Hybrid shortest path and ray bending method for traveltime and raypath calculations

Abstract

The shortest path method (SPM) is a robust ray‐tracing technique that is particularly useful in 3-D tomographic studies because the method is well suited for a strongly heterogeneous seismic velocity structure. We test the accuracy of its traveltime calculations with a seismic velocity structure for which the nearly exact solution is easily found by conventional ray shooting. The errors in the 3-D SPM solution are strongly dependent on the choice of search directions in the “forward star,” and these errors appear to accumulate with traveled distance. We investigate whether these traveltime errors can be removed most efficiently by an SPM calculation on a finer grid or by additional ray bending. Testing the hybrid scheme on a realistic ray‐tracing example, we find that in an efficient mix ray bending and SPM account for roughly equal amounts of computation time. The hybrid method proves to be an order of magnitude more efficient than SPM without ray bending in our example. We advocate the hybrid ray‐tracing technique, which offers an efficient approach to find raypaths and traveltimes for large seismic refraction studies with high accuracy.