|
<< Click to Display Table of Contents >> Navigation: Migration > Kirchhoff PreSTM - 2D/3D - Offset Mode (big data) |
This module performs Kirchhoff pre-stack time migration (PreSTM) for 2D, 2.5D, and 3D seismic datasets, producing output as offset-class common image gathers (CIGs). Unlike the standard gather-in/gather-out migration modules, this variant is designed for large datasets: it reads input traces directly from a SEG-Y file on disk and writes migrated results to a new SEG-Y output file, avoiding the need to hold the entire dataset in memory. Each output CIG contains migrated traces grouped by offset range, allowing velocity analysis and amplitude-versus-offset (AVO) studies on the migrated image domain.
The migration applies the Kirchhoff summation algorithm, using an RMS velocity model to compute travel-time curves. An anti-alias filter and optional stretching-factor mute suppress migration noise and mute operator stretch artifacts. Execution can be distributed across multiple CPU threads or GPU devices to accelerate processing of large 3D volumes.
Note: This module is deprecated. For new projects, use the current Kirchhoff PreSTM modules that offer improved performance and additional features.
A handle to the input SEG-Y file containing the pre-stack seismic data to be migrated. The module reads traces directly from this file during processing, making it suitable for datasets too large to fit in RAM. The file must be opened prior to running this module using an appropriate file reader node in the processing flow. The sampling interval and number of samples are taken from this file and must match those of the velocity model.
A sorted trace header index (gather index vector) describing the geometry and ordering of the input seismic data. This index defines the CMP bin positions, offsets, and record positions that the migration engine uses to identify which input traces contribute to each output image bin. For 2D migration the index must represent a 2D CMP geometry; for 3D migration it must represent a 3D CMP grid. The index is typically generated by geometry application modules upstream in the processing flow.
The RMS velocity model used for Kirchhoff travel-time computation. This model must have the same time sampling interval and number of samples as the seismic data. The velocity values are used to calculate diffraction travel-time curves for each output image point. Accurate velocities are essential for correctly positioning reflectors after migration; errors in the velocity model will cause under- or over-migration of events.
An optional trace vector defining the output image point grid. If provided, the migrated output will be placed at the bin positions defined by this geometry rather than at the input CMP positions. This is useful when the output image grid differs from the acquisition grid — for example, when migrating to a regular subsurface grid that is finer or differently oriented than the original CMP spacing. If left empty, the output image positions are taken directly from the input sorted headers.
Selects the dimensionality of the migration operator. Choose 2D for a single 2D seismic line, 3D for a full 3D survey, or 2.5D for a pseudo-3D approximation applied to 2D data. The selected dimension determines how aperture distances and bin correlations are computed. The input geometry must be consistent with the chosen dimension: 2D geometry for 2D or 2.5D modes, and 3D geometry for 3D mode. Default: 2D.
Full file path for the output SEG-Y file that will receive the migrated common image gathers. The file will be created automatically if it does not exist. If a file already exists at the specified path, the module will prompt whether to overwrite it. Ensure sufficient disk space is available, as the output size is proportional to the number of output bins multiplied by the number of offset classes and the trace length.
The maximum lateral distance (in metres) from an output image point within which input traces are summed during migration. Larger apertures include more traces in the summation, improving the imaging of steeply dipping reflectors and wide-angle energy, but increasing computation time. Smaller apertures speed up processing but may under-image steep dips. As a starting guide, set the aperture to at least the expected maximum lateral displacement of dipping events at the deepest target — typically 500–5000 m for most surveys. Default: 3000 m.
When enabled, the migration uses the full CMP aperture for each output image point, meaning all input traces within the defined aperture radius are considered regardless of their position relative to the CMP fold pattern. When disabled, the aperture may be trimmed based on the actual data coverage. Enabling this option produces the most complete image but may introduce migration noise at the edges of the survey. Default: On.
A taper distance (in metres) defining a boundary zone around the edge of the survey. Image points within this distance from the survey border are tapered to reduce migration edge effects and wrap-around noise. Set this to zero to disable edge tapering. Use values in the range of one to two CMP intervals for minimal tapering, or larger values (several hundred metres) to suppress stronger edge noise. Default: 0 m.
When enabled, the Stretching Factor mute is applied during migration. Kirchhoff migration stretches wavelets at large offsets and dips, distorting the frequency content of migrated traces. The stretching factor mute suppresses samples where the operator stretch exceeds the specified threshold, improving the frequency fidelity of migrated gathers. Enable this option when AVO analysis or accurate amplitude preservation is required. Default: Off.
The maximum allowable operator stretch expressed as a percentage. Samples for which the migration operator stretches the wavelet by more than this value are muted. Lower values (e.g., 10–30%) apply a stricter mute, removing more stretched samples and producing cleaner gathers at the cost of some coverage loss at large offsets. Higher values allow more stretched energy through. This parameter is only active when Use Stretching Factor is enabled. Default: 0%.
The near-surface replacement velocity in metres per second (m/s). This value represents the assumed velocity of the shallow layer used in computing the anti-alias filter cutoff frequency. It is typically set to the velocity of the uppermost layer or the weathering layer velocity. For marine surveys, use the water velocity (approximately 1500 m/s). For land surveys, use the near-surface replacement velocity from statics processing. Default: 1500 m/s.
A dimensionless coefficient controlling the strength of the anti-aliasing filter applied to each trace before summation. The anti-aliasing filter prevents spatial aliasing artifacts that arise when the migration operator has frequency content above the spatial Nyquist limit. A value of 1.0 applies a standard anti-alias filter; lower values relax the filtering and preserve more high-frequency content at the risk of aliasing noise; higher values apply stronger low-pass filtering. For most surveys a value of 0.5 to 1.0 is appropriate.
The maximum frequency (in Hz) used in the Rho filter (the ramp or derivative filter applied as part of the Kirchhoff migration amplitude correction). The Rho filter boosts higher frequencies to correct for the migration integral; however, boosting frequencies beyond the useful bandwidth amplifies noise. Set this value to the highest frequency present in the seismic data with useful signal, typically the upper end of the bandwidth estimated from the frequency spectrum. Default: 120 Hz.
The maximum dip angle (in degrees from vertical) at which energy will be imaged during migration. Traces contributing at angles greater than this value are excluded from the summation. This acts as a dip filter that limits migration noise at large angles. A value of 90 degrees includes all angles and is equivalent to using no dip limit. Reduce this value (e.g., to 40–60 degrees) if steep-dip migration noise is problematic or if the target geology is predominantly flat-lying. Default: 90 degrees.
A scalar multiplier applied to the RMS velocity model before migration. Use this factor to globally scale the velocity field — for example, set to 0.95 to reduce all velocities by 5%, or to 1.05 to increase them by 5%. This can be used to test the sensitivity of the migrated image to velocity perturbations without modifying the velocity model itself. A value of 1.0 uses the velocity model unchanged. Default: 1.0.
The CMP bin spacing (in metres) along the inline direction. This value is used to convert the aperture distance from metres into a number of bins for the aperture search. It must match the actual CMP spacing of the input dataset along the inline axis. Incorrect values will cause the aperture to be applied over too few or too many bins, leading to under-migration or excessive computation. Default: 12.5 m.
The CMP bin spacing (in metres) along the crossline direction. Used in the same way as the inline interval to determine the aperture in bins along the crossline axis. For 2D surveys this parameter is not applicable but should still be set consistently. Default: 25.0 m.
This group contains optional advanced parameters for 3D grid-based migration. When these settings are enabled, a BinGrid object is used to map input and velocity data onto a regular 3D grid, allowing migration onto a grid that differs from the acquisition geometry. Leave these settings at their defaults for standard 2D and 3D migration workflows.
When enabled, the module uses the BinGrid objects provided below to map input seismic and velocity data to a regular 3D grid before migration. This is useful when the input data is defined on an irregular or rotated grid and the output image is required on a regular Cartesian grid. When disabled (default), the bin correlation is performed using the nearest-bin search based on coordinate proximity. Default: Off.
A BinGrid object describing the regular 3D bin grid of the input seismic data. Required only when 3D grid define is enabled. The BinGrid provides the mapping between world coordinates and inline/crossline bin indices for the input dataset, enabling precise bin-to-bin correlation during aperture search.
A BinGrid object describing the regular 3D bin grid of the velocity model. Required only when 3D grid define is enabled. This grid is used to extract velocity values at the correct spatial positions during migration. If the velocity model and seismic data are on the same grid, the same BinGrid can be used for both inputs.
This parameter group controls the definition of the offset-class common image gathers (CIGs) produced by the migration. The offset range is divided into a series of classes defined by the minimum offset, maximum offset, and offset increment. Each class collects energy from input traces whose offset falls within that window, producing one output trace per class per image point. Correct CIG parameters are essential for subsequent velocity analysis and AVO interpretation on the migrated gathers.
The smallest offset (in metres) of the first output CIG offset class. Traces with offsets below this value are not included in the migrated output. Set this to zero or to the minimum usable offset in the dataset. Default: 0 m.
The largest offset (in metres) of the last output CIG offset class. Traces with offsets above this value are not included. Set this to the maximum usable offset in the dataset, or to a lower value to exclude far-offset traces with poor signal-to-noise ratio. Together with the minimum offset and offset increment, this value determines the total number of output CIG classes: (MaxOffset - MinOffset) / OffsetStep + 1. Default: 3000 m.
The width of each offset class (in metres). Input traces whose offset falls within a window of this width around each class centre are summed into that class. Smaller increments produce more offset classes with finer offset resolution, which is beneficial for AVO analysis but increases output file size and computation time. A typical value is 50–100 m for standard surveys. Default: 50 m.
An additional offset search margin (in metres) added symmetrically around each CIG class when collecting input traces. Input traces within the offset range [class_offset - OffsetAperture, class_offset + OffsetAperture + OffsetStep] are considered for the class. A larger aperture increases the number of traces contributing to each class, improving gather fold but potentially blending energy from adjacent offset classes. Set this to approximately half the offset increment for standard operation. Default: 25 m.