Ray footprint and redundancy in seismic tomography

The joint inversion of different wave types (e.g., direct, reflected and he ad waves) or different recording settings (e.g., VSP's, surface and cross-w ell surveys) generally provides better tomographic images than inverting th em separately. However, in the joint inversion result we cannot distinguish immediately the local contribution from each information source in the mod el space. We present here a method for analysing such a contribution, and a lso for detecting possible redundancy in the available data. The redundancy can be decreased, to reduce the computational cost, or increased, to impro ve statistically the signal/noise ratio. In particular, it can be increased at the target area either by modifying the tomographic grid, or by adding new sources and receivers at the proper recording locations. After presenti ng a 2D and a 3D synthetic example, we apply this method to a real data vin tage from the North Sea, including a 3D surface profile and a circular 3D V SP.