Did you smooth your well logs the right way for seismic interpretation?

Correlations between physical properties and seismic reflection data are useful to determine the geological nature of seismic reflections and the lateral extent of geological strata. The difference in resolution between well logs and seismic data is a major hurdle faced by seismic interpreters when tying both data sets. In general, log data have a resolution of at least two orders of magnitude greater than seismic data. Smoothing physical property logs improves correlation at the seismic scale. Three different approaches were used and compared to smooth a density log: binomial filtering, seismic wavelet filtering and discrete wavelet transform (DWT) filtering. Regression plots between the density logs and the acoustic impedance show that the data smoothed with the DWT is the only method that preserves the original relationship between the raw density data and the acoustic impedance. Smoothed logs were then used to generate synthetic seismograms that were tied to seismic data at the borehole site. Best ties were achieved using the synthetic seismogram computed with the density log processed with the DWT. The good performance of the DWT is explained by its adaptive multi-scale characteristic which preserved significant local changes of density on the high-resolution data series that were also pictured at the seismic scale. Since synthetic seismograms are generated using smoothed logs, the choice of the smoothing method impacts on the quality of seismic-to-well ties. This ultimately can have economical implications during hydrocarbon exploration or exploitation phases.