Input, output, and difference of the deblending tool on a single cross spread.
Seismic data is now often acquired using simultaneous or blended sources (Abma and Foster, 2020); that is, the acquisition crew does not wait until the energy from one source has fully died out before firing the next source. This speeds up acquisition and reduces costs, but it means that energy from one source will often contaminate another. This extraneous-source energy is referred to as “crosstalk”.
One could proceed with conventional processing without worrying about crosstalk, but it’s better to remove it early in the processing, a task called “deblending”. That is the purpose of this tool.
The tool works on cross-spread ensembles. It’s fast and easy to use, with only a few parameters. As each cross spread is processed independently, a job can be run over any number of nodes in parallel.
Not all of the crosstalk will be removed. Rather the aim is to significantly reduce its amplitude so that it is no longer a serious problem. The highest priority is to not distort the signal.
For quality-control purposes you can also request that only the removed crosstalk (that is, the difference) be output. Alternatively, you can use Seismic Compare on the input and output data to view the difference. There are also options for generating synthetically blended data for testing purposes.
References
Abma, R. and M. A. Foster, 2020, Simultaneous Source Seismic Acquisition, Society of Exploration Geophysicists.
Trickett, S., 2024, Ordering cross-spread gathers: Geophysics, 88 (accepted for publication). [Website Link]
Vermeer, G. J. O., 2005, Processing orthogonal geometry – what is missing?: 75th Annual International Meeting, SEG, Expanded Abstracts, 2201-2204. [Website Link]