Trim statics - Calculate

<< Click to Display Table of Contents >>

Navigation:  Phase >

Trim statics - Calculate

 

Description

Trim statics - Calculate measures and stores the short-period residual time shifts (trim statics) needed to align individual traces within a gather. It works by cross-correlating each input trace against a reference (typically a stack or super-gather model) within a user-defined time window, then finding the time shift that maximises the similarity. The resulting per-trace shift values are stored in the output trace headers and optionally in a trim file, ready to be applied by the companion module Trim statics - Apply.

Two algorithms are available. The Simple algorithm uses a single cross-correlation pass with overlap control and optional trend-deviation removal. The Advanced algorithm performs multiple iterations, refines the reference after each pass, and offers additional controls such as trace-window median filtering, offset-sorted processing, and sub-sample interpolation. Use the Advanced algorithm when the gather alignment is poor or when high-accuracy static estimation is required.

Input data

Input DataItem

The primary seismic dataset whose gathers require trim static correction. Connect the processing sequence data item here.

Input SEG-Y data handle

Handle to the SEG-Y file associated with the primary dataset, required for random-access trace reading during cross-correlation.

Input trace headers

Trace header table for the primary dataset, providing geometry coordinates and other attributes used during gather selection and static storage.

Input gather

The current gather (e.g., CMP or shot gather) passed from the preceding processing step. Trim statics are calculated gather by gather.

Input stack line

Optional stack line used as the reference model for cross-correlation when no super-gather model is provided.

Input crooked line

Crooked-line geometry definition for 2D surveys processed along non-straight acquisition lines.

Input bin grid

3D bin-grid definition specifying the inline/crossline layout of the survey. Required for 3D datasets.

Input sorted headers

Pre-sorted trace index used to efficiently locate neighbouring gathers when building the model or super-gather.

Model

Container for optional model inputs that can improve the reference trace quality.

Super gather (optional)

An optional super-gather formed by stacking neighbouring CMPs, used as the cross-correlation reference. When connected, the super-gather provides a higher signal-to-noise reference than a single stack trace, improving static estimation in noisy data.

Partial xcor summing for each bin

When enabled (default: on), partial cross-correlation sums are accumulated and averaged over the analysis window rather than computing a single global correlation. This reduces sensitivity to localised amplitude anomalies and produces more robust static estimates.

Parameters

Algorithm { Simple, Advanced }

Selects the cross-correlation algorithm. Default: Advanced.

The Simple algorithm performs a single-pass cross-correlation using a sliding window with configurable overlap. It exposes the Overlapping and Remove deviations from trend detection parameters. Use the Simple algorithm for a quick first-pass or when processing speed is more important than precision.

The Advanced algorithm performs multiple iterations of cross-correlation, updating the reference stack after each pass. It provides Trace window, Median, Sort by offset, Interpolation type, and Number of iterations controls. Use the Advanced algorithm when high-accuracy static estimation is required.

Selection type { Constant time, Time variant }

Controls how the analysis window is defined. Default: Constant time.

When set to Constant time, the From time and To time parameters define a single fixed window applied to every trace. When set to Time variant, the Time variant collection allows you to define multiple time intervals with independent window parameters, enabling the calculation to adapt to different depth zones.

Time variant

Collection of time-variant analysis windows. Only active when Selection type is set to Time variant. Each entry in the collection defines a distinct time interval for static estimation, with its own From time, To time, Analysis window step, and Maximum time shift. Add multiple entries to calculate trim statics optimised for different zones of the record.

From time

Start of the analysis window, in seconds. Default: 0.1 s. Active when Selection type is Constant time. Set this to the top of the zone of interest, typically just below any noisy shallow arrivals.

To time

End of the analysis window, in seconds. Default: 3.0 s. Active when Selection type is Constant time. Set this to the base of the target zone. Avoid including deep reflections where the signal-to-noise ratio is very low.

Max time shift

Maximum allowable trim static shift, in seconds. Default: 0.025 s. The cross-correlation search is limited to the range [−MaxShift, +MaxShift]. Set this value to slightly exceed the expected residual static magnitude. Setting it too large may produce cycle-skips on noisy data; setting it too small will clip the correction on traces with large residuals.

Overlapping

Window overlap length, in seconds. Default: 0.01 s. Active for the Simple algorithm only. Adjacent sub-windows within the analysis range are overlapped by this amount to reduce edge effects in the cross-correlation. Increase this value if you observe step artefacts in the estimated statics.

Remove deviations from trend detection

Enables automatic rejection of outlier static values that deviate significantly from the estimated spatial trend. Default: on. Active for the Simple algorithm only. When enabled, statics that are inconsistent with their neighbours are replaced by interpolated values, preventing isolated noisy traces from distorting the correction field. Disable this option only if you have already pre-cleaned the gather.

Time window

Length of the individual cross-correlation sub-window, in seconds. Default: 3.0 s. The cross-correlation is computed within this sliding window, which is stepped across the range From time–To time. A longer window averages more of the gather and gives more stable estimates; a shorter window is better for resolving time-varying statics.

Trace window

Number of neighbouring traces used when building the reference for each trace, as a half-width (traces on each side). Default: 0 (use the full gather stack). Active for the Advanced algorithm only. Setting a non-zero value creates a local reference from adjacent traces, which can improve performance on gathers with a large number of traces or strong amplitude variations across offset.

Median

Half-width of the median filter (in traces) applied to the estimated static field after each iteration. Default: -1 (disabled). Active for the Advanced algorithm only. When set to a positive integer, the median filter smooths out spike artefacts in the computed statics while preserving genuine short-wavelength variation. A value of 1–3 is typically sufficient to suppress isolated outliers.

Sort by offset

When enabled, traces within the gather are sorted by offset before cross-correlation. Default: off. Active for the Advanced algorithm only. Sorting by offset can improve the spatial consistency of the reference trace when the Trace window option is used, because neighbouring traces in offset space have similar moveout and amplitude characteristics.

Interpolation type { 8-points, Linear }

Sub-sample interpolation method used to refine the peak of the cross-correlation function. Default: 8-points. Active for the Advanced algorithm only. The 8-points (sinc) interpolation provides sub-sample accuracy and is recommended for most workflows. Select Linear if the data sample rate is already very fine and sub-sample precision is not required.

Number of iterations

Number of cross-correlation iterations performed by the Advanced algorithm. Default: 3. After each iteration the estimated statics are applied to the traces and the reference stack is recomputed, progressively improving alignment. Increasing the number of iterations generally improves quality but increases computation time. Values of 3–5 are typical; further iterations yield diminishing returns.

Save parameters to trim gather headers

When enabled, the calculated trim static shifts are written into the trace headers of the output gather in addition to any trim file output. Default: on. Storing shifts in headers allows downstream modules to read them directly without requiring a separate trim file, which simplifies single-pass workflows.

Settings

Auto-connection

When enabled, g-Platform automatically connects compatible data items from the preceding module in the sequence. Disable this option if you need to manually wire specific input connections.

Bad data values option { Fix, Notify, Continue }

Determines how the module responds when NaN or infinity values are detected in the input data. Fix replaces bad values with zero before processing. Notify logs a warning and continues. Continue silently proceeds without modification.

Execute on { CPU, GPU }

Selects the compute device for processing. CPU execution is the default and is available on all systems. GPU execution can significantly accelerate cross-correlation on large datasets when a compatible CUDA device is present.

Distributed execution

Options for distributing the gather-by-gather static calculation across multiple processing nodes in a cluster environment.

Bulk size

Minimum number of gathers dispatched to each processing node as a single work unit. Larger values reduce scheduling overhead at the cost of less fine-grained load balancing.

Limit number of threads on nodes

Caps the number of CPU threads used per cluster node. Useful when the node hosts other concurrent jobs and full-core utilisation would cause resource contention.

Job suffix

Text appended to the distributed job name for identification in cluster job queues. Useful when running multiple simultaneous processing flows.

Set custom affinity

Enables manual specification of CPU core affinity for the processing threads. When disabled, the operating system scheduler assigns cores automatically.

Affinity

The specific CPU core or core range to which the processing threads are pinned. Only active when Set custom affinity is enabled.

Number of threads

Number of parallel CPU threads used for processing. Set to 0 to use all available logical cores. Reducing this value can free resources for other concurrent processes.

Skip

When enabled, this module is bypassed and input data is passed through unchanged. Use Skip to temporarily disable static calculation without restructuring the processing sequence.

Output data

Output DataItem

The primary output dataset, passed to the next module in the sequence. Trace amplitudes are unchanged; the output differs from the input only in that trim static values have been added to the trace headers.

Output SEG-Y data handle

SEG-Y file handle propagated downstream for use by subsequent read/write modules.

Output trace headers

Updated trace header table containing the newly computed trim static shift values in the appropriate header fields, ready for use by the Trim statics - Apply module.

Output gather

The output gather with unchanged trace amplitudes and updated headers. Connect this output to Trim statics - Apply or to a save module to store the calculated shifts.

Output stack line

Stack line passed through unchanged for use by downstream modules.

Output crooked line

Crooked-line definition passed through to downstream modules.

Output bin grid

Bin grid passed through to downstream modules, unchanged.

Output sorted headers

Sorted trace index passed through to downstream modules.

Information

Graphics

Custom actions