Detection of seismic refraction signals using a variance fractal dimension technique

Abstract

Seismic signals in deep crustal surveys are often contaminated with various types of noise, mainly caused by the low signal‐to‐noise (S/N) earth environment. A variance fractal dimension (VFD) technique is investigated and tested with real data sets for detection of seismic refraction signals from background noise. The data tested in this study were collected during the 1992 Lithoprobe Abitibi‐Grenville Transect high‐resolution refraction and wide‐angle reflection seismic experiments. The sharpness of transition features on the VFD trajectory is used as a criterion for distinguishing specific seismic phases. The window size and window interval applied in the application of VFD technique were determined using synthetic seismic data for generation of the optimum VFD trajectory. The window size of 48 samples and the window interval of 8 sample intervals were chosen to calculate the fractal dimension values and create the trajectories for detecting phases Pg, Pn, PmP, and ground roll. The VFD technique was also tested and applied for automatic detection of the first breaks in the high‐resolution seismic reflection data collected during the 1990 Lithoprobe regional and high‐resolution seismic surveys. The sharp transition features corresponding to the first arrivals in the seismic reflection data are distinct and provide us with a robust and powerful tool for separating the seismic signals from noise.