Blended acquisition & iterative deblending

Part 8 — Compressed sensing & modern methods

Learning objectives

State the blended-acquisition productivity gain (2–4×)
Describe the iterative fit–subtract–refit deblending loop
Quote required jitter (±60–200 ms) and why under-jitter fails
Link OBN recording to simultaneous-source tolerance

Blended (simultaneous-source) acquisition has two or more sources firing within seconds of each other. Every receiver records the sum. Deblending separates the superposed shot records in processing. Classical single-source shooting holds 6–10 s between shots; blended cuts that to 2–4 s per shot pair, with ±100–200 ms random jitter.

Why jitter is the trick

Sort the blended record into shot-1’s time frame. Shot-1 events align coherently across traces; shot-2 events appear at the jittered offsets, which DIFFER per shot, so they scatter as trace-to-trace outliers. A coherence filter (f–x median, curvelet soft-threshold) extracts shot-1; subtract it from the blended record to get a shot-2-dominated residual. Sort into shot-2’s frame, filter, subtract, iterate. Three to five iterations typically converge.

Productivity gain

Two blended sources ≈ 2× shots/day. Three vessels ≈ 2.8×, four vessels ≈ 3.4× — diminishing returns because de-blend cross-talk rises with source count. Marine streamer sweet spot: 2 vessels. OBN projects routinely push 4–6 source vessels because nodes record continuously regardless of source cadence, and the coherence filters can separate many overlapped shots. Some ultra-simultaneous-source (USS) programmes have shot 10+ sources in parallel.

Jitter requirements

Below ±60 ms the jittered shot-2 events start to look coherent in shot-1 frame, and de-blending leaks. ±100–200 ms is industry standard. Above ±300 ms the inter-shot gap must widen to accommodate, eating into productivity. For OBN the jitter can go to a full second because the nodes continue recording regardless.

References

Beasley, C. J., Chambers, R. E., Jiang, Z. (1998). A new look at simultaneous sources. SEG Annual Meeting Expanded Abstracts, 133–135.
Berkhout, A. J. (2008). Changing the mindset in seismic data acquisition. The Leading Edge, 27(7), 924–938.
Mahdad, A., Doulgeris, P., Blacquière, G. (2011). Separation of blended data by iterative estimation and subtraction of blending interference noise. Geophysics, 76(3), Q9–Q17.