Statistical multi-trace wavelet estimation (prototype)
●Wavelet estimation based on the statistical analysis of multiple seismic traces and wells
Initial wavelet is computed direct from the seismic data around multiple wells. Then, amplitude and phase spectrum are optimized by the interface using the elastic well logs. It is also possible to make an automatic optimization by a MCMC method
The Plugin is under developing for the software OpendTect. In the figures below we summarize the workflow of the statistical wavelet estimation.
For a given radius around the wells, the software computes the best time shift for each trace that provides the maximum correlation with the synthetics. The number of seismic traces actually used in the wavelet estimation depends on the parameter “correlation cut”. By increasing the correlation cut the user selects only the best traces.
After selecting the traces, the wavelet scale factor and phase can be obtained by observing the histograms of the best values.