Hunting for microseismic reflections using multiplets

Method

To identify multiplets in seismograms, we adapt a method from earthquake seismology which involves identifying clear direct arrivals (master waveforms), performing normalized cross-correlations of them with the microseismic recordings taken during the treatment, and thresholding the peak correlations to identify and group arrivals into multiplets (Eisner et al., 2008; Shelly et al., 2011; Brown et al., 2008).

We assessed the peak on the average correlations across all the 36 channels in each cross-correlation record. To accommodate small differences in source locations within each multiplet, we tested two positive and two negative receiver linear moveouts and chose the moveout yielding the highest peak cross-correlation. We used the peak magnitude to decide whether the record was part of a multiplet and, if so, the peak location was used for subsequent alignment and stacking of the records in the multiplet.