This study compares the uncertainty and bias associated with Multi-channel Analysis of Surface Waves (MASW) dispersion estimates obtained from both an impact and harmonic source. Each source was used to collect dispersion data from a common receiver array consisting of 24, 4.5-Hz geophones with a uniform spacing of 0.91 m. Source-offset distances ranging from 3- to 20-times the receiver spacing were used for each source type. The sledgehammer and vibroseis sources both produced very similar dispersion curves for frequencies greater than 20 Hz and wavelengths less than 20m. Coefficients of variation (COV) for the dispersion data in this range were generally less than 8% for the vibroseis and less than 13% for the sledgehammer. However, a significant bias developed between the mean sledgehammer and vibroseis dispersion data at frequencies less than approximately 20 Hz and wavelengths greater than approximately 20m. This bias was also accompanied by a decrease in signal-to-noise ratio (less than 10dB) and an increase in the COV's (14-54%) for the sledgehammer data. COV's for the vibroseis data remained less than 12% out to wavelengths greater than 150 m. While these results are site-specific, they reinforce the need to quantify dispersion uncertainty and data quality in a meaningful way. As additional data is gathered, meaningful estimates of dispersion uncertainty and data quality will allow recommendations to be developed for limiting depth of exploration based on a set level of phase velocity uncertainty.
展开▼
