Turning-ray tomography and tomostatics

Babatunde Arenrin, Gary F. Margrave, John C. Bancroft

Turning-ray tomography is a good tool for estimating near surface velocity structure, especially in areas where conventional refraction statics fails such as in the case of a hidden layer. The velocity model from turning-ray tomography can be used for statics correction, wave equation datuming and prestack depth migration. In this research we apply turning-ray tomography to the statics problem of the Hussar 2D seismic line. This process is referred to as tomostatics. The traveltime tomography algorithm is similar to the constrained damped simultaneous iterative reconstruction technique (CDSIRT). The two-point problem for ray-tracing interpolation was used for forward modelling. To verify results from tomostatics, we compared datasets after tomostatics with datasets using the delay-time approach of a conventional refraction statics. The inversion result converged after 50 iterations and was used for statics correction. This inverted velocity model is reliable to a depth of about 750 meters, i.e. from the surface location to about one-fifth of the farthest offset (the recommended depth of sounding for turning rays). Our results show that the velocity model from turning-ray tomography reveals a hidden, slow velocity layer between two fast velocity layers that conventional refraction statics would not detect. The hidden layer is in agreement with the interval velocities from well logs. As we would expect, the stacked section, after applying tomostatics, shows better continuity of events compared to the stacked section from conventional refraction statics.