178 Seismic Imaging The depth conversion is not calibrated in depth. This must be carried out, and it can be performed using the depth conversion of seismic horizons by a geo-statistical method (Omre H. 1987, Sandjivy L. and Shtuka A. 2009) which simultaneously uses Bayesian co-kriging and a multilayer model, and which handles the following sources of uncertainty: 1. Velocity model uncertainty: this relates to regional (trend or low frequency) uncertainty and local uncertainty (high frequency velocity fluctuations that are invisible on the seismic). 2. Time interpretation uncertainty. Even when one considers that interpretation is unbiased and calibrated, there are fluctuations that cannot be observed by the seismic because of limited resolution. 3. Well marker uncertainty. The depth converted horizons are calibration points for the velocity model extracted from the acoustic impedance distribution. The updated velocity model thus obtained must be consistent with an underlying interval velocity model obtained by the geostatistical method for the time-to-depth conversion of seismic horizons with a resolution comparable to the resolution of the acoustic impedance sections. Some results have already been obtained using seismic lines extracted from the 3D seismic survey recorded on the Zira area (Mari J.L. and Yven B., 2014). The P-wave velocity distribution Vp and the results of elastic inversion Ip (Ip = AI = ρ.Vp) and Is (Is = ρ.Vs) are used both to compute the shear wave velocity Vs and the density ρ distributions in depth. The values of densities obtained are realistic for the sedimentary layers present. They vary between 2.25 and 2.70 g/cm3. In the Callovo-Oxfordian, the highest value of the density is 2.48 g/cm3. Figures 7.12 to 7.14 show the results obtained for the 3 profiles: 07EST10, IL405, XL217. For each profile, we show the seismic line after depth conversion, the distribution of P-wave and S-wave velocities in depth, and the distribution of density in depth.
RkJQdWJsaXNoZXIy NjA3NzQ=