|
<< Click to Display Table of Contents >> Navigation: Diffraction > Diffraction amplitude correction |
Diffraction amplitude correction normalizes the amplitudes of diffraction-domain gathers to compensate for uneven illumination, variable fold, and lateral amplitude trends that arise during diffraction imaging workflows. Without this correction, bins with more contributing traces will appear artificially brighter than sparse bins, making it difficult to interpret true diffractor strength.
The module applies up to three independent normalization stages that can be enabled or combined: a vertical (time-domain) amplitude balancing that smooths the time-variant mean level across all traces; an optional fold-weighted correction that accounts for the number of traces contributing to each sample; and an optional lateral (spatial) trend correction that removes large-scale amplitude gradients across the survey area using the bin grid geometry.
Use this module after diffraction imaging or migration to prepare the output gather for attribute extraction, stacking, or further AVO/AVA analysis. It is particularly important for 3D surveys where acquisition geometry causes strong illumination variations across the grid.
The primary seismic data source. Connect the SEG-Y handle or internal data item that contains the diffraction-domain gathers to be corrected. This item provides trace geometry, header information, and the data connector used to drive the module execution.
The seismic gather that will be amplitude-corrected. This is the main waveform data input — typically a partial stack, common-image gather, or diffraction image gather produced by a preceding diffraction imaging step. The output gather will have identical dimensions (number of traces and samples) but with balanced amplitudes.
This parameter group controls whether the amplitude correction takes into account the varying number of traces (fold) that contribute to each sample. When fold is uneven across the gather — which is typical after diffraction imaging — higher-fold samples receive systematically stronger amplitudes. Enabling fold consideration corrects this bias using a regression-based scaling derived from the fold gather.
Controls whether the fold gather is used to weight the amplitude correction. The default value is No.
No — the module computes a per-sample mean amplitude across all non-zero traces and smooths it over a time window (controlled by the Window parameter). Each trace sample is then divided by this smoothed mean. This mode is suitable when fold information is not available or when the fold variation is modest.
Yes — the module applies a linear regression to estimate the relationship between the square root of input amplitude and the fold value at each sample position. The derived regression slope is then used to scale each trace by its local fold value. Connect a fold gather to the Gather Fold input when using this mode. This option provides a more accurate correction in surveys with strongly variable illumination.
A gather containing the fold values for each trace and sample position. This input is required when Fold Consideration Mode is set to Yes. The fold gather must have the same trace count and sample count as the input gather. It is typically produced by the diffraction imaging or stacking module that generated the input data. When Fold Consideration Mode is set to No, this input is not used.
This group controls the time-domain amplitude smoothing that forms the normalization reference for the main correction pass. The module computes the mean amplitude level at each time sample across the gather (in the No-fold-mode path) and then averages this level over a sliding time window before using it as the divisor. A wider window produces a smoother, more gradual gain function, while a narrower window allows the correction to follow faster amplitude changes with depth or time.
The half-length of the time window (in seconds) used to smooth the per-sample mean amplitude before applying the vertical normalization. The smoothing is applied symmetrically: for each time sample, the mean amplitude is averaged over samples within this distance on either side. The default value is 0.3 s. The minimum allowed value is 0 s (no smoothing — each sample is normalized individually by the local mean).
Use smaller values (0.05–0.1 s) when you need to preserve short-scale amplitude trends that are geologically meaningful. Use larger values (0.5–1.0 s) when the primary goal is to remove long-wavelength gain variations caused by spherical divergence or attenuation that were not corrected upstream. A value of 0.3 s is a reasonable starting point for most post-stack diffraction gathers with 4 ms sampling.
This group controls an optional spatial amplitude equalization that removes large-scale lateral amplitude trends across the survey grid. When enabled, the module maps average trace energy onto the bin grid, spatially smooths this energy map over an inline-by-crossline window, and then divides each trace by its smoothed local energy relative to the global survey mean. This is useful when acquisition footprint, near-surface absorption, or geometrical effects create systematic amplitude gradients from one part of the survey to another.
Enables or disables the lateral amplitude trend correction. The default value is No.
No — no spatial correction is applied. Only the vertical normalization (and optionally fold-mode) is used. This is appropriate when the survey has uniform illumination or when lateral trends have already been removed.
Yes — the module reads the bin grid to determine each trace's inline and crossline position, accumulates the mean absolute amplitude per bin, spatially smooths this map using the Inline window and Crossline window parameters, and then scales each trace so that the local smoothed energy equals the global survey mean. A bin grid must be connected when this option is selected. This mode is recommended for 3D surveys with strong acquisition footprint or irregular sampling.
The 3D bin grid that defines the inline/crossline geometry of the survey. This input is required when Detect Trend is set to Yes. The bin grid is used to map each trace's XY coordinate to its corresponding inline and crossline bin index, so that the spatial energy map can be computed and smoothed on the regular grid. Connect the same bin grid that was used during the diffraction imaging step. When Detect Trend is set to No, this input is ignored.
The half-width of the spatial smoothing window applied to the energy map in the inline direction, measured in number of crosslines. The default value is 150 crosslines. The minimum allowed value is 1. A larger value averages the energy map over a wider inline band and removes only coarse spatial trends. A smaller value allows the correction to follow finer-scale illumination variations. This parameter is only active when Detect Trend is set to Yes.
Set this value to be large enough that the smoothed map captures only the acquisition-related trend and not genuine geological amplitude contrasts. For a typical land 3D survey with bin spacing of 25 m, a window of 150 bins corresponds to 3750 m, which is generally sufficient to remove acquisition footprint while preserving local amplitude anomalies.
The half-width of the spatial smoothing window applied to the energy map in the crossline direction, measured in number of inlines. The default value is 150 inlines. The minimum allowed value is 1. This parameter works identically to the Inline window but acts perpendicular to the inline direction. Set it independently from the Inline window when the survey has asymmetric acquisition geometry (for example, longer receiver lines than source lines) or when the illumination gradient is known to be stronger in one spatial direction. This parameter is only active when Detect Trend is set to Yes.