We implement a technique to suppress surface-wave noise in microseismic data based upon the distinct characteristics that microseismic signals and noise display in the − p domain. In order to determine an estimated orientation that best accounts for the polarity pattern of any microseismic events with double-couple source mechanisms, we employ a scanning approach similar to semblance analysis, to test all possible double-couple orientations. We then correct the polarity of the data traces according to this pattern, prior to conducting signal-noise separation in the − p domain. We apply this noise suppression technique to microseismic data recorded by a star-shaped array, over a Bakken Shale reservoir during a hydraulic fracturing process. Our technique significantly improves the signal-to-noise ratios of the microseismic events. We illustrate that the enhancement in signal-to-noise ratio also results in improved imaging of the microseismic hypocenter.