Source signature measurement & QC

Part 1 — Sources

Learning objectives

Describe how the far-field signature is recorded in the field (near-field hydrophones)
Explain the median ± MAD envelope method for outlier detection
Set a reasonable MAD threshold (typically 2.5–3.0)
Decide what to do with flagged shots (exclude from signature averaging; investigate array)

Every acquisition system records a reference signature for every shot. In a marine source array, that’s near-field hydrophones (NFH) mounted on each gun, from which the far-field signature can be computed. On land, it’s a calibrated “field monitor” recording of the vibroseis ground force or a downhole NFH on a dynamite shot.

Why per-shot signatures matter

Processing assumes a stationary, repeatable source wavelet. In practice the wavelet varies shot-to-shot due to coupling, charge variability, bubble-timing jitter, and occasional gun failures. A bad shot injects energy into the data that doesn’t match the assumed wavelet, and processing steps that rely on wavelet stationarity (decon, inversion, 4D differencing) degrade accordingly.

Median ± MAD flagging

Don’t use mean ± standard deviation — it’s too sensitive to the outliers you’re trying to detect. Use median and median absolute deviation (MAD):

\mathrm{MAD}(x) = \mathrm{median}\bigl(|x_i - \mathrm{median}(x)|\bigr)

Scale by 1.4826 and you have an outlier-robust estimate of the standard deviation for Gaussian data. A shot is flagged if its max-across-samples deviation from the median signature exceeds k·MAD, for k typically 2.5–3.0. Drop the flagged shots from the shot-averaged signature used by processing.

What the field crew does with the flags

Flagged shots trigger an investigation. A cluster of flagged shots in one area → inspect the source array. A single outlier after a change in sea state → probably a wave-slap on a gun; re-shoot. An outlier during a vibroseis point → check ground-force feedback and vibrator phase lock. Real-time flagging keeps bad data from contaminating the stack before the crew leaves the area.

References

Ziolkowski, A. (1970). A method for calculating the output pressure waveform from an air gun. Geophysical Journal International, 21(2), 137–161.
Dragoset, B. (1990). Air-gun array specs: a tutorial. Geophysics, 55(11), 1426–1440.
Pritchett, W. C. (1990). Acquiring Better Seismic Data. Chapman & Hall.
Sallas, J. J., Weber, R. M. (1982). Comments on the digital-filter equivalent of a vibrator. Geophysics, 47(11), 1577–1582.