Automatic passive data selection in time domain for imaging near-surface surface waves

Abstract

The passive surface wave method, as a valid supplement to the active survey method for the deeper imaging, has gained growing attentions from geophysicists and civil engineering communities, particularly in urban applications. The multi-channel analysis of passive surface waves (MAPS), as a newly developed method, combines seismic interferometry and multi-channel analysis of surface waves and shows high potential in extracting high-frequency surface-wave energy in urban environment. The unknown source-distribution of ambient noise and the complicated anthropogenic environment, however, make it difficult to obtain high-quality dispersion images of passive surface waves. We proposed a data selection technique in time domain for selective stacking of cross-correlations and applied it to improve the MAPS method. The proposed technique sorts the time segments with the defined acausal-to-causal ratio factor, and automatically detects the optimum time segments with the highest signal-to-noise ratio (SNR). Three real-world applications have demonstrated the feasibility of the proposed data selection technique. The results indicated that our method is able to preserve the time segments with coherent signals, and reject the time segments polluted by non-stationary noise sources and/or spatial aliasing. Compared with the conventional MAPS method, the improved one can broaden interested frequency band of surface waves and deblur the measured image of dispersion energy. Finally, we discussed the influence of velocity window parameter for SNR calculation on the performance of the proposed technique, and suggested that the velocity window should be appropriately wide to guarantee the accuracy of surface-wave imaging.