Pre-stack Depth Imaging as a Tool of Accurate Structure Delineation in Presence of Complex Surface Topography
By
Turki Z Al-Rowaili1, Alex Litvin2
(1) Saudi Aramco, Dhahran, Saudi Arabia (2) Paradigm Geophysical, Woking, United Kingdom
In Saudi Arabia, complex surface topography features and in particular sand
dunes, lead to distortion of target horizons on seismic time images. These
distortions can lead to incorrect time and final depth maps unless corrections
are made during
processing
.
In this paper we show benefits of pre-stack depth imaging workflow that includes tomographic iterative velocity modelling procedure. The workflow is presented for 2D data, but it can be applied with minor modifications to 3D data as well.
Initial velocity modelling uses CMP gathers on floating datum to derive layer velocities using ray tracing based approach - coherency inversion. Ray tracing is performed from floating datum representing a smoothed version of the topography. An initial velocity-depth model is used for the first pass of pre-stack depth migration. This process generates a depth section and depth image gathers in each bin location. Residual moveout on depth image gathers is analysed along the model horizons and the initial interval velocity - depth model is used in tomographic model update procedure. Low effective offset for shallow reflectors does not allow direct update of the shallow velocity model. We introduce a tomographic procedure that uses better quality deeper reflections along with shallow reflections to update the shallow velocity model. The updated velocity model is used for the next iteration of pre-stack depth migration and at this stage tomography is used to update velocity in deeper layers of the model. This way we generate a final velocity model and final depth image free of artificial distortions observed on time images.