for exploration efficiency and environmental restrictions in urban areas, making dispersion measurement from ultra-short noise important and necessary. Considering the limitations of linear stacking, we compared the dispersion images obtained by linear stacking, phase-weighted stacking (PWS), and time-frequency phase-weighted stacking using ultra-short ambient noise to find an appropriate stacking way. Synthetic experiments show that PWS can obtain better cross-correlation functions and dispersion spectra with fewer noise segments. Two field cases further verify the reliability of applying PWS in dispersion measurement from several minutes of the continuous noise record by analyzing the phase velocities error or comparing the inversion results with the drilling data. Both synthetic experiments and field examples illustrate PWS can be used as an efficient and economical method to meet the needs of ultra-short-time imaging and real-time monitoring of urban underground structures.