|
<< Click to Display Table of Contents >> Navigation: Statics > Refraction statics - cross correlation |
This module computes refraction static corrections by cross-correlating adjacent common-source and common-receiver gathers within a correlation window that brackets the first-break arrivals. For each source or receiver line, a static shift is estimated by finding the lag at which neighbouring gather waveforms align most coherently, then combining the forward and reverse estimates to obtain a stable shift at every station location. A long-wavelength trend is removed using an exponential moving average (EMA) filter so that the final statics represent only the short-wavelength near-surface component.
The module operates interactively. Source and receiver gathers are displayed on a map and in gather windows, and first-break picks can be loaded, computed automatically using the Auto-picking LMO and Auto-picking FB actions, or edited manually. The resulting source and receiver static corrections are written to the output statics correction and azimuthal statics correction items for use by downstream static application modules. This module supports distributed (cluster) execution for large 3D datasets.
Optional LMO (linear moveout) picking item from a previous session. Connect to initialise the LMO velocity trend curves from saved picks. When connected, the LMO trend is displayed in the source and receiver gather panels.
Optional azimuthal statics correction item from a previous run or from the azimuthal refraction statics module. When connected, the module starts from these corrections and refines them rather than computing from scratch. Use in iterative workflows where multiple passes improve convergence.
The seismic dataset to be processed. Connect the output DataItem from the preceding module in the sequence, providing access to raw prestack traces for gather display and cross-correlation.
Handle to the SEG-Y file providing random-access reading of individual source, receiver, and offset gathers for interactive display and batch cross-correlation.
Trace header table providing source and receiver coordinates and line numbers. Used to build the source, receiver, and CMP trees for interactive selection and for cross-correlation gather assembly.
Spatial aperture (m) of the exponential moving average filter used to estimate and remove the long-wavelength trend from the raw static shifts. Default: 10000 m. Increase this value to remove a broader long-wavelength background and preserve only very short-wavelength near-surface anomalies. Decrease it to retain more of the medium-wavelength component. Set to a length that matches the scale of the near-surface variations you wish to correct.
Start of the cross-correlation window measured from the surface, in seconds. Default: 0.0 s. Set this slightly before the first-break arrival time for the nearest offsets to ensure the window begins just above the refracted energy.
End of the cross-correlation window in seconds. Default: 0.6 s. Set this to a time that includes the first-break arrivals for the maximum offset used in the analysis. The window should be long enough to contain the refraction moveout across the gather but not so long that it includes reflections that would bias the correlation.
Maximum allowed static correction (s) at any station. Default: 0.02 s. Estimated shifts that exceed this value are clipped to the limit, preventing unrealistically large corrections. Set to the maximum plausible near-surface delay for your survey; typical values are 0.01–0.05 s.
When enabled (default), the LMO correction applied before cross-correlation uses a single constant velocity across all offsets. Disable to use an offset-dependent (non-constant) LMO velocity, which is appropriate when near-surface velocity varies significantly across the receiver spread.
Minimum normalised cross-correlation coefficient required for a static shift estimate to be accepted. Expressed as a fraction from 0 to 1. Default: 0.1. Estimates with a cross-correlation peak below this value are rejected and the station is assigned a zero shift. Increase this threshold to reject poorly correlated picks on noisy data; decrease it if too many stations are being rejected.
Container grouping parameters that control automatic LMO trend picking. The LMO (linear moveout) velocity trend is picked to align the first-break arrivals before cross-correlation, improving the quality of the shift estimates.
Minimum normalised amplitude a sample must exceed to be considered a first-break candidate during LMO auto-picking. Expressed as a fraction from 0 to 1. Default: 0.1. Increase on high signal-to-noise data to avoid picking noise spikes; decrease on low signal-to-noise data to avoid missing weak first breaks.
Maximum number of samples by which a first-break pick may deviate from the predicted LMO trend before it is rejected during auto-picking. Default: 5 samples. Valid range: 1–50. Increase to allow the auto-picker to follow larger lateral variations in first-break time; decrease to suppress outlier picks on data with consistent refractor moveout.
Automatically picks the LMO velocity trend for the currently displayed source or receiver gather. Keyboard shortcut: Alt+L. The trend is used to apply a linear moveout correction before cross-correlation, aligning the first breaks and improving shift estimation. Review and edit the resulting LMO curve in the gather display before running the correlation.
Deletes all LMO picks for the current session, resetting the LMO curves to zero. Use before restarting LMO picking with different parameters.
Loads previously saved LMO picks into the session. Use to resume a picking session or to import LMO picks generated in a separate interactive run.
Saves the current LMO picks to disk. Run this action after completing LMO picking to preserve the results for future sessions.
Automatically picks first-break arrivals for all source and receiver gathers using the LMO trend as a guide. The picks are displayed as FB picking overlays on the gather windows. Review the auto-picks before running the cross-correlation to verify their quality on representative gathers.
Exports the first-break picks and LMO curves to an external ASCII file. Use to share picks with other software or to archive first-break picking results alongside the static corrections.
Resets all computed static shifts to zero for the current session. Use when restarting the static estimation after changing correlation parameters, to ensure the new estimates are computed from scratch rather than accumulated on top of previous results.