Time reverse locating of noise source based on isotime point imaging

Abstract

In seismic exploration, a noise source is simply described as a source producing stochastic vibration continuously. The first arrival or events of a noise source are generally unpickable. Hence it's more appropriate to use the time reverse imaging (TRI) for locating noise source rather than the travel time inversion. In this paper, we introduce TRI into the noise source locating. Hybrid imaging condition and random selection of receiver positions in back propagation are used to improve the effect and reliability of imaging result. For identifying the source, the local maxima of imaging result are extracted, combining with a statistical threshold. Besides, we propose an imaging method based on isotime point: taking imaging points as sources, and calculating the travel time around them respectively; picking several points on one or more isotime lines (which are called isotime points), and using the back propagated wavefield sequences at these points to conduct imaging. Again, a threshold is used to identify sources. Combination of results of the conventional method and that from the isotime point imaging helps eliminate artifacts. Homogenous and complex models are illustrated to verify the effectiveness of the proposed method and its resistance to white noise.