In case that the subsurface is not horizontal layered system, 2d seismic data of one CMP gather are not from the same plane and 2d velocity inversion is not accurate. In order to obtain more accurate interval velocities, this article invests a 3d inversion technique from conventional 2d line profiles. Given a 2d seismic time reflections and stacking velocities (or pre-stack CMP gather), time gradients of stacking velocities along the seismic line and normal to the seismic line are calculated. A 3d time model is then established, rays of the CMP gather extend to 3d space through Snell’s law and interactive modeling and inversion method. Eventually, 3-dimensional interval velocities and reflector positions corresponding to the input stacking velocities are calculated. Results of the synthetic model and real data example show that the 3d inversion technique is more accurate than 2d inversion in the case of 3d structures. In case of dipping multilayered subsurface with different dipping angles and azimuths and with that dipping angles and azimuths don’t change too much within one CMP gather, this method can give minimally biased velocity estimates.
Interval velocity, 3D inversion, CMP gather
Geophysicists are encountering more complicated structures with development of seismic exploration. Velocity model is one of the most important factors for geologists to know more details about complicated subsurface structures. One of the reasons is that seismic velocity determines the quality of seismic image, which is the basis of geologists to understand the geology structures of the whole region. Another reason is that the result map, which gives the closest look into the true subsurface structures, is also dependent closely on the accurate velocity field. So velocity analysis of seismic data through different methods is one of the most important tasks for geophysicists. At present, For a 2d seismic data, the velocity inversion methods only consider seismic waves traveling in 2d space. In the case of 3d multilayered structures, these kinds of methods are inaccurate in velocity inversion estimates. Sorin (1993) proposed a 3d velocity inversion technique which given conventional 2d seismic stack profiles and ray-path parameters of interchanging shot-geophone ray in routine 2D line. In this article, we invests a 3d inversion technique from conventional 2d line profiles and 2D stacking velocity or pre-stack CMP trace gather.
A six layer model with different dipping angles and different azimuths, shown in Figure 1, is established to test the proposed velocity inversion method. In the middle of the model, three seismic lines with 1km interval are designed along north-south and west-east respectively. The map of the seismic lines is shown in Figure 2. Zero-offset time sections of these lines and the stacking velocities at the intersections of the lines are obtained by ray-tracing the synthetic model. The stacking velocities are used to calculate interval velocities with the inversion technique proposed above. Table 1 shows the interval velocities of Line 02 (north-south) at intersect B by 2d and 3d inversion technique.