- Copyright © 2001 Society of Exploration Geophysicists
Conventional methods of 3-D survey design concentrate on properties of acquisition geometry such as fold, offset distributions, and azimuth distributions and combinations of these properties such as fold within given offset and/or azimuth ranges. This, in a rough and ready way, enables one to make comparative statements about the relative merits of one design over another but without saying whether either will be good enough for a given target in a given area. This is because the approach neglects the “seismogram component”—i.e., the local earth response. In particular it ignores the characteristics of shot-generated noise that leak through the imperfect stack which is the consequence of most 3-D survey designs. This leakage gives rise to a so-called “footprint” on stacked data and further results such as migrated volumes derived from the stacked data.
First, we describe briefly and qualitatively what footprints are and how they arise. Once the crucial contribution of shot-generated noise to footprints is understood, it is immediately clear that cost-effective 3-D survey design must take account of the characteristics inherent in shot-generated noise in the area to be surveyed, particularly at target reflection times.
We then describe three possible ways in which synthetic raw 3-D data may be generated using existing 2-D data to test any proposed acquisition geometry. Two use the 2-D data directly; the other still requires 2-D data but just for guidance.
Footprints defined and explained
The footprint is one of the most commonly seen artifacts on 3-D data, especially for land and transition zone surveys. A footprint is most easily recognized on horizontal (time slice) views due to a well-defined periodicity clearly related to source and receiver layout. Figure 1 shows the layout for the Zamzama 3-D survey acquired by BHP in Pakistan. Figure 2 shows the same time slice from a stack of the real 3-D …