Impact of adaptive filtering-based noise reduction on the quality of f-c images in MASW

This study investigates the effects of adaptive filtering, based on the least mean square algorithm, on the quality of multichannel analysis of surface waves (MASW) data and frequency-phase velocity (f-c) images. Specifically, the adaptive filtering method aims to decrease the noise and improve the determination of fundamental modes in f-c images. Both sample and field seismic data with varying noise levels were evaluated using adaptive filtering, resulting in significant improvements in f-c image quality. The filtered data displayed a wider frequency range of f-c images, and reduced noise compared to the original noisy data. Based on these findings, adaptive filtering can be considered an effective tool for enhancing the MASW data, particularly in environments with nonstationary noise, by improving the reliability of dispersion curve extraction from f-c images.