|
<< Click to Display Table of Contents >> Navigation: Statics > Statics - surface consistent (RnD) |
This module computes surface-consistent static corrections using an iterative cross-correlation algorithm (IMS method). For each iteration it assembles a model trace for every CDP bin by stacking neighbouring gathers within the Model aperture, cross-correlates each input trace against its bin model within the Correlation length window, decomposes the resulting trace shifts into surface-consistent source and receiver components using a least-squares solver, and updates the static corrections. The process repeats for the specified Number of iterations. Optionally, the algorithm can also correct for phase rotation (surface-consistent phase rotation) or simultaneously estimate both time and phase corrections.
This module is the research and development variant of the production surface-consistent statics module. It includes extended options for velocity update during iteration, improving statics, azimuthal term handling, and variable V0 matrices that are not available in the production version. It supports distributed (cluster) execution and is suitable for large 3D land surveys. The output statics correction item is passed to a downstream statics application module.
When disabled (default), the module reads trace data via the in-memory SEG-Y data handle and trace headers. When enabled, it reads directly from a trace vector file on disk via the Input traces data handle, which is useful for very large datasets that exceed available RAM. Switching this flag changes which input items are visible and required.
Handle to the SEG-Y file used for random-access reading of traces during model assembly and cross-correlation. Visible when Use trace vector on disk is disabled.
Trace header table providing source, receiver, and CMP geometry. Used to build the bin and surface-consistent decomposition structure. Visible when Use trace vector on disk is disabled.
On-disk trace vector file handle for large-dataset mode. Visible when Use trace vector on disk is enabled.
Optional LMO picking item. When connected, a linear moveout correction based on the picked velocity trend is applied to the data before cross-correlation, improving static estimation on data with strong refraction moveout.
Number of traces read per I/O batch when reading from the on-disk trace vector. Default: 200000 traces. Visible when Use trace vector on disk is enabled. Adjust to balance throughput and memory usage.
Optional statics correction item from a previous run. When connected and Use previous result is enabled in the Misc parameters container, the module starts from these corrections rather than from zero, which is useful for iterating the solution across multiple processing passes.
Container grouping parameters that control how the reference model trace is assembled from neighbouring CDP bins.
Number of neighbouring CDPs on each side of the central bin included in the model stack. Default: 1. Increasing this value raises the fold of the model trace, improving its signal-to-noise ratio. Too large a value can smear lateral velocity variations into the model and degrade static accuracy.
Maximum structural dip (ms per trace) applied when assembling the model stack along the inline direction. Default: 0.004 ms/tr. Set to the maximum expected dip of the dominant reflectors; this prevents time-alignment errors in the model on steeply dipping datasets. Active when Use MF model is disabled.
Maximum time shift (s) allowed for aligning the CMP model before cross-correlation. Default: 0.01 s. Active when Use MF model is disabled.
Number of adjacent traces summed into each super-bin trace to increase the model signal-to-noise ratio. Default: 1 (no super-binning). Increase on low-fold 3D data.
When enabled, the model is assembled using a multi-focusing (5D) dip-scan approach instead of the simple CMP stack. Activates the Minimum dip, Maximum dip, Delta dip, Number of directions, and Windows selection parameters. This mode provides better model quality in the presence of conflicting dips but requires significantly more computation.
Minimum slope (s/tr) scanned during the MF model assembly. Default: -0.5. Active when Use MF model is enabled.
Maximum slope (s/tr) scanned during the MF model assembly. Default: 1.0. Active when Use MF model is enabled.
Dip step (s/tr) between scanned directions in the MF model. Default: 0.008. Active when Use MF model is enabled.
Number of 2D directional scans used to build the 3D MF model. Default: 1. Active when Use MF model is enabled.
Time half-window (s) used for selecting the best-fitting dip direction in the MF model. Default: 0.020 s. Active when Use MF model is enabled.
Container grouping the primary cross-correlation and iteration parameters.
Start of the time window used for cross-correlation (s). Default: -0.001 s (use the full record). Set to the two-way time of the shallowest reflector to include in the statics estimation. Excluding very shallow noise improves convergence.
End of the time window used for cross-correlation (s). Default: -0.001 s (use the full record). Set to the deepest reflector time to include; excluding noise below the target horizon improves the stability of the solution.
Length (s) of each sub-window used for computing cross-correlation functions within the analysis time range. Default: 0.2 s. Longer windows give more stable cross-correlations but average over more of the record. Shorter windows allow time-variant shift estimation within the analysis range. Typical values: 0.1–0.5 s.
Number of samples used for smoothing the velocity function in the time direction when velocity-based NMO undercorrection is removed before cross-correlation. Default: 10 samples.
Maximum surface-consistent static correction allowed per source or receiver (s). Default: 0.04 s. Estimated shifts exceeding this limit are clipped. Set to the maximum plausible near-surface delay for your survey. Typical values: 0.02–0.1 s.
Minimum cross-correlation coefficient (0 to 1) required for a shift estimate to be included in the surface-consistent decomposition. Default: 0 (accept all). Increase to reject poorly correlated pairs on noisy data.
Number of outer IMS iterations. Default: 5. Each iteration updates the model with the current static corrections and re-estimates the shifts. More iterations improve convergence but increase computation time. Typical values: 3–10. Monitor convergence using the convergence chart display.
Number of IRLS (iteratively re-weighted least squares) sub-iterations applied within each outer iteration to down-weight outlier cross-correlation measurements. Default: 6. More IRLS iterations suppress the influence of noisy or incorrectly correlated traces but increase computation time per outer iteration.
Selects what is estimated and applied. SC static shift (default) estimates only time shifts. SC phase rotation estimates only constant phase rotations (useful for correcting wavelet polarity or phase variations between source or receiver types). SC static and phase rotation estimates both simultaneously.
Angular step (degrees) used when searching for the best-fit phase rotation. Default: 5°. Active when Type includes phase rotation. Smaller values increase resolution of the phase scan at the cost of longer computation.
Container grouping parameters for removing the long-wavelength trend from the computed static shifts, so that only the short-wavelength residual is preserved.
When enabled (default), the long-wavelength trend defined by the Grid size and Smooth parameters is subtracted from the computed static corrections before output. Disable to output the full absolute static shifts without trend removal.
Spatial grid cell size (m) used when interpolating the static shifts onto a regular grid for trend estimation. Default: 100 m. Set to approximately the average station spacing.
Smoothing radius (m) of the spatial filter applied to the gridded statics before subtracting the trend. Default: 2000 m. This controls the wavelength boundary between the trend (removed) and the residual (kept). Increase to remove more of the long-wavelength component; decrease to retain longer-wavelength features in the output.
Container grouping advanced algorithmic options for the surface-consistent decomposition.
Controls whether a CMP (structure) term is included in the surface-consistent decomposition to absorb residual NMO or dip-related structure. Don't use fits only source and receiver terms. Use residual adds a residual CMP term. Use cross-correlation (default) also includes a CMP cross-correlation window parameter. Including the CMP term can improve the decomposition but risks absorbing long-wavelength geology into the static correction.
Time window (s) used for computing the CMP cross-correlation term. Default: 0.010 s. Active when CMP term is set to Use cross-correlation.
Minimum CDP fold required for a bin to be included in the surface-consistent decomposition. Default: 10. Bins with fewer traces than this value are excluded, preventing low-fold edge bins from biasing the solution.
When enabled (default), a residual moveout (RMO) term is added to the decomposition model to account for residual velocity errors. This prevents velocity errors from being incorrectly absorbed into the static corrections.
When enabled (default), limits the magnitude of the RMO term to prevent it from absorbing genuine static variation. Active when Use a residual RMO term is enabled.
When enabled, the cross-correlation solver considers multiple local maxima in the correlation function instead of only the global maximum. Default: off. Use on data where cycle-skipping between local maxima is a problem, typically when the maximum static shift is large relative to the dominant period.
When enabled, cross-correlations are computed between all trace pairs in the dataset rather than between each trace and the bin model only. Default: off. This significantly increases computation time but can improve accuracy on datasets with highly variable amplitudes between shots or receivers.
Maximum offset range (m) over which the RMO correction is applied. Default: 0 m (full range). Active when Use a residual RMO term is enabled.
Minimum offset (m) included in RMO estimation. Default: 150 m. Near-offset traces are excluded because they carry little moveout information for constraining the RMO term. Active when Use a residual RMO term is enabled.
Container grouping memory and caching parameters for large-dataset performance tuning.
Maximum number of traces per CDP bin used in the model assembly. Default: 100. Bins with more traces than this limit have their traces sub-sampled, reducing computation time and memory usage on high-fold 3D surveys.
Maximum RAM (GB) allocated for the trace cache. Default: 8 GB. Increase on workstations with large available RAM to reduce disk reads and speed processing. Visible when Setup block size manually is disabled.
When enabled, the spatial processing block size is specified manually using Block size Inline(CDP) and Block size Crossline instead of being derived from the RAM cache size. Enable when you need precise control over the block geometry, for example when running on a cluster with shared memory limits.
Number of CDP bins in the inline direction per processing block. Default: 1500. Active when Setup block size manually is enabled.
Number of CDP bins in the crossline direction per processing block. Default: 1. Active when Setup block size manually is enabled.
When enabled (default), a trace cache is used to avoid re-reading data from disk during iteration. Disable only when disk I/O bandwidth is not a bottleneck and RAM is limited.
Container grouping parameters that control intermediate stack generation during the iterative computation, used for visual QC of convergence.
When enabled, intermediate stack sections are computed every N iterations and displayed in the stack view. Default: off. Enable to monitor how the stack quality improves with each iteration.
Number of outer iterations between stack updates in the display. Default: 1. Set to a higher value to reduce display overhead on long runs.
Inline number of the QC stack section displayed during processing. Default: 0. Set to an inline through a representative area of the survey.
Crossline number of the QC stack section displayed during processing. Default: 0.
Container grouping parameters for automated velocity re-picking during the static iteration. When enabled, the module re-picks velocities from semblance panels at regular iteration intervals and updates the NMO model, improving the statics–velocity coupling.
Enables automatic velocity re-picking every N iterations. Default: off.
When enabled, a separate VRMS velocity model output is created for each iteration, allowing you to track how the velocity evolves during the static-velocity loop. Default: off.
Number of outer static iterations between automatic velocity re-picks. Default: 1. Active when Enable velocity update is on.
Half-width velocity corridor (m/s) around the current velocity function within which the auto-picker searches for the new semblance maximum. Default: 100 m/s. Active when Enable velocity update is on.
Step (in trace number) between velocity analysis locations during auto-pick. Default: 20 traces. Active when Enable velocity update is on.
Time step (s) between velocity picks within each gather. Default: 0.1 s. Active when Enable velocity update is on.
Velocity step (m/s) between tested velocities in the semblance scan during auto-pick. Default: 50 m/s. Active when Enable velocity update is on.
Time smoothing window (s) applied to the semblance panel before auto-picking. Default: 0.05 s. Active when Enable velocity update is on.
NMO stretch mute threshold (fraction) applied before semblance calculation during velocity auto-pick. Default: 0.5 (50%). Active when Enable velocity update is on.
Running-window normalization length (s) applied before semblance calculation. Default: 0.1 s. Active when Enable velocity update is on.
Spatial smoothing applied to the updated velocity field. None (default) applies no smoothing. Vertical smooths only in the time direction. Full smooths in both time and spatial directions. Active when Enable velocity update is on.
Time smoothing window (s) for the updated velocity. Default: 0.05 s. Active when Smooth type is Vertical or Full.
Spatial smoothing radius (traces) in the inline direction for the updated velocity. Default: 50 traces. Active when Smooth type is Full.
Spatial smoothing radius (traces) in the crossline direction for the updated velocity. Default: 50 traces. Active when Smooth type is Full.
Container grouping NMO parameters applied to the input gathers before cross-correlation when the input data is not already NMO-corrected.
Indicates whether the input gathers are already NMO-corrected. NMO Gathers (default) assumes NMO has already been applied. Gathers without NMO applies NMO internally using the Vrms model.
VRMS velocity model used for internal NMO correction. Required when Input gathers type is set to Gathers without NMO.
NMO stretch mute threshold (fraction). Default: 1% (0.01). Samples where NMO stretch exceeds this fraction are muted before cross-correlation. Set to 0.5–0.6 for typical data.
Smoothing applied to the VRMS model in the time direction before NMO application, in samples. Default: 10 samples.
When enabled (default), a bulk datum correction is applied using V0 and the datum elevation before the surface-consistent decomposition. This removes the large-amplitude datum shift from the trace shifts, improving the stability of the cross-correlation.
Datum elevation (m) used for the bulk static correction applied before the surface-consistent decomposition. Default: 0 m.
Near-surface replacement velocity (m/s) used for the datum bulk correction. Default: 1500 m/s. Set to the estimated near-surface velocity for your survey.
Optional matrix data item providing a spatially variable near-surface replacement velocity field. When Use variable V0 (matrix) is enabled, this matrix overrides the constant V0 value. Use when near-surface velocity varies significantly across the survey.
When enabled, the spatially variable V0 matrix is used instead of the constant V0 for datum correction. Default: off.
Container grouping parameters for an additional automatic quality-control step that detects and removes bad-trace contributions from the static estimation.
When enabled, the module applies an additional pass to identify and exclude traces that contribute anomalous cross-correlation measurements from the final surface-consistent decomposition. Default: off.
Criterion used to identify bad traces in the improving static step. By semblance (default) assesses each trace by its semblance contribution. By semblance row maximums ratio compares the ratio of semblance maxima across rows. By semblance differences uses the gradient of the semblance panel.
Minimum slope (s/trace) defining the valid alignment corridor when checking for bad traces. Default: -0.004 s/tr.
Maximum slope (s/trace) of the alignment corridor. Default: 0.004 s/tr.
Slope increment (s/trace) between evaluated directions. Default: 0.004 s/tr.
Fraction of traces (0 to 1) allowed to be classified as bad within a gather before the improving step is skipped for that gather. Default: 0.5 (50%).
Number of neighbouring traces used on each side for the semblance-based bad-trace criterion. Default: 1.
Half-length (s) of the time window used for the semblance calculation in the bad-trace detection step. Default: 0.04 s.
Container grouping miscellaneous options including iterative mode and output path settings.
When enabled, the statics from the Input static item are used as the starting solution for this run, effectively continuing a previous iteration. Default: off. Enable in iterative workflows where you apply statics, re-pick velocities, and then run a further round of statics estimation.
Optional output directory path for saving intermediate static correction files from each iteration. Leave empty to skip saving intermediate results. Set a path to archive the full iteration history for later review or to restart from an intermediate state.
Velocity model data item that will receive the updated VRMS field after velocity auto-picking. Connect to the NMO module for subsequent processing using the improved velocity model.
SEG-Y trace cache configuration. Adjust cache size to balance memory usage against I/O throughput during iterative processing.
Selects CPU or GPU execution for the cross-correlation computations.
Cluster execution configuration for running on distributed compute nodes.
Minimum number of gathers per compute node chunk during distributed execution.
When enabled, limits CPU threads per cluster node to the Number of threads value.
Optional suffix appended to the cluster job name for identification.
When enabled, sets the CPU affinity mask explicitly.
CPU affinity mask used when Set custom affinity is enabled.
Maximum CPU thread count for this module.
Container grouping options for running external scripts before or after the module executes.
Path to a script executed before the module starts. Use for pre-processing setup or file preparation steps.
Path to a script executed after the module completes. Use for post-processing cleanup or result archiving.
When enabled, bypasses this module during sequence execution.
Collection of updated VRMS velocity model items produced when Enable velocity update is active. One model per velocity update iteration is stored when Create vrms models for iterations is also enabled.
The computed surface-consistent static corrections for all sources and receivers. Connect to a downstream statics application module to apply the corrections to the seismic traces.
Exports the current source and receiver static shifts to an ASCII file. Use for archiving, sharing with external software, or manual editing.
Loads static corrections from an .rstat binary file into the output static item. Use to restore results from a previous run or to import statics computed by another module.
Computes and stores a reference stack section used as the baseline for iteration quality comparison in the visualization display.
Clears all intermediate and base stack sections from memory. Use to free RAM before starting a new iteration or when restarting the computation with different parameters.