Abstract
The stackarray approach proposed in Anstey (1986a) is a large improvement over seismic acquisition techniques with shot intervals that are 3 or 4 times as large as group intervals. The stackarray criterion (Anstey, 1986b) prescribes “an even, continuous, uniform succession of geophones across the CMP gather,” and Morse and Hildebrandt (1989) illustrate the effectiveness of this regular spatial sampling, using some interesting experiments. However, the stackarray criterion does not explain the residual ground roll that remains even after a two-trace mix or simulation of overlapping geophone arrays as shown in Figures 14b and 13b, respectively, of Morse and Hildebrandt. Morse and Hildebrandt state that these events are not suppressed due to “the unusually low fold of the stack.” In the following I will argue that Morse and Hildebrandt's argument does not satisfactorily explain the presence of residual ground roll and that the residual ground roll most likely results from not using a shot array in data acquisition. I put forward a hypothesis as to the nature of residual ground roll, and this hypothesis can be easily tested using the basic data recorded by Morse and Hildebrandt.
