Here are two different expanded abstracts from the 1991 SEG annual convention: [ seg1991depth1.pdf ] [ seg1991depth2.pdf ]
Here's a manuscript (unpublished) [ reflection_tomo/ ] explaining a method of reflection tomography after depth migration.
For example, A plot [ tomoin.gif ] of picked zero-offset reflection times (black lines) and of picked prestack stacking "velocities" (in color) describe the change in prestack times with offset. This interval velocity model [ tomoout.gif ] , in depth, was estimated from the picked traveltime information. Lines show the estimated locations of picked reflections in depth.
This paper [ seg1994amp.pdf ] from the 1994 SEG convention in Los Angeles, explains an application of amplitude tomography to surface seismic data.
Here are the original figures. [ seg94a/ ]
Here is a method to solve 3D anisotropic diving wave tomography [ divingtomo.pdf ] .
See an example of input first-arrival times [ divein.gif ] from the Husky Canadian Foothills dataset and the output velocity model [ diveout.gif ] with sample raypaths, estimated by diving wave tomography. Input station numbers are 20m apart. Surface-consistent static shifts were also estimated, but negligible in this case. Smooth near-surface velocity changes were able to model and explain all first arrival times well.
Here is an expanded abstract using wavepaths presented at the 1990 SEG annual convention: [ seg1990wave.pdf ]
Interval velocities [ xwell.gif ] were inverted from first-arrival times between four wells. Sample raypaths are included. The wells are not coplanar.