Fast and novel microseismic detection using time-frequency analysis

An automatic algorithm for fast and robust detection of microseismic events recorded on surface arrays is presented based upon incorporating time-frequency thresholding and image processing techniques. To build a pseudocharacteristic function, time-frequency coefficients of the seismic trace are thresholded and resulting spectrums are transferred into histograms to detect high-energy arrivals associated seismic events. The main novelty of this methodology is incorporating a simple but powerful thresholding scheme that does not rely on any prior knowledge about noise level and an image processing step on the resulting spectrum that makes this method fast and adaptive. This approach is well suited for real time processing and detection of noisy microearthquakes