One and two dimensional velocity inversion in the domain of intercept time and ray parameter : an example in the Nankai Trough




Wood, Warren Theodore, 1962-

Journal Title

Journal ISSN

Volume Title



The intercept time equation in τ-p (Diebold and Stoffa, 1981) [mathematical equation] is used as the basis for exact 1D and 2D τ-p velocity analysis. Intercept time curves for an initial model are superimposed on the τ-p seismic data. Model parameters such as layer velocity, thickness, and dip are adjusted until the intercept time curves are coincident with the reflections in the data. Normal moveout in the domain of τ and p is applied during the analysis so as to check the picks of the reflections. When a reflection has been moved out correctly, it has been properly modeled, (i.e. the velocity, thickness, and dip of the layer have been determined). Once all the reflections have been imaged, the analysis is complete. One attractive feature of this method is that all of the calculations can be done quickly, so the analysis can be done interactively on a computer with a graphics screen. This velocity analysis method was then applied to long offset seismic data collected in the Nankai Trough. Eight expanding spread profiles (ESPs) and five split spread profiles (SSPs) were collected in two different areas and analyzed to accurately determine sediment velocities in the trough and on the accretionary wedge. The results are a series of 1D earth models in the ESP area and 2D models in the SSP area. The analysis of the ESP data clearly shows a low velocity zone associated with a bottom simulating reflector but does not show evidence of a large (200 - 300 m/sec) velocity reversal at the decollement as expected. The analysis of the SSPs in an area about 100 km away, however, does point to a trend of decreasing velocity just above the decollement with increasing distance under the wedge.