Bay Of Biscay Reprocessing

Bay-of-Biscay-map-SMLThe Bay of Biscay legacy dataset consists of approximately 6,400 km of 2D marine data acquired between 1968 and 2003. The data was acquired by various companies, using a variety of sources. In total 14 vintages were reprocessed in this Bay of Biscay 2014 project; ranging from the A vintage recorded in 1968 with an aquapulse source to the 2003 academic project Marconi with an airgun source.

The main objective for reprocessing the data was to provide improved imaging and better multiple removal compared to the 2003-2005 reprocessing, thus allowing for better structural analysis of the region.

The datasets within the project were extremely challenging, incorporating some surveys which were shot as academic projects or with experimental acquisition techniques. These vintages were shot with shot-point intervals recorded in seconds; and one survey was composed of varying shot intervals along each line, including a change in the near offset during the line. Of the more conventionally acquired surveys we had data with airgun, watergun, vaporchoc, aquapulse, maxipulse and mini flexichoc sources. Transcription errors on the supplied input also complicated the data retrieval.

Some surveys were only available as pre-processed CDPs without any indication of what had been applied; so after careful analysis of the seismic, some ‘best guesses’ had to be made. These complications plus, a lack of support material available for some vintages provided a distinct challenge processing the data. In addition to this, much of the data was low fold (lowest fold of 8), short cable data, which provided further challenges whilst dealing with offset related processing techniques.

Surface Related Multiple Attenuation (SRME) and High Resolution Radon offered far more effective multiple removal techniques than available when the data was previously processed (2003-2005). These processes used in conjunction with Kirchhoff Pre-Stack Time Migration (PSTM) & dense automatic velocity analysis provided much improved imaging.

The final migrated sections show distinct improvements on the legacy data available across all vintages. This verified the robustness of the processing sequence selected. Efforts were made to standardize the sequence. However, processes varied in success depending on the vintage. (Careful QC after each processing stage confirmed that the sequence was applicable to each vintage.) Once final stacks had been created, the datasets were phase matched and static shifts applied to ensure data consistency across the survey.

Processing Comparison
Click and swipe the images below to discover the imaging uplift…


ECORS-7 Legacy Processing, 2004, 30 fold (before) and ECORS-7 2014 Processing (after)


GUI84-34 Legacy Processing, 2004, 48 fold (before) and GUI84-34 2014 Processing (after)

Example processing sequence for this project:

• Reformat
• Shot & Trace Edit
• Geometry Assignment
• Velocity Analysis 2km
• Source Designature
• Amplitude Compensation
• Surface Related Multiple Elimination (SRME)
• Swell Noise Attenuation
• 1st Pass Parabolic Radon Multiple Attenuation
• Velocity Analysis 1km
• Linear FK Filter
• Tau-p Deconvolution
• Trace Balance
• Band Pass Filter
• 1st Pass Kirchhoff PSTM
• Velocity Analysis 1km
• 2nd Pass Kirchhoff PSTM
• Residual Velocity Analysis
• 2nd Pass Parabolic Radon Multiple Attenuation
• Outer Trace Mute
• Gun & Cable Static
• Static Shift and Phase Rotation (Survey Matching)
• Stack
• Anti-Alias Filter
• FK Coherency Enhancement
• Time Variant Band Pass Filter
• Equalization