Description

Stereo tomography multiline (test) jointly updates depth-domain velocity models for multiple 2D seismic lines using stereo tomographic inversion. All input lines share a common inversion framework so that velocity updates are laterally consistent at line crossings. This module differs from Stereo tomography 3D in that it works with a set of 2D lines rather than a true 3D volume, and it provides additional options to enforce or suppress the multiline constraint, substitute the constrained area back to the original model, and include or exclude lateral side constraints. This module is marked as a test version.

Input data

Tomo item multiline

Multi-line tomo dataset assembled by Set multi line tomo item. Contains tomo parameters for all lines to be jointly inverted.

Horizons for constrains

Optional horizon picks used as hard geological constraints to bound the velocity update zone within a specific stratigraphic interval.

Parameters

Layer constraint

Container grouping horizon-based layer constraint settings.

Layer from { }

Top horizon of the constrained velocity update zone.

Layer to { }

Bottom horizon of the constrained velocity update zone.

Use layer constraint

When enabled, restricts velocity updates to the zone between Layer from and Layer to. Default: off.

Don't use multiline constraint

When enabled, disables the multiline intersection constraint so each line is inverted independently using the legacy single-line algorithm. Default: off. Enable for backward compatibility with older workflows or to diagnose multiline constraint artefacts.

Use side constraint

When enabled, applies lateral side constraints to regularise velocity updates at the edges of the inversion grid. Default: off. Useful for preventing edge artefacts in areas with sparse data coverage.

Substitute constraint area to original

When enabled (default), the velocity values in the constrained horizon zone are replaced with the values from the original input model after each iteration, preventing the inversion from updating velocities inside already-resolved intervals. Disable to allow the inversion to modify velocity in all zones including those bounded by horizons.

Factor ML

Multi-line weighting factor controlling the strength of inter-line consistency constraint relative to single-line data fit. Default: 1.

Datum

Processing datum elevation (m) read from the input data (not editable).

Velocity from topo to datum(0 - don't fille)

Velocity (m/s) used to fill the datum-to-topography interval in the output model. Default: 0 m/s (do not fill).

Global iteration count

Number of outer tomographic update cycles. Default: 3.

Time step

Ray tracing time step (s). Default: 0.01 s.

Localization iteration count

Number of localisation sub-iterations per global iteration. Default: 10.

Joint inversion iteration count

Number of LSQR iterations per global iteration. Default: 5.

Regularization weight

Weight of the spatial smoothness regularisation in the inversion. Default: 100. Higher values produce smoother models.

Outliers threshold

Standard deviation multiplier for outlier classification. Default: 3.

Outliers start

Global iteration from which outlier rejection is applied. Default: 1.

Maximum depth

Maximum depth (m) of the output model. Default: 4000 m.

Initial velocity min and V0

Minimum (surface) velocity (m/s) of the initial gradient starting model. Default: 1500 m/s.

Initial velocity max

Maximum (base) velocity (m/s) of the initial gradient starting model. Default: 5000 m/s.

Init step X

Initial inversion grid cell size in the X direction (m). Default: 500 m.

Init step Y

Initial inversion grid cell size in the Y direction (m). Default: 500 m.

Adding bound

Extra boundary margin (m) added to the inversion grid around the survey footprint. Default: 2500 m. Prevents edge effects in the velocity model near the survey boundary.

Output dz

Depth sampling interval (m) of the output velocity model. Default: 15 m.

Calculate alter cost

When enabled, computes an alternative cost diagnostic alongside the main inversion misfit. Default: off.

Sealing Factor

Grid refinement factor applied between global iterations. Default: 0.5 (halves the grid cell size each iteration). Range: 0.1–5.

Smooth model factor

Spatial smoothing factor applied to each velocity update. Default: 1.

LSQR parameters

Container grouping LSQR solver convergence parameters.

Conlim

Maximum condition number limit for the LSQR solver. Default: 10000.

Percitnlim

Percentage of maximum iterations used as an early stopping check. Default: 70.

Min vertical shift

LSQR atol convergence tolerance. Default: 0.000001.

Max vertical shift

LSQR btol convergence tolerance. Default: 0.000001.

Tomo item read params

Container grouping parameters that filter tomo data points before inversion.

Maximum points

Maximum tomo data points per line. Default: 5000. Reduce if memory errors occur.

Min angle

Minimum event incidence angle (degrees) to include. Events below this angle are rejected.

Max angle

Maximum event incidence angle (degrees) to include. Default: 90 degrees.

Leave/Remove event { None, Leave, Remove }

Controls whether events in the Min/Max angle range are kept or removed. Default: None.

Factor

Decimation factor for tomo data points. Default: 1 (no decimation).

Settings

Execute on { CPU, GPU }

Selects whether the inversion runs on CPU or GPU.

Distributed execution

Configuration for distributed (cluster) execution.

Bulk size

Minimum gathers per chunk during distributed execution.

Limit number of threads on nodes

When enabled, restricts CPU threads per node to the Number of threads value.

Job suffix

Optional text appended to the distributed job name.

Set custom affinity

When enabled, applies the Affinity mask for CPU core assignment.

Affinity

CPU affinity mask used when Set custom affinity is enabled.

Number of threads

Maximum CPU threads used during execution.

Skip

When enabled, this module is bypassed and data passes through unchanged.

Output data

Output gathers

Intermediate per-iteration depth velocity model gathers for convergence monitoring.

Final output gathers

Final updated depth velocity model at the common processing datum. Save via Save seismic by gather for use in the next migration iteration.

Final output gathers original datum

Final updated velocity model in the topography datum reference frame.

Tomo item multiline with updated velocities

The multi-line tomo item updated with the new velocity model, ready for use as input to the next iteration of Set multi line tomo item and stereo tomography.

Current line output depth velocity

Depth velocity model for the currently active line, available for interactive display during processing.

Information

Working points count

Number of tomo data points used after filtering and decimation.

Graphics

Custom actions