The automated selection of pilot sweeps in high-resolution seismic: some field examples

The quality of seismic data is determined by various factors among which we may recognize subsurface characteristics, wavefield sampling, recording equipment, and source characteristics. The improvement of data quality requires optimization of each or any of these factors. As far as subsurface characteristics are concemed (as they are contained in Greens function), but few options remain to improve on data quality. One should furthermore question whether any alterations to subsurface characteristics (e.g., through the improvement of surface coupling conditions) need be addressed within the context of data quality optimization. The optimization of spatial- and temporal wavefield sampling for given subsurface build up (e.g., through the application of non-linear spatial grids in 3D acquisition) has been subject of recent research and will not be dealt with in this paper. Also the selection of recording equipment (receivers, seismograph, etc.) and recording parameters (filters, gains) may be taken into account in the optimization of the acquisition process, although it might be disputed whether the introduetion of high-dynamie-range seismographs and geophones still necessitate the optimization of recording parameters. The optimization of the source function for given subsurface characteristics, either through the selection of the most suitable source, or - for given source - through optimization of source parameters (e.g., stack-fold, drive level) has been reported by numerous authors (e.g., Miller et al., 1994, v. d. Veen et al., 1998). The improvement of data quality through the optimization of vibratory source signals (sweeps) was e.g., discussed by Brouwer et al., 1997. In the paper presented here we address the practical implementation of the optimization of sweep signals proposed in the latter by Brouwer 'et al., 1997.