Abstract:
A method for sorting large volumes of seismic data portions (206) into a defined order in which the data portions (206) are allocated to leaf files (S7) of a B-Tree structure and stored in a corresponding scratch memory space (S3, 210) corresponding to the allocated leaf file; a full leaf file is read out into a sorting memory space (S13, 214) wherein it is sorted into a respective suborder. All of the sorted data portions are output (S14) in the respective sub-orders for all of the full leaf files in leaf order to provide a final output data stream, where the data portions in their respective sub-orders for all of the full leaf files in leaf order are in the defined overall order.