Complex stratigraphy and topology are challenges for conventional hypocentre location. A method is developed which may be used to determine the hypocentre of seismic and microseismic events located within regions of variable topology and stratigraphy. The program requires a velocity model (supplied as a 3-D grid of values), receiver locations (either on the surface or buried), and a set of observed first-arrival pick times. The velocity field is transformed into a set of travel-time cubes, one per receiver location, computed using a fast-marching eikonal equation solver. Candidate hypocentre locations may be anywhere within the search space defined by the velocity field. Each candidate hypocentre location is evaluated according to a cost function. The simplest cost function is the square of the cumulative errors between observed first-break arrivals and the predicted arrival times based on the velocity field. The global minimum for all candidate locations is considered the best estimate of the hypocentre. The global minimum is determined using a pattern search technique. More complex cost functions may be designed which make use of the seismic trace data so as to reduce the reliance on human-picked first break times. The method is illustrated in a short MATLAB program.
View full article as PDF (0.29 Mb)