An example of 3‐D spatial sampling effects on time slices

Abstract

Horizontal resolution has been discussed by several earlier papers. Denham and Sheriff (1980), Berkhout (1980) and Safar (1985) indicated that horizontal resolution is affected by the spatial sampling interval, dominant wavelength and migration aperture. In this paper, the effect of the spatial sampling interval on horizontal resolution is focused on experimentally. The dominant wavelength and migration aperture are held constant. The spatial sampling criterion to avoid aliasing of dipping events is well understood in reflection seismology, even for the 3-D case. Less well understood (and certainly less quantified) are the criteria for spatial sampling for the unambiguous interpretation of stratigraphic events in time-slices. Intuitively, the expectation is that the required spatial sampling will be related to the lateral width of geologically homogenous stratigraphic units. To test this experimentally, 3-D physical model seismic reflection data were acquired over the well-known SALNEL model. We processed the data utilizing several trace spacings to simulate land and marine geometries. For each different trace spacing, we have used DISCO 3-D modules to strip off a time-slice corresponding to the depth of the interest in our model. The data processing sequence was minimal, as these physical model data were relatively noise-free. Due to the acquisition scheme, no CMP sorting was required. The data were corrected for NMO, stacked and then migrated using a 3D finite-differences algorithm. Following data processing, a horizontal time slice was produced for the 420 ms horizons (equivalent to 1800 feet in depth). This is shown in Figure 2. The cross-line CMP number (West-East) is labeled from 1 to 240 while the inline CMP number (South-North) ranges from 1 to 250 (not shown). The time slice shown in Figure 2 shows the success of 3D method. All of the features shown in the schematic (Figure 1) are present and easily interpretable. The spatial sampling in this survey clearly either satisfied or exceeded the minimum necessary sampling criterion. Note that the sinuous dark gray event in the middle-right of Figure 2, roughly resembling an oxbow lake, is the result of pull-down due to a meander (of the same shape) at a shallower horizon in the model. For the larger sand bodies, with a width of about 400 feet, and a bin size of 50 feet, there are 8 samples across the width of feature.