Multi-Valued Two-Point Raytracing

Imaging in complex geological areas some times require multi-valued traveltime and amplitude information[1],[2].<br>Traditionally two-point raytracing is used to calculate the shortest traveltime and raypath between two points. However if<br>there exists more than one ray solution, corresponding to multiple branches of a traveltime curve, this method is<br>computationally expensive and less successful. Here we present an efficient way to compute the minimum traveltime<br>based on a triangulated network by adaptive wavefront construction. We search for the minimum traveltime for each<br>arrival with maximum amplitude from multi-valued arrivals. The main contributions of this work compared to our previous<br>algorithm[4] are two fold: (1) obtain more ray information from wavefront construction, which will improve the accuracy<br>for two-point raytracing, and (2) decrease sufficiently the computational time of minimal traveltime searching on each<br>branch. Since we use the triangulated network for wavefront construction and implementing Fermat’s principle, we are<br>able to increase the efficiency, especially for multi-valued traveltimes. This, in turn, increases the efficiency for velocity<br>model building, traveltime table generation for 3-D prestack Kirchhoff migration, AVO analysis, tomographic analysis, and<br>seismic data acquisition planning. The algorithms are tested and illustrated on a synthetic model. Finally the algorithms<br>are tested by generating the traveltime and amplitude maps on the Mahogany subsalt model.