Interactive seismic event recognition and its applications

Abstract

Many methods for automatic seismic event picking have been proposed over the years. All fully automatic methods lack robustness. Given that present-day machines can at best mimic human picking procedures, it is hardly surprising that they will fail when things get tough. Attenuation, sensor coupling variations, random scattering in the overburden and the competition of multiple arrivals for a given time slot all complicate the task of designing a robust picking machine. In this paper, we set out the architecture of a semi-automatic seismic event recognition system, in which robustness is achieved by allowing an observer to train the machine interactively both to accomplish a given event recognition task and to cope with variations in the nature of that task. The event-recognition task itself is accomplished by moving from conventional envelope threshold detection to phase sensitive windowed cross-correlation. The art of enhancing the precision of time picking lies in supervising the transition between the two as rising signal-to-noise levels permit. The scheme has been used on a standard PC to extract arrival times from full waveform sonic logs, reflection and tomographic in-mine borehole radar surveys, and 3D seismic reflection surveys for coal mining.