Factorial kriging analysis: Geostatistical filtering applied to reservoir characterization

摘要

This paper discusses factorial kriging analysis (FKA) for filtering seismic attributes. The great advantage of this technique is that the filter parameterization uses geologic knowledge about the field. The general theory of FKA was developed by Matheron (1982) and was used in the mid-1980s for separation of magnetic anomalies. FKA relies on the assumption that a regionalized phenomenon can be seen as a linear sum of varied independent sub-phenomena acting at different scales, each of which presents its own vari-ogram or covariance model which will, when linearly summed, comprise the variogram or covariance model of the regionalized phenomenon.