| Seismic ray tracing has a broad application in real problems, such as seismic exploration,seismic tomography, earthquake location and etc. With the increase of actual demand and fastcomputing development, it is required a higher tomographic resolution in real application.For this reason, it is necessary to develop a fast and efficient algorithm for multi-valued raytracing in complex media. Therefore, we will concentrate ourselves in the development offast and efficient algorithm for multi-valued ray tracing method to meet such declaration. Forthis purpose, this research work includes two parts:(1) the multi-valued ray tracing based onphase space method and (2) the multi-valued ray tracing based on the stacked traveltime field.The phase space method includes two main classes: one is based on gridded network andthe other is based on traditional two-point ray tracing algorithm. Basically, in the phase spacemethod a third variable is introduced so that it is able to distinguish the multi-valuedwavefront, which is intercepted in real space, but unfold in the phase space. In suchcircumstance, it is easy to trace multi-valued rays. Previous researches have some problemsin handling reflected wavefront propagation at the interface. To overcome this problem we inthis work put forward a ghost imaging method to continually expand the reflected wavefrontso that it is easily to calculate the multi-valued reflection. This is major contribution in thisresearch work.Meanwhile, we also introduce some fashionable algorithms for multi-valued ray tracing,these include a ‘raylet’ method and an extreme value solution method. The former is used tocalculate multi-valued direct arrivals and the latter is applied to obtain the multi-valuedreflected arrivals. But both methods are based on the stacked traveltime fields between theforward modeling from the source and backward modeling from the receiver, in which theextramum solution is found to locate the multi-valued rays. |
PDF Full Text Request