|
<< Click to Display Table of Contents >> Navigation: Migration > Beam PostSDM (2D) |
Beam PostSDM (2D) performs post-stack depth migration of 2D seismic data using the Gaussian beam method. It takes a time-domain stacked section and an interval depth velocity field as inputs and produces a depth-migrated image by decomposing the wavefield into a set of Gaussian beams, tracing each beam through the velocity model, and accumulating their contributions at the correct depth positions.
Gaussian beam migration is a ray-based depth imaging technique that offers better handling of multipathing and lateral velocity variations than standard Kirchhoff migration, while remaining more computationally efficient than full-wave equation methods. The output is a depth-domain stack with the same lateral sampling as the input, suitable for structural interpretation and depth conversion quality control.
This module processes one 2D inline at a time. It requires an interval velocity model in the depth domain — a time-domain velocity field will cause an error. Optionally, angle and semblance gathers produced by a previous processing step can be supplied to guide beam formation when using the angle-guided or semblance-weighted beam modes.
The 2D post-stack time section to be depth-migrated. This should be a fully stacked (zero-offset equivalent) seismic gather in the time domain. The number of traces in this input must match the number of traces in the Interval Depth Velocity input exactly; a mismatch will cause the module to stop with an error.
The interval velocity model in the depth domain, provided as a 2D field v(x, z). This field is used both to trace rays through the subsurface and to determine beam parameters such as the minimum surface velocity and beam sampling interval. The input must be in the depth domain — a time-domain velocity field will be rejected. The depth sampling and spatial coverage of this input determine the geometry of the output migrated image.
A gather containing emergence angle values (in radians) at each time sample, used to guide beam formation in the angle-guided migration modes (Flag1 = 1 or 2). This input is required when Flag1 is set to 1 or 2, and is ignored when standard beam migration mode is selected (Flag1 = 0). Typical source for this input is an angle estimation step run prior to migration.
A gather containing semblance (coherency) values used to weight the beam contributions in mode Flag1 = 1, where beams are formed directly from the angle and semblance gathers rather than from the conventional slant-stack decomposition. Like Input Angles, this input is only required when Flag1 is set to 1 or 2. Higher semblance values indicate more coherent events and lead to stronger beam contributions.
The maximum ray emergence angle, in degrees, measured from vertical (normal incidence). Beams are traced symmetrically from -Max Angle Aperture to +Max Angle Aperture. Larger aperture angles allow more steeply dipping reflectors to be imaged, at the cost of including more migration noise from steeply travelling rays. The default value is 90 degrees, which means all angles are included. Reduce this value (e.g., to 45–60 degrees) when the velocity model is not reliable at steep angles or when migration noise is excessive.
The minimum frequency, in Hz, used in the beam migration. This value sets the low-frequency limit for the Gaussian beam decomposition and also controls the beam sampling density and beam half-width (when Beam Half-Width is set to automatic). The default value of -1 causes the module to derive an automatic minimum frequency from the time sampling interval of the input stack (approximately 0.025 / dt). Set this explicitly to match the low-frequency content of your data, typically in the range of 5–15 Hz for reflection seismic data.
The maximum frequency, in Hz, used in the beam migration. This value sets the high-frequency limit for the beam decomposition and influences the beam width at depth. The default value of -1 causes the module to use ten times the minimum frequency automatically. Set this to match the highest usable signal frequency in your data, typically 60–120 Hz for typical reflection seismic surveys. Using a maximum frequency that is too high can introduce high-frequency noise artefacts in the migrated image.
The half-width of the Gaussian beams at the surface, in metres. This parameter controls the spatial extent of each beam and determines how broadly or narrowly each beam illuminates the subsurface. Wider beams provide smoother, more stable migration at the cost of lateral resolution; narrower beams give sharper images but can be more sensitive to velocity errors. The default value of -1 causes the module to calculate the optimal beam half-width automatically from the average velocity and the minimum frequency (bwh = average velocity / Frequency minimum). This automatic value is a good starting point; adjust manually only if the image shows beam-related artefacts or insufficient resolution.
The horizontal distance between adjacent CMP (Common Mid-Point) traces in the input stack, in metres. This value defines the lateral sampling interval dx used internally for beam tracing and output grid construction. It must be set accurately to match the actual trace spacing in the input data. The default value is 25 m. Setting this incorrectly will cause geometric distortion in the migrated image. A value of zero or less will cause an error.
Selects the beam formation mode used during migration. Three modes are available:
0 — Standard Gaussian beam migration. Beams are formed by a conventional slant-stack (plane-wave decomposition) of the input stack traces within a spatial window around each beam centre. The Input Angles and Input Semblance are not used. This is the default and most robust mode.
1 — Angle-guided beam migration. Beams are constructed directly from the Input Angles and Input Semblance gathers, bypassing the conventional slant-stack. This mode uses pre-computed dip information to form more targeted beams, which can improve imaging of complex structures. Both Input Angles and Input Semblance must be connected.
2 — Pseudo-stack beam migration. A synthetic pseudo-stack is first constructed from the single-trace Input Angles and the Input Stack using angle-based time shifts, creating a locally corrected multi-trace gather. The standard slant-stack beams are then formed from this pseudo-stack. Both Input Angles and Input Semblance must be connected. This mode can improve lateral continuity in areas with complex dip variations.
The half-length of the time window, in seconds, applied when constructing the pseudo-stack in Flag1 = 2 mode. For each sample in the angle gather, the module shifts and accumulates stack amplitudes within a window of ±Time Window around that sample. A larger window captures more samples during the pseudo-stack construction, providing smoother results, but may smear sharp events. The default value is 0.01 s (10 ms). This parameter has no effect in Flag1 = 0 or Flag1 = 1 modes.
The length of the taper, in samples, applied to the edges of the time window during pseudo-stack construction in Flag1 = 2 mode. The taper reduces edge artefacts when accumulating shifted amplitudes by smoothly tapering the contribution of samples near the boundary of the time window to zero. The default value is 1 sample. Increase this value if ringing or edge effects are visible in the migrated output when using Flag1 = 2. This parameter has no effect in Flag1 = 0 or Flag1 = 1 modes.
The depth-migrated 2D seismic section resulting from the Gaussian beam migration. The output has the same number of traces as the input stack, with the vertical axis converted from two-way time to depth. The depth sampling interval and maximum depth are inherited from the Interval Depth Velocity input. Amplitudes represent the migrated reflectivity image at each (x, z) position in the subsurface.