Coherent Wavefield Reconstruction Improves Event Location With Dense Seismic Arrays

We propose a novel reconstruction method for coherent wavefields recorded by dense seismic arrays. The inherent spatiotemporal coherence in collocated time series is quantified by means of the semblance norm. Using field data recorded by the 1800-station large number Seismic Survey in Oklahoma (LASSO) array and realistic simulations, we demonstrate that the proposed method can reconstruct the wavefields well and produce more coherent and regularized waveform data with high signal-to-noise ratio (SNR). We further examine the effectiveness of the reconstructed and enhanced data with stacking-based seismic location. The comparison of imaging results for two synthetic (SNR <1) and four field events (M
$_{\text {L}}~0.08$
–2.03) demonstrates the superiority of reconstructed waveforms regarding source energy focusing and imaging resolution. Polarity-uncorrected traces of reconstructed waveforms produce high-resolution source images, and the corresponding short-term average to long-term average traces yield more stable source images with lower imaging resolution, suggesting the method’s applicability to a wide range of common imaging and monitoring tasks.