|
<< Click to Display Table of Contents >> Navigation: Multiples > Radon - TauP forward transform |
This module performs a forward Radon (Tau-P) transform, converting seismic gather data from the offset-time (t-x) domain into the slowness-intercept time (tau-p) domain. The resulting Radon spectrum can be used for multiple attenuation, noise separation, data interpolation, and AVO analysis. Four transform types are supported: parabolic, Foster-Mosher pseudo-hyperbolic, linear tau-p, and absolute-value linear tau-p.
The input gather must be sorted by offset in ascending order before applying this transform. The output Radon spectrum is a 2D gather with tau (intercept time) on the vertical axis and p (slowness or curvature) on the horizontal axis. Use this module as the first step in a transform-filter-inverse workflow for interbed multiple attenuation or dip separation.
The seismic data container providing access to the gather collection. Typically connects to the main seismic dataset.
The seismic gather (typically a CMP gather) to be transformed. Traces must be sorted by offset in ascending order. NMO correction is not required for the forward transform itself, but the output spectrum interpretation depends on the type of moveout present in the data.
The minimum slowness (p) value defining the lower bound of the Radon spectrum. For the linear tau-p transform p is in s/m; for the parabolic transform p is in s/m². A negative P min allows the spectrum to capture events with opposite moveout polarity (e.g., up-going energy). Default: -0.5. Set this value to encompass the full range of dips or curvatures present in your data.
The maximum slowness (p) value defining the upper bound of the Radon spectrum. For multiple attenuation, P max should be large enough to capture the residual moveout of the multiples at the maximum offset. Default: 1.0. Together with P min and Delta P, this determines the number of Radon traces in the output spectrum.
The slowness sampling interval used to discretize the Radon spectrum between P min and P max. Smaller values produce a higher-resolution spectrum with more output traces, at the cost of computation time. A good starting value is approximately equal to the dominant signal time interval of your data. Default: 0.008.
The reference offset (in metres) used to normalize offsets in the transform computation. For the parabolic transform, this defines the offset at which the curvature is measured. Typically set to the maximum offset or a representative large offset in the gather. Default: 500 m. This parameter affects the scaling of p values and should match across forward and inverse transform steps.
Advanced options group. Contains the Ignore sorting checkbox. Enable this only if you are certain the gather traces are in correct offset order but the sort verification is incorrectly rejecting the data. By default, the module validates ascending offset order and stops with an error if offsets are not sorted.
The minimum frequency (Hz) used in the least-squares Radon computation. Frequencies below this value are excluded from the transform. Set this to the low-cut frequency of your usable signal bandwidth to improve stability and reduce low-frequency noise in the spectrum. Default: 0 Hz.
The maximum frequency (Hz) used in the Radon computation. Frequencies above this value are excluded. Set to the Nyquist frequency or the high-cut of your usable bandwidth. Default: 100 Hz.
A stabilization factor (in percent) added to the diagonal of the normal equations in the least-squares inversion. Larger values improve numerical stability but reduce spectral resolution. Typical values range from 0.1% to 5%. Default: 0.5%.
Selects the moveout curve model used in the Radon transform. Parabolic is the most widely used option for NMO-corrected data where residual moveout is approximately parabolic. Foster-Mosher hyperbolic models hyperbolic moveout using a pseudo-hyperbolic approximation, suitable for non-NMO-corrected data. Linear tau-p is appropriate for plane-wave decomposition. Abs linear tau-p uses the absolute offset value, which is useful for split-spread acquisitions where positive and negative offsets are treated symmetrically. Default: Linear tau-p.
A list of trace header fields to copy from the first input trace to all traces in the output Radon spectrum. Use this to propagate key identification headers (such as CDP number, inline/crossline, or shot coordinates) so the spectrum can be linked back to the original geometry during inverse transform or visualization.
Selects whether processing runs on the CPU or GPU. GPU execution can significantly accelerate the least-squares Radon computation for large gathers.
The output Radon spectrum gather. The horizontal axis represents the slowness parameter p, the vertical axis represents intercept time tau. Each trace in the output corresponds to one p value. This gather is passed to a filter step (e.g., muting in the Radon domain) and then to the inverse Radon transform module.
The number of time samples added to the top and bottom of the gather to accommodate edge effects during the transform. This value is automatically calculated based on the P range and offset extent, and is stored in the output gather headers for use by the inverse transform module.