- Copyright © 2004 Society of Exploration Geophysicists
This article has been inspired by recent papers discussing the number of channels that might have to be recorded in future (single-sensor) seismic data acquisition. This article addresses this issue and some related aspects of seismic data acquisition.
What station spacing do we really need?
Current 3D acquisition geometries tend to have very large station spacing, something in the order of 50 or 60 m. Often shot-station spacing is even larger. As a consequence, not even the desired part of the wavefield is sampled without aliasing.
In contrast, there is a lot of talk on the use of single-sensor geometries that require drastically smaller station spacing (the single-sensor hype). This is driven by two main factors. In the first place, there is the view, inspired by Ongkiehong, that single-sensor recording is the ultimate in seismic data recording and every geophysicist would like to have it (the WesternGeco push). Secondly, the fact that digital sensors are inherently single sensors, so if you want to use them, you cannot do anything else but use single sensors (the I/O and Sercel push). Given this large discrepancy between the routine use of large station spacing and the ideal of very small station spacing, we should try to answer the question posed above: “What station spacing do we really need?”
The change from 50-m station spacing to 25- or 20-m in 2D acquisition provided much-improved data in the late 1970s and early 1980s. (In 1980, a seasoned interpreter complained to me that his new data were much more difficult to interpret than the old data because of the extra detail.) The main reasons for this dramatic improvement are that the desired wavefield is finally sampled without aliasing, there is less loss of high frequencies across the arrays, noise removal is more successful in processing than in the field, and there is less …