Structure-constrained relative acoustic impedance using stratigraphic coordinates |

Although seismic-derived acoustic impedance is a powerful tool in many aspects mentioned above, it is trace-based, which can cause errors in the presence of dipping layers. According to the convolutional model, seismic traces are considered normal-incidence 1D seismograms, which is strictly true only in the case of horizontal layers. When the subsurface exhibits dipping layers, the convolutional model will no longer hold true, because the seismic waveform will be sampled vertically instead of normally to the reflector (Guo and Marfurt, 2010), introducing a possible bias in the acoustic-impedance result.

In this paper, I propose to approach this problem and improve the accuracy of impedance estimates by employing the stratigraphic coordinate system (Karimi and Fomel, 2014,2011) for impedance inversion. In stratigraphic coordinates, the vertical direction stays normal to reflectors (Mallet, 2014), conforming to the assumption of the convolutional model.

In the following sections, I start by briefly reviewing the algorithm used to generate stratigraphic coordinates. Then, I explain the proposed methodology for impedance inversion. I use a field-data example to test the proposed approach and to verify that, in the presence of dipping layers, seismic-derived impedance becomes biased and can be improved significantly by the use of stratigraphic coordinates.

Structure-constrained relative acoustic impedance using stratigraphic coordinates |

2015-05-06