Multiscale matched-field processing for noise-source localization in exploration geophysics

In the study, we investigate the use of ambient-noise data to locate microseismic sources at the exploration scale. We develop a multiscale matched-field processing (MFP) approach to localize seismic sources at frequencies below 10 Hz. An application to an actual data set acquired over a hydrocarbon field is presented to determine the reliability of the MFP procedure. The data used were continuously recorded over five days, at a total of 397 stations on a 1-km-per-side square seismic network. The MFP results show: (1) a dominant and stable surface source associated with human activities (road and exploration platforms) around the reservoir; and (2) weaker sources at depth below the seismic network that are related to the injection/extraction process.