GNGTS 2018 - 37° Convegno Nazionale

GNGTS 2018 S essione 3.1 563 complex structures. Therefore, we used it to improve our starting velocity field. Then we used the new velocity field to depth migrate the MS29 line and as input to the velocity-depth model building for the horizon-based tomography approach. The iterative working flow for the grid tomography approach can be summarized into four steps: 1. Build of the initial velocity model. We converted the RMS velocity field (used for the pre-stack time migration) into interval velocities through Dix equations. Then we laterally and vertically smoothed the velocity model, starting from the water bottom layer, to avoid artifacts due to lateral variations. We performed a first PSDM with these velocities. 2. Calculate structural attributes (continuity and dip) from the depth migrated section and of the travel-time error through an automatic picking. These attributes act as source for the autopicking process and as ray shooting for the tomography. Therefore, the travel-time error (residual move-out, RMO) is calculated by an automatic picking on the migrated gathers and it is then used by the tomography to update the velocity model 3. Use the grid tomography to update the velocity model. Tomography tends to minimize the errors between modeled and real situation by small changes in the velocity parameter. 4. Depth migration with the updated velocities and quality control on the common image gathers (CIG). We repeated these steps until the quality control on the CIG showed a satisfactory flatness of the reflections. The first step is performed only once, and the flow is stopped when a satisfactory result is reached. Horizon-based tomography. This method requires a set of picked time horizons, an initial trial interval velocity field and the CMP time gathers in order to build the initial velocity-depth model through a layer-by-layer coherence inversion. Fig. 2 - a) Interval velocity in depth obtained after the Dix conversion of the RMS velocity. b) Velocity field updated by 2D grid tomography. c) Initial velocity-depth model obtained by coherence inversion. d) Velocity-depth model updated by horizon-based tomography.