|
<< Click to Display Table of Contents >> Navigation: Statics > MF statics - surface consistent |
MF statics - surface consistent estimates and solves for surface-consistent static corrections using a cross-correlation engine based on the IMS (Iterative Multi-Surface) algorithm. The module decomposes the total static shift on each trace into separate source, receiver, and CMP terms, then iteratively updates these terms by cross-correlating traces against a continuously improving reference stack. The surface-consistent decomposition ensures that statics are geologically meaningful and can be applied to improve gather alignment before stacking.
Optionally, an NMO correction can be applied internally before the cross-correlation to improve the quality of the reference stack on non-NMO-corrected gathers. The module can also use a Multi-Focusing (MF) model to build the reference stack with enhanced dip and azimuth coverage. The solved statics are output as a surface-consistent static correction object that can be applied downstream.
The primary seismic dataset on which surface-consistent statics are to be estimated. Connect the main pre-stack data flow here.
Handle to the SEG-Y file for random-access trace reading. Required for efficient block-by-block processing of the survey during each iteration.
Pre-sorted trace index used to organise traces by CMP for efficient stack building and spatial block processing.
Optional surface-consistent static correction from a previous run. Connect this input and enable Use previous result to initialise the solver with previously computed shifts, which can accelerate convergence and improve stability on difficult datasets.
Container grouping parameters that control how the reference stack (model) is constructed for cross-correlation.
Number of neighbouring CMP bins (half-width) included when summing traces to build the reference stack for each bin. Default: 1 CMP. Increasing this value improves the signal-to-noise ratio of the reference stack but may degrade static resolution in areas of rapid lateral variation. Must be equal to or greater than the Super bin aperture value.
Structural dip correction applied to the reference trace when summing across neighbouring CMPs for a dip-guided stack, in ms/trace. Default: 0.004 ms/trace. Active when Use MF model is disabled. Set to the approximate regional dip of the primary reflectors to avoid smearing in the stacked reference. Visible only when Use MF model is off.
Maximum time shift applied to align traces from adjacent CMPs before stacking, in seconds. Default: 0.01 s. Visible only when Use MF model is off. Limits the dip-based shift to prevent cycle-skip artefacts in the reference stack.
Number of adjacent CMP bins (half-width) used to form a super-bin stack by summing nearby traces before cross-correlation. Default: 1 CMP. Must be less than or equal to the Model aperture. A super-bin increases the reference fold and can improve correlation quality on low-fold data.
When enabled, the module uses the Multi-Focusing (5D) model to build the reference stack, taking into account multiple dip and azimuth directions simultaneously. Default: off. Enabling this option activates the Minimum dip, Maximum dip, Delta dip, Number of directions, and Windows selection parameters in place of the simpler Dip and Max CMP shift parameters. Use the MF model on structurally complex datasets where a single-dip reference is insufficient.
Minimum dip value for the MF model dip scan, in s/trace. Default: -0.5 s/trace. Active only when Use MF model is enabled.
Maximum dip value for the MF model dip scan, in s/trace. Default: 1.0 s/trace. Active only when Use MF model is enabled.
Dip increment for the MF model dip scan, in s/trace. Default: 0.008 s/trace. Active only when Use MF model is enabled. Finer increments improve dip resolution but increase computation time.
Number of azimuthal directions included in the MF model scan. Default: 1. Active only when Use MF model is enabled. Increasing the number of directions improves performance on 3D surveys with significant azimuthal anisotropy.
Half-length of the time window used for dip selection during MF model stacking, in seconds. Default: 0.020 s. Active only when Use MF model is enabled.
Container grouping parameters that control the Common-Offset (CO) stacking strategy used to build the cross-correlation model. CO stacking organises traces by offset class to produce a more balanced reference.
Minimum CO stack aperture, in metres. Default: 100 m. Defines the smallest spatial aperture (half-width in metres) used when building the CO reference stack.
Maximum CO stack aperture, in metres. Default: 1500 m. The CO stacking can adaptively increase its aperture up to this maximum to improve reference quality in low-fold areas.
Aperture increment step for adaptive CO stack aperture search, in metres. Default: 150 m.
Additional aperture extension (in metres) applied when the standard CO aperture does not gather sufficient traces. Default: 150 m.
Number of cross-correlation iterations performed within each CO offset class before moving to the next. Default: 1.
Reserved parameter, currently not used. Default: 3.
Bin aperture used for CO cross-correlation, in metres. Default: 150 m.
Maximum number of traces included in a single CO stack aperture. Default: 1000. Limits memory usage and computation time per bin when fold is very high.
Enables the enhanced CO stacking mode that adaptively adjusts the aperture to maximise reference quality. Default: on. Recommended for most datasets.
Read-only display showing the current iteration number during processing. Updated automatically by the module as it progresses through the solver iterations.
Start of the cross-correlation window, in seconds. Default: -0.001 s (use full trace). Set this to the top of the zone of interest to focus the static solution on the target reflections. Excluding shallow noisy arrivals from the calculation window generally improves static quality.
End of the cross-correlation window, in seconds. Default: -0.001 s (use full trace). Set this to the base of the target zone.
Length of the cross-correlation window used to compute the similarity between each trace and the reference, in seconds. Default: 0.2 s. A longer window produces more stable cross-correlation estimates; a shorter window can better resolve time-variant behaviour. Set to cover at least two or three dominant reflection events.
Maximum static shift that can be estimated for any source or receiver, in seconds. Default: 0.04 s. The cross-correlation peak search is restricted to the range [−MaxStaticShift, +MaxStaticShift]. Set this to slightly exceed the largest expected near-surface delay. Too large a value increases the risk of cycle-skips on noisy traces.
Minimum cross-correlation coefficient required for a static estimate to be accepted, in the range 0–1. Default: 0.0 (accept all). Estimates with a correlation coefficient below this threshold are rejected and the trace is assigned a zero static shift for that iteration. Increase this threshold to suppress solutions from noisy or misaligned traces.
Total number of surface-consistent solver iterations. Default: 6. After each iteration the estimated statics are applied and the reference stack is recomputed. Increasing the number of iterations improves convergence on datasets with large initial statics, but the gain per additional iteration diminishes after 5–10 passes.
Number of Iteratively Re-weighted Least Squares (IRLS) sub-iterations performed during each main solver iteration to improve robustness to outlier traces. Default: 1. Increasing this value makes the solver less sensitive to traces with large anomalous shifts, at the cost of additional computation.
Container grouping parameters that control the removal of long-wavelength trends from the estimated static field after each iteration.
When enabled, a long-wavelength trend is estimated from the static field and removed after each iteration, leaving only the short-wavelength near-surface component. Default: on. Trend removal prevents the solver from absorbing structural dip or other coherent long-wavelength signals into the static solution. Disable this option only if the long-wavelength component of the statics is known to be real near-surface delay.
Grid cell size used to compute the long-wavelength trend surface, in metres. Default: 100 m. Active when Remove trend is enabled. Defines the sampling resolution of the estimated trend; set it to match the typical bin spacing of the survey.
Spatial smoothing radius applied to the trend surface before subtraction, in metres. Default: 2000 m. Active when Remove trend is enabled. Increase this value to remove only the very long-wavelength component of the statics field; decrease it for more aggressive trend removal.
Container grouping optional advanced solver parameters for CMP and RMO (Residual Moveout) terms.
When enabled, the solver includes a CMP-dependent static term in addition to the surface-consistent source and receiver terms. Default: on. The CMP term captures any residual time shift that is common to all traces in a bin and cannot be attributed to the source or receiver alone. Disable this option if you need a purely surface-consistent solution (source + receiver terms only).
Minimum CMP fold required for a bin to contribute to the solution. Default: 10. Bins with fewer than this number of live traces are excluded from the static estimation to prevent unreliable cross-correlation results from influencing the solution. Increase this value on surveys with noisy shallow data.
When enabled, the solver estimates and removes a residual NMO (moveout) term from each gather before computing the surface-consistent static shifts. Default: on. This prevents residual moveout from being misinterpreted as a static shift, which is particularly important on gathers that have been NMO-corrected but still have residual curvature.
Limits the magnitude of the estimated residual RMO term to prevent over-correction. Default: on. Active when Use a residual RMO term is enabled.
When enabled, the solver considers multiple local maxima of the cross-correlation function and selects the one that is most consistent with the surface-consistent model. Default: off. Use this option when cycle-skip artefacts are observed in the initial solution.
Spatial aperture (in metres) used when estimating the residual RMO term. Default: 0 m (use all offsets in the gather). Active when Use a residual RMO term is enabled.
Minimum offset threshold for RMO estimation, in metres. Default: 150 m. Near-offset traces where moveout curvature is negligible are excluded from the RMO estimation to avoid introducing noise.
Container grouping parameters that control memory allocation and block-processing strategy.
Expected maximum CMP fold, used together with RAM cache size to compute the optimal block size automatically. Default: 100. Set this to the actual maximum fold of the survey to ensure the automatic block-size calculation reserves sufficient memory.
Amount of RAM allocated for trace caching during block processing, in GB. Default: 8 GB. This value, together with Max binning fold, determines the block sizes automatically when Setup block size manually is off. Increase this value on machines with large RAM to allow the processing of larger spatial blocks in each pass, reducing the number of I/O passes required.
When enabled, the Block size Inline and Block size Crossline parameters are exposed for manual entry, overriding the automatic calculation. Default: off. Use manual block sizing when the automatic calculation does not respect specific hardware or dataset constraints.
Number of CMPs processed in a single inline block. Default: 1500 CMPs. Active when Setup block size manually is on. Larger blocks require more RAM but reduce the number of I/O passes.
Number of CMPs processed in a single crossline block. Default: 1 CMP. Active when Setup block size manually is on.
Enables in-memory caching of blocks of traces to speed up repeated random-access reads during iterative processing. Default: on. Disable this option only on machines with very limited RAM where the cache itself becomes a bottleneck.
Container grouping options that control the generation of intermediate stacks for QC during processing.
When enabled, intermediate stacked sections are created and displayed in the graphics views at the interval defined by Show additional graphics every N iteration. Default: off. Enable this during initial tests to monitor convergence of the static solution.
Interval (in iterations) at which intermediate stacked sections are refreshed in the Inline and Crossline graphics views. Default: 100000 (effectively disabled). Active when Create stacks is on. Set to a smaller value such as 1–5 to refresh the stacks after each iteration for real-time QC.
Inline number for which the intermediate stack is displayed in the Inline graphics view. Default: 0. Use the Create base stack custom action to populate the display for the selected inline.
Crossline number for which the intermediate stack is displayed in the Crossline graphics view. Default: 0.
Container grouping NMO correction parameters. An optional internal NMO correction can be applied to non-NMO-corrected input gathers before cross-correlation to improve reference stack quality.
Indicates whether the input gathers have already been NMO-corrected. Options: NMO Gathers (default) or Gathers without NMO. When set to Gathers without NMO, the module internally applies an NMO correction using the provided Vrms model before building the reference stack, then removes it before computing the final static shifts.
RMS velocity model used for the internal NMO correction. Required when Input gathers type is set to Gathers without NMO. Connect a velocity model object from the project.
Maximum NMO stretch limit, in percent. Default: 1%. Traces stretched beyond this limit during the internal NMO correction are muted. A value of 1% is very tight; for typical preprocessing workflows a value of 20–50% is more common.
When enabled, a bulk datum shift is applied to each trace before cross-correlation. Default: on. The datum elevation and replacement velocity are specified by the Datum and V0 parameters below.
Datum elevation used for the bulk datum shift, in metres (depth). Default: 0 m. Active when Shift to datum is enabled.
Replacement velocity used for the datum correction, in m/s. Default: 1500 m/s. Set this to the velocity of the replacement medium (typically the water velocity for marine data or the near-surface refraction velocity for land data).
When enabled, the solver is initialised with the static values loaded from the Input static connection rather than starting from zero. Default: off. Use this option when continuing a previously saved run or when applying an external pilot static as a starting point.
Optional directory path where intermediate static files are saved at the end of each iteration. Saving iteration results allows you to resume processing from any completed iteration and to examine the convergence of the static solution. Leave blank to skip saving intermediate results.
Exports the current solved surface-consistent static field to a plain-text ASCII file. Use this to archive the solution or to import it into external tools for quality control and editing.
Loads a previously saved static correction from a binary .rstat file and initialises the solver state. After importing, enable Use previous result to continue the solve from this starting point.
Builds and displays stacked sections for the inline and crossline numbers specified in the Visualization settings. Use this action before running the solver to assess the initial data quality, or interactively between runs to monitor the improvement in stack coherency.
Clears all stacked sections from the Inline and Crossline graphics views, freeing the associated memory.