GNGTS 2014 - Atti del 33° Convegno Nazionale

72 GNGTS 2014 S essione 3.1 Examples. For the following examples we used the 2D synthetic dataset Sigsbee2b. The Sigsbee2b dataset simulates a conventional marine experiment with a geology similar to those commonly found in the Gulf of Mexico environment, characterized by the presence of big salt bodies. The dataset, originally built for the purpose of testing multiples prediction algorithms, contains both surface related and internal multiple events. For the following tests, we extracted 219 CSGs (Common Shot Gathers) of the original dataset. The sources go from xs = 6027 m to xs = 16000 m and the receivers offset from 0m to 8000 m. Both sources and receivers are placed at zero-depth, at the interface between air and water. Fig. 2 (left-hand-side panel) shows a close-up of the original stratigraphic model. In Fig. 2 (right-hand-side panel) it is possible to see the image generated by the migration of the of the new virtual datasets D nl below ( B | A ) (see Fig. 1b). Sources and receivers x -coordinates are the same as the original dataset while the depth of the new acquisition surface is z datum = 5000 m. The receivers datuming procedure has been performed using the (smooth) migration velocity model being the salt interfaces the only detailed information. As expected the main contribution of the non-linear term can be noticed beneath the salt body, where the strongest intra-salt multiples generate. The panels on the top of Fig. 3 show another area of the Sigsbee2b stratigraphic model and the corresponding RTM image obtained from the original dataset. In order to test the effectiveness of the imaging of wave elds of the type D nl above ( B | A ), we performed two different tests. We first modeled a seismic experiment generated with sources at depth z = 0 and the receivers placed at datum depth z datum = 5000 m. For the FD modeling we used the Sigsbee2b stratigraphic model. Fig. 3 (bottom left panel) shows the result of the internal multiple migration: the yellow arrows point to those interfaces better imaged by the internal multiples with respect to the results of RTM fed with the original data. In a more realistic scenario, however, we have at our disposal only the data recorded at the surface. We thus retrieved D nl above ( B | A ) by performing the receivers datuming pro-cedure explained in the previous sections, followed by the up/downgoingwave eld separation and the Interferometric sources datuming. The receivers datuming procedure was performed using the stratigraphic velocity model. Fig. 3 (bottom right panel) shows the results. Some of the interfaces are well imaged (bottom of the salt, seabed, etc...) and there are some little improvements with respect to the RTM image (the interfaces pointed by the red arrows in Fig. 3). However, compared with both the images in Fig. 3 (top right and bottom left) the results are worse, especially in the region near the side of the salt body and the additional information provided with respect to the RTM image is not so significant as for the previous case. Fig. 2 – Left: Sigsbee2b stratigraphic model. Right: migration result of the intra-salt multiples. The blue box highlights the area where the internal multiples provide illumination. The illumination provided by the multiples can be useful to complement the image coming from a conventional migration.