Adaptive Threshold Based Shearlet Transform Noise Attenuation Method for Surface Microseismic Data

Surface microseismic data is more informative than downhole data due to the potentially larger receiver aperture and higher number of receivers. However, the presence of strong random noise introduced from the acquisition environment decreases the signal-to-noise ratio (SNR) in the surface microseismic data, which brings great difficulty to identify and locate the microseismic events. Thus, noise attenuation is a crucial step in the processing of surface microseismic data. Shearlet transform is a new multiscale transform which can adaptively capture the geometrical characteristic of multidimensional signals and represent signals containing edges optimally. In this paper, a random noise attenuation technique for surface microseismic data based on shearlet transform by using an efficient adaptive threshold is proposed. The threshold changes adaptively according to the energy distribution of each decomposion direction. Usually, the energy from signal’s directions is much higher than that from directions of the noise.

Therefore, we apply a small threshold to the coefficients associated with signal’s directions and a large threshold is chosen for directions of the noise to achieve a better denoising result. Experimental results show that the proposed algorithm significantly improves the SNR of the microseismic data and effectively preserves the signal of interest.