|
<< Click to Display Table of Contents >> Navigation: Velocity > VRMS to stereo 3D tomo parameters |
The VRMS to stereo 3D tomo parameters module converts a 3D RMS velocity field into the tomographic parameter set required for stereo-tomography. Starting from a picked or interpolated VRMS volume, the module derives normal-ray (zero-offset) travel times, source-receiver offset information, and an initial depth velocity model. These are assembled into a tomo item that can be directly used as input to a 3D stereotomography inversion workflow for depth model building.
The module supports two processing modes selected by the Version 1 toggle. In Version 1 mode, the module samples the VRMS volume along user-defined inline and crossline grids to extract NIP (Normal Incidence Point) ray parameters at each grid node. In the alternative mode, the full VRMS volume is processed directly. In both modes, the Dix parameters container controls how interval velocity is derived from the input RMS velocities, and the Create tomo params container defines the geometry and initial model for the stereotomography preparation step.
An optional multiline tomo item that provides pre-existing stereotomography parameters to be merged or extended. If not connected, the module builds all tomo parameters from scratch using the input VRMS volume.
An optional correlation or semblance gather associated with the VRMS picks. When connected, the correlation values are used to weight the quality of the tomo parameters — picks with high semblance receive more weight in the inversion.
Optional horizon picks in the time domain. When connected, the module can use these to constrain the tomo parameter extraction to specific reflectors rather than sampling the VRMS volume at all times.
The input 3D RMS velocity volume. Each trace in this gather represents the RMS velocity as a function of two-way time at one CMP location. This is the primary input from which all tomo parameters are derived. The volume should cover the full area of interest and must be spatially complete (no large gaps) for reliable tomo parameter extraction.
The time window (in seconds) over which the correlation or semblance is averaged when evaluating the quality of each VRMS pick. Default: 0.05 s (50 ms). A larger window smooths the quality estimate and reduces sensitivity to pick noise, while a smaller window preserves local quality variations.
When enabled (default: true), uses the Version 1 processing mode in which VRMS picks are sampled along a regular grid of inlines and crosslines defined in the Version 1 parameters container. When disabled, the full VRMS volume is processed. Version 1 mode is more efficient for large volumes and produces a more uniform distribution of tomo parameters.
The minimum RMS velocity (in m/s) used as a validity threshold. Default: 1500 m/s. VRMS picks below this value are treated as invalid and excluded from the tomo parameter set. Increase this if the input VRMS field contains unrealistically low values in poorly-resolved zones.
The maximum RMS velocity (in m/s) used as a validity threshold. Default: 6000 m/s. VRMS picks above this value are excluded. Adjust this to match the maximum expected velocity in the survey area.
This container defines the grid of inline and crossline positions at which VRMS picks are sampled when Version 1 mode is active.
Controls which directions are sampled for tomo parameters. Default: IL's (inline sections only). Set to XL's to sample crosslines only, or IL's and XL's for a full 3D grid. Using both directions provides better azimuthal coverage for the stereotomography inversion but doubles computation time.
The first inline number to include in the tomo parameter grid. Default: 10. Set to the first inline of the area of interest.
The first crossline number to include in the tomo parameter grid. Default: 10.
The last inline number to include. Default: 999999 (process all inlines up to the last in the dataset).
The last crossline number to include. Default: 999999.
The spacing between sampled inlines in inline number units. Default: 25. Larger steps reduce the number of tomo parameters extracted and speed up processing; smaller steps provide denser spatial coverage and improve the resolution of the tomography model.
The spacing between sampled crosslines in crossline number units. Default: 25.
Defines how the inline and crossline grid constraints are combined when selecting which bins contribute tomo parameters. Default: Logical AND. With Logical AND, only bins that satisfy both inline and crossline grid criteria are included. With Logical OR, bins satisfying either criterion are included. Use AND to restrict parameters to the intersection of the inline and crossline grids, or OR to include parameters from both directions independently.
Determines whether the inline/crossline start, stop, and step values are interpreted as sequence numbers (integer inline/crossline identifiers in the survey bin grid) or as real-valued coordinate positions. Default: Seq.
This container groups all parameters that define the geometry and initial velocity model used when preparing the stereotomography parameter set.
An optional label string assigned to the output tomo item, used to identify this set of tomo parameters in the project. Useful when combining tomo parameters from multiple lines or surveys.
The near-surface reference velocity (in m/s) used in the initial velocity model when Initial velocity type is set to Gradient. Default: 1500 m/s. This value represents the velocity at the datum level and is used as the starting point of the linear velocity gradient model.
The spatial grid step in the X (inline) direction used for the stereotomography parameter grid, in metres. Default: 250 m. This defines the horizontal resolution at which tomo parameters are computed. Use a step consistent with the inline bin spacing of the survey and the desired spatial resolution of the tomography model.
The spatial grid step in the Y (crossline) direction for the tomo parameter grid, in metres. Default: 250 m.
The time grid step used for sampling VRMS picks along the time axis when building the tomo parameter set, in seconds. Default: 0.1 s (100 ms). Finer sampling produces more tomo parameter entries along the time axis but increases computation and inversion time. A step of 50–100 ms is typical for most applications.
The maximum two-way time (in seconds) at which tomo parameters are extracted. Default: 5 s. Set this to the maximum time of interest for the depth model, which is typically determined by the deepest target horizon. VRMS picks beyond this time will not contribute to the output tomo item.
A list of source-receiver offset values (in metres) at which stereotomography ray parameters are computed. Default offsets: 100, 300, 600, 1000, 1500, 2300, 3000, 4000, 5000 m. These offsets define the multi-offset geometry of the tomo parameters and determine the angular diversity of the NIP (Normal Incidence Point) rays in the output tomo item. The list should cover the usable offset range of the survey. Add more entries for surveys with very large offsets or remove entries that are beyond the usable aperture.
Individual offset value entries in the Offset collection list. Each entry specifies one source-receiver offset in metres at which tomo parameters are evaluated.
Selects how the initial depth velocity model is constructed. Default: Gradient. Gradient uses a linear velocity model defined by the Gradient velocity model from and to parameters. Dix derives the initial model by applying the Dix inversion to the input VRMS field, which typically gives a better starting model if the VRMS picks are of good quality.
The velocity at the shallow end of the gradient model (m/s), used when Initial velocity type is Gradient. Default: 1500 m/s. This is the velocity at the datum level.
The velocity at the deep end of the gradient model (m/s), used when Initial velocity type is Gradient. Default: 6000 m/s. A linear velocity gradient is applied between the from and to values over the depth range from the datum to Max depth.
The depth sample interval (in metres) of the initial velocity model. Default: 5 m. Use a value consistent with the desired depth resolution of the output model. A smaller dz produces a finer depth grid but increases memory usage.
The maximum depth (in metres) of the initial velocity model. Default: 10000 m. Set this to the deepest target depth in the survey. The depth model is built from the datum down to this depth.
The reference datum elevation (in metres) for the initial velocity model. Default: 0 m (sea level). Set to the processing datum of the survey.
A container of boolean flags that allow automatic detection of key model parameters from the input data.
When enabled (default: true), the datum elevation is automatically determined from the elevation information in the input trace headers, overriding the manually specified Datum value.
When enabled (default: true), the maximum time is automatically determined from the record length of the input VRMS data, overriding the manually specified Max time value.
When enabled (default: true), the minimum velocity of the gradient model is automatically estimated from the minimum valid VRMS value in the input data.
When enabled (default: true), the maximum velocity of the gradient model is automatically estimated from the maximum valid VRMS value in the input data.
This container groups all parameters controlling the Dix inversion step when Initial velocity type is set to Dix. The Dix parameters also define the interval velocity bounds and quality control settings applied when the module constructs the initial model from the VRMS input.
When enabled, VRMS values outside the range [Min VRMS to consider, Max VRMS to consider] are excluded before the Dix inversion. This prevents outlier velocity picks from producing physically unrealistic interval velocities. Enable this if the VRMS field contains erroneous picks.
The minimum acceptable VRMS value (in m/s) used by the Dix inversion quality filter. Default: 1500 m/s. VRMS picks below this value are rejected when Don't use bad Vrms values is enabled.
The maximum acceptable VRMS value (in m/s) for the Dix inversion. Default: 4500 m/s. Picks above this value are excluded when Don't use bad Vrms values is enabled.
Parameters controlling the iterative Dix inversion solver used to derive interval velocities from the input VRMS field.
The number of global (outer) iterations of the Dix inversion solver. Default: 2. Each global iteration refines the entire 3D interval velocity model. More iterations improve convergence but increase computation time.
The number of local (inner) iterations per global iteration. Default: 5. The inner loop refines the velocity at each depth node. The product of global and local iterations determines the total solver depth.
The convergence tolerance for the Dix inversion solver. Default: 1e-6. The solver stops when the residual norm falls below this threshold. Leave at the default unless the solver is diverging or converging too slowly.
The spatial step in the X (inline) direction used during the Dix inversion, in metres. Default: 200 m. This controls the spatial density of the inversion computation grid. Use a value consistent with the inline spacing of the velocity picks.
The spatial step in the Y (crossline) direction used during the Dix inversion, in metres. Default: 200 m.
A table of time-based interval velocity constraints used to limit the Dix inversion result. Each entry defines an allowed velocity range [MinVelocity, MaxVelocity] at a specific start time. Add rows to apply depth-varying velocity constraints — for example, to prevent unrealistically fast shallow velocities or unrealistically slow deep velocities.
Individual rows in the Interval Velocity-Time table, each defining a velocity constraint at a given time.
The start time (in seconds) of a velocity constraint row in the Interval Velocity-Time table. The constraint applies from this time down to the start time of the next row (or to the end of the record if there is no next row).
The minimum allowed interval velocity (in m/s) for this time interval. Default: 1500 m/s. The Dix inversion result is clipped to this value if it falls below the minimum.
The maximum allowed interval velocity (in m/s) for this time interval. Default: 7000 m/s. The Dix inversion result is clipped to this value if it exceeds the maximum.
When enabled (default: true), prevents the interval velocity from decreasing more rapidly than the fraction specified in Fraction of decrease. This ensures that the interval velocity profile is geologically plausible by suppressing large velocity inversions caused by noisy VRMS picks.
Controls how much the interval velocity is allowed to decrease from one time step to the next, as a fraction of the previous value. Default: 1.0 (no constraint on decrease). A value of 0.99 limits the decrease to at most 1% per time step, effectively enforcing near-monotonic behaviour. This parameter is only active when Limit decrease of VINT is enabled.
Advanced regularisation and solver parameters for the Dix inversion. Leave at defaults unless directed otherwise by g-Platform support.
The Tikhonov regularisation weight applied in the X (lateral) direction. Default: 10. Larger values produce a spatially smoother interval velocity field at the cost of reduced resolution.
The Tikhonov regularisation weight applied in the Z (depth) direction. Default: 10. Larger values enforce a smoother velocity variation with depth.
An internal solver parameter controlling the step size scaling of the iterative inversion. Default: 1.5. Leave at default unless the solver is oscillating or diverging.
An internal solver damping parameter. Default: 1.0. Valid range: 0.9–1.1. Leave at default unless instructed otherwise.
Selects whether the computation is performed on the CPU or GPU. The multi-threaded Dix inversion and tomo parameter extraction typically run efficiently on multi-core CPUs.
When enabled, distributes tomo parameter computation across multiple nodes in a cluster environment, enabling efficient processing of very large 3D volumes.
The number of gathers sent to each compute node per batch in distributed mode.
Sets the maximum number of CPU threads per node in distributed mode.
An optional text label appended to the distributed job name.
When enabled, allows manual specification of the CPU core affinity mask.
The CPU core affinity mask when Set custom affinity is enabled.
The number of parallel CPU threads used for processing. Set to match the available cores on the machine.
When enabled, this module is bypassed and input data passes through unchanged.
The output stereotomography parameter item. This contains the set of NIP ray parameters (zero-offset times, source-receiver offsets, ray azimuths, and initial velocities) assembled from the input VRMS volume. Connect this output to the input of a 3D stereotomography inversion module to use these parameters as the starting point for depth model building.
The initial depth-domain interval velocity model derived from the Dix inversion of the input VRMS field. This can be used as the starting velocity model for PSDM or as a quality-control output to inspect the initial depth velocity before running the stereotomography inversion.