|
<< Click to Display Table of Contents >> Navigation: Interpretation > Horizon slice extractor |
The Horizon slice extractor extracts amplitude (or RMS) values from a seismic volume along a picked horizon surface, producing a two-dimensional horizon slice map. For each bin in the survey grid, the module interpolates the horizon time from the supplied horizon picks, then samples the seismic amplitude within a configurable time window centered on that horizon. The resulting map provides a spatial image of seismic attributes that follows the geological reflector rather than a fixed time level.
An optional fold matrix (MFFold) can be supplied together with a mathematical formula, which allows the extracted amplitude values to be scaled or weighted by any per-bin quantity — for example, migration aperture fold or a coverage mask. A custom action is also available to apply this formula-based scaling directly to the full trace cube and export the result as a SEG-Y file. Use this module when you need to study lateral amplitude variations, bright-spot distributions, or any seismic attribute along a specific geological horizon.
The seismic trace dataset from which amplitude values will be extracted. This should be a post-stack volume (or the stack you wish to analyse) with traces associated to bin coordinates. The bin headers must be populated; if they are empty, run the Binning module first to assign each trace to its survey bin before using this module.
The survey bin grid that defines the spatial layout of inline and crossline cells. The extracted slice map is built on this grid, so it must match the geometry of the seismic traces. The output matrix will have the same inline and crossline extent as this bin grid.
A set of horizon picks (X, Y, time/depth coordinates) that define the geological surface along which the slice will be extracted. The module uses kriging interpolation to estimate the horizon time at every bin location between the supplied picks. Picks do not need to cover every bin — the interpolation fills gaps up to the distance defined by the MaxDist parameter.
An optional two-dimensional matrix (for example, a migration fold map or a coverage weight grid) whose per-bin values are passed into the Mathematical expression formula as the variable A. When this item is connected and a formula is set, each extracted amplitude is multiplied by the scalar result of the formula evaluated at that bin. If you do not need amplitude weighting or scaling, leave this input disconnected.
A mathematical formula that is evaluated once per bin using the MFFold value at that bin as the variable $A$. The scalar result of the formula is used to scale the extracted amplitude at each bin. The default expression $A$ simply passes the fold value through unchanged. You can use standard arithmetic expressions, for example 1/$A$ to normalise by fold, or ($A$ > 10) ? 1 : 0 to mask bins with low fold. This parameter only has an effect when the MFFoldItem input is connected.
This group contains parameters that control how the horizon time is estimated at bin locations that do not have a direct horizon pick. The module uses a fast kriging interpolator to spread horizon values from picked points to surrounding bins.
The maximum search radius (in metres) within which horizon picks are used to interpolate the horizon time at a given bin. Any bin farther than this distance from all available picks will not receive an extracted value and will be left as zero in the output map. The default value is 50 m. Increase this value for sparse pick grids to fill larger gaps; decrease it to restrict extrapolation in areas of uncertain horizon coverage.
The maximum number of nearby horizon picks used by the kriging interpolator when estimating the horizon time at each bin. Using more points can improve interpolation accuracy when the horizon has strong local curvature, but increases computation time. The default value is 100. For most surveys this default is sufficient.
This group controls the time window and spatial neighbourhood over which amplitude values are averaged when building the horizon slice. Increasing these windows smooths the output map but reduces spatial and temporal resolution. Set all windows to zero to extract only the single interpolated sample at the exact horizon time without any averaging.
The half-width of the time window (in seconds) centred on the interpolated horizon time. Amplitude samples from all time steps within this window are combined (averaged or RMS-summed, as set by the Calculation type parameter) to produce the output value for each bin. The default value is 0.01 s (10 ms). Increase this window to capture a broader zone around the reflector — useful when the horizon pick uncertainty is high or to compute an integrated attribute over a thin layer.
Determines how amplitude samples within the Average Window are combined into a single output value per bin. Two options are available:
Average (default) — computes the arithmetic mean of all samples within the window. The output retains the sign of the seismic amplitudes, making this choice appropriate for studying peak or trough character.
RMS — computes the root-mean-square amplitude over the window. The output is always positive and highlights energy levels rather than polarity. Use RMS when looking for bright spots or AVO anomalies where you care about amplitude strength rather than sign.
A constant time offset (in seconds) added to the interpolated horizon time at every bin before amplitude extraction. Use a positive value to shift the extraction window downward in time (below the horizon), or a negative value to shift it upward. This is useful when you want to extract amplitudes from a stratigraphic interval a fixed time interval above or below the picked surface — for example, to study a layer that is consistently 20 ms beneath the main reflector. The default value is 0 s (no shift).
The number of adjacent inline bins on each side of the current bin to include in the amplitude calculation. When set to a value greater than zero, the module collects samples from a spatial neighbourhood of (2 × Inline window + 1) inlines and combines them using the selected Calculation type. This spatial averaging reduces noise in the slice at the cost of lateral resolution in the inline direction. The default value is 0 (no spatial averaging). When either Inline window or Crossline window is non-zero, the time averaging window is applied to each trace in the neighbourhood.
The number of adjacent crossline bins on each side of the current bin to include in the amplitude calculation. Works together with the Inline window to define a rectangular spatial averaging patch. A setting of 2 means that traces from 2 crosslines on each side (5 crosslines total) are included. The default value is 0 (no spatial averaging). Increase this parameter in areas of poor signal-to-noise ratio where spatial stacking of nearby bins helps stabilise the slice values.