Model-Driven Data Interpolation vs. Velocity Knowledge Uncertainties

B-07 Model-driven data interpolation vs. velocity knowledge uncertainties. P. MAZZUCCHELLI 1 N. BIENATI 2 U. SPAGNOLINI 1 Abstract 1 Model-driven data interpolation performed by continuation algorithms (i.e. 3D SMO – Shot MoveOut) exploits the redundancy of the prestack data to regularize/densify survey geometries. Their usual Kirchhoff-type kernel implies that the accuracy of interpolation results is directly linked to the coverage of the operator itself [2] then coverage must be properly defined. A simple scalar index (e.g. fold of coverage) is not enough as it does not account for uneven dip illumination. Moreover the sensitivity of the continuation operators to velocity