Microseismic Monitoring and Denoising

摘要

Microseismic data recorded by surface arrays are often strongly contaminated by unwanted noise. This background noise makes the detection and location of small magnitude events difficult. The focus of this dissertation is to develop methods for improving the detection and location of microseismic events through multidisciplinary approaches. A method for automatic discrimination of microseismic events based on their source depths using machine learning techniques is presented. We also introduce four different methods for automatic denoising of seismic data. These methods are based on the time-frequency thresholding approach. We have improved the efficiency and performance of the thresholding-based method for seismic data that can improve detection of small events and arrival time picking resulting in increased location accuracy. All of these methods are automatic and data driven and are applied to single channel data analysis; they do not require large arrays of seismometers or coherency of arrivals across an array. Hence, these methods can be applied to every type of seismic data and they can be combined with other array based methods. Results from application of this algorithm to synthetic and real seismic data show that it holds a great promise for improving microseismic event detection.