Pre or post -stack T/X domain

  • Predictive or spiking deconvolution can use up to 10 design windows to compress non-stationary wavelets.
  • Pre-stack design can be based on single traces, on trace averages grouped by offset or on whole shot or CMP gathers.
  • Prediction gaps can be linked to and changed automatically with water depth
  • Phase only or amplitude only operators can be specified
  • The high and low frequency roll off in the amplitude spectrum can be amortised to produce a reliable phase spectrum estimate for use in correcting the wavelet to zero phase.

Surface consistent deconvolution

Data may be decomposed into source, receiver, offset and CMP consistent components and their amplitude spectrum used to determine surface consistent operators capable of compressing the wavelet associated with data in each of these domains. The technique is used with land data when processing to preserve amplitude and phase response.

 

Pre-stack tau-p domain

  • p traces can be generated using the linear Radon transform of input records. They represent plane wave components having constant multiple periodicity with time.
  • Predictive deconvolution is often more successful than in the T/X domain as periodicity is more faithfully preserved in the autocorrelation function.
  • Benefits are greatest for longer multiple periods of ~100-500ms
  • Prediction gap can be linked to water depth and to p value
  • Spatial interpolation of input may be necessary in order to prevent aliasing

Q Compensation

A time and space varying amplitude and phase correction can be used to compensate for increasing loss of high frequency components the deeper the data. It reduces non-stationarity in the wavelet allowing subsequent use of single window deconvolution.

 

Maximum entropy deconvolution

Can be used to spectrally balance data

 

Phase (Minimum Entropy) deconvolution

Can be used to maximize resolution –a phase only operator converts to zero phase

 

Wavelet estimation-zero phase correction

  • The average wavelet can be estimated from the windowed average amplitude spectrum of post-stack data. A white reflectivity spectrum is assumed. Wavelet phase may be estimated or assumed to be minimum or zero phase.
  • A phase only operator derived from the average wavelet is used to change the phase spectrum to zero.
 More Info